SEPA
September 2024
United States Office of Chemical Safety and
Environmental Protection Agency Pollution Prevention
Risk Evaluation for
Tris(2-chloroethyl) Phosphate (TCEP)
Systematic Review Supplemental File:
Data Quality Evaluation Information for
Human Health Hazard Epidemiology
CASRN: 115-96-8
CI
September 2024
-------
This supplemental file contains the data quality evaluation results for epidemiology data sources that met the PECO screen-
ing criteria in the Risk Evaluation for Tris(2-chloroethyl) phosphate (TCEP). EPA conducted data quality evaluation based on
author-reported descriptions and results; additional analyses (e.g., statistical analyses performed during data integration into the
risk evaluation) potentially conducted by EPA are not contained in this supplemental file. EPA used the TSCA systematic review
process described in the Draft Systematic Review Protocol Supporting TSCA Risk Evaluations for Chemical Substances(also
referred to as '2021 Draft Systematic Review Protocol'). Any updated steps in the systematic review process since the publica-
tion of the 2021 Draft Systematic Review Protocol are described in Risk Evaluation for Tris(2-chloroethyl) phosphate (TCEP)
- Systematic Review Protocol.
Page 2 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Table of Contents
Table of Contents
HERO ID
Reference
Page
Tris(2-chloroethyl) phosphate (TCEP)
6957526
Araki, A., Bamai, Y. A., Bastiaensen, M., Eede, V.d., N., Kawai, T., Tsuboi, T., Miyashita, C., Itoh, S., Goudarzi, H., Konno, S., Covad,
A., Kishi, R. (2020). Combined exposure to phthalate esters and phosphate flame retardants and plasticizers and their associations with
wheeze and allergy symptoms among school children. Environmental Research 183:109212.
2994738
Canbaz, D., Velzen, van, M. J., Hallner, E., Zwinderman, A. H., Wickman, M., Leonards, P. E., Ree, van, R., Rijt, van, L. S. (2015).
Exposure to organophosphate and polybrominated diphenyl ether flame retardants via indoor dust and childhood asthma. Indoor Air
26(3):403-413.
12
11581665
Foster, S. A., Kile, M. L., Hystad, P., Diamond, M. L., Jantunen, L. M., Mandhane, P. J., Moraes, T. J., Navaranjan, G., Scott, J. A.,
Simons, E., Subbarao, P., Takaro, T. K., Turvey, S. E., Brook, J. R. (2024). Organophosphate ester flame retardants and plasticizers in
house dust and mental health outcomes among Canadian mothers: A nested prospective cohort study in CHILD. Environmental Research
240(Pt 1): 117451.
14
4161719
Hoffman, K., Lorenzo, A., Butt, C. M., Hammel, S. C., Henderson, B. B., Roman, S. A., Scheri, R. P., Stapleton, H. M., Sosa, J. A.
(2017). Exposure to flame retardant chemicals and occurrence and severity of papillary thyroid cancer: A case-control study. Environment
International 107:235-242.
22
6747922
11364983
11364830
Li, Y., Fu, Y., Hu, K., Zhang, Y., Chen, J., Zhang, S., Zhang, B., Liu, Y. (2020). Positive correlation between human exposure to
organophosphate esters and gastrointestinal cancer in patients from Wuhan, China. Ecotoxicology and Environmental Safety 196:110548.
Liao, K., Zhao, Y., Qu, J., Yu, W., Hu, S., Fang, S., Zhao, M., Jin, H. (2023). Organophosphate esters concentrations in human serum and
their associations with Sjogren syndrome. Environmental Pollution 331(Pt 1): 121941.
Liu, M., Li, A., Meng, L., Zhang, G., Guan, X., Zhu, J., Li, Y., Zhang, Q., Jiang, G. (2022). Exposure to Novel Brominated Flame Retar-
dants and Organophosphate Esters and Associations with Thyroid Cancer Risk: A Case-Control Study in Eastern China. Environmental
Science & Technology 56(24):17825-17835.
30
33
38
7537904
11364495
Liu, Y., Li, Y., Dong, S., Han, L., Guo, R., Fu, Y., Zhang, S., Chen, J. (2021). The risk and impact of organophosphate esters on
the development of female-specific cancers: Comparative analysis of patients with benign and malignant tumors. Journal of Hazardous
Materials 404(Pt B): 124020.
Mendy, A., Percy, Z., Braun, J. M., Lanphear, B., Guardia, La, M. J., Hale, R. C., Yolton, K., Chen, A. (2024). Prenatal exposure to
replacement flame retardants and organophosphate esters and childhood adverse respiratory outcomes. Environmental Research 240(Pt
2):117523.
48
50
Metabolite: bis-2-chloroethyl phosphate (BCEP)
Page 3 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Table of Contents
7274557
Crawford, K. A., Hawley, N., Calafat, A. M., Jayatilaka, N. K., Froehlich, R. J., Has, P., Gallagher, L. G., Savitz, D. A., Braun, J. M.,
Werner, E. F., Romano, M. E. (2020). Maternal urinary concentrations of organophosphate ester metabolites: associations with gestational
weight gain, early life anthropometry, and infant eating behaviors among mothers-infant pairs in Rhode Island. Environmental Health: A
Global Access Science Source 19(1):97.
53
11364495
10081087
Mendy, A., Percy, Z., Braun, J. M., Lanphear, B., Guardia, La, M. J., Hale, R. C., Yolton, K., Chen, A. (2024). Prenatal exposure to
replacement flame retardants and organophosphate esters and childhood adverse respiratory outcomes. Environmental Research 240(Pt
2):117523.
Percy, Z., Vuong, A. M., Xu, Y., Xie, C., Ospina, M., Calafat, A. M., Lanphear, B. P., Braun, J. M., Cecil, K. M., Dietrich, K. N., Chen,
A., Yolton, K. (2021). Prenatal exposure to a mixture of organophosphate esters and intelligence among 8-year-old children of the HOME
Study. NeuroToxicology 87:149-155.
59
62
Metabolite: BCEP
10078361
Kang, H., Lee, J., Lee, J. P., Choi, K. (2019). Urinary metabolites of organophosphate esters (OPEs) are associated with chronic kidney
disease in the general US population, NHANES 2013-2014. Environment International 131:5034-5034.
66
Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
11577122
Hernandez-Castro, I., Eckel, S. P., Howe, C. G., Niu, Z. Z., Kannan, K., Robinson, M., Foley, H. B., Grubbs, B., Al-Marayati, L.,
Lerner, D., Lurvey, N., Aung, M. T., Habre, R., Dunton, G. F., Farzan, S. F., Breton, C. V., Bastain, T. M. (2023). Sex-specific effects of
prenatal organophosphate ester (OPE) metabolite mixtures and adverse infant birth outcomes in the maternal and developmental risks from
environmental and social stressors (MADRES) pregnancy cohort. Environmental Research 226:115703.
75
11581666
Hernandez-Castro, I., Eckel, S. P., Howe, C. G., Niu, Z., Kannan, K., Robinson, M., Foley, H. B., Yang, T., Vigil, M. J., Chen, X., Grubbs,
B., Lerner, D., Lurvey, N., Al-Marayati, L., Habre, R., Dunton, G. F., Farzan, S. F., Aung, M. T., Breton, C. V., Bastain, T. M. (2023).
Prenatal exposures to organophosphate ester metabolite mixtures and children's neurobehavioral outcomes in the MADRES pregnancy
cohort. Environmental Health 22(1):66.
80
11365039
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame
Retardants Impair the Lung Function via the IL-6/JAK/STAT Signaling Pathway. Environmental Science & Technology 56(24):17858-
17869.
Metabolite: bis (2-chloroethyl) phosphate (BCEP)
85
11577121
Yang, W. L., Braun, J. M., Vuong, A. M., Percy, Z., Xu, Y. Y., Xie, C. C., Deka, R., Calafat, A. M., Ospina, M., Burris, H. H., Yolton, K.,
Cecil, K. M., Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and
neonatal anthropometric measures: The HOME study*. Environmental Pollution 316(Part 1): 120516.
105
Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
Page 4 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Table of Contents
11577123 Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., 125
Croen, L. A., Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas,
M. R., Kuiper, J. R., McEvoy, C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S.,
Schmidt, R. J., Simhan, H., Starling, A. P., Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations
of organophosphate ester flame retardant exposures during pregnancy with gestational duration and fetal growth: The Environmental
Influences on Child Health Outcomes (ECHO) program. Environmental Health Perspectives 132(1):017004.
Metabolite: Bis-2-chloroethyl phosphate (BCEP)
11581667 Percy, Z., Chen, A., Sucharew, H., Yang, W., Vuong, A. M., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Cecil, K. M., Xu, Y., 155
Yolton, K. (2023). Early-life exposure to a mixture of organophosphate esters and child behavior. International Journal of Hygiene and
Environmental Health 250:114162.
11581664 Percy, Z., Chen, A., Yang, W., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Xie, C., Cecil, K. M., Vuong, A. M., Xu, Y., Yolton, 160
K. (2022). Childhood urinary organophosphate esters and cognitive abilities in a longitudinal cohort study. Environmental Research 215(Pt
1): 114265.
Page 5 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 6957526 Table: 1 of 2
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Araki, A., Bamai, Y. A., Bastiaensen, M., Eede, V.d., N., Kawai, T., Tsuboi, T., Miyashita, C., Itoh, S., Goudarzi, H., Konno, S., Covaci, A., Kishi, R.
(2020). Combined exposure to phthalate esters and phosphate flame retardants and plasticizers and their associations with wheeze and allergy symptoms
among school children. Environmental Research 183:109212.
Lung/Respiratory-wheeze
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
6957526
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
High The study was conducted among elementary school students in Sapporo, Japan in 2008.
The survey was sent to 6393 school children across 12 public elementary schools, and
44008 students responded, with 951 students interested in participating. Only 681 fam-
ilies were still at the same elementary school in 2009, of those only 128 families were
able to be contacted for a home visit, meaning there was an overall participation rate
of 2.9%Despite a low participation rate, the reasons for exclusion are clearly defined at
each step and are unlikely to introduce significant bias.
High Of the final 128 participants, none were excluded from the data analysis after home
visits/data collection occurred. There is no missing exposure or outcome data.
Medium Participants in this cross-sectional study appear to be similar in terms of baseline charac-
teristics. Most variables are controlled for in statistical analyses. Height and weight are
not controlled for, but the listed mean + standard deviations imply that there is a some-
what large range.
Domain 2: Exposure Characterization
Metric 4:
Metric 5:
Measurement of Exposure
Exposure Levels
Metric 6: Temporality
High Exposure is reported as TCEP measured from urine.
Medium The study reports on a range of exposures from the LOQ to 1.13 nM of TCEP. In statisti-
cal analyses, TCEP concentrations are split into tertiles.
Low Exposure was measured after the onset of symptoms, so the temporality of the exposure
and outcome is uncertain. However, due to the ubiquity of TCEP in the environment, it
may be reasonable to assume that they were exposed to the same amount of TCEP prior
to exposure assessment. However, the study does not check to see if children had moved
recently, which could have altered exposure levels.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
Medium Outcomes were assessed via the International Study of Asthma and Allergies in Child-
hood (ISAAC) questionnaire, which was filled out by study investigators and the parents
of children. The questionnaire asks for symptoms common for wheeze instead of actual
medical diagnoses. Parents may also be impacted by the desirability bias. However, this
is likely to have a non-differential effect.
High A description of the the outcomes is clearly mentioned in the methods. Results are pre-
sented as odds ratios with 95% confidence intervals and p-values.
Continued on next page ...
Page 6 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 6957526 Table: 1 of 2
... continued from previous page
Araki, A., Bamai, Y. A., Bastiaensen, M., Eede, V.d., N., Kawai, T., Tsuboi, T., Miyashita, C., Itoh, S., Goudarzi, H., Konno, S., Covaci, A., Kishi, R.
(2020). Combined exposure to phthalate esters and phosphate flame retardants and plasticizers and their associations with wheeze and allergy symptoms
among school children. Environmental Research 183:109212.
Lung/Respiratory-wheeze
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
6957526
Domain
Metric
Rating
Comments
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High The results are reported as odds ratios that are adjusted for sex, grade, annual income,
and the dampness index of the child's home. They were all included based on a priori
evidence.
Medium Information on confounders were obtained via questionnaire. Sex and grade were mea-
sured via questionnaire and were unlikely to be reported incorrectly. Annual income
is also self-reported, and could fall victim to desirability bias. There were also miss-
ing values for income (14.8%) to which the mean annual household income was as-
signed. Finally, a "dampness" index was calculated by study investigator observation of
dampness-related problems in each dwelling, such as condensation and visible mold.
Medium The intent of this study was to measure co-exposures of phthalates and phosphate flame
retardants and allergic outcomes. Statistical models were used in this study to examine
in urine individual phthalates, combinations of phthalates, metabolites of PFRs and a
combination of certain PFRs.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High The use of a cross-sectional design to understand the relationship between TCEP and
wheeze/allergy symptoms is an appropriate study design, and calculating odds ratios via
logistic regression is an appropriate method.
Medium The number of participants in each tertile is >20, which could be sufficiently large to
detect an effect. The authors do not calculate statistical power, however they do mention
that the power may not be enough to detect a significant effect. They may be referring to
the multipollutant models in this case.
Medium The description of the analysis is thorough and allows for replication given the study
data.
High The model building is transparent and it is clear why variables were chosen.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric
16:
Use of Biomarker of Exposure
High
TCEP is a parent compound.
Metric
17:
Effect Biomarker
N/A
Not applicable - no biomarker of effect.
Metric
18:
Method Sensitivity
Medium
The % of samples below the detection limit is stated to be 14.8%, which is not incredi-
bly high but is sufficient to address the research hypothesis.
Metric
19:
Biomarker Stability
Medium
There is no documented stability data, but is clarified that spot urine samples were col-
lected in polypropylene containers and refrigerated until the study visit, and then stored
a -20 deg C until the day of analysis.
Metric 20: Sample Contamination Medium There is no documentation of potential sample contamination.
Continued on next page ...
Page 7 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 6957526 Table: 1 of 2
... continued from previous page
Study Citation: Araki, A., Bamai, Y. A., Bastiaensen, M., Eede, V.d., N., Kawai, T., Tsuboi, T., Miyashita, C., Itoh, S., Goudarzi, H., Konno, S., Covaci, A., Kishi, R.
(2020). Combined exposure to phthalate esters and phosphate flame retardants and plasticizers and their associations with wheeze and allergy symptoms
among school children. Environmental Research 183:109212.
Health Lung/Respiratory-wheeze
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
Linked HERO ID(s): No linked references.
HERO ID: 6957526
Domain Metric Rating Comments
Metric 21: Method Requirements High TCEP was measured using liquid chromatography with tandem mass spectrometry (LC-
MS/MS).
Metric 22: Matrix Adjustment Medium TCEP concentrations were creatinine-adjusted, and ranges are reported without adjust-
ment and with adjustment. However, the statistical model only uses adjusted concentra-
tions of TCEP.
Additional Comments: This cross-sectional study measures the association between TCEP measured in spot urine samples and parent-reported symptoms of wheeze. Associations
were analyzed using logistic regression and the study calculated odds ratios. No significant associations in single-pollutant models for TCEP were reported.
The study has some deficiencies in outcome assessment, as there is no verifiable method used in this study.
Overall Quality Determination High
Page 8 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 6957526 Table: 2 of 2
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Araki, A., Bamai, Y. A., Bastiaensen, M., Eede, V.d., N., Kawai, T., Tsuboi, T., Miyashita, C., Itoh, S., Goudarzi, H., Konno, S., Covaci, A., Kishi, R.
(2020). Combined exposure to phthalate esters and phosphate flame retardants and plasticizers and their associations with wheeze and allergy symptoms
among school children. Environmental Research 183:109212.
Immune/Hematological-eczema, allergic rhinoconjunctivitis
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
6957526
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
High The study was conducted among elementary school students in Sapporo, Japan in 2008.
The survey was sent to 6393 school children across 12 public elementary schools, and
44008 students responded, with 951 students interested in participating. Only 681 fam-
ilies were still at the same elementary school in 2009, of those only 128 families were
able to be contacted for a home visit, meaning there was an overall participation rate
of 2.9%Despite a low participation rate, the reasons for exclusion are clearly defined at
each step and are unlikely to introduce significant bias.
High Of the final 128 participants, none were excluded from the data analysis after home
visits/data collection occurred. There is no missing exposure or outcome data.
Medium Participants in this cross-sectional study appear to be similar in terms of baseline charac-
teristics. Most variables are controlled for in statistical analyses. Height and weight are
not controlled for, but the listed mean + standard deviations imply that there is a some-
what large range.
Domain 2: Exposure Characterization
Metric 4:
Metric 5:
Measurement of Exposure
Exposure Levels
Metric 6: Temporality
High Exposure is reported as TCEP measured from urine.
Medium The study reports on a range of exposures from the LOQ to 1.13 nM of TCEP. In statisti-
cal analyses, TCEP concentrations are split into tertiles.
Low Exposure was measured after the onset of symptoms, so the temporality of the exposure
and outcome is uncertain. However, due to the ubiquity of TCEP in the environment, it
may be reasonable to assume that they were exposed to the same amount of TCEP prior
to exposure assessment. However, the study does not check to see if children had moved
recently, which could have altered exposure levels.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
Medium Outcomes were assessed via the International Study of Asthma and Allergies in Child-
hood (ISAAC) questionnaire, which was filled out by the parents. The questionnaire
asks for symptoms common for wheeze, eczema, and allergic rhinoconjunctivitis. There
was no medical diagnosis by a physician.
High A description of the the outcomes is clearly mentioned in the methods. Results are pre-
sented as odds ratios with 95% confidence intervals and p-values.
Domain 4: Potential Confounding / Variability Control
Continued on next page ...
Page 9 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 6957526 Table: 2 of 2
... continued from previous page
Araki, A., Bamai, Y. A., Bastiaensen, M., Eede, V.d., N., Kawai, T., Tsuboi, T., Miyashita, C., Itoh, S., Goudarzi, H., Konno, S., Covaci, A., Kishi, R.
(2020). Combined exposure to phthalate esters and phosphate flame retardants and plasticizers and their associations with wheeze and allergy symptoms
among school children. Environmental Research 183:109212.
Immune/Hematological-eczema, allergic rhinoconjunctivitis
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
6957526
Domain
Metric
Rating
Comments
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High The results are reported as odds ratios that are adjusted for sex, grade, annual income,
and the dampness index of the child's home. They were all included based on a priori
evidence.
Medium Information on confounders were obtained via questionnaire. Sex and grade were mea-
sured via questionnaire and were unlikely to be reported incorrectly. Annual income
is also self-reported, and could fall victim to desirability bias. There were also miss-
ing values for income (14.8%) to which the mean annual household income was as-
signed. Finally, a "dampness" index was calculated by study investigator observation of
dampness-related problems in each dwelling, such as condensation and visible mold.
Medium The intent of this study was to measure co-exposures of phthalates and phosphate flame
retardants and allergic outcomes. Statistical models were used in this study to examine
in urine individual phthalates, combinations of phthalates, metabolites of PFRs and a
combination of certain PFRs.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High The use of a cross-sectional design to understand the relationship between TCEP and
wheeze/allergy symptoms is an appropriate study design, and calculating odds ratios via
logistic regression is an appropriate method.
Medium The number of participants in each tertile is >20, which could be sufficiently large to
detect an effect. The authors do not calculate statistical power, however they do mention
that the power may not be enough to detect a significant effect. They may be referring to
the multipollutant models in this case.
Medium The description of the analysis is thorough and allows for replication given the study
data.
High The model building is transparent and it is clear why variables were chosen.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric
16:
Use of Biomarker of Exposure
High
TCEP is a parent compound.
Metric
17:
Effect Biomarker
N/A
Not applicable - no biomarker of effect.
Metric
18:
Method Sensitivity
Medium
The % of samples below the detection limit is stated to be 14.8%, which is not incredi-
bly high but is sufficient to address the research hypothesis.
Metric
19:
Biomarker Stability
Medium
There is no documented stability data, but is clarified that spot urine samples were col-
lected in polypropylene containers and refrigerated until the study visit, and then stored
a -20 deg C until the day of analysis.
Metric 20: Sample Contamination Medium There is no documentation of potential sample contamination.
Continued on next page ...
Page 10 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 6957526 Table: 2 of 2
... continued from previous page
Study Citation: Araki, A., Bamai, Y. A., Bastiaensen, M., Eede, V.d., N., Kawai, T., Tsuboi, T., Miyashita, C., Itoh, S., Goudarzi, H., Konno, S., Covaci, A., Kishi, R.
(2020). Combined exposure to phthalate esters and phosphate flame retardants and plasticizers and their associations with wheeze and allergy symptoms
among school children. Environmental Research 183:109212.
Health Immune/Hematological-eczema, allergic rhinoconjunctivitis
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
Linked HERO ID(s): No linked references.
HERO ID: 6957526
Domain
Metric
Rating
Comments
Metric 21:
Metric 22:
Method Requirements
Matrix Adjustment
High
Medium
TCEP was measured using liquid chromatography with tandem mass spectrometry (LC-
MS/MS).
TCEP concentrations were creatinine-adjusted, and ranges are reported without adjust-
ment and with adjustment.However, the statistical model only uses adjusted concentra-
tions of TCEP.
Additional Comments: This cross-sectional study measures the association between TCEP measured in spot urine samples and parent-reported symptoms of allergic rhinoconjunc-
tivitis and eczema. Associations were analyzed using logistic regression and the study calculated odds ratios. No significant associations in single-pollutant
models for TCEP were reported. The study has some deficiencies in outcome assessment, as there is no verifiable method used in this study.
Overall Quality Determination High
Page 11 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 2994738 Table: 1 of 1
Study Citation: Canbaz, D., Velzen, van, M. J., Hallner, E., Zwinderman, A. H., Wickman, M., Leonards, P. E., Ree, van, R., Rijt, van, L. S. (2015). Exposure to
organophosphate and polybrominated diphenyl ether flame retardants via indoor dust and childhood asthma. Indoor Air 26(3):403-413.
Health Immune/Hematological-Asthma
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
Linked HERO ID(s): No linked references.
HERO ID: 2994738
Domain Metric Rating Comments
Domain 1: Study Participation
Metric 1: Participant Selection Medium The authors reported participant selection in both this study and in Almqvist et al., 2003.
However, not all of the elements were reported, such as participation rate at all phases of
the study.
Metric 2: Attrition Medium Exclusion and missing values were reported in the Materials and Methods Section, Table
1, and Table SI.
Metric 3: Comparison Group High Study matches controls based on sex, atopic background of the parents, and socioeco-
nomic status. Study states that the asthmatic children and their matched controls did not
differ significantly according to several sociodemographic characteristics outlined in
Table SI.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure
Metric 5:
Metric 6:
Exposure Levels
Temporality
High Samples were filtered, sealed, and stored appropriately. Analyzed by GC-EI-MS with
QA detailed in Brandsma et al., 2014. Additional details in Almqvist et al., 2003
Medium Table 3 provides the range and distribution of exposure for both cases and controls.
High House dust adequate to capture short half-life of TCEP. Dust was collected and analyzed
when children were two months of age; exposure precedes disease.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
High Asthma defined according to set of criteria, including doctor's diagnosis and asthma
medicine prescription.
Low Not all data shown from analyses, including the multivariate linear regression analysis.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium
High
Medium
Study reports multivariate linear regression analyses to adjust for covariates; however,
the data was not shown in the main paper or supplemental. Study noted that the results
did not differ from those reported.
Questionnaire and doctors' diagnoses used.
Adjusted for in the multivariate linear regression analyses; data not shown but stated to
not affect results.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Low Case-control used but no logistic regression conducted; no odds ratios.
Medium 110 cases; 110 controls; adequate sample size
Continued on next page ...
Page 12 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 2994738 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Canbaz, D., Velzen, van, M. J., Hallner, E., Zwinderman, A. H., Wickman, M., Leonards, P. E., Ree, van, R., Rijt, van, L. S. (2015). Exposure to
organophosphate and polybrominated diphenyl ether flame retardants via indoor dust and childhood asthma. Indoor Air 26(3):403-413.
Immune/Hematological-Asthma
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
2994738
Domain
Metric
Rating
Comments
Metric 14:
Metric 15:
Reproducibility of Analyses
Statistical Analysis
Medium Sufficient summary of analyses
High Methods are transparent
Additional Comments: None
Overall Quality Determination
Medium
Page 13 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581665 Table: 1 of 1
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Foster, S. A., Kile, M. L., Hystad, P., Diamond, M. L., Jantunen, L. M., Mandhane, P. J., Moraes, T. J., Navaranjan, G., Scott, J. A., Simons, E., Subbarao,
P., Takaro, T. K., Turvey, S. E., Brook, J. R. (2024). Organophosphate ester flame retardants and plasticizers in house dust and mental health outcomes
among Canadian mothers: A nested prospective cohort study in CHILD. Environmental Research 240(Pt 1):117451.
Neurological/Behavioral-Maternal depression scores during pregnancy (at 18 and 36 weeks gestation) and postpartum (at 6 months and 1 year postpartum)
using the Centre for Epidemiologic Studies for Depression Scale (CES-D); andMaternal stress scores during pregnancy (at 18 and 36 weeks gestation) and
postpartum (at 6 months and 1 year postpartum) using the Perceived Stress Scale (PSS)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11581665
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Medium This study analyzed associations between TCEP in household dust collected shortly
after birth and maternal mental health during pregnancy and postpartum using a sub-
sample of from the CHILD (Canadian Healthy Infant Longitudinal Development)
study.The "CHILD Cohort Study recruited adult pregnant women who were in their
second trimester from health centers located across Canada." Eligibility criteria included
being able to communicate in English and the infant being born at 35 weeks gestation
or later.The analysis sample was compiled for a nested case-cohort study that initially
focused on exposure to phthalates and respiratory outcomes in children. Of the 3,628
CHILD participants, 2,319 were potentially eligible for inclusion based on having house
dust samples, along with having genotyping from a related study and the child having
been clinically assessed at age 5 years. Out of 3628 participants, 2319 met these criteria,
and a random sample of 436 of these 2319 eligible children were selected for inclusion.
Additionally, all remaining children with asthma or recurrent wheeze (n=290), were in-
cluded. Eight otherwise eligible participants were excluded due to issues analyzing their
house dust samples. The final sample comprised 718 participants.Distributions of char-
acteristics are presented in Supplemental Table SI; The authors stated "characteristics
of this sub-sample were similar to those in the full cohort, although missingness in the
sub-sample was generally lower than in the full cohort"; statistical significance testing of
differences was not discussed. There was no direct evidence of selection bias. However,
a potentially important concern is that selectivity related to oversampling for childhood
respiratory illnesses was not addressed in the analyses, e.g., by using selection proba-
bility weights or analyzing associations separately in the random sample. A publication
from this cohort found house dust TCEP to be non-linearly associated with lower odds
of asthma and recurrent wheeze at age 5 years (significant for Q3 vs Q1; Navaranjan
et al, 2021 HEROID 10134087), suggesting that selection based on child respiratory
outcomes may not be independent of TCEP. Childhood asthma has been associated
with maternal asthma and maternal postpartum depression (e.g., Blais et al. 2019 PMID
30292921; Alcala et al 2023 PMID 36128727). Given these relationships, potential bias
(due to collider stratification) is a concern: the TCEP-maternal depression association
may differ among children with respiratory outcomes vs the general population. How-
ever, as there was no direct evidence that oversampling for childhood respiratory illness
affected the validity or significance of associations between house dust TCEP and mater-
nal mental health outcomes, a medium rating was selected.
Continued on next page ...
Page 14 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581665 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Foster, S. A., Kile, M. L., Hystad, P., Diamond, M. L., Jantunen, L. M., Mandhane, P. J., Moraes, T. J., Navaranjan, G., Scott, J. A., Simons, E., Subbarao,
P., Takaro, T. K., Turvey, S. E., Brook, J. R. (2024). Organophosphate ester flame retardants and plasticizers in house dust and mental health outcomes
among Canadian mothers: A nested prospective cohort study in CHILD. Environmental Research 240(Pt 1):117451.
Neurological/Behavioral-Maternal depression scores during pregnancy (at 18 and 36 weeks gestation) and postpartum (at 6 months and 1 year postpartum)
using the Centre for Epidemiologic Studies for Depression Scale (CES-D); andMaternal stress scores during pregnancy (at 18 and 36 weeks gestation) and
postpartum (at 6 months and 1 year postpartum) using the Perceived Stress Scale (PSS)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11581665
Domain
Metric
Rating
Comments
Metric 2:
Attrition
Metric 3:
Comparison Group
Low The study analyzed maternal mental health outcomes measured at four time points,
over which the proportion of participants with available data varied (57.1% overall,
and 77.4%, 67.8%, 63.6%, and 70.2% at 18 weeks' gestation, 36 weeks' gestation, 6
months postpartum, and one year postpartum, respectively). A potentially important
concern is that over time, the proportion of women who did not report mental health
scores increased. The proportion of non-reporters at 18 weeks gestation vs. 6 months
postpartum increased from 6.3 % to 17.3% for depression scores and from 6.3% to
17.0% for stress scores. This >10% increase in missing outcome reporting is similar
in magnitude to the prevalence of clinically relevant depression (12.4% at 6 months
postpartum) and several times higher than the prevalence of elevated stress (2.2% at 6
months). The authors made an effort to address missingness using multiple imputation
by chained equations. However, imputation models used fixed characteristics (e.g. age,
education) which may not adequately predict the significant within-person changes in
mental health scores across time points. Potential bias as a consequence of the >10%
increase over time in outcome non-reporting is an important concern.
Medium The analysis compared mental health scores among women from households with
higher vs. lower levels of TCEP in house dust. Although potential bias due to over-
sampling for childhood respiratory illnesses cannot be excluded as noted under Metric
1, the nested prospective cohort design minimizes potential differences between groups,
and differences between groups were considered as potential confounding variables.
There was no evidence for additional concerns related to the comparison group.
Domain 2: Exposure Characterization
Continued on next page .
Page 15 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581665 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Foster, S. A., Kile, M. L., Hystad, P., Diamond, M. L., Jantunen, L. M., Mandhane, P. J., Moraes, T. J., Navaranjan, G., Scott, J. A., Simons, E., Subbarao,
P., Takaro, T. K., Turvey, S. E., Brook, J. R. (2024). Organophosphate ester flame retardants and plasticizers in house dust and mental health outcomes
among Canadian mothers: A nested prospective cohort study in CHILD. Environmental Research 240(Pt 1):117451.
Neurological/Behavioral-Maternal depression scores during pregnancy (at 18 and 36 weeks gestation) and postpartum (at 6 months and 1 year postpartum)
using the Centre for Epidemiologic Studies for Depression Scale (CES-D); andMaternal stress scores during pregnancy (at 18 and 36 weeks gestation) and
postpartum (at 6 months and 1 year postpartum) using the Perceived Stress Scale (PSS)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11581665
Domain
Metric
Rating
Comments
Metric 4:
Measurement of Exposure
Low TCEP and several other organophosphate ester (OPE) flame retardants were measured in
finely sieved house dust using gas chromatography. Samples were collected at a single
time point (3-4 months postpartum) by vacuuming using a standardized protocol. Lab-
oratory analyses were based on homogenized equal aliquots combined from two areas:
(i) the floor of the most used living area; and (ii) the child's mattress sheet or cover plus
either an adjacent 2-meter area of carpeting or the whole floor if uncarpeted. Only 1.4%
of samples had TCEP concentrations below the LOD; values below the LOD (1.4%)
were imputed using the NDExpo web application. Samples were collected from each
area using new thimbles that were transported in sterile glass bottles (Navaranjan et al,
2021 HEROID 10134087). Samples were stored at room temperature for two weeks,
then at -80 degrees C prior to analysis. There was no evidence to suggest concerns re-
lated to sample degradation. Dust sample collection was standardized, and standard
amounts of homogenized aliquots were analyzed. The analyses used gas chromatogra-
phy [Agilent GC-MSD (6890-5975)] operating in electron impact mode using selective
ion monitoring. TCEP exposure was analyzed as concentrations (ng/g) in dust. Quality
control procedures included using standard reference material and analyzing control
samples and laboratory blanks. Blanks with OPE concentrations below detection limits
were appropriately included. A mixture of isotopically labelled OPEs, including TCEP-
dl2, were added as surrogate standards to selected blanks (81-112% recovery). Z-scores
were calculated and analyses were conducted for TCEP exposure individually, and for
summary exposure to OPEs based on the sum score index method.There were several
concerns with the exposure assessment for TCEP. First, it is uncertain extent to which
TCEP in the pooled dust sample with 50% from infant sleeping areas reflects maternal
exposure, particularly as OPE concentrations in a subsample were considerably higher
in bedrooms vs living areas (Navaranjan et al 2921, 10134087). The authors speculated
that exposures of mothers and infants might be similar because "parents often share a
room with their child for the first three to six months of life, and their mattress would
likely also have flame retardants additives". Second, individual and household factors
that may influence the extent to which these dust measures represent maternal exposure
were not taken into account (e.g., cleaning frequency, frequency with which mothers
slept in children's rooms, age of infant mattress vs. maternal mattress). Third, the au-
thors acknowledge that hand wipes, wristbands, or urine samples, rather than dust sam-
ples, might have better characterized internal dose. Finally, the TCEP exposure metrics
were analyzed as concentrations that did not account for variability in room size/floor
area such as by calculating TCEP dust loadings, which may yield different associations
(e.g., Mendy et al 2024, HEROID 11364495). Thus, there is substantial uncertainty in
the extent to which the TCEP dust measures used adequately captured maternal expo-
sures.
Continued on next page ...
Page 16 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581665 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Foster, S. A., Kile, M. L., Hystad, P., Diamond, M. L., Jantunen, L. M., Mandhane, P. J., Moraes, T. J., Navaranjan, G., Scott, J. A., Simons, E., Subbarao,
P., Takaro, T. K., Turvey, S. E., Brook, J. R. (2024). Organophosphate ester flame retardants and plasticizers in house dust and mental health outcomes
among Canadian mothers: A nested prospective cohort study in CHILD. Environmental Research 240(Pt 1):117451.
Neurological/Behavioral-Maternal depression scores during pregnancy (at 18 and 36 weeks gestation) and postpartum (at 6 months and 1 year postpartum)
using the Centre for Epidemiologic Studies for Depression Scale (CES-D); andMaternal stress scores during pregnancy (at 18 and 36 weeks gestation) and
postpartum (at 6 months and 1 year postpartum) using the Perceived Stress Scale (PSS)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11581665
Domain
Metric
Rating
Comments
Metric 5: Exposure Levels
Metric 6: Temporality
Medium TCEP exposure was analyzed using a continuous variable - concentrations per gram of
sieved house dust (ng/g) converted to Z-scores. Distribution data provided suggested
that the exposure variable was highly non-normal (the minimum, geometric mean, and
maximum were 375, 5150 and 81200 ng/g, respectively). However, there was no direct
evidence that a transformation to improve normality would have meaningfully influ-
enced results. Thus, the range and distribution of exposure were sufficient and a contin-
uous measure of exposure was used, which meets the requirements for a Medium rating
for this metric.
Low Exposure was measured in house dust collected at only a single timepoint of 3-4 months
postpartum. Two of the four maternal mental health assessments occurred prior to the
exposure measurement, having taken place at 18- and 36 weeks' gestation. However,
temporality was appropriate for the remaining two measures. Exposure was estimated
using a single dust sample, and there was no information on variability in house dust
TCEP over time, relative to the timing of outcome measures.
Domain 3: Outcome Assessment
Metric 7: Outcome Measurement or Medium Maternal mental health outcomes were measured in pregnancy and postpartum (18
Characterization weeks' gestation, 36 weeks' gestation, 6 months postpartum, and one year postpartum).
Measures were characterized using two self-administered questionnaires: the Centre
for Epidemiologic Studies for Depression Scale (CES-D) was used to quantify maternal
depression, and the Perceived Stress Scale (PSS) was used to quantify maternal stress.
Both questionnaires have been reported to have good reliability and internal consistency
in other studies in the general population. Primary analyses used continuous scores. Di-
chotomized scores used in supplementary analyses were informed by the literature but
were not based on clinically validated cutoffs. One minor limitation in outcome charac-
terization is that the authors did not discuss the utility and validity of these instruments
to measure changes in mental health occuring specifically around the time of pregnancy.
Both depression and perceived stress scores decreased from gestation to postpartum:
for example, at 18 weeks' gestation vs. 6 months postpartum, mean CES-D scores were
9.3 vs. 8.3, respectively, and the prevalence of clinically relevant scores declined from
16.4% to 12.4%, respectively. The authors did not discuss the extent to which dimin-
ished scores were expected, nor whether the apparent declines might be attributable,
at least in part, to increases in missing responses (from 6.3% to 17.3%, see metric 2).
A related limitation is that repeated mental health scores were analyzed without char-
acterizing whether the depression or perceived stress was persistent, of new onset, or
transient. Despite these limitations, there was no evidence of important biases in out-
come assessment methods that would influence the validity of the assessed associations
with TCEP.
Continued on next page ...
Page 17 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581665 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Foster, S. A., Kile, M. L., Hystad, P., Diamond, M. L., Jantunen, L. M., Mandhane, P. J., Moraes, T. J., Navaranjan, G., Scott, J. A., Simons, E., Subbarao,
P., Takaro, T. K., Turvey, S. E., Brook, J. R. (2024). Organophosphate ester flame retardants and plasticizers in house dust and mental health outcomes
among Canadian mothers: A nested prospective cohort study in CHILD. Environmental Research 240(Pt 1):117451.
Neurological/Behavioral-Maternal depression scores during pregnancy (at 18 and 36 weeks gestation) and postpartum (at 6 months and 1 year postpartum)
using the Centre for Epidemiologic Studies for Depression Scale (CES-D); andMaternal stress scores during pregnancy (at 18 and 36 weeks gestation) and
postpartum (at 6 months and 1 year postpartum) using the Perceived Stress Scale (PSS)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11581665
Domain
Metric
Rating
Comments
Metric 8: Reporting Bias
Medium Results were shown or described for all analyses that had been specified in the Materi-
als and Methods section of the paper. Effect estimates were presented with confidence
intervals. Numbers of participants included were provided for the primary analyses of
continuous outcomes. Numbers in the dichotomous outcome groups (clinically relevant
vs not clinically relevant CES-D; and Low to Moderate vs high PSS score) were shown
in Table 1, but numbers of cases by TCEP exposure status were not shown for supple-
mental analyses of dichotomous outcomes.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment Medium Covariates were selected a priori based on the literature; use of directed acyclic graphs
or other qualitative methods was not discussed. Models adjusted for study center, ma-
ternal age, household income, ethnicity, marital status, prenatal smoking, depression
medication use, and season of dust collection. Maternal education and self-reported
depression were considered as covariates but excluded from final models due to high
collinearity; parity and delivery mode were excluded as they were less important in the
literature and not significantly associated with outcomes. A potential concern is that the
authors did not discuss potential confounding or effect modification by maternal physi-
cal health issues potentially related to both TCEP exposure and maternal mental health
(e.g., gestational diabetes, pregnancy-induced hypertension, or maternal asthma). How-
ever, excluding these variables would be appropriate if they are intermediates. There was
no evidence of important bias.
Metric 10: Covariate Characterization Medium Covariates were measured using questionnaires, but the authors didn't specify whether
the questionnaires were validated. The ambient air pollution co-exposure measures were
assigned based on postal codes. Though details were limited, there was no evidence of
important deficiencies or biases.
Metric 11: Co-exposure Counfounding Medium Models adjusted for annual averages for ambient air pollutants (PM2.5, N02, and 03).
A heatmap of Spearman correlations indicated that TCEP was not highly correlated with
other OPEs measured in this study, suggesting that co-exposure confounding by other
OPEs was unlikely. Potential confounding by other OPEs was not explicitly evaluated,
although associations with exposure to multiple OPEs was analyzed using summed Z-
scores for 14 individual OPEs.
Domain 5: Analysis
Continued on next page ...
Page 18 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581665 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Foster, S. A., Kile, M. L., Hystad, P., Diamond, M. L., Jantunen, L. M., Mandhane, P. J., Moraes, T. J., Navaranjan, G., Scott, J. A., Simons, E., Subbarao,
P., Takaro, T. K., Turvey, S. E., Brook, J. R. (2024). Organophosphate ester flame retardants and plasticizers in house dust and mental health outcomes
among Canadian mothers: A nested prospective cohort study in CHILD. Environmental Research 240(Pt 1):117451.
Neurological/Behavioral-Maternal depression scores during pregnancy (at 18 and 36 weeks gestation) and postpartum (at 6 months and 1 year postpartum)
using the Centre for Epidemiologic Studies for Depression Scale (CES-D); andMaternal stress scores during pregnancy (at 18 and 36 weeks gestation) and
postpartum (at 6 months and 1 year postpartum) using the Perceived Stress Scale (PSS)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11581665
Domain
Metric
Rating
Comments
Metric 12:
Study Design and Methods
High
The study analyzed data nested within a prospective cohort. The primary analyses used
linear mixed models for repeated continuous outcome measures and linear regression for
timepoint-specific continuous outcome measures. Supplemental models of dichotomized
outcomes were analyzed using mixed logistic regression and timepoint-specific logistic
models. These methods are appropriate for the research question of assessing the asso-
ciation between exposure to TCEP and maternal depression and stress. Oversampling
for childhood respiratory illness (see metric 1) could have been better addressed in the
analyses, such as by using sample weights. Additional important deficiencies in design
were not noted.
Metric 13:
Statistical Power
Medium
Repeated measures analyses included 718 participants; sample sizes for timepoint spe-
cific analyses ranged from 594 to 676. There was variability in both TCEP exposure and
mental health outcome variables, which were used as continuous measures in primary
analyses. Although power calculations weren't presented, the number of participants
appears to be adequate to detect an effect. However, relatively few women overall re-
ported scores that were classified as elevated stress, which might have limited the power
to detect differences in stress between groups when the outcome was dichotomized.
Metric 14:
Reproducibility of Analyses
Medium
The description of analyses was adequate to be conceptually reproducible with access to
the data.
Continued on next page ...
Page 19 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581665 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Foster, S. A., Kile, M. L., Hystad, P., Diamond, M. L., Jantunen, L. M., Mandhane, P. J., Moraes, T. J., Navaranjan, G., Scott, J. A., Simons, E., Subbarao,
P., Takaro, T. K., Turvey, S. E., Brook, J. R. (2024). Organophosphate ester flame retardants and plasticizers in house dust and mental health outcomes
among Canadian mothers: A nested prospective cohort study in CHILD. Environmental Research 240(Pt 1):117451.
Neurological/Behavioral-Maternal depression scores during pregnancy (at 18 and 36 weeks gestation) and postpartum (at 6 months and 1 year postpartum)
using the Centre for Epidemiologic Studies for Depression Scale (CES-D); andMaternal stress scores during pregnancy (at 18 and 36 weeks gestation) and
postpartum (at 6 months and 1 year postpartum) using the Perceived Stress Scale (PSS)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11581665
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Descriptive data were presented stratified by elevated CES-D and PSS scores. Both
crude and adjusted associations were presented and the covariates for the adjusted mod-
els were specified a priori; both complete case analyses and analyses using multiple
imputation were conducted. Associations between house dust TCEP concentrations and
continuous mental health outcome scores were analyzed using linear mixed models for
repeated outcome measures, as well as using linear regression for timepoint specific
outcome measures. CES-D scores were natural log-transformed to meet the assump-
tions of linear regression. Dichotomized mental health outcomes were analyzed using
both mixed logistic regression and timepoint-specific logistic models. Results were pre-
sented as beta coefficients or odds ratios with confidence intervals. There were several
potential concerns related to data analysis, but no direct evidence of important error or
bias. One potential concern is that TCEP exposure Z-scores were analyzed continu-
ously and without transformation, assuming a linear dose-response relationship. There
was no direct evidence of non-linearity in these associations. However, TCEP was non-
linearly associated with childhood respiratory illnesses within this cohort (Navaranjan
et al, 10134087). A second potential concern is that model diagnostics and sensitivity
analyses were discussed only for analyses of summed OPE exposures. It is uncertain
whether findings for TCEP were robust after exclusions such as removing participants
who had moved homes during the study period. Finally, despite the aim of analyzing
whether OPEs influenced depression and stress around pregnancy, sensitivity analyses
for summed OPEs did not include evaluating the impact of excluding participants report-
ing depression prior to pregnancy.
Continued on next page ...
Page 20 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581665 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Foster, S. A., Kile, M. L., Hystad, P., Diamond, M. L., Jantunen, L. M., Mandhane, P. J., Moraes, T. J., Navaranjan, G., Scott, J. A., Simons, E., Subbarao,
P., Takaro, T. K., Turvey, S. E., Brook, J. R. (2024). Organophosphate ester flame retardants and plasticizers in house dust and mental health outcomes
among Canadian mothers: A nested prospective cohort study in CHILD. Environmental Research 240(Pt 1):117451.
Neurological/Behavioral-Maternal depression scores during pregnancy (at 18 and 36 weeks gestation) and postpartum (at 6 months and 1 year postpartum)
using the Centre for Epidemiologic Studies for Depression Scale (CES-D); andMaternal stress scores during pregnancy (at 18 and 36 weeks gestation) and
postpartum (at 6 months and 1 year postpartum) using the Perceived Stress Scale (PSS)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11581665
Domain
Metric
Rating
Comments
Additional Comments:
This study analyzed associations between TCEP exposure, as well as summed organophosphate ester flame retardant (OPE) exposure, in household dust
and maternal depression and perceived stress scores using a subset of 718 participants from the Canadian CHILD cohort. Household dust was measured at
3-4 months postpartum; mental health scores were obtained from widely-used self-administered instruments (CES-D, PSS) at weeks 18 and 36 gestation,
and months 6 and 12 postpartum. Both timepoint specific and repeated measures analyses were conducted. Although most of the relevant findings were null
or not statistically significant for TCEP, there was a statistically significant positive association between dust TCEP levels and odds of high perceived stress
(when perceived stress was dichotomized) that was observed only at 6 months postpartum (shown in Table S10), and a small but statistically significant
positive association between summed OPE levels and perceived stress in linear mixed models. However, relatively few women (n=16 to 18) reported
scores that were classified as elevated stress. This study had several limitations, including the use of pooled dust samples comprised of 50% from the
main living area and 50% from the infant's mattress and adjacent floor. It is uncertain to what extent these measures capture infant vs. maternal exposure.
There was also an increase of more than 10% in non-reporting of mental health outcomes between pregnancy and postpartum; it is uncertain whether
the apparent postpartum declines in maternal depression and perceived stress reflect improved mental health vs. selective attrition. The timing of two of
the four maternal mental health measures preceded exposure sample collection. However, the authors presented timepoint specific measures to address
this temporality concern. Other limitations included oversampling for childhood respiratory illnesses without incorporating sample weights despite an
association between these outcomes and TCEP in the CHILD cohort, and assuming a linear dose-response relationship.
Overall Quality Determination
Medium
Page 21 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 4161719 Table: 1 of 4
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Hoffman, K., Lorenzo, A., Butt, C. M., Hammel, S. C., Henderson, B. B., Roman, S. A., Scheri, R. P., Stapleton, H. M., Sosa, J. A. (2017). Exposure to
flame retardant chemicals and occurrence and severity of papillary thyroid cancer: A case-control study. Environment International 107:235-242.
Endocrine-thyroid cancer (papillary)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
4161719
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Metric 2:
Metric 3:
Participant Selection
Attrition
Comparison Group
High Participant selection was well described, including timing of study, inclusion and exclu-
sion criteria and case ascertainment.
High Supplemental Figure 1 provides detailed information and shows only 1 of 71 cases
excluded.
High Cases and controls were recruited from the same Health Care Center during the same
time period and were matched by age and gender.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6:
Temporality
High Flame retardants were measured in house dust. Collection methods and analysis of dust
was fully reported. TCEP concentrations in household dust were measured by GC/EI-
MS.
Low Only 2 levels of flame retardant exposure in the home were reported for cases and con-
trols; TCEP concentrations above the median concentration were reported for both
groups. Median TCEP concentrations were shown in the box plot in Fig. 1 but con-
centrations were not reported. Detection limits were not reported for organophosphate
flame retardants.
Medium Exposure at the same residence for at least 2 years prior to diagnosis; however, it is
unclear whether exposure duration was sufficient for thyroid cancer.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
High The thyroid cancer outcome was assessed by medical review of clinical and pathology
data; pathological stage was assessed based on tumor size, location, metastasis.
Medium Health outcomes and exposures are reported, except for median exposure levels of each
of the flame retardants. Adjusted ORs with CIs reported for overall incidence as well
as measures of tumor aggressiveness (i.e., pathologic stage). Some data were reported
for PBDEs but not the organophosphate flame retardants, such as number of detects,
detection limits.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
High Covariate adjustments were described (age, income, BMI). Regression analyses were
adjusted for participant age and household income. BMI was both included and ex-
cluded as a covariate. Other potential confounders were considered, but not applied
because they did not alter effect estimates (race, employment status, and smoking). Ion-
izing radiation exposure was considered but no participants reported prior exposure.
Continued on next page ...
Page 22 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 4161719 Table: 1 of 4
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Hoffman, K., Lorenzo, A., Butt, C. M., Hammel, S. C., Henderson, B. B., Roman, S. A., Scheri, R. P., Stapleton, H. M., Sosa, J. A. (2017). Exposure to
flame retardant chemicals and occurrence and severity of papillary thyroid cancer: A case-control study. Environment International 107:235-242.
Endocrine-thyroid cancer (papillary)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
4161719
Domain
Metric
Rating
Comments
Metric 10:
Metric 11:
Covariate Characterization
Co-exposure Counfounding
Medium
Low
A questionnaire was administered to participants, but the method for assessing covari-
ates was not described, although the study indicated that protocols were approved by an
institutional review board. Did not report if the questionnaire was validated or the types
of data collected in the questionnaire, but there is no evidence of confounding.
Many flame retardants were measured in household dust samples. The authors acknowl-
edge and provide data on those most highly correlated. Statistical analyses modeled
each flame retardant separately. The authors indicate that component analyses were con-
ducted to systematically assess FR mixtures but did not provide them in the article. They
stated that they "did not provide any additional insights."
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High
Medium
Medium
High
The case-control study design was appropriate for the research question and applicable
statistical methods were used (logistic regression models). However, the co-exposures
were not accounted for in the analysis.
The number of cases and controls was adequate to detect an effect.
The description of the analysis was sufficient to be reproducible with access to the ana-
lytical data.
Model assumptions were adequately described.
Additional Comments: For TCEP, dust levels in the home are used as surrogates for exposure, measured using MS/GC. PBDEs only were measured in serum (not organophosphate
flame retardants). This case-control study used an adequate number of age- and gender-matched pairs. Multiple flame retardants were present in dust and in
large ranges, depending on the flame retardant: adjustments were not made for multiple comparisons (separate models were assessed for each compound).
Overall Quality Determination High
Page 23 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 4161719 Table: 2 of 4
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Hoffman, K., Lorenzo, A., Butt, C. M., Hammel, S. C., Henderson, B. B., Roman, S. A., Scheri, R. P., Stapleton, H. M., Sosa, J. A. (2017). Exposure to
flame retardant chemicals and occurrence and severity of papillary thyroid cancer: A case-control study. Environment International 107:235-242.
Endocrine-thyroid cancer (papillary)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
4161719
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Metric 2:
Metric 3:
Participant Selection
Attrition
Comparison Group
High Participant selection was well described, including timing of study, inclusion and exclu-
sion criteria and case ascertainment.
High Supplemental Figure 1 provides detailed information and shows only 1 of 71 cases
excluded.
High Cases and controls were recruited from the same Health Care Center during the same
time period and were matched by age and gender.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6:
Temporality
High Flame retardants were measured in house dust. Collection methods and analysis of dust
was fully reported. TCEP concentrations in household dust were measured by GC/EI-
MS.
Low Only 2 levels of flame retardant exposure in the home were reported for cases and con-
trols; TCEP concentrations above the median concentration were reported for both
groups. Median TCEP concentrations were shown in the box plot in Fig. 1 but con-
centrations were not reported. Detection limits were not reported for organophosphate
flame retardants.
Medium Exposure at the same residence for at least 2 years prior to diagnosis; however, it is
unclear whether exposure duration was sufficient for thyroid cancer.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
High The thyroid cancer outcome was assessed by medical review of clinical and pathology
data; pathological stage was assessed based on tumor size, location, metastasis.
Medium Health outcomes and exposures are reported, except for median exposure levels of each
of the flame retardants. Adjusted ORs with CIs reported for overall incidence as well
as measures of tumor aggressiveness (i.e., pathologic stage). Some data were reported
for PBDEs but not the organophosphate flame retardants, such as number of detects,
detection limits.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
High Covariate adjustments were described (age, income, BMI). Regression analyses were
adjusted for participant age and household income. BMI was both included and ex-
cluded as a covariate. Other potential confounders were considered, but not applied
because they did not alter effect estimates (race, employment status, and smoking). Ion-
izing radiation exposure was considered but no participants reported prior exposure.
Continued on next page ...
Page 24 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 4161719 Table: 2 of 4
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Hoffman, K., Lorenzo, A., Butt, C. M., Hammel, S. C., Henderson, B. B., Roman, S. A., Scheri, R. P., Stapleton, H. M., Sosa, J. A. (2017). Exposure to
flame retardant chemicals and occurrence and severity of papillary thyroid cancer: A case-control study. Environment International 107:235-242.
Endocrine-thyroid cancer (papillary)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
4161719
Domain
Metric
Rating
Comments
Metric 10:
Metric 11:
Covariate Characterization
Co-exposure Counfounding
Medium
Low
A questionnaire was administered to participants, but the method for assessing covari-
ates was not described, although the study indicated that protocols were approved by an
institutional review board. Did not report if the questionnaire was validated or the types
of data collected in the questionnaire, but there is no evidence of confounding.
Many flame retardants were measured in household dust samples. The authors acknowl-
edge and provide data on those most highly correlated. Statistical analyses modeled
each flame retardant separately. The authors indicate that component analyses were con-
ducted to systematically assess FR mixtures but did not provide them in the article. They
stated that they "did not provide any additional insights."
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High
Medium
Medium
High
The case-control study design was appropriate for the research question and applicable
statistical methods were used (logistic regression models). However, the co-exposures
were not accounted for in the analysis.
The number of cases and controls was adequate to detect an effect.
The description of the analysis was sufficient to be reproducible with access to the ana-
lytical data.
Model assumptions were adequately described.
Additional Comments: For TCEP, dust levels in the home are used as surrogates for exposure, measured using MS/GC. PBDEs only were measured in serum (not organophosphate
flame retardants). This case-control study used an adequate number of age- and gender-matched pairs. Multiple flame retardants were present in dust and in
large ranges, depending on the flame retardant: adjustments were not made for multiple comparisons (separate models were assessed for each compound).
Overall Quality Determination High
Page 25 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 4161719 Table: 3 of 4
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Hoffman, K., Lorenzo, A., Butt, C. M., Hammel, S. C., Henderson, B. B., Roman, S. A., Scheri, R. P., Stapleton, H. M., Sosa, J. A. (2017). Exposure to
flame retardant chemicals and occurrence and severity of papillary thyroid cancer: A case-control study. Environment International 107:235-242.
Endocrine-thyroid cancer (papillary)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
4161719
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Metric 2:
Metric 3:
Participant Selection
Attrition
Comparison Group
High Participant selection was well described, including timing of study, inclusion and exclu-
sion criteria and case ascertainment.
High Supplemental Figure 1 provides detailed information and shows only 1 of 71 cases
excluded.
High Cases and controls were recruited from the same Health Care Center during the same
time period and were matched by age and gender.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6:
Temporality
High Flame retardants were measured in house dust. Collection methods and analysis of dust
was fully reported. TCEP concentrations in household dust were measured by GC/EI-
MS.
Low Only 2 levels of flame retardant exposure in the home were reported for cases and con-
trols; TCEP concentrations above the median concentration were reported for both
groups. Median TCEP concentrations were shown in the box plot in Fig. 1 but con-
centrations were not reported. Detection limits were not reported for organophosphate
flame retardants.
Medium Exposure at the same residence for at least 2 years prior to diagnosis; however, it is
unclear whether exposure duration was sufficient for thyroid cancer.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
High The thyroid cancer outcome was assessed by medical review of clinical and pathology
data; pathological stage was assessed based on tumor size, location, metastasis.
Medium Health outcomes and exposures are reported, except for median exposure levels of each
of the flame retardants. Adjusted ORs with CIs reported for overall incidence as well
as measures of tumor aggressiveness (i.e., pathologic stage). Some data were reported
for PBDEs but not the organophosphate flame retardants, such as number of detects,
detection limits.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
High Covariate adjustments were described (age, income, BMI). Regression analyses were
adjusted for participant age and household income. BMI was both included and ex-
cluded as a covariate. Other potential confounders were considered, but not applied
because they did not alter effect estimates (race, employment status, and smoking). Ion-
izing radiation exposure was considered but no participants reported prior exposure.
Continued on next page ...
Page 26 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 4161719 Table: 3 of 4
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Hoffman, K., Lorenzo, A., Butt, C. M., Hammel, S. C., Henderson, B. B., Roman, S. A., Scheri, R. P., Stapleton, H. M., Sosa, J. A. (2017). Exposure to
flame retardant chemicals and occurrence and severity of papillary thyroid cancer: A case-control study. Environment International 107:235-242.
Endocrine-thyroid cancer (papillary)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
4161719
Domain
Metric
Rating
Comments
Metric 10:
Metric 11:
Covariate Characterization
Co-exposure Counfounding
Medium
Low
A questionnaire was administered to participants, but the method for assessing covari-
ates was not described, although the study indicated that protocols were approved by an
institutional review board. Did not report if the questionnaire was validated or the types
of data collected in the questionnaire, but there is no evidence of confounding.
Many flame retardants were measured in household dust samples. The authors acknowl-
edge and provide data on those most highly correlated. Statistical analyses modeled
each flame retardant separately. The authors indicate that component analyses were con-
ducted to systematically assess FR mixtures but did not provide them in the article. They
stated that they "did not provide any additional insights."
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High
Medium
Medium
High
The case-control study design was appropriate for the research question and applicable
statistical methods were used (logistic regression models). However, the co-exposures
were not accounted for in the analysis.
The number of cases and controls was adequate to detect an effect.
The description of the analysis was sufficient to be reproducible with access to the ana-
lytical data.
Model assumptions were adequately described.
Additional Comments: For TCEP, dust levels in the home are used as surrogates for exposure, measured using MS/GC. PBDEs only were measured in serum (not organophosphate
flame retardants). This case-control study used an adequate number of age- and gender-matched pairs. Multiple flame retardants were present in dust and in
large ranges, depending on the flame retardant: adjustments were not made for multiple comparisons (separate models were assessed for each compound).
Overall Quality Determination High
Page 27 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 4161719 Table: 4 of 4
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Hoffman, K., Lorenzo, A., Butt, C. M., Hammel, S. C., Henderson, B. B., Roman, S. A., Scheri, R. P., Stapleton, H. M., Sosa, J. A. (2017). Exposure to
flame retardant chemicals and occurrence and severity of papillary thyroid cancer: A case-control study. Environment International 107:235-242.
Cancer/Carcinogenesis-papillary thyroid cancer
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
4161719
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Metric 2:
Metric 3:
Participant Selection
Attrition
Comparison Group
High Participant selection was well described, including timing of study, inclusion and exclu-
sion criteria and case ascertainment.
High Supplemental Figure 1 shows only 1 of 71 cases excluded and shows only 1 of 71 cases
excluded.
High Cases and controls were recruited from the same Health Care Center during the same
time period and were matched by age and gender.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6:
Temporality
High Flame retardants were measured in house dust. Collection methods and analysis of dust
was fully reported. TCEP concentrations in household dust were measured by GC/EI-
MS.
Low Levels of flame retardant exposure in the home were reported for cases and controls;
TCEP concentrations above the median concentration were reported for both groups.
Median TCEP concentrations were shown in the box plot in Fig. 1 but concentrations
were not reported. Detection limits were not reported for organophosphate flame retar-
dants.
Medium Exposure at the same residence for at least 2 years prior to diagnosis; however, it is
unclear whether exposure duration was sufficient for thyroid cancer.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
High The thyroid cancer outcome was assessed by medical review of clinical and pathology
data; pathological stage was assessed based on tumor size, location, metastasis.
High Health outcomes and exposures are reported, except for median exposure levels of each
of the flame retardants. Adjusted ORs with CIs reported for overall incidence as well
as measures of tumor aggressiveness (i.e., pathologic stage). Some data were reported
for PBDEs but not the organophosphate flame retardants, such as number of detects,
detection limits.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
High Covariate adjustments were described (age, income, BMI). Regression analyses were
adjusted for participant age and household income. BMI was both included and ex-
cluded as a covariate. Other potential confounders were considered, but not applied
because they did not alter effect estimates (race, employment status, and smoking). Ion-
izing radiation exposure was considered but no participants reported prior exposure.
Continued on next page ...
Page 28 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 4161719 Table: 4 of 4
... continued from previous page
Study Citation: Hoffman, K., Lorenzo, A., Butt, C. M., Hammel, S. C., Henderson, B. B., Roman, S. A., Scheri, R. P., Stapleton, H. M., Sosa, J. A. (2017). Exposure to
flame retardant chemicals and occurrence and severity of papillary thyroid cancer: A case-control study. Environment International 107:235-242.
Health Cancer/Carcinogenesis-papillary thyroid cancer
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
Linked HERO ID(s): No linked references.
HERO ID: 4161719
Domain
Metric
Rating
Comments
Metric 10:
Metric 11:
Covariate Characterization
Co-exposure Counfounding
Medium
Low
A questionnaire was administered to participants, but the method for assessing covari-
ates was not described, although the study indicated that protocols were approved by an
institutional review board. Did not report if the questionnaire was validated or the types
of data collected in the questionnaire, but there is no evidence of confounding.
Many flame retardants were measured in household dust samples. The authors acknowl-
edge and provide data on those most highly correlated. Statistical analyses modeled
each flame retardant separately. The authors indicate that component analyses were con-
ducted to systematically assess FR mixtures but did not provide them in the article. They
stated that they "did not provide any additional insights."
Domain 5: Analysis
Metric
12:
Study Design and Methods
High
The case-control study design was appropriate for the research question and applicable
statistical methods were used (logistic regression models). However, the co-exposures
were not accounted for in the analysis.
Metric
13:
Statistical Power
Medium
The number of cases and controls was adequate to detect an effect.
Metric
14:
Reproducibility of Analyses
Medium
The description of the analysis was sufficient to be reproducible with access to the ana-
lytical data.
Metric
15:
Statistical Analysis
High
Model assumptions were adequately described.
Additional Comments: For TCEP, dust levels in the home are used as surrogates for exposure, measured using MS/GC. PBDEs only ere measured in serum (not organophosphate
flame retardants). This case-control study used an adequate number of age- and gender-matched pairs. Multiple flame retardants were present in dust and in
large ranges, depending on the flame retardant: adjustments were not made for multiple comparisons (separate models were assessed for each compound).
Overall Quality Determination High
Page 29 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 6747922 Table: 1 of 1
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Li, Y., Fu, Y., Hu, K., Zhang, Y., Chen, J., Zhang, S., Zhang, B., Liu, Y. (2020). Positive correlation between human exposure to organophosphate esters
and gastrointestinal cancer in patients from Wuhan, China. Ecotoxicology and Environmental Safety 196:110548.
Cancer/Carcinogenesis-Gastrointestinal cancer
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
6747922
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
Medium Most key elements of study design and participation selection were described. Con-
firmed through pathology, cancer cases (n=74) were selected from a hospital in Wuhan,
China. However, information on how controls (n=62) were selected was less transparent.
It only mentions that they were "healthy enough to donate blood".
Medium There was minimal subject withdrawal from the study, and the outcome and exposure
data seem to have been largely complete.
Medium There were some evidence that cases and controls were similar. For example, cases and
controls on average had similar age and were recruited in the same time frame.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure High Quantitative measurement of chemicals/organophosphate esters were measured in
plasma using HPLC-MS/MS and electrospray positive ionization methods.
Metric 5: Exposure Levels Low Range of exposure in the population were limited. Several chemicals, including TCEP,
were not readily detected among cases and controls.
Metric 6: Temporality Low Temporality was established, exposures were assessed after cancer was diagnosed. How-
ever, it was unclear whether exposures, including TCEP, fall within relevant exposures
windows for the outcome of interest.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
Medium Cases were assessed using appropriate methods; however, not enough information of
how it was done is provided. Confirmed through pathology, cancer cases (n=74) were
selected from a hospital in Wuhan, China.
Medium An appropriate description of the outcome was reported in the abstract, introduction and
methods. The odds ratio as the main effect estimates were reported with their respective
confidence intervals (CI). Numbers of cases and controls were detailed for the logistic
regressions, but they seemed to be very low based on the reported odds ratio and 95%
CI.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment Low There was some adjustments for potential confounders in the logistic regression mod-
els. To test the main hypothesis, the authors adjusted for age, gender and stage, not for
smoking and other potential confounders.
Continued on next page ...
Page 30 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 6747922 Table: 1 of 1
... continued from previous page
Study Citation: Li, Y., Fu, Y., Hu, K., Zhang, Y., Chen, J., Zhang, S., Zhang, B., Liu, Y. (2020). Positive correlation between human exposure to organophosphate esters
and gastrointestinal cancer in patients from Wuhan, China. Ecotoxicology and Environmental Safety 196:110548.
Health Cancer/Carcinogenesis-Gastrointestinal cancer
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
Linked HERO ID(s): No linked references.
HERO ID: 6747922
Domain
Metric
Rating
Comments
Metric 10:
Covariate Characterization
Medium
A few potential confounders were obtained and assessed through clinical records. How-
ever, the authors do not mention why other important potential confounders were in-
cluded in the study.
Metric 11:
Co-exposure Counfounding
Low
There were direct evidence of potential co-exposures, but were not appropriately ad-
justed for in the statistical approaches and/or in the logistic regression models.
Domain 5: Analysis
Metric 12:
Study Design and Methods
High
The study design chosen was appropriate for the research question to investigate the
associations between TCEP exposures and gastrointestinal cancer. The study used a
variety of statistical approaches, including logistic regression models, appropriate to
address the research question.
Metric 13:
Statistical Power
Medium
The number of participants (cases, n=74; controls, n=62) was adequate to detect an
effect in the population; however, not adequate for a comprehensive subgroup analysis.
Thus, this metric is borderline medium with a lean towards low.
Metric 14:
Reproducibility of Analyses
Low
Description of statistical analyses is limited, thus it would be difficult to reproduce the
authors' approach. Information about the treatment of missing values not provided, and
the language discussing the Mann-Whitney U test is not entirely clear (unsure about
"comparing the two groups" - cases and controls?).
Metric 15:
Statistical Analysis
Low
The statistical models (logistic regression models) building process was appropriate
and model assumptions were met (with respect to the outcome) to address the research
question. However, the authors do not mentioned how they treated the independent
variables and covariates in the models.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric
16:
Use of Biomarker of Exposure
Metric
17:
Effect Biomarker
Metric
18:
Method Sensitivity
Metric
19:
Biomarker Stability
Metric
20:
Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High
N/A
Medium
Medium
High
High
N/A
Biomarkers of exposure/organophosphate esters (including TCEP) were measured in
plasma using HPLC-MS/MS and electrospray positive ionization methods.
An effect biomarker was not assessed.
Analytical methods to detect and measure biomarkers were sensitive.
Some description of the sample collection and storage (at -20 degrees Celsius) until use
for analytical purposes was provided.
Some information on sample handling to avoid or reduce potential contamination of the
samples was described. Quality control measures were also taken into account to ensure
reliable data.
Information on instrumentation(s) that allowed for identification of the biomarkers,
including TCEP, with a high degree of confidence and the required sensitivity was pro-
vided.
Matrix adjustment is not necessary for plasma biomarker samples.
Continued on next page ...
Page 31 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 6747922 Table: 1 of 1
... continued from previous page
Study Citation: Li, Y., Fu, Y., Hu, K., Zhang, Y., Chen, J., Zhang, S., Zhang, B., Liu, Y. (2020). Positive correlation between human exposure to organophosphate esters
and gastrointestinal cancer in patients from Wuhan, China. Ecotoxicology and Environmental Safety 196:110548.
Health Cancer/Carcinogenesis-Gastrointestinal cancer
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
Linked HERO ID(s): No linked references.
HERO ID: 6747922
Domain Metric Rating Comments
Additional Comments: In this study, the authors examined the associations between investigated the associations between organophosphate esters, including TCEP, and the increase
risk for gastrointestinal cancer in Wuhan, China. Overall, most of the methodology used in this study were apparently adequate. However, there were a
number of chemicals that gave high frequencies of not detected values among the studied population and were not possible to be linked with the outcome
of interest. In terms of the results, there was an increased risk for gastrointestinal cancer and exposure to organophosphate esters. Nonetheless, TCEP was
not associated with gastrointestinal cancer due high frequencies of not detected values.
Overall Quality Determination Medium
Page 32 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364983 Table: 1 of 1
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liao, K., Zhao, Y., Qu, J., Yu, W., Hu, S., Fang, S., Zhao, M., Jin, H. (2023). Organophosphate esters concentrations in human serum and their associations
with Sjogren syndrome. Environmental Pollution 331(Pt 1): 121941.
Immune/Hematological-Sjogren's syndrome
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364983
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Low
Metric 2: Attrition
Metric 3: Comparison Group
Medium
Low
This study, conducted in China, analyzed associations between several organophosphate
esters (OPEs) including tris (2-chloroethyl) phosphate [TCEP], and Sjogren's syndrome
(SjS). The analysis sample included 145 healthy volunteers (controls) and 138 patients
with SjS (cases) recruited from a hospital in Hangzhou, Zhejiang Province, China in
2021-2022. Information on participant selection was sparse. Recruitment methods and
refusal rates were not described. There are concerns related to participant selection, but
no direct evidence of selection bias.
Of 318 participants recruited for the study, 35 participants (11 %) were excluded due to
missing demographic information or insufficient serum for OPE analysis. Attrition was
not described by case status. However, the overall attrition of 11 % is relatively low.
Controls were recruited from the same hospital as cases during the same time period.
Controls were described as individuals "from the physical examination population,"
without much additional detail. Controls were matched to cases based on gender, which
is appropriate and important because Sjogren's Syndrome is an autoimmune disease
that is substantially more common in women than in men. A potential limitation is that
matching wasn't also performed based on age. The mean age in cases was significantly
higher than controls (61.46 ± 14.80 vs. 48.4± 10.57 years); however, the statistical
analyses adjusted for age. Eligibility criteria were not specified for controls. It is there-
fore uncertain whether controls were screened to exclude diagnoses of other conditions
potentially associated with OPE exposure. This uncertainty, common to hospital-based
case control studies, is a concern, but there is no direct evidence of bias.
Domain 2: Exposure Characterization
Continued on next page .
Page 33 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364983 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liao, K., Zhao, Y., Qu, J., Yu, W., Hu, S., Fang, S., Zhao, M., Jin, H. (2023). Organophosphate esters concentrations in human serum and their associations
with Sjogren syndrome. Environmental Pollution 331(Pt 1): 121941.
Immune/Hematological-Sjogren's syndrome
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364983
Domain
Metric
Rating
Comments
Metric 4: Measurement of Exposure
Medium
Metric 5: Exposure Levels
Metric 6: Temporality
Medium
Low
Exposure to several OPEs, including tris (2-chloroethyl) phosphate [TCEP], was as-
sessed using serum samples. Samples were analyzed using high performance liquid
chromatography with mass spectrometry, using internal standards and quality controls.
The authors stated that no OPEs were detected in procedural blanks (ultra-pure water).
The mean (SE) % recovery for TCEP was 100 ±11. TCEP was detected in 95% of
serum samples. Values below detection limits were replaced by the LOD divided by the
square root of 2. There is limited data on the utility or validity of estimating TCEP expo-
sure using serum samples. Issues such as the half-life of TCEP in serum, or the utility of
estimating exposure using serum with or without including measures of TCEP metabo-
lites, were not discussed.The authors did not adjust TCEP concentrations for serum lipid
levels or adjust for lipids in statistical models. Serum TCEP concentrations in studies
cited in the Discussion section were lipid-adjusted. However, OPE chemicals are often
partly hydrophilic, so it is uncertain whether lipid adjustment is important for the analy-
ses using serum measures of TCEP to estimate exposure.
TCEP exposure was analyzed using both a log transformed continuous variable and
quartiles. TCEP was detectable in serum of 95% of participants. The mean was 0.54
ng/mL, and the range was from
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364983 Table: 1 of 1
... continued from previous page
Study Citation: Liao, K., Zhao, Y., Qu, J., Yu, W., Hu, S., Fang, S., Zhao, M., Jin, H. (2023). Organophosphate esters concentrations in human serum and their associations
with Sjogren syndrome. Environmental Pollution 331(Pt 1): 121941.
Health Immune/Hematological-Sjogren's syndrome
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
Linked HERO ID(s): No linked references.
HERO ID: 11364983
Domain Metric Rating Comments
Metric 7:
Outcome Measurement or
Characterization
Medium
The outcome assessed in this study, Sjogren's syndrome (SjS), is a rare autoimmune dis-
ease. The authors described cases as diagnosed based on American-European Consensus
Group classification criteria, which include the presence of auto-antibodies (SjS-related
antigen A or B), or focal lymphocytic sialadenitis detected in salivary gland biopsies.
They stated that these criteria have high sensitivity (93.5%) and specificity (94%). The
study did not discuss whether the cases included were limited to primary SjS, or whether
a proportion had secondary SjS (i.e. associated with another underlying autoimmune
disorder such as rheumatoid arthritis). There was no mention of assessing autoantibodies
in controls or screening controls for either SjS or other autoimmune disorders.
Metric 8:
Reporting Bias
High
Results for associations described in the Materials and Methods section of the paper
were presented. Effect estimates were reported with measures of variability (confidence
intervals).
Domain 4: Potential Confounding / Variability Control
Metric 9:
Covariate Adjustment
Medium
Controls were matched to cases based on gender during participant selection because
Sjogren's occurs disproportionately in women, but the statistical methods didn't in-
clude matched pairs analyses. Conditional logistic regression could have been used to
account for the matching, but the study appears to have used ordinary logistic regression
and gender was omitted from the pooled model. However, some analyses were strati-
fied by gender. The authors considered other potential confounders using exclusion by
single-factor screening to develop a final model that included age, smoking, and alcohol
consumption. Other potential confounders (BMI, education, household income, parity,
and diet) had similar distributions in cases and controls and were excluded from the final
models. Residual confounding cannot be ruled out, but there was no evidence of sub-
stantial confounding bias.
Metric 10:
Covariate Characterization
Medium
Covariates were assessed using questionnaires. The validity of the questionnaires used
to measure confounders was not discussed. However, there was no evidence that covari-
ate measures had poor validity.
Metric 11:
Co-exposure Counfounding
Medium
TCEP was not strongly correlated with other OPE chemicals measured in this study
(Spearman's R <0.50). Joint effects of multiple OPE chemicals were also examined in
a supplementary mixtures analysis using weighted quantile sum regression. Potential
confounding by non-OPE co-exposures is possible. The authors mentioned infectious
agents, UV radiation, and drugs as potential SjS triggers, but didn't assess these expo-
sures in participants.
Domain 5: Analysis
Continued on next page ...
Page 35 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364983 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liao, K., Zhao, Y., Qu, J., Yu, W., Hu, S., Fang, S., Zhao, M., Jin, H. (2023). Organophosphate esters concentrations in human serum and their associations
with Sjogren syndrome. Environmental Pollution 331(Pt 1): 121941.
Immune/Hematological-Sjogren's syndrome
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364983
Domain
Metric
Rating
Comments
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High The authors only made a few mentions of the study type, and were inconsistent in the
study type specified. In the abstract and in one sentence in the paper, the authors state
that this is a cross-sectional study. However, in another sentence in the paper, the authors
describe the "case-control population" and in another sentence the authors refer to the
limitations of their "case-cohort design". Despite these issues with terminology and
some lacking details, the methods described in the paper are generally consistent with a
case-control design, which would be appropriate because Sjogren's syndrome is a rare
disease. Logistic regression was used, which is appropriate for a case-control study.
Medium This study sample included 145 controls and 138 cases, and there was variability in
serum TCEP concentrations (median, IQR = 0.45, 0.26-0.76 ng/mL). However, the
power to detect associations was lower in sex-stratified analyses (males: 45 cases, 51
controls; females 93 cases, 94 controls). The authors mentioned the relatively small
study sample size as a potential limitation.
Medium The statistical analysis methods were clearly described and appear to be readily repro-
ducible.
High Descriptive data were provided for case and control characteristics. Logistic regression
models were used to analyze the odds of Sjogren's syndrome associated with exposure
to TCEP and other OPE chemicals. Crude and adjusted odds ratios with 95% confidence
intervals and p-values were shown. In the overall sample, TCEP exposure was analyzed
both using quartiles and using a log transformed continuous variable. Results using
quartiles suggested that the association was non-linear and U-shaped; the utility of sex-
stratified analyses that used only a continuous exposure variable may be limited.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Effect Biomarker
Metric 18: Method Sensitivity
Medium TCEP concentrations in this study were measured in participant serum. The authors
did not cite evidence to support the validity of serum TCEP as a measure of exposure.
However, data on the use of serum vs other media (e.g. whole blood, urine, hair) to
estimate OPE exposures is limited at this time. This study did not measure metabolites
of TCEP [e.g., bis(2-chloroethyl) phosphate (BCEP)] along with the parent chemical.
It is uncertain whether including metabolites along with the parent compound would
meaningfully influence TCEP exposure estimation based on serum measures.
N/A Although biomarkers of effect (autoantibodies) are one of the criteria for SjS diagnosis,
the biomarker metrics weren't evaluated for biomarkers of effect for this study because
additional non-biomarker components are included in the diagnostic criteria, and au-
toantibody biomarkers weren't assessed in controls.
Medium TCEP was detected in 95% of serum samples from controls and 84% of samples from
cases.
Continued on next page ...
Page 36 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364983 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liao, K., Zhao, Y., Qu, J., Yu, W., Hu, S., Fang, S., Zhao, M., Jin, H. (2023). Organophosphate esters concentrations in human serum and their associations
with Sjogren syndrome. Environmental Pollution 331(Pt 1): 121941.
Immune/Hematological-Sjogren's syndrome
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364983
Domain
Metric
Rating
Comments
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
Medium Substantial details about sample storage weren't discussed. However, there is evidence
that the timeframe of sample storage was relatively short based on participants in the
study being recruited in 2021-2022 and the study being published in 2023.
Medium The authors stated that procedural blanks had no detectable levels of TCEP. However,
there was limited information on sample collection and handling protocols.
Medium The study reported using HPLC-MS with internal standards and quality controls. The
mean (SE) % recovery reported for matrix spiked TCEP samples was 100 ± 11. There
was no evidence to suggest important deficiencies in measurement methods.
Medium The authors did not adjust TCEP concentrations for serum lipid levels or adjust for
lipids in statistical models. In the discussion section, serum TCEP concentrations de-
scribed in other studies were lipid-adjusted. However, as OPE chemicals are often partly
hydrophilic, it is unclear whether lipid adjustment is important for analyses using serum
measures of TCEP.
Additional Comments: This study assessed the association between serum TCEP concentrations and Sjogren's syndrome (SjS) in a study of 138 SjS cases and 145 controls in
Hangzhou, China. Increased serum levels of a combination of several organophosphate esters were found to be associated with an increased odds of SjS.
However, the only statistically significant associations found between TCEP serum levels and SjS were for an analysis using exposure quartiles in which
the observed relationship with SjS was non-linear. Compared to the lowest TCEP exposure quartile (first quartile), the odds of SjS were significantly lower
in the second and third quartile and higher in the fourth quartile in the crude analyses. After adjusting for potential confounders, the inverse associations
observed for the second and third quartile remained statistically significant, but the positive association for the fourth quartile was attenuated and no
longer statistically significant (p = 0.143). When TCEP was log-transformed and treated as a continuous variable, no statistically significant increase
in odds of SjS was found in the crude or adjusted models. Strengths of this study include a clinical assessment of SjS that was reported to have high
specificity and sensitivity. However, the authors did not specify whether cases were limited to primary SjS vs both primary and secondary disease. Other
limitations include a lack of details on potential temporal changes in exposure, lack of detailed health information on the controls and lack of description
of diagnosis date and treatments for the cases. The potential for residual confounding of the assessed associations cannot be ruled out. Sjogren's syndrome
occurs disproportionately in females. Although controls were matched to cases based on gender during participant selection, the statistical analyses didn't
account for this matching, and some but not all analyses were stratified based on gender. Controls and cases weren't matched by age, but the analyses
adjusted for age, and other potential confounders were also considered in modeling. Some aspects of the study design terminology and methods could
have been discussed with more clarity and detail. There are substantial temporality concerns relevant to interpreting the findings of this study, particularly
the limitation that the exposure (serum TCEP levels) was only assessed at a single timepoint, which didn't precede the development of the outcome (SjS
status).
Overall Quality Determination Medium
Page 37 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364830 Table: 1 of 2
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liu, M., Li, A., Meng, L., Zhang, G., Guan, X., Zhu, J., Li, Y., Zhang, Q., Jiang, G. (2022). Exposure to Novel Brominated Flame Retardants and
Organophosphate Esters and Associations with Thyroid Cancer Risk: A Case-Control Study in Eastern China. Environmental Science & Technology
56(24): 17825-17835.
Endocrine-Papillary thyroid cancer; Thyroid hormone levels: triiodothyronine (T3) free T3, thyroxine (T4), free FT4, thyroid stimulating hormone (TSH)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364830
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
Medium Cross-sectional study examining exposure to flame retardant chemicals and associa-
tions with papillary thyroid cancer, and thyroid hormones among controls. Initially, 350
patients with papillary thyroid cancer were recruited, along with 784 controls, from a
hospital in Shandong Province between 2020 and 2021. The analysis sample included
242 cases and 239 controls. Eligibility criteria were age >18 years, more than 6 months
residing in the area, sufficient serum for analysis, and availability of triglyceride and
cholesterol measures. Information on several aspects of participant selection was lim-
ited, including the number of individuals excluded for each reason. Recruitment meth-
ods and refusal rates were not described. The study did not describe how cases were
identified and selected, did not discuss time elapsed since initial cancer diagnosis, and
did not mention whether cases had other diagnoses (i.e., health status might have influ-
enced TCEP exposure-related behaviors such as dietary habits). Indeed, 12 cases had
diabetes. There are several concerns related to participant selection, but no direct evi-
dence of selection bias.
Low The study initially recruited 350 thyroid cancer patients and 784 controls. After exclu-
sions based on eligibility, the sample included 242 (69.1%) cases and 239 (30.5%) con-
trols. It was unclear why participants, in particular a high proportion of controls, were
excluded: attrition due to specific eligibility criteria was not described. Though the lack
of information is a concern, there was no evidence of bias associated with exclusions of
either cases or controls.
Low Controls were recruited from the same hospital as cases, during the same time period.
Controls were randomly selected individuals receiving a physical exam, matched to
cases on age (within 5 years) and sex, and screened to ensure healthy thyroid function
(including ultrasounds). Eligibility criteria for controls was the same as cases. Infor-
mation on several aspects of control selection was limited. The study did not describe
whether physical exams used to identify controls were routine well checks, and there
was limited information on health status. Because two controls and 12 cases had dia-
betes, models adjusted for diabetes, but diagnoses with other conditions is uncertain.
There are several concerns related to control selection, but no direct evidence of bias.
Domain 2: Exposure Characterization
Continued on next page .
Page 38 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364830 Table: 1 of 2
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liu, M., Li, A., Meng, L., Zhang, G., Guan, X., Zhu, J., Li, Y., Zhang, Q., Jiang, G. (2022). Exposure to Novel Brominated Flame Retardants and
Organophosphate Esters and Associations with Thyroid Cancer Risk: A Case-Control Study in Eastern China. Environmental Science & Technology
56(24): 17825-17835.
Endocrine-Papillary thyroid cancer; Thyroid hormone levels: triiodothyronine (T3) free T3, thyroxine (T4), free FT4, thyroid stimulating hormone (TSH)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364830
Domain
Metric
Rating
Comments
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6: Temporality
Medium A single serum sample was used to measure levels of nine OPEs, including tris (2-
chloroethyl) phosphate [TCEP], in all study participants. Samples were analyzed using
gas chromatography—triple quadrupole mass spectrometry with standards and quality
controls that included procedural blanks. The mean (SE) % recovery for TCEP during
calibration was 119 ± 12.4; TCEP was detected in 64% of cases and 74% of controls.
Values below method detection limits (MDL) were replaced by MDL/2. Use of serum
samples to estimate TCEP exposure is not well-established, and issues such as half-life
were not discussed. However, there is no evidence that measures were not appropriate.
Medium TCEP exposure was analyzed using approximate exposure quartiles, or natural log trans-
formed continuous variables. Levels were detectable in 74% of controls and 64% of
cases. Though concentrations were low, there was no evidence of insufficient variability
to estimate associations.
Low As is common with case control studies of cancer outcomes, temporality of exposure
and outcome is uncertain. There was no information on the timing of sample collection
relative to date of diagnosis: disease-related changes in metabolism, cancer treatments,
or shifts in exposure-related behaviors among cases might have affected TCEP mea-
sures. In addition, residential mobility among participants beyond the 6-month eligibil-
ity requirement was not described. Changes in residence might affect whether current
measures reflect exposure levels in indoor environments during the period relevant for
cancer genesis. There were several concerns, but no direct evidence that exposure lev-
els were influenced by factors such as changes in metabolism, by cancer treatments, or
shifts in behavior.
Domain 3: Outcome Assessment
Metric 7: Outcome Measurement or Medium Outcomes included diagnosed papillary thyroid cancer and serum levels of several thy-
Characterization r°id hormones. (1) Thyroid cancer: Cases were recruited from a hospital and described
as having diagnosed cancer. An important strength is that controls were screened to en-
sure normal thyroid function, though criteria used were not specified. Including a single
type of thyroid cancer was also a strength. However, heterogeneity in disease stage or
tumor characteristics, potentially related to etiology, was not discussed. As noted earlier,
there was no discussion of time since cancer diagnosis. (2) Thyroid hormones: Levels
of thyroid hormones were measured in serum samples collected from controls using an
electro-chemiluminescence immunoassay analyzer. Cancer cases were not included in
these analyses. Measures included triiodothyronine (T3) free T3, thyroxine (T4), free
FT4, and thyroid stimulating hormone (TSH). The distribution of hormone levels indi-
cated substantial variability, and that a few participants had values slightly outside of
reference ranges. Nonetheless, there was no evidence that this issue would meaningfully
influence results.
Continued on next page ...
Page 39 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364830 Table: 1 of 2
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liu, M., Li, A., Meng, L., Zhang, G., Guan, X., Zhu, J., Li, Y., Zhang, Q., Jiang, G. (2022). Exposure to Novel Brominated Flame Retardants and
Organophosphate Esters and Associations with Thyroid Cancer Risk: A Case-Control Study in Eastern China. Environmental Science & Technology
56(24): 17825-17835.
Endocrine-Papillary thyroid cancer; Thyroid hormone levels: triiodothyronine (T3) free T3, thyroxine (T4), free FT4, thyroid stimulating hormone (TSH)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364830
Domain
Metric
Rating
Comments
Metric 8:
Reporting Bias
Medium
All measured outcomes were analyzed as described in the study aims. Results are re-
ported with measures of variability. However, the number of cases included was not
characterized.
Domain 4: Potential Confounding / Variability Control
Metric 9:
Covariate Adjustment
Medium
Models analyzing TCEP and thyroid cancer adjusted for age, sex, BMI, current smoking
status, current alcohol drinking, and diabetes status. Models analyzing thyroid hormones
among controls adjusted for the same variables. Confounders appear to have been se-
lected a priori. Residual confounding by factors such as past smoking or alcohol use,
weight or dietary changes, and socioeconomic factors is a potential concern, but there is
no evidence of bias.
Metric 10:
Covariate Characterization
Medium
The validity of questionnaires used to measure confounders were not discussed. How-
ever, there was no evidence that covariate measures had poor validity.
Metric 11:
Co-exposure Counfounding
Medium
Numerous OPEs were measured in participant serum samples. Weighted quantile sum
regression was used to analyze associations with a chemical mixture including co-
exposures. Moreover, TCEP was not strongly correlated with other chemicals measured
in this study.
Domain 5: Analysis
Metric 12:
Study Design and Methods
High
This case-control study in China analyzed associations between thyroid cancer and
several organophosphate esters (OPEs) and flame retardants (FRs), including tris (2-
chloroethyl) phosphate [TCEP], The use of a case control study design was appropriate
for the research question given that thyroid cancer is a rare disease. A crosss-sectional
analysis was also appropriate for evaluating how TCEP exposure was associated with
thyroid hormone levels among adults with normal thyroid function.
Metric 13:
Statistical Power
Medium
The analysis sample included 242 (87 male, 155 female) cases and 239 (88 male, 151
female) controls. Concentrations of TCEP were low, but there was variability in ex-
posure. Analysis of thyroid hormone levels were limited to controls, but power was
increased by the use of continuous outcome measures. Power to detect associations was
lower in sex-stratified than in pooled analyses. However, there was no evidence of inade-
quate sensitivity.
Metric 14:
Reproducibility of Analyses
Medium
Analysis methods were clearly described and readily reproducible.
Continued on next page ...
Page 40 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364830 Table: 1 of 2
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liu, M., Li, A., Meng, L., Zhang, G., Guan, X., Zhu, J., Li, Y., Zhang, Q., Jiang, G. (2022). Exposure to Novel Brominated Flame Retardants and
Organophosphate Esters and Associations with Thyroid Cancer Risk: A Case-Control Study in Eastern China. Environmental Science & Technology
56(24): 17825-17835.
Endocrine-Papillary thyroid cancer; Thyroid hormone levels: triiodothyronine (T3) free T3, thyroxine (T4), free FT4, thyroid stimulating hormone (TSH)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364830
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Descriptive data were provided for case and control characteristics, and the distribution
of thyroid hormones was shown. Logistic regression models that included matching
variables were used to analyze thyroid cancer, and linear regression was used to analyze
thyroid hormone outcomes. TCEP was analyzed using quartiles or natural log trans-
formed continuous variables. Thyroid hormone outcomes were also log transformed in
linear regression models.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Effect Biomarker
Metric 18: Method Sensitivity
Metric 19: Biomarker Stability
Medium TCEP concentrations were measured in patient serum. As there is limited data on the
use of serum to estimate of OPE exposures at this time, the authors did not cite evidence
to directly support the validity of serum TCEP as a measure of exposure. However, they
cited other studies in China with comparable levels of TCEP in serum, as well as studies
with considerably higher levels. Given the short half-life of OPE metabolites in urine,
the authors suggested use of serum may be an advantage as this medium may reflect a
longer exposure time window but did not cite supporting evidence of utility of a single
serum sample to quantify exposure. A potential limitation is that metabolites of TCEP
[major metabolite bis(2-chloroethyl) phosphate (BCEP)] were not assayed in addition
to the parent chemical. There is however no evidence that exposure rankings based
on metabolites vs the parent compound might differ. Despite potential concerns, there
was no evidence that measures had poor validity, and no evidence of either substantial
exposure misclassification or bias.
High Serum measures of thyroid hormones were analyzed to estimate associations between
TCEP and thyroid function in healthy adults (controls). A strength of the approach was
measurement of both total and free levels of T3 and T4, along with thyroid stimulating
hormone (TSH). Thyroid hormones are typically used as diagnostic criteria for thyroid
conditions. However, outcomes were based on a single serum measure in a sample of
moderate size. Assay details were not provided, but a standard clinical method and
analyzer was used.
Medium The method detection limit was provided for TCEP exposure and was low: 8.0 x 10—2
ng/mL. Detection limits for thyroid hormones were not described, but there was no
evidence of deficiencies.
Medium Samples were not likely to have been stored for a lengthy period. There is no data on
stability, but no evidence of degradation.
Continued on next page ...
Page 41 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364830 Table: 1 of 2
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liu, M., Li, A., Meng, L., Zhang, G., Guan, X., Zhu, J., Li, Y., Zhang, Q., Jiang, G. (2022). Exposure to Novel Brominated Flame Retardants and
Organophosphate Esters and Associations with Thyroid Cancer Risk: A Case-Control Study in Eastern China. Environmental Science & Technology
56(24): 17825-17835.
Endocrine-Papillary thyroid cancer; Thyroid hormone levels: triiodothyronine (T3) free T3, thyroxine (T4), free FT4, thyroid stimulating hormone (TSH)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364830
Domain
Metric
Rating
Comments
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
Medium Lab procedures including precleaning glass containers using pesticide-grade solvents.
The study does not discuss whether contamination was an issue of concern for TCEP
or discuss detection of TCEP in procedural blanks. However, MDL for TCEP was low
(8.0 x 10—2 ng/mL) and there was a moderate proportion of samples with levels below
MDL (26% in controls, 36% in cases). There was no evidence to suggest contamination
problems.
Medium The laboratory methods described appeared to be appropriate. Samples were analyzed
using gas chromatography—triple quadrupole mass spectrometry with reference stan-
dards for calibration, and the use of quality controls that included procedural blanks.
The mean (SE) % recovery for TCEP during calibration was 119 ± 12.4.
Medium TCEP exposure variables were analyzed adjusted for lipid weight, as in numerous other
studies. Total cholesterol and triglycerides were used for this adjustment.
Additional Comments: This hospital-based case control (242 cases, 239 controls) study in Shandong, China analyzed the relationship between serum levels of tris (2-chloroethyl)
phosphate [TCEP] and papillary thyroid cancer. The study found an association between TCEP and thyroid cancer that was negative overall, with null and
non-linear associations after stratifying by sex. The authors cited evidence of potential sex-specific effects on thyroid function of the chemicals examined in
this study. Strengths of this analysis include screening to ensure that controls were cancer free and had healthy thyroid function. Important concerns include
limited detail on the health status of controls, presence of other diagnoses in cases (12 had diabetes), and the use of prevalent thyroid cancer cases without
discussion of diagnosis dates, treatments, or behavior changes that might have affected organophosphate chemical metabolism. Exposure measurements
were measured cross-sectionally, which limits interpretations of temporality. The study also analyzed the cross-sectional association between serum TCEP
and levels of thyroid hormones among controls. Controls were free of thyroid disease, but eligibility criteria did not clearly specify excluding other
diagnoses. Two controls had diabetes, which was included as a covariate in the analyses. Consistent with hypothyroid activity, TCEP was negatively
associated with T3, free T3, T4, and free T4 and positively associated with TCEP in females (p<0.05 except for total T4). Among male participants and
in the overall sample, there was also a significant (p<0.05) negative association between TCEP and free T4. In contrast to females, however, there was a
non-significant negative association with TSH among men. The authors stated that the use of serum to estimate TCEP exposure in this study may be an
advantage relative to studies that measure short-lived TCEP metabolites in urine samples. However, there is limited evidence on the validity of biomarkers
of TCEP exposure, and TCEP metabolites were not measured in serum along with the parent compound.
Overall Quality Determination Medium
Page 42 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364830 Table: 2 of 2
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liu, M., Li, A., Meng, L., Zhang, G., Guan, X., Zhu, J., Li, Y., Zhang, Q., Jiang, G. (2022). Exposure to Novel Brominated Flame Retardants and
Organophosphate Esters and Associations with Thyroid Cancer Risk: A Case-Control Study in Eastern China. Environmental Science & Technology
56(24): 17825-17835.
Cancer/Carcinogenesis-Papillary thyroid cancer
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364830
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
Medium Cross-sectional study examining exposure to flame retardant chemicals and associa-
tions with papillary thyroid cancer, and thyroid hormones among controls. Initially, 350
patients with papillary thyroid cancer were recruited, along with 784 controls, from a
hospital in Shandong Province between 2020 and 2021. The analysis sample included
242 cases and 239 controls. Eligibility criteria were age >18 years, more than 6 months
residing in the area, sufficient serum for analysis, and availability of triglyceride and
cholesterol measures. Information on several aspects of participant selection was lim-
ited, including the number of individuals excluded for each reason. Recruitment meth-
ods and refusal rates were not described. The study did not describe how cases were
identified and selected, did not discuss time elapsed since initial cancer diagnosis, and
did not mention whether cases had other diagnoses (i.e., health status might have influ-
enced TCEP exposure-related behaviors such as dietary habits). Indeed, 12 cases had
diabetes. There are several concerns related to participant selection, but no direct evi-
dence of selection bias.
Low The study initially recruited 350 thyroid cancer patients and 784 controls. After exclu-
sions based on eligibility, the sample included 242 (69.1%) cases and 239 (30.5%) con-
trols. It was unclear why participants, in particular a high proportion of controls, were
excluded: attrition due to specific eligibility criteria was not described. Though the lack
of information is a concern, there was no evidence of bias associated with exclusions of
either cases or controls.
Low Controls were recruited from the same hospital as cases, during the same time period.
Controls were randomly selected individuals receiving a physical exam, matched to
cases on age (within 5 years) and sex, and screened to ensure healthy thyroid function
(including ultrasounds). Eligibility criteria for controls was the same as cases. Infor-
mation on several aspects of control selection was limited. The study did not describe
whether physical exams used to identify controls were routine well checks, and there
was limited information on health status. Because two controls and 12 cases had dia-
betes, models adjusted for diabetes, but diagnoses with other conditions is uncertain.
There are several concerns related to control selection, but no direct evidence of bias.
Domain 2: Exposure Characterization
Continued on next page .
Page 43 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364830 Table: 2 of 2
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liu, M., Li, A., Meng, L., Zhang, G., Guan, X., Zhu, J., Li, Y., Zhang, Q., Jiang, G. (2022). Exposure to Novel Brominated Flame Retardants and
Organophosphate Esters and Associations with Thyroid Cancer Risk: A Case-Control Study in Eastern China. Environmental Science & Technology
56(24): 17825-17835.
Cancer/Carcinogenesis-Papillary thyroid cancer
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364830
Domain
Metric
Rating
Comments
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6: Temporality
Medium A single serum sample was used to measure levels of nine OPEs, including tris (2-
chloroethyl) phosphate [TCEP], in all study participants. Samples were analyzed using
gas chromatography—triple quadrupole mass spectrometry with standards and quality
controls that included procedural blanks. The mean (SE) % recovery for TCEP during
calibration was 119 ± 12.4; TCEP was detected in 64% of cases and 74% of controls.
Values below method detection limits (MDL) were replaced by MDL/2. Use of serum
samples to estimate TCEP exposure is not well-established, and issues such as half-life
were not discussed. However, there is no evidence that measures were not appropriate.
Medium TCEP exposure was analyzed using approximate exposure quartiles, or natural log trans-
formed continuous variables. Levels were detectable in 74% of controls and 64% of
cases. Though concentrations were low, there was no evidence of insufficient variability
to estimate associations.
Low As is common with case control studies of cancer outcomes, temporality of exposure
and outcome is uncertain. There was no information on the timing of sample collection
relative to date of diagnosis: disease-related changes in metabolism, cancer treatments,
or shifts in exposure-related behaviors among cases might have affected TCEP mea-
sures. In addition, residential mobility among participants beyond the 6-month eligibil-
ity requirement was not described. Changes in residence might affect whether current
measures reflect exposure levels in indoor environments during the period relevant for
cancer genesis. There were several concerns, but no direct evidence that exposure lev-
els were influenced by factors such as changes in metabolism, by cancer treatments, or
shifts in behavior.
Domain 3: Outcome Assessment
Metric 7: Outcome Measurement or
Characterization
Metric 8: Reporting Bias
Medium Outcomes included diagnosed papillary thyroid cancer. Cases were recruited from a
hospital and described as having diagnosed cancer. An important strength is that con-
trols were screened to ensure normal thyroid function, though criteria used were not
specified. Including a single type of thyroid cancer was also a strength. However, het-
erogeneity in disease stage or tumor characteristics, potentially related to etiology, was
not discussed. As noted earlier, there was no discussion of time since cancer diagnosis.
Medium All measured outcomes were analyzed as described in the study aims. Results are re-
ported with measures of variability. However, the number of cases included was not
characterized.
Domain 4: Potential Confounding / Variability Control
Continued on next page ...
Page 44 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364830 Table: 2 of 2
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liu, M., Li, A., Meng, L., Zhang, G., Guan, X., Zhu, J., Li, Y., Zhang, Q., Jiang, G. (2022). Exposure to Novel Brominated Flame Retardants and
Organophosphate Esters and Associations with Thyroid Cancer Risk: A Case-Control Study in Eastern China. Environmental Science & Technology
56(24): 17825-17835.
Cancer/Carcinogenesis-Papillary thyroid cancer
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364830
Domain
Metric
Rating
Comments
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium Models analyzing TCEP and thyroid cancer adjusted for age, sex, BMI, current smoking
status, current alcohol drinking, and diabetes status. Models analyzing thyroid hormones
among controls adjusted for the same variables. Confounders appear to have been se-
lected a priori. Residual confounding by factors such as past smoking or alcohol use,
weight or dietary changes, and socioeconomic factors is a potential concern, but there is
no evidence of bias.
Medium The validity of questionnaires used to measure confounders were not discussed. How-
ever, there was no evidence that covariate measures had poor validity.
Medium Numerous OPEs were measured in participant serum samples. Weighted quantile sum
regression was used to analyze associations with a chemical mixture including co-
exposures. Moreover, TCEP was not strongly correlated with other chemicals measured
in this study.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High This case-control study in China analyzed associations between thyroid cancer and
several organophosphate esters (OPEs) and flame retardants (FRs), including tris (2-
chloroethyl) phosphate [TCEP], The use of a case control study design was appropriate
for the research question given that thyroid cancer is a rare disease. A crosss-sectional
analysis was also appropriate for evaluating how TCEP exposure was associated with
thyroid hormone levels among adults with normal thyroid function.
Medium The analysis sample included 242 (87 male, 155 female) cases and 239 (88 male, 151
female) controls. Concentrations of TCEP were low, but there was variability in ex-
posure. Analysis of thyroid hormone levels were limited to controls, but power was
increased by the use of continuous outcome measures. Power to detect associations was
lower in sex-stratified than in pooled analyses. However, there was no evidence of inade-
quate sensitivity.
Medium Analysis methods were clearly described and readily reproducible.
High Descriptive data were provided for case and control characteristics, and the distribution
of thyroid hormones was shown. Logistic regression models that included matching
variables were used to analyze thyroid cancer, and linear regression was used to analyze
thyroid hormone outcomes. TCEP was analyzed using quartiles or natural log trans-
formed continuous variables. Thyroid hormone outcomes were also log transformed in
linear regression models.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Continued on next page ...
Page 45 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364830 Table: 2 of 2
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Liu, M., Li, A., Meng, L., Zhang, G., Guan, X., Zhu, J., Li, Y., Zhang, Q., Jiang, G. (2022). Exposure to Novel Brominated Flame Retardants and
Organophosphate Esters and Associations with Thyroid Cancer Risk: A Case-Control Study in Eastern China. Environmental Science & Technology
56(24): 17825-17835.
Cancer/Carcinogenesis-Papillary thyroid cancer
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364830
Domain
Metric
Rating
Comments
Metric 16: Use of Biomarker of Exposure
Metric 17: Effect Biomarker
Metric 18: Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
Medium TCEP concentrations were measured in patient serum. As there is limited data on the
use of serum to estimate of OPE exposures at this time, the authors did not cite evidence
to directly support the validity of serum TCEP as a measure of exposure. However, they
cited other studies in China with comparable levels of TCEP in serum, as well as studies
with considerably higher levels. Given the short half-life of OPE metabolites in urine,
the authors suggested use of serum may be an advantage as this medium may reflect a
longer exposure time window but did not cite supporting evidence of utility of a single
serum sample to quantify exposure. A potential limitation is that metabolites of TCEP
[major metabolite bis(2-chloroethyl) phosphate (BCEP)] were not assayed in addition
to the parent chemical. There is however no evidence that exposure rankings based
on metabolites vs the parent compound might differ. Despite potential concerns, there
was no evidence that measures had poor validity, and no evidence of either substantial
exposure misclassification or bias.
High Serum measures of thyroid hormones were analyzed to estimate associations between
TCEP and thyroid function in healthy adults (controls). A strength of the approach was
measurement of both total and free levels of T3 and T4, along with thyroid stimulating
hormone (TSH). Thyroid hormones are typically used as diagnostic criteria for thyroid
conditions. However, outcomes were based on a single serum measure in a sample of
moderate size. Assay details were not provided, but a standard clinical method and
analyzer was used.
Medium The method detection limit was provided for TCEP exposure and was low: 8.0 x 10—2
ng/mL. Detection limits for thyroid hormones were not described, but there was no
evidence of deficiencies.
Medium Samples were not likely to have been stored for a lengthy period. There is no data on
stability, but no evidence of degradation.
Medium Lab procedures including precleaning glass containers using pesticide-grade solvents.
The study does not discuss whether contamination was an issue of concern for TCEP
or discuss detection of TCEP in procedural blanks. However, MDL for TCEP was low
(8.0 x 10—2 ng/mL) and there was a moderate proportion of samples with levels below
MDL (26% in controls, 36% in cases). There was no evidence to suggest contamination
problems.
Medium The laboratory methods described appeared to be appropriate. Samples were analyzed
using gas chromatography—triple quadrupole mass spectrometry with reference stan-
dards for calibration, and the use of quality controls that included procedural blanks.
The mean (SE) % recovery for TCEP during calibration was 119 ± 12.4.
Medium TCEP exposure variables were analyzed adjusted for lipid weight, as in numerous other
studies. Total cholesterol and triglycerides were used for this adjustment.
Continued on next page ...
Page 46 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364830 Table: 2 of 2
... continued from previous page
Study Citation: Liu, M., Li, A., Meng, L., Zhang, G., Guan, X., Zhu, J., Li, Y., Zhang, Q., Jiang, G. (2022). Exposure to Novel Brominated Flame Retardants and
Organophosphate Esters and Associations with Thyroid Cancer Risk: A Case-Control Study in Eastern China. Environmental Science & Technology
56(24): 17825-17835.
Health Cancer/Carcinogenesis-Papillary thyroid cancer
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
Linked HERO ID(s): No linked references.
HERO ID: 11364830
Domain Metric Rating Comments
Additional Comments: This hospital-based case control (242 cases, 239 controls) study in Shandong, China analyzed the relationship between serum levels of tris (2-chloroethyl)
phosphate [TCEP] and papillary thyroid cancer. The study found an association between TCEP and thyroid cancer that was inverse overall, with null and
non-linear associations after stratifying by sex. The authors cited evidence of potential sex-specific effects on thyroid function of the chemicals examined in
this study. Strengths of this analysis include screening to ensure that controls were cancer free and had healthy thyroid function. Important concerns include
limited detail on the health status of controls, presence of other diagnoses in cases (12 had diabetes), and the use of prevalent thyroid cancer cases without
discussion of diagnosis dates, treatments, or behavior changes that might have affected organophosphate chemical metabolism. Exposure measurements
were measured cross-sectionally, which limits interpretations of temporality. The authors stated that the use of serum to estimate TCEP exposure in this
study may be an advantage relative to studies that measure short-lived TCEP metabolites in urine samples. However, there is limited evidence on the
validity of biomarkers of TCEP exposure, and TCEP metabolites were not measured in serum along with the parent compound.
Overall Quality Determination Medium
Page 47 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 7537904 Table: 1 of 1
Study Citation: Liu, Y., Li, Y., Dong, S., Han, L., Guo, R., Fu, Y., Zhang, S., Chen, J. (2021). The risk and impact of organophosphate esters on the development of
female-specific cancers: Comparative analysis of patients with benign and malignant tumors. Journal of Hazardous Materials 404(Pt B): 124020.
Health Cancer/Carcinogenesis-Benign breast tumor; breast cancer; benign tumor of the uterus; cervical cancer.
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
7537904
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
Medium Study population recruited from a 3A hospital in Wuhan, China. Number of participants
reported but participation rate not provided. The study included female patients with
tumor/cancer disease but other inclusion and exclusion criteria not specified. Cancer
cases was confirmed by two independent pathologists after histopathological analysis.
High No subject was reported withdrawal from the study. The outcome data and exposure
measurement were complete for the study participants.
Low The study recruited patients with malignant tumor (cancer cases) and benign tumor.
Comparison was made between cancer and benign tumor patients only. The control
group of non-tumor population was not recruited. The similarity between benign and
cancer groups is a concern, and lack of description of similarity.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure
Metric 5:
Metric 6:
Exposure Levels
Temporality
High Exposure chemicals were extracted from plasma samples using liquid-liquid extraction
method, and directly measured using HPLC-MS/MS.
Medium Continuous exposure chemical concentrations were measured for each patient. The
range is adequate to detect the exposure-outcome association.
Low Due to the nature of cross-sectional design of this study, temporality is not well-
established. The study only reported age of participants, other temporality supporting
information was not provided. It is unclear whether exposures fall within relevant expo-
sure windows for the outcomes.
Domain 3: Outcome Assessment
Metric 7: Outcome Measurement or
Characterization
Metric 8: Reporting Bias
High Cancer diagnosis performed using postoperative histopathological analysis of tumors,
then confirmed by 2 independent pathologists. The measurement of outcomes has high
level of certainty.
High Measured outcomes were reported in the methods section. Number of subjects in each
case group and reference group provided. The effect estimates were reported as ORs
with 95% CI in both adjusted and unadjusted model. Spearman correlations of each
chemical in each group were reported. The results are fully tabulated in this study.
Domain 4: Potential Confounding / Variability Control
Continued on next page ...
Page 48 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 7537904 Table: 1 of 1
... continued from previous page
Study Citation: Liu, Y., Li, Y., Dong, S., Han, L., Guo, R., Fu, Y., Zhang, S., Chen, J. (2021). The risk and impact of organophosphate esters on the development of
female-specific cancers: Comparative analysis of patients with benign and malignant tumors. Journal of Hazardous Materials 404(Pt B): 124020.
Health Cancer/Carcinogenesis-Benign breast tumor; breast cancer; benign tumor of the uterus; cervical cancer.
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
Linked HERO ID(s): No linked references.
HERO ID: 7537904
Domain
Metric
Rating
Comments
Metric 9:
Covariate Adjustment
Medium
The covariates include age, hypertension, liver function, urine routine tests results, and
renal function. Covariates were adjusted in the adjusted model. Distribution of covari-
ates among study population was provided, but a description of how covariates were
selected into the model was not reported.
Metric 10:
Covariate Characterization
Medium
The covariate assessment methods or validation is not provided, but there is little to no
evidence to indicate poor validity.
Metric 11:
Co-exposure Counfounding
Low
Co-exposure of other OPE chemicals was measured. But the statistical model didn't
make appropriate adjustment for co-exposure chemicals when evaluating the association
between target chemical and outcomes.
Domain 5: Analysis
Metric 12:
Study Design and Methods
Low
The cross-sectional design of this study to evaluate the association between OPE con-
centrations and cancer status is not appropriate, since cancer is usually a long-term dis-
ease. The statistical methods used to assess the correlation of chemicals and exposure-
outcome association is appropriate.
Metric 13:
Statistical Power
Low
Overall, 258 participants included in this study but divided into 4 subgroups based on
tumor/cancer type. The number of participants in each subgroups ranged from 45 to 78.
The number of participants are limited so it's a concern if there is sufficient statistical
power to detect the effect of interest.
Metric 14:
Reproducibility of Analyses
Medium
Sufficient information of study design and analysis were provided. The description is
adequate to understand what has been done and conceptually reproducible with the
access of original data.
Metric 15:
Statistical Analysis
High
The binary logistic regression model to calculate odd ratios is appropriate. Lack of
description of inclusion and exclusion of variables in adjusted model is a concern, but
overall the analysis is transparent and appropriate.
Additional Comments: Overall, this cross-sectional study evaluated the association between organophosphate esters (OPEs) concentrations and female-specific tumor, and com-
pared the OPE levels in benign and malignant disease patients. The exposure and outcome were measured using well-established methods with high
certainty, and the analysis methods is overall appropriate. However, some concerns exist and could downgrade this study. The number of participants is
small indicates potential lack of statistical power to detect the effect of interest. The cross-sectional design is not appropriate since cancer is usually not a
acute disease, reverse causality could be an issue for this study. No health or non-cancer participants as reference group or control, so similarity of groups
is a concern.
Overall Quality Determination Low
Page 49 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364495 Table: 1 of 1
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Mendy, A., Percy, Z., Braun, J. M., Lanphear, B., Guardia, La, M. J., Hale, R. C., Yolton, K., Chen, A. (2024). Prenatal exposure to replacement flame
retardants and organophosphate esters and childhood adverse respiratory outcomes. Environmental Research 240(Pt 2): 117523.
Lung/Respiratory-Wheeze, respiratory infections, hay fever or allergies, forced expiratory volume in Is (FEV1), peak expiratory flow (PEF)
Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
No linked references.
11364495
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
Medium Participants in this prospective birth cohort study were pulled from the Health Outcomes
and Measures of the Environment (HOME) study. Pregnant women were recruited from
March 2003-January 2006 in the greater Cincinnati, Ohio metropolitan area. Details of
the HOME study are available in Braun et al. 2017 (HEROID: 6749104). In Braun et
al. 2017 inclusion criteria and recruitment are extensively detailed and there is no evi-
dence to suggest selection bias occurred in the original creation of the HOME study401
mother-child pairs were included in the original HOME study. Inclusion criteria for the
analysis in this study were: having data on dust and urinary organophosphate esters and
replacement brominated flame retardants, having data on covariates, and data on respira-
tory outcomes through follow-up until 5 years. 342 mother-child pairs were included in
the present study.In general, there is no significant evidence for selection bias. However,
no data is presented comparing those included in the study to participants who were ex-
cluded after originally participating in the HOME study.
Medium There were n=59 pregnant mother-child pair participants out of 401 who were included
in the original HOME study but not the present analysis; all exclusion reasons were
related to missing data, but exact breakdowns are not provided (ex; how many were
excluded due to missing covariate data vs. missing exposure or outcome data). However,
there is no reason to suggest that exclusion of those participants was inappropriate.
High Participants were all recruited from the same eligible population of pregnant women
in the Greater Cincinnati, Ohio metropolitan area. The study presents a table of de-
mographic characteristics split by exposure to all relevant exposures, including TCEP.
Significant differences were reported by race/ethnicity and serum cotinine (significant
differences were reported additionally for serum cotinine when assessing BCEP). All
these factors were adjusted for in statistical analysis.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure Medium Exposure to TCEP was measured in dust samples. Dust was sampled from a 1-m
squared area floor area for median sampling time of 275 seconds in the "main activity
room" of participants home at around 20 weeks gestation. TCEP was quantified using
high-performance liquid chromatography mass spectrometry.
Metric 5: Exposure Levels Medium Exposure ranges were likely large enough to allow an exposure-response estimate. For
example, TCEP in dust (5th-95th percentile): 0.06 - 9.31 ug/g dust.
Continued on next page ...
Page 50 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364495 Table: 1 of 1
... continued from previous page
Study Citation: Mendy, A., Percy, Z., Braun, J. M., Lanphear, B., Guardia, La, M. J., Hale, R. C., Yolton, K., Chen, A. (2024). Prenatal exposure to replacement flame
retardants and organophosphate esters and childhood adverse respiratory outcomes. Environmental Research 240(Pt 2): 117523.
Health Lung/Respiratory-Wheeze, respiratory infections, hay fever or allergies, forced expiratory volume in Is (FEV1), peak expiratory flow (PEF)
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
Linked HERO ID(s): No linked references.
HERO ID: 11364495
Domain
Metric
Rating
Comments
Metric 6:
Temporality
Medium
Exposure was measured in the homes of mothers during pregnancy, and outcomes were
measured every 6 months after birth. Temporality is thus established, but it is uncertain
whether or not dust concentrations of TCEP during development represent a etiologi-
cally relevant time period for respiratory outcomes of children.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
High
High
Respiratory outcomes included wheeze, respiratory infections, hay fever or allergies,
forced expiratory volume in Is (FEV1), and peak expiratory flow (PEF). Wheeze, res-
piratory infections, and hay fever or allergies were measured busting standardized ques-
tions, based on NHANES questionnaires. Questionnaires were provided every 6 months
until the age of 5 years old. FEV1 and PEF were measured at age 5 years old using
a portable spirometer, and measures were reported to follow the American Thoracic
Society criteria and conducted by trained research assistants. In general, there is no sig-
nificant evidence of outcome misclassification.
All measured outcomes outlined in the methods are reported in the results in a way that
would allow for detailed extraction (effect estimates and 95% confidence intervals).
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium
High
Medium
Covariates included child sex, child race/ethnicity, median household income at base-
line, duration of breastfeeding, gestational age, maternal and paternal allergy and
asthma, birth weight was retrieved from medical records, and prenatal exposure to
smoking. No information is provided on why these covariates were chosen, perhaps
based on a priori knowledge.
The majority of covariates were measured using questionnaires from the HOME study.
Birth weight was retrieved from medical records, and prenatal smoking was measured
using serum cotinine at ages 16 and 26 weeks of gestation.
The study measured and analyzed other co-exposures, including TCIPP, TDCIPP, TPHP,
EH-TBB, and BEH-TEBP. While these were not adjusted for in analyses of TCEP, cor-
relation coefficients were presented across all exposures.
Domain 5: Analysis
Metric 12:
Metric 13:
Study Design and Methods
Statistical Power
High
Medium
The prospective birth cohort study design is an appropriate design to assess the impact
of prenatal TCEP exposure on respiratory outcomes. The use of generalized estimating
equation analysis is appropriate to account for repeated measures.
The number of participants in the analysis (n=342) is likely large enough to detect an
effect.
The description of the analysis is sufficiently detailed so that the results can be repro-
duced.
Metric 14:
Reproducibility of Analyses
Medium
Continued on next page ...
Page 51 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364495 Table: 1 of 1
... continued from previous page
Study Citation: Mendy, A., Percy, Z., Braun, J. M., Lanphear, B., Guardia, La, M. J., Hale, R. C., Yolton, K., Chen, A. (2024). Prenatal exposure to replacement flame
retardants and organophosphate esters and childhood adverse respiratory outcomes. Environmental Research 240(Pt 2): 117523.
Health Lung/Respiratory-Wheeze, respiratory infections, hay fever or allergies, forced expiratory volume in Is (FEV1), peak expiratory flow (PEF)
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Parent compound
Linked HERO ID(s): No linked references.
HERO ID: 11364495
Domain Metric Rating Comments
Metric 15: Statistical Analysis High The association between exposure and outcome was primarily measured using gener-
alized estimating equations. The study chose a binomial distribution with a log link
function for the models due to the repeated measures of the binary respiratory outcomes.
Further, "unstructured working matrix and robust variance estimators were specified
to estimate the beta coefficients exponentiated to obtain the relative risks and their cor-
responding 95% confidence intervals." Lung function parameters only had a single
measurement and were thus assessed using linear regression modeling. Sensitivity anal-
yses to assess the potential for effect modification by sex.Concentrations of TCEP were
log-transformed to account for non-normality. TCEP was measured in 78.8% of all dust
samples, but it is not specified how values below the LOD were handled. Analyses were
presented separately for dust concentrations (ug/g dust) and dust loadings (mg/mA3).
Additional Comments: This prospective birth cohort study assessed the association between prenatal measurements of TCEP in dust samples and BCEP in maternal urine to
repeated measures of respiratory outcomes among children. In general there is a limited risk of bias in the study due to well-conducted exposure assessment
and outcome assessments. Significant positive associations were reported for TCEP and respiratory outcomes - however, these associations tended to be
inconsistent across methods of exposure measurement (dust concentrations, dust loadings, urinary BCEP).
Overall Quality Determination High
Page 52 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 7274557 Table: 1 of 2
Crawford, K. A., Hawley, N., Calafat, A. M., Jayatilaka, N. K., Froehlich, R. J., Has, P., Gallagher, L. G., Savitz, D. A., Braun, J. M., Werner, E. F., Romano,
M. E. (2020). Maternal urinary concentrations of organophosphate ester metabolites: associations with gestational weight gain, early life anthropometry,
and infant eating behaviors among mothers-infant pairs in Rhode Island. Environmental Health: A Global Access Science Source 19(1 ):97.
Reproductive /Developmental-Gestational weight gain among pregnant women; infant gestational age at delivery; infant anthropometric measure-
ments at birth and 6 weeks postpartum, including birth weight and length, head and abdominal circumference, and four body composition (iliac, subscapular,
triceps, and thigh skinfold thickness).
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
7274557
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
High Overall, all key elements of study design and participation were reported. In this pilot
study, 62 women and infants were recruited. Inclusion and exclusion criteria as well as
participant selection were fully described.
High A minimal number of subjects (6/62) were excluded from further analysis due to with-
drawal, miscarriage, and lost to follow up. Exposure and outcome measurements were
complete among pregnant women and infants. The exclusion or loss of follow up are not
likely to introduce bias since the participated population is likely to represent the general
eligible population.
High Differences in baseline characteristics were reported and adequately considered and
adjusted in the statistical analysis. Identified effect modification by infant sex were
considered in further analysis. Participants were similar since they were recruited from
the same setting using same inclusion criteria.
Domain 2: Exposure Characterization
Metric 4:
Metric 5:
Metric 6:
Measurement of Exposure
Exposure Levels
Temporality
High Exposure levels of the target chemical were derived from metabolite concentrations in
urine samples. Collection and quantification methods were fully described.
Medium Continuous individual exposure levels measured from OPE metabolite concentrations in
pooled urine samples were used.
Medium The authors reported that OPE metabolites in urine samples showed good reproducibil-
ity and good intraclass correlation. They used pooled urine samples collected at ges-
tation weeks 12, 28 and 35 to represent the exposure window. OPE chemicals have
relatively short half lives and very low bioaccumulation rate, so the OPE chemicals are
commonly used to indicate persistent exposure. It is not clear whether the exposures fell
within relevant exposure windows for the outcomes of interest.
Domain 3: Outcome Assessment
Metric 7: Outcome Measurement or
Characterization
Medium Gestational weight gain (GWG) among nine pregnant women were substituted because
of missing information. Newborn infants' anthropometry outcomes were measured
twice by staff. A third measurement was applied if differences were out of pre-specified
range. The same measurements were applied at 6 weeks postpartum. Significant mea-
surement error is not likely to be present but not using the gold-standard.
Continued on next page ...
Page 53 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 7274557 Table: 1 of 2
... continued from previous page
Crawford, K. A., Hawley, N., Calafat, A. M., Jayatilaka, N. K., Froehlich, R. J., Has, P., Gallagher, L. G., Savitz, D. A., Braun, J. M., Werner, E. F., Romano,
M. E. (2020). Maternal urinary concentrations of organophosphate ester metabolites: associations with gestational weight gain, early life anthropometry,
and infant eating behaviors among mothers-infant pairs in Rhode Island. Environmental Health: A Global Access Science Source 19(1 ):97.
Reproductive /Developmental-Gestational weight gain among pregnant women; infant gestational age at delivery; infant anthropometric measure-
ments at birth and 6 weeks postpartum, including birth weight and length, head and abdominal circumference, and four body composition (iliac, subscapular,
triceps, and thigh skinfold thickness).
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
7274557
Domain
Metric
Rating
Comments
Metric 8:
Reporting Bias
High All of the measured outcomes were described in detail. Effect measurements with 95%
CI and medians with interquartile ranges were reported. Continuous exposure levels
were used and each analysis was tabulated or graphed for data extraction.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High Covariates including maternal age at delivery, income, pre-pregnancy BMI, parity and
infant sex were appropriately adjusted in the linear regression model and mixed effect
model.
Medium Covariate information was collected through a questionnaire at enrollment and medical
records. Medical records are a well-established and reliable source, while the question-
naire is less-established. There is little to no concern about validity or confounding.
Medium Co-exposure of 2 other OPE metabolites in urine samples were evaluated and adjusted
in this study. Even though the authors mentioned residual confounding by unmeasured
co-exposure may be present, there is no direct evidence that it would introduce signifi-
cant bias to the effect.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High The study design and analytical models used were appropriate to analyze the relation-
ship between exposure and outcomes. Up to 3 urine metabolite measurements were used
to represent the exposure level in the pregnancy window. Linear regression models were
used for the continuous variables.
Low The authors reported that this pilot study with a smaller sample size (56 maternal-infant
pairs) may not have sufficient statistical power to detect some effects. However, the
power was sufficient to detect an association between OPE exposure and infant anthro-
pometry by sex, which was also reported by Hoffman et al 2018.
Medium The analysis methods, model selection, and data processing methods were reported and
sufficient to understand and reproduce.
High Model assumptions were met and the method was transparent. Variable were appropri-
ately transformed.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure Medium BCEP is a metabolite of TCEP and used as a biomarker of TCEP exposure. There might
be other parent compounds but TCEP is one of the most common OPE detected.
Metric 17: Effect Biomarker N/A Not applicable - no biomarker of effect.
Metric 18: Method Sensitivity Medium Analytical methods were fully described and appropriate. The LODs and detection
frequency were reported in the supplemental table SI.
Continued on next page ...
Page 54 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 7274557 Table: 1 of 2
... continued from previous page
Crawford, K. A., Hawley, N., Calafat, A. M., Jayatilaka, N. K., Froehlich, R. J., Has, P., Gallagher, L. G., Savitz, D. A., Braun, J. M., Werner, E. F., Romano,
M. E. (2020). Maternal urinary concentrations of organophosphate ester metabolites: associations with gestational weight gain, early life anthropometry,
and infant eating behaviors among mothers-infant pairs in Rhode Island. Environmental Health: A Global Access Science Source 19(1 ):97.
Reproductive /Developmental-Gestational weight gain among pregnant women; infant gestational age at delivery; infant anthropometric measure-
ments at birth and 6 weeks postpartum, including birth weight and length, head and abdominal circumference, and four body composition (iliac, subscapular,
triceps, and thigh skinfold thickness).
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
7274557
Domain
Metric
Rating
Comments
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High
High
High
N/A
Sample storage and shipping condition was reported and no reported loss.
There is no direct evidence to show the samples had contamination concerns. The ana-
lytical methods were described and the quality assurance used were within the lab limits,
according to the authors.
Target analytes were separated on a ultra-high-performanceliquid chromatography sys-
tem and quantified using mass spectrometry.
The matrix adjustment information is not reported.
Additional Comments: Overall, this is a high-quality pilot study to evaluate the association between gestational exposure to target chemicals and reported health outcomes. The
models and analytical methods applied were clear, fully described and appropriate. Strengths and limitations were discussed and not likely to introduce
significant bias to the study.
Overall Quality Determination High
Page 55 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 7274557 Table: 2 of 2
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Crawford, K. A., Hawley, N., Calafat, A. M., Jayatilaka, N. K., Froehlich, R. J., Has, P., Gallagher, L. G., Savitz, D. A., Braun, J. M., Werner, E. F., Romano,
M. E. (2020). Maternal urinary concentrations of organophosphate ester metabolites: associations with gestational weight gain, early life anthropometry,
and infant eating behaviors among mothers-infant pairs in Rhode Island. Environmental Health: A Global Access Science Source 19(1 ):97.
Neurological/Behavioral-Infant feeding behaviors including general appetite, enjoyment of food, food responsiveness, slowness in eating, and satiety
responsiveness.
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
7274557
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
High Overall, all key elements of study design and participation were reported. In this pilot
study, 62 women were recruited. Inclusion and exclusion criteria as well as participant
selection were fully described.
High A minimal number of subjects (6/62) were excluded from further analysis due to with-
drawal, miscarriage, and loss to follow up. Exposure and outcome measurements were
complete among pregnant women and infants. The exclusion or loss to follow up are not
likely to introduce bias since the participated population is likely to represent the general
eligible population.
High Differences in baseline characteristics were reported and adequately considered and ad-
justed in statistical analysis. Identified effect modification by infant sex was considered
in further analysis. Participants were similar since they were recruited from the same
setting using same inclusion criteria.
Domain 2: Exposure Characterization
Metric 4:
Metric 5:
Metric 6:
Measurement of Exposure
Exposure Levels
Temporality
High Exposure levels of target chemical were derived from metabolite concentrations in urine
samples. Collection and quantification methods were fully described.
Medium Continuous individual exposure levels were measured from OPE metabolite concentra-
tions in pooled urine samples.
Medium The authors reported that OPE metabolites in urine samples showed good reproducibil-
ity and good intraclass correlation. They used pooled urine samples collected at ges-
tation weeks 12, 28 and 35 to represent the exposure window. OPE chemicals have
relatively short half lives and very low bioaccumulation rates, so the OPE chemicals are
commonly used to indicate persistent exposure. It is not clear whether the exposures fell
within relevant exposure windows for the outcomes of interests.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
Medium Infant feeding behaviors were evaluated by Baby Eating Behavior Questionnaire
(BEBQ) completed by mothers. The validation is not reported but there is no direct
evidence that the method has poor validity or significant misclassification.
High All of the measured outcomes were described in detail. Effect measurements with 95%
CI and medians with interquartile range reported. Continuous exposure levels were used
and each analysis was tabulated or graphed for data extraction.
Domain 4: Potential Confounding / Variability Control
Continued on next page ...
Page 56 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 7274557 Table: 2 of 2
... continued from previous page
Crawford, K. A., Hawley, N., Calafat, A. M., Jayatilaka, N. K., Froehlich, R. J., Has, P., Gallagher, L. G., Savitz, D. A., Braun, J. M., Werner, E. F., Romano,
M. E. (2020). Maternal urinary concentrations of organophosphate ester metabolites: associations with gestational weight gain, early life anthropometry,
and infant eating behaviors among mothers-infant pairs in Rhode Island. Environmental Health: A Global Access Science Source 19(1 ):97.
Neurological/Behavioral-Infant feeding behaviors including general appetite, enjoyment of food, food responsiveness, slowness in eating, and satiety
responsiveness.
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
7274557
Domain
Metric
Rating
Comments
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High Covariates including maternal age at delivery, income, pre-pregnancy BMI, parity and
infant sex were appropriately adjusted in the linear regression model and mixed effect
model.
Medium Covariate information was collected through a questionnaire at enrollment and medical
records. Medical records are a well-established and reliable source, while the question-
naire is less-established. There is little to no concern about validity or confounding.
Medium Co-exposure of other 2 OPE metabolites in urine samples were evaluated and adjusted
in this study. Even though the authors mentioned that residual confounding by unmea-
sured co-exposure may be present, there is no direct evidence that would introduce
significant bias to the effect.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High The study design and analytical models used were appropriate to catch the relationship
between exposure and outcomes. Up to 3 urine metabolite measurements were applied
to represent the exposure level during pregnancy. Linear regression models were used
for the continuous variables.
Low The authors reported that this pilot study with a smaller sample size (56 maternal-infant
pairs) may not have sufficient statistical power to detect some effects. However, the
power was sufficient to detect an association between OPE exposure and infant anthro-
pometry by sex, which was also reported by Hoffman et al 2018.
Medium The description of analysis methods, model selection, and data processing methods were
reported and sufficient to understand and reproduce.
High Model assumptions were met and the method was transparent. Variable were appropri-
ately transformed.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric
16:
Use of Biomarker of Exposure
Medium
BCEP is a metabolite of TCEP and used as a biomarker of TCEP exposure. There might
be other parent compounds but TCEP is one of the most common OPE detected.
Metric
17:
Effect Biomarker
N/A
Not applicable - no biomarker of effect.
Metric
18:
Method Sensitivity
Medium
Analytical methods were fully described and appropriate. The LODs and detection
frequency were reported in the supplemental table SI.
Metric
19:
Biomarker Stability
High
Sample storage and shipping condition was reported and no reported loss.
Metric
20:
Sample Contamination
High
There is no direct evidence to show the samples had contamination concerns. The ana-
lytical methods were described and the quality assurance used were within the lab limits,
according to the authors.
Continued on next page ...
Page 57 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 7274557 Table: 2 of 2
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Crawford, K. A., Hawley, N., Calafat, A. M., Jayatilaka, N. K., Froehlich, R. J., Has, P., Gallagher, L. G., Savitz, D. A., Braun, J. M., Werner, E. F., Romano,
M. E. (2020). Maternal urinary concentrations of organophosphate ester metabolites: associations with gestational weight gain, early life anthropometry,
and infant eating behaviors among mothers-infant pairs in Rhode Island. Environmental Health: A Global Access Science Source 19(1 ):97.
Neurological/Behavioral-Infant feeding behaviors including general appetite, enjoyment of food, food responsiveness, slowness in eating, and satiety
responsiveness.
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
7274557
Domain
Metric
Rating
Comments
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High Target analytes were separated on a ultra-high-performanceliquid chromatography sys-
tem and quantified using mass spectrometry.
N/A The matrix adjustment information is not reported.
Additional Comments:
Overall, this is a high-quality pilot study to evaluate the association between gestational exposure to target chemicals and reported health outcomes. Despite
the small pilot-scale sample size, the samples are a good represent of eligible general population. The models and analytical methods applied were clear,
fully described and appropriate. Strengths and limitations were discussed and not likely to introduce significant bias to the study.
Overall Quality Determination
High
Page 58 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364495 Table: 1 of 1
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Mendy, A., Percy, Z., Braun, J. M., Lanphear, B., Guardia, La, M. J., Hale, R. C., Yolton, K., Chen, A. (2024). Prenatal exposure to replacement flame
retardants and organophosphate esters and childhood adverse respiratory outcomes. Environmental Research 240(Pt 2): 117523.
Lung/Respiratory-Wheeze, respiratory infections, hay fever or allergies, forced expiratory volume in Is (FEV1), peak expiratory flow (PEF)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
11364495
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
Medium Participants in this prospective birth cohort study were pulled from the Health Outcomes
and Measures of the Environment (HOME) study. Pregnant women were recruited from
March 2003-January 2006 in the greater Cincinnati, Ohio metropolitan area. Details of
the HOME study are available in Braun et al. 2017 (HEROID: 6749104). In Braun et
al. 2017 inclusion criteria and recruitment are extensively detailed and there is no evi-
dence to suggest selection bias occurred in the original creation of the HOME study401
mother-child pairs were included in the original HOME study. Inclusion criteria for the
analysis in this study were: having data on dust and urinary organophoshate esters and
replacement brominated flame retardants, having data on covariates, and data on respira-
tory outcomes through follow-up until 5 years. 342 mother-child pairs were included in
the present study.In general, there is no significant evidence for selection bias. However,
no data is presented comparing those included in the study to participants who were ex-
cluded after originally participating in the HOME study.
Medium There were n=59 pregnant mother-child pair participants out of 401 who were included
in the original HOME study but not the present analysis; all exclusion reasons were
related to missing data, but exact breakdowns are not provided (ex; how many were
excluded due to missing covariate data vs. missing exposure or outcome data). However,
there is no reason to suggest that exclusion of those participants was inappropriate.
High Participants were all recruited from the same eligible population of pregnant women
in the Greater Cincinnati, Ohio metropolitan area. The study presents a table of de-
mographic characteristics split by exposure to all relevant exposures, including TCEP.
Significant differences were reported by race/ethnicity and serum cotinine (significant
differences were reported additionally for serum cotinine when assessing BCEP). All
these factors were adjusted for in statistical analysis.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure Medium Exposure to TCEP was measured via the metabolite of BCEP. Urine was sampled at 16
and 26 weeks of gestation and within 48 hours of delivery. BCEP was quantified in spot
urine samples using high-performance liquid chromatography mass spectrometry. The
multitude of measurements allows for consistency across samples and a higher certainty
in exposure classification.
Metric 5: Exposure Levels Medium Exposure ranges were likely large enough to allow an exposure-response estimate. For
example, BCEP in urine average (5th-95th percentile): 0.13 - 6.98 ug/L.
Continued on next page ...
Page 59 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364495 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Mendy, A., Percy, Z., Braun, J. M., Lanphear, B., Guardia, La, M. J., Hale, R. C., Yolton, K., Chen, A. (2024). Prenatal exposure to replacement flame
retardants and organophosphate esters and childhood adverse respiratory outcomes. Environmental Research 240(Pt 2): 117523.
Lung/Respiratory-Wheeze, respiratory infections, hay fever or allergies, forced expiratory volume in Is (FEV1), peak expiratory flow (PEF)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
11364495
Domain
Metric
Rating
Comments
Metric 6:
Temporality
High
Exposure was measured at 16 weeks gestation, 26 weeks gestation, and at delivery.
Temporality is thus established, and the introduction highlights the importance of devel-
opmental lung formation in predicting respiratory outcomes.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
High
High
Respiratory outcomes included wheeze, respiratory infections, hay fever or allergies,
forced expiratory volume in Is (FEV1), and peak expiratory flow (PEF). Wheeze, res-
piratory infections, and hay fever or allergies were measured busting standardized ques-
tions, based on NHANES questionnaires. Questionnaires were provided every 6 months
until the age of 5 years old. FEV1 and PEF were measured at age 5 years old using
a portable spirometer, and measures were reported to follow the American Thoracic
Society criteria and conducted by trained research assistants. In general, there is no sig-
nificant evidence of outcome misclassification.
All measured outcomes outlined in the methods are reported in the results in a way that
would allow for detailed extraction (effect estimates and 95% confidence intervals).
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium
High
Medium
Covariates included child sex, child race/ethnicity, median household income at base-
line, duration of breastfeeding, gestational age, maternal and paternal allergy and
asthma, birth weight was retrieved from medical records, and prenatal exposure to
smoking. No information is provided on why these covariates were chosen.
The majority of covariates were measured using questionnaires from the HOME study.
Birth weight was retrieved from medical records, and prenatal smoking was measured
using serum cotinine at ages 16 and 26 weeks of gestation.
The study measured and analyzed other co-exposures, including TCIPP, TDCIPP, TPHP,
EH-TBB, and BEH-TEBP. While these were not adjusted for in analyses of TCEP, cor-
relation coefficients were presented across all exposures.
Domain 5: Analysis
Metric 12:
Metric 13:
Study Design and Methods
Statistical Power
High
Medium
The prospective birth cohort study design is an appropriate design to assess the impact
of prenatal TCEP exposure on respiratory outcomes. The use of generalized estimating
equation analysis is appropriate to account for repeated measures.
The number of participants in the analysis (n=342) is likely large enough to detect an
effect.
The description of the analysis is sufficiently detailed so that the results can be repro-
duced.
Metric 14:
Reproducibility of Analyses
Medium
Continued on next page ...
Page 60 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11364495 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Mendy, A., Percy, Z., Braun, J. M., Lanphear, B., Guardia, La, M. J., Hale, R. C., Yolton, K., Chen, A. (2024). Prenatal exposure to replacement flame
retardants and organophosphate esters and childhood adverse respiratory outcomes. Environmental Research 240(Pt 2): 117523.
Lung/Respiratory-Wheeze, respiratory infections, hay fever or allergies, forced expiratory volume in Is (FEV1), peak expiratory flow (PEF)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
11364495
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High The association between exposure and outcome was primarily measured using gener-
alized estimating equations. The study chose a binomial distribution with a log link
function for the models due to the repeated measures of the binary respiratory outcomes.
Further, "unstructured working matrix and robust variance estimators were specified
to estimate the beta coefficients exponentiated to obtain the relative risks and their cor-
responding 95% confidence intervals." Lung function parameters only had a single
measurement and were thus assessed using linear regression modeling. Sensitivity anal-
yses to assess the potential for effect modification by sex.Concentrations of BCEP were
log-transformed to account for non-normality. BCEP was detected in roughly 86% of
samples across all three measured time periods, and samples below the LOD were re-
placed with LOD/sqrt(2). Analyses were presented separately for the urine sampling
period.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16:
Use of Biomarker of Exposure
High
BCEP is a highly sensitive and specific biomarker of exposure to TCEP.
Metric 17:
Effect Biomarker
N/A
Effect biomarkers were not assessed.
Metric 18:
Method Sensitivity
Medium
The LOD is stated to be 0.10 ug/L, which is likely low enough to detect BCEP in a
sufficient percentage of samples.
Metric 19:
Biomarker Stability
Medium
Urine samples were frozen at -20 degrees Celsius until analysis. No stability data is
reported.
Metric 20:
Sample Contamination
Medium
There is no information provided on sample contamination, but there is no reason to
suspect contamination occurred.
Metric 21:
Method Requirements
High
Samples were analyzed using high performance liquid chromatography with mass spec-
trometry.
Metric 22:
Matrix Adjustment
Medium
Samples were adjusted for urinary dilution by standardizing for specific gravity. Results
are only presented as adjusted.
Additional Comments: This prospective birth cohort study assessed the association between prenatal measurements of TCEP in dust samples and BCEP in maternal urine to
repeated measures of respiratory outcomes among children. In general there is a limited risk of bias in the study due to well-conducted exposure assessment
and outcome assessments. Significant positive associations were reported for TCEP and respiratory outcomes - however, these associations tended to be
inconsistent across methods of exposure measurement (dust concentrations, dust loadings, urinary BCEP).
Overall Quality Determination High
Page 61 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10081087 Table: 1 of 1
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Vuong, A. M., Xu, Y., Xie, C., Ospina, M., Calafat, A. M., Lanphear, B. P., Braun, J. M., Cecil, K. M., Dietrich, K. N., Chen, A., Yolton,
K. (2021). Prenatal exposure to a mixture of organophosphate esters and intelligence among 8-year-old children of the HOME Study. NeuroToxicology
87:149-155.
Neurological/Behavioral-Weschler Intelligence Scale for Children-IV scores (FSIQ, Verbal Comprehension, Perceptual Reasoning, Working Memory,
Processing Speed)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
10081087
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
Medium The HOME study recruited pregnant women from Cincinnati, Ohio from 2003 to 2006.
Inclusion criteria were women 18 years of age or older, at 13-19 weeks of gestation, and
living in a home built before 1978. Women were excluded if they were taking medica-
tion for thyroid disorders or seizures, HIV positive, had bipolar disorder, schizophrenia,
diabetes, active cancer, not fluent in English, or planning to move outside the Cincinnati
area. 5184 women were approached about the study, 1263 met the eligibility criteria and
468 women agreed to participate (37% participation rate). Entry into this specific analy-
sis of HOME study participants also required that women deliver a live, singleton infant,
have at least one urinary OPE measurement during pregnancy, and have a child intel-
ligence measurement at 8 or 12 years of age. In total, 233 participants participated in
the final analysis.There was no comparison between women who joined and those who
did not join the study and limited the methods of study advertisement and recruitment.
However, there is no evidence of selection bias.
Medium The participation rate after follow up at 8-12 years of age was 49.8% (n=233). Sociode-
mographic characteristics of the remaining participants and the original participants
were reported to be similar. There was no evidence that missingness was related to both
exposure and outcome.
Medium Demographic characteristics are not presented in association with exposure or outcome
levels, and thus there is no direct evidence that groups were similar. However, there is no
reason to suspect that groups were wildly different, as key demographic characteristics
were adjusted for in statistical analysis. Recruitment methods and inclusion/exclusion
do not appear to have been performed differently in subsets of participants.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure
High TCEP was measured via the urinary metabolite BCEP in spot urine samples collected at
3 times over the course of pregnancy (around 16 weeks' gestation, 26 weeks' gestation,
and within 48 hours of delivery). BCEP was quantified using high performance liquid
chromatography with tandem mass spectrometry. The use of multiple measurement time
points increases the confidence in reported exposure levels.
Continued on next page ...
Page 62 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10081087 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Vuong, A. M., Xu, Y., Xie, C., Ospina, M., Calafat, A. M., Lanphear, B. P., Braun, J. M., Cecil, K. M., Dietrich, K. N., Chen, A., Yolton,
K. (2021). Prenatal exposure to a mixture of organophosphate esters and intelligence among 8-year-old children of the HOME Study. NeuroToxicology
87:149-155.
Neurological/Behavioral-Weschler Intelligence Scale for Children-IV scores (FSIQ, Verbal Comprehension, Perceptual Reasoning, Working Memory,
Processing Speed)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
10081087
Domain
Metric
Rating
Comments
Metric 5:
Exposure Levels
Metric 6:
Temporality
Medium The median value for samples collected at 16 weeks was 0.76 ng/mL (25th - 75th per-
centile: 0.45 - 1.34 ng/mL). The median value for samples collected at 26 weeks was
0.65 ng/mL (25th - 75th percentile: 0.34 - 1.31 ng/mL). The median value for samples
collected at delivery was 0.62 ng/mL (25th - 75th percentile: 0.35 - 1.27 ng/mL). Gen-
erally, the levels of exposure at all three time points likely have sufficient variation to
detect an effect. Concentrations of BCEP were modeled continuously in statistical mod-
els.
Medium Exposure measures were taken three times before the outcome during pregnancy, so
temporality was established and there is some degree of certainty that exposure levels
were consistent across pregnancy. However, due to the short half-life of BCEP and a
long follow-up time, it's unclear whether the exposure was consistent across the first 8
or 12 years of life, and it is unclear whether or not gestation represents the only etiologi-
cally relevant time period for neurodevelopmental effects.
Domain 3: Outcome Assessment
Metric 7: Outcome Measurement or
Characterization
Metric 8: Reporting Bias
High The outcome was assessed by using Wechsler Intelligence Scale for Children-IV
(WISC-IV) for children at 8 years old. A small number of children (n=12) were assessed
at 12 years old using the same questionnaire. The tests were administered by research
assistants trained by an experienced and reliable gold standard examiner and blinded to
all exposure variables. Training exercises for examiners are described in detail. Examin-
ers were re-assessed every six months to ensure accuracy in assessment.
High All analyses mentioned in the methods are reported in the results. Effect estimates were
reported as regression coefficient with 95% confidence intervals.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High Covariates were identified through a directed acyclic graph and kept minimum for analy-
sis. The covariates were income, maternal BMI, maternal IQ as measured by WASI, and
maternal race.
Medium Covariate information was measured using standardized questionnaire. Maternal intelli-
gence was assessed with WASI and examination methods are extensively detailed.
Medium Co-exposures included other OPEs (BDCIPP, DPHP, DNBP) and 10 PBDE congeners
(BDE-17, -28, -66, -85, -99, -100, -153, -154, -183). Detailed methods are provided for
the quantification of each co-exposure. Exposure to PBDEs was adjusted for in statisti-
cal analyses, while co-exposure to other OPEs was addressed via BKMR modeling.
Domain 5: Analysis
Continued on next page ...
Page 63 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10081087 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Vuong, A. M., Xu, Y., Xie, C., Ospina, M., Calafat, A. M., Lanphear, B. P., Braun, J. M., Cecil, K. M., Dietrich, K. N., Chen, A., Yolton,
K. (2021). Prenatal exposure to a mixture of organophosphate esters and intelligence among 8-year-old children of the HOME Study. NeuroToxicology
87:149-155.
Neurological/Behavioral-Weschler Intelligence Scale for Children-IV scores (FSIQ, Verbal Comprehension, Perceptual Reasoning, Working Memory,
Processing Speed)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
10081087
Domain
Metric
Rating
Comments
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High
Medium
Medium
High
The study design was a prospective birth cohort with a long follow-up time to evaluate
the association between maternal urinary BCEP exposure and child intelligence. The use
of repeated measures analysis is also appropriate to assess changes in exposure levels of
BCEP across pregnancy.
The sample size (N=233 mother-child dyads) is likely large enough to detect an effect.
The methods are described in sufficient detail so that they would able to be reproduced
given access to the analytic data.
The association between prenatal concentration of BCEP and child intelligence was
assessed using generalized estimating equations to account for the three different mea-
surements during the pregnancy. BCEP concentrations were ln-transformed to normalize
the distribution. As BCEP had more than 10% of samples with concentrations below the
limit of detection, a multiple imputation approach was used to impute left-censored data
via Markov Chain Monte Carlo algorithm to yield 10 imputed data sets. Each data set
was used for modeling and the results were pooled using Rubin's rules. The results were
presented as regression coefficient with 95% confidence interval.Sensitivity analysis was
conducted to include urinary specific gravity as a covariate and to standardize OPE con-
centrations by specific gravity.In general, the statistical analysis method was appropriate,
and model assumptions were met.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric
16:
Use of Biomarker of Exposure
High
The biomarker BCEP is derived from the parent compound TCEP and is appropriate to
use as a proxy for exposure to TCEP.
Metric
17:
Effect Biomarker
N/A
0
Metric
18:
Method Sensitivity
Medium
The limit of detection was 0.1 ng/mL and is likely low enough to detect chemicals in a
sufficient percentage of samples.
Metric
19:
Biomarker Stability
Medium
Samples were collected and kept frozen at -20 degrees Celsius for storage. No stability
data was reported but there was no evidence of sample loss
Metric
20:
Sample Contamination
Medium
There is no specific information about contamination, but quality control procedures are
detailed.
Metric
21:
Method Requirements
High
BCEP was quantified via high-performance liquid chromatography with isotope
dilution-electrospray ionization tandem mass spectrometry.
Metric
22:
Matrix Adjustment
Medium
BCEP concentrations were only presented as adjusted for specific gravity.
Additional Comments: The HOME prospective birth cohort had a good design to evaluate any associations between prenatal exposure to BCEP and child intelligence at ages 8 or
12. There were no significant sources of potential bias, and the exposure assessment was well-conducted with urine sampled three times during pregnancy
to ensure consistent exposure during the gestational period. BCEP was reported to be associated with a slight increase in child full-scale IQ.
Continued on next page ...
Page 64 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10081087 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Vuong, A. M., Xu, Y., Xie, C., Ospina, M., Calafat, A. M., Lanphear, B. P., Braun, J. M., Cecil, K. M., Dietrich, K. N., Chen, A., Yolton,
K. (2021). Prenatal exposure to a mixture of organophosphate esters and intelligence among 8-year-old children of the HOME Study. NeuroToxicology
87:149-155.
Neurological/Behavioral-Weschler Intelligence Scale for Children-IV scores (FSIQ, Verbal Comprehension, Perceptual Reasoning, Working Memory,
Processing Speed)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis-2-chloroethyl phosphate (BCEP)
No linked references.
10081087
Domain Metric
Rating
Comments
Overall Quality Determination
High
Page 65 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10078361 Table: 1 of 3
Study Citation: Kang, H., Lee, J., Lee, J. P., Choi, K. (2019). Urinary metabolites of organophosphate esters (OPEs) are associated with chronic kidney disease in the
general US population, NHANES 2013-2014. Environment International 131:5034-5034.
Health Renal/Kidney-estimated Glomerular Filtration Rate (eGFR)
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: BCEP
No linked references.
10078361
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
High Study used data from NHANES 2013-2014. Cohort is designed to be representative
sample of general US population. Analyzed random selection of approximately 1/3 of
participants with stored spot urine samples. Inclusion/exclusion criteria were appropriate
for the outcome assessment: excluded pregnant women, those with missing eGFR, ACR
or BMI data. Low risk of selection bias.
High Of the 2666 stored urine samples chosen to measure OPE metabolites, 1578 were used
in the analysis for this study. The exclusion of participants was appropriately addressed
based on age, current pregnancy (1,660). BMI, ACR, or eGFR data was missing for only
82 (and subsequently excluded). Exposure and outcome data largely complete.
High Baseline characteristics of groups assessed and adjusted for in statistical modelling, in-
cluding sex, age, race/ethnicity, BMI, smoking status, poverty income ratio, physical and
current CKD. Large established cohort with documented setting, methods of selection.
(Documented previously) Appropriate inclusion/exclusion criteria for study reported.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6: Temporality
High The investigators used an established method for analyzing the exposure (biomarker):
the analytes were separated using reverse-phase high performance liquid chromatogra-
phy. The analytes were measured using isotope dilution-electroscopy ionization tandem
mass spectrometry. The study also measured urine creatinine to account for urinary dilu-
tion and concentration of the biomarkers.
Medium The authors reported LOD for the analytes which were detected in >75% of samples,
and also the detection frequency, geometric mean and 25, 50 and 75 percentile. The
distributions of metabolites were reported for unadjusted, adjusted using a traditional
creatinine adjustment and using a novel creatinine adjustment.
Low Cross-sectional study precludes temporal causality. In addition, this study utilized spot
urine sample. The metabolites have short half-life and thus it is difficult to assess long-
term exposure.
Domain 3: Outcome Assessment
Metric 7: Outcome Measurement or
Characterization
High This study used eGFR and ACR parameters to classify participants as chronic kidney
disease (CKD) patients; the parameters previously established using an extended defini-
tion of CKD. eGFR was calculated using an appropriate formula. The ACR was calcu-
lated as the urinary albumin/creatinine ratio.
Continued on next page ...
Page 66 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10078361 Table: 1 of 3
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Kang, H., Lee, J., Lee, J. P., Choi, K. (2019). Urinary metabolites of organophosphate esters (OPEs) are associated with chronic kidney disease in the
general US population, NHANES 2013-2014. Environment International 131:5034-5034.
Renal/Kidney-estimated Glomerular Filtration Rate (eGFR)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: BCEP
No linked references.
10078361
Domain
Metric
Rating
Comments
Metric 8:
Reporting Bias
High
The authors report all of the study's measured outcomes as described in the methods. N
is reported for each analysis, and 95% CI reported for associations.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High This study accounted for several potential confounders and the analyses adjusted for:
sex, age, race/ethnicity, BMI, poverty income ratio, smoking history, and physical activ-
ity.
Medium NHANES uses a validated survey instrument for obtaining potential confounder data.
The urinary creatinine (and thus urine dilution) was determined using established lab-
oratory methods. Use of creatine to establish urine dilution does have limitations and
warrants further examination. This study used two different methods, one traditional and
another "novel" approach. These different methods revealed changes in the association
between kidney function and metabolite concentration. The authors state "Since kidney
function can directly influence urine hydration or creatinine excretion, however, the use
of creatinine-adjustment for urine dilution may induce a collider stratification biaslead-
ing to possible confounding".
Medium Correlation co-efficients were calculated for the OPE metabolites. Co-exposures to pol-
lutants (biomarkers -metabolites) were assessed in the analysis using a multi-pollutant
model.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
Low Cross-sectional design used spot urine levels to establish association with exposure
and a chronic disease outcome. Limited simply by the design. However, the statistical
methods used for this study were appropriate for the data.
Medium The sample size (n = 1578) was adequate to detect an effect, power of study not re-
ported.
Medium The authors clearly describe the analytic methods and models used for the analyses.
They could be reproduced.
High The models and methods used to analyze the associations between exposure and out-
comes were appropriate, and the inclusion/exclusion of variables was clearly stated.
Model assumptions were met.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure Low The authors note that the metabolism of OPE is not completely understood. OPE is me-
tabolized into diester metabolites and excreted though urine; this is what was measured
in this study. No discussion of validation or assessment of accuracy and precision.
Continued on next page ...
Page 67 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10078361 Table: 1 of 3
... continued from previous page
Study Citation: Kang, H., Lee, J., Lee, J. P., Choi, K. (2019). Urinary metabolites of organophosphate esters (OPEs) are associated with chronic kidney disease in the
general US population, NHANES 2013-2014. Environment International 131:5034-5034.
Health Renal/Kidney-estimated Glomerular Filtration Rate (eGFR)
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: BCEP
No linked references.
10078361
Domain
Metric
Rating
Comments
Metric 17: Effect Biomarker
Metric 18: Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
Low
Medium
Low
Low
High
High
Little information available on OPE toxic effect on kidney outcomes. There is some
evidence of potential health effects on humans and animals (disruption of thyroid and
sex hormones, development and behavior, and allergic response) but mechanism of
action is not well understood.
The LOD are low enough to detect the metabolites in a sufficient percentage of samples.
The analytic methods are described in detail and use established methods. The LOD and
% are reported.
Storage history and stability data for analytes is not reported. OPE metabolites have a
short half life, but this would be reflected in the cross section of samples taken and not a
reflection of the stability in properly stored samples.
There is no information regarding storage or potential contamination and subsequent
steps to correct issues. However, there is no reason to suspect there are any issues with
this large and well-established cohort's sampling methods.
This study used LC-MS/MS for mesuring the metabolites.
This study reported both adjusted and unadjusted matrix concentrations (used 2 different
methods of urinary creatinine adjustment) and this is supported and discussed at length
due to the importance of urine dilution in studies of kidney function.
Additional Comments: This is a large cross sectional study using NHANES 2013-2014 data to examine the associations between urinary OPE's and CKD. This study used
appropriate participant selection, analytic methods and examined important potential confounders. However, the cross-sectional design used spot urine
levels to establish association with exposure and a chronic disease outcome. There is limited ability to determine causal effect of exposure on the outcome.
The mechanisms of OPE and metabolites' effect on CKD need further investigating.
Overall Quality Determination Medium
Page 68 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10078361 Table: 2 of 3
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Kang, H., Lee, J., Lee, J. P., Choi, K. (2019). Urinary metabolites of organophosphate esters (OPEs) are associated with chronic kidney disease in the
general US population, NHANES 2013-2014. Environment International 131:5034-5034.
Renal/Kidney-estimated Glomerular Filtration Rate (eGFR)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: BCEP
No linked references.
10078361
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
High Study used data from NHANES 2013-2014. Cohort is designed to be representative
sample of general US population. Analyzed random selection of approximately 1/3 of
participants with stored spot urine samples. Inclusion/exclusion criteria were appropriate
for the outcome assessment: excluded pregnant women, those with missing eGFR, ACR
or BMI data. Low risk of selection bias.
High Of the 2666 stored urine samples chosen to measure OPE metabolites, 1578 were used
in the analysis for this study. The exclusion of participants was appropriately addressed
based on age, current pregnancy (1,660). BMI, ACR, or eGFR data was missing for only
82 (and subsequently excluded). Exposure and outcome data largely complete.
High Baseline characteristics of groups assessed and adjusted for in statistical modelling, in-
cluding sex, age, race/ethnicity, BMI, smoking status, poverty income ratio, physical and
current CKD. Large established cohort with documented setting, methods of selection.
(Documented previously) Appropriate inclusion/exclusion criteria for study reported.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6: Temporality
High The investigators used an established method for analyzing the exposure (biomarker):
the analytes were separated using reverse-phase high performance liquid chromatogra-
phy. The analytes were measured using isotope dilution-electroscopy ionization tandem
mass spectrometry. The study also measured urine creatinine to account for urinary dilu-
tion and concentration of the biomarkers.
Medium The authors reported LOD for the analytes which were detected in >75% of samples,
and also the detection frequency, geometric mean and 25, 50 and 75 percentile. The
distributions of metabolites were reported for unadjusted, adjusted using a traditional
creatinine adjustment and using a novel creatinine adjustment.
Low Cross-sectional study precludes temporal causality. In addition, this study utilized spot
urine sample. The metabolites have short half-life and thus it is difficult to assess long-
term exposure.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
High This study used eGFR and ACR parameters to classify participants as chronic kidney
disease (CKD) patients; the parameters previously established using an extended defini-
tion of CKD. eGFR was calculated using an appropriate formula. The ACR was calcu-
lated as the urinary albumin/creatinine ratio.
High The authors report all of the study's measured outcomes as described in the methods. N
is reported for each analysis, and 95% CI reported for associations.
Continued on next page ...
Page 69 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10078361 Table: 2 of 3
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Kang, H., Lee, J., Lee, J. P., Choi, K. (2019). Urinary metabolites of organophosphate esters (OPEs) are associated with chronic kidney disease in the
general US population, NHANES 2013-2014. Environment International 131:5034-5034.
Renal/Kidney-estimated Glomerular Filtration Rate (eGFR)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: BCEP
No linked references.
10078361
Domain
Metric
Rating
Comments
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High This study accounted for several potential confounders and the analyses adjusted for:
sex, age, race/ethnicity, BMI, poverty income ratio, smoking history, and physical activ-
ity.
Medium NHANES uses a validated survey instrument for obtaining potential confounder data.
The urinary creatinine (and thus urine dilution) was determined using established lab-
oratory methods. Use of creatine to establish urine dilution does have limitations and
warrants further examination. This study used two different methods, one traditional and
another "novel" approach. These different methods revealed changes in the association
between kidney function and metabolite concentration. The authors state "Since kidney
function can directly influence urine hydration or creatinine excretion, however, the use
of creatinine-adjustment for urine dilution may induce a collider stratification biaslead-
ing to possible confounding".
Medium Correlation co-efficients were calculated for the OPE metabolites. Co-exposures to pol-
lutants (biomarkers -metabolites) were assessed in the analysis using a multi-pollutant
model.
Domain 5: Analysis
Metric 12:
Study Design and Methods
Low
Cross-sectional design used spot urine levels to establish association with exposure
and a chronic disease outcome. Limited simply by the design. However, the statistical
methods used for this study were appropriate for the data.
Metric 13:
Statistical Power
Medium
The sample size (n = 1578) was adequate to detect an effect, power of study not re-
ported.
Metric 14:
Reproducibility of Analyses
Medium
The authors clearly describe the analytic methods and models used for the analyses.
They could be reproduced.
Metric 15:
Statistical Analysis
High
The models and methods used to analyze the associations between exposure and out-
comes were appropriate, and the inclusion/exclusion of variables was clearly stated.
Model assumptions were met.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16:
Use of Biomarker of Exposure
Low
The authors note that the metabolism of OPE is not completely understood. OPE is me-
tabolized into diester metabolites and excreted though urine; this is what was measured
in this study. No discussion of validation or assessment of accuracy and precision.
Metric 17:
Effect Biomarker
Low
Little information available on OPE toxic effect on kidney outcomes. There is some
evidence of potential health effects on humans and animals (disruption of thyroid and
sex hormones, development and behavior, and allergic response) but mechanism of
action is not well understood.
Continued on next
page...
Page 70 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10078361 Table: 2 of 3
... continued from previous page
Study Citation: Kang, H., Lee, J., Lee, J. P., Choi, K. (2019). Urinary metabolites of organophosphate esters (OPEs) are associated with chronic kidney disease in the
general US population, NHANES 2013-2014. Environment International 131:5034-5034.
Health Renal/Kidney-estimated Glomerular Filtration Rate (eGFR)
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: BCEP
No linked references.
10078361
Domain
Metric
Rating
Comments
Metric 18: Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
Medium
Low
Low
High
High
The LOD are low enough to detect the metabolites in a sufficient percentage of samples.
The analytic methods are described in detail and use established methods. The LOD and
% are reported.
Storage history and stability data for analytes is not reported. OPE metabolites have a
short half life, but this would be reflected in the cross section of samples taken and not a
reflection of the stability in properly stored samples.
There is no information regarding storage or potential contamination and subsequent
steps to correct issues. However, there is no reason to suspect there are any issues with
this large and well-established cohort's sampling methods.
This study used LC-MS/MS for mesuring the metabolites.
This study reported both adjusted and unadjusted matrix concentrations (used 2 different
methods of urinary creatinine adjustment) and this is supported and discussed at length
due to the importance of urine dilution in studies of kidney function.
Additional Comments: This is a large cross sectional study using NHANES 2013-2014 data to examine the associations between urinary OPE's and CKD. This study used
appropriate participant selection, analytic methods and examined important potential confounders. However, the cross-sectional design used spot urine
levels to establish association with exposure and a chronic disease outcome. There is limited ability to determine causal effect of exposure on the outcome.
The mechanisms of OPE and metabolites' effect on CKD need further investigating.
Overall Quality Determination Medium
Page 71 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10078361 Table: 3 of 3
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Kang, H., Lee, J., Lee, J. P., Choi, K. (2019). Urinary metabolites of organophosphate esters (OPEs) are associated with chronic kidney disease in the
general US population, NHANES 2013-2014. Environment International 131:5034-5034.
Renal/Kidney-estimated Glomerular Filtration Rate (eGFR)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: BCEP
No linked references.
10078361
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
High Study used data from NHANES 2013-2014. Cohort is designed to be representative
sample of general US population. Analyzed random selection of approximately 1/3 of
participants with stored spot urine samples. Inclusion/exclusion criteria were appropriate
for the outcome assessment: excluded pregnant women, those with missing eGFR, ACR
or BMI data. Low risk of selection bias.
High Of the 2666 stored urine samples chosen to measure OPE metabolites, 1578 were used
in the analysis for this study. The exclusion of participants was appropriately addressed
based on age, current pregnancy (1,660). BMI, ACR, or eGFR data was missing for only
82 (and subsequently excluded). Exposure and outcome data largely complete.
High Baseline characteristics of groups assessed and adjusted for in statistical modelling, in-
cluding sex, age, race/ethnicity, BMI, smoking status, poverty income ratio, physical and
current CKD. Large established cohort with documented setting, methods of selection.
(Documented previously) Appropriate inclusion/exclusion criteria for study reported.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6: Temporality
High The investigators used an established method for analyzing the exposure (biomarker):
the analytes were separated using reverse-phase high performance liquid chromatogra-
phy. The analytes were measured using isotope dilution-electroscopy ionization tandem
mass spectrometry. The study also measured urine creatinine to account for urinary dilu-
tion and concentration of the biomarkers.
Medium The authors reported LOD for the analytes which were detected in >75% of samples,
and also the detection frequency, geometric mean and 25, 50 and 75 percentile. The
distributions of metabolites were reported for unadjusted, adjusted using a traditional
creatinine adjustment and using a novel creatinine adjustment.
Low Cross-sectional study precludes temporal causality. In addition, this study utilized spot
urine sample. The metabolites have short half-life and thus it is difficult to assess long-
term exposure.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
High This study used eGFR and ACR parameters to classify participants as chronic kidney
disease (CKD) patients; the parameters previously established using an extended defini-
tion of CKD. eGFR was calculated using an appropriate formula. The ACR was calcu-
lated as the urinary albumin/creatinine ratio.
High The authors report all of the study's measured outcomes as described in the methods. N
is reported for each analysis, and 95% CI reported for associations.
Continued on next page ...
Page 72 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10078361 Table: 3 of 3
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Kang, H., Lee, J., Lee, J. P., Choi, K. (2019). Urinary metabolites of organophosphate esters (OPEs) are associated with chronic kidney disease in the
general US population, NHANES 2013-2014. Environment International 131:5034-5034.
Renal/Kidney-estimated Glomerular Filtration Rate (eGFR)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: BCEP
No linked references.
10078361
Domain
Metric
Rating
Comments
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High This study accounted for several potential confounders and the analyses adjusted for:
sex, age, race/ethnicity, BMI, poverty income ratio, smoking history, and physical activ-
ity.
Medium NHANES uses a validated survey instrument for obtaining potential confounder data.
The urinary creatinine (and thus urine dilution) was determined using established lab-
oratory methods. Use of creatine to establish urine dilution does have limitations and
warrants further examination. This study used two different methods, one traditional and
another "novel" approach. These different methods revealed changes in the association
between kidney function and metabolite concentration. The authors state "Since kidney
function can directly influence urine hydration or creatinine excretion, however, the use
of creatinine-adjustment for urine dilution may induce a collider stratification biaslead-
ing to possible confounding".
Medium Correlation co-efficients were calculated for the OPE metabolites. Co-exposures to pol-
lutants (biomarkers -metabolites) were assessed in the analysis using a multi-pollutant
model.
Domain 5: Analysis
Metric 12:
Study Design and Methods
Low
Cross-sectional design used spot urine levels to establish association with exposure
and a chronic disease outcome. Limited simply by the design. However, the statistical
methods used for this study were appropriate for the data.
Metric 13:
Statistical Power
Medium
The sample size (n = 1578) was adequate to detect an effect, power of study not re-
ported.
Metric 14:
Reproducibility of Analyses
Medium
The authors clearly describe the analytic methods and models used for the analyses.
They could be reproduced.
Metric 15:
Statistical Analysis
High
The models and methods used to analyze the associations between exposure and out-
comes were appropriate, and the inclusion/exclusion of variables was clearly stated.
Model assumptions were met.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16:
Use of Biomarker of Exposure
Low
The authors note that the metabolism of OPE is not completely understood. OPE is me-
tabolized into diester metabolites and excreted though urine; this is what was measured
in this study. No discussion of validation or assessment of accuracy and precision.
Metric 17:
Effect Biomarker
Low
Little information available on OPE toxic effect on kidney outcomes. There is some
evidence of potential health effects on humans and animals (disruption of thyroid and
sex hormones, development and behavior, and allergic response) but mechanism of
action is not well understood.
Continued on next
page...
Page 73 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 10078361 Table: 3 of 3
... continued from previous page
Study Citation: Kang, H., Lee, J., Lee, J. P., Choi, K. (2019). Urinary metabolites of organophosphate esters (OPEs) are associated with chronic kidney disease in the
general US population, NHANES 2013-2014. Environment International 131:5034-5034.
Health Renal/Kidney-estimated Glomerular Filtration Rate (eGFR)
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: BCEP
No linked references.
10078361
Domain
Metric
Rating
Comments
Metric 18: Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
Medium
Low
Low
High
High
The LOD are low enough to detect the metabolites in a sufficient percentage of samples.
The analytic methods are described in detail and use established methods. The LOD and
% are reported.
Storage history and stability data for analytes is not reported. OPE metabolites have a
short half life, but this would be reflected in the cross section of samples taken and not a
reflection of the stability in properly stored samples.
There is no information regarding storage or potential contamination and subsequent
steps to correct issues. However, there is no reason to suspect there are any issues with
this large and well-established cohort's sampling methods.
This study used LC-MS/MS for mesuring the metabolites.
This study reported both adjusted and unadjusted matrix concentrations (used 2 different
methods of urinary creatinine adjustment) and this is supported and discussed at length
due to the importance of urine dilution in studies of kidney function.
Additional Comments: This is a large cross sectional study using NHANES 2013-2014 data to examine the associations between urinary OPE's and CKD. This study used
appropriate participant selection, analytic methods and examined important potential confounders. However, the cross-sectional design used spot urine
levels to establish association with exposure and a chronic disease outcome. There is limited ability to determine causal effect of exposure on the outcome.
The mechanisms of OPE and metabolites' effect on CKD need further investigating.
Overall Quality Determination Medium
Page 74 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577122 Table: 1 of 1
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Hernandez-Castro, I., Eckel, S. P., Howe, C. G., Niu, Z. Z., Kannan, K., Robinson, M., Foley, H. B., Grubbs, B., Al-Marayati, L., Lerner, D., Lurvey,
N., Aung, M. T., Habre, R., Dunton, G. F., Farzan, S. F., Breton, C. V., Bastain, T. M. (2023). Sex-specific effects of prenatal organophosphate ester
(OPE) metabolite mixtures and adverse infant birth outcomes in the maternal and developmental risks from environmental and social stressors (MADRES)
pregnancy cohort. Environmental Research 226:115703.
Reproductive /Developmental-Gestational age at birth, gestational-age adjusted birth weight z-scores
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11577122
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2:
Attrition
Metric 3: Comparison Group
Medium This study used data from 421 mother-infant pairs in the ongoing MADRES (Maternal
And Developmental Risks from Environmental and Social Stressors) cohort to analyze
associations between maternal urinary concentrations of the TCEP metabolite bis(2-
chloroethyl) phosphate (BCEP) and birth outcomes. This cohort has recruited predom-
inately low-income Hispanic/Latino mother-child pairs from Los Angeles; participants
in this study were recruited from 2015 and 2019. Eligible women were aged 18 years or
older, had no significant disabilities or HIV, were recruited prior to 30 weeks' gestation,
and had singleton pregnancies. Participation rates for the cohort were not described.
However, there was no evidence to suggest that participation would have been associated
with either TCEP exposure or birth outcomes. A subset of 421 participants with OPE
metabolite concentrations, birth outcome data, and information on key covariates was
created for this analysis and the study reported that this subset was similar to the full
cohort on "key demographic characteristics" (supporting data shown in related HEROID
11581666, parent cohort N=774).
Medium The analysis sample included MADRES participants with available urinary organophos-
phate ester (OPE) metabolites. Urinary OPEs, including BCEP, were measured in
all available third trimester spot urine samples collected from women recruited from
November 2015 to October 2019. Of 426 participants, 5 were excluded due to missing
information on key covariates. The authors stated that the subset of 421 was similar to
the parent cohort (supporting data shown in related HEROID 11581666, parent cohort
N=774). Overall, the analysis sample included more than half of the parent cohort, with
minimal attrition among those with available exposure data. There was no evidence of
any bias.
Medium Analyses compared birth outcome measures among participants with lower vs. higher
BCEP in maternal urine during gestation. While descriptive data stratified by BCEP
were not presented, there was no evidence to suggest the comparison group with lower
BCEP differed from those with higher concentrations, and many potential demographic
differences (ex: age, income, education) were considered as potential covariates.
Domain 2: Exposure Characterization
Continued on next page .
Page 75 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577122 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Hernandez-Castro, I., Eckel, S. P., Howe, C. G., Niu, Z. Z., Kannan, K., Robinson, M., Foley, H. B., Grubbs, B., Al-Marayati, L., Lerner, D., Lurvey,
N., Aung, M. T., Habre, R., Dunton, G. F., Farzan, S. F., Breton, C. V., Bastain, T. M. (2023). Sex-specific effects of prenatal organophosphate ester
(OPE) metabolite mixtures and adverse infant birth outcomes in the maternal and developmental risks from environmental and social stressors (MADRES)
pregnancy cohort. Environmental Research 226:115703.
Reproductive /Developmental-Gestational age at birth, gestational-age adjusted birth weight z-scores
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11577122
Domain
Metric
Rating
Comments
Metric 4:
Measurement of Exposure
Medium
Prenatal TCEP exposure was estimated using BCEP concentrations from a single mater-
nal urine sample collected at a mean of 31.5 weeks gestation, adjusted for dilution using
specific gravity. BCEP was detected in 68.4% of samples. Values below the LOD of
0.02 ng/mL were imputed as LOD divided by the square root of 2. Use of a biomarker
is a strength. A potential limitation is the use of a single spot urine sample, given that
within-person variability in BCEP throughout pregnancy is unknown. Other studies ex-
amining TCEP biomarkers reported detecting unmetabolized TCEP in 37% of samples
(e.g., Hou et al. 2020 HEROID 10143372).; excluding this factor may also introduce
measurement error. Though some measurement error is likely, there was no evidence of
important error or bias in exposure estimates.
Metric 5:
Exposure Levels
Medium
BCEP was analyzed as a continuous variable. The distribution indicated adequate vari-
ability, with a median (IQR) of 0.53 (0.03, 1.62) ng/mL.
Metric 6:
Temporality
Medium
BCEP concentrations were measured at a mean (SD) of 31.5 (2) weeks' gestation, pre-
ceding child behavior measures characterized at age 36 months. BCEP concentrations
were not measured in early or mid-pregnancy. However, there was no evidence of spe-
cific vulnerable windows or other concerns for temporality. It is uncertain whether there
may be potentially sensitive windows for adverse effects of TCEP on birth weight or
gestational age
Domain 3: Outcome Assessment
Metric 7:
Outcome Measurement or
Characterization
Medium
The outcomes analyzed in this study were gestational age at birth, birth weight, and
gestational-age adjusted birth weight z-scores. Gestational age was estimated using the
following hierarchy: first trimester ultrasound (59.6%), second trimester ultrasound,
and physician clinical estimate in medical records; the estimate for one subject used
last menstrual period. 79% of participants had multiple gestational age estimates, with
Spearman correlations reported as ranging from 0.52 to 0.91. Correlations of 0.52 sug-
gest important variability in estimates. Though there was no further information on the
extent of variation in gestational age estimates, more accurate estimates were prioritized
in the selection hierarchy. 9.3% of participants were classified as preterm (<37 weeks'
gestation). Birth weight was extracted from medical records and used to calculate sex-
specific birthweight for gestational age z-scores using a nationally representative U.S.
sample. Because unadjusted birth weight was not analyzed, any measurement error in
gestational age is integrated. Despite concerns, there was no evidence of important er-
rors in outcome characterization.
Metric 8:
Reporting Bias
High
Results were presented or described for all analyses included as aims.
Continued on next page ...
Page 76 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577122 Table: 1 of 1
... continued from previous page
Hernandez-Castro, I., Eckel, S. P., Howe, C. G., Niu, Z. Z., Kannan, K., Robinson, M., Foley, H. B., Grubbs, B., Al-Marayati, L., Lerner, D., Lurvey,
N., Aung, M. T., Habre, R., Dunton, G. F., Farzan, S. F., Breton, C. V., Bastain, T. M. (2023). Sex-specific effects of prenatal organophosphate ester
(OPE) metabolite mixtures and adverse infant birth outcomes in the maternal and developmental risks from environmental and social stressors (MADRES)
pregnancy cohort. Environmental Research 226:115703.
Reproductive /Developmental-Gestational age at birth, gestational-age adjusted birth weight z-scores
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11577122
Domain
Metric
Rating
Comments
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10:
Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium Covariates were selected based on prior literature and using a directed acyclic graph
(DAG). All multivariate models adjusted for recruitment site, season of sample col-
lection, gestational age at sample collection, maternal age, pre-pregnancy BMI, parity,
race/ethnicity, maternal hypertensive disorders of pregnancy, income, and education).
Infant sex was included except for models stratified by sex or analyzing sex-specific
birthweight for gestational age; smoking during pregnancy was addressed by exclusion
in sensitivity analysis due to its low prevalence. Gestational diabetes was also included
as a covariate in sensitivity analyses. Although all analyses of birth weight included
gestational age, which is a potential intermediate, the finding of no association between
TCEP and gestational age mitigates concern of important bias. Gestational hypertension
is also not established as a potential intermediate. While there are potential concerns
related to overadjustment, there was no evidence of important bias or of residual con-
founding.
Medium Covariate data came from interviewer-administered questionnaire or medical records.
Validation was not discussed explicitly, but there was no evidence to suggest important
error or bias.
Medium Correlations between BCEP and other OPEs (DPHP, DNBP, DIBP, BDCIPP, BBOEP,
BCIPP, BMPP, BEHP, DPRP) measured in this study were weak (Spearman's r <0.25).
Pair-wise interactions among OPE metabolites were examined. Co-exposure to other
OPEs during pregnancy was also evaluated using Bayesian methods to analyze chemical
mixtures. Though confounding by other exposures cannot be ruled out there was no
evidence indicative of important co-exposure confounding.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
High A prospective cohort design was used to analyze the relationship between prenatal
TCEP exposure and birth outcomes, with exposure measured in the third trimester. This
design was appropriate. Linear regression was appropriately used for analysis.
Medium The analysis sample included 421 participants with a mean (SD) gestational age of
39.1 (1.5) weeks, birthweight of 3303 (475) grams, and birthweight for gestational age
z-score of -0.1 (1.0). There was variability in exposure. There was no indication of
inadequate statistical power.
Medium The description of the analysis is sufficient to understand precisely what was done and
to replicate findings.
Continued on next page ...
Page 77 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577122 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Hernandez-Castro, I., Eckel, S. P., Howe, C. G., Niu, Z. Z., Kannan, K., Robinson, M., Foley, H. B., Grubbs, B., Al-Marayati, L., Lerner, D., Lurvey,
N., Aung, M. T., Habre, R., Dunton, G. F., Farzan, S. F., Breton, C. V., Bastain, T. M. (2023). Sex-specific effects of prenatal organophosphate ester
(OPE) metabolite mixtures and adverse infant birth outcomes in the maternal and developmental risks from environmental and social stressors (MADRES)
pregnancy cohort. Environmental Research 226:115703.
Reproductive /Developmental-Gestational age at birth, gestational-age adjusted birth weight z-scores
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11577122
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Descriptive data were presented for the sample as a whole. Urinary BCEP was described
using percentiles, the geometric means, and the minimum and maximum. To satisfy
model assumptions and evaluate potentially non-linear dose-response, primary mod-
els for individual exposures analyzed associations with natural-log transformed CBCL
scores using either BCEP tertiles or using generalized additive models. Both crude
and adjusted effect estimates were presented with 95% confidence intervals. Interac-
tion terms and stratified models were used to assess effect modification by child sex. In
addition, Bayesian kernel machine regression (BKMR) models were used to analyze
OPE mixtures, and Bayesian semiparametric regression models were used to analyze
synergistic or antagonistic interactions among pairs of chemicals. Both sets of models
allowed for non-linear dose-response patterns. Results were presented graphically and
using posterior inclusion probabilities with 95% credible intervals. Sensitivity anal-
yses to assess robustness included excluding children whose mothers smoked during
pregnancy, analyzing CBCL t-scores vs raw scores, and adjusting BKMR smoothness
parameters. There was no evidence of deficiencies in the data analysis methods used.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Medium
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
N/A
Medium
Medium
Medium
Exposure was estimated using urinary concentrations of the major TCEP metabolite
BCEP measured in a spot urine sample from the third trimester of pregnancy. Unme-
tabolized TCEP, detected in 37% of urine samples in another study (Hou et al. 2020
HEROID 10143372), was not measured. BCEP concentrations in a single spot urine
sample characterize TCEP exposure in pregnancy with some measurement error, but
there was no evidence of bias.
No biomarkers of effect were assessed.
The LOD was 0.02 ng/mL; 68.41% of samples had concentrations above detection lim-
its.
Aliquoted samples were stored at -80 degrees Celsius prior to shipment. Storage dura-
tion was not reported, but there was no evidence of degradation or inappropriate han-
dling. Quality control methods are detailed.
The authors reported that "trace" levels (not quantified) of all OPE diester metabolites
were found in procedural blanks. The trace concentrations measured in blanks were sub-
tracted from sample values for each batch. Further details were not provided. Despite
this evidence of some contamination, there was nothing to suggest important error, or
that the method used to address this issue was biased.
Continued on next page ...
Page 78 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577122 Table: 1 of 1
... continued from previous page
Hernandez-Castro, I., Eckel, S. P., Howe, C. G., Niu, Z. Z., Kannan, K., Robinson, M., Foley, H. B., Grubbs, B., Al-Marayati, L., Lerner, D., Lurvey,
N., Aung, M. T., Habre, R., Dunton, G. F., Farzan, S. F., Breton, C. V., Bastain, T. M. (2023). Sex-specific effects of prenatal organophosphate ester
(OPE) metabolite mixtures and adverse infant birth outcomes in the maternal and developmental risks from environmental and social stressors (MADRES)
pregnancy cohort. Environmental Research 226:115703.
Reproductive /Developmental-Gestational age at birth, gestational-age adjusted birth weight z-scores
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11577122
Domain
Metric
Rating
Comments
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High BCEP was analyzed using HPLC with mass spectrometry. Internal standards were used
for each OPE. An overview of laboratory methods was provided in limited detail. The
authors stated that each batch analyzed included spiked quality control samples and
replicates of quality control standard reference materials. Repeated analysis of pooled
urine QC samples across batches showed coefficient of variation of ±12-31%. Though
precision was variable, there was no evidence of important error.
Medium Dilution was addressed using specific gravity adjusted OPE exposure variables. There
was no rationale provided, but there was no evidence the approach was inappropriate.
Results were only presented as matrix-adjusted.
Additional Comments: This study analyzed the relationship between several OPEs and birth outcomes among 421 mother-infant pairs from the Los Angeles MADRES cohort of
predominately low-income Latino subjects. BCEP, a major metabolite of TCEP, was measured in a third trimester spot urine sample. The study found no
association between urinary BCEP and either gestational age at birth weight for gestational age overall or stratified by infant sex. Strengths of this study
include the prospective design. A potential limitation includes variability in gestational age estimated for the same participant based on alternative methods
(e.g. ultrasound from different trimesters), but there was no evidence of important error or bias.
Overall Quality Determination Medium
Page 79 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581666 Table: 1 of 1
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Hernandez-Castro, I., Eckel, S. P., Howe, C. G., Niu, Z., Kannan, K., Robinson, M., Foley, H. B., Yang, T., Vigil, M. J., Chen, X., Grubbs, B., Lerner,
D., Lurvey, N., Al-Marayati, L., Habre, R., Dunton, G. F., Farzan, S. F., Aung, M. T., Breton, C. V., Bastain, T. M. (2023). Prenatal exposures to
organophosphate ester metabolite mixtures and children's neurobehavioral outcomes in the MADRES pregnancy cohort. Environmental Health 22(1 ):66.
Neurological/Behavioral-Internalizing Problems, Externalizing Problems, and Total Problems scores from the Child Behavior Checklist (CBCL)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11581666
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2:
Attrition
Metric 3: Comparison Group
Medium This study used data from the ongoing MADRES (Maternal and Developmental Risks
from Environmental and Social Stressors) cohort to analyze associations between the
TCEP metabolite bis(2-chloroethyl) phosphate (BCEP) in maternal urine and child be-
havioral outcomes at age 36 months. Urine samples were collected in the third trimester
between 2017 and 2019. MADRES includes predominately low-income Hispanic/Latino
mother-child pairs recruited through community health practices and advertising in ur-
ban Los Angeles. Eligible women are > aged 18 years, have singleton pregnancies, and
are recruited prior to 30 weeks' gestation. Exclusion criteria included disabilities, cur-
rent incarceration, and HIV positivity. Participation rates were not described. However,
there was no evidence to suggest that participation was associated with either TCEP ex-
posure or child behavior. A total of n=204 mother-infant pairs were included in the final
analysis, compared to the n=426 full cohort of MADRES participants. The study reports
that the distribution of OPE metabolites was similar between the two subsets.
Medium Of 774 MADRES cohort births through August 2022, 426 had prenatal urinary
organophosphate ester (OPE) measures. From this OPE subgroup, the final analysis
sample of 204 children excluded 181 children who either had not reached 36 months or
had passed that age before the Child Behavior Checklist (CBCL) was added to the study,
and another 41 who were missing CBCL data. Descriptive data showed that the analy-
sis sample and urinary OPE subsample had characteristics similar to the parent cohort
(e.g. 54 to 56% < high school education, mean gestational age 39 weeks in all groups).
TCEP concentrations were also similar in the final analysis sample and OPE subsam-
ple (68.6 vs 68.3% detection rates; median [IQR] 0.47 (0.02-1.60) vs. 0.53 (0.02-1.62).
Though attrition was high (the analysis sample included 26.3% of the parent cohort),
there was no evidence of bias.
Medium Analyses compared child behavioral outcomes among participants with lower vs. higher
BCEP in maternal urine during gestation. Though descriptive data on participant char-
acteristics were not stratified by urinary BCEP concentrations, there was no evidence to
suggest the comparison group with lower BCEP differed from those with higher concen-
trations and key demographic characteristics were considered as potential covariates in
statistical analysis.
Domain 2: Exposure Characterization
Continued on next page .
Page 80 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581666 Table: 1 of 1
... continued from previous page
Hernandez-Castro, I., Eckel, S. P., Howe, C. G., Niu, Z., Kannan, K., Robinson, M., Foley, H. B., Yang, T., Vigil, M. J., Chen, X., Grubbs, B., Lerner,
D., Lurvey, N., Al-Marayati, L., Habre, R., Dunton, G. F., Farzan, S. F., Aung, M. T., Breton, C. V., Bastain, T. M. (2023). Prenatal exposures to
organophosphate ester metabolite mixtures and children's neurobehavioral outcomes in the MADRES pregnancy cohort. Environmental Health 22(1 ):66.
Neurological/Behavioral-Internalizing Problems, Externalizing Problems, and Total Problems scores from the Child Behavior Checklist (CBCL)
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11581666
Domain
Metric
Rating
Comments
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6: Temporality
Medium Prenatal TCEP exposure was estimated by BCEP concentrations in a single third
trimester urine sample adjusted for dilution using specific gravity. BCEP was detected
in 68.6% of samples. In several models, exposure was categorized in tertiles, and values
below LOD were used as the referent. Analyses also imputed values below the LOD of
0.02 ng/mL as LOD divided by the square root of 2. Use of the main TCEP metabolite
as a biomarker was a strength of exposure measurement. Using a single spot urine sam-
ple was a possible limitation: as a single measure may misclassify habitual exposure
given uncertain variability in and sources of urinary BCEP in this population. Another
potential limitation is that unmetabolized TCEP—which was detected along with BCEP
in 37% of subjects in another study - was not measured (e.g., Hou et al. 2020 HEROID
10143372). Use of organophosphate ester measures from late vs early pregnancy is also
a possible limitation; other studies found adverse associations with toddler neurobehav-
ioral outcomes only for first trimester measures (e.g., Wang et al 2023 PMID37856202).
However, there was no evidence that third trimester concentrations would systematically
differ from those in early gestation. While some non-differential misclassification of
exposure is likely, there was no evidence of bias.
Medium BCEP was analyzed using both natural log-transformed continuous measures and us-
ing tertiles. While the range of exposures was relatively small, the range is likely large
enough to allow for sufficient contrast (25th-75th percentiles: 0.02-1.60 ng/mL). No
concerns.
Medium BCEP concentrations were measured prenatally, preceding childbirth and all observa-
tions of child behavior. While it is unclear whether the exact relevant time window for
developmental effects due to TCEP exposure was captured, there were no specific con-
cerns for temporality.
Domain 3: Outcome Assessment
Metric 7: Outcome Measurement or Medium Child behavioral outcomes were characterized based on maternal responses to the Child
Characterization Behavior Checklist (CBCL), which was orally administered at 36 months. The CBCL
is a validated instrument widely used for evaluating emotional and behavioral problems
in children aged 1.5 to 5 years. To facilitate comparisons with previous studies, primary
analyses used raw composite scores for three common CBCL metrics: internalizing,
externalizing and total problems, adjusting for prematurity-corrected age at CBCL ad-
ministration and child sex. CBCL T-scores, standardized vs. a US population, were also
analyzed. Additional CBCL subscales (e.g. emotionally reactive, attention problems)
were not analyzed in this paper. Interviewer training was not discussed, nor was the
CBCL validity or reliability within the study population. However, there was no evi-
dence of inadequacies in outcome assessment.
Continued on next page ...
Page 81 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581666 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Hernandez-Castro, I., Eckel, S. P., Howe, C. G., Niu, Z., Kannan, K., Robinson, M., Foley, H. B., Yang, T., Vigil, M. J., Chen, X., Grubbs, B., Lerner,
D., Lurvey, N., Al-Marayati, L., Habre, R., Dunton, G. F., Farzan, S. F., Aung, M. T., Breton, C. V., Bastain, T. M. (2023). Prenatal exposures to
organophosphate ester metabolite mixtures and children's neurobehavioral outcomes in the MADRES pregnancy cohort. Environmental Health 22(1 ):66.
Neurological/Behavioral-Internalizing Problems, Externalizing Problems, and Total Problems scores from the Child Behavior Checklist (CBCL)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11581666
Domain
Metric
Rating
Comments
Metric 8: Reporting Bias
High
Results were presented or described for all analyses included as aims.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment Medium Covariates were selected based on prior literature and a minimally sufficient set devel-
oped using a directed acyclic graph (DAG). Models adjusted for maternal age, parity,
pre-pregnancy BMI, race/ethnicity, income, education, child sex, child adjusted age at
CBCL administration, and study design or sample collection variables that changed ef-
fect estimates by 10% or more (recruitment site, specimen collection season, gestational
age at sample collection). Maternal smoking during pregnancy was analyzed by exclu-
sion in a sensitivity analysis due to low prevalence of this behavior. Potential confound-
ing or modification by factors such as size at birth, gestational diabetes or hypertension
was not discussed; however, these variables might also be intermediates and thus appro-
priately excluded. There was no evidence of important bias due to residual confounding.
Metric 10: Covariate Characterization Medium Covariates were characterized using information from interviewer-administered ques-
tionnaires and medical records. Validation was not discussed, but there was no evidence
of important error or bias.
Metric 11: Co-exposure Counfounding Medium Correlations with BCEP were weak (Spearman's r <0.25) for other organophosphate
ester (OPE) metabolites measured in this study, which included diphenyl phosphate
(DPHP), a composite of di-n-butyl phosphate and di-isobutyl phosphate which co-eluted
(DNBP + DIBP), bis(l,3,-dichloro-2-propyl) phosphate (BDCIPP), bis(butoxethyl)
phosphate (BBOEP), bis(l-chloro-2-propyl) phosphate (BCIPP), bis(2-ethylhexyl)
phosphate (BEHP), bis(2-methylphenyl) phosphate (BMPP), and dipropyl phosphate
(DPRP). Co-exposure to other OPEs during pregnancy was evaluated using Bayesian
methods to analyze chemical mixtures as well as synergistic interactions among chemi-
cals. Confounding by postnatal exposure to these chemicals was not evaluated. However
there was no evidence of bias.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
High A prospective cohort design was used to analyze the relationship between prenatal
TCEP exposure and child behavior outcomes at age 36 months. The study design was
appropriate.
Medium The sample included 204 mother-child pairs. There was variability in both exposure
(BCEP median [IQR] = 0.47 [0.02-1.60] and CBCL outcome variables (e.g., for in-
ternalizing problems median 8.0 [IQR = 12.0]). There was no evidence of inadequate
statistical power.
Medium The description of the analysis is sufficient to understand precisely what was done and
to replicate findings.
Continued on next page ...
Page 82 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581666 Table: 1 of 1
... continued from previous page
Hernandez-Castro, I., Eckel, S. P., Howe, C. G., Niu, Z., Kannan, K., Robinson, M., Foley, H. B., Yang, T., Vigil, M. J., Chen, X., Grubbs, B., Lerner,
D., Lurvey, N., Al-Marayati, L., Habre, R., Dunton, G. F., Farzan, S. F., Aung, M. T., Breton, C. V., Bastain, T. M. (2023). Prenatal exposures to
organophosphate ester metabolite mixtures and children's neurobehavioral outcomes in the MADRES pregnancy cohort. Environmental Health 22(1 ):66.
Neurological/Behavioral-Internalizing Problems, Externalizing Problems, and Total Problems scores from the Child Behavior Checklist (CBCL)
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11581666
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Descriptive data were presented for the sample as a whole. Urinary BCEP was described
using percentiles, the geometric means, and the minimum and maximum. To satisfy
model assumptions and evaluate potentially non-linear dose-response, primary mod-
els for individual exposures analyzed associations with natural-log transformed CBCL
scores using either BCEP tertiles or using generalized additive models. Both crude
and adjusted effect estimates were presented with 95% confidence intervals. Interac-
tion terms and stratified models were used to assess effect modification by child sex. In
addition, Bayesian kernel machine regression (BKMR) models were used to analyze
OPE mixtures, and Bayesian semiparametric regression models were used to analyze
synergistic or antagonistic interactions among pairs of chemicals. Both sets of models
allowed for non-linear dose-response patterns. Results were presented graphically and
using posterior inclusion probabilities with 95% credible intervals. Sensitivity anal-
yses to assess robustness included excluding children whose mothers smoked during
pregnancy, analyzing CBCL t-scores vs raw scores, and adjusting BKMR smoothness
parameters. Analyses appeared to be appropriate and there was no evidence of deficien-
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
High
N/A
Medium
Medium
Medium
Urinary concentrations of BCEP, the primary metabolite of TCEP, were used to estimate
exposure.
No biomarkers of effect were assessed.
BCEP was detected in 68.63% of urine samples; values below the LOD of 0.02 ng/mL
were imputed as LOD divided by the square root of two. Though a substantial propor-
tion of samples were below LOD, there was no evidence of inadequate sensitivity.
Processed samples were stored at -80 degrees C prior to shipment. The duration of pro-
cessing time at room temperature, and total storage time prior to analysis, were not
described, but there was no evidence of instability or degradation. Quality control meth-
ods are detailed.
Urine samples were collected in sterile specimen containers, and aliquoted in glass
tubes. The authors reported that "trace" levels (not quantified) of all OPE diester
metabolites were found in procedural blanks, and that the concentrations measured in
blanks were subtracted from sample values for each batch. Further details, such as vari-
ability in trace amounts across batches and the number of batches analyzed, were not
provided. However, there was no evidence of inappropriate handling of samples, or of
bias resulting from the approach used to address trace levels in blanks.
Continued on next page ...
Page 83 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581666 Table: 1 of 1
... continued from previous page
Hernandez-Castro, I., Eckel, S. P., Howe, C. G., Niu, Z., Kannan, K., Robinson, M., Foley, H. B., Yang, T., Vigil, M. J., Chen, X., Grubbs, B., Lerner,
D., Lurvey, N., Al-Marayati, L., Habre, R., Dunton, G. F., Farzan, S. F., Aung, M. T., Breton, C. V., Bastain, T. M. (2023). Prenatal exposures to
organophosphate ester metabolite mixtures and children's neurobehavioral outcomes in the MADRES pregnancy cohort. Environmental Health 22(1 ):66.
Neurological/Behavioral-Internalizing Problems, Externalizing Problems, and Total Problems scores from the Child Behavior Checklist (CBCL)
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11581666
Domain
Metric
Rating
Comments
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
Medium BCEP was measured using HPLC with mass spectrometry. Unspecified internal stan-
dards were used. Each batch analyzed included spiked quality control samples and repli-
cates of quality control standard reference materials. Standardized reference materials
had coefficients of variation of up to ± 12-40%; average recoveries of matrix spiked
samples ranged from 70.4-133% (coefficients of variation ± 9-19%). Though details
were limited and precision was variable, there was no evidence of important error or bias
in BCEP measures.
Medium Samples were adjusted for dilution using specific gravity measured in room temperature
urine samples using a digital handheld refractometer. Specific gravity-adjusted concen-
trations were used in analysis. A rationale for this approach was not discussed, but there
was no evidence for concern. Results were only shown for concentrations adjusted for
dilution.
Additional Comments: This study used data on 204 mother-infant pairs from the Los Angeles MADRES cohort of predominately low-income Latino subjects. TCEP exposure
was estimated using BCEP in a single third trimester spot urine sample collected in 2017-2019; child behavior was assessed using scores on the Child
Behavior Checklist at age 36 months. The study found positive, non-significant associations between BCEP and CBCL internalizing, externalizing and total
problems scores. Sex differences were not significant. In addition, using BKMR models, there were stronger positive associations between DNBP+DIBP
(co-eluting OPEs) and both internalizing and total problems scores at higher quartiles of BCEP. A significant interaction, tested for the strongest interaction,
was confirmed using generalized additive models. Strengths of this study include the prospective design, and that characteristics of the analysis sample
were similar to the parent cohort.Potential limitations include that the analysis sample comprised 204 of more than 700 MADRES participants, and that
exposure estimation was based on a single spot sample from the third trimester. There was no evidence of important error or bias.
Overall Quality Determination Medium
Page 84 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 1 of 5
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
Medium Participants were drawn from the National Health and Nutrition Examination Survey
(NHANES) study (2011-2012 cycle), a study on a nationally representative sample
conducted by the CDC. This cycle included both urinary measures of organophosphate
flame retardants and lung function spirometry measures. All NHANES participants
aged 20 years or older (n=5,560) were eligible if they had data on urinary organophos-
phate flame retardant (OPFR) metabolites and pulmonary parameters. From the 5,560
adults included in the 2011-2012 cycle, 4,178 without OPFR metabolites were excluded,
384 with missing or inadequate pulmonary parameters were excluded, and 11 pregnant
women were excluded. This resulted in a final sample of 987 participants. No informa-
tion was provided to demonstrate the exposure-outcome distribution of the participants
is likely representative of the exposure-outcome distributions in the population of per-
sons eligible for inclusion in the study, however, the study is intended to be a nationally
representative sample and risk of selection bias is low. There was nothing to suggest the
subset with available OPFR measures was selective. Further details on recruitment and
participant selection can be found on the CDC NHANES website.
Medium Reasons for exclusion were appropriately outlined, and only those with complete expo-
sure and outcome information were included in the analysis. Details such as exclusions
due to ineligibility (e.g., due to heart attack or stroke) vs. inadequate quality of spirome-
try were not provided, but there was no evidence of bias.
High All participants were drawn from NHANES, a nationally representative study sam-
ple, using the same inclusion and exclusion criteria. The study was restricted to adults
(>=20 years old) and pregnant women were excluded.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure Medium A metabolite of TCEP, BCEP, was measured in urine to indicate exposure to TCEP.
Samples were analyzed using automated off-line solid phase extraction, reversed-phase
high-performance liquid chromatography followed by isotope dilution-electrospray ion-
ization tandem mass spectrometry. Limits of detection and QA/QC procedures were
described, and further details can be found on the CDC NHANES website. The extent
to which chronic TCEP exposure may be misclassified using a single spot urine concen-
tration of BCEP is uncertain. The authors did not discuss the half-life of urinary BCEP
or include measures of the parent compound along with the metabolite (Wang et al 2020
HEROID 7276658; Hou et al 10143372).
Continued on next page ...
Page 85 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 1 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 5:
Metric 6:
Exposure Levels
Temporality
Medium Exposure to TCEP was evaluated based on urinary BCEP concentrations, analyzed
continuously. BCEP was detectable in 85.31% of urine samples, and values below LOD
were imputed as the LOD divided by the square root of 2. No concerns about exposure
distribution (BCEP median [25th-75th percentile]: 0.51 ug/L [0.21, 1.03]). Urinary
creatinine was included in models to account for urine dilution.
Medium Urine samples used to estimate OPFR exposure and spirometry used to characterize the
respiratory function outcomes were collected during the same NHANES study visit.
While temporality cannot be established due to the cross-sectional design given the
short half-life (<30 days) of the exposure biomarker, there is no evidence of reverse
causation.
Domain 3: Outcome Assessment
Metric 7:
Outcome Measurement or
Characterization
Metric 8: Reporting Bias
High Spirometry testing procedures were performed according to the recommendations of the
American Thoracic Society (ATS). Measurements included forced expiratory volume
first second (FEV1), forced vital capacity (FVC), peak expiratory flowrate (PEF), forced
expiratory flow at 25-75% of the FVC (FEF25-75%), and FEV1/FVC. Measurements
were performed in a standing position (except for physical limitation). "A maximum of
eight spirometry tests were conducted to obtain three acceptable curves. The two high-
est values for FEV1 and FVC were used for quality ratings." Only those subjects with
FEV1 and FVC measurements rated as A or B (according to ATS data collection stan-
dards) were included. Further details can be found in the NHANES Respiratory Health
Spirometry Procedures Manual. Blinding was not mentioned, however, spirometry data
was collected at the same visit as urine was collected, indirectly indicating spirometry
technicians would be unaware of exposure status.
High All outcomes outlined in the abstract, introduction, and methods were provided in the
results or supplementary materials. Risk estimates were reported with p-values and
confidence intervals.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment Medium Study authors provided a simple model adjusted for age, sex, and race (model 1) and
a model additionally adjusted for BMI, serum cotinine, smoking status, physical ac-
tivity, family poverty/income ratio, educational level, and urinary creatinine. Study
authors conducted a stratified analysis to evaluate interaction by smoking status (never,
former, current). Examining the robustness of findings in sensitivity analyses was not
discussed (e.g. potential influence of participants with diagnoses respiratory illness such
as asthma).
Continued on next page ...
Page 86 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 1 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 10:
Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium Covariates were collected by trained interviewers during NHANES study visits; these
included self-reported sociodemographic measures. Complete information on collection
covariates can be found in the NHANES (2011-2012) Procedure Manuals. There was no
evidence of covariates with inadequate validity.
Medium Other OPFR metabolites were measured in urine. However, correlations among metabo-
lites were not provided, and models did not adjust for other metabolites. There was no
direct evidence of confounding by these or other co-exposures.
Domain 5: Analysis
Metric
12:
Study Design and Methods
High
The study design was appropriate to examine the research question (i.e., the association
between exposure to TCEP and pulmonary function). Appropriate statistical methods
(linear regression models) were used to evaluate the exposure-outcome relationship.
Metric
13:
Statistical Power
Medium
There was variability in exposure, and the number of participants (n=987) was likely
sufficient to detect an effect.
Metric
14:
Reproducibility of Analyses
Medium
The description of the analysis is sufficient to understand precisely what has been done
and to be conceptually reproducible with access to the analytic data.
Metric
15:
Statistical Analysis
High
Linear regression models were adequately described. Study authors noted that exposure
data was natural log transformed to normalize the distribution. Minor potential concerns
include that the non-linearity of dose-response relationships and sex-specific effects
were not discussed.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Medium
N/A
Medium
Medium
Medium
Supplemental material (Table SI) and the Pubchem dashboard indicate BCEP is the
direct metabolite of TCEP. No concerns about accuracy or precision of metabolite mea-
surements in urine. However, the extent to which a single spot urine BCEP accurately
classified TCEP exposure in non-occupationally exposed populations is not fully estab-
lished.
No biomarkers of effect were assessed.
The LOD (0.10 ug/L) was low enough to detect chemicals in a sufficient percentage of
the samples (85.31%) to address the research question, and the method was adequately
described.
Samples were collected during an NHANES study visit and documentation on storage
is provided in the NHANES Laboratory Procedures Manual. There was no information
provided on biomarker stability, but no evidence for concern was documented.
There is no specific information on sample contamination, but study authors note that
samples were collected "according to the analytic guidelines of the NHANES." Specific
information may be found in the NHANES Laboratory Procedures Manual.
Continued on next page ...
Page 87 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 1 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 21:
Metric 22:
Method Requirements
Matrix Adjustment
High
Medium
Samples were analyzed using automated off-line solid phase extraction, reversed-phase
high-performance liquid chromatography followed by isotope dilution-electrospray
ionization tandem mass spectrometry.
Model 1 adjusted for age, sex and race but excluded urinary creatinine, while Model 2
provides additionally adjusted for urinary creatinine and several other covariates. Asso-
ciations with and without creatinine were shown. However, the influence of creatinine
adjustment vs other confounding cannot be isolated.
Additional Comments:
This was a well-conducted study based on available NHANES data in adults, using the 2011-2012 cycle. There were no large concerns with participant
selection or methods relating to exposure measurement, outcome ascertainment, and statistical analysis. Minor concerns include that temporality cannot
be fully established due to the cross-sectional design, and that co-exposure confounding was not explored.
Overall Quality Determination
Medium
Page 88 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 2 of 5
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
Medium Participants were drawn from the National Health and Nutrition Examination Survey
(NHANES) study (2011-2012 cycle), a study on a nationally representative sample
conducted by the CDC. This cycle included both urinary measures of organophosphate
flame retardants and lung function spirometry measures. All NHANES participants
aged 20 years or older (n=5,560) were eligible if they had data on urinary organophos-
phate flame retardant (OPFR) metabolites and pulmonary parameters. From the 5,560
adults included in the 2011-2012 cycle, 4,178 without OPFR metabolites were excluded,
384 with missing or inadequate pulmonary parameters were excluded, and 11 pregnant
women were excluded. This resulted in a final sample of 987 participants. No informa-
tion was provided to demonstrate the exposure-outcome distribution of the participants
is likely representative of the exposure-outcome distributions in the population of per-
sons eligible for inclusion in the study, however, the study is intended to be a nationally
representative sample and risk of selection bias is low. There was nothing to suggest the
subset with available OPFR measures was selective. Further details on recruitment and
participant selection can be found on the CDC NHANES website.
Medium Reasons for exclusion were appropriately outlined, and only those with complete expo-
sure and outcome information were included in the analysis. Details such as exclusions
due to ineligibility (e.g., due to heart attack or stroke) vs. inadequate quality of spirome-
try were not provided, but there was no evidence of bias.
High All participants were drawn from NHANES, a nationally representative study sam-
ple, using the same inclusion and exclusion criteria. The study was restricted to adults
(>=20 years old) and pregnant women were excluded.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure Medium A metabolite of TCEP, BCEP, was measured in urine to indicate exposure to TCEP.
Samples were analyzed using automated off-line solid phase extraction, reversed-phase
high-performance liquid chromatography followed by isotope dilution-electrospray ion-
ization tandem mass spectrometry. Limits of detection and QA/QC procedures were
described, and further details can be found on the CDC NHANES website. The extent
to which chronic TCEP exposure may be misclassified using a single spot urine concen-
tration of BCEP is uncertain. The authors did not discuss the half-life of urinary BCEP
or include measures of the parent compound along with the metabolite (Wang et al 2020
HEROID 7276658; Hou et al 10143372).
Continued on next page ...
Page 89 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 2 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 5:
Metric 6:
Exposure Levels
Temporality
Medium Exposure to TCEP was evaluated based on urinary BCEP concentrations, analyzed
continuously. BCEP was detectable in 85.31% of urine samples, and values below LOD
were imputed as the LOD divided by the square root of 2. No concerns about exposure
distribution (BCEP median [25th-75th percentile]: 0.51 ug/L [0.21, 1.03]). Urinary
creatinine was included in models to account for urine dilution.
Medium Urine samples used to estimate OPFR exposure and spirometry used to characterize the
respiratory function outcomes were collected during the same NHANES study visit.
While temporality cannot be established due to the cross-sectional design given the
short half-life (<30 days) of the exposure biomarker, there is no evidence of reverse
causation.
Domain 3: Outcome Assessment
Metric 7:
Outcome Measurement or
Characterization
Metric 8: Reporting Bias
High Spirometry testing procedures were performed according to the recommendations of the
American Thoracic Society (ATS). Measurements included forced expiratory volume
first second (FEV1), forced vital capacity (FVC), peak expiratory flowrate (PEF), forced
expiratory flow at 25-75% of the FVC (FEF25-75%), and FEV1/FVC. Measurements
were performed in a standing position (except for physical limitation). "A maximum of
eight spirometry tests were conducted to obtain three acceptable curves. The two high-
est values for FEV1 and FVC were used for quality ratings." Only those subjects with
FEV1 and FVC measurements rated as A or B (according to ATS data collection stan-
dards) were included. Further details can be found in the NHANES Respiratory Health
Spirometry Procedures Manual. Blinding was not mentioned, however, spirometry data
was collected at the same visit as urine was collected, indirectly indicating spirometry
technicians would be unaware of exposure status.
High All outcomes outlined in the abstract, introduction, and methods were provided in the
results or supplementary materials. Risk estimates were reported with p-values and
confidence intervals.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment Medium Study authors provided a simple model adjusted for age, sex, and race (model 1) and
a model additionally adjusted for BMI, serum cotinine, smoking status, physical ac-
tivity, family poverty/income ratio, educational level, and urinary creatinine. Study
authors conducted a stratified analysis to evaluate interaction by smoking status (never,
former, current). Examining the robustness of findings in sensitivity analyses was not
discussed (e.g. potential influence of participants with diagnoses respiratory illness such
as asthma).
Continued on next page ...
Page 90 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 2 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 10:
Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium Covariates were collected by trained interviewers during NHANES study visits; these
included self-reported sociodemographic measures. Complete information on collection
covariates can be found in the NHANES (2011-2012) Procedure Manuals. There was no
evidence of covariates with inadequate validity.
Medium Other OPFR metabolites were measured in urine. However, correlations among metabo-
lites were not provided, and models did not adjust for other metabolites. There was no
direct evidence of confounding by these or other co-exposures.
Domain 5: Analysis
Metric
12:
Study Design and Methods
High
The study design was appropriate to examine the research question (i.e., the association
between exposure to TCEP and pulmonary function). Appropriate statistical methods
(linear regression models) were used to evaluate the exposure-outcome relationship.
Metric
13:
Statistical Power
Medium
There was variability in exposure, and the number of participants (n=987) was likely
sufficient to detect an effect.
Metric
14:
Reproducibility of Analyses
Medium
The description of the analysis is sufficient to understand precisely what has been done
and to be conceptually reproducible with access to the analytic data.
Metric
15:
Statistical Analysis
High
Linear regression models were adequately described. Study authors noted that exposure
data was natural log transformed to normalize the distribution. Minor potential concerns
include that the non-linearity of dose-response relationships and sex-specific effects
were not discussed.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Medium
N/A
Medium
Medium
Medium
Supplemental material (Table SI) and the Pubchem dashboard indicate BCEP is the
direct metabolite of TCEP. No concerns about accuracy or precision of metabolite mea-
surements in urine. However, the extent to which a single spot urine BCEP accurately
classified TCEP exposure in non-occupationally exposed populations is not fully estab-
lished.
No biomarkers of effect were assessed.
The LOD (0.10 ug/L) was low enough to detect chemicals in a sufficient percentage of
the samples (85.31%) to address the research question, and the method was adequately
described.
Samples were collected during an NHANES study visit and documentation on storage
is provided in the NHANES Laboratory Procedures Manual. There was no information
provided on biomarker stability, but no evidence for concern was documented.
There is no specific information on sample contamination, but study authors note that
samples were collected "according to the analytic guidelines of the NHANES." Specific
information may be found in the NHANES Laboratory Procedures Manual.
Continued on next page ...
Page 91 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 2 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 21:
Metric 22:
Method Requirements
Matrix Adjustment
High
Medium
Samples were analyzed using automated off-line solid phase extraction, reversed-phase
high-performance liquid chromatography followed by isotope dilution-electrospray
ionization tandem mass spectrometry.
Model 1 adjusted for age, sex and race but excluded urinary creatinine, while Model 2
provides additionally adjusted for urinary creatinine and several other covariates. Asso-
ciations with and without creatinine were shown. However, the influence of creatinine
adjustment vs other confounding cannot be isolated.
Additional Comments:
This was a well-conducted study based on available NHANES data in adults, using the 2011-2012 cycle. There were no large concerns with participant
selection or methods relating to exposure measurement, outcome ascertainment, and statistical analysis. Minor concerns include that temporality cannot
be fully established due to the cross-sectional design, and that co-exposure confounding was not explored.
Overall Quality Determination
Medium
Page 92 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 3 of 5
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
Medium Participants were drawn from the National Health and Nutrition Examination Survey
(NHANES) study (2011-2012 cycle), a study on a nationally representative sample
conducted by the CDC. This cycle included both urinary measures of organophosphate
flame retardants and lung function spirometry measures. All NHANES participants
aged 20 years or older (n=5,560) were eligible if they had data on urinary organophos-
phate flame retardant (OPFR) metabolites and pulmonary parameters. From the 5,560
adults included in the 2011-2012 cycle, 4,178 without OPFR metabolites were excluded,
384 with missing or inadequate pulmonary parameters were excluded, and 11 pregnant
women were excluded. This resulted in a final sample of 987 participants. No informa-
tion was provided to demonstrate the exposure-outcome distribution of the participants
is likely representative of the exposure-outcome distributions in the population of per-
sons eligible for inclusion in the study, however, the study is intended to be a nationally
representative sample and risk of selection bias is low. There was nothing to suggest the
subset with available OPFR measures was selective. Further details on recruitment and
participant selection can be found on the CDC NHANES website.
Medium Reasons for exclusion were appropriately outlined, and only those with complete expo-
sure and outcome information were included in the analysis. Details such as exclusions
due to ineligibility (e.g., due to heart attack or stroke) vs. inadequate quality of spirome-
try were not provided, but there was no evidence of bias.
High All participants were drawn from NHANES, a nationally representative study sam-
ple, using the same inclusion and exclusion criteria. The study was restricted to adults
(>=20 years old) and pregnant women were excluded.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure Medium A metabolite of TCEP, BCEP, was measured in urine to indicate exposure to TCEP.
Samples were analyzed using automated off-line solid phase extraction, reversed-phase
high-performance liquid chromatography followed by isotope dilution-electrospray ion-
ization tandem mass spectrometry. Limits of detection and QA/QC procedures were
described, and further details can be found on the CDC NHANES website. The extent
to which chronic TCEP exposure may be misclassified using a single spot urine concen-
tration of BCEP is uncertain. The authors did not discuss the half-life of urinary BCEP
or include measures of the parent compound along with the metabolite (Wang et al 2020
HEROID 7276658; Hou et al 10143372).
Continued on next page ...
Page 93 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 3 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 5:
Metric 6:
Exposure Levels
Temporality
Medium Exposure to TCEP was evaluated based on urinary BCEP concentrations, analyzed
continuously. BCEP was detectable in 85.31% of urine samples, and values below LOD
were imputed as the LOD divided by the square root of 2. No concerns about exposure
distribution (BCEP median [25th-75th percentile]: 0.51 ug/L [0.21, 1.03]). Urinary
creatinine was included in models to account for urine dilution.
Medium Urine samples used to estimate OPFR exposure and spirometry used to characterize the
respiratory function outcomes were collected during the same NHANES study visit.
While temporality cannot be established due to the cross-sectional design given the
short half-life (<30 days) of the exposure biomarker, there is no evidence of reverse
causation.
Domain 3: Outcome Assessment
Metric 7:
Outcome Measurement or
Characterization
Metric 8: Reporting Bias
High Spirometry testing procedures were performed according to the recommendations of the
American Thoracic Society (ATS). Measurements included forced expiratory volume
first second (FEV1), forced vital capacity (FVC), peak expiratory flowrate (PEF), forced
expiratory flow at 25-75% of the FVC (FEF25-75%), and FEV1/FVC. Measurements
were performed in a standing position (except for physical limitation). "A maximum of
eight spirometry tests were conducted to obtain three acceptable curves. The two high-
est values for FEV1 and FVC were used for quality ratings." Only those subjects with
FEV1 and FVC measurements rated as A or B (according to ATS data collection stan-
dards) were included. Further details can be found in the NHANES Respiratory Health
Spirometry Procedures Manual. Blinding was not mentioned, however, spirometry data
was collected at the same visit as urine was collected, indirectly indicating spirometry
technicians would be unaware of exposure status.
High All outcomes outlined in the abstract, introduction, and methods were provided in the
results or supplementary materials. Risk estimates were reported with p-values and
confidence intervals.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment Medium Study authors provided a simple model adjusted for age, sex, and race (model 1) and
a model additionally adjusted for BMI, serum cotinine, smoking status, physical ac-
tivity, family poverty/income ratio, educational level, and urinary creatinine. Study
authors conducted a stratified analysis to evaluate interaction by smoking status (never,
former, current). Examining the robustness of findings in sensitivity analyses was not
discussed (e.g. potential influence of participants with diagnoses respiratory illness such
as asthma).
Continued on next page ...
Page 94 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 3 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 10:
Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium Covariates were collected by trained interviewers during NHANES study visits; these
included self-reported sociodemographic measures. Complete information on collection
covariates can be found in the NHANES (2011-2012) Procedure Manuals. There was no
evidence of covariates with inadequate validity.
Medium Other OPFR metabolites were measured in urine. However, correlations among metabo-
lites were not provided, and models did not adjust for other metabolites. There was no
direct evidence of confounding by these or other co-exposures.
Domain 5: Analysis
Metric
12:
Study Design and Methods
High
The study design was appropriate to examine the research question (i.e., the association
between exposure to TCEP and pulmonary function). Appropriate statistical methods
(linear regression models) were used to evaluate the exposure-outcome relationship.
Metric
13:
Statistical Power
Medium
There was variability in exposure, and the number of participants (n=987) was likely
sufficient to detect an effect.
Metric
14:
Reproducibility of Analyses
Medium
The description of the analysis is sufficient to understand precisely what has been done
and to be conceptually reproducible with access to the analytic data.
Metric
15:
Statistical Analysis
High
Linear regression models were adequately described. Study authors noted that exposure
data was natural log transformed to normalize the distribution. Minor potential concerns
include that the non-linearity of dose-response relationships and sex-specific effects
were not discussed.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Medium
N/A
Medium
Medium
Medium
Supplemental material (Table SI) and the Pubchem dashboard indicate BCEP is the
direct metabolite of TCEP. No concerns about accuracy or precision of metabolite mea-
surements in urine. However, the extent to which a single spot urine BCEP accurately
classified TCEP exposure in non-occupationally exposed populations is not fully estab-
lished.
No biomarkers of effect were assessed.
The LOD (0.10 ug/L) was low enough to detect chemicals in a sufficient percentage of
the samples (85.31%) to address the research question, and the method was adequately
described.
Samples were collected during an NHANES study visit and documentation on storage
is provided in the NHANES Laboratory Procedures Manual. There was no information
provided on biomarker stability, but no evidence for concern was documented.
There is no specific information on sample contamination, but study authors note that
samples were collected "according to the analytic guidelines of the NHANES." Specific
information may be found in the NHANES Laboratory Procedures Manual.
Continued on next page ...
Page 95 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 3 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 21:
Metric 22:
Method Requirements
Matrix Adjustment
High
Medium
Samples were analyzed using automated off-line solid phase extraction, reversed-phase
high-performance liquid chromatography followed by isotope dilution-electrospray
ionization tandem mass spectrometry.
Model 1 adjusted for age, sex and race but excluded urinary creatinine, while Model 2
provides additionally adjusted for urinary creatinine and several other covariates. Asso-
ciations with and without creatinine were shown. However, the influence of creatinine
adjustment vs other confounding cannot be isolated.
Additional Comments:
This was a well-conducted study based on available NHANES data in adults, using the 2011-2012 cycle. There were no large concerns with participant
selection or methods relating to exposure measurement, outcome ascertainment, and statistical analysis. Minor concerns include that temporality cannot
be fully established due to the cross-sectional design, and that co-exposure confounding was not explored.
Overall Quality Determination
Medium
Page 96 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 4 of 5
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
Medium Participants were drawn from the National Health and Nutrition Examination Survey
(NHANES) study (2011-2012 cycle), a study on a nationally representative sample
conducted by the CDC. This cycle included both urinary measures of organophosphate
flame retardants and lung function spirometry measures. All NHANES participants
aged 20 years or older (n=5,560) were eligible if they had data on urinary organophos-
phate flame retardant (OPFR) metabolites and pulmonary parameters. From the 5,560
adults included in the 2011-2012 cycle, 4,178 without OPFR metabolites were excluded,
384 with missing or inadequate pulmonary parameters were excluded, and 11 pregnant
women were excluded. This resulted in a final sample of 987 participants. No informa-
tion was provided to demonstrate the exposure-outcome distribution of the participants
is likely representative of the exposure-outcome distributions in the population of per-
sons eligible for inclusion in the study, however, the study is intended to be a nationally
representative sample and risk of selection bias is low. There was nothing to suggest the
subset with available OPFR measures was selective. Further details on recruitment and
participant selection can be found on the CDC NHANES website.
Medium Reasons for exclusion were appropriately outlined, and only those with complete expo-
sure and outcome information were included in the analysis. Details such as exclusions
due to ineligibility (e.g., due to heart attack or stroke) vs. inadequate quality of spirome-
try were not provided, but there was no evidence of bias.
High All participants were drawn from NHANES, a nationally representative study sam-
ple, using the same inclusion and exclusion criteria. The study was restricted to adults
(>=20 years old) and pregnant women were excluded.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure Medium A metabolite of TCEP, BCEP, was measured in urine to indicate exposure to TCEP.
Samples were analyzed using automated off-line solid phase extraction, reversed-phase
high-performance liquid chromatography followed by isotope dilution-electrospray ion-
ization tandem mass spectrometry. Limits of detection and QA/QC procedures were
described, and further details can be found on the CDC NHANES website. The extent
to which chronic TCEP exposure may be misclassified using a single spot urine concen-
tration of BCEP is uncertain. The authors did not discuss the half-life of urinary BCEP
or include measures of the parent compound along with the metabolite (Wang et al 2020
HEROID 7276658; Hou et al 10143372).
Continued on next page ...
Page 97 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 4 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 5:
Metric 6:
Exposure Levels
Temporality
Medium Exposure to TCEP was evaluated based on urinary BCEP concentrations, analyzed
continuously. BCEP was detectable in 85.31% of urine samples, and values below LOD
were imputed as the LOD divided by the square root of 2. No concerns about exposure
distribution (BCEP median [25th-75th percentile]: 0.51 ug/L [0.21, 1.03]). Urinary
creatinine was included in models to account for urine dilution.
Medium Urine samples used to estimate OPFR exposure and spirometry used to characterize the
respiratory function outcomes were collected during the same NHANES study visit.
While temporality cannot be established due to the cross-sectional design given the
short half-life (<30 days) of the exposure biomarker, there is no evidence of reverse
causation.
Domain 3: Outcome Assessment
Metric 7:
Outcome Measurement or
Characterization
Metric 8: Reporting Bias
High Spirometry testing procedures were performed according to the recommendations of the
American Thoracic Society (ATS). Measurements included forced expiratory volume
first second (FEV1), forced vital capacity (FVC), peak expiratory flowrate (PEF), forced
expiratory flow at 25-75% of the FVC (FEF25-75%), and FEV1/FVC. Measurements
were performed in a standing position (except for physical limitation). "A maximum of
eight spirometry tests were conducted to obtain three acceptable curves. The two high-
est values for FEV1 and FVC were used for quality ratings." Only those subjects with
FEV1 and FVC measurements rated as A or B (according to ATS data collection stan-
dards) were included. Further details can be found in the NHANES Respiratory Health
Spirometry Procedures Manual. Blinding was not mentioned, however, spirometry data
was collected at the same visit as urine was collected, indirectly indicating spirometry
technicians would be unaware of exposure status.
High All outcomes outlined in the abstract, introduction, and methods were provided in the
results or supplementary materials. Risk estimates were reported with p-values and
confidence intervals.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment Medium Study authors provided a simple model adjusted for age, sex, and race (model 1) and
a model additionally adjusted for BMI, serum cotinine, smoking status, physical ac-
tivity, family poverty/income ratio, educational level, and urinary creatinine. Study
authors conducted a stratified analysis to evaluate interaction by smoking status (never,
former, current). Examining the robustness of findings in sensitivity analyses was not
discussed (e.g. potential influence of participants with diagnoses respiratory illness such
as asthma).
Continued on next page ...
Page 98 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 4 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 10:
Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium Covariates were collected by trained interviewers during NHANES study visits; these
included self-reported sociodemographic measures. Complete information on collection
covariates can be found in the NHANES (2011-2012) Procedure Manuals. There was no
evidence of covariates with inadequate validity.
Medium Other OPFR metabolites were measured in urine. However, correlations among metabo-
lites were not provided, and models did not adjust for other metabolites. There was no
direct evidence of confounding by these or other co-exposures.
Domain 5: Analysis
Metric
12:
Study Design and Methods
High
The study design was appropriate to examine the research question (i.e., the association
between exposure to TCEP and pulmonary function). Appropriate statistical methods
(linear regression models) were used to evaluate the exposure-outcome relationship.
Metric
13:
Statistical Power
Medium
There was variability in exposure, and the number of participants (n=987) was likely
sufficient to detect an effect.
Metric
14:
Reproducibility of Analyses
Medium
The description of the analysis is sufficient to understand precisely what has been done
and to be conceptually reproducible with access to the analytic data.
Metric
15:
Statistical Analysis
High
Linear regression models were adequately described. Study authors noted that exposure
data was natural log transformed to normalize the distribution. Minor potential concerns
include that the non-linearity of dose-response relationships and sex-specific effects
were not discussed.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Medium
N/A
Medium
Medium
Medium
Supplemental material (Table SI) and the Pubchem dashboard indicate BCEP is the
direct metabolite of TCEP. No concerns about accuracy or precision of metabolite mea-
surements in urine. However, the extent to which a single spot urine BCEP accurately
classified TCEP exposure in non-occupationally exposed populations is not fully estab-
lished.
No biomarkers of effect were assessed.
The LOD (0.10 ug/L) was low enough to detect chemicals in a sufficient percentage of
the samples (85.31%) to address the research question, and the method was adequately
described.
Samples were collected during an NHANES study visit and documentation on storage
is provided in the NHANES Laboratory Procedures Manual. There was no information
provided on biomarker stability, but no evidence for concern was documented.
There is no specific information on sample contamination, but study authors note that
samples were collected "according to the analytic guidelines of the NHANES." Specific
information may be found in the NHANES Laboratory Procedures Manual.
Continued on next page ...
Page 99 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 4 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 21:
Metric 22:
Method Requirements
Matrix Adjustment
High
Medium
Samples were analyzed using automated off-line solid phase extraction, reversed-phase
high-performance liquid chromatography followed by isotope dilution-electrospray
ionization tandem mass spectrometry.
Model 1 adjusted for age, sex and race but excluded urinary creatinine, while Model 2
provides additionally adjusted for urinary creatinine and several other covariates. Asso-
ciations with and without creatinine were shown. However, the influence of creatinine
adjustment vs other confounding cannot be isolated.
Additional Comments:
This was a well-conducted study based on available NHANES data in adults, using the 2011-2012 cycle. There were no large concerns with participant
selection or methods relating to exposure measurement, outcome ascertainment, and statistical analysis. Minor concerns include that temporality cannot
be fully established due to the cross-sectional design, and that co-exposure confounding was not explored.
Overall Quality Determination
Medium
Page 100 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 5 of 5
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1: Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
Medium Participants were drawn from the National Health and Nutrition Examination Survey
(NHANES) study (2011-2012 cycle), a study on a nationally representative sample
conducted by the CDC. This cycle included both urinary measures of organophosphate
flame retardants and lung function spirometry measures. All NHANES participants
aged 20 years or older (n=5,560) were eligible if they had data on urinary organophos-
phate flame retardant (OPFR) metabolites and pulmonary parameters. From the 5,560
adults included in the 2011-2012 cycle, 4,178 without OPFR metabolites were excluded,
384 with missing or inadequate pulmonary parameters were excluded, and 11 pregnant
women were excluded. This resulted in a final sample of 987 participants. No informa-
tion was provided to demonstrate the exposure-outcome distribution of the participants
is likely representative of the exposure-outcome distributions in the population of per-
sons eligible for inclusion in the study, however, the study is intended to be a nationally
representative sample and risk of selection bias is low. There was nothing to suggest the
subset with available OPFR measures was selective. Further details on recruitment and
participant selection can be found on the CDC NHANES website.
Medium Reasons for exclusion were appropriately outlined, and only those with complete expo-
sure and outcome information were included in the analysis. Details such as exclusions
due to ineligibility (e.g., due to heart attack or stroke) vs. inadequate quality of spirome-
try were not provided, but there was no evidence of bias.
High All participants were drawn from NHANES, a nationally representative study sam-
ple, using the same inclusion and exclusion criteria. The study was restricted to adults
(>=20 years old) and pregnant women were excluded.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure Medium A metabolite of TCEP, BCEP, was measured in urine to indicate exposure to TCEP.
Samples were analyzed using automated off-line solid phase extraction, reversed-phase
high-performance liquid chromatography followed by isotope dilution-electrospray ion-
ization tandem mass spectrometry. Limits of detection and QA/QC procedures were
described, and further details can be found on the CDC NHANES website. The extent
to which chronic TCEP exposure may be misclassified using a single spot urine concen-
tration of BCEP is uncertain. The authors did not discuss the half-life of urinary BCEP
or include measures of the parent compound along with the metabolite (Wang et al 2020
HEROID 7276658; Hou et al 10143372).
Continued on next page ...
Page 101 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 5 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 5:
Metric 6:
Exposure Levels
Temporality
Medium Exposure to TCEP was evaluated based on urinary BCEP concentrations, analyzed
continuously. BCEP was detectable in 85.31% of urine samples, and values below LOD
were imputed as the LOD divided by the square root of 2. No concerns about exposure
distribution (BCEP median [25th-75th percentile]: 0.51 ug/L [0.21, 1.03]). Urinary
creatinine was included in models to account for urine dilution.
Medium Urine samples used to estimate OPFR exposure and spirometry used to characterize the
respiratory function outcomes were collected during the same NHANES study visit.
While temporality cannot be established due to the cross-sectional design given the
short half-life (<30 days) of the exposure biomarker, there is no evidence of reverse
causation.
Domain 3: Outcome Assessment
Metric 7:
Outcome Measurement or
Characterization
Metric 8: Reporting Bias
High Spirometry testing procedures were performed according to the recommendations of the
American Thoracic Society (ATS). Measurements included forced expiratory volume
first second (FEV1), forced vital capacity (FVC), peak expiratory flowrate (PEF), forced
expiratory flow at 25-75% of the FVC (FEF25-75%), and FEV1/FVC. Measurements
were performed in a standing position (except for physical limitation). "A maximum of
eight spirometry tests were conducted to obtain three acceptable curves. The two high-
est values for FEV1 and FVC were used for quality ratings." Only those subjects with
FEV1 and FVC measurements rated as A or B (according to ATS data collection stan-
dards) were included. Further details can be found in the NHANES Respiratory Health
Spirometry Procedures Manual. Blinding was not mentioned, however, spirometry data
was collected at the same visit as urine was collected, indirectly indicating spirometry
technicians would be unaware of exposure status.
High All outcomes outlined in the abstract, introduction, and methods were provided in the
results or supplementary materials. Risk estimates were reported with p-values and
confidence intervals.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment Medium Study authors provided a simple model adjusted for age, sex, and race (model 1) and
a model additionally adjusted for BMI, serum cotinine, smoking status, physical ac-
tivity, family poverty/income ratio, educational level, and urinary creatinine. Study
authors conducted a stratified analysis to evaluate interaction by smoking status (never,
former, current). Examining the robustness of findings in sensitivity analyses was not
discussed (e.g. potential influence of participants with diagnoses respiratory illness such
as asthma).
Continued on next page ...
Page 102 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 5 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 10:
Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium Covariates were collected by trained interviewers during NHANES study visits; these
included self-reported sociodemographic measures. Complete information on collection
covariates can be found in the NHANES (2011-2012) Procedure Manuals. There was no
evidence of covariates with inadequate validity.
Medium Other OPFR metabolites were measured in urine. However, correlations among metabo-
lites were not provided, and models did not adjust for other metabolites. There was no
direct evidence of confounding by these or other co-exposures.
Domain 5: Analysis
Metric
12:
Study Design and Methods
High
The study design was appropriate to examine the research question (i.e., the association
between exposure to TCEP and pulmonary function). Appropriate statistical methods
(linear regression models) were used to evaluate the exposure-outcome relationship.
Metric
13:
Statistical Power
Medium
There was variability in exposure, and the number of participants (n=987) was likely
sufficient to detect an effect.
Metric
14:
Reproducibility of Analyses
Medium
The description of the analysis is sufficient to understand precisely what has been done
and to be conceptually reproducible with access to the analytic data.
Metric
15:
Statistical Analysis
High
Linear regression models were adequately described. Study authors noted that exposure
data was natural log transformed to normalize the distribution. Minor potential concerns
include that the non-linearity of dose-response relationships and sex-specific effects
were not discussed.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Medium
N/A
Medium
Medium
Medium
Supplemental material (Table SI) and the Pubchem dashboard indicate BCEP is the
direct metabolite of TCEP. No concerns about accuracy or precision of metabolite mea-
surements in urine. However, the extent to which a single spot urine BCEP accurately
classified TCEP exposure in non-occupationally exposed populations is not fully estab-
lished.
No biomarkers of effect were assessed.
The LOD (0.10 ug/L) was low enough to detect chemicals in a sufficient percentage of
the samples (85.31%) to address the research question, and the method was adequately
described.
Samples were collected during an NHANES study visit and documentation on storage
is provided in the NHANES Laboratory Procedures Manual. There was no information
provided on biomarker stability, but no evidence for concern was documented.
There is no specific information on sample contamination, but study authors note that
samples were collected "according to the analytic guidelines of the NHANES." Specific
information may be found in the NHANES Laboratory Procedures Manual.
Continued on next page ...
Page 103 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11365039 Table: 5 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Zhu, H., Zhang, H., Lu, K., Yang, S., Tang, X., Zhou, M., Sun, G., Zhang, Z., Chu, H. (2022). Chlorinated Organophosphate Flame Retardants Impair the
Lung Function via the IL-6/IAK/STAT Signaling Pathway. Environmental Science & Technology 56(24): 17858-17869.
Lung/Respiratory-Spirometry measures: Forced expiratory volume first second (FEV1), Forced vital capacity (FVC), Peak expiratory flow (PEF), Forced
expiratory flow at 25-75% of FVC (FEF25-75%), FEV1/FVC
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCEP)
No linked references.
11365039
Domain
Metric
Rating
Comments
Metric 21:
Metric 22:
Method Requirements
Matrix Adjustment
High
Medium
Samples were analyzed using automated off-line solid phase extraction, reversed-phase
high-performance liquid chromatography followed by isotope dilution-electrospray
ionization tandem mass spectrometry.
Model 1 adjusted for age, sex and race but excluded urinary creatinine, while Model 2
provides additionally adjusted for urinary creatinine and several other covariates. Asso-
ciations with and without creatinine were shown. However, the influence of creatinine
adjustment vs other confounding cannot be isolated.
Additional Comments:
This was a well-conducted study based on available NHANES data in adults, using the 2011-2012 cycle. There were no large concerns with participant
selection or methods relating to exposure measurement, outcome ascertainment, and statistical analysis. Minor concerns include that temporality cannot
be fully established due to the cross-sectional design, and that co-exposure confounding was not explored.
Overall Quality Determination
Medium
Page 104 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 1 of 4
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2:
Attrition
Metric 3: Comparison Group
High This prospective birth cohort study used data from the Health Outcomes and Measures
of the Environment (HOME) study, which was conducted in the Greater Cincinnati
Metropolitan Area, Ohio. Pregnant women were recruited from March 2003-January
2006 Participants were recruited "using the medical scheduling systems of nine prenatal
practices affiliated with three hospitals," with eligibility determined "using clinic records
and phone interviews with women" (Braun et al., 2017; HEROID: 6749104). Inclusion
criteria were: 1) age >18 years, 2) 13-19 weeks pregnancy, 3) residing in a home built
in or before 1978, 5) fluent in English, 6) planning to live in the study area for the next
year, 7) planning to continue prenatal care and deliver at the participating hospitals.
Exclusion criteria were: living in a mobile or trailer home, were on medications for thy-
roid or seizure disorders, or diagnosed with bipolar disorder, schizophrenia, diabetes,
or cancer requiring radiation or chemotherapy. The participation rate at each step was
reported in Braun et al. 2017. From a sampling frame of 8878 women, 1263 were eli-
gible and 340 women were included in the final analytic sample. The study compares
those included in the original cohort (n=389) and those finally included in the present
analysis (n=340) and reports that maternal characteristics were comparable between the
two groups. There is no evidence to suggest substantial selection bias and key elements
of the study design were reported.
Medium Details on attrition are available in Braun et al. 2017 (HEROID: 6749104). Of 468
women enrolled, 67 dropped out before delivery due to concerns about the time com-
mitment, having family members unwilling to participate, or loss of contact. 389 women
delivered singleton births, but only 340 were included in the final analysis for having
no congenital malformations, having one spot urine sample quantified for OPE metabo-
lites, and having neonatal anthropometry abstracted from medical records. While it is
not clear why these measures were not available for all 389 women, exclusion from the
analysis sample is appropriate and there is no evidence that any attrition in the study was
related to exposure and outcome.
High The study reports on demographic characteristics and reports whether differences in
characteristics were statistically significant by exposure and outcome variables. While
significant differences were reported, these factors were considered as potential covari-
ates in statistical analysis. There is no evidence to suggest that participants were selected
differently by outcome or exposure group.
Domain 2: Exposure Characterization
Continued on next page .
Page 105 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 1 of 4
... continued from previous page
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6: Temporality
High Exposure was assessed using an established biomarker of TCEP exposure (the urinary
metabolite BCEP) in urine samples collected from mothers twice during pregnancy at
an average of 16 ± 2 and 26 ± 3 weeks of gestation. BCEP was quantified using high-
performance liquid chromatography with tandem mass spectrometry. LOD information
is reported. Overall, exposure to TCEP was well-conducted and the use of two measure-
ments increases the certainty in exposure concentrations across pregnancy.
Medium The range of exposure is adequate, with no evidence of insufficient variability to esti-
mate associations. BCEP was characterized into tertiles and also analyzed as continuous
variable.
Medium The study includes measurements of urinary metabolites at two time points: 16 and 26
weeks, for pregnancy outcomes (gestational age and neonatal anthropometric measures).
Temporality is therefore clearly established, with exposure preceding outcome. While
the exposure is measured during pregnancy both time points captured are during the sec-
ond trimester, meaning the first trimester, a relevant time period and critical window for
fetal development was missed. Additionally, it is unclear whether the second trimester is
the most etiologically relevant time window for developmental effects of BCEP.
Domain 3: Outcome Assessment
Metric 7: Outcome Measurement or
Characterization
Metric 8: Reporting Bias
Medium Medical charts were used to obtain gestational age and infant anthropometric parameters
at birth (birth weight (g), length (cm), and head circumference (cm). Standardized birth
weight, length, and head circumference z-scores were calculated using values from the
2010 Olsen growth charts. Ponderal index (PI) was calculated as lOOx (weight/length).
One limitation is that gestational age was estimated by last menstrual period (LMP) for
330 participants, but by other measures (ultrasound (n = 7) or Ballard scores (n = 2)
for very limited cases). While it is a widely used method to estimate gestational age,
LMP has limitations as it relies on participant recall and also assumes regular 28-day
menstrual cycles with conception occurring on the 14th day.Preterm birth was defined as
birth prior to 37 completed weeks of gestation.
Medium All measured outcomes were analyzed and reported as described in the study aims.
Effect estimates are reported with confidence intervals. While the number of ex-
posed/unexposed are not reported in each analysis, the number of observations used
are reported consistently for each analysis.
Domain 4: Potential Confounding / Variability Control
Continued on next page ...
Page 106 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 1 of 4
... continued from previous page
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High Potential covariates were identified a priori through use of a directed acyclic graph
(DAG). Considered covariates included household income, infant sex, marital status,
maternal age, maternal blood lead, maternal serum cotinine, maternal education, mater-
nal race, parity, pre-pregnancy BMI. Effect modification by sex was also explored. Key
covariates were thus considered, and potential residual confounding is unlikely to have a
significant impact on effect estimates.
Medium The exact methodology for obtaining information on covariates was not described but
given the description of other methods the use of questionnaires or interviews is likely.
Braun et al. 2017 (HEROID: 6749104) also reports that clinical information was ab-
stracted from medical records.
Medium Other OPEs were examined, including the urinary metabolites BDCIPP, DNBP, and
DPHP. Models accounted for each OPE metabolite separately.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
High This cohort study analyzed associations between gestational exposure to BCEP with
infant anthropometric measurements. The use of a cohort study design was appropriate
for the research question given that gestational exposure was measured and participants
were prospectively followed for pregnancy outcomes. Appropriate statistical analy-
ses were used to account for exposure measurement twice during pregnancy, including
multiple informant models (two exposure windows as informants) using non-standard
generalized estimating equations (GEE) to examine whether OPE concentrations in dif-
ferent windows was related to pregnancy outcomes, as well as using a modified Poisson
regression with robust error variance combined with multiple informant models to exam-
ine period-specific risk ratios (RR) of individual OPE metabolites with preterm birth.
Medium The analysis sample included 340 mother-infant dyads, which is likely a large enough
sample size to detect an effect.
Medium Analysis methods were clearly described and conceptually reproducible.
Continued on next page ...
Page 107 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 1 of 4
... continued from previous page
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Descriptive data were provided for participant characteristics as well as for all outcomes
by participant characteristics. All models included adjustment for covariates, as well
as stratified analyses for potential mediators, such as infant sex. Appropriate statisti-
cal analyses were used to account for exposure measurement twice during pregnancy,
including multiple informant models (two exposure windows as informants) using non-
standard generalized estimating equations (GEE), as well as using a modified Poisson
regression with robust error variance combined with multiple informant models to ex-
amine period-specific risk ratios (RR) of individual OPE metabolites with preterm birth.
Cox proportional hazards regression analyses with gestational age as the underlying
time scale were used to estimate hazard ratios (HRs) assessing the occurrence of preterm
birth according to OPE metabolite concentrations.BCEP was analyzed using loglO
transformed continuous variables or tertiles.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Metric 19:
Effect Biomarker
Method Sensitivity
Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High
N/A
Medium
Medium
Medium
High
Medium
BCEP is derived from parent chemical TCEP and is an appropriate biomarker of expo-
sure.
No biomarkers of effect were assessed.
The method detection limit was provided for TCEP exposure: 0.1 jUg/L.
Samples were stored at or below -20 °C until further analysis. They were shipped
overnight on dry ice for quantification of OPEs. There is no data on stability, but no
evidence of degradation and quality control methods are detailed.
Maternal urine samples were collected in polypropylene specimen cups. There is no
information included about contamination.However, the CDC laboratory that conducted
the analysis is certified by CLIA and quality control methods are detailed. There was no
evidence to suggest contamination problems.
The laboratory methods described appeared to be appropriate. Samples were analyzed
using reversed-phase high-performance liquid chromatography, and detection by isotope
dilution-electrospray ionization tandem mass spectrometry.
BCEP concentrations were adjusted for specific gravity: specific gravity standardized
concentrations were calculated to account for hydration status during pregnancy. How-
ever, study only provides results for matrix-adjusted concentrations.
Continued on next page ...
Page 108 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 1 of 4
... continued from previous page
Study Citation: Yang, W. L., Braun, J. M., Vuong, A. M., Percy, Z., Xu, Y. Y., Xie, C. C., Deka, R., Calafat, A. M., Ospina, M., Burris, H. H., Yolton, K., Cecil, K. M.,
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Health Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
Linked HERO ID(s): No linked references.
HERO ID: 11577121
Domain Metric Rating Comments
Additional Comments: This cohort (340 mother-infant dyads) study in the Greater Cincinnati Area (Ohio, US) analyzed the relationship between gestational exposure to bis(2-
chloroethyl) phosphate (BCEP) and gestational age at birth and newborn anthropometric measures.The study found an association between BCEP and
several outcomes: increased BCEP concentrations in maternal urine at 16 weeks or 26 weeks of gestation were associated with longer gestation and
reduced risk of preterm birth. BCEP was also negatively associated with weight and length z-scores at birth in females. However, the associations
were not statistically significant after adjustment for multiple comparisons. Strengths of this analysis include exposure measurements at two time points
during pregnancy, to examine windows of susceptibility (although both occurred in the second trimester), as well as extensive data on covariates. Another
limitation is short-lived BCEP in urine samples., although this is a well-established biomarker.
Overall Quality Determination High
Page 109 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 2 of 4
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2:
Attrition
Metric 3: Comparison Group
High This prospective birth cohort study used data from the Health Outcomes and Measures
of the Environment (HOME) study, which was conducted in the Greater Cincinnati
Metropolitan Area, Ohio. Pregnant women were recruited from March 2003-January
2006 Participants were recruited "using the medical scheduling systems of nine prenatal
practices affiliated with three hospitals," with eligibility determined "using clinic records
and phone interviews with women" (Braun et al., 2017; HEROID: 6749104). Inclusion
criteria were: 1) age >18 years, 2) 13-19 weeks pregnancy, 3) residing in a home built
in or before 1978, 5) fluent in English, 6) planning to live in the study area for the next
year, 7) planning to continue prenatal care and deliver at the participating hospitals.
Exclusion criteria were: living in a mobile or trailer home, were on medications for thy-
roid or seizure disorders, or diagnosed with bipolar disorder, schizophrenia, diabetes,
or cancer requiring radiation or chemotherapy. The participation rate at each step was
reported in Braun et al. 2017. From a sampling frame of 8878 women, 1263 were eli-
gible and 340 women were included in the final analytic sample. The study compares
those included in the original cohort (n=389) and those finally included in the present
analysis (n=340) and reports that maternal characteristics were comparable between the
two groups. There is no evidence to suggest substantial selection bias and key elements
of the study design were reported.
Medium Details on attrition are available in Braun et al. 2017 (HEROID: 6749104). Of 468
women enrolled, 67 dropped out before delivery due to concerns about the time com-
mitment, having family members unwilling to participate, or loss of contact. 389 women
delivered singleton births, but only 340 were included in the final analysis for having
no congenital malformations, having one spot urine sample quantified for OPE metabo-
lites, and having neonatal anthropometry abstracted from medical records. While it is
not clear why these measures were not available for all 389 women, exclusion from the
analysis sample is appropriate and there is no evidence that any attrition in the study was
related to exposure and outcome.
High The study reports on demographic characteristics and reports whether differences in
characteristics were statistically significant by exposure and outcome variables. While
significant differences were reported, these factors were considered as potential covari-
ates in statistical analysis. There is no evidence to suggest that participants were selected
differently by outcome or exposure group.
Domain 2: Exposure Characterization
Continued on next page .
Page 110 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 2 of 4
... continued from previous page
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6: Temporality
High Exposure was assessed using an established biomarker of TCEP exposure (the urinary
metabolite BCEP) in urine samples collected from mothers twice during pregnancy at
an average of 16 ± 2 and 26 ± 3 weeks of gestation. BCEP was quantified using high-
performance liquid chromatography with tandem mass spectrometry. LOD information
is reported. Overall, exposure to TCEP was well-conducted and the use of two measure-
ments increases the certainty in exposure concentrations across pregnancy.
Medium The range of exposure is adequate, with no evidence of insufficient variability to esti-
mate associations. BCEP was characterized into tertiles and also analyzed as continuous
variable.
Medium The study includes measurements of urinary metabolites at two time points: 16 and 26
weeks, for pregnancy outcomes (gestational age and neonatal anthropometric measures).
Temporality is therefore clearly established, with exposure preceding outcome. While
the exposure is measured during pregnancy both time points captured are during the sec-
ond trimester, meaning the first trimester, a relevant time period and critical window for
fetal development was missed. Additionally, it is unclear whether the second trimester is
the most etiologically relevant time window for developmental effects of BCEP.
Domain 3: Outcome Assessment
Metric 7: Outcome Measurement or
Characterization
Metric 8: Reporting Bias
Medium Medical charts were used to obtain gestational age and infant anthropometric parameters
at birth (birth weight (g), length (cm), and head circumference (cm). Standardized birth
weight, length, and head circumference z-scores were calculated using values from the
2010 Olsen growth charts. Ponderal index (PI) was calculated as lOOx (weight/length).
One limitation is that gestational age was estimated by last menstrual period (LMP) for
330 participants, but by other measures (ultrasound (n = 7) or Ballard scores (n = 2)
for very limited cases). While it is a widely used method to estimate gestational age,
LMP has limitations as it relies on participant recall and also assumes regular 28-day
menstrual cycles with conception occurring on the 14th day.Preterm birth was defined as
birth prior to 37 completed weeks of gestation.
Medium All measured outcomes were analyzed and reported as described in the study aims.
Effect estimates are reported with confidence intervals. While the number of ex-
posed/unexposed are not reported in each analysis, the number of observations used
are reported consistently for each analysis.
Domain 4: Potential Confounding / Variability Control
Continued on next page ...
Page 111 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 2 of 4
... continued from previous page
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High Potential covariates were identified a priori through use of a directed acyclic graph
(DAG). Considered covariates included household income, infant sex, marital status,
maternal age, maternal blood lead, maternal serum cotinine, maternal education, mater-
nal race, parity, pre-pregnancy BMI. Effect modification by sex was also explored. Key
covariates were thus considered, and potential residual confounding is unlikely to have a
significant impact on effect estimates.
Medium The exact methodology for obtaining information on covariates was not described but
given the description of other methods the use of questionnaires or interviews is likely.
Braun et al. 2017 (HEROID: 6749104) also reports that clinical information was ab-
stracted from medical records.
Medium Other OPEs were examined, including the urinary metabolites BDCIPP, DNBP, and
DPHP. Models accounted for each OPE metabolite separately.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
High This cohort study analyzed associations between gestational exposure to BCEP with
infant anthropometric measurements. The use of a cohort study design was appropriate
for the research question given that gestational exposure was measured and participants
were prospectively followed for pregnancy outcomes. Appropriate statistical analy-
ses were used to account for exposure measurement twice during pregnancy, including
multiple informant models (two exposure windows as informants) using non-standard
generalized estimating equations (GEE) to examine whether OPE concentrations in dif-
ferent windows was related to pregnancy outcomes, as well as using a modified Poisson
regression with robust error variance combined with multiple informant models to exam-
ine period-specific risk ratios (RR) of individual OPE metabolites with preterm birth.
Medium The analysis sample included 340 mother-infant dyads, which is likely a large enough
sample size to detect an effect.
Medium Analysis methods were clearly described and conceptually reproducible.
Continued on next page ...
Page 112 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 2 of 4
... continued from previous page
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Descriptive data were provided for participant characteristics as well as for all outcomes
by participant characteristics. All models included adjustment for covariates, as well
as stratified analyses for potential mediators, such as infant sex. Appropriate statisti-
cal analyses were used to account for exposure measurement twice during pregnancy,
including multiple informant models (two exposure windows as informants) using non-
standard generalized estimating equations (GEE), as well as using a modified Poisson
regression with robust error variance combined with multiple informant models to ex-
amine period-specific risk ratios (RR) of individual OPE metabolites with preterm birth.
Cox proportional hazards regression analyses with gestational age as the underlying
time scale were used to estimate hazard ratios (HRs) assessing the occurrence of preterm
birth according to OPE metabolite concentrations.BCEP was analyzed using loglO
transformed continuous variables or tertiles.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Metric 19:
Effect Biomarker
Method Sensitivity
Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High
N/A
Medium
Medium
Medium
High
Medium
BCEP is derived from parent chemical TCEP and is an appropriate biomarker of expo-
sure.
No biomarkers of effect were assessed.
The method detection limit was provided for TCEP exposure: 0.1 jUg/L.
Samples were stored at or below -20 °C until further analysis. They were shipped
overnight on dry ice for quantification of OPEs. There is no data on stability, but no
evidence of degradation and quality control methods are detailed.
Maternal urine samples were collected in polypropylene specimen cups. There is no
information included about contamination.However, the CDC laboratory that conducted
the analysis is certified by CLIA and quality control methods are detailed. There was no
evidence to suggest contamination problems.
The laboratory methods described appeared to be appropriate. Samples were analyzed
using reversed-phase high-performance liquid chromatography, and detection by isotope
dilution-electrospray ionization tandem mass spectrometry.
BCEP concentrations were adjusted for specific gravity: specific gravity standardized
concentrations were calculated to account for hydration status during pregnancy. How-
ever, study only provides results for matrix-adjusted concentrations.
Continued on next page ...
Page 113 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 2 of 4
... continued from previous page
Study Citation: Yang, W. L., Braun, J. M., Vuong, A. M., Percy, Z., Xu, Y. Y., Xie, C. C., Deka, R., Calafat, A. M., Ospina, M., Burris, H. H., Yolton, K., Cecil, K. M.,
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Health Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
Linked HERO ID(s): No linked references.
HERO ID: 11577121
Domain Metric Rating Comments
Additional Comments: This cohort (340 mother-infant dyads) study in the Greater Cincinnati Area (Ohio, US) analyzed the relationship between gestational exposure to bis(2-
chloroethyl) phosphate (BCEP) and gestational age at birth and newborn anthropometric measures.The study found an association between BCEP and
several outcomes: increased BCEP concentrations in maternal urine at 16 weeks or 26 weeks of gestation were associated with longer gestation and
reduced risk of preterm birth. BCEP was also negatively associated with weight and length z-scores at birth in females. However, the associations
were not statistically significant after adjustment for multiple comparisons. Strengths of this analysis include exposure measurements at two time points
during pregnancy, to examine windows of susceptibility (although both occurred in the second trimester), as well as extensive data on covariates. Another
limitation is short-lived BCEP in urine samples., although this is a well-established biomarker.
Overall Quality Determination High
Page 114 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 3 of 4
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2:
Attrition
Metric 3: Comparison Group
High This prospective birth cohort study used data from the Health Outcomes and Measures
of the Environment (HOME) study, which was conducted in the Greater Cincinnati
Metropolitan Area, Ohio. Pregnant women were recruited from March 2003-January
2006 Participants were recruited "using the medical scheduling systems of nine prenatal
practices affiliated with three hospitals," with eligibility determined "using clinic records
and phone interviews with women" (Braun et al., 2017; HEROID: 6749104). Inclusion
criteria were: 1) age >18 years, 2) 13-19 weeks pregnancy, 3) residing in a home built
in or before 1978, 5) fluent in English, 6) planning to live in the study area for the next
year, 7) planning to continue prenatal care and deliver at the participating hospitals.
Exclusion criteria were: living in a mobile or trailer home, were on medications for thy-
roid or seizure disorders, or diagnosed with bipolar disorder, schizophrenia, diabetes,
or cancer requiring radiation or chemotherapy. The participation rate at each step was
reported in Braun et al. 2017. From a sampling frame of 8878 women, 1263 were eli-
gible and 340 women were included in the final analytic sample. The study compares
those included in the original cohort (n=389) and those finally included in the present
analysis (n=340) and reports that maternal characteristics were comparable between the
two groups. There is no evidence to suggest substantial selection bias and key elements
of the study design were reported.
Medium Details on attrition are available in Braun et al. 2017 (HEROID: 6749104). Of 468
women enrolled, 67 dropped out before delivery due to concerns about the time com-
mitment, having family members unwilling to participate, or loss of contact. 389 women
delivered singleton births, but only 340 were included in the final analysis for having
no congenital malformations, having one spot urine sample quantified for OPE metabo-
lites, and having neonatal anthropometry abstracted from medical records. While it is
not clear why these measures were not available for all 389 women, exclusion from the
analysis sample is appropriate and there is no evidence that any attrition in the study was
related to exposure and outcome.
High The study reports on demographic characteristics and reports whether differences in
characteristics were statistically significant by exposure and outcome variables. While
significant differences were reported, these factors were considered as potential covari-
ates in statistical analysis. There is no evidence to suggest that participants were selected
differently by outcome or exposure group.
Domain 2: Exposure Characterization
Continued on next page .
Page 115 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 3 of 4
... continued from previous page
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6: Temporality
High Exposure was assessed using an established biomarker of TCEP exposure (the urinary
metabolite BCEP) in urine samples collected from mothers twice during pregnancy at
an average of 16 ± 2 and 26 ± 3 weeks of gestation. BCEP was quantified using high-
performance liquid chromatography with tandem mass spectrometry. LOD information
is reported. Overall, exposure to TCEP was well-conducted and the use of two measure-
ments increases the certainty in exposure concentrations across pregnancy.
Medium The range of exposure is adequate, with no evidence of insufficient variability to esti-
mate associations. BCEP was characterized into tertiles and also analyzed as continuous
variable.
Medium The study includes measurements of urinary metabolites at two time points: 16 and 26
weeks, for pregnancy outcomes (gestational age and neonatal anthropometric measures).
Temporality is therefore clearly established, with exposure preceding outcome. While
the exposure is measured during pregnancy both time points captured are during the sec-
ond trimester, meaning the first trimester, a relevant time period and critical window for
fetal development was missed. Additionally, it is unclear whether the second trimester is
the most etiologically relevant time window for developmental effects of BCEP.
Domain 3: Outcome Assessment
Metric 7: Outcome Measurement or
Characterization
Metric 8: Reporting Bias
Medium Medical charts were used to obtain gestational age and infant anthropometric parameters
at birth (birth weight (g), length (cm), and head circumference (cm). Standardized birth
weight, length, and head circumference z-scores were calculated using values from the
2010 Olsen growth charts. Ponderal index (PI) was calculated as lOOx (weight/length).
One limitation is that gestational age was estimated by last menstrual period (LMP) for
330 participants, but by other measures (ultrasound (n = 7) or Ballard scores (n = 2)
for very limited cases). While it is a widely used method to estimate gestational age,
LMP has limitations as it relies on participant recall and also assumes regular 28-day
menstrual cycles with conception occurring on the 14th day.Preterm birth was defined as
birth prior to 37 completed weeks of gestation.
Medium All measured outcomes were analyzed and reported as described in the study aims.
Effect estimates are reported with confidence intervals. While the number of ex-
posed/unexposed are not reported in each analysis, the number of observations used
are reported consistently for each analysis.
Domain 4: Potential Confounding / Variability Control
Continued on next page ...
Page 116 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 3 of 4
... continued from previous page
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High Potential covariates were identified a priori through use of a directed acyclic graph
(DAG). Considered covariates included household income, infant sex, marital status,
maternal age, maternal blood lead, maternal serum cotinine, maternal education, mater-
nal race, parity, pre-pregnancy BMI. Effect modification by sex was also explored. Key
covariates were thus considered, and potential residual confounding is unlikely to have a
significant impact on effect estimates.
Medium The exact methodology for obtaining information on covariates was not described but
given the description of other methods the use of questionnaires or interviews is likely.
Braun et al. 2017 (HEROID: 6749104) also reports that clinical information was ab-
stracted from medical records.
Medium Other OPEs were examined, including the urinary metabolites BDCIPP, DNBP, and
DPHP. Models accounted for each OPE metabolite separately.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
High This cohort study analyzed associations between gestational exposure to BCEP with
infant anthropometric measurements. The use of a cohort study design was appropriate
for the research question given that gestational exposure was measured and participants
were prospectively followed for pregnancy outcomes. Appropriate statistical analy-
ses were used to account for exposure measurement twice during pregnancy, including
multiple informant models (two exposure windows as informants) using non-standard
generalized estimating equations (GEE) to examine whether OPE concentrations in dif-
ferent windows was related to pregnancy outcomes, as well as using a modified Poisson
regression with robust error variance combined with multiple informant models to exam-
ine period-specific risk ratios (RR) of individual OPE metabolites with preterm birth.
Medium The analysis sample included 340 mother-infant dyads, which is likely a large enough
sample size to detect an effect.
Medium Analysis methods were clearly described and conceptually reproducible.
Continued on next page ...
Page 117 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 3 of 4
... continued from previous page
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Descriptive data were provided for participant characteristics as well as for all outcomes
by participant characteristics. All models included adjustment for covariates, as well
as stratified analyses for potential mediators, such as infant sex. Appropriate statisti-
cal analyses were used to account for exposure measurement twice during pregnancy,
including multiple informant models (two exposure windows as informants) using non-
standard generalized estimating equations (GEE), as well as using a modified Poisson
regression with robust error variance combined with multiple informant models to ex-
amine period-specific risk ratios (RR) of individual OPE metabolites with preterm birth.
Cox proportional hazards regression analyses with gestational age as the underlying
time scale were used to estimate hazard ratios (HRs) assessing the occurrence of preterm
birth according to OPE metabolite concentrations.BCEP was analyzed using loglO
transformed continuous variables or tertiles.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Metric 19:
Effect Biomarker
Method Sensitivity
Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High
N/A
Medium
Medium
Medium
High
Medium
BCEP is derived from parent chemical TCEP and is an appropriate biomarker of expo-
sure.
No biomarkers of effect were assessed.
The method detection limit was provided for TCEP exposure: 0.1 jUg/L.
Samples were stored at or below -20 °C until further analysis. They were shipped
overnight on dry ice for quantification of OPEs. There is no data on stability, but no
evidence of degradation and quality control methods are detailed.
Maternal urine samples were collected in polypropylene specimen cups. There is no
information included about contamination.However, the CDC laboratory that conducted
the analysis is certified by CLIA and quality control methods are detailed. There was no
evidence to suggest contamination problems.
The laboratory methods described appeared to be appropriate. Samples were analyzed
using reversed-phase high-performance liquid chromatography, and detection by isotope
dilution-electrospray ionization tandem mass spectrometry.
BCEP concentrations were adjusted for specific gravity: specific gravity standardized
concentrations were calculated to account for hydration status during pregnancy. How-
ever, study only provides results for matrix-adjusted concentrations.
Continued on next page ...
Page 118 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 3 of 4
... continued from previous page
Study Citation: Yang, W. L., Braun, J. M., Vuong, A. M., Percy, Z., Xu, Y. Y., Xie, C. C., Deka, R., Calafat, A. M., Ospina, M., Burris, H. H., Yolton, K., Cecil, K. M.,
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Health Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
Linked HERO ID(s): No linked references.
HERO ID: 11577121
Domain Metric Rating Comments
Additional Comments: This cohort (340 mother-infant dyads) study in the Greater Cincinnati Area (Ohio, US) analyzed the relationship between gestational exposure to bis(2-
chloroethyl) phosphate (BCEP) and gestational age at birth and newborn anthropometric measures.The study found an association between BCEP and
several outcomes: increased BCEP concentrations in maternal urine at 16 weeks or 26 weeks of gestation were associated with longer gestation and
reduced risk of preterm birth. BCEP was also negatively associated with weight and length z-scores at birth in females. However, the associations
were not statistically significant after adjustment for multiple comparisons. Strengths of this analysis include exposure measurements at two time points
during pregnancy, to examine windows of susceptibility (although both occurred in the second trimester), as well as extensive data on covariates. Another
limitation is short-lived BCEP in urine samples., although this is a well-established biomarker.
Overall Quality Determination High
Page 119 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 4 of 4
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2:
Attrition
Metric 3: Comparison Group
High This prospective birth cohort study used data from the Health Outcomes and Measures
of the Environment (HOME) study, which was conducted in the Greater Cincinnati
Metropolitan Area, Ohio. Pregnant women were recruited from March 2003-January
2006 Participants were recruited "using the medical scheduling systems of nine prenatal
practices affiliated with three hospitals," with eligibility determined "using clinic records
and phone interviews with women" (Braun et al., 2017; HEROID: 6749104). Inclusion
criteria were: 1) age >18 years, 2) 13-19 weeks pregnancy, 3) residing in a home built
in or before 1978, 5) fluent in English, 6) planning to live in the study area for the next
year, 7) planning to continue prenatal care and deliver at the participating hospitals.
Exclusion criteria were: living in a mobile or trailer home, were on medications for thy-
roid or seizure disorders, or diagnosed with bipolar disorder, schizophrenia, diabetes,
or cancer requiring radiation or chemotherapy. The participation rate at each step was
reported in Braun et al. 2017. From a sampling frame of 8878 women, 1263 were eli-
gible and 340 women were included in the final analytic sample. The study compares
those included in the original cohort (n=389) and those finally included in the present
analysis (n=340) and reports that maternal characteristics were comparable between the
two groups. There is no evidence to suggest substantial selection bias and key elements
of the study design were reported.
Medium Details on attrition are available in Braun et al. 2017 (HEROID: 6749104). Of 468
women enrolled, 67 dropped out before delivery due to concerns about the time com-
mitment, having family members unwilling to participate, or loss of contact. 389 women
delivered singleton births, but only 340 were included in the final analysis for having
no congenital malformations, having one spot urine sample quantified for OPE metabo-
lites, and having neonatal anthropometry abstracted from medical records. While it is
not clear why these measures were not available for all 389 women, exclusion from the
analysis sample is appropriate and there is no evidence that any attrition in the study was
related to exposure and outcome.
High The study reports on demographic characteristics and reports whether differences in
characteristics were statistically significant by exposure and outcome variables. While
significant differences were reported, these factors were considered as potential covari-
ates in statistical analysis. There is no evidence to suggest that participants were selected
differently by outcome or exposure group.
Domain 2: Exposure Characterization
Continued on next page .
Page 120 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 4 of 4
... continued from previous page
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6: Temporality
High Exposure was assessed using an established biomarker of TCEP exposure (the urinary
metabolite BCEP) in urine samples collected from mothers twice during pregnancy at
an average of 16 ± 2 and 26 ± 3 weeks of gestation. BCEP was quantified using high-
performance liquid chromatography with tandem mass spectrometry. LOD information
is reported. Overall, exposure to TCEP was well-conducted and the use of two measure-
ments increases the certainty in exposure concentrations across pregnancy.
Medium The range of exposure is adequate, with no evidence of insufficient variability to esti-
mate associations. BCEP was characterized into tertiles and also analyzed as continuous
variable.
Medium The study includes measurements of urinary metabolites at two time points: 16 and 26
weeks, for pregnancy outcomes (gestational age and neonatal anthropometric measures).
Temporality is therefore clearly established, with exposure preceding outcome. While
the exposure is measured during pregnancy both time points captured are during the sec-
ond trimester, meaning the first trimester, a relevant time period and critical window for
fetal development was missed. Additionally, it is unclear whether the second trimester is
the most etiologically relevant time window for developmental effects of BCEP.
Domain 3: Outcome Assessment
Metric 7: Outcome Measurement or
Characterization
Metric 8: Reporting Bias
Medium Medical charts were used to obtain gestational age and infant anthropometric parameters
at birth (birth weight (g), length (cm), and head circumference (cm). Standardized birth
weight, length, and head circumference z-scores were calculated using values from the
2010 Olsen growth charts. Ponderal index (PI) was calculated as lOOx (weight/length).
One limitation is that gestational age was estimated by last menstrual period (LMP) for
330 participants, but by other measures (ultrasound (n = 7) or Ballard scores (n = 2)
for very limited cases). While it is a widely used method to estimate gestational age,
LMP has limitations as it relies on participant recall and also assumes regular 28-day
menstrual cycles with conception occurring on the 14th day.Preterm birth was defined as
birth prior to 37 completed weeks of gestation.
Medium All measured outcomes were analyzed and reported as described in the study aims.
Effect estimates are reported with confidence intervals. While the number of ex-
posed/unexposed are not reported in each analysis, the number of observations used
are reported consistently for each analysis.
Domain 4: Potential Confounding / Variability Control
Continued on next page ...
Page 121 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 4 of 4
... continued from previous page
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High Potential covariates were identified a priori through use of a directed acyclic graph
(DAG). Considered covariates included household income, infant sex, marital status,
maternal age, maternal blood lead, maternal serum cotinine, maternal education, mater-
nal race, parity, pre-pregnancy BMI. Effect modification by sex was also explored. Key
covariates were thus considered, and potential residual confounding is unlikely to have a
significant impact on effect estimates.
Medium The exact methodology for obtaining information on covariates was not described but
given the description of other methods the use of questionnaires or interviews is likely.
Braun et al. 2017 (HEROID: 6749104) also reports that clinical information was ab-
stracted from medical records.
Medium Other OPEs were examined, including the urinary metabolites BDCIPP, DNBP, and
DPHP. Models accounted for each OPE metabolite separately.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
High This cohort study analyzed associations between gestational exposure to BCEP with
infant anthropometric measurements. The use of a cohort study design was appropriate
for the research question given that gestational exposure was measured and participants
were prospectively followed for pregnancy outcomes. Appropriate statistical analy-
ses were used to account for exposure measurement twice during pregnancy, including
multiple informant models (two exposure windows as informants) using non-standard
generalized estimating equations (GEE) to examine whether OPE concentrations in dif-
ferent windows was related to pregnancy outcomes, as well as using a modified Poisson
regression with robust error variance combined with multiple informant models to exam-
ine period-specific risk ratios (RR) of individual OPE metabolites with preterm birth.
Medium The analysis sample included 340 mother-infant dyads, which is likely a large enough
sample size to detect an effect.
Medium Analysis methods were clearly described and conceptually reproducible.
Continued on next page ...
Page 122 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 4 of 4
... continued from previous page
Yang, W. L., Braun, J. M„ Vuong, A. M„ Percy, Z„ Xu, Y. Y„ Xie, C. C„ Deka, R„ Calafat, A. M„ Ospina, M„ Burris, H. H„ Yolton, K„ Cecil, K. M„
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
No linked references.
11577121
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Descriptive data were provided for participant characteristics as well as for all outcomes
by participant characteristics. All models included adjustment for covariates, as well
as stratified analyses for potential mediators, such as infant sex. Appropriate statisti-
cal analyses were used to account for exposure measurement twice during pregnancy,
including multiple informant models (two exposure windows as informants) using non-
standard generalized estimating equations (GEE), as well as using a modified Poisson
regression with robust error variance combined with multiple informant models to ex-
amine period-specific risk ratios (RR) of individual OPE metabolites with preterm birth.
Cox proportional hazards regression analyses with gestational age as the underlying
time scale were used to estimate hazard ratios (HRs) assessing the occurrence of preterm
birth according to OPE metabolite concentrations.BCEP was analyzed using loglO
transformed continuous variables or tertiles.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Metric 19:
Effect Biomarker
Method Sensitivity
Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High
N/A
Medium
Medium
Medium
High
Medium
BCEP is derived from parent chemical TCEP and is an appropriate biomarker of expo-
sure.
No biomarkers of effect were assessed.
The method detection limit was provided for TCEP exposure: 0.1 jUg/L.
Samples were stored at or below -20 °C until further analysis. They were shipped
overnight on dry ice for quantification of OPEs. There is no data on stability, but no
evidence of degradation and quality control methods are detailed.
Maternal urine samples were collected in polypropylene specimen cups. There is no
information included about contamination.However, the CDC laboratory that conducted
the analysis is certified by CLIA and quality control methods are detailed. There was no
evidence to suggest contamination problems.
The laboratory methods described appeared to be appropriate. Samples were analyzed
using reversed-phase high-performance liquid chromatography, and detection by isotope
dilution-electrospray ionization tandem mass spectrometry.
BCEP concentrations were adjusted for specific gravity: specific gravity standardized
concentrations were calculated to account for hydration status during pregnancy. How-
ever, study only provides results for matrix-adjusted concentrations.
Continued on next page ...
Page 123 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577121 Table: 4 of 4
... continued from previous page
Study Citation: Yang, W. L., Braun, J. M., Vuong, A. M., Percy, Z., Xu, Y. Y., Xie, C. C., Deka, R., Calafat, A. M., Ospina, M., Burris, H. H., Yolton, K., Cecil, K. M.,
Lanphear, B. P., Chen, A. M. (2022). Associations of gestational exposure to organophosphate esters with gestational age and neonatal anthropometric
measures: The HOME study*. Environmental Pollution 316(Part 1):120516.
Health Reproductive /Developmental-Gestational age at birth, newborn weight, length, head circumference, preterm birth, ponderal index
Outcome(s) and
Reported Health
Effect(s):
Chemical: Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: bis (2-chloroethyl) phosphate (BCEP)
Linked HERO ID(s): No linked references.
HERO ID: 11577121
Domain Metric Rating Comments
Additional Comments: This cohort (340 mother-infant dyads) study in the Greater Cincinnati Area (Ohio, US) analyzed the relationship between gestational exposure to bis(2-
chloroethyl) phosphate (BCEP) and gestational age at birth and newborn anthropometric measures.The study found an association between BCEP and
several outcomes: increased BCEP concentrations in maternal urine at 16 weeks or 26 weeks of gestation were associated with longer gestation and
reduced risk of preterm birth. BCEP was also negatively associated with weight and length z-scores at birth in females. However, the associations
were not statistically significant after adjustment for multiple comparisons. Strengths of this analysis include exposure measurements at two time points
during pregnancy, to examine windows of susceptibility (although both occurred in the second trimester), as well as extensive data on covariates. Another
limitation is short-lived BCEP in urine samples., although this is a well-established biomarker.
Overall Quality Determination High
Page 124 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 1 of 5
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2:
Attrition
High Participants in this study were all recruited from cohorts that were enrolled in the "En-
vironmental influences on Child Health Outcomes" (ECHO) program at the NIH, where
existing cohorts enrolled their participants into the ECHO program set of cohorts. All
cohorts included were birth cohorts with urine samples from pregnant mothers and mea-
sures of birth outcomes. Individual cohorts followed their own protocols prior to 2019,
but in late 2019 ECHO instituted a common protocol for cohorts to follow. Data for this
study was collected prior to initiation of the common protocol. In total 16 separate co-
horts contributed data to this study, with data collected and submitted to ECHO prior to
March 2022. Study details are available for each cohort, including the number of par-
ticipants included, a description of the sample, eligibility criteria, and their strategy for
obtaining birth outcome information. Of the ECHO cohort participants eligible based on
the availability of maternal urine samples for organophosphate ester (OPE) quantifica-
tion (n=12,873), n=7,048 (54.8%) participants had urinary OPE and dilution data. The
number of urine samples included and excluded from each cohort was provided. Partic-
ipants were excluded if they had no information on children (n=82), had multiple births
(n=10), or had missing outcome data (n=309). One additional child born with a gesta-
tional age of >42 weeks was also excluded, as the formula used for calculating birth
weight for gestational age z-scores did not accommodate gestations of that duration. In
total, the final sample was n=6,646 mother-child dyads. The study compared the charac-
teristics of participants included to those excluded; while no statistical analysis was done
across these two groups, there do not appear to be significant differences in characteris-
tics (e.g. socio-demographics, BMI, tobacco use) of included vs. excluded participants.
Overall, key study information for the overall ECHO program and for individual cohorts
was well-reported and indicated that significant selection bias was unlikely.
Medium The study indicates that 309 out of 7,048 mother-child dyads with urinary OPE mea-
sures had missing birth outcome data (gestational age, birth weight, biological sex at
birth) and were thus excluded from the study. Another 82 pregnancies had no available
child information, and 11 additional children were ineligible (multiple births, length
of gestation). The study does not provide details as to why some dyads were missing
outcome data, but there is no evidence that any cause would be related to their outcome
status.
Continued on next page ...
Page 125 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 1 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, L, Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 3: Comparison Group
Medium Since this study was a pooled analyses of multiple birth cohorts, participants were in-
herently drawn from different background populations and locations across the United
States. Key demographic and study design characteristics that may have differed across
cohorts were considered as potential covariates (e.g., maternal race/ethnicity, education,
age, marital status, along with timing, season, and year of sample collection). Clustering
by cohort was addressed in the statistical model. The authors did not discuss adjusting
for region or examining potential modification by cohort or region. Instead, to examine
the potential influence of heterogeneity across cohorts, the authors included a sensi-
tivity analysis using a "leave-one-cohort-out" approach to assess the influence of each
cohort. While there was generally little or no impact of leaving out any one cohort,
for the TCEP metabolite BCEP (alternative acronym BCETP) omitting one of the two
largest cohorts (CANDLE, predominantly African American mothers from Memphis,
TN) strengthened a positive association between low concentrations and birth weight
z-scores, which became statistically significant. While there was a change in magnitude
and statistical significance, there was no change in direction, and no evidence of bias
due to cohort heterogeneity.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure Medium In all participants, TCEP was measured via the urinary metabolite BCEP. Urine samples
were collected between 2007 and 2020. Measures used a single spot or first morning
urine samples collected from each participant primarily during the second (44.1%) and
third (55.5%) trimesters of pregnancy. BCEP is an appropriate biomarker of exposure
to TCEP. BCEP was quantified using high-performance liquid chromatography with
triple quadrupole mass spectrometry, and quality control measures are well-detailed.
The LOD was specified to be 0.02 ng/mL and the percentage of samples above the LOD
was specified to be 69%. Machine readings were used for values below LOD. Urinary
dilution was addressed using creatinine or specific gravity, depending on the cohort.
Exposure measures at only a single time point may not reflect the full scope of TCEP
exposure across pregnancy due to variability. In addition, other studies examining TCEP
biomarkers reported detecting unmetabolized TCEP in 37% of samples (e.g., Hou et al.
2020 HEROID 10143372); it is uncertain whether or to what extent excluding this factor
may introduce error. Though some error is likely, there was no evidence of important
error or bias in exposure estimates.
Continued on next page ...
Page 126 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 1 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 5:
Exposure Levels
Medium
While the 25th percentile of exposure was below the limit of detection, the median
value was 0.52 ng/mL and the 75th and 95th percentiles were 1.58 and 8.22 ng/mL,
respectively, which should allow for sufficient contrast compared to participants with
no detectable BCEP exposure. In statistical analyses, BCEP was categorized into three
groups: undetected, and then low or high detectable exposure depending on whether
measurements fell above or below the median of dilution-adjusted samples above the
LOD.
Temporality is established as exposure was measured during the second or third
trimester of pregnancy for outcomes that were measured at birth. However, it is un-
certain whether exposures during the second or third trimester reflect the etiologically
relevant time window for birth outcome effects in relation to TCEP exposure.
Metric 6:
Temporality
Medium
Domain 3: Outcome Assessment
Metric 7:
Outcome Measurement or
Characterization
Metric 8: Reporting Bias
Medium Outcomes collected prior to 2019 were ascertained using cohort-specific protocols; the
majority of cohorts used medical record abstraction for outcomes, with a small pro-
portion of parent reporting. Children born in or after 2019 had outcomes ascertained
following the universal ECHO protocol of assigning outcomes based on medical record
abstraction.All instances of birth weight were measured using medical records. Gesta-
tional age was largely abstracted from or estimated using data from medical records;
methods used by hospitals to estimate gestational age were not always specified. Some
cohorts specified using 1st or 2nd trimester ultrasound to calculate gestational age, us-
ing last menstrual period in cases of missing data. While there was some variation in
the estimation of gestational age, there was no evidence that this variation might signifi-
cantly impact effect estimates. Gestational age was further categorized as preterm (<37
weeks), early term (37-38 weeks), full term (39-40 weeks), and late/post-term (41-42
weeks). Sex-specific birth weight for gestational age z-scores were calculated based on
a United states reference population (Aris et al. 2019, not in HERO; PMID: 31201230).
Birth weight for gestational age was also categorized as small for gestational age (<10th
percentile) and large for gestational age (>90th percentile), as well as term low birth
weight (birth weight of <2,500 g at >=37 weeks).
Medium All specified outcomes and results that were outlined in the methods were reported in
the results. Effect estimates are presented with confidence intervals. Tabulations of
outcomes by level of exposure were not provided.
Continued on next page ...
Page 127 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 1 of 5
... continued from previous page
Study Citation:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10:
Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium Potential covariates were selected using a DAG to identify ad include confounders and
precision variables, and to exclude intermediates. Covariates included were maternal
race/ethnicity, maternal age at delivery, maternal education, maternal marital status,
maternal pre-pregnancy BMI, maternal smoking during pregnancy, parity, infant sex,
time of urine sample, season of urine sample, and year of urine sample. A potential
limitation is that the authors did not mention evaluating the influence of gestational
age adjustment on associations with size at birth. Though gestational age might be an
intermediate or collider, there was no direct evidence of important error or bias.
Medium Covariates were collected separately by each cohort within the ECHO program and
then harmonized by the ECHO Data Analysis Center. While it is not specified how
each of these covariates were measured, there is no evidence that insensitive instru-
ments/methods/measures were used.
Medium The study considered co-exposure to other organophosphate esters, including TPHP,
TBUP, TIBP, TDCPP, TBOEP, TCPP, TMPP, TEHP, and TPRP. These co-exposures
were measured via urinary metabolites at the same time and through the same process as
BCEP. Spearman correlations between BECP and these biomarkers were <0.20. These
co-exposures were considered in sensitivity analyses by simultaneously adjusting for
these OPEs, and results were similar.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
High The cohort study design is appropriate for assessing the association between prenatal
concentrations of TCEP metabolites in urine and birth outcomes. The use of linear and
logistic regression for continuous and categorical outcomes respectively is appropriate to
address the research question. Regressions were performed using generalized estimating
equations accounting for clustering at the cohort level, which is appropriate given the
fact that cohorts were combined into one group in this study. Potential non-linear associ-
ations were examined by categorizing BCEP into three groups of similar size.
Medium The number of participants in the study (n=6,646 mother-child dyads) is likely suffi-
ciently large to detect an effect.
Medium The description of the analysis is sufficient to understand precisely what has been done
and to be conceptually reproducible with access to the analytic data.
Continued on next page ...
Page 128 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 1 of 5
... continued from previous page
Study Citation:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Risk estimates were calculated using linear regression models for continuous outcomes
and multinomial logistic regression models for categorical outcomes. BCEP concentra-
tions were categorized, with values below LOD as the referent, and a lower and higher
exposure group defined by categorizing remaining samples at their median. Inclusion
of covariates was dependent on a DAG created to exclude variables that may be causal
intermediates. Covariates with <20% of missing values were addressed via multiple
imputation by chained equations. Sensitivity analyses are detailed, including sex stratifi-
cation, a leave-one-cohort-out approach, and a co-pollutant model.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High
N/A
Medium
Medium
Medium
Medium
Medium
Urinary concentrations of BCEP are a valid and appropriate biomarker of exposure to
TCEP.
No biomarkers of effect were assessed.
The limit of detection was specified to be 0.02 ng/mL. BCEP was detected in 69% of
samples. There was no evidence that the detection limit was inadequate.
The study reports that urine samples were shipped on dry ice to the Human Health Ex-
posure Analysis Resource (HHEAR) at the NYU Grossman School of Medicine. While
no other specific storage information is provided, there is no evidence that inappropriate
methods were used and the biomarker is not expected to have a high likelihood of stabil-
ity.
No information is provided potential contamination.
BCEP was quantified using high-performance liquid chromatography with triple
quadrupole mass spectrometry, and quality control (QC) measures are well-detailed.
One QC measure included the use of blinded duplicate samples (n=127 samples for
BCEP). For BCEP, 66% of duplicate pairs were both above LOD; the remaining pairs
were not concordant with respect to detection. Details on the magnitude of disparities in
estimated concentrations were not provided.
BCEP was adjusted for urinary dilution, measured as either specific gravity or creatinine
depending on the source cohort, using published methods. Results were presented only
for dilution-standardized exposure variables.
Continued on next page ...
Page 129 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 1 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Additional Comments:
This prospective study used data from 16 cohorts participating in the ECHO (Environmental influences on Child Health Outcomes) program to estimate
the association between maternal urinary organophosphate ester concentrations and two types of birth outcome: birth weight and duration of gestation.
Spot urine concentrations of BCEP from the second or third trimester were used to estimate exposure. There were no major concerns for bias. There
was some degree of uncertainty due to the use of only a single timepoint measurement of exposure and variability in ways in which gestational age was
assessed. Birth weight was analyzed using birth weight for gestational age z-scores continuously or classified as small or large size for gestational age, but
not in analyses that excluded gestational age adjustment. Among girls but not boys, compared to concentrations below detection, a high level of detectable
BCEP was associated with significantly shorter mean gestational age and increased odds of preterm birth. In the population overall, a low concentration
of detectable BCEP was associated with significantly reduced odds of late/post term birth; the association with high concentration was null. In addition,
overall, a high concentration of detectable BCEP was associated with significantly lower odds of small for gestational age (SGA) birth. Both low and high
detectable BCEP were associated with reduced odds of SGA among boys. Among girls, low-level detectable BCEP was associated with increased odds of
large for gestational age (LGA) birth. Findings from this study suggest that effects of BCEP may be sex-specific and that BCEP may reduce the length of
gestation particularly in girls but may not reduce birth weight relative to the duration of gestation.
Overall Quality Determination
Medium
Page 130 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 2 of 5
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2:
Attrition
High Participants in this study were all recruited from cohorts that were enrolled in the "En-
vironmental influences on Child Health Outcomes" (ECHO) program at the NIH, where
existing cohorts enrolled their participants into the ECHO program set of cohorts. All
cohorts included were birth cohorts with urine samples from pregnant mothers and mea-
sures of birth outcomes. Individual cohorts followed their own protocols prior to 2019,
but in late 2019 ECHO instituted a common protocol for cohorts to follow. Data for this
study was collected prior to initiation of the common protocol. In total 16 separate co-
horts contributed data to this study, with data collected and submitted to ECHO prior to
March 2022. Study details are available for each cohort, including the number of par-
ticipants included, a description of the sample, eligibility criteria, and their strategy for
obtaining birth outcome information. Of the ECHO cohort participants eligible based on
the availability of maternal urine samples for organophosphate ester (OPE) quantifica-
tion (n=12,873), n=7,048 (54.8%) participants had urinary OPE and dilution data. The
number of urine samples included and excluded from each cohort was provided. Partic-
ipants were excluded if they had no information on children (n=82), had multiple births
(n=10), or had missing outcome data (n=309). One additional child born with a gesta-
tional age of >42 weeks was also excluded, as the formula used for calculating birth
weight for gestational age z-scores did not accommodate gestations of that duration. In
total, the final sample was n=6,646 mother-child dyads. The study compared the charac-
teristics of participants included to those excluded; while no statistical analysis was done
across these two groups, there do not appear to be significant differences in characteris-
tics (e.g. socio-demographics, BMI, tobacco use) of included vs. excluded participants.
Overall, key study information for the overall ECHO program and for individual cohorts
was well-reported and indicated that significant selection bias was unlikely.
Medium The study indicates that 309 out of 7,048 mother-child dyads with urinary OPE mea-
sures had missing birth outcome data (gestational age, birth weight, biological sex at
birth) and were thus excluded from the study. Another 82 pregnancies had no available
child information, and 11 additional children were ineligible (multiple births, length
of gestation). The study does not provide details as to why some dyads were missing
outcome data, but there is no evidence that any cause would be related to their outcome
status.
Continued on next page ...
Page 131 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 2 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, L, Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 3: Comparison Group
Medium Since this study was a pooled analyses of multiple birth cohorts, participants were in-
herently drawn from different background populations and locations across the United
States. Key demographic and study design characteristics that may have differed across
cohorts were considered as potential covariates (e.g., maternal race/ethnicity, education,
age, marital status, along with timing, season, and year of sample collection). Clustering
by cohort was addressed in the statistical model. The authors did not discuss adjusting
for region or examining potential modification by cohort or region. Instead, to examine
the potential influence of heterogeneity across cohorts, the authors included a sensi-
tivity analysis using a "leave-one-cohort-out" approach to assess the influence of each
cohort. While there was generally little or no impact of leaving out any one cohort,
for the TCEP metabolite BCEP (alternative acronym BCETP) omitting one of the two
largest cohorts (CANDLE, predominantly African American mothers from Memphis,
TN) strengthened a positive association between low concentrations and birth weight
z-scores, which became statistically significant. While there was a change in magnitude
and statistical significance, there was no change in direction, and no evidence of bias
due to cohort heterogeneity.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure Medium In all participants, TCEP was measured via the urinary metabolite BCEP. Urine samples
were collected between 2007 and 2020. Measures used a single spot or first morning
urine samples collected from each participant primarily during the second (44.1%) and
third (55.5%) trimesters of pregnancy. BCEP is an appropriate biomarker of exposure
to TCEP. BCEP was quantified using high-performance liquid chromatography with
triple quadrupole mass spectrometry, and quality control measures are well-detailed.
The LOD was specified to be 0.02 ng/mL and the percentage of samples above the LOD
was specified to be 69%. Machine readings were used for values below LOD. Urinary
dilution was addressed using creatinine or specific gravity, depending on the cohort.
Exposure measures at only a single time point may not reflect the full scope of TCEP
exposure across pregnancy due to variability. In addition, other studies examining TCEP
biomarkers reported detecting unmetabolized TCEP in 37% of samples (e.g., Hou et al.
2020 HEROID 10143372); it is uncertain whether or to what extent excluding this factor
may introduce error. Though some error is likely, there was no evidence of important
error or bias in exposure estimates.
Continued on next page ...
Page 132 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 2 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 5:
Exposure Levels
Medium
While the 25th percentile of exposure was below the limit of detection, the median
value was 0.52 ng/mL and the 75th and 95th percentiles were 1.58 and 8.22 ng/mL,
respectively, which should allow for sufficient contrast compared to participants with
no detectable BCEP exposure. In statistical analyses, BCEP was categorized into three
groups: undetected, and then low or high detectable exposure depending on whether
measurements fell above or below the median of dilution-adjusted samples above the
LOD.
Temporality is established as exposure was measured during the second or third
trimester of pregnancy for outcomes that were measured at birth. However, it is un-
certain whether exposures during the second or third trimester reflect the etiologically
relevant time window for birth outcome effects in relation to TCEP exposure.
Metric 6:
Temporality
Medium
Domain 3: Outcome Assessment
Metric 7:
Outcome Measurement or
Characterization
Metric 8: Reporting Bias
Medium Outcomes collected prior to 2019 were ascertained using cohort-specific protocols; the
majority of cohorts used medical record abstraction for outcomes, with a small pro-
portion of parent reporting. Children born in or after 2019 had outcomes ascertained
following the universal ECHO protocol of assigning outcomes based on medical record
abstraction.All instances of birth weight were measured using medical records. Gesta-
tional age was largely abstracted from or estimated using data from medical records;
methods used by hospitals to estimate gestational age were not always specified. Some
cohorts specified using 1st or 2nd trimester ultrasound to calculate gestational age, us-
ing last menstrual period in cases of missing data. While there was some variation in
the estimation of gestational age, there was no evidence that this variation might signifi-
cantly impact effect estimates. Gestational age was further categorized as preterm (<37
weeks), early term (37-38 weeks), full term (39-40 weeks), and late/post-term (41-42
weeks). Sex-specific birth weight for gestational age z-scores were calculated based on
a United states reference population (Aris et al. 2019, not in HERO; PMID: 31201230).
Birth weight for gestational age was also categorized as small for gestational age (<10th
percentile) and large for gestational age (>90th percentile), as well as term low birth
weight (birth weight of <2,500 g at >=37 weeks).
Medium All specified outcomes and results that were outlined in the methods were reported in
the results. Effect estimates are presented with confidence intervals. Tabulations of
outcomes by level of exposure were not provided.
Continued on next page ...
Page 133 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 2 of 5
... continued from previous page
Study Citation:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10:
Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium Potential covariates were selected using a DAG to identify ad include confounders and
precision variables, and to exclude intermediates. Covariates included were maternal
race/ethnicity, maternal age at delivery, maternal education, maternal marital status,
maternal pre-pregnancy BMI, maternal smoking during pregnancy, parity, infant sex,
time of urine sample, season of urine sample, and year of urine sample. A potential
limitation is that the authors did not mention evaluating the influence of gestational
age adjustment on associations with size at birth. Though gestational age might be an
intermediate or collider, there was no direct evidence of important error or bias.
Medium Covariates were collected separately by each cohort within the ECHO program and
then harmonized by the ECHO Data Analysis Center. While it is not specified how
each of these covariates were measured, there is no evidence that insensitive instru-
ments/methods/measures were used.
Medium The study considered co-exposure to other organophosphate esters, including TPHP,
TBUP, TIBP, TDCPP, TBOEP, TCPP, TMPP, TEHP, and TPRP. These co-exposures
were measured via urinary metabolites at the same time and through the same process as
BCEP. Spearman correlations between BECP and these biomarkers were <0.20. These
co-exposures were considered in sensitivity analyses by simultaneously adjusting for
these OPEs, and results were similar.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
High The cohort study design is appropriate for assessing the association between prenatal
concentrations of TCEP metabolites in urine and birth outcomes. The use of linear and
logistic regression for continuous and categorical outcomes respectively is appropriate to
address the research question. Regressions were performed using generalized estimating
equations accounting for clustering at the cohort level, which is appropriate given the
fact that cohorts were combined into one group in this study. Potential non-linear associ-
ations were examined by categorizing BCEP into three groups of similar size.
Medium The number of participants in the study (n=6,646 mother-child dyads) is likely suffi-
ciently large to detect an effect.
Medium The description of the analysis is sufficient to understand precisely what has been done
and to be conceptually reproducible with access to the analytic data.
Continued on next page ...
Page 134 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 2 of 5
... continued from previous page
Study Citation:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Risk estimates were calculated using linear regression models for continuous outcomes
and multinomial logistic regression models for categorical outcomes. BCEP concentra-
tions were categorized, with values below LOD as the referent, and a lower and higher
exposure group defined by categorizing remaining samples at their median. Inclusion
of covariates was dependent on a DAG created to exclude variables that may be causal
intermediates. Covariates with <20% of missing values were addressed via multiple
imputation by chained equations. Sensitivity analyses are detailed, including sex stratifi-
cation, a leave-one-cohort-out approach, and a co-pollutant model.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High
N/A
Medium
Medium
Medium
Medium
Medium
Urinary concentrations of BCEP are a valid and appropriate biomarker of exposure to
TCEP.
No biomarkers of effect were assessed.
The limit of detection was specified to be 0.02 ng/mL. BCEP was detected in 69% of
samples. There was no evidence that the detection limit was inadequate.
The study reports that urine samples were shipped on dry ice to the Human Health Ex-
posure Analysis Resource (HHEAR) at the NYU Grossman School of Medicine. While
no other specific storage information is provided, there is no evidence that inappropriate
methods were used and the biomarker is not expected to have a high likelihood of stabil-
ity.
No information is provided potential contamination.
BCEP was quantified using high-performance liquid chromatography with triple
quadrupole mass spectrometry, and quality control (QC) measures are well-detailed.
One QC measure included the use of blinded duplicate samples (n=127 samples for
BCEP). For BCEP, 66% of duplicate pairs were both above LOD; the remaining pairs
were not concordant with respect to detection. Details on the magnitude of disparities in
estimated concentrations were not provided.
BCEP was adjusted for urinary dilution, measured as either specific gravity or creatinine
depending on the source cohort, using published methods. Results were presented only
for dilution-standardized exposure variables.
Continued on next page ...
Page 135 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 2 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Additional Comments:
This prospective study used data from 16 cohorts participating in the ECHO (Environmental influences on Child Health Outcomes) program to estimate
the association between maternal urinary organophosphate ester concentrations and two types of birth outcome: birth weight and duration of gestation.
Spot urine concentrations of BCEP from the second or third trimester were used to estimate exposure. There were no major concerns for bias. There
was some degree of uncertainty due to the use of only a single timepoint measurement of exposure and variability in ways in which gestational age was
assessed. Birth weight was analyzed using birth weight for gestational age z-scores continuously or classified as small or large size for gestational age, but
not in analyses that excluded gestational age adjustment. Among girls but not boys, compared to concentrations below detection, a high level of detectable
BCEP was associated with significantly shorter mean gestational age and increased odds of preterm birth. In the population overall, a low concentration
of detectable BCEP was associated with significantly reduced odds of late/post term birth; the association with high concentration was null. In addition,
overall, a high concentration of detectable BCEP was associated with significantly lower odds of small for gestational age (SGA) birth. Both low and high
detectable BCEP were associated with reduced odds of SGA among boys. Among girls, low-level detectable BCEP was associated with increased odds of
large for gestational age (LGA) birth. Findings from this study suggest that effects of BCEP may be sex-specific and that BCEP may reduce the length of
gestation particularly in girls but may not reduce birth weight relative to the duration of gestation.
Overall Quality Determination
Medium
Page 136 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 3 of 5
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2:
Attrition
High Participants in this study were all recruited from cohorts that were enrolled in the "En-
vironmental influences on Child Health Outcomes" (ECHO) program at the NIH, where
existing cohorts enrolled their participants into the ECHO program set of cohorts. All
cohorts included were birth cohorts with urine samples from pregnant mothers and mea-
sures of birth outcomes. Individual cohorts followed their own protocols prior to 2019,
but in late 2019 ECHO instituted a common protocol for cohorts to follow. Data for this
study was collected prior to initiation of the common protocol. In total 16 separate co-
horts contributed data to this study, with data collected and submitted to ECHO prior to
March 2022. Study details are available for each cohort, including the number of par-
ticipants included, a description of the sample, eligibility criteria, and their strategy for
obtaining birth outcome information. Of the ECHO cohort participants eligible based on
the availability of maternal urine samples for organophosphate ester (OPE) quantifica-
tion (n=12,873), n=7,048 (54.8%) participants had urinary OPE and dilution data. The
number of urine samples included and excluded from each cohort was provided. Partic-
ipants were excluded if they had no information on children (n=82), had multiple births
(n=10), or had missing outcome data (n=309). One additional child born with a gesta-
tional age of >42 weeks was also excluded, as the formula used for calculating birth
weight for gestational age z-scores did not accommodate gestations of that duration. In
total, the final sample was n=6,646 mother-child dyads. The study compared the charac-
teristics of participants included to those excluded; while no statistical analysis was done
across these two groups, there do not appear to be significant differences in characteris-
tics (e.g. socio-demographics, BMI, tobacco use) of included vs. excluded participants.
Overall, key study information for the overall ECHO program and for individual cohorts
was well-reported and indicated that significant selection bias was unlikely.
Medium The study indicates that 309 out of 7,048 mother-child dyads with urinary OPE mea-
sures had missing birth outcome data (gestational age, birth weight, biological sex at
birth) and were thus excluded from the study. Another 82 pregnancies had no available
child information, and 11 additional children were ineligible (multiple births, length
of gestation). The study does not provide details as to why some dyads were missing
outcome data, but there is no evidence that any cause would be related to their outcome
status.
Continued on next page ...
Page 137 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 3 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, L, Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 3: Comparison Group
Medium Since this study was a pooled analyses of multiple birth cohorts, participants were in-
herently drawn from different background populations and locations across the United
States. Key demographic and study design characteristics that may have differed across
cohorts were considered as potential covariates (e.g., maternal race/ethnicity, education,
age, marital status, along with timing, season, and year of sample collection). Clustering
by cohort was addressed in the statistical model. The authors did not discuss adjusting
for region or examining potential modification by cohort or region. Instead, to examine
the potential influence of heterogeneity across cohorts, the authors included a sensi-
tivity analysis using a "leave-one-cohort-out" approach to assess the influence of each
cohort. While there was generally little or no impact of leaving out any one cohort,
for the TCEP metabolite BCEP (alternative acronym BCETP) omitting one of the two
largest cohorts (CANDLE, predominantly African American mothers from Memphis,
TN) strengthened a positive association between low concentrations and birth weight
z-scores, which became statistically significant. While there was a change in magnitude
and statistical significance, there was no change in direction, and no evidence of bias
due to cohort heterogeneity.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure Medium In all participants, TCEP was measured via the urinary metabolite BCEP. Urine samples
were collected between 2007 and 2020. Measures used a single spot or first morning
urine samples collected from each participant primarily during the second (44.1%) and
third (55.5%) trimesters of pregnancy. BCEP is an appropriate biomarker of exposure
to TCEP. BCEP was quantified using high-performance liquid chromatography with
triple quadrupole mass spectrometry, and quality control measures are well-detailed.
The LOD was specified to be 0.02 ng/mL and the percentage of samples above the LOD
was specified to be 69%. Machine readings were used for values below LOD. Urinary
dilution was addressed using creatinine or specific gravity, depending on the cohort.
Exposure measures at only a single time point may not reflect the full scope of TCEP
exposure across pregnancy due to variability. In addition, other studies examining TCEP
biomarkers reported detecting unmetabolized TCEP in 37% of samples (e.g., Hou et al.
2020 HEROID 10143372); it is uncertain whether or to what extent excluding this factor
may introduce error. Though some error is likely, there was no evidence of important
error or bias in exposure estimates.
Continued on next page ...
Page 138 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 3 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 5:
Exposure Levels
Medium
While the 25th percentile of exposure was below the limit of detection, the median
value was 0.52 ng/mL and the 75th and 95th percentiles were 1.58 and 8.22 ng/mL,
respectively, which should allow for sufficient contrast compared to participants with
no detectable BCEP exposure. In statistical analyses, BCEP was categorized into three
groups: undetected, and then low or high detectable exposure depending on whether
measurements fell above or below the median of dilution-adjusted samples above the
LOD.
Temporality is established as exposure was measured during the second or third
trimester of pregnancy for outcomes that were measured at birth. However, it is un-
certain whether exposures during the second or third trimester reflect the etiologically
relevant time window for birth outcome effects in relation to TCEP exposure.
Metric 6:
Temporality
Medium
Domain 3: Outcome Assessment
Metric 7:
Outcome Measurement or
Characterization
Metric 8: Reporting Bias
Medium Outcomes collected prior to 2019 were ascertained using cohort-specific protocols; the
majority of cohorts used medical record abstraction for outcomes, with a small pro-
portion of parent reporting. Children born in or after 2019 had outcomes ascertained
following the universal ECHO protocol of assigning outcomes based on medical record
abstraction.All instances of birth weight were measured using medical records. Gesta-
tional age was largely abstracted from or estimated using data from medical records;
methods used by hospitals to estimate gestational age were not always specified. Some
cohorts specified using 1st or 2nd trimester ultrasound to calculate gestational age, us-
ing last menstrual period in cases of missing data. While there was some variation in
the estimation of gestational age, there was no evidence that this variation might signifi-
cantly impact effect estimates. Gestational age was further categorized as preterm (<37
weeks), early term (37-38 weeks), full term (39-40 weeks), and late/post-term (41-42
weeks). Sex-specific birth weight for gestational age z-scores were calculated based on
a United states reference population (Aris et al. 2019, not in HERO; PMID: 31201230).
Birth weight for gestational age was also categorized as small for gestational age (<10th
percentile) and large for gestational age (>90th percentile), as well as term low birth
weight (birth weight of <2,500 g at >=37 weeks).
Medium All specified outcomes and results that were outlined in the methods were reported in
the results. Effect estimates are presented with confidence intervals. Tabulations of
outcomes by level of exposure were not provided.
Continued on next page ...
Page 139 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 3 of 5
... continued from previous page
Study Citation:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10:
Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium Potential covariates were selected using a DAG to identify ad include confounders and
precision variables, and to exclude intermediates. Covariates included were maternal
race/ethnicity, maternal age at delivery, maternal education, maternal marital status,
maternal pre-pregnancy BMI, maternal smoking during pregnancy, parity, infant sex,
time of urine sample, season of urine sample, and year of urine sample. A potential
limitation is that the authors did not mention evaluating the influence of gestational
age adjustment on associations with size at birth. Though gestational age might be an
intermediate or collider, there was no direct evidence of important error or bias.
Medium Covariates were collected separately by each cohort within the ECHO program and
then harmonized by the ECHO Data Analysis Center. While it is not specified how
each of these covariates were measured, there is no evidence that insensitive instru-
ments/methods/measures were used.
Medium The study considered co-exposure to other organophosphate esters, including TPHP,
TBUP, TIBP, TDCPP, TBOEP, TCPP, TMPP, TEHP, and TPRP. These co-exposures
were measured via urinary metabolites at the same time and through the same process as
BCEP. Spearman correlations between BECP and these biomarkers were <0.20. These
co-exposures were considered in sensitivity analyses by simultaneously adjusting for
these OPEs, and results were similar.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
High The cohort study design is appropriate for assessing the association between prenatal
concentrations of TCEP metabolites in urine and birth outcomes. The use of linear and
logistic regression for continuous and categorical outcomes respectively is appropriate to
address the research question. Regressions were performed using generalized estimating
equations accounting for clustering at the cohort level, which is appropriate given the
fact that cohorts were combined into one group in this study. Potential non-linear associ-
ations were examined by categorizing BCEP into three groups of similar size.
Medium The number of participants in the study (n=6,646 mother-child dyads) is likely suffi-
ciently large to detect an effect.
Medium The description of the analysis is sufficient to understand precisely what has been done
and to be conceptually reproducible with access to the analytic data.
Continued on next page ...
Page 140 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 3 of 5
... continued from previous page
Study Citation:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Risk estimates were calculated using linear regression models for continuous outcomes
and multinomial logistic regression models for categorical outcomes. BCEP concentra-
tions were categorized, with values below LOD as the referent, and a lower and higher
exposure group defined by categorizing remaining samples at their median. Inclusion
of covariates was dependent on a DAG created to exclude variables that may be causal
intermediates. Covariates with <20% of missing values were addressed via multiple
imputation by chained equations. Sensitivity analyses are detailed, including sex stratifi-
cation, a leave-one-cohort-out approach, and a co-pollutant model.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High
N/A
Medium
Medium
Medium
Medium
Medium
Urinary concentrations of BCEP are a valid and appropriate biomarker of exposure to
TCEP.
No biomarkers of effect were assessed.
The limit of detection was specified to be 0.02 ng/mL. BCEP was detected in 69% of
samples. There was no evidence that the detection limit was inadequate.
The study reports that urine samples were shipped on dry ice to the Human Health Ex-
posure Analysis Resource (HHEAR) at the NYU Grossman School of Medicine. While
no other specific storage information is provided, there is no evidence that inappropriate
methods were used and the biomarker is not expected to have a high likelihood of stabil-
ity.
No information is provided potential contamination.
BCEP was quantified using high-performance liquid chromatography with triple
quadrupole mass spectrometry, and quality control (QC) measures are well-detailed.
One QC measure included the use of blinded duplicate samples (n=127 samples for
BCEP). For BCEP, 66% of duplicate pairs were both above LOD; the remaining pairs
were not concordant with respect to detection. Details on the magnitude of disparities in
estimated concentrations were not provided.
BCEP was adjusted for urinary dilution, measured as either specific gravity or creatinine
depending on the source cohort, using published methods. Results were presented only
for dilution-standardized exposure variables.
Continued on next page ...
Page 141 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 3 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Additional Comments:
This prospective study used data from 16 cohorts participating in the ECHO (Environmental influences on Child Health Outcomes) program to estimate
the association between maternal urinary organophosphate ester concentrations and two types of birth outcome: birth weight and duration of gestation.
Spot urine concentrations of BCEP from the second or third trimester were used to estimate exposure. There were no major concerns for bias. There
was some degree of uncertainty due to the use of only a single timepoint measurement of exposure and variability in ways in which gestational age was
assessed. Birth weight was analyzed using birth weight for gestational age z-scores continuously or classified as small or large size for gestational age, but
not in analyses that excluded gestational age adjustment. Among girls but not boys, compared to concentrations below detection, a high level of detectable
BCEP was associated with significantly shorter mean gestational age and increased odds of preterm birth. In the population overall, a low concentration
of detectable BCEP was associated with significantly reduced odds of late/post term birth; the association with high concentration was null. In addition,
overall, a high concentration of detectable BCEP was associated with significantly lower odds of small for gestational age (SGA) birth. Both low and high
detectable BCEP were associated with reduced odds of SGA among boys. Among girls, low-level detectable BCEP was associated with increased odds of
large for gestational age (LGA) birth. Findings from this study suggest that effects of BCEP may be sex-specific and that BCEP may reduce the length of
gestation particularly in girls but may not reduce birth weight relative to the duration of gestation.
Overall Quality Determination
Medium
Page 142 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 4 of 5
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2:
Attrition
High Participants in this study were all recruited from cohorts that were enrolled in the "En-
vironmental influences on Child Health Outcomes" (ECHO) program at the NIH, where
existing cohorts enrolled their participants into the ECHO program set of cohorts. All
cohorts included were birth cohorts with urine samples from pregnant mothers and mea-
sures of birth outcomes. Individual cohorts followed their own protocols prior to 2019,
but in late 2019 ECHO instituted a common protocol for cohorts to follow. Data for this
study was collected prior to initiation of the common protocol. In total 16 separate co-
horts contributed data to this study, with data collected and submitted to ECHO prior to
March 2022. Study details are available for each cohort, including the number of par-
ticipants included, a description of the sample, eligibility criteria, and their strategy for
obtaining birth outcome information. Of the ECHO cohort participants eligible based on
the availability of maternal urine samples for organophosphate ester (OPE) quantifica-
tion (n=12,873), n=7,048 (54.8%) participants had urinary OPE and dilution data. The
number of urine samples included and excluded from each cohort was provided. Partic-
ipants were excluded if they had no information on children (n=82), had multiple births
(n=10), or had missing outcome data (n=309). One additional child born with a gesta-
tional age of >42 weeks was also excluded, as the formula used for calculating birth
weight for gestational age z-scores did not accommodate gestations of that duration. In
total, the final sample was n=6,646 mother-child dyads. The study compared the charac-
teristics of participants included to those excluded; while no statistical analysis was done
across these two groups, there do not appear to be significant differences in characteris-
tics (e.g. socio-demographics, BMI, tobacco use) of included vs. excluded participants.
Overall, key study information for the overall ECHO program and for individual cohorts
was well-reported and indicated that significant selection bias was unlikely.
Medium The study indicates that 309 out of 7,048 mother-child dyads with urinary OPE mea-
sures had missing birth outcome data (gestational age, birth weight, biological sex at
birth) and were thus excluded from the study. Another 82 pregnancies had no available
child information, and 11 additional children were ineligible (multiple births, length
of gestation). The study does not provide details as to why some dyads were missing
outcome data, but there is no evidence that any cause would be related to their outcome
status.
Continued on next page ...
Page 143 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 4 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, L, Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 3: Comparison Group
Medium Since this study was a pooled analyses of multiple birth cohorts, participants were in-
herently drawn from different background populations and locations across the United
States. Key demographic and study design characteristics that may have differed across
cohorts were considered as potential covariates (e.g., maternal race/ethnicity, education,
age, marital status, along with timing, season, and year of sample collection). Clustering
by cohort was addressed in the statistical model. The authors did not discuss adjusting
for region or examining potential modification by cohort or region. Instead, to examine
the potential influence of heterogeneity across cohorts, the authors included a sensi-
tivity analysis using a "leave-one-cohort-out" approach to assess the influence of each
cohort. While there was generally little or no impact of leaving out any one cohort,
for the TCEP metabolite BCEP (alternative acronym BCETP) omitting one of the two
largest cohorts (CANDLE, predominantly African American mothers from Memphis,
TN) strengthened a positive association between low concentrations and birth weight
z-scores, which became statistically significant. While there was a change in magnitude
and statistical significance, there was no change in direction, and no evidence of bias
due to cohort heterogeneity.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure Medium In all participants, TCEP was measured via the urinary metabolite BCEP. Urine samples
were collected between 2007 and 2020. Measures used a single spot or first morning
urine samples collected from each participant primarily during the second (44.1%) and
third (55.5%) trimesters of pregnancy. BCEP is an appropriate biomarker of exposure
to TCEP. BCEP was quantified using high-performance liquid chromatography with
triple quadrupole mass spectrometry, and quality control measures are well-detailed.
The LOD was specified to be 0.02 ng/mL and the percentage of samples above the LOD
was specified to be 69%. Machine readings were used for values below LOD. Urinary
dilution was addressed using creatinine or specific gravity, depending on the cohort.
Exposure measures at only a single time point may not reflect the full scope of TCEP
exposure across pregnancy due to variability. In addition, other studies examining TCEP
biomarkers reported detecting unmetabolized TCEP in 37% of samples (e.g., Hou et al.
2020 HEROID 10143372); it is uncertain whether or to what extent excluding this factor
may introduce error. Though some error is likely, there was no evidence of important
error or bias in exposure estimates.
Continued on next page ...
Page 144 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 4 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 5:
Exposure Levels
Medium
While the 25th percentile of exposure was below the limit of detection, the median
value was 0.52 ng/mL and the 75th and 95th percentiles were 1.58 and 8.22 ng/mL,
respectively, which should allow for sufficient contrast compared to participants with
no detectable BCEP exposure. In statistical analyses, BCEP was categorized into three
groups: undetected, and then low or high detectable exposure depending on whether
measurements fell above or below the median of dilution-adjusted samples above the
LOD.
Temporality is established as exposure was measured during the second or third
trimester of pregnancy for outcomes that were measured at birth. However, it is un-
certain whether exposures during the second or third trimester reflect the etiologically
relevant time window for birth outcome effects in relation to TCEP exposure.
Metric 6:
Temporality
Medium
Domain 3: Outcome Assessment
Metric 7:
Outcome Measurement or
Characterization
Metric 8: Reporting Bias
Medium Outcomes collected prior to 2019 were ascertained using cohort-specific protocols; the
majority of cohorts used medical record abstraction for outcomes, with a small pro-
portion of parent reporting. Children born in or after 2019 had outcomes ascertained
following the universal ECHO protocol of assigning outcomes based on medical record
abstraction.All instances of birth weight were measured using medical records. Gesta-
tional age was largely abstracted from or estimated using data from medical records;
methods used by hospitals to estimate gestational age were not always specified. Some
cohorts specified using 1st or 2nd trimester ultrasound to calculate gestational age, us-
ing last menstrual period in cases of missing data. While there was some variation in
the estimation of gestational age, there was no evidence that this variation might signifi-
cantly impact effect estimates. Gestational age was further categorized as preterm (<37
weeks), early term (37-38 weeks), full term (39-40 weeks), and late/post-term (41-42
weeks). Sex-specific birth weight for gestational age z-scores were calculated based on
a United states reference population (Aris et al. 2019, not in HERO; PMID: 31201230).
Birth weight for gestational age was also categorized as small for gestational age (<10th
percentile) and large for gestational age (>90th percentile), as well as term low birth
weight (birth weight of <2,500 g at >=37 weeks).
Medium All specified outcomes and results that were outlined in the methods were reported in
the results. Effect estimates are presented with confidence intervals. Tabulations of
outcomes by level of exposure were not provided.
Continued on next page ...
Page 145 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 4 of 5
... continued from previous page
Study Citation:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10:
Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium Potential covariates were selected using a DAG to identify ad include confounders and
precision variables, and to exclude intermediates. Covariates included were maternal
race/ethnicity, maternal age at delivery, maternal education, maternal marital status,
maternal pre-pregnancy BMI, maternal smoking during pregnancy, parity, infant sex,
time of urine sample, season of urine sample, and year of urine sample. A potential
limitation is that the authors did not mention evaluating the influence of gestational
age adjustment on associations with size at birth. Though gestational age might be an
intermediate or collider, there was no direct evidence of important error or bias.
Medium Covariates were collected separately by each cohort within the ECHO program and
then harmonized by the ECHO Data Analysis Center. While it is not specified how
each of these covariates were measured, there is no evidence that insensitive instru-
ments/methods/measures were used.
Medium The study considered co-exposure to other organophosphate esters, including TPHP,
TBUP, TIBP, TDCPP, TBOEP, TCPP, TMPP, TEHP, and TPRP. These co-exposures
were measured via urinary metabolites at the same time and through the same process as
BCEP. Spearman correlations between BECP and these biomarkers were <0.20. These
co-exposures were considered in sensitivity analyses by simultaneously adjusting for
these OPEs, and results were similar.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
High The cohort study design is appropriate for assessing the association between prenatal
concentrations of TCEP metabolites in urine and birth outcomes. The use of linear and
logistic regression for continuous and categorical outcomes respectively is appropriate to
address the research question. Regressions were performed using generalized estimating
equations accounting for clustering at the cohort level, which is appropriate given the
fact that cohorts were combined into one group in this study. Potential non-linear associ-
ations were examined by categorizing BCEP into three groups of similar size.
Medium The number of participants in the study (n=6,646 mother-child dyads) is likely suffi-
ciently large to detect an effect.
Medium The description of the analysis is sufficient to understand precisely what has been done
and to be conceptually reproducible with access to the analytic data.
Continued on next page ...
Page 146 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 4 of 5
... continued from previous page
Study Citation:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Risk estimates were calculated using linear regression models for continuous outcomes
and multinomial logistic regression models for categorical outcomes. BCEP concentra-
tions were categorized, with values below LOD as the referent, and a lower and higher
exposure group defined by categorizing remaining samples at their median. Inclusion
of covariates was dependent on a DAG created to exclude variables that may be causal
intermediates. Covariates with <20% of missing values were addressed via multiple
imputation by chained equations. Sensitivity analyses are detailed, including sex stratifi-
cation, a leave-one-cohort-out approach, and a co-pollutant model.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High
N/A
Medium
Medium
Medium
Medium
Medium
Urinary concentrations of BCEP are a valid and appropriate biomarker of exposure to
TCEP.
No biomarkers of effect were assessed.
The limit of detection was specified to be 0.02 ng/mL. BCEP was detected in 69% of
samples. There was no evidence that the detection limit was inadequate.
The study reports that urine samples were shipped on dry ice to the Human Health Ex-
posure Analysis Resource (HHEAR) at the NYU Grossman School of Medicine. While
no other specific storage information is provided, there is no evidence that inappropriate
methods were used and the biomarker is not expected to have a high likelihood of stabil-
ity.
No information is provided potential contamination.
BCEP was quantified using high-performance liquid chromatography with triple
quadrupole mass spectrometry, and quality control (QC) measures are well-detailed.
One QC measure included the use of blinded duplicate samples (n=127 samples for
BCEP). For BCEP, 66% of duplicate pairs were both above LOD; the remaining pairs
were not concordant with respect to detection. Details on the magnitude of disparities in
estimated concentrations were not provided.
BCEP was adjusted for urinary dilution, measured as either specific gravity or creatinine
depending on the source cohort, using published methods. Results were presented only
for dilution-standardized exposure variables.
Continued on next page ...
Page 147 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 4 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Additional Comments:
This prospective study used data from 16 cohorts participating in the ECHO (Environmental influences on Child Health Outcomes) program to estimate
the association between maternal urinary organophosphate ester concentrations and two types of birth outcome: birth weight and duration of gestation.
Spot urine concentrations of BCEP from the second or third trimester were used to estimate exposure. There were no major concerns for bias. There
was some degree of uncertainty due to the use of only a single timepoint measurement of exposure and variability in ways in which gestational age was
assessed. Birth weight was analyzed using birth weight for gestational age z-scores continuously or classified as small or large size for gestational age, but
not in analyses that excluded gestational age adjustment. Among girls but not boys, compared to concentrations below detection, a high level of detectable
BCEP was associated with significantly shorter mean gestational age and increased odds of preterm birth. In the population overall, a low concentration
of detectable BCEP was associated with significantly reduced odds of late/post term birth; the association with high concentration was null. In addition,
overall, a high concentration of detectable BCEP was associated with significantly lower odds of small for gestational age (SGA) birth. Both low and high
detectable BCEP were associated with reduced odds of SGA among boys. Among girls, low-level detectable BCEP was associated with increased odds of
large for gestational age (LGA) birth. Findings from this study suggest that effects of BCEP may be sex-specific and that BCEP may reduce the length of
gestation particularly in girls but may not reduce birth weight relative to the duration of gestation.
Overall Quality Determination
Medium
Page 148 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 5 of 5
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2:
Attrition
High Participants in this study were all recruited from cohorts that were enrolled in the "En-
vironmental influences on Child Health Outcomes" (ECHO) program at the NIH, where
existing cohorts enrolled their participants into the ECHO program set of cohorts. All
cohorts included were birth cohorts with urine samples from pregnant mothers and mea-
sures of birth outcomes. Individual cohorts followed their own protocols prior to 2019,
but in late 2019 ECHO instituted a common protocol for cohorts to follow. Data for this
study was collected prior to initiation of the common protocol. In total 16 separate co-
horts contributed data to this study, with data collected and submitted to ECHO prior to
March 2022. Study details are available for each cohort, including the number of par-
ticipants included, a description of the sample, eligibility criteria, and their strategy for
obtaining birth outcome information. Of the ECHO cohort participants eligible based on
the availability of maternal urine samples for organophosphate ester (OPE) quantifica-
tion (n=12,873), n=7,048 (54.8%) participants had urinary OPE and dilution data. The
number of urine samples included and excluded from each cohort was provided. Partic-
ipants were excluded if they had no information on children (n=82), had multiple births
(n=10), or had missing outcome data (n=309). One additional child born with a gesta-
tional age of >42 weeks was also excluded, as the formula used for calculating birth
weight for gestational age z-scores did not accommodate gestations of that duration. In
total, the final sample was n=6,646 mother-child dyads. The study compared the charac-
teristics of participants included to those excluded; while no statistical analysis was done
across these two groups, there do not appear to be significant differences in characteris-
tics (e.g. socio-demographics, BMI, tobacco use) of included vs. excluded participants.
Overall, key study information for the overall ECHO program and for individual cohorts
was well-reported and indicated that significant selection bias was unlikely.
Medium The study indicates that 309 out of 7,048 mother-child dyads with urinary OPE mea-
sures had missing birth outcome data (gestational age, birth weight, biological sex at
birth) and were thus excluded from the study. Another 82 pregnancies had no available
child information, and 11 additional children were ineligible (multiple births, length
of gestation). The study does not provide details as to why some dyads were missing
outcome data, but there is no evidence that any cause would be related to their outcome
status.
Continued on next page ...
Page 149 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 5 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, L, Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 3: Comparison Group
Medium Since this study was a pooled analyses of multiple birth cohorts, participants were in-
herently drawn from different background populations and locations across the United
States. Key demographic and study design characteristics that may have differed across
cohorts were considered as potential covariates (e.g., maternal race/ethnicity, education,
age, marital status, along with timing, season, and year of sample collection). Clustering
by cohort was addressed in the statistical model. The authors did not discuss adjusting
for region or examining potential modification by cohort or region. Instead, to examine
the potential influence of heterogeneity across cohorts, the authors included a sensi-
tivity analysis using a "leave-one-cohort-out" approach to assess the influence of each
cohort. While there was generally little or no impact of leaving out any one cohort,
for the TCEP metabolite BCEP (alternative acronym BCETP) omitting one of the two
largest cohorts (CANDLE, predominantly African American mothers from Memphis,
TN) strengthened a positive association between low concentrations and birth weight
z-scores, which became statistically significant. While there was a change in magnitude
and statistical significance, there was no change in direction, and no evidence of bias
due to cohort heterogeneity.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure Medium In all participants, TCEP was measured via the urinary metabolite BCEP. Urine samples
were collected between 2007 and 2020. Measures used a single spot or first morning
urine samples collected from each participant primarily during the second (44.1%) and
third (55.5%) trimesters of pregnancy. BCEP is an appropriate biomarker of exposure
to TCEP. BCEP was quantified using high-performance liquid chromatography with
triple quadrupole mass spectrometry, and quality control measures are well-detailed.
The LOD was specified to be 0.02 ng/mL and the percentage of samples above the LOD
was specified to be 69%. Machine readings were used for values below LOD. Urinary
dilution was addressed using creatinine or specific gravity, depending on the cohort.
Exposure measures at only a single time point may not reflect the full scope of TCEP
exposure across pregnancy due to variability. In addition, other studies examining TCEP
biomarkers reported detecting unmetabolized TCEP in 37% of samples (e.g., Hou et al.
2020 HEROID 10143372); it is uncertain whether or to what extent excluding this factor
may introduce error. Though some error is likely, there was no evidence of important
error or bias in exposure estimates.
Continued on next page ...
Page 150 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 5 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 5:
Exposure Levels
Medium
While the 25th percentile of exposure was below the limit of detection, the median
value was 0.52 ng/mL and the 75th and 95th percentiles were 1.58 and 8.22 ng/mL,
respectively, which should allow for sufficient contrast compared to participants with
no detectable BCEP exposure. In statistical analyses, BCEP was categorized into three
groups: undetected, and then low or high detectable exposure depending on whether
measurements fell above or below the median of dilution-adjusted samples above the
LOD.
Temporality is established as exposure was measured during the second or third
trimester of pregnancy for outcomes that were measured at birth. However, it is un-
certain whether exposures during the second or third trimester reflect the etiologically
relevant time window for birth outcome effects in relation to TCEP exposure.
Metric 6:
Temporality
Medium
Domain 3: Outcome Assessment
Metric 7:
Outcome Measurement or
Characterization
Metric 8: Reporting Bias
Medium Outcomes collected prior to 2019 were ascertained using cohort-specific protocols; the
majority of cohorts used medical record abstraction for outcomes, with a small pro-
portion of parent reporting. Children born in or after 2019 had outcomes ascertained
following the universal ECHO protocol of assigning outcomes based on medical record
abstraction.All instances of birth weight were measured using medical records. Gesta-
tional age was largely abstracted from or estimated using data from medical records;
methods used by hospitals to estimate gestational age were not always specified. Some
cohorts specified using 1st or 2nd trimester ultrasound to calculate gestational age, us-
ing last menstrual period in cases of missing data. While there was some variation in
the estimation of gestational age, there was no evidence that this variation might signifi-
cantly impact effect estimates. Gestational age was further categorized as preterm (<37
weeks), early term (37-38 weeks), full term (39-40 weeks), and late/post-term (41-42
weeks). Sex-specific birth weight for gestational age z-scores were calculated based on
a United states reference population (Aris et al. 2019, not in HERO; PMID: 31201230).
Birth weight for gestational age was also categorized as small for gestational age (<10th
percentile) and large for gestational age (>90th percentile), as well as term low birth
weight (birth weight of <2,500 g at >=37 weeks).
Medium All specified outcomes and results that were outlined in the methods were reported in
the results. Effect estimates are presented with confidence intervals. Tabulations of
outcomes by level of exposure were not provided.
Continued on next page ...
Page 151 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 5 of 5
... continued from previous page
Study Citation:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10:
Covariate Characterization
Metric 11: Co-exposure Counfounding
Medium Potential covariates were selected using a DAG to identify ad include confounders and
precision variables, and to exclude intermediates. Covariates included were maternal
race/ethnicity, maternal age at delivery, maternal education, maternal marital status,
maternal pre-pregnancy BMI, maternal smoking during pregnancy, parity, infant sex,
time of urine sample, season of urine sample, and year of urine sample. A potential
limitation is that the authors did not mention evaluating the influence of gestational
age adjustment on associations with size at birth. Though gestational age might be an
intermediate or collider, there was no direct evidence of important error or bias.
Medium Covariates were collected separately by each cohort within the ECHO program and
then harmonized by the ECHO Data Analysis Center. While it is not specified how
each of these covariates were measured, there is no evidence that insensitive instru-
ments/methods/measures were used.
Medium The study considered co-exposure to other organophosphate esters, including TPHP,
TBUP, TIBP, TDCPP, TBOEP, TCPP, TMPP, TEHP, and TPRP. These co-exposures
were measured via urinary metabolites at the same time and through the same process as
BCEP. Spearman correlations between BECP and these biomarkers were <0.20. These
co-exposures were considered in sensitivity analyses by simultaneously adjusting for
these OPEs, and results were similar.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
High The cohort study design is appropriate for assessing the association between prenatal
concentrations of TCEP metabolites in urine and birth outcomes. The use of linear and
logistic regression for continuous and categorical outcomes respectively is appropriate to
address the research question. Regressions were performed using generalized estimating
equations accounting for clustering at the cohort level, which is appropriate given the
fact that cohorts were combined into one group in this study. Potential non-linear associ-
ations were examined by categorizing BCEP into three groups of similar size.
Medium The number of participants in the study (n=6,646 mother-child dyads) is likely suffi-
ciently large to detect an effect.
Medium The description of the analysis is sufficient to understand precisely what has been done
and to be conceptually reproducible with access to the analytic data.
Continued on next page ...
Page 152 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 5 of 5
... continued from previous page
Study Citation:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Metric 15: Statistical Analysis
High Risk estimates were calculated using linear regression models for continuous outcomes
and multinomial logistic regression models for categorical outcomes. BCEP concentra-
tions were categorized, with values below LOD as the referent, and a lower and higher
exposure group defined by categorizing remaining samples at their median. Inclusion
of covariates was dependent on a DAG created to exclude variables that may be causal
intermediates. Covariates with <20% of missing values were addressed via multiple
imputation by chained equations. Sensitivity analyses are detailed, including sex stratifi-
cation, a leave-one-cohort-out approach, and a co-pollutant model.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
High
N/A
Medium
Medium
Medium
Medium
Medium
Urinary concentrations of BCEP are a valid and appropriate biomarker of exposure to
TCEP.
No biomarkers of effect were assessed.
The limit of detection was specified to be 0.02 ng/mL. BCEP was detected in 69% of
samples. There was no evidence that the detection limit was inadequate.
The study reports that urine samples were shipped on dry ice to the Human Health Ex-
posure Analysis Resource (HHEAR) at the NYU Grossman School of Medicine. While
no other specific storage information is provided, there is no evidence that inappropriate
methods were used and the biomarker is not expected to have a high likelihood of stabil-
ity.
No information is provided potential contamination.
BCEP was quantified using high-performance liquid chromatography with triple
quadrupole mass spectrometry, and quality control (QC) measures are well-detailed.
One QC measure included the use of blinded duplicate samples (n=127 samples for
BCEP). For BCEP, 66% of duplicate pairs were both above LOD; the remaining pairs
were not concordant with respect to detection. Details on the magnitude of disparities in
estimated concentrations were not provided.
BCEP was adjusted for urinary dilution, measured as either specific gravity or creatinine
depending on the source cohort, using published methods. Results were presented only
for dilution-standardized exposure variables.
Continued on next page ...
Page 153 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11577123 Table: 5 of 5
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Oh, J., Buckley, J. P., Li, X., Gachigi, K. K., Kannan, K., Lyu, W. J., Ames, J. L., Barrett, E. S., Bastain, T. M., Breton, C. V., Buss, C., Croen, L. A.,
Dunlop, A. L., Ferrara, A., Ghassabian, A., Herbstman, J. B., Hernandez-Castro, I., Hertz-Picciotto, I., Kahn, L. G., Karagas, M. R., Kuiper, J. R., McEvoy,
C. T., Meeker, J. D., Morello-Frosch, R., Padula, A. M., Romano, M. E., Sathyanarayana, S., Schantz, S., Schmidt, R. J., Simhan, H., Starling, A. P.,
Tylavsky, F. A., Volk, H. E., Woodruff, T. J., Zhu, Y. Y., Bennett, D. H. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child Health Outcomes (ECHO) program. Environmental Health
Perspectives 132(1):017004.
Reproductive /Developmental-Birth anthropometry (birth weight for gestational age z-score, term low birth weight, small for gestational age, large
for gestational age)Length of gestation (gestational age at birth, preterm birth, early term birth, late/post-term birth)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis(2-chloroethyl) phosphate (BCETP)
No linked references.
11577123
Domain
Metric
Rating
Comments
Additional Comments:
This prospective study used data from 16 cohorts participating in the ECHO (Environmental influences on Child Health Outcomes) program to estimate
the association between maternal urinary organophosphate ester concentrations and two types of birth outcome: birth weight and duration of gestation.
Spot urine concentrations of BCEP from the second or third trimester were used to estimate exposure. There were no major concerns for bias. There
was some degree of uncertainty due to the use of only a single timepoint measurement of exposure and variability in ways in which gestational age was
assessed. Birth weight was analyzed using birth weight for gestational age z-scores continuously or classified as small or large size for gestational age, but
not in analyses that excluded gestational age adjustment. Among girls but not boys, compared to concentrations below detection, a high level of detectable
BCEP was associated with significantly shorter mean gestational age and increased odds of preterm birth. In the population overall, a low concentration
of detectable BCEP was associated with significantly reduced odds of late/post term birth; the association with high concentration was null. In addition,
overall, a high concentration of detectable BCEP was associated with significantly lower odds of small for gestational age (SGA) birth. Both low and high
detectable BCEP were associated with reduced odds of SGA among boys. Among girls, low-level detectable BCEP was associated with increased odds of
large for gestational age (LGA) birth. Findings from this study suggest that effects of BCEP may be sex-specific and that BCEP may reduce the length of
gestation particularly in girls but may not reduce birth weight relative to the duration of gestation.
Overall Quality Determination
Medium
Page 154 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581667 Table: 1 of 1
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Chen, A., Sucharew, H., Yang, W., Vuong, A. M., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Cecil, K. M., Xu, Y., Yolton, K. (2023).
Early-life exposure to a mixture of organophosphate esters and child behavior. International lournal of Hygiene and Environmental Health 250:114162.
Neurological/Behavioral-Behavioral Assessment System for Children, 2nd Edition (BASC-2): scales for internalizing problems, externalizing problems,
and Behavioral Symptom Index (BSI)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis-2-chloroethyl phosphate (BCEP)
No linked references.
11581667
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2:
Attrition
Metric 3:
Comparison Group
Medium This cohort study examined organophosphate ester (OPE) mixtures and child behavior
among participants recruited between March 2003 and February 2006 for the Health
Outcomes and Measures of the Environment (HOME) Study, a pregnancy and birth
cohort based in Cincinnati, Ohio. Eligibility criteria for pregnant women included:
being at least 18 years of age, being at 16 +/- 3 weeks of gestation, living in a home
built before 1978, being fluent in English, not being diagnosed with various conditions
(bipolar disorder, schizophrenia, diabetes, or cancer), not taking medications for thy-
roid disorder of seizures, being HIV negative, and not planning to move outside of the
Greater Cincinatti Area. Participants provided urine samples that were used to estimate
organophosphate ester (OPE) metabolites at multiple study visits during pregnancy and
early childhood. For this study of organophosphate exposure and child behavior, the
authors included mother-child pairs with at least one measure of urinary OPE metabo-
lites from pregnancy or childhood, as well as at least one parent-completed behavioral
assessment at age 3 or 8 years. Of 389 women who gave birth to singletons, 170 did
not meet these criteria, resulting in a final analytical sample of 219 mother-child pairs.
Included and excluded participants were not compared, but there was no evidence to
suggest inclusion might be associated with OPE exposure. The authors provided suffi-
cient information pertaining to participant selection, and there are no major concerns of
selection bias.
Medium Of 389 pregnant women in the HOME cohort who gave birth to singleton infants, 170
were excluded because they did not have at least one measure of OPEs for the mother
or child, or at least one parent assessment of children using the Behavioral Assessment
System for Children (BASC-2) at age 3 or 8 years. There were no additional exclusions
due to missing covariate data. The analysis sample of 219 children included 56.2%
of singleton births. While there was attrition, there was no evidence of differences in
included vs. excluded participants that might induce bias. No major concerns were
noted.
Medium All participants were recruited from the same eligible population using the same cri-
teria. The authors provided clear details about the study setting and key inclusion and
exclusion details. There was no evidence for concerns related to the adequacy of the
comparison group of participants with lower levels of OPE metabolites.
Domain 2: Exposure Characterization
Continued on next page .
Page 155 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581667 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Chen, A., Sucharew, H., Yang, W., Vuong, A. M., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Cecil, K. M., Xu, Y., Yolton, K. (2023).
Early-life exposure to a mixture of organophosphate esters and child behavior. International lournal of Hygiene and Environmental Health 250:114162.
Neurological/Behavioral-Behavioral Assessment System for Children, 2nd Edition (BASC-2): scales for internalizing problems, externalizing problems,
and Behavioral Symptom Index (BSI)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis-2-chloroethyl phosphate (BCEP)
No linked references.
11581667
Domain
Metric
Rating
Comments
Metric 4: Measurement of Exposure
Metric 5:
Exposure Levels
Metric 6:
Temporality
High Spot urines or urine samples collected in diaper inserts were used to measure OPE
metabolite concentrations that included the TCEP metabolite BCEP. While associa-
tions were analyzed for an OPE mixture and not for TCEP individually, there were no
notable limitations with respect to the measurement approach. Samples were collected
from participating mothers (weeks 16 and 26 of gestation, at delivery) and children (ages
1,2,3 and 5 years). The number of samples collected varied by participant. While the
number of samples per mother-child pair was not reported, sample sizes across time
points ranged from n=153 to n=217. Intra-class correlations (ICC) for repeated mea-
sures were low (ICC = 0.19 for children, 0.27 for mothers), indicating that there was
substantial variation in concentrations within individuals over time. BCEP was quan-
tified using HPLC with mass spectrometry. The authors reported an LOD of 0.1 ug/L,
and the percent of samples with detectable levels ranged from 83.9% at 26 weeks of
gestation to 95.5% in children at age 1 year. Values below LOD were imputed as LOD
divided by the square root of 2. To account for dilution, laboratory technicians measured
the specific gravity of all urine samples using an ATAGO PAL-10S pocket refractometer.
Although exposure at each time point was represented by a single sample, an important
strength of this study was the availability of repeated measures from both pregnancy and
early childhood.
Medium Table 1 provided medians and IQRs for individual OPE metabolites, including BCEP, at
each timepoint. The median (IQR) in ug/L at was 0.60 (0.34, 1.07) at 16 weeks' gesta-
tion, and 1.13 (0.47, 2.73) at age 1 year. The range and distribution of exposure reported
in this study appeared sufficient to develop an exposure-response estimate, and the au-
thors utilized a continuous measure of exposure for their analyses. However, it was a
limitation that associations between OPE exposures and behavioral outcomes were pre-
sented only for metabolite mixtures, and not for individual metabolites.
High Temporal sequencing was appropriate. OPE exposure measures from pregnancy were
used to estimate associations with behavioral outcomes measured at both ages 3 and
8 years. Exposure measures from childhood (ages 1 to 5 years) were used to estimate
associations with outcomes at age 8 years. While the importance of specific exposure
windows for BCEP and child neurodevelopment is unclear, analysis of both prenatal and
postnatal exposure measures, along with appropriate sequencing, were strengths of this
study.
Domain 3: Outcome Assessment
Continued on next page .
Page 156 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581667 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Chen, A., Sucharew, H., Yang, W., Vuong, A. M., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Cecil, K. M., Xu, Y., Yolton, K. (2023).
Early-life exposure to a mixture of organophosphate esters and child behavior. International lournal of Hygiene and Environmental Health 250:114162.
Neurological/Behavioral-Behavioral Assessment System for Children, 2nd Edition (BASC-2): scales for internalizing problems, externalizing problems,
and Behavioral Symptom Index (BSI)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis-2-chloroethyl phosphate (BCEP)
No linked references.
11581667
Domain
Metric
Rating
Comments
Metric 7: Outcome Measurement or
Characterization
Metric 8: Reporting Bias
Medium Outcomes were assessed using the BASC-2 questionnaire, which was completed by
caregivers (98% by mothers) at the 3- or 8-year study visits after receiving instructions
from study staff. The BASC-2 is widely used 160-item parent-reported assessment of
children's behavior in public and home settings. This assessment provides four com-
posite scales; three were assessed in this study (Internalizing Problems, Externalizing
Problems, and the Behavioral Symptom Index (BSI)). BASC-2 scores are age normal-
ized to a mean of 50 and a SD of 10. The authors cited literature reporting that the in-
ternal consistency of the BASC-2 ranges from r = 0.87-0.95 across ages 3 and 8 years,
and correlations with other measures of child behavior from r = 0.65-0.84. A potential
limitation is that the behavioral measures used in this study did not include assessments
by teachers or personnel with specialized training. Some measurement error in parental
reporting is likely, but there was no evidence of systematic bias, and models adjusted
for variables that included maternal depression and income. There was no evidence of
inadequate validity or important bias in outcome assessment.
High The authors provided details about the measured outcomes, and associated effect esti-
mates are reported along with 95% confidence intervals. No concerns of reporting bias
are noted in this study.
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment Medium The authors detailed that a number of variables were considered as potential con-
founders including breastfeeding status, maternal race, household income at baseline,
maternal depression at the time of BASC-2 measurement, maternal education at base-
line, maternal age at baseline, marital status at baseline, and the caregiving environment.
The final covariate set was selected using a directed acyclic graph, and included ma-
ternal depression, breastfeeding status, maternal race (white or non-white), household
income, and caregiving environment. The authors did not explicitly state that BASC-2
scores used were normed for child sex as well as child age at administration, and there
was no stratification by or adjustment for child sex. Covariates did not include indicators
of schooling or behavioral interventions. Nonetheless, there was no direct evidence of
important residual confounding or that potential intermediates were included as covari-
ates.
Metric 10: Covariate Characterization Medium All variables considered as covariates were assessed using questionnaires and methods
that included well-established instruments. The caregiving environment was assessed
using the Home Observation and Measurement of the Environment, which is a ques-
tionnaire and interview obtaining information on the quality and quantity of caregiving
in the home at 12 months of age. Maternal depression was measured using the Beck's
Depression Inventory. There was no evidence that the methods employed to measure
covariates had poor validity.
Continued on next page ...
Page 157 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581667 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Chen, A., Sucharew, H., Yang, W., Vuong, A. M., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Cecil, K. M., Xu, Y., Yolton, K. (2023).
Early-life exposure to a mixture of organophosphate esters and child behavior. International lournal of Hygiene and Environmental Health 250:114162.
Neurological/Behavioral-Behavioral Assessment System for Children, 2nd Edition (BASC-2): scales for internalizing problems, externalizing problems,
and Behavioral Symptom Index (BSI)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis-2-chloroethyl phosphate (BCEP)
No linked references.
11581667
Domain
Metric
Rating
Comments
Metric 11: Co-exposure Counfounding
Medium Exposure was assessed as a mixture of three OPE metabolites: BCEP, BDCIPP, bis(l,3-
dichloro-2-propyl) phosphate [BDCIPP], and diphenyl phosphate [DPHP], However,
correlations among these OPE metabolites were not reported, and potential confounding
of associations with one metabolite by others was not directly assessed. The authors
cited a study in which an OPE metabolite not measured in this study - isopropyl-phenyl
phenyl phosphate (ip-PPP) - was associated with child behavior at age 8 years. The
authors noted the possibility that confounding by co-exposures as a potential concern.
However, there was no direct evidence of such bias.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High The cohort study design implemented was appropriate for the research question being
examined. Urinary OPE concentrations were measured prenatally and in early life, and
associations with BASC-2 scores in children at later ages were examined. The authors
also applied appropriate statistical methods to evaluate the primary research questions.
No concerns were noted for the general study design.
Medium There was variability in both exposure measures; statistical power was increased by ana-
lyzing both exposure and outcome using continuous variables, as well as the availability
of repeated measures. There was no evidence that the number of participants in this
study (n=219 mother-child pairs) was inadequate to detect associations.
Medium The authors provided a sufficient description of their analytical methods to allow others
to conceptually reproduce their results with access to the pertinent data. No major con-
cerns are noted, as the description of analyses performed was clear.
Low Descriptive data were provided for participant characteristics and OPE exposure vari-
ables, but not for BASC-2 scores. In this paper, the authors estimated associations
between mixtures of natural log transformed OPE metabolite concentrations and be-
havioral outcomes using two methods: structural equation modeling and quantile g-
computation. These analysis methods are sound for the analysis of mixtures. Analyses
of associations between behavioral outcomes and individual OPEs were not discussed.
The exclusive analysis of mixtures negates the utility of the study for evaluating po-
tential effects specific to TCEP or other individual OPEs. Factor loadings for the latent
OPE exposure variable provided limited information suggesting that BCEP and BDCIPP
were potential drivers of associations. BCEP was the reference chemical against which
factor loadings for BDCIPP and DPHP were reported; loadings were highest for BCEP
at age 2 years, and lower for BCEP than for BDCIPP at other time points. However, as-
sociations for OPE mixtures at age 2 years were not significant; patterns of associations
were inconsistent over time. While there were no major concerns for the analysis of
OPE mixtures, the analysis strategy provided little meaningful information on potential
effects of TCEP individually, warranting a low rating.
Continued on next page ...
Page 158 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581667 Table: 1 of 1
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Chen, A., Sucharew, H., Yang, W., Vuong, A. M., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Cecil, K. M., Xu, Y., Yolton, K. (2023).
Early-life exposure to a mixture of organophosphate esters and child behavior. International lournal of Hygiene and Environmental Health 250:114162.
Neurological/Behavioral-Behavioral Assessment System for Children, 2nd Edition (BASC-2): scales for internalizing problems, externalizing problems,
and Behavioral Symptom Index (BSI)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis-2-chloroethyl phosphate (BCEP)
No linked references.
11581667
Domain
Metric
Rating
Comments
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric 16: Use of Biomarker of Exposure
Metric 17:
Metric 18:
Effect Biomarker
Method Sensitivity
Metric 19: Biomarker Stability
Metric 20: Sample Contamination
Metric 21: Method Requirements
Metric 22: Matrix Adjustment
Medium
N/A
Medium
Medium
Medium
Medium
Medium
Concentrations of the TCEP metabolite BCEP were quantified in urine samples ob-
tained from mothers or their children. Multiple samples were obtained from participants,
though the mean and range in number of repeated measures was not provided in this
manuscript. A single urine sample was obtained for each period of interest; ICCs for
repeated measures were 0.27 for prenatal and 0.19 for postnatal BCEP measures.
No biomarkers of effect were assessed.
The authors reported the LOD for TCEP as 0.1 ug/L, and this LOD was low enough to
detect TCEP in a sufficient percentage of samples. The authors reported the percent of
samples with detectable BCEP as ranging from 83.9% to 95.5%. No concerns of method
sensitivity are noted.
The authors note that samples were aliquoted and stored at -20 degrees Celsius until
they were shipped overnight to the CDC laboratory. Further details were not provided.
No concerns of biomarker stability were noted.
The authors did not discuss the potential for contamination in this study. However, par-
ent compounds were not measured, and there was no evidence of inappropriate handling
or processing that might lead to contamination of samples with OPE metabolites.
Appropriate methods were implemented to quantify exposure concentrations in urine
samples including automated off-line solid-phase extraction, reversed phase high-
performance liquid chromatography, and isotope dilution-electrospray ionization tandem
mass spectrometry. Total free and conjugated concentrations of metabolites were mea-
sured. A reference publication with further analytic details and quality control methods
were cited. No concerns are noted pertaining to the exposure quantification methods.
The authors reported their results using specific gravity adjusted urinary concentrations
of TCEP. All results reported are associated with this matrix-adjusted concentration.
Additional Comments: This study examined participants from the Health Outcomes and Measures of the Environment (HOME) Study, a pregnancy and birth cohort from Cincin-
nati, Ohio. It examined the association between urinary concentrations of a mixture of OPE metabolites that included BCEP, a metabolite of TCEP, and
scores from several components of BASC-2. The final analysis examined 219 mother-child pairs. A strength of this study is that authors obtained urine
samples for OPE exposure quantification at several timepoints. Associations with a mixture of three OPE metabolites measured at multiple timepoints in
pregnancy and in early childhood were reported. The direction and magnitude of associations varied, and few were statistically significant. The authors
provided sufficient details pertaining to numerous aspects of their methodology. While this study had important strengths, the lack of analyses of associa-
tions with individual OPE metabolites was a major concern. All associations were reported for a mixture of three OPE metabolites, making it impossible
to determine the extent to which there were any TCEP-specific effects. The study did not provide associations with TCEP for data extraction.
Overall Quality Determination Medium
Page 159 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581664 Table: 1 of 2
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Chen, A., Yang, W., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Xie, C., Cecil, K. M., Vuong, A. M., Xu, Y., Yolton, K. (2022).
Childhood urinary organophosphate esters and cognitive abilities in a longitudinal cohort study. Environmental Research 215(Pt 1): 114265.
Neurological/Behavioral-Cognitive abilities/IQ (WISC, FSIQ)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis-2-chloroethyl phosphate (BCEP)
No linked references.
11581664
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
High All key elements of the study design were present. Risk of participant and selection bias
is low; methods of participant selection and inclusion/exclusion criteria are well-stated.
The exposure-outcome distribution of the participants is likely representative of the
exposure-outcome distributions in the population of persons eligible for inclusion in the
study.
Medium Moderate exclusion from the analysis sample. Excluded participant demographics (i.e.,
missing IQ data) were presented and exclusion reasoning was adequately addressed.
High Key elements of the study design are reported (i.e., setting, inclusion/exclusion criteria,
and methods of participant selection), and subjects were indicated as similar by popu-
lation and timeframe recruitment using the same inclusion/exclusion criteria and were
of similar age and health status. Potential confounding and stratification variables were
addressed and controlled by statistical analysis including effect modification by both
individual- and neighborhood-level SES.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6: Temporality
Medium Exposure was consistently assessed via spot urine and blood sample using well-
established methods, sampling strategy was adequately described, and LOD was stated.
Slight potential for exposure misclassification given that method of urine collection was
based on participant's toilet-training status, but not expected to change effect estimate.
Medium The range and distribution of exposure is sufficient to develop an exposure-response
estimate and 3 or more levels of exposure (i.e., ages 1, 2, 3, and 5) are reported.
Medium Temporality for exposure is established (maternal and early life exposure), but it is un-
clear whether exposures fall within relevant exposure windows for the outcome of inter-
est.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
High The outcome was assessed using well-established methods stated as "gold standard".
Researchers assessing participant cognitive abilities were blinded, trained, and evaluated
without notice every six months. Low concern for reporting biases or outcome misclas-
sification.
High A description of all measured outcomes are reported in the methods and abstract that
allows for detailed extraction. Effect estimates are reported with a 95% confidence inter-
val, and sample size information to be used in exposure-response or descriptive analyses
are presented.
Continued on next page ...
Page 160 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581664 Table: 1 of 2
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Chen, A., Yang, W., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Xie, C., Cecil, K. M., Vuong, A. M., Xu, Y., Yolton, K. (2022).
Childhood urinary organophosphate esters and cognitive abilities in a longitudinal cohort study. Environmental Research 215(Pt 1): 114265.
Neurological/Behavioral-Cognitive abilities/IQ (WISC, FSIQ)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis-2-chloroethyl phosphate (BCEP)
No linked references.
11581664
Domain
Metric
Rating
Comments
Domain 4: Potential Confounding / Variability Control
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High Specific covariates were discussed and adjustments were made for potential confounders
(e.g. age, sex, socioeconomic status) in the final analyses through the use of statistical
models including adjustment in multivariate models and stratification. All methods to
address confounding are well-documented.
High Potential confounders were assessed using valid and reliable methodology where appro-
priate. SES was assessed at both the individual-level and neighborhood-level. Assess-
ment methodology and reasoning was well-documented.
Medium Co-exposures to pollutant that are not the target exposure were not addressed in the
study.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High
Medium
Medium
High
The study design chosen was appropriate to address the research question to assess
the associations between exposure to three OPE metabolites (including BCEP) and
cognitive ability in children. Statistical methods appropriate to address the research
question were used, including repeated measures analyses and regression modeling.
The number of participants are adequate to detect an effect in the exposed population.
Reproducibility of analyses is likely given description of analysis.
Model assumptions were adequately described, and the data fit normality after adjust-
ments.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric
16:
Use of Biomarker of Exposure
High
BCEP biomarker of interest is metabolite of TCEP.
Metric
17:
Effect Biomarker
N/A
Biomarkers of effect were not assessed.
Metric
18:
Method Sensitivity
Low
LOD is sufficient and analytical methods for measuring biomarker are adequately re-
ported. LOQ is not reported.
Metric
19:
Biomarker Stability
Medium
Samples appeared to have been handled correctly during collection, processing and
storage.
Metric
20:
Sample Contamination
Medium
Documentation of sampling is provided, but no information on contamination is re-
ported.
Metric
21:
Method Requirements
Medium
Instrumentation that allows for identification of the biomarker with a high degree of
confidence and the required sensitivity were used.
Metric
22:
Matrix Adjustment
Medium
Only adjusted results are provided.
Continued on next page ...
Page 161 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581664 Table: 1 of 2
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Chen, A., Yang, W., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Xie, C., Cecil, K. M., Vuong, A. M., Xu, Y., Yolton, K. (2022).
Childhood urinary organophosphate esters and cognitive abilities in a longitudinal cohort study. Environmental Research 215(Pt 1): 114265.
Neurological/Behavioral-Cognitive abilities/IQ (WISC, FSIQ)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis-2-chloroethyl phosphate (BCEP)
No linked references.
11581664
Domain
Metric
Rating
Comments
Additional Comments:
This study did not find any statistically significant associations between early-life urinary OPE metabolites and cognitive abilities at 8 years old. This
study collected urine from participants of a larger birth cohort study (HOMES) at ages 1, 2, 3, and 5. Blood was collected at age 5 and cognitive abilities
were measured at age 8 through the administration of an IQ test. Maternal IQ was also tested at the same time as the child participant. Biomonitoring
media and IQ collections and analytical processes were well documented and seemingly valid. Various potential covariates and confounding variables were
considered and adjusted for.
Overall Quality Determination
High
Page 162 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581664 Table: 2 of 2
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Chen, A., Yang, W., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Xie, C., Cecil, K. M., Vuong, A. M., Xu, Y., Yolton, K. (2022).
Childhood urinary organophosphate esters and cognitive abilities in a longitudinal cohort study. Environmental Research 215(Pt 1): 114265.
Neurological/Behavioral-Cognitive abilities/IQ (WISC, FSIQ)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis-2-chloroethyl phosphate (BCEP)
No linked references.
11581664
Domain
Metric
Rating
Comments
Domain 1: Study Participation
Metric 1:
Participant Selection
Metric 2: Attrition
Metric 3: Comparison Group
High All key elements of the study design were present. Risk of participant and selection bias
is low; methods of participant selection and inclusion/exclusion criteria are well-stated.
The exposure-outcome distribution of the participants is likely representative of the
exposure-outcome distributions in the population of persons eligible for inclusion in the
study.
Medium Moderate exclusion from the analysis sample. Excluded participant demographics (i.e.,
missing IQ data) were presented and exclusion reasoning was adequately addressed.
High Key elements of the study design are reported (i.e., setting, inclusion/exclusion criteria,
and methods of participant selection), and subjects were indicated as similar by popu-
lation and timeframe recruitment using the same inclusion/exclusion criteria and were
of similar age and health status. Potential confounding and stratification variables were
addressed and controlled by statistical analysis including effect modification by both
individual- and neighborhood-level SES.
Domain 2: Exposure Characterization
Metric 4: Measurement of Exposure
Metric 5: Exposure Levels
Metric 6: Temporality
Medium Exposure was consistently assessed via spot urine and blood sample using well-
established methods, sampling strategy was adequately described, and LOD was stated.
Slight potential for exposure misclassification given that method of urine collection was
based on participant's toilet-training status, but not expected to change effect estimate.
Medium The range and distribution of exposure is sufficient to develop an exposure-response
estimate and 3 or more levels of exposure (i.e., ages 1, 2, 3, and 5) are reported.
Medium Temporality for exposure is established (maternal and early life exposure), but it is un-
clear whether exposures fall within relevant exposure windows for the outcome of inter-
est.
Domain 3: Outcome Assessment
Metric 7:
Metric 8:
Outcome Measurement or
Characterization
Reporting Bias
High The outcome was assessed using well-established methods stated as "gold standard".
Researchers assessing participant cognitive abilities were blinded, trained, and evaluated
without notice every six months. Low concern for reporting biases or outcome misclas-
sification.
High A description of all measured outcomes are reported in the methods and abstract that
allows for detailed extraction. Effect estimates are reported with a 95% confidence inter-
val, and sample size information to be used in exposure-response or descriptive analyses
are presented.
Domain 4: Potential Confounding / Variability Control
Continued on next page ...
Page 163 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581664 Table: 2 of 2
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Chen, A., Yang, W., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Xie, C., Cecil, K. M., Vuong, A. M., Xu, Y., Yolton, K. (2022).
Childhood urinary organophosphate esters and cognitive abilities in a longitudinal cohort study. Environmental Research 215(Pt 1): 114265.
Neurological/Behavioral-Cognitive abilities/IQ (WISC, FSIQ)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis-2-chloroethyl phosphate (BCEP)
No linked references.
11581664
Domain
Metric
Rating
Comments
Metric 9: Covariate Adjustment
Metric 10: Covariate Characterization
Metric 11: Co-exposure Counfounding
High Specific covariates were discussed and adjustments were made for potential confounders
(e.g. age, sex, socioeconomic status) in the final analyses through the use of statistical
models including adjustment in multivariate models and stratification. All methods to
address confounding are well-documented.
High Potential confounders were assessed using valid and reliable methodology where appro-
priate. SES was assessed at both the individual-level and neighborhood-level. Assess-
ment methodology and reasoning was well-documented.
Medium Co-exposures to pollutant that are not the target exposure were not addressed in the
study.
Domain 5: Analysis
Metric 12: Study Design and Methods
Metric 13: Statistical Power
Metric 14: Reproducibility of Analyses
Metric 15: Statistical Analysis
High
Medium
Medium
High
The study design chosen was appropriate to address the research question to assess
the associations between exposure to three OPE metabolites (including BCEP) and
cognitive ability in children. Statistical methods appropriate to address the research
question were used, including repeated measures analyses and regression modeling.
The number of participants are adequate to detect an effect in the exposed population.
Reproducibility of analyses is likely given description of analysis.
Model assumptions were adequately described, and the data fit normality after adjust-
ments.
Domain 6: Other (if applicable) Considerations for Biomarker Selection and Measurement (Lakind et al. 2014)
Metric
16:
Use of Biomarker of Exposure
High
BCEP biomarker of interest is metabolite of TCEP.
Metric
17:
Effect Biomarker
N/A
Biomarkers of effect were not assessed.
Metric
18:
Method Sensitivity
Low
LOD is sufficient and analytical methods for measuring biomarker are adequately re-
ported. LOQ is not reported.
Metric
19:
Biomarker Stability
Medium
Samples appeared to have been handled correctly during collection, processing and
storage.
Metric
20:
Sample Contamination
Medium
Documentation of sampling is provided, but no information on contamination is re-
ported.
Metric
21:
Method Requirements
Medium
Instrumentation that allows for identification of the biomarker with a high degree of
confidence and the required sensitivity were used.
Metric
22:
Matrix Adjustment
Medium
Only adjusted results are provided.
Continued on next page ...
Page 164 of 165
-------
Tris(2-chloroethyl) phosphate (TCEP)
Human Health Hazard Epidemology Evaluation
HERO ID: 11581664 Table: 2 of 2
... continued from previous page
Study Citation:
Health
Outcome(s) and
Reported Health
Effect(s):
Chemical:
Linked HERO ID(s):
HERO ID:
Percy, Z., Chen, A., Yang, W., Braun, J. M., Lanphear, B., Ospina, M., Calafat, A. M., Xie, C., Cecil, K. M., Vuong, A. M., Xu, Y., Yolton, K. (2022).
Childhood urinary organophosphate esters and cognitive abilities in a longitudinal cohort study. Environmental Research 215(Pt 1): 114265.
Neurological/Behavioral-Cognitive abilities/IQ (WISC, FSIQ)
Tris(2-chloroethyl) phosphate (TCEP)- Metabolite: Bis-2-chloroethyl phosphate (BCEP)
No linked references.
11581664
Domain
Metric
Rating
Comments
Additional Comments:
This study did not find any statistically significant associations between early-life urinary OPE metabolites and cognitive abilities at 8 years old. This
study collected urine from participants of a larger birth cohort study (HOMES) at ages 1, 2, 3, and 5. Blood was collected at age 5 and cognitive abilities
were measured at age 8 through the administration of an IQ test. Maternal IQ was also tested at the same time as the child participant. Biomonitoring
media and IQ collections and analytical processes were well documented and seemingly valid. Various potential covariates and confounding variables were
considered and adjusted for.
Overall Quality Determination
High
Page 165 of 165
------- |