EPA/600/R-13/098F | August 2013 | www.epa.gov/ncea
United States
Environmental Protection
Agency
Fish Consumption in
Connecticut, Florida,
Minnesota, and North Dakota
Office of Research and Development
National Center for Environmental Assessment, Washington, DC
-------
EPA/600/R-13/098F
July 2013
www.epa.gov/ncea
Fish Consumption in Connecticut, Florida, Minnesota, and
North Dakota
National Center for Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
ABSTRACT
Fish consumption rates derived from national surveys may not accurately reflect
consumption rates in a particular population such as recreational anglers. Many state and local
health agencies in the United States have conducted area-specific surveys to study fish
consumption patterns in local populations, assess exposure to environmental contaminants, or
evaluate compliance with fish advisories. The National Center for Environmental Assessment
(NCEA) of the Environmental Protection Agency's (EPA) Office of Research and Development
(ORD) has conducted an analysis of data from fish consumption surveys from the states of
Connecticut, Florida, Minnesota, and North Dakota. The primary objective of the analysis was
to identify populations within these state that consume more fish and shellfish than either the
state's or the Nation's general population and thus may be at higher risk from exposure to
contaminants in fish. EPA was particularly interested in estimating each state's fish and shellfish
consumption for recreational anglers, low income populations, children, and ethnic groups. The
report provides distribution offish consumption rates for different age cohorts, ethnic groups,
socioeconomic status, fish types (i.e., freshwater, marine, estuarine), and fish sources (i.e., store-
bought versus self-caught).
Fish and shellfish intake for those who consume both self-caught and store-bought fish is
higher than for those who reported eating only bought or only self-caught. In Connecticut, mean
fish consumption per kilogram of bodyweight ranged from 0.23 g/kg-day to 0.84 g/kg-day. In
Minnesota, mean fish consumption per kilogram of bodyweight ranged from 0.11 g/kg-day to
0.69 g/kg-day. The highest values observed in the Connecticut and Minnesota data corresponded
to females 16-29 years of age. In Florida, mean fish consumption per kilogram of bodyweight
ranged from 0.64 g/kg-day to 2.34 g/kg-day. In North Dakota, mean fish consumption per
kilogram of bodyweight ranged from 0.20 g/kg-day to 0.70 g/kg-day. The highest values
observed in the Florida and North Dakota data corresponded to children 1-5 years of age. The
Florida data showed a statistically significant increase in the percentage of the population that
reported eating fish and shellfish with increased household income and education. This trend
was not observed in the other states. Some minor differences were observed between Whites,
ii
-------
non-Hispanic and other ethnic groups, but these differences were not statistically significant and
not consistent across states. Because of differences in survey methodologies for the collection,
processing, and analysis of the data, comparisons across states should be interpreted with
caution.
Preferred Citation:
U.S. EPA (Environmental Protection Agency). (2013) Fish consumption in Connecticut, Florida, Minnesota, and
North Dakota. National Center for Environmental Assessment, Washington, DC; EPA/600/R-13/098F. Available
online at http://epa.gov/ncea.
in
-------
CONTENTS
LIST OF TABLES vi
LIST OF FIGURES xxi
PREFACE xxiii
AUTHORS, CONTRIBUTORS, AND REVIEWERS xxiv
EXECUTIVE SUMMARY ES-1
1. INTRODUCTION 1-1
1.1. OVERVIEW OF THE SURVEYS 1-1
1.2. DIFFERENCES AMONG THE SURVEYS 1-5
1.3. OVERVIEW OF THE REPORT 1-6
2. METHODS AND STATISTICAL ANALYSES 2-1
2.1. CONVERSIONS BETWEEN RAW AND AS-CONSUMED WEIGHT 2-1
2.1.1. Florida Survey 2-1
2.1.2. Connecticut Survey 2-2
2.1.3. Minnesota and North Dakota Survey 2-3
2.2. VARIABLES USED FOR THE DATA ANALYSIS 2-3
2.2.1. Dependent Variables 2-4
2.2.2. Independent Variables 2-4
2.3. IDENTIFICATION AND HANDLING OF OUTLIERS 2-7
2.4. SAMPLE SIZES 2-15
2.5. SUMMARY STATISTICS AND STATISTICAL ANALYSIS METHODS 2-17
2.5.1. Distribution of the Consumption Data 2-18
2.5.2. CalculationofPercentiles 2-21
2.5.3. Calculation of Confidence Intervals and Significance Tests 2-22
2.6. ANALYSIS WEIGHTS 2-23
3. RESULTS AND DISCUSSION 3-1
3.1. DIFFERENCES BY INCOME 3-1
3.2. DIFFERENCES BY GENDER AND AGE 3-4
3.3. DIFFERENCES BY RACE-ETHNICITY AND TARGETED POPULATIONS.... 3-13
3.4. DIFFERENCES BY EDUCATION 3-17
3.5. DIFFERENCES BY FISH AND SHELLFISH SOURCE 3-18
3.6. PROCEDURAL DIFFERENCES AMONG SURVEYS 3-20
3.7. COMPARISONS AMONG STATES 3-23
3.8. LIMITATIONS OF THE ANALYSIS 3-26
4. REFERENCES 4-1
APPENDIX A. DATABASE DEVELOPMENT A-l
A.I. INTRODUCTION A-l
A.2. FLORIDA DATA A-2
A.2.1. Description of the Survey and the Original Data Files A-2
A.2.2. Processing of the Data Files A-3
iv
-------
CONTENTS (continued)
A. 3. CONNECTICUT DAT A A-ll
A.3.1. Description of the Survey and the Original Data Files A-ll
A.3.2. Processing the Data Files A-13
A.4. MINNESOTA AND NORTH DAKOTA DATA A-23
A.4.1. Description of the Survey and the Original Data Files A-23
A. 5. COMBINED DAT A A-26
A.5.1. Description of the Combined Data Files A-26
A.6. CALCULATION OF WEIGHTS A-34
APPENDIX B. QA/QC PROCESS B-l
APPENDIX C. CALCULATION OF PERCENTERS C-l
APPENDIX D. GLOSSARY D-l
APPENDIX E. TABLES OF FISH AND SHELLFISH CONSUMPTION E-l
E.I. ORGANIZATION OF THE TABLES E-l
E.2. MEAN FISH AND SHELLFISH CONSUMPTION FOR THE STATE'S
GENERAL POPULATION, AS-CONSUMED WEIGHT PER DAY E-3
E.3. MEAN FISH AND SHELLFISH CONSUMPTION FOR THE GENERAL
POPULATION, AS-CONSUMED WEIGHT PER KILOGRAM
BODYWEIGHT PER DAY E-7
E.4. PERCENTAGE OF THE GENERAL POPULATION RESPONDENTS
THAT REPORTED EATING FISH AND SHELLFISH E-13
E.5. MEAN FISH AND SHELLFISH CONSUMPTION FOR THE GENERAL
POPULATION, CONSUMERS ONLY, AS-CONSUMED WEIGHT PER
DAY E-15
E.6. MEAN FISH AND SHELLFISH CONSUMPTION FOR THE GENERAL
POPULATION, CONSUMERS ONLY, AS-CONSUMED WEIGHT PER
KILOGRAM BODYWEIGHT PER DAY E-20
E.7. GEOMETRIC MEAN FISH AND SHELLFISH CONSUMPTION FOR THE
GENERAL POPULATION, CONSUMERS ONLY, AS-CONSUMED
WEIGHT PER DAY E-27
E.8. GEOMETRIC MEAN FISH AND SHELLFISH CONSUMPTION FOR THE
GENERAL POPULATION, CONSUMERS ONLY, AS-CONSUMED
WEIGHT PER KILOGRAM BODYWEIGHT PER DAY E-32
E.9. PLOTS OF CONFIDENCE INTERVALS FOR THE GENERAL
POPULATION E-40
E. 10. FISH AND SHELLFISH CONSUMPTION FOR TARGETED
POPULATIONS E-62
E.ll. SPECIES EATEN AND CAUGHT E-66
E.12. ADDITIONAL DETAILS FOR TARGETED POPULATIONS E-73
-------
LIST OF TABLES
1-1. Summary offish consumption surveys 1-2
2-1. Factors for converting the Florida data from uncooked to as-consumed
weights 2-2
2-2. Targeted populations 2-7
2-3. State level estimates that changed by more than 3% as a result of removing
outliers 2-15
2-4. Sample sizes for the general population and targeted populations 2-16
2-5. Weighted totals for the general population, in thousands 2-16
2-6. Skewness of consumption variables, general population (as-consumed weight).... 2-18
3-1. Percent eating fish and shellfish, per capita, by state and income (with 95%
CIs) 3-2
3-2. Geometric mean consumption, consumers only, by state and income (as-
consumed g/day with 95% CIs) 3-3
3-3. Mean consumption, consumers only, by state and income (as-consumed g/day
with 95% CIs) 3-4
3-4. Percent eating fish and shellfish, per capita, by state and age/gender (with
95% CIs) 3-5
3-5. Mean consumption, consumers only, by state and age/gender (as-consumed
g/day) (with 95% CIs) 3-7
3-6. Mean consumption per bodyweight, general population, per capita, by state
and age/gender (as-consumed g/kg-day) (with 95% CIs) 3-9
3-7. Percent eating fish and shellfish, per capita, by state and gender (with 95%
CIs) 3-11
3-8. Mean consumption, general population, per capita, by state and gender (as-
consumed g/day with 95% CIs) 3-11
3-9. Mean consumption per bodyweight, general population, per capita, by state
and gender (as-consumed g/kg-day with 95% CIs) 3-12
3-10. Percent eating fish and shellfish, per capita, by state and race-ethnicity (with
95% CIs) 3-14
VI
-------
LIST OF TABLES (continued)
3-11. Mean consumption, consumers only, by state and race-ethnicity (as-consumed
g/day with 95% CIs) 3-15
3-12. Mean consumption, consumers only, by state and targeted populations (as
consumed, g/day with 95% CIs) 3-16
3-13. Mean consumption, per capita, by state and targeted populations (as consumed,
g/day with 95% CIs) 3-17
3-14. Percent eating fish and shellfish, per capita, by state and education (with 95%
CIs) 3-18
3-15. Mean consumption, consumers only, by state and angler status (as-consumed
g/day with 95% CIs) 3-19
3-16. Illustrative results for two recall periods 3-21
3-17. Percent eating fish and shellfish, per capita, by state and subpopulation (with
95% CIs) 3-24
A-1. Abbreviations used throughout the data documentation report A-1
A-2. RSA/Household records with inconsistent and missing values (out of 8,740
records) A-6
A-3. Individual records with inconsistent and missing values (out of 17,213
records) A-6
A-4. Fish consumption records with inconsistent and missing values (out of 16,099
records) A-6
A-5. Annual seafood consumption estimates for the two samples (kg/yr) A-8
A-6. Factors for converting the Florida data from uncooked to as-consumed
weights A-10
A-7. Variables with differences between the INDIV and GDAY files for the same
household (out of 827 households) A-12
A-8. Sample sizes by population from the processed files and Table 5 of the study
report A-17
A-9. Total fish consumption (g/day) by population from the processed files and
Table 11 of the Study report A-18
vn
-------
LIST OF TABLES (continued)
A-10. Factors for converting the Connecticut data from as-consumed to uncooked
weights A-19
A-ll. Crosswalk of variables in original and combined data files A-27
A-12. Number of records in the processed data files A-28
A-13. Common income categories A-29
A-14. Common education categories A-30
A-15. Common race-ethnicity categories A-31
C-l. Example data C-2
C-2. Percentile estimates from the macro and from PROC Univariate for the
example data C-3
C-3. Percentile estimates from the macro and from PROC Univariate for the
example data C-5
D-1. Abbreviations used throughout the data documentation report D-l
E-l. Mean consumption, general population, per capita, by state and adult/child
(as-consumed g/day with 95% CIs) E-3
E-2. Mean consumption, general population, per capita, by state and age/gender (5
categories) (as-consumed g/day with 95% CIs) E-4
E-3. Mean consumption, general population, per capita, by state and education (as-
consumed g/day with 95% CIs) E-5
E-4. Mean consumption, general population, per capita, by state and income (as-
consumed g/day with 95% CIs) E-6
E-5. Mean consumption, general population, per capita, by state and race-ethnicity
(as-consumed g/day with 95% CIs) E-7
E-6. Mean consumption per bodyweight, general population, per capita, by state
and adult/child (as-consumed g/kg-day with 95% CIs) E-8
E-7. Mean consumption per bodyweight, general population, per capita, by state
and age/gender (9 categories) (as-consumed g/kg-day with 95% CIs) E-9
E-8. Mean consumption per bodyweight, general population, per capita, by state
and age/gender (5 categories) (as-consumed g/kg-day with 95% CIs) E-10
viii
-------
LIST OF TABLES (continued)
E-9. Mean consumption per bodyweight, general population, per capita, by state
and education (as-consumed g/kg-day with 95% CIs) E-ll
E-10. Mean consumption per bodyweight, general population, per capita, by state
and income (as-consumed g/kg-day with 95% CIs) E-12
E-ll. Mean consumption per bodyweight, general population, per capita, by state
and race-ethnicity (as-consumed g/kg-day with 95% CIs) E-13
E-12. Percent eating fish and shellfish, per capita, by state and adult/child (with 95%
CIs) E-14
E-13. Percent eating fish and shellfish, per capita, by state and age/gender (5
categories) (with 95% CIs) E-15
E-14. Mean consumption, consumers only, by state and adult/child (as-consumed
g/day with 95% CIs) E-16
E-15. Mean consumption, consumers only, by state and age/gender (5 categories)
(as-consumed g/day with 95% CIs) E-17
E-16. Mean consumption, consumers only, by state and education (as-consumed
g/day with 95% CIs) E-18
E-17. Mean consumption, consumers only, by state and income (as-consumed g/day
with 95% CIs) E-19
E-18. Mean consumption, consumers only, by state and gender (as-consumed g/day
with 95% CIs) E-19
E-19. Mean consumption per bodyweight, general population, consumers only, by
state and adult/child (as-consumed g/kg-day with 95% CIs) E-20
E-20. Mean consumption per bodyweight, general population, consumers only, by
state and age/gender (9 categories) (as-consumed g/kg-day with 95% CIs) E-21
E-21. Mean consumption per bodyweight, general population, consumers only, by
state and age/gender (5 categories) (as-consumed g/kg-day with 95% CIs) E-22
E-22. Mean consumption per bodyweight, general population, consumers only, by
state and education (as-consumed g/kg-day with 95% CIs) E-23
E-23. Mean consumption per bodyweight, general population, consumers only, by
state and income (as-consumed g/kg-day with 95% CIs) E-24
IX
-------
LIST OF TABLES (continued)
E-24. Mean consumption per bodyweight, general population, consumers only, by
state and race-ethnicity (as-consumed g/kg-day with 95% CIs) E-25
E-25. Mean consumption per bodyweight, general population, consumers only, by
state and gender (as-consumed g/kg-day with 95% CIs) E-26
E-26. Mean consumption per bodyweight, general population, consumers only, by
state and angler status (as-consumed g/kg-day with 95% CIs) E-26
E-27. Geometric mean consumption, general population, consumers only, by state
and adult/child (as-consumed g/day with 95% CIs) E-27
E-28. Geometric mean consumption, general population, consumers only, by state
and age/gender (9 categories) (as-consumed g/day with 95% CIs) E-28
E-29. Geometric mean consumption, general population, consumers only, by state
and education (as-consumed g/day with 95% CIs) E-29
E-30. Geometric mean consumption, general population, consumers only, by state
and income (as-consumed g/day with 95% CIs) E-30
E-31. Geometric mean consumption, general population, consumers only, by state
and race-ethnicity (as-consumed g/day with 95% CIs) E-31
E-32. Geometric mean consumption, general population, consumers only, by state
and gender (as-consumed g/day with 95% CIs) E-31
E-33. Geometric mean consumption, general population, consumers only, by state
and angler status (as-consumed g/day with 95% CIs) E-32
E-34. Geometric mean consumption per bodyweight, general population, consumers
only, by state and adult/child (as-consumed g/kg-day with 95% CIs) E-33
E-3 5. Geometric mean consumption per bodyweight, general population, consumers
only, by State and age/gender (9 categories) (as-consumed g/kg-day with 95%
CIs) E-34
E-36. Geometric mean consumption per bodyweight, general population, consumers
only, by state and age/gender (5 categories) (as-consumed g/kg-day with 95%
CIs) E-35
E-37. Geometric mean consumption per bodyweight, general population, consumers
only, by state and education (as-consumed g/kg-day with 95% CIs) E-36
E-3 8. Geometric mean consumption per bodyweight, general population, consumers
only, by state and income (as-consumed g/kg-day with 95% CIs) E-37
-------
LIST OF TABLES (continued)
E-39. Geometric mean consumption per bodyweight, general population, consumers
only, by state and race-ethnicity (as-consumed g/kg-day with 95% CIs) E-38
E-40. Geometric mean consumption per bodyweight, general population, consumers
only, by state and gender (as-consumed g/kg-day with 95% CIs) E-39
E-41. Geometric mean consumption per bodyweight, general population, consumers
only, by state and angler status (as-consumed g/kg-day with 95% CIs) E-40
E-42. Mean consumption per bodyweight, per capita, by state and subpopulation
(as-consumed g/kg-day with 95% CIs) E-63
E-43. Mean consumption per bodyweight, consumers only, by state and
subpopulation (as-consumed g/kg-day with 95% CIs) E-64
E-44. Geometric mean consumption, consumers only, by state and subpopulation
(as-consumed g/day with 95% CIs) E-65
E-45. Geometric mean consumption per bodyweight, consumers only, by state and
subpopulation (as-consumed g/kg-day with 95% CIs) E-66
E-46. Total and caught fish consumption for the Connecticut general population
(weighted, as-consumed g/day) E-67
E-47. Total and caught fish consumption for the Florida general population
(weighted, as-consumed g/day) E-69
E-48. Total and caught fish consumption for the Minnesota general population
(weighted, as-consumed g/day) E-71
E-49. Total and caught fish consumption for the North Dakota general population
(weighted, as-consumed g/day) E-72
E-50. Consumption, per capita, by state and income (as-consumed g/day) E-76
E-51. Consumption, consumers only, by state and income (as-consumed g/day) E-77
E-52. Consumption, per capita, by state and race-ethnicity (as-consumed g/day) E-78
E-53. Consumption, consumers only, by state and race-ethnicity (as-consumed
g/day) E-79
E-54. Consumption, per capita, by state, adult/child and race-ethnicity (as-consumed
g/day) E-80
XI
-------
LIST OF TABLES (continued)
E-55. Consumption, consumers only, by state, adult/child, and race-ethnicity (as-
consumed g/day) E-82
E-56. Consumption, per capita, by state and gender (as-consumed g/day) E-84
E-57. Consumption, consumers only, by state and gender (as-consumed g/day) E-85
E-58. Consumption, per capita, by state, adult/child and gender (as-consumed g/day)....E-86
E-59. Consumption, consumers only, by state, adult/child and gender (as-consumed
g/day) E-87
E-60. Consumption, per capita, by state and education (as-consumed g/day) E-88
E-61. Consumption, consumers only, by state and education (as-consumed g/day) E-89
E-62. Consumption, per capita, by state and age-gender category (as-consumed
g/day) E-90
E-63. Consumption, consumers only, by state and age-gender category (as-
consumed g/day) E-92
E-64. Consumption, per capita, by state and age-gender category (as-consumed
g/day) E-94
E-65. Consumption, consumers only, by state and age-gender category (as-
consumed g/day) E-95
E-66. Consumption, per capita, by state and acquisition method (as-consumed
g/day) E-96
E-67. Consumption, consumers only, by state and acquisition method (as-consumed
g/day) E-97
E-68. Consumption, per capita, by state acquisition method, and income (as-
consumed g/day) E-98
E-69. Consumption, consumers only, by state, acquisition method, and income (as-
consumed g/day) E-100
E-70. Consumption, per capita, by state and habitat (as-consumed g/day) E-102
E-71. Consumption, consumers only, by state and habitat (as-consumed g/day) E-103
E-72. Consumption, per capita, by state and fish/shellfish type (as-consumed g/day) ...E-104
xn
-------
LIST OF TABLES (continued)
E-73. Consumption, consumers only, by state and fish/shellfish type (as-consumed
g/day) E-105
E-74. Consumption, consumers only, by state and type offish consumed (as-
consumed g/day) E-106
E-75. Consumption, consumers only, by state and fresh/estuarine fish consumption
(as-consumed g/day) E-107
E-76. Fish consumption, per capita, by state and income (uncooked g/day) E-108
E-77. Fish consumption, consumers only, by state and income (uncooked g/day) E-109
E-78. Fish consumption, per capita, by state and race-ethnicity (uncooked g/day) E-l 10
E-79. Fish consumption, consumers only, by state and race-ethnicity (uncooked
g/day) E-l 11
E-80. Fish consumption, per capita, by state, adult/child and race-ethnicity
(uncooked g/day) E-l 12
E-81. Fish consumption, consumers only, by state, adult/child, and race-ethnicity
(uncooked g/day) E-l 14
E-82. Fish consumption, per capita, by state and gender (uncooked g/day) E-l 16
E-83. Fish consumption, consumers only, by state and gender (uncooked g/day) E-l 17
E-84. Fish consumption, per capita, by state, adult/child and gender (uncooked
g/day) E-l 18
E-85. Fish consumption, consumers only, by state, adult/child and gender (uncooked
g/day) E-l 19
E-86. Fish consumption, per capita, by state and education (uncooked g/day) E-120
E-87. Fish consumption, consumers only, by state and education (uncooked g/day) E-121
E-88. Fish consumption, per capita, by state and age-gender category (uncooked
g/day) E-122
E-89. Fish consumption, consumers only, by state and age-gender category
(uncooked g/day) E-124
E-90. Fish consumption, per capita, by state and age-gender category (uncooked
g/day) E-126
xiii
-------
LIST OF TABLES (continued)
E-91. Fish consumption, consumers only, by state and age-gender category
(uncooked g/day) E-127
E-92. Fish consumption, per capita, by state and acquisition method (uncooked
g/day) E-128
E-93. Fish consumption, consumers only, by state and acquisition method
(uncooked g/day) E-129
E-94. Fish consumption, per capita, by state acquisition method, and income
(uncooked g/day) E-130
E-95. Fish consumption, consumers only, by state, acquisition method, and income
(uncooked g/day) E-132
E-96. Fish consumption, per capita, by state and habitat (uncooked g/day) E-134
E-97. Fish consumption, consumers only, by state and habitat (uncooked g/day) E-135
E-98. Fish consumption, per capita, by state and fish/shellfish type (uncooked g/day)..E-136
E-99. Fish consumption, consumers only, by state and fish/shellfish type (uncooked
g/day) E-137
E-100. Fish consumption, consumers only, by state and type offish consumed
(uncooked g/day) E-138
E-101. Fish consumption, consumers only, by state and fresh/estuarine fish
consumption (uncooked g/day) E-139
E-102. Fish consumption per kg, per capita, by state and income (as-consumed g/kg-
day) E-140
E-103. Fish consumption per kg, consumers only, by state and income (as-consumed
g/kg-day) E-141
E-104. Fish consumption per kg, per capita, by state and race-ethnicity (as-consumed
g/kg-day) E-142
E-105. Fish consumption per kg, consumers only, by state and race-ethnicity (as-
consumed g/kg-day) E-143
E-106. Fish consumption per kg, per capita, by state, adult/child and race-ethnicity
(as-consumed g/kg-day) E-144
xiv
-------
LIST OF TABLES (continued)
E-107. Fish consumption per kg, consumers only, by state, adult/child, and race-
ethnicity (as-consumed g/kg-day) E-146
E-108. Fish consumption per kg, per capita, by state and gender (as-consumed g/kg-
day) E-148
E-109. Fish consumption per kg, consumers only, by state and gender (as-consumed
g/kg-day) E-149
E-l 10. Fish consumption per kg, per capita, by state, adult/child and gender (as-
consumed g/kg-day) E-l 50
E-l 11. Fish consumption per kg, consumers only, by state, adult/child and gender (as-
consumed g/kg-day) E-l 51
E-l 12. Fish consumption per kg, per capita, by state and education (as-consumed
g/kg-day) E-152
E-l 13. Fish consumption per kg, consumers only, by state and education (as-
consumed g/kg-day) E-l 53
E-l 14. Fish consumption per kg, per capita, by state and age-gender category (as-
consumed g/kg-day) E-l 54
E-l 15. Fish consumption per kg, consumers only, by state and age-gender category
(as-consumed g/kg-day) E-l56
E-l 16. Fish consumption per kg, per capita, by state and age-gender category (as-
consumed g/kg-day) E-l 58
E-l 17. Fish consumption per kg, consumers only, by state and age-gender category
(as-consumed g/kg-day) E-l59
E-l 18. Fish consumption per kg, per capita, by state and acquisition method (as-
consumed g/kg-day) E-160
E-l 19. Fish consumption per kg, consumers only, by state and acquisition method
(as-consumed g/kg-day) E-l 61
E-120. Fish consumption per kg, per capita, by state acquisition method, and income
(as-consumed g/kg-day) E-162
E-121. Fish consumption per kg, consumers only, by state, acquisition method, and
income (as-consumed g/kg-day) E-l64
xv
-------
LIST OF TABLES (continued)
E-122. Fish consumption per kg, per capita, by state and habitat (as-consumed g/kg-
day) E-166
E-123. Fish consumption per kg, consumers only, by state and habitat (as-consumed
g/kg-day) E-167
E-124. Fish consumption per kg, per capita, by state and fish/shellfish type (as-
consumed g/kg-day) E-168
E-125. Fish consumption per kg, consumers only, by state and fish/shellfish type (as-
consumed g/kg-day) E-169
E-126. Fish consumption per kg, consumers only, by state and type offish consumed
(as-consumed g/kg-day) E-170
E-127. Fish consumption per kg, consumers only, by state and fresh/estuarine fish
consumption (as-consumed g/kg-day) E-171
E-128. Fish consumption per kg, per capita, by state and income (uncooked g/kg-day)..E-172
E-129. Fish consumption per kg, consumers only, by state and income (uncooked
g/kg-day) E-173
E-130. Fish consumption per kg, per capita, by state and race-ethnicity (uncooked
g/kg-day) E-174
E-131. Fish consumption per kg, consumers only, by state and race-ethnicity
(uncooked g/kg-day) E-175
E-132. Fish consumption per kg, per capita, by state, adult/child and race-ethnicity
(uncooked g/kg-day) E-176
E-133. Fish consumption per kg, consumers only, by state, adult/child, and race-
ethnicity (uncooked g/kg-day) E-178
E-134. Fish consumption per kg, per capita, by state and gender (uncooked g/kg-day)...E-180
E-135. Fish consumption per kg, consumers only, by state and gender (uncooked
g/kg-day) E-181
E-136. Fish consumption per kg, per capita, by state, adult/child and gender
(uncooked g/kg-day) E-182
E-137. Fish consumption per kg, consumers only, by state, adult/child and gender
(uncooked g/kg-day) E-183
xvi
-------
LIST OF TABLES (continued)
E-138. Fish consumption per kg, per capita, by state and education (uncooked g/kg-
day) E-184
E-139. Fish consumption per kg, consumers only, by state and education (uncooked
g/kg-day) E-185
E-140. Fish consumption per kg, per capita, by state and age-gender category
(uncooked g/kg-day) E-186
E-141. Fish consumption per kg, consumers only, by state and age-gender category
(uncooked g/kg-day) E-188
E-142. Fish consumption per kg, per capita, by state and age-gender category
(uncooked g/kg-day) E-190
E-143. Fish consumption per kg, consumers only, by state and age-gender category
(uncooked g/kg-day) E-191
E-144. Fish consumption per kg, per capita, by state and acquisition method
(uncooked g/kg-day) E-192
E-145. Fish consumption per kg, consumers only, by state and acquisition method
(uncooked g/kg-day) E-193
E-146. Fish consumption per kg, per capita, by state acquisition method, and income
(uncooked g/kg-day) E-194
E-147. Fish consumption per kg, consumers only, by state, acquisition method, and
income (uncooked g/kg-day) E-196
E-148. Fish consumption per kg, per capita, by state and habitat (uncooked g/kg-day)...E-198
E-149. Fish consumption per kg, consumers only, by state and habitat (uncooked
g/kg-day) E-199
E-150. Fish consumption per kg, per capita, by state and fish/shellfish type (uncooked
g/kg-day) E-200
E-151. Fish consumption per kg, consumers only, by state and fish/shellfish type
(uncooked g/kg-day) E-201
E-152. Fish consumption per kg, consumers only, by state and type offish consumed
(uncooked g/kg-day) E-202
E-153. Fish consumption per kg, consumers only, by state and fresh/estuarine fish
consumption (uncooked g/kg-day) E-203
xvii
-------
LIST OF TABLES (continued)
E-154. Fish consumption, per capita, by state and subpopulation (as-consumed g/day) ..E-204
E-155. Fish consumption, consumers only, by state and subpopulation (as-consumed
g/day) E-205
E-156. Fish consumption, per capita, by state, subpopulation, and adult/child (as-
consumed g/day) E-206
E-157. Fish consumption, consumers only, state, subpopulation, and adult/child (as-
consumed g/day) E-208
E-158. Fish consumption, per capita, by state, subpopulation, and gender (as-
consumed g/day) E-210
E-159. Fish consumption, consumers only, state, subpopulation, and gender (as-
consumed g/day) E-212
E-160. Fish consumption, consumers only, by state, subpopulation, and
fresh/estuarine fish consumption (as-consumed g/day) E-214
E-161. Fish consumption, per capita, by state and subpopulation (as-consumed g/day) ..E-216
E-162. Fish consumption, consumers only, by state and subpopulation (as-consumed
g/day) E-217
E-163. Fish consumption, per capita, by state, subpopulation, and adult/child (as-
consumed g/day) E-218
E-164. Fish consumption, consumers only, state, subpopulation, and adult/child (as-
consumed g/day) E-220
E-165. Fish consumption, per capita, by state, subpopulation, and gender (as-
consumed g/day) E-222
E-166. Fish consumption, consumers only, state, subpopulation, and gender (as-
consumed g/day) E-224
E-167. Fish consumption, consumers only, by state, subpopulation, and
fresh/estuarine fish consumption (as-consumed g/day) E-226
E-168. Fish consumption per kg, per capita, by state and subpopulation (as-consumed
g/kg-day) E-228
E-169. Fish consumption per kg, consumers only, by state and subpopulation (as-
consumed g/kg-day) E-229
xvin
-------
LIST OF TABLES (continued)
E-170. Fish consumption per kg, per capita, by state, subpopulation, and adult/child
(as-consumed g/kg-day) E-230
E-171. Fish consumption per kg, consumers only, state, subpopulation, and
adult/child (as-consumed g/kg-day) E-232
E-172. Fish consumption per kg, per capita, by state, subpopulation, and gender (as-
consumed g/kg-day) E-234
E-173. Fish consumption per kg, consumers only, state, subpopulation, and gender
(as-consumed g/kg-day) E-236
E-174. Fish consumption per kg, consumers only, by state, subpopulation, and
fresh/estuarine fish consumption (as-consumed g/kg-day) E-238
E-175. Fish consumption per kg, per capita, by state and subpopulation (as-consumed
g/kg-day) E-240
E-176. Fish consumption per kg, consumers only, by state and subpopulation (as-
consumed g/kg-day) E-241
E-177. Fish consumption per kg, per capita, by state, subpopulation, and adult/child
(as-consumed g/kg-day) E-242
E-178. Fish consumption per kg, consumers only, state, subpopulation, and
adult/child (as-consumed g/kg-day) E-244
E-179. Fish consumption per kg, per capita, by state, subpopulation, and gender (as-
consumed g/kg-day) E-246
E-180. Fish consumption per kg, consumers only, state, subpopulation, and gender
(as-consumed g/kg-day) E-248
E-181. Fish consumption per kg, consumers only, by state, subpopulation, and
fresh/estuarine fish consumption (as-consumed g/kg-day) E-250
E-182. Total and caught fish consumption for the Connecticut angler population (as-
consumed g/day) E-252
E-183. Total and caught fish consumption for the Connecticut aquaculture students
population (as-consumed g/day) E-254
E-184. Total and caught fish consumption for the Connecticut Asian population (as-
consumed g/day) E-255
xix
-------
LIST OF TABLES (continued)
E-185. Total and caught fish consumption for the Connecticut commercial fishermen
population (as-consumed g/day) E-257
E-186. Total and caught fish consumption for the Connecticut EFNEP participant
population (as-consumed g/day) E-259
E-187. Total and caught fish consumption for the Connecticut WIC participant
population (as-consumed g/day) E-260
E-188. Total and caught fish consumption for the Minnesota American Indian
population (as-consumed g/day) E-262
E-189. Total and caught fish consumption for the Minnesota angler population (as-
consumed g/day) E-263
E-190. Total and caught fish consumption for the Minnesota families with new
mothers (as-consumed g/day) E-263
E-191. Total and caught fish consumption for the North Dakota American Indian
population (as-consumed g/day) E-264
E-192. Total and caught fish consumption for the North Dakota angler population
(as-consumed g/day) E-264
xx
-------
LIST OF FIGURES
2-1. Log-transformed bodyweight (pounds) versus age (years) 2-10
2-2. Q-Q plot of weight residuals (outliers in red) 2-11
2-3. Log-transformed bodyweight (pounds) versus age (years) (outliers in red) 2-12
2-4. Log-transformed fish consumption as raw g/kg-day versus age (outliers in
red) 2-13
2-5. Q-Q plot offish consumption residuals (outliers in red) 2-14
2-6. Histogram offish and shellfish consumption, consumers only, as-consumed
g/day 2-19
2-7. Histogram of log-transformed fish and shellfish consumption, consumers
only, as-consumed g/day 2-20
C-l. Comparison of the cumulative inverse cumulative distribution function
estimates from the macro andPROC Univariate C-4
E-l. Mean consumption, Connecticut general population, per capita, with 95%
confidence intervals E-42
E-2. Mean consumption per bodyweight, Connecticut general population, per
capita, with 95% confidence intervals E-43
E-3. Percent eating fish and shellfish during a year period, Connecticut general
population, per capita, with 95% confidence intervals E-44
E-4. Mean consumption, Connecticut general population, consumers only, with
95% confidence intervals E-45
E-5. Mean consumption per bodyweight, Connecticut general population,
consumers only, with 95% confidence intervals E-46
E-6. Mean consumption, Florida general population, per capita, with 95%
confidence intervals E-47
E-7. Mean consumption per bodyweight, Florida general population, per capita,
with 95% confidence intervals E-48
E-8. Percent eating fish and shellfish during a 7-day period, Florida general
population, per capita, with 95% confidence intervals E-49
E-9. Mean consumption, Florida general population, consumers only, with 95%
confidence intervals E-50
xxi
-------
LIST OF FIGURES (continued)
E-10. Mean consumption per bodyweight, Florida general population, consumers
only, with 95% confidence intervals E-51
E-l 1. Mean consumption, Minnesota general population, per capita, with 95%
confidence intervals E-52
E-12. Mean consumption per bodyweight, Minnesota general population, per capita,
with 95% confidence intervals E-53
E-13. Percent eating fish and shellfish during a roughly year-long period, Minnesota
general population, per capita, with 95% confidence intervals E-54
E-14. Mean consumption, Minnesota general population, consumers only, with 95%
confidence intervals E-55
E-l5. Mean consumption per bodyweight, Minnesota general population, consumers
only, with 95% confidence intervals E-56
E-l6. Mean consumption, North Dakota general population, per capita, with 95%
confidence intervals E-57
E-l7. Mean consumption per bodyweight, North Dakota general population, per
capita, with 95% confidence intervals E-58
E-18. Percent eating fish and shellfish during a roughly year-long period, North
Dakota general population, per capita, with 95% confidence intervals E-59
E-l9. Mean consumption, North Dakota general population, consumers only, with
95% confidence intervals E-60
E-20. Mean consumption per bodyweight, North Dakota general population,
consumers only, with 95% confidence intervals E-61
xxn
-------
PREFACE
The Exposure Factors Program of the of the U.S. Environmental Protection Agency's
(EPA) Office of Research and Development (ORD) has three main goals: (1) provide updates to
the Exposure Factors Handbook (U.S. EPA, 2011) and the Child-Specific Exposure Factors
Handbook (U.S. EPA, 2008); (2) identify exposure factors data gaps and needs in consultation
with clients; and (3) develop companion documents to assist clients in the use of exposure
factors data. The activities under each goal are supported by and respond to the needs of the
various EPA Program Offices.
The Exposure Factors Handbook: 2011 Edition (U.S. EPA, 2011) and the Child-Specific
Exposure Factors Handbook (U.S. EPA, 2008) provide a summary of statistical data on various
exposure factors used in assessing environmental exposures to both adults and children. Fish
consumption is one of the factors included in both handbooks. The National Center for
Environmental Assessment (NCEA) of EPA's Office of Research and Development (ORD) has
conducted an analysis of data from fish consumption surveys from the states of Connecticut,
Florida, Minnesota, and North Dakota. These states were selected from a previous effort aimed
at collecting available fish consumption rate data from state-wide surveys. These states were
selected for analysis based on sample size, study design, and data availability. The primary
objective of the analysis was to identify populations that consume more fish and shellfish than
the general population within each state and relative to the national general population and thus
may be at higher risk from exposure to contaminants in fish. EPA was particularly interested in
estimating fish and shellfish consumption for recreational anglers, low income populations,
children, and ethnic groups within each state. Results of this analysis were incorporated into
both the Exposure Factors Handbook: 2011 Edition (U.S. EPA, 2011) and the Child-Specific
Exposure Factors Handbook (M.S. EPA, 2008).
xxin
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
The National Center for Environmental Assessment (NCEA), Office of Research and
Development was responsible for the preparation of this report. This report was prepared by
Westat under EPA contract No. GS-23F-8144H. Jacqueline Moya served as Task Order
Manager, providing overall direction and technical assistance. Sherry Selevan (retired) and
Cheryl Itkin served as Task Order Manager on previous drafts of this report.
AUTHORS
Westat
Robert Clickner
John Rogers
EPA
Jacqueline Moya
Laurie Schuda
Cheryl Itkin
The following EPA individuals reviewed an earlier draft of this document and provided
valuable comments:
Denis Borum, Office of Water
Peter P. Egeghy, Office of Research and Development, National Exposure Research Lab
Ann Johnson, Office of Policy, Economics, and Innovation
Henry Kahn, Office of Research and Development, National Center for Environmental
Assessment
This document was reviewed by an external panel of experts. The panel was composed
of the following individuals:
Joanna Burger, Ph.D.
Rutgers University
Piscataway, NJ 08854-8082
XXIV
-------
Margy Gassel, Ph.D.
Consultant
Albany, CA 94706
Patricia Guenther, Ph.D., R.D.
U.S. Department of Agriculture
Center for Nutrition Policy and Promotion
Alexandria, VA 22134
Koenraad Marie'n, Ph.D.
Washington Department of Health
Office of Environmental Health Assessments
Olympia, WA 98504-7846
Nayak Polissar, Ph.D.
The Mountain-Whisper-Light Statistical Consulting
Seattle, WA 98112
Alan Stern, Dr.P.H., D.A.B.T. (Chair)
University of Medicine & Dentistry of NJ-School of Public Health
Piscataway, NJ 08854
XXV
-------
EXECUTIVE SUMMARY
Many state and local health agencies throughout the United States conduct area-specific
surveys that monitor and evaluate contaminant levels in local fish and study local populations for
fish consumption behavior. Summary information on these fish consumption surveys, which are
also available to the public, were compiled, summarized, and published by the National Center
for Environmental Assessment (NCEA), Office of Research and Development (ORD) of the
Environmental Protection Agency (EPA) in the Exposure Factors Handbook: 2011 Edition
(U.S. EPA, 2011).
Three surveys covering four states were selected for further analysis because they
contained data for specific targeted populations, had adequate sample size, and obtainable raw
data. In addition, they were selected because they identified recreational anglers and collected
information on the amount offish consumed that was purchased versus self-caught. These
surveys covered Connecticut, Florida, and a combined survey of Minnesota and North Dakota.
The fish consumption data were extracted from each of the database files provided by the states
and a single file was created in order to analyze the data consistently for each individual state.
The primary objective of the analysis was to identify populations that consume more fish
and shellfish within the state's general population and relative to the national general population
and thus may be at higher risk of exposure to contaminants in fish and shellfish. To achieve this
objective, the study focused on estimating fish and shellfish consumption for not only the general
population within the state, but also for recreational anglers, low income populations, children,
and other targeted populations.
The three surveys had similar procedures for selecting participants. The Florida survey
selected participants using a random sample within strata defined by counties. The Connecticut
and Minnesota/North Dakota surveys selected participants randomly from a state's general
population mailing list. The Connecticut and Minnesota/North Dakota surveys also collected
data from targeted populations of interest (e.g., anglers, Native American tribes). The
respondents that were randomly selected from the state's general population are referred to as the
"general" population. The means and percentiles for the general populations in each state were
statistically weighted in order to estimate the values for the states' general population. The
means and percentiles from the targeted populations were not weighted, since they statistically
represent only themselves.
Tables at the end of each section and in the Appendices show the fish and shellfish
consumption rates for subgroups classified by demographic characteristics and by the source of
the fish and shellfish consumed (i.e., freshwater versus marine, and store-bought versus self-
caught). The measurement units used to describe consumption rates are: (1) grams offish and
ES-1
-------
shellfish consumed per day (g/day), and (2) grams per kilogram of bodyweight per day (g/kg-
day). Both of these measures are presented for consumer-only offish and shellfish and for the
entire surveyed population (i.e., per capita). Consumer-only intake is defined as the quantity of
fish and shellfish consumed by individuals during the survey period. These estimates are
generated by averaging intake across only the individuals in the survey who consumed fish and
shellfish. Per capita intake rates are generated by averaging consumer-only intakes over the
entire population (including those individuals that reported no intake).
The three surveys had different questionnaires for collecting the fish consumption
information, procedures for classifying the fish consumed into species groups, and for processing
the data. The primary difference between surveys was the method for collection offish
consumption data. Florida used a telephone survey approach, while Connecticut, North Dakota
and Minnesota used mail surveys. In Connecticut, the respondents were asked how often each
type of seafood was eaten, without specifying a recall period. However, a long recall period was
implied, since one of the preceded response options was the number of times per year. In the
Minnesota/North Dakota survey, respondents were asked the rate offish or shellfish
consumption in the previous 12 months. In Florida, the respondents were asked for their fish
consumption during the "last 7 days" prior to the telephone interview.
The difference in data collection procedures may result in different consumption patterns.
Therefore, comparisons among the states should be interpreted with caution. Alternatively,
differences in reported fish and shellfish consumption across states may be due to geography or
other factors. For example, the fishing habits of the populations vary by state. In particular,
Florida residents have easy access to marine fishing, whereas, Minnesota and North Dakota
residents have good access to freshwater fishing, but not marine fishing.
This report focuses primarily on differences among groups of respondents within states,
where the groups are defined by categorical independent variables for income, age and gender,
race-ethnicity, education, and source offish consumed. General observations are summarized
below.
Overall, the percentage of respondents surveyed in the four states (Connecticut, Florida,
Minnesota, and North Dakota) that consumed fish and shellfish was higher for adults than
children. The consumption rate in g/day was higher for adults than children, but the consumption
rate in g/kg-day was generally lower for adults than children.
Anglers had higher consumption rates than the general population in Connecticut,
Minnesota, and North Dakota. Anglers were not sampled in the Florida survey. In Connecticut,
Asians, commercial fishermen, Expanded Food and Nutrition Education Program (EFNEP)
participants, and Women, Infants and Children Program (WIC) participants also had higher fish
consumption rates than the state's general population. In all four states, respondents that ate both
ES-2
-------
self-caught and store-bought fish and shellfish consumed more fish and shellfish than other
respondents.
In the Florida and Minnesota/North Dakota surveys, the percentage of the population that
reported eating fish increased with increasing household income. The Florida data showed a
statistically significant difference: for households with annual incomes below $20,000, 45% of
residents reported they ate fish or shellfish during the previous week, whereas households with
annual incomes above $50,000 the percentage increased to 57%.
Because age and bodyweight are closely related, fish and shellfish consumption per
kilogram bodyweight adjusts in part for differences associated with age. The results from these
four states suggest that a combination of females, non-whites, children, and those that consume
both self-caught and store-bought fish will have the highest fish and shellfish consumption rate
per kilogram bodyweight.
For race-ethnicity, the data suggest that Asians and American Indians may consume more
fish and shellfish than Whites, Blacks, or Hispanics on a per capita and per consumer basis.
Some significant differences between Whites and other race-ethnicity groups were found, with
Whites generally being more likely to consume fish and shellfish, but having a lower
consumption rate than other race-ethnicity groups. However, it is difficult to make
generalizations because the differences among race-ethnicity categories were not consistent
across states and were often not significant.
In the Florida survey, there was a significant increase in the percentage of the population
that reported eating fish with an increase in the education level of the head of the household. A
similar, though non-significant trend was seen in Minnesota and North Dakota. Although
comparisons across states should be done with caution, given the differences in survey
methodologies, some observations can be made based on per capita estimates. The per capita
fish and shellfish intake (averaged across the entire population of those who eat fish and those
that do not) is not directly affected by the length of the recall period and thus is more comparable
between states. The estimated average per capita intake in grams of "as-consumed" (generally
cooked) fish and shellfish per day is 27.10 g/day in Florida, 26.46 g/day in Connecticut, 18.06
g/day in Minnesota, and 18.99 g/day in North Dakota. Given that the Florida consumption is
likely to be underestimated because away-from-home consumption was not obtained for some
respondents, Florida residents may consume as much or more fish and shellfish per capita than
residents of the other three states. The estimates from the four states are higher than those for the
entire U.S. general population as calculated from the USDA's 1994-1998 Continuing Survey of
Food Intakes by Individuals (CSFII) (USDA, 1998). The average per capita consumption of
fish and shellfish for the U.S. population, based on the CSFII survey, is 12.83 g/day.
ES-3
-------
Although fish and shellfish consumption data from National Health and Nutrition
Examination Survey (NHANES) 2003-2006 are reported in the Exposure Factors Handbook:
2011 Edition (U.S. EPA, 2011), they are presented in g/kg-day and therefore not directly
comparable. The average per capita fish and shellfish consumption from NHANES 2003-2006
for all ages combined was estimated to be 0.22 g/kg-day (U.S. EPA, 2011). The average per
capita fish and shellfish intake from Florida, Connecticut, Minnesota, and North Dakota in g/kg-
day can be estimated by calculating a weighted average from all age groups and genders
combined. These are calculated to be 0.47 g/kg-day, 0.42 g/kg-day, 0.29 g/kg-day, and 0.33
g/kg-day for Florida, Connecticut, Minnesota, and North Dakota, respectively. These are higher
than the value obtained from NHANES 2003-2006 of 0.22 g/kg-day. United States regional
estimates based on NHANES 1999-2002 have also been reported (Mahaffey et al., 2005). The
average per capita 24-hour consumption estimates reported are: 20.8 g/day for the Atlantic Coast
region (which includes most of Connecticut and almost half of Florida); 13.5 g/day for the Gulf
Coast region (which includes the rest of Florida); and 11.5 g/day for the Midwest (which
includes Minnesota and North Dakota).
ES-4
-------
1. INTRODUCTION
1.1. OVERVIEW OF THE SURVEYS
Many state and local health agencies throughout the U.S. conduct area-specific surveys
that monitor and evaluate contaminant levels in local fish as well as survey local populations for
fish consumption behavior. These fish consumption surveys were compiled, summarized, and
published by the National Center for Environmental Assessment (NCEA), Office of Research
and Development (ORD) of the Environmental Protection Agency (EPA) in the Exposure
Factors Handbook: 2011 Edition (U.S. EPA, 2011).
Three studies covering four states for which raw data were available were selected for
further analysis. These were selected because they contained data for specific targeted
populations, had adequate sample size, and obtainable raw data. In addition, they were selected
because they identified recreational anglers and collected information on the amount offish
consumed that was purchased versus self-caught. Appendix A provides additional information
about each survey and describes the development of the combined database with fish
consumption data from all three studies. Reports on those three studies (referred to as the study
reports) are:
Steven A. Benson, Charlene R. Crocker, John Erjavec, Robert R. Jensen, Carolyn M.
Nyberg, Constance Y. Wixo, and Jill M. Zola. (2001) Fish Consumption Survey:
Minnesota and North Dakota, prepared by of the Energy & Research Center, University
of North Dakota, Grand Forks, ND.
N.C. Balcom, C.M. Capacchione, and D.W. Hirsch. (1999) Quantification of Fish and
Seafood Consumption Rates for Connecticut; submitted to the CT Dept. of Environmental
Protection, Contract No. CWF-332-R. CT Sea Grant Publication No. CT-SG-99- 02.
Robert L. Degner, Charles M. Adams, Susan D. Moss, and Stephanie K. Mack. (1994)
Per Capita Fish and Shellfish Consumption in Florida, University of Florida, Gainsville,
FL.
Although these reports provide state-wide estimates offish consumption rates,
information provided in the surveys allowed for a more thorough examination of the variability
in the survey population. Thus, EPA obtained the raw data from these studies and conducted
further analysis of the data. The primary objective of the EPA's analysis was to identify
populations that consume more fish and shellfish than the general population within each state
and relative to the national data and thus may be at higher risk to exposure from contaminants in
fish and shellfish. To achieve this objective, the study focused on estimating fish and shellfish
1-1
-------
intake for not only the general population within the state, but also for recreational anglers, low
income populations, children, and other targeted populations.
Table 1-1 presents summary information for the three studies that are discussed in this
report. The three surveys had different questionnaires and procedures for collecting the fish
consumption information, classifying the fish consumed into species groups, and processing the
data. The primary difference between surveys was the method for collection offish consumption
data. All three studies used household surveys in which data was collected on fish and shellfish
consumption for all or most members of the household.
Table 1-1. Summary offish consumption surveys
Study Location
Florida
Connecticut
Minnesota/North Dakota
Population/sample/mode
General population
Saltwater anglers
Freshwater anglers
Commercial fishermen
Southeast Asians
Food stamp
recipients/WIC
participants
New mothers
Tribal members
Fish
Recall
ROD
Phone interviews
Convenience
In-person interviews
All fish for respondent.
Fish eaten at home for
other HH members
7 days
Randomized HH list
Mailed questionnaire
Convenience
Mailed questionnaire
Randomized license list
Mailed questionnaire
Convenience
Mailed questionnaire
Convenience
In-person interviews
All fish for all family
members
Unspecified, asked rate of
consumption
Randomized HH list
Mailed questionnaire
Randomized license list
Mailed questionnaire
Randomized list (MN only)
Mailed questionnaire
Convenience
Direct contact canvassing
All fish for up to five family
members
Past 12 months, asked rate of
consumption
1-2
-------
Table 1-1 Summary offish consumption surveys (continued)
Study Location
Information on seasonal
patterns in fish
consumption
How fish consumption
was asked
How fish consumption
was reported
Self -caught vs. store-
bought categories of each
fish type
Information on location
of caught fish
Data collection period
Number of households
Number of individuals
Florida
Calls uniformly
distributed over 1 year,
call date in the file
Type offish/seafood,
amount eaten, how cooked
in the last 7 days
Asked specifically for 24
fish types and 13 shellfish
types and room for
'others'
Uncooked edible weight
by person by fish
Percent caught by fish
Asked but not provided in
the data file, respondent
county known
March 1993 to March
1994
8,740
17,213
Connecticut
As reported by respondent
1 - FFQ (how often/much
fish/seafood types eaten)
list 19 fish dishes + space
for Other
2-10-day diary including
type, amount, preparation
(for a subset of
respondents)
Apparently as-consumed
weight of edible fish by
person by fish (g/day),
response units not
carefully defined
Source checkboxes: Self-
caught, store-bought,
restaurant
Text description, name of
waterbody, town
July 1996 to May 1997
810
2,080
Minnesota/North Dakota
As reported by respondent
Number of meals, including
usual portion size for five
categories of purchased
fish/swordfish/shark, breaded
fish products, canned tuna,
shellfish, other fish) and seven
categories of caught fish
(panfish, walleye/sauger,
pike/muskie, bass, salmon/lake
trout, stream trout, other self-
caught fish)
Apparently, as-eaten
consumption by person by
category offish, as reported by
the respondent
Non-purchased or (store-
bought or restaurant)
One of several geographic
areas
October to November 2000
1,568
4,262
HH stands for household; FFQ stands for Food Frequency Questionnaire; RDD stand for Random Digit Dialing.
There are several approaches for collecting fish consumption data. The approach used
depends on the purpose of the data collection. More detailed information regarding approaches
used for collecting fish consumption data and their advantages and limitations can be found in
1-3
-------
the report entitled Consumption Surveys for Fish and Shellfish: A Review and Analysis of Survey
Methods (U.S. EPA, 1992). Five approaches reviewed by the EPA included: (1) recalled
information collected by telephone; (2) recalled information collected by in-person interviews;
(3) recalled information using mailed questionnaires; (4) diaries maintained by anglers; and (5)
on-site creel surveys (U.S. EPA, 1992). Each approach has its advantages and limitations and
usually a combination of various approaches is used to improve the validity of the data collected.
Survey instruments used by the three studies are described below.
A mail survey instrument was used to collect fish consumption data for the
Minnesota/North Dakota study (Dillman, 1978; Dillman, 2000; U.S. EPA, 1998). The
questionnaire design in this study was aimed at obtaining the most recent years' intake offish
from all sources, personal information from each respondent, and enough sociodemographic
information to categorize responses. In the Minnesota/North Dakota survey, respondents were
asked the rate offish or shellfish consumption in the past 12 months. An advisory board was
selected to aid in the development of the survey design.
In the Connecticut study, a mailed survey questionnaire was also used to collect the fish
consumption data, with a telephone follow-up (Peters and Houseknecht, 1992; West et al., 1993).
A food frequency questionnaire was used in the Connecticut survey as a more accurate approach
to collecting food consumption data than a 24-hour recall survey. The Connecticut respondents
were asked how often each type of seafood was eaten, without specifying a recall period.
However, a long recall period was implied, since one of the preceded response options was the
number of times per year.
For the Florida study, a telephone survey was chosen for its relatively low cost and the
ease of probability sampling (Peters and Houseknecht, 1992; Dubois and Boivin, 1990; Block,
1982). To enhance the accuracy of the respondents recall, an approach known as aided recall
was used. In this approach, the questionnaires used six commonly eaten types of finfish and five
major types of shellfish as memory cues to screen for finfish and shellfish consumers. To
improve portion size estimates, a range of portion sizes were offered to respondents (Block et al.,
1986; Block, 1982). The respondents were asked for their fish consumption during the last 7
days prior to the telephone interview (Anderson, 1988). Thus, these data represent short-term
consumption patterns and comparisons among the states should be interpreted with caution.
It should be noted that fishing habits of the populations in the states covered by the
surveys are quite different. In particular, Florida residents have easy access to marine fishing
and Minnesota and North Dakota residents have good access to freshwater fishing. Since the
consumption differences between states may be due in part to different procedures used in the
different surveys and the focus of the analysis is to identify subgroups with high fish and
shellfish consumption, the statistical analysis focuses on differences between subgroups within
1-4
-------
states rather than comparisons between states. At the same time, the data from the surveys were
combined using common categories for demographic variables and a common definition of
consumption in order to allow a general comparison of per capita consumption across states.
1.2. DIFFERENCES AMONG THE SURVEYS
The three surveys differed primarily in their procedures for the collection of the fish
consumption data. In Connecticut, the respondents were asked how often each type of seafood
was eaten, without specifying a recall period. However, a long recall period was implied by the
option of responding using the number of times per year. In the Minnesota/North Dakota survey,
respondents were asked the rate offish or shellfish consumption in the past 12 months. In
Florida, the respondents were asked for their fish consumption during the "last 7 days" prior to
the telephone interview. All of these values were converted to a consumption rate and, using the
raw data and the respondents reported weight, a consumption rate per kilogram bodyweight.In
the Connecticut and the Minnesota/North Dakota surveys, the questionnaires collected data on
all fish and shellfish consumption for all or most respondents in the household. However, in
Florida, the questionnaire did not collect away-from-home fish and shellfish consumption data
for children. Away-from-home meals are defined in the Florida survey as meals not prepared in
the household (e.g., fast food places, restaurants, ready-to-eat meals, TV-dinners). As a result,
the consumption estimates presented in this report for children in Florida tend to underestimate
their fish and shellfish consumption. Also, the lower proportion of children eating fish and
shellfish in Florida is due, in part, to not collecting away-from-home consumption.
In all three surveys, some populations were less likely to respond to the questionnaire
than others. In addition, the Minnesota/North Dakota survey collected data on up to five
household members and the Florida survey did not collect data on some household members if
the primary meal preparer was not available. All three surveys collected data from a random
sample of the state's populations stratified by county, referred to in this report as the state's
general population. The fish and shellfish consumption estimates for the general population were
weighted in order to calculate estimates for the state's population, after adjusting for different
sampling rates in different counties and different response rates.
The Minnesota/North Dakota and Connecticut surveys also collected questionnaire data
from members of targeted populations of interest. The consumption estimates for these
populations were not weighted.
1-5
-------
1.3. OVERVIEW OF THE REPORT
This report has three sections, this Introduction (Section 1) and:
• Section 2 describes data processing and statistical analysis methodologies, including:
o The conversion between raw and as-consumed weight offish and shellfish;
o The calculation of analytical weights;
o The identification and handling of outliers; and
o The calculation of summary statistics (including percentiles).
• Section 3 provides a discussion of the results from the surveys.
• Appendices A though D describe the database development, QA/QC process, and
calculation of percentiles, and provide a glossary of terms.
• Appendix E provides extensive tables and graphics, including:
o Tables offish and shellfish consumption, as-consumed, for the general population;
and
o Plots of consumption estimates with 95% confidence intervals and significance tests
for the general population and its populations defined by state, state, age, gender,
education, income, race-ethnicity, and fish type consumed. These plots provide basic
summary statistics for the three surveys.
Because there are a large number of data tables, Appendix E begins with a listed
directory to the tables within the Appendix organized to help the reader find the tables by
consumption characteristics - consumers only, per capita, amount consumed, percent
consumers—and state, age, gender, education, income, race-ethnicity, and type offish
consumed.
1-6
-------
2. METHODS AND STATISTICAL ANALYSES
The quality of the fish and shellfish consumption estimates depends on the quality of the
original data, the processing of the data to obtain consumption estimates for individuals, and the
processing of the data to calculate the estimates within states and demographic subgroups.
Appendix A has information on the surveys and provides details on the processing of the data.
An important part of the processing is the conversion between raw weight and as-consumed (or
cooked) weight offish and shellfish. Section 2.1 describes how this conversion was done for
each survey. Section 2.2 describes the variables used in the analysis. Section 2.3 describes the
identification and handling of outliers. Section 2.4 tabulates the sample sizes for the analysis
tables. Section 2.5 discusses the summary statistics calculated and presented in the tables.
Finally, Section 2.6 describes the analysis weights used for the estimation offish and shellfish
consumption for the general populations in each state. All calculations were performed using
SAS version 9.1.3.
2.1. CONVERSIONS BETWEEN RAW AND AS-CONSUMED WEIGHT
The weight offish consumed can be expressed as the weight offish as caught, as cleaned
and prepared for cooking, and as cooked. As a general rule depending on cooking practices, the
cooked weight is less than the uncooked weight due to moisture loss. The data from the three
surveys were processed to estimate the raw weight and the as-consumed weight (raw or cooked
weight, depending on how the fish or shellfish was consumed). Where possible, the procedures
used by the original investigators were used for this conversion. In general, there was inadequate
data to distinguish between weight loses due to trimming the raw fish or removing the shell from
the shellfish (refuse factors) and due to the cooking method. The following sections describe the
conversions to calculate the raw and as-consumed weights offish and shellfish for each of the
surveys. Additional details are provided in Appendix A.
2.1.1. Florida Survey
The Florida survey questionnaire requested both the quantity eaten and the cooking
method used. The as-consumed weight offish was calculated from the respondent's estimate of
the volume offish consumed compared to a slice of bread and the typical density offish. The
weight of shellfish was requested in units likely to be familiar to the respondent, such as pounds
for lobster, number and size of shrimp, and cups of canned fish. The respondent's replies were
converted to the as-consumed weight in the data files received from Florida. The reported as-
consumed quantity was converted to uncooked weight using conversion factors documented in
the study report and shown in Table 2-1. The conversion factors depend on the type of food
2-1
-------
(several types of shellfish, other shellfish, or any finfish) and the cooking method. An assumed
value was used when the cooking method was not provided.
Table 2-1. Factors for converting the Florida data from uncooked to as-
consumed weights
Cooking Method
Fried
Broiled
Steamed
Broiled or poached
Raw
Microwave baked
Conventional oven
Grilled or smoked
Other3
Salad
Shrimp
0.75
0.78
0.79
0.80
1.00
0.80
0.85
0.78
0.77
Clams
0.57
0.60
0.62
0.63
1.00
0.66
0.64
0.60
0.61
Blue
crab
0.75
0.78
0.79
0.80
1.00
0.80
0.85
0.78
0.77
Crab
meat
0.57
0.60
0.62
0.63
1.00
0.66
0.64
0.60
0.61
Imita-
tion
crab
meat
0.75
0.78
0.78
0.78
1.00
0.87
0.79
0.78
0.82
Conch
0.57
0.60
0.62
0.63
1.00
0.66
0.64
0.60
0.61
Cray-
fish
0.75
0.78
0.79
0.80
1.00
0.80
0.85
0.78
0.77
Misc.
shell-
fish
0.67
0.70
0.72
0.72
1.00
0.75
0.75
0.70
0.71
Finfish
0.75
0.78
0.78
0.78
1.00
0.87
0.79
0.78
0.82
"Other includes: other cooking methods, don't know, and missing data values.
2.1.2. Connecticut Survey
The Connecticut data files provide the quantity of consumed fish as "cooked weight of
edible fish." However, for the few items that were clearly eaten uncooked (e.g., oysters, sushi),
the quantity is assumed to be the uncooked weight. This weight is referred to as the as-
consumed weight. The Connecticut respondents assessed the quantity of a food item by
selecting a shape of the food item from a set of pictures and a corresponding thickness.
Connecticut collected cooking method information for only a subset of the respondents and this
information was apparently not used in the conversion to as-consumed consumption. When
converting the reported shape and thickness to the as-consumed weight some assumptions were
made such as assuming that certain dishes were eaten uncooked or cooked or that the reported
shape corresponded to an uncooked or cooked food item. The conversion factors used were
provided in an appendix to the study report. If the respondent reported the as-consumed
quantity, then no conversion was needed.
2-2
-------
For this analysis, the as-consumed weights in the data file were converted to uncooked
weights. The conversion used the conversion factors in the report, if provided. In a few cases
judgment was used to decide what conversion factor to use. Otherwise, a conversion factor from
the EPA Mercury Study Report to Congress (U.S. EPA, 1997) was used. The conversion factors
used to calculate the uncooked weight from the as-consumed weight for each fish or shellfish are
shown in Table A-10 in Appendix A. If the conversion factor did not come directly from the
Connecticut study report, a comment column explains the source of the conversion factor. In
some cases the conversion factors were based on the report Agriculture Handbook (USDA,
1987).
2.1.3. Minnesota and North Dakota Survey
The Minnesota/North Dakota questionnaire asked for the "usual" portion size in ounces.
Pictures of the food items with associated weights were provided as a guide. The pictures
showed the food on a plate, suggesting that the weight provided with each picture was the as-
consumed quantity. The respondent was asked the weight of the fish, and if that weight was the
cooked or raw weight. The questionnaire did not inquire if the item was cooked before being
eaten. The Minnesota and North Dakota data file has programming code to calculate fish
consumption. However, this code ignores the uncooked versus cooked responses. A distinction
between cooked and uncooked weights was made for this analysis. For the purposes of
converting from uncooked to as-consumed or visa versa, it was assumed that all reported
consumption was eaten cooked because the questionnaire was not specific enough to identify
food items that might have been consumed uncooked. All food items were converted from as-
consumed to uncooked weight by dividing the as-consumed weight by 0.75, the value used in the
EPA Mercury Study Report to Congress (U.S EPA, 1997).
2.2. VARIABLES USED FOR THE DATA ANALYSIS
The combined database has fish consumption data for Florida, Connecticut, Minnesota,
and North Dakota. The variables were divided into three files, one with household information,
a second with information about individuals within the sampled households, and a third with fish
consumption information for each type or species offish or shellfish eaten by each individuals.
It also has information on individual fish and shellfish consumption of both self-caught and
purchased (or store-bought) fish and shellfish. The total consumption for a specific fish or
shellfish is the sum of the self-caught and store-bought quantities. The fish consumption rates
were expressed in two units, grams of uncooked fish (i.e., weight before any cooking[g]) and
grams of as-consumed fish (i.e., weight offish as prepared for eating) per day (g/day). The data
also include the type offish (e.g., finfish or shellfish) and fish habitat (freshwater, estuarine, or
2-3
-------
marine). The files with household and individual characteristics include demographic
information.
2.2.1. Dependent Variables
For the purpose of analysis, the variables in the data file can be divided into dependent
variables that are the primary focus of the analysis, and independent variables that may be
associated with differences in the dependent variables. The primary dependent variables are the:
(1) consumption rate in grams offish or shellfish per day; (2) consumption rate per kilogram of
bodyweight (g/kg-day), and (3) percentage of respondents that reported consuming fish and
shellfish.
The data from the combined database were processed to calculate the total fish and
shellfish consumption for each individual by summing the consumption across all (or a selected
subset of) fish and shellfish species. For those individuals that did not report any fish or shellfish
consumption, the total is zero. For some tables the consumption was calculated separately for
(1) fish and shellfish, (2) fish habitat (freshwater, estuarine, and marine), and (3) fish source,
store-bought or self-caught.
2.2.2. Independent Variables
Summary statistics for the dependent variables were calculated for selected populations
defined by state, income categories, race-ethnicity categories, age and gender categories, and
education categories. The following paragraphs describe the independent variables that were
used.
Household income was derived from the original data files. Different surveys used
different categories for classifying income. However, the categories used in the surveys could be
combined to create the following common income categories across the three surveys: $0 to
$20,000, $20,000 to $50,000, and greater than $50,000. Fifteen percent of households did not
provide income information. Households without income information were placed into an
"Unknown" category.
Race-ethnicity was also defined differently in the three surveys. Categories were
combined to create the following common set of generic race-ethnicity categories:
• American Indian
• Asian (includes a general Asian category, Asian Indian, Cambodian, Chinese, Filipino,
Hmong, Laotian, and Vietnamese)
2-4
-------
Black non-Hispanic
Hispanic (include a general Hispanic category and Central American, Dominican,
Mexican, Puerto Rican, and South American)
White non-Hispanic, and
Unknown
The population was divided into age and gender groups using several different variables
to create several sets of age and gender categories, including: adults (18 and older) and children
(aged < 18), ages 1 to <6, 6 to <11 and 11 to <16 for children, ages 16 to <30, 30 to <50, and 15
to <45 to cover women of child-bearing age, females 45 and older, 50 and over, and men aged 16
to <30, 30 to <50, 15 to <45, 45 and older, and 50 and over. In the tables, these age categories
are written as 1-5, 6-10, 11-15, 16-29, 30-49, 50+, 15-44, and 45+, respectively.
Education of the head of household was defined differently in the different surveys.
The categories in the different surveys were receded to create the following common categories
for the highest level of education:
• Some High School: Completion of some high school or 11 or fewer years of education.
• High School: Completion of high school, 12 years of education, or a GED.
• Some College: Completion of "Some College" or 13 to 15 years of education.
• College Graduate: Completion of 16 or more years of education, a 2-4 year degree,
graduate or post-graduate education.
Some judgment was used to define the common education categories. In addition,
different respondents were asked about their education in the different surveys. In the
Minnesota/North Dakota study education was requested for the first two listed household
members, the respondent and the second listed member. The combined files have the education
of the respondent. The Connecticut study asked for the education of the head of household. The
Florida study asked for the education of the randomly selected adult. Since the randomly
selected adult may not be the head of household and the respondent (the person that filled out the
questionnaire) in the Minnesota/North Dakota study was not randomly selected and might not be
an adult, the measures of education level associated with the household are not completely
equivalent between studies. The education is thus a general way to classify households rather
2-5
-------
than a classification related to a specific definition of household education or an indication of the
education of individuals within the household.
Based on their reported fish consumption, respondents were classified as eating only self-
caught fish, both self-caught fish and self-caught and store-bought fish and shellfish, or only
store-bought fish and shellfish. A second classification grouped freshwater and estuarine self-
caught fish to create the classifications: (1) exclusively, (2) occasionally, or (3) never eats
freshwater and estuarine self-caught fish and shellfish.
Finally, the Connecticut and Minnesota/North Dakota surveys collected fish consumption
information from populations of interest, referred to here a targeted populations. The
questionnaires for the targeted populations were collected in addition to the questionnaires for
the general population. The general population and the populations are described in Table 2-2.
The last two columns of Table 2-2 indicate if the respondents were selected randomly from an
identifiable targeted population and, if so, the response rate reported in the survey reports.
2-6
-------
Table 2-2. Targeted populations
State
Populations
Sampled from:
Statistical Response
Sample Rate
Florida General population
General population
Angler/Recreational fisherman
Connecticut Aquaculture students
Asians
Commercial fishermen
EFNEP3 participants
WICb participants
Minnesota General population
American Indians
Angler/Recreational fishermen
Families with new mothers
North Dakota General population
American Indians
Angler/Recreational fishermen
Random digit dialing procedure Yes
Stratified random sample of
household addresses
Yes
Lists of saltwater anglers
Interviewers compiled lists of
anglers at sites with shore- and
vessel-based fishing opportunities
along Long Island Sound
No
Students at a vocational aquaculture
school No
Southeast Asian households
identified by one field interviewer No
List of licensed commercial
fishermen Yes
Identified through Connecticut
EFNEP offices No
Identified through Connecticut WIC
offices No
Stratified random sample of
household addresses Yes
Identified through members of the
Bois Forte Tribe No
List of licensed fishermen Yes
List of women who gave birth in
1999 Yes
Stratified random sample of
household addresses Yes
Identified through members of the
Spirit Lake Nation and three
Affiliated Tribes No
Two different lists of licensed
fishermen Yes
Not reported
6%
10%
21%
21%
15%
21%
21%
"Expanded Food and Nutrition Education Program.
bWomen Infants and Children Program.
2.3. IDENTIFICATION AND HANDLING OF OUTLIERS
Individuals with particularly unusual observations were removed from the data before the
analysis tables were prepared. Outliers are observations that are particularly unusual compared
to expected values or compared to other observed values. Outliers may be correct observations
2-7
-------
for very unusual individuals or incorrect observations. Removing incorrect observations will
improve the quality of the results by removing a source of bias. However, removing correct
observations for unusual individuals can result in increased bias and reduced quality of the
results. Unfortunately, without additional data collection, it is not possible to determine whether
an unusual observation is an error or is a correct value for an unusual individual. There may also
be observations that are incorrect but are not unusual. Such observations are difficult to identify
without additional data collection. The procedures for identifying outliers focused on identifying
highly unusual observations that were unlikely to be correct. Although the primary observations
are estimates offish consumption, important outcomes of the research is an analysis offish
consumption by age and an analysis offish consumption per kilogram bodyweight. As a result,
unusual combinations of age and bodyweight may also affect the fish consumption estimates.
Decisions about how to detect and handle outliers are somewhat subjective. The objective was
to remove the values that were most unusual on the assumption that these values were most
likely to be incorrect and removing them would decrease the bias in the fish consumption
estimates.
A common procedure is to remove cases from analyses if they have suspect or missing
data for the variables in that analysis. Thus, for example, a bivariate analysis of bodyweight by
gender would exclude cases with suspect or missing data for either of these variables. However,
a multiple regression analysis of, say, fish consumption on weight, gender, age, geographic
region, and ethnicity would necessarily exclude all cases with suspect or missing data for any of
these variables. The regression analysis would therefore likely be based on fewer cases than the
bivariate tabulation. An alternative approach would be to remove all cases with any missing or
suspect data from any of the variables in any of the analyses. All analyses would be based on the
same number of observations. This latter approach is what was used. As detailed below, a total
of 13 observations were removed from all analyses due to outlying observations.
The highest eight age values were 98, 110, 120, 140, 190, 190, 341, and 731 years. There
were seven values of 98 years. The National Health and Nutrition Examination Survey
(NHANES) (CDC, 2002) is an ongoing national survey that collects health data on Americans.
The NHANES data provides a comparison dataset against which to judge whether the age ranges
and weight ranges in the various fish consumption surveys are reasonable. In the 2001-2002
NHANES data the highest ages are reported as >85 years and 2.14% of respondents were in this
bracket. In the data from the three studies, less than 2% of respondents have ages over 84 years.
In the NHANES data, the oldest age when the respondent's heaviest weight was attained was
103 years. NHANES respondents were asked their age when various medical conditions were
diagnosed. The oldest age of a reported medical condition was 98 years. Based on this
information from the NHANES survey, seven values of 110 years or greater are assumed to be
2-8
-------
unusually high, likely to be incorrect, and therefore outliers. One was from Florida, three were
from Minnesota, and three were from North Dakota. Five of these records were also missing
bodyweight data.
The six largest bodyweights reported were 400, 416, 420, 500, 530, and 686 pounds. The
value of 500 pounds appeared in the data four times. In the 2001-2002 NHANES data (CDC,
2002), the highest reported bodyweight for a respondent was 434 pounds; the self-reported
greatest bodyweight attained by a respondent was 500 pounds. Based on the distribution of the
bodyweights in the study data and the NHANES data, six values over 425 pounds were judged to
be outliers. One was from Connecticut, one was from North Dakota, and four were from Florida.
If children have an exponential bodyweight growth rate up to approximately age 18 and
fairly constant bodyweight after age 18, a plot of log-transformed bodyweight versus age would
show a linear increase up to age 18 and a constant level for ages above 18. The distribution of
the log-transformed bodyweights around the mean for a selected age would be expected to be
roughly normally distributed. Figure 2-1 shows a plot of log-transformed bodyweight versus
age. The plot shows the basic pattern described above. However, there are some unusual
combinations of bodyweight and age and the linear relationship expected at the lowest ages may
not describe the data. The unusual combinations of bodyweight and age were identified by: (1)
fitting a non-parametric smooth relationship between log-transformed weight and age; and (2)
classifying bodyweights that differed significantly from the predicted mean for the observed age
as outliers.
The smooth relationship was fit using SAS PROC LOESS with two iterations, one for an
initial fit and a second iteration to reduce the effect of possible outliers. Separate smooth curves
were fit for each study to accommodate differences between populations and differences in how
the data were collected and reported.
2-9
-------
Figure 2-1
100-
10-
1 - o
I
20
i
40
i
60
i
80
100
Age
Figure 2-1. Log-transformed bodyweight (pounds) versus age (years).
Classifying the bodyweights as unusual assumed that the residuals around the predicted
log-transformed bodyweights could be reasonably described by a smooth, roughly normal
distribution. Figure 2-2 shows a Q-Q plot of the sorted residuals versus the corresponding values
that would be obtained if the data had a standard normal distribution. If the residuals have a
normal distribution, the data will fall on a straight line in the Q-Q plot. In Figure 2-2, most of
the data fall on a straight line. However, there are several high and low residuals that are
inconsistent with the assumption that the log-transformed bodyweights at a fixed age have a
normal distribution. There are 21,293 residuals. If the residuals do have a normal distribution,
the probability that all residuals are within 4.72 standard deviations of the mean is only 5%. One
criterion for identifying outliers from a normal distribution is to classify all values outside 4.72
standard deviations from the mean as an outlier (Barnett and Lewis, 1996). In this case that
corresponds to classifying residuals outside -0.50 to 0.50 as outliers. Since the distribution of the
data might differ somewhat from a normal distribution, as indicated by the slight curvature in the
2-10
-------
middle of portion of Figure 2-2, wider limits were used. The most extreme residuals, judged to
be those outside the range from -0.62 to 0.62, were assumed to be outliers and were removed
from the analysis. The cut-off of 0.62 was chosen because it corresponded to a gap in the Q-Q
plot curve that separated the lower group of outliers from the majority of the residuals. The red
values in Figure 2-2 were classified as outliers. Figure 2-3 shows the same data in Figure 2-1
with observations classified as outliers shown in red.
Figure 2-2
1 -
o-
to
111
cr
-1 -
i
-4
Corresponding normal quantile
Classification o OK + Outlier
Figure 2-2. Q-Q plot of weight residuals (outliers in red).
2-11
-------
100-
10-
1 - +
I
20
i
40
i
60
i
80
100
Age
Classification o OK + Outlier
Figure 2-3. Log-transformed bodyweight (pounds) versus age (years)
(outliers in red).
Similar procedures were used to identify outliers in raw fish consumption (g/kg-day).
SAS PROC LOESS was used to fit a smooth curve to the fish consumption data. The normality
assumption for the residuals appeared to be reasonable. There were only one or two observations
that appeared to be inconsistent with the overall curve. Based on 13,270 residuals, the
probability that all residuals are within 4.63 standard deviations of the mean is only 5%. In this
case, a range of 4.63 standard deviations corresponds to residuals between -1.97 and 1.97.
However, to accommodate the fact that the distribution might not be quite normal, a cutoff of
2.20 was used resulting in classifying the lowest two residuals as outliers. Figure 2-4 shows the
log-transformed fish consumption per day per kilogram versus age with the two outliers shown
in red. Figure 2-5 shows the Q-Q plot for the residuals when fitting the log-transformed fish
consumption in g/kg-day.
2-12
-------
Figure 2-4
re
Q
<3
re
ce
100-
10-
1 -
0.1 -
0.01 -
0.001 -
o _ o
£°o o
00
I
20
i
40
i
60
\
80
Age
Classification o OK + Outlier
100
Figure 2-4. Log-transformed fish consumption as raw g/kg-day versus age
(outliers in red).
2-13
-------
Figure 2-5
to
ill
cc
Q
<3
(Tj
Si
o'
bo
o
0-
-2-
-2
Corresponding normal quantile
Classification o OK + Outlier
Figure 2-5. Q-Q plot of fish consumption residuals (outliers in red).
Only the relationships between age and bodyweight and age and raw consumption (g/kg-
day) were used to identify outliers. A review of the other consumption variables found no other
outliers. A total of 44 individuals out of 23,566 (0.2%) were classified having unusual
bodyweight or fish consumption and removed from the analysis based on the procedures above.
Two individuals were removed due to unusual fish consumption values; 29 were removed due to
unusual combinations of age and weight; and 13 were removed due to unreasonably high ages or
bodyweights.
For most estimates, removing these outliers is not expected to have an important effect on
the results. Removing outliers will generally have the greatest effect on estimates of extreme
percentiles. Using data with and without the outliers, the minimum, maximum, mean and
percentiles were calculated for the raw fish consumption in both g/day and g/kg-day for fish and
shellfish consumers in the general population from each state and per capita. Most percentile
estimates and means changed by much less than 3% as a result of removing the outlying values.
2-14
-------
Differences that were greater than 3% are shown in Table 2-3. Even if the removed data values
are correct, the estimates of extreme percentiles can be sensitive to the observed values and are
not very precise.
Table 2-3. State level estimates that changed by more than 3% as a result of
removing outliers
Variable
Fish consumption
(g/day)
Fish consumption per
kilogram (g/kg-day)
State Statistic
CT 10th percentile
Minimum
MN 5th percentile
CT 10th percentile
Minimum
FL 99th percentile
Maximum
Minimum
ND 10th percentile
5th percentile
Population
Per capita
Consumers
Consumers
Per capita
Consumers
Consumers
Per capita and
consumers
Consumers
Per capita
Consumers
Estimate:
All Data
0.0106
0.0388
0.0590
0.0002
0.0007
7.618
127.31
0.0030
0.0452
0.0414
Estimate:
With Outliers
Removed
0.1131
0.4242
0.4899
0.0023
0.0070
7.127
38.29
0.0047
0.0466
0.0438
Ratio
10.67
10.93
1.07
11.50
10.50
0.94
0.30
1.58
1.03
1.06
2.4. SAMPLE SIZES
The summary tables in Section 3 and Appendix E tabulate fish consumption by state
broken down by demographic characteristics or by the type offish and shellfish consumed.
Table 2-4 shows the number of survey respondents by state for the general population and
targeted populations. The first two columns list the state and sampled population. The third
column has the number of respondents in the combined data file. The fourth through seventh
columns have the number of respondents for which summary statistics are presented after
removing outliers. Sample sizes are shown separately for estimates offish and shellfish
consumption in g/day and g/kg-day on a per capita basis and for consumers only. The difference
between the sample sizes for consumption (g/day) and (g/kg-day) are due to missing
bodyweights for some individuals. The sample sizes are for statistics that are broken out by state
and demographic characteristics. Since an individual can consume multiple types offish, the
sample size column in the tabulations by type offish consumed will generally not add to the
numbers shown in Table 2-4.
2-15
-------
Table 2-4. Sample sizes for the general population and targeted populations
Number of records after removing outliers
State
CT
FL
MN
ND
Sampled population
Angler/recreational
fishermen
Aquaculture students
Asians
Commercial fishermen
EFNEP participants
General population
WIC participants
General population
American Indians
Angler/recreational
fishermen
General population
Families with new mothers
American Indians
Angler/recreational
fishermen
General population
Number of
respondents
in data file
267
25
402
178
71
433
704
17,213
221
1,172
843
415
134
872
605
Consumption
(g/day)
Consumers
Per capita only
266
25
402
178
71
431
703
17,181
221
1,171
841
415
133
871
602
257
19
396
171
60
369
557
8,566
196
1,127
796
352
78
825
570
Consumption
Per capita
250
25
396
173
67
420
699
15,367
216
1,152
837
401
106
854
575
(g/day-kg)
Consumers
only
244
19
393
166
58
362
553
7,757
192
1,109
793
341
64
808
546
The summary statistics for the general population were weighted. Table 2-5 shows the
weighted totals for the general population in both g/day and g/kg-day on a per capita basis and
for fish consumers.
Table 2-5. Weighted totals for the general population, in thousands
State
CT
FL
MN
ND
Consumption (g/day)
Per capita Consumers only
3,378 2,854
15,952 7,912
4,900 4,623
639 606
Consumption (g/kg-day)
Per capita Consumers only
3,296 2,804
14,827 7,490
4,897 4,621
610 580
2-16
-------
In tables displaying demographic characteristics, if the demographic characteristic was
missing for some respondents, an "Unknown" category was created so the totals would be
consistent across tables.
2.5. SUMMARY STATISTICS AND STATISTICAL ANALYSIS METHODS
The tables in Section 3 show summary statistics for the dependent variables. Separate
tables show statistics for per capita consumption and for consumer only. Separate statistics are
shown for each category of the independent variables. The following summary statistics are
presented in various tables:
• Sample size: The number of respondents used to calculate the summary statistics.
• Weighted population estimate, in thousands: The size of the general population
represented by the respondents. The population size is not calculated for the targeted
populations.
• Arithmetic mean (or average): For the population, this is the arithmetic mean fish
consumption across all individuals (including zero consumption for those that ate no fish
or shellfish). The arithmetic mean is useful for summarizing average consumption and
estimating long-term exposure.
• Geometric mean: The geometric mean cannot be calculated when there are zeroes in the
data. Therefore, the geometric mean is calculated only for respondents that reported
consuming fish or shellfish. The geometric mean is an approximate estimate of the
median consumption and, for fish and shellfish consumers, provides a better test of
differences among population groups.
• Percent of respondents: Reported consuming fish or shellfish during a 7-day (in Florida)
or year-long period.
• Minimum: The minimum reported consumption offish and shellfish.
• Percentiles: The 5th, 10th, 25th, 50th, 75th, 90th, 95th, and 99th percentiles of the distribution
offish consumption quantities. Lower percentiles for the population (i.e., per capita) may
be zero because some respondents did not report eating any fish or shellfish. For
example, if 40% of the respondents consumed fish or shellfish, then the 5* , 10* , and 25*
percentiles will be zero. All percentiles for consumers are greater than zero. As explained
in Appendix C, extreme percentiles that cannot be estimated are shown as a dot,
indicating a missing value.
• Maximum: The maximum reported consumption offish and shellfish.
• 95% Confidence interval for the mean.
2-17
-------
Section 2.5.1 discusses the skewness and distribution of the consumption data and
usefulness of the geometric mean. Section 2.5.2 discusses the calculation of percentiles.
Summary tables in Appendix E provide means for subgroups, 95% confidence intervals for the
means, and significance tests for differences of means among the subgroups. Section 2.5.3
discusses the calculation of the confidence intervals and significance tests.
2.5.1. Distribution of the Consumption Data
Table 2-6 shows the skewness values for as-consumed fish and shellfish consumption in
both g/day and g/kg-day or both for consumers-only offish and shellfish and per capita. For a
symmetric distribution the skewness is zero. Except for the log-transformed variables, all of the
skewness values in Table 2-6 are greater than 3.0, indicating the data have a skewed distribution
with a long tail on the high side of the distribution. For example, Figure 2-6 is a histogram of as-
consumed fish and shellfish consumption for consumers in Florida. The distribution is highly
skewed with a maximum observation (2,338) much higher than the range with most of the data.
Figure 2-7 shows a histogram of the log-transformed data. The log-transformed fish and
shellfish consumption data is reasonably normally distributed.
Table 2-6. Skewness of consumption variables, general population (as-
consumed weight)
State
CT
FL
MN
ND
Per capita
g/day
5.4
11.1
8.4
3.5
g/kg-day
5.2
9.5
9.9
5.9
Fish and Shellfish Consumers
g/day
5.3
9.5
8.5
3.5
g/kg-day
5.2
8.3
9.8
5.9
Log10(g/day) Log
-0.6
-0.1
-0.3
-0.1
io(g/kg-day)
-0.7
-0.1
-0.1
0.1
2-18
-------
80
60
«
20-
I
500
1000 1500
Grams per day, cooked
2000
Figure 2-6. Histogram of fish and shellfish consumption, consumers only, as-
consumed g/day.
2-19
-------
HI
O
12.5-
10.0-
7.5-
5.0-
2.5-
n
1 1
Log1Q(Grams per day, cooked)
Figure 2-7. Histogram of log-transformed fish and shellfish consumption,
consumers only, as-consumed g/day.
For consumers, the distributions of both the consumption in g/day and the consumption in
g/kg-day are skewed. However, the distribution of the log-transformed values (see Figures 2-5
and 2-7) is reasonably normally distributed. Because many statistical methods, such as analysis
of variance, have higher power when the data are normally distributed, comparing the mean of
the log-transformed consumption, or equivalently the geometric mean of the consumption, may
be more useful for comparing different categories of respondents. The arithmetic mean is more
useful for summarizing average consumption and estimating long-term exposure. However,
estimates of the arithmetic mean can be influenced by the small number of high consumption
values in the data file.
All of the statistics for the general population are weighted to estimate state population
values. For targeted populations, the statistics are unweighted. Means were calculated using
SAS PROC MEANS. Geometric means were calculated as: GeomMean =
2-20
-------
10x(Mean(loglO(data))), where the mean of the log-transformed data was calculated using SAS
PROC MEANS.
2.5.2. Calculation of Percentiles
There are many different methods for calculating percentiles. SAS provides several
different algorithms for calculating percentiles. However only one of those algorithms (SAS
PROC Univariate with PCTLDEF = 5) is implemented for weighted data. The disadvantage of
this option is that:
The estimated percentiles are always equal to the observed values even though values
between the observed values are possible responses. In this case interpolation between
the observed values can be used to improve the estimate.
When the number of observations is small the SAS estimates of extreme percentiles are
all equal to the minimum or maximum. However, with small sample sizes the upper
percentiles of the population may be greater than the observed maximum and lower
percentiles of the population may be less than the observed minimum.
To remedy these problems, a macro was used to calculate weighted percentile estimates
with the following characteristics:
Percentile estimates interpolate between the observed values to create a continuous
(rather than discrete) distribution function; and
Extreme percentiles are reported as missing when the expected percentile is greater than
the maximum or less than the minimum observed value.
Details of the algorithm are shown in Appendix C.
The percentiles of the consumption distribution are affected by the measurement method.
If the data are meant to represent the long-term consumption rate, the ideal consumption estimate
might be obtained using a detailed consumption diary with measurements offish and shellfish
weight, completed over an extended time period. Measurements based on a one-time
questionnaire with approximate weights (judged perhaps in terms of the equivalent number of
bread slices) will be less precise, sometimes over- and sometimes under-estimating the correct
long-term fish consumption. As a result, the measurements based on the survey questionnaire
are likely to be more spread out (have a higher standard deviation) than the true values. In
2-21
-------
particular, the estimated upper percentiles may over-estimate the long-term consumption rate for
those with the highest consumption rate.
2.5.3. Calculation of Confidence Intervals and Significance Tests
Appendix E provides confidence intervals and significance tests for differences in fish
and shellfish consumption among subgroups defined by selected analysis variables. The
confidence interval calculations assume the estimates have a normal distribution. This
assumption is valid if the sample size is large or the log-transformed data have a reasonably
normal distribution. When the data are highly skewed and the sample sizes in the subgroups are
small (perhaps <100), the confidence intervals and significance tests are approximate. The
significance tests evaluate if there are mean (or geometric mean) differences among all
categories being tested versus a null hypothesis that the means (or geometric means) in all
groups are identical.
Q-Q plots were used to evaluate the distribution of the data. Formal tests of normality to
evaluate if the data deviated from a normal distribution were not done. The confidence intervals
have the nominal coverage (in this case 95%) if the data has a normal distribution or if, due to
the central limit theorem, the sample size is large enough that the distribution of the mean has a
reasonably normal distribution. As long as the data are roughly normally distributed, as are the
log-transformed values, the results are not sensitive to the exact distribution. Since the log-
transformed fish consumption values have a roughly normal distribution, confidence intervals for
the geometric mean calculated from the log-transformed values should have the nominal
coverage. Confidence intervals for the arithmetic mean will be approximate, particularly when
based on small sample sizes.
Confidence intervals were calculated using SAS PROC SURVEYMEANS. Significance
tests for consumption differences among subgroups defined by categorical variables were
calculated using PROC SURVEYREG using analysis of variance. Significance tests for
differences among subgroups in the percentage of respondents that eat fish were calculated using
PROC SURVEYLOGISTIC. The significance tests excluded the "Unknown" category from the
calculations. However, the mean for the "Unknown" category is still shown in the tables along
with its confidence interval. Note that confidence intervals based on a small number of
respondents can be very imprecise. If there is only one respondent in a subgroup, the confidence
interval has width of zero, incorrectly implying great precision. Confidence intervals for
subgroups with only one respondent were set to missing (indicated by a period in the output).
Significance tests comparing the percentages of respondents that eat fish and shellfish among
subgroups can be adversely affected if all respondents in a subgroup have the same response. In
these cases, subgroups in which all responses were identical were combined with another
2-22
-------
subgroup to create an "Other" group for assessing significance. The tables indicate which
groups were combined to assess significance. For race-ethnicity in North Dakota, it was not
possible to assess significance because all responses were identical in all but one subgroup.
^-values below 0.05 are described as statistically significant. If the differences among
subgroups are statistically significant, apparent trends across ordered subgroups may be
described without a formal test for a linear trend.
2.6. ANALYSIS WEIGHTS
All of the surveys collected some fish consumption data from members of randomly
selected households within the surveyed states (referred to as the general population). Based on
the sample selection methods described in the survey reports, statistical weights were calculated
to calculate weighted estimates of population statistics, that is, summary statistics that describe
the entire population of the state. The weighted results are presented for the general population.
Details of the weight calculation can be found in Appendix A, Section A.6. Note that the
weights adjust for missing household members in Minnesota and North Dakota (if there are more
than five household members) and Florida (if the primary meal preparer was not available).
The Connecticut and Minnesota/North Dakota surveys also collected data from target
populations of particular interest. The targeted populations are listed in Table 2-2. In some
cases the populations were selected using a probability sample from a list of individuals (such as
a sample of those with fishing licenses in Minnesota). In such cases, weighted estimates could
be calculated for the populations. However, without additional information for calculating non-
response adjustments, the weighted estimates would be the same as the unweighted estimates.
Members of other populations were selected using non-probability methods, such as those that
happened to go to the WIC office at the time the interviewer was present. It is difficult to
describe the population represented by these respondents and there is no basis for constructing
weights. Unweighted results are presented for all targeted populations.
2-23
-------
3. RESULTS AND DISCUSSION
The combined database used in this analysis has information on fish consumption in four
states, Connecticut, Florida, Minnesota, and North Dakota. The data are useful for identifying
consumption patterns within states among respondents from the general population including
various age groups, gender, and different targeted populations. The survey estimates may differ
between states because of differences in fish consumption or because of differences in how the
data were collected in the three surveys. The discussion is divided into five characteristics of
interest within states, such as differences by income, gender and age, race-ethnicity and targeted
populations, education, and fish source (i.e., self-caught and store-bought). A selected number
of tables are presented in Section 3. This is followed by Section 3.7 discussing the differences
between data collection methods in the different surveys and the resulting effect on the
consumption estimates. The results and discussion focus primarily on patterns of differences
among subgroups within states. However, a general discussion of consumption differences
between states is also included at the end of this Section 3. Additional tables and plots showing
more detailed results are included in Appendix E.
3.1. DIFFERENCES BY INCOME
The Florida data show a statistically significant difference in the percentage of the
population that report eating fish among income categories (p <0.0001), corresponding to an
increase in the percentage with increasing household income. Differences in fish consumption
by income are shown in Table 3-1. For households with annual incomes below $20,000, 45.1%
of residents ate fish or shellfish in the previous week. For households with annual income above
$50,000 that percentage increased to 56.7%. Differences among income categories in Minnesota
and North Dakota show a similar pattern, but are only marginally significant (p = 0.067) in
Minnesota. Table 3-2 shows that for Minnesotans that eat fish and shellfish, increasing income
is associated with a significant decrease in the geometric mean grams offish and shellfish eaten
per day (p = 0.012). Table 3-3 shows that differences in the arithmetic mean were not
significant. Other differences associated with income were not significant and not consistent
across states.
5-1
-------
Table 3-1. Percent eating fish and shellfish, per capita, by state and income (with
95% CIs)
State
CT
FL
MN
ND
Subgroup
$0-20,000
$20,000-50,000
$50,000-
Unknown
$0-20,000
$20,000-50,000
$50,000-
Unknown
$0-20,000
$20,000-50,000
$50,000-
Unknown
$0-20,000
$20,000-50,000
$50,000-
Unknown
N
41
155
219
16
3,746
7,353
3,417
2,665
89
328
327
97
53
252
239
58
Weighted
N/1,000
312
1,179
1,778
119
3,408
6,814
3,250
2,480
373
1,802
2,155
570
56
268
251
63
Percent
86.8
85.3
83.8
73.4
45.1
50.0
56.7
45.4
91.0
91.3
97.9
92.9
94.0
93.3
97.1
93.1
Lower
Conf.
Limit
74.7
76.0
76.2
50.1
41.8
48.2
54.1
42.0
82.5
85.0
95.0
86.2
87.6
89.4
95.0
86.0
Upper
Conf.
Limit
98.8
94.7
91.4
96.7
48.4
51.7
59.4
48.8
99.6
97.6
100.8
99.7
100.5
97.3
99.3
100.2
/7-value
0.9161
O.0001
0.0668
0.1729
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
5-2
-------
Table 3-2. Geometric mean consumption, consumers only, by state and income (as-
consumed g/day with 95% CIs)
State
CT
FL
MN
ND
Subgroup
$0-20,000
$20,000-50,000
$50,000-
Unknown
$0-20,000
$20,000-50,000
$50,000-
Unknown
$0-20,000
$20,000-50,000
$50,000-
Unknown
$0-20,000
$20,000-50,000
$50,000-
Unknown
N
36
135
186
12
1,707
3,709
1,960
1,190
79
302
321
94
50
235
231
54
Weighted
N/1,000
270
1,006
1,490
87
1,537
3,404
1,844
1,126
339
1,645
2,109
530
53
250
244
59
Percent
16.72
18.59
16.54
25.13
35.75
35.95
34.83
32.08
19.16
10.49
9.88
10.98
13.71
10.04
12.93
12.24
Lower
Conf.
Limit
11.07
13.90
13.02
16.80
33.30
34.17
32.98
29.83
13.39
7.79
7.58
8.08
8.44
8.14
10.72
8.07
Upper
Conf.
Limit
25.25
24.85
21.01
37.58
38.38
37.82
36.77
34.49
27.41
14.12
12.88
14.94
22.26
12.38
15.60
18.55
/7-value
0.8365
0.6722
0.0118
0.0517
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
5-3
-------
Table 3-3. Mean consumption, consumers only, by state and income (as-
consumed g/day with 95% CIs)
State
CT
FL
MN
ND
Subgroup
$0-20,000
$20,000-50,000
$50,000-
Unknown
$0-20,000
$20,000-50,000
$50,000-
Unknown
$0-20,000
$20,000-50,000
$50,000-
Unknown
$0-20,000
$20,000-50,000
$50,000-
Unknown
N
36
135
186
12
1,707
3,709
1,960
1,190
79
302
321
94
50
235
231
54
Weighted
N/1,000
270
1,006
1,490
87
1,537
3,404
1,844
1,126
339
1,645
2,109
530
53
250
244
59
Percent
27.12
37.52
28.10
30.70
57.65
56.27
53.93
46.81
31.26
21.11
16.71
14.99
26.08
18.80
19.78
20.69
Lower
Conf.
Limit
16.49
22.60
23.14
20.06
51.15
52.28
50.35
43.28
17.29
11.06
12.42
11.33
10.03
14.40
16.12
11.67
Upper
Conf.
Limit
37.76
52.44
33.06
41.33
64.15
60.27
57.52
50.33
45.24
31.17
21.00
18.66
42.13
23.19
23.44
29.70
/7-value
0.5102
0.5398
0.1734
0.5251
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
3.2. DIFFERENCES BY GENDER AND AGE
By far the most statistically significant patterns in the data were differences by age.
These differences vary somewhat among states and genders. The percentages of individuals
eating fish and shellfish by state, age, and gender are presented in Table 3-4. As would be
expected, the percentage of individuals eating fish and shellfish is less for children than adults.
As children grow, the grams offish and shellfish consumed per day increases to around age 18,
after which the consumption rate continues to increase more slowly until about age 50. On a per
kilogram bodyweight basis, fish and shellfish consumption is generally higher for younger
children.
5-4
-------
Table 3-4. Percent eating fish and shellfish, per capita, by state and
age/gender (with 95% CIs)
State Subgroup
CT Child 1-5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL Child 1-5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
MN Child 1-5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
Lower
Weighted Conf.
N N/1,000 Percent Limit
28
28
22
17
88
79
14
81
63
11
1,107
943
865
1,636
2,546
2,367
1,702
2,673
2,347
995
47
47
68
47
133
162
55
120
156
6
274
259
201
141
656
579
119
600
461
99
1,138
962
849
1,518
2,296
2,142
1,567
2,411
2,127
941
437
299
337
331
723
854
275
731
852
62
47.6
80.0
86.1
79.9
87.1
90.9
70.5
92.9
90.5
76.1
37.6
39.3
42.8
48.7
56.4
55.7
45.9
52.9
54.4
39.2
97.4
88.4
92.8
96.0
95.0
94.9
92.3
96.0
99.8
1.5
25.1
61.2
67.4
57.8
80.4
83.6
48.3
87.3
82.4
48.4
33.7
35.5
38.4
45.8
54.2
53.5
43.0
50.9
52.2
34.3
92.0
73.0
83.5
89.1
89.3
90.7
77.8
91.8
99.5
-2.0
Upper
Conf.
Limit /7-value
70.1 0.0001
98.8
104.9
102.0
93.9
98.1
92.7
98.6
98.6
103.9
41.4 O.0001
43.2
47.2
51.5
58.3
57.8
48.7
54.8
56.6
44.1
102.8 0.0001
103.8
102.1
103.0
100.6
99.0
106.8
100.3
100.1
5.0
3-5
-------
Table 3-4. Percent eating fish and shellfish, per capita, by state and
age/gender (with 95% CIs) (continued)
State Subgroup
ND Child 1-5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
Weighted
N N/1,000 Percent
31
46
58
45
99
102
37
92
85
7
30
44
54
47
105
116
39
99
97
7
91.5
92.4
97.2
85.6
98.4
93.6
100.0
97.8
94.3
75.2
Lower
Conf.
Limit
82.6
80.7
93.5
73.2
96.1
89.1
100.0
94.8
89.4
39.5
Upper
Conf.
Limit /7-value
100.4 0.0942
104.0
100.9
98.0
100.7
98.1
100.0
100.9
99.2
111.0
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
The Florida data show a significant difference in the percentage of the population that
reports eating fish and shellfish among nine age and gender categories (p < 0.0001),
corresponding to increasing percentages of consumers with increasing age, up to age 49. The
percentage for those 50 and over is similar to that for ages 30-49. An estimated 37.6% of
children aged 1-5 consume fish or shellfish during a one week period. For adults aged 30-49,
56.4% of females and 52.9% of males consume fish or shellfish in a week. Due to the survey
procedures used, the Florida data will underestimate the percentage of children that consume fish
and shellfish. Thus, the actual percentages for children may be higher, possibly making the trend
versus age less significant.
In the Connecticut data, the percentage of the population that eats fish and shellfish is
lower (47.6%) for children aged 1-5 and higher (roughly 80-90%) for other age groups. This
difference is statistically significant (p <0.0001). Among fish and shellfish consumers, adults
consume significantly more fish and shellfish than children (p <0.0001). Consumption generally
increases from the younger to older age groups, with a slight decrease for the oldest age group.
On a per bodyweight basis, consumption for fish consumers varies among categories. However,
the differences are statistically significant (p = 0.0006 on a per capita basis andp = 0.0004 for
fish and shellfish consumers). For consumers, consumption per kilogram per day has a generally
decreasing trend from the youngest to the oldest age group.
-------
Of those that do consume fish and shellfish, differences among age and gender categories
are significant (p <0.0001): children in Florida aged 1-5 consume 28.98 g/day; adults aged 30-
49 consume about 60 g/day (61.17 for females and 66.19 for males) (see Table 3-5). Above age
50, consumption drops off. The per capita fish and shellfish consumption is a combination of the
percentage of respondents that consume fish and the amount consumed. The per capita
consumption also shows significant differences (p <0.0001) corresponding to a significant
increase to age 50 followed by a modest decrease for consumers over 50. The decrease appears
to be due to a decrease in the quantity consumed per day rather than a decrease in the percentage
of people that consume fish and shellfish.
Because bodyweight is closely related to age, the patterns in fish and shellfish
consumption in g/kg of bodyweight versus age are quite different than for consumption on a
g/day basis. In Florida, differences in per capita fish and shellfish consumption per kilogram
bodyweight per day were statistically significant (p <0.0001), decreasing from 0.89 g/kg-day for
children aged 1-5 to 0.37 g/kg-day for children 11-15, increasing slightly for middle aged adults
and decreasing slightly to 0.41 g/kg-day for females 50 and over and 0.38 g/kg-day for males 50
and over (see Table 3-6). Similar patterns were seen for consumers offish and shellfish (see
Appendix E). Any adjustment for not collecting away-from-home fish and shellfish
consumption for children will make the differences more significant.
Table 3-5. Mean consumption, consumers only, by state and age/gender (as-
consumed g/day) (with 95% CIs)
State Subgroup
CT Child 1-5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
N
14
22
19
14
77
72
10
75
57
9
Weighted
N/1,000
131
207
173
113
571
526
84
557
417
75
Mean
8.92
14.48
16.35
44.94
36.16
31.45
16.92
44.63
32.19
9.02
Lower
Conf.
Limit
5.60
9.82
8.78
7.85
21.93
24.28
11.80
33.43
24.74
-0.72
Upper
Conf.
Limit
12.25
19.14
23.91
82.03
50.38
38.62
22.05
55.82
39.64
18.76
/7-value
O.0001
5-7
-------
Table 3-5. Mean consumption, consumers only, by state and age/gender (as-
consumed g/day) (with 95% CIs) (continued)
State Subgroup
FL Child 1-5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
MN Child 1-5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
ND Child 1-5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
N
421
376
365
791
1,446
1,315
785
1,406
1,272
389
46
43
63
44
128
150
52
115
154
1
28
43
56
39
97
95
37
90
80
5
Weighted
N/1,000
428
378
364
739
1,292
1,192
719
1,275
1,156
369
425
265
313
318
686
810
254
702
851
1
28
41
53
40
103
108
39
97
91
6
Mean
28.98
30.52
40.58
54.62
61.17
47.18
72.00
66.19
56.85
41.83
8.17
10.76
12.89
38.44
16.31
24.53
8.00
21.16
20.88
1.05
11.91
15.98
20.32
12.10
18.08
25.87
16.36
18.75
24.00
31.88
Lower
Conf.
Limit
26.53
27.51
35.47
45.97
56.31
44.64
60.33
60.90
52.46
34.22
4.77
6.09
8.08
-8.98
13.58
17.89
5.77
15.16
16.98
-
3.20
9.76
11.86
7.92
13.76
17.75
9.96
15.03
17.02
5.11
Upper
Conf.
Limit
31.44
33.54
45.70
63.28
66.02
49.71
85.66
71.48
61.06
49.43
11.57
15.43
17.70
85.87
19.04
31.18
12.22
27.15
24.79
-
20.62
22.19
28.78
16.27
22.40
34.00
22.77
22.47
32.97
58.65
/7-value
O.0001
0.0001
0.0155
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
-------
Table 3-6. Mean consumption per bodyweight, general population, per
capita, by state and age/gender (as-consumed g/kg-day) (with 95% CIs)
State Subgroup
CT Child 1-5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL Child 1-5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
MN Child 1-5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
N
26
26
21
17
85
77
14
80
63
11
1,102
938
864
1,537
2,264
2,080
1,638
2,540
2,206
198
47
46
68
47
132
162
55
120
155
Weighted
N/1,000
253
239
193
141
634
563
119
594
461
99
1,134
956
848
1,477
2,178
2,025
1,551
2,383
2,090
185
437
298
337
331
722
854
275
731
851
Mean
0.32
0.51
0.27
0.67
0.46
0.43
0.16
0.47
0.35
0.09
0.89
0.44
0.37
0.44
0.53
0.41
0.44
0.43
0.38
0.35
0.57
0.33
0.22
0.67
0.24
0.34
0.10
0.24
0.24
Lower
Conf.
Limit
0.13
0.33
0.13
0.06
0.32
0.33
0.08
0.37
0.27
-0.03
0.75
0.37
0.29
0.35
0.49
0.38
0.36
0.39
0.35
0.26
0.30
0.17
0.15
-0.20
0.20
0.25
0.06
0.16
0.20
Upper
Conf.
Limit ^-value
0.50 0.0006
0.69
0.41
1.28
0.61
0.54
0.25
0.57
0.42
0.21
1.02 O.0001
0.50
0.44
0.52
0.57
0.44
0.52
0.47
0.41
0.45
0.83 0.0001
0.50
0.29
1.53
0.29
0.43
0.14
0.31
0.29
3-9
-------
Table 3-6. Mean consumption per bodyweight, general population, per
capita, by state and age/gender (as-consumed g/kg-day) (with 95% CIs)
(continued)
State Subgroup
Unknown
ND Child 1-5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
N
5
30
44
55
42
95
99
36
90
81
3
Weighted
N/1,000
61
30
42
52
43
101
112
38
97
92
3
Mean
0.00
0.67
0.51
0.40
0.18
0.28
0.38
0.22
0.22
0.29
0.11
Lower
Conf.
Limit
-0.00
0.25
0.28
0.22
0.12
0.21
0.25
0.13
0.17
0.19
-0.08
Upper
Conf.
Limit
0.01
1.08
0.75
0.58
0.24
0.35
0.50
0.31
0.26
0.39
0.29
/7-value
0.0617
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
When looking at gender alone, a slightly higher percentage of Florida females consume
fish and shellfish than males (51.5% versus 49.1%,p = 0.0001). Per capita consumption offish
and shellfish is higher for males than for females (29.15 versus 26.29 g/day,p = 0.0113).
However, because males generally weigh more than females, females consume slightly more fish
and shellfish per kilogram bodyweight per day than males (0.50 versus 0.44 g/kg-day, p =
0.0040). These results are shown in Tables 3-7, 3-8 and 3-9.
3-10
-------
Table 3-7. Percent eating fish and shellfish, per capita, by state and gender
(with 95% CIs)
State
CT
FL
MN
ND
Subgroup
Male
Female
Male
Female
Unknown
Male
Female
Male
Female
N
205
226
8,262
8,110
809
422
419
288
314
Weighted
N/1,000
1,617
1,771
7,662
7,517
774
2,497
2,403
306
332
Percent
85.1
83.4
49.1
51.5
36.1
95.3
93.4
96.3
93.5
Lower
Conf.
Limit
78.6
77.5
47.7
50.1
30.4
91.4
89.5
94.1
90.2
Upper
Conf.
Limit
91.7
89.3
50.4
52.9
41.8
99.1
97.3
98.6
96.7
/7-value
0.5786
0.0001
0.3933
0.1378
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
Table 3-8. Mean consumption, general population, per capita, by state and
gender (as-consumed g/day with 95% CIs)
State
CT
FL
MN
ND
Subgroup
Male
Female
Male
Female
Unknown
Male
Female
Male
Female
N
205
206
8,262
8,110
809
422
419
288
314
Weighted
N/1,000
1,617
1,771
7,662
7,517
774
2,497
2,403
306
332
Mean
27.45
25.56
29.15
26.29
14.73
16.35
19.85
20.21
17.85
Lower
Conf.
Limit
21.65
19.54
27.29
24.44
10.22
13.43
12.19
16.02
14.13
Upper
Conf.
Limit
33.25
31.59
31.01
28.14
19.25
19.27
27.51
24.40
21.58
p-value
0.5193
0.0113
0.3630
0.1626
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
3-11
-------
Table 3-9. Mean consumption per bodyweight, general population, per
capita, by state and gender (as-consumed g/kg-day with 95% CIs)
State
CT
FL
MN
ND
Subgroup
Male
Female
Male
Female
Unknown
Male
Female
Male
Female
N
201
219
7,911
7,426
30
419
418
276
299
Weighted
N/1,000
1,581
1,715
7,568
7,229
30
2,495
2,402
293
317
Mean
0.39
0.43
0.44
0.50
0.41
0.26
0.36
0.32
0.32
Lower
Conf.
Limit
0.31
0.34
0.41
0.46
0.12
0.22
0.22
0.22
0.25
Upper
Conf.
Limit
0.46
0.53
0.47
0.53
0.70
0.31
0.50
0.41
0.40
p-value
0.2782
0.0040
0.1897
0.9048
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
The estimates for other states are less precise and the patterns in the data by age are less
distinct than in the Florida data, most likely due to differences in sample size. In Minnesota and
North Dakota the percentage of the population that consumes fish and shell fish is high and
relatively constant for all age groups, with a slight, but inconsistent increase with age. The
differences among nine age and gender groups are significant in Minnesota (p <0.0001) and
marginally significant in North Dakota (p = 0.094). Per capita fish and shellfish consumption is
higher in adults than children, with significant differences in Minnesota (p = 0.0001) and North
Dakota (p = 0.039). The grams offish and shellfish consumed per day generally increases from
the youngest to the oldest respondents. The consumption per kilogram bodyweight generally
decreases from the youngest to the oldest age groups, with significant differences among
age/gender categories in Minnesota (p <0.0001) and significant or marginally significant
differences in North Dakota (p = 0.062 per capita, p = 0.0001 for differences in the geometric
mean for consumers only). However the decrease is not as dramatic as in the Florida data.
In the Connecticut, Minnesota, and North Dakota data, a general trend line for
consumption can be drawn between the youngest age groups, (1-5 and 6-10) and the older aged
groups (30-49 and 50 and over). However, the consumption for intermediate aged groups, those
11-15 and 16-29, are not always consistent with the overall trend. The differences may be due
to differences in the survey procedures, imprecision in the estimates, or to real patterns in the
data that are not easy to generalize. One possible explanation is that young women increase their
consumption offish and shellfish at an earlier age than men who may not increase fish and
3-12
-------
shellfish consumption until around age 30. In addition, there is a wider confidence interval for
females age 16-29 compared to that of other age groups. More investigation would be required
to provide additional insight into the patterns across age categories.
Across all states there were generally consistent and significant age and gender trends,
with more adults reporting consumption offish and shellfish than children and adults consuming
more fish and shellfish than children on a g/day basis. The fish and shellfish consumption per
kilogram bodyweight is generally higher for children than adults. However the difference is not
always significant. The consumption broken out by age categories is generally consistent with
this summary, with the youngest children generally having the lowest overall consumption but
the highest consumption per kilogram bodyweight. There are also consistent differences
between genders, with males consuming more fish and shellfish, but less on a per kilogram
bodyweight basis. However, the differences are significant only in Florida.
3.3. DIFFERENCES BY RACE-ETHNICITY AND TARGETED POPULATIONS
In Connecticut, Minnesota, and North Dakota members of targeted populations were
sampled separately in addition to being represented in the state's general population sample. In
most cases, the populations were defined by race-ethnicity or factors closely related to race-
ethnicity. As a result, this section discusses both the results from the state's general population
and from the targeted populations.
In the Florida survey, 50.9% of non-Hispanic Whites (hereafter Whites) reported eating
fish or shellfish in the previous week (Table 3-10). A lower percentage of non-Hispanic Black
(46.2%) and Hispanic (45.0%) respondents reported eating fish and shellfish during the previous
week. These differences are significant (p = 0.0341). For those that consumed fish and shellfish,
the consumption rate offish and shellfish was less for Whites than for minorities (p = 0.0726;
Table 3-11). Similar patterns were observed on a per kilogram bodyweight basis (p = 0.0025;
Appendix E). The higher percentage of Whites eating fish and shellfish and the lower daily
consumption combine so that the per capita fish and shellfish consumption is similar for Whites,
Blacks, and Hispanics.
In Connecticut, the number of minority respondents from the general population is small.
However, differences among race-ethnicity groups are statistically significant (p <0.0001) for all
of the dependent variables. The general population respondents included 9 classified as Black of
which only 3 reported fish or shellfish consumption (Tables 3-10 and 3-11). Thus sample sizes
for Blacks are particularly small. In Connecticut, the estimated fish consumption and percentage
of respondents consuming fish and shellfish is less for Blacks than other race-ethnicity groups.
However, the sample size for Blacks was small in the Connecticut survey. Therefore,
generalizations cannot be made with regard to this minority group. While a higher percentage of
3-13
-------
Whites consume fish and shellfish than Hispanics and other minorities, Whites consume less fish
and shellfish than Hispanics and other minorities.
Table 3-10. Percent eating fish and shellfish, per capita, by state and race-
ethnicity (with 95% CIs)
State
CT
FL
MN
ND
Subgroup
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
White, Non-Hispanic
Black, Non-Hispanic
Other
Unknown
N
380
9
20
20
2
12,957
1,842
1,673
382
327
779
1
3
19
39
551
2
17
32
Weighted
N/1,000
2,968
66
178
155
21
11,887
1,690
1,719
330
325
4,473
1
50
173
204
585
2
16
36
Percent
88.0
33.5
70.9
54.7
43.4
50.9
46.2
45.0
50.3
42.7
93.8
0.0
100.0
100.0
100.0
94.8
100.0
100.0
93.5
Lower
Conf.
Limit
83.2
-22.9
32.5
19.4
43.4
49.6
42.1
40.8
41.2
33.2
90.3
-
100.0
100.0
100.0
92.6
100.0
100.0
85.3
Upper
Conf.
Limit
92.8
89.9
109.4
90.0
43.4
52.3
50.2
49.1
59.4
52.2
97.3
-
100.0
100.0
100.0
97.0
100.0
100.0
101.7
/7-value
0.0001
0.0341
0.0010
-
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population
3-14
-------
Table 3-11. Mean consumption, consumers only, by state and race-ethnicity
(as-consumed g/day with 95% CIs)
State
CT
FL
MN
ND
Subgroup
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
White, Non-Hispanic
Hispanic
Other
Unknown
White, Non-Hispanic
Black, Non-Hispanic
Other
Unknown
N
338
3
15
12
1
6,607
867
762
191
139
735
3
19
39
521
2
17
30
Weighted
N/1,000
2,612
22
126
85
9
6,053
780
773
166
139
4,197
50
173
204
555
2
16
33
Mean
30.67
13.03
41.85
47.01
0.99
53.17
57.53
59.17
71.11
57.92
17.17
45.01
56.04
22.25
20.16
15.72
20.89
17.37
Lower
Conf.
Limit
25.65
8.45
18.18
0.39
-
50.77
50.59
48.01
50.70
37.09
14.43
-10.06
-45.60
12.13
15.91
15.03
14.32
8.07
Upper
Conf.
Limit
35.67
17.61
65.52
93.63
-
55.58
64.47
70.33
91.51
78.76
19.91
100.08
157.68
32.38
24.41
16.41
27.46
26.66
/7-value
O.OOOl
0.0726
0.5354
0.0578
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
Table 3-12 and 3-13 show that anglers had higher consumption rates than the general
population in Connecticut, Minnesota and North Dakota. In Connecticut, Asians, commercial
fishermen, Expanded Food and Nutrition Education Program (EFNEP) and Women, Infants and
Children Program (WIC) participants also had higher fish consumption rates than the state's
general population. In Minnesota, new mothers had lower fish consumption rates than the state's
general population. American Indians in Minnesota consumed less fish and shellfish than other
groups (Table 3-12 and 3-13). No significance tests were performed to evaluate these
differences due to differences in how the populations were sampled.
3-15
-------
Table 3-12. Mean consumption, consumers only, by state and targeted
populations (as consumed, g/day with 95% CIs)
State
CT
FL
MN
ND
Subpopulation
General
Angler
Aquaculture Student
Asians
Commercial Fishermen
EFNEP Participants
WIC Participants
General
General
American Indians
Anglers
New Mothers
General
American Indians
Anglers
Weighted
N N/1,000
369 2,854
257
19
396
171
60
557
8,566 7,912
796 4,623
196
1,127
352
570 606
78
825
Mean
31.41
49.16
19.33
57.37
48.78
59.95
45.85
54.65
19.15
13.20
21.68
16.83
20.01
26.04
20.57
Lower
Conf.
Limit
25.84
40.87
4.23
48.50
36.47
18.67
38.57
52.14
14.66
8.80
17.90
12.75
16.28
11.50
18.32
Upper
Conf.
Limit
36.98
57.44
34.42
66.23
61.08
101.22
53.14
57.16
23.63
17.60
25.46
20.91
23.75
40.58
22.81
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
3-16
-------
Table 3-13. Mean consumption, per capita, by state and targeted
populations (as consumed, g/day with 95% CIs)
State
CT
FL
MN
ND
Subpopulation
General
Angler
Aquaculture Student
Asians
Commercial Fishermen
EFNEP Participants
WIC Participants
General
General
American Indians
Anglers
New Mothers
General
American Indians
Anglers
N
431
266
25
402
178
71
703
17,181
841
221
1,171
415
602
133
871
Weighted
N/1,000
3388
-
-
-
15,952
4,900
-
-
639
-
-
Mean
26.46
47.49
14.69
56.51
46.86
50.66
36.33
27.10
18.06
11.70
20.87
14.28
18.99
15.27
19.48
Lower
Conf.
Limit
21.30
39.31
2.32
47.73
34.78
15.51
30.35
25.61
13.76
7.83
17.17
10.61
15.39
5.74
17.26
Upper
Conf.
Limit
31.62
55.68
27.05
65.29
58.94
85.81
42.31
28.60
22.37
15.58
24.57
17.94
22.58
24.80
21.70
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
Overall, there were often significant differences among race-ethnicity groups. The
differences were not always consistent except that a higher percentage of Whites reported
consuming fish or shellfish and Whites generally consumed less than minorities. The differences
between the general populations and the targeted populations in each state are generally
consistent with this summary.
3.4. DIFFERENCES BY EDUCATION
The Florida data show significant differences in the percentage of the population that
consumed fish and shellfish among education groups (p < 0.0001). The estimates correspond to
an increase in the percentage of the population that reported eating fish with an increase in
education of the head of household. When the head of household had less than a high school
education, 40.7% of occupants consumed fish or shellfish in a period of a week (Table 3-14).
For college graduates, the percentage increased to 53.6%. Similar, though non-significant,
trends were seen in the Minnesota and North Dakota percentages.
3-17
-------
Table 3-14. Percent eating fish and shellfish, per capita, by state and
education (with 95% CIs)
State
CT
FL
MN
ND
Subgroup
0-11 years
High School
Some College
College Graduate
0-11 years
High School
Some College
College Graduate
Unknown
0-11 years
High School
Some College
College Graduate
Unknown
0-11 years
High School
Some College
College Graduate
Unknown
N
13
89
66
263
1,744
5,677
5,261
4,367
132
46
236
260
256
43
31
143
195
196
37
Weighted
N/1,000
97
682
504
2,105
1,523
5,118
4,948
4,240
123
214
1,332
1,330
1,808
215
35
144
212
206
42
Percent
100.0
85.6
89.3
81.9
40.7
47.3
51.5
53.6
39.4
86.2
92.9
95.3
95.0
99.7
87.6
97.4
93.9
96.8
87.2
Lower
Conf.
Limit
100.0
73.0
81.3
75.1
36.2
45.2
49.0
51.3
28.5
70.9
84.6
91.4
90.9
99.1
74.1
94.8
89.2
94.2
77.7
Upper
Conf.
Limit
100.0
98.2
97.3
88.7
45.3
49.3
54.0
56.0
50.3
101.6
101.1
99.3
99.2
100.3
101.1
99.9
98.7
99.4
96.6
/7-value
0.3444
O.0001
0.4294
0.1680
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
3.5. DIFFERENCES BY FISH AND SHELLFISH SOURCE
For those that consumed fish and shellfish, respondents were classified by the source of
fish and shellfish they eat, using the categories: eats only self-caught fish, only store-bought fish,
or both store-bought and self-caught fish. A second set of categories was also defined to focus
on whether respondents consume self-caught fish from freshwater or estuaries (classified as
exclusively, sometimes, and never). Fish consumption was then summarized for these categories
3-18
-------
of respondents. Since the differences using just the store-bought and self-caught categories were
more significant, only those results are discussed in this report.
In all four states, using consumption in g/day or g/kg-day, and using the mean or the
geometric mean, fish consumption is highest for those that consume both self-caught and store-
bought fish and shellfish as opposed to either only self-caught or only store-bought fish and
shellfish. In all cases the difference is statistically significant. As shown in Table 3-15, those
that eat both caught and bought fish in Florida consume 111.97 g/day compared to 49.64 g/day
for those that consume only purchased fish and shellfish. In Connecticut, those that eat both self-
caught and bought fish consume 38.47 g/day compared to 29.80 g/day for those that consume
only purchased fish and shellfish. In Minnesota and North Dakota these numbers are 24.30 and
23.31 g/day for consumers of bought and self-caught fish and 12.23 and 13.50 g/day for
consumers of purchased fish and shellfish only. In all but Florida the number of respondents that
consume only self-caught fish was relatively small. In Florida, the consumption rate for those
that eat only self-caught fish and shellfish is similar to that for those that eat only store-bought
fish and shellfish.
Table 3-15. Mean consumption, consumers only, by state and angler status
(as-consumed g/day with 95% CIs)
State
CT
FL
MN
ND
Subgroup
Eats Caught Only
Eats Caught and Bought
Eats Bought Only
Eats Caught Only
Eats Caught and Bought
Eats Bought Only
Eats Caught Only
Eats Caught & Bought
Eats Bought Only
Eats Caught Only
Eats Caught and Bought
Eats Bought Only
N
1
74
294
600
802
7,164
38
556
202
33
376
161
Weighted
N/1,000
9
559
2286
493
667
6,752
221
2,747
1,655
36
403
167
Mean
0.99
38.47
29.80
45.59
111.97
49.64
6.80
24.31
12.23
13.31
23.31
13.50
Lower
Conf.
Limit
-
25.44
24.51
40.35
97.35
47.34
2.84
17.70
9.27
6.08
18.71
7.88
Upper
Conf.
Limit
-
51.51
35.10
50.83
126.60
51.95
10.76
30.91
15.19
20.55
27.90
19.12
/7-value
O.0001
0.0001
0.0001
0.0054
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
3-19
-------
3.6. PROCEDURAL DIFFERENCES AMONG SURVEYS
Although there are multiple differences between the survey methods used for the three
surveys, two differences in the Florida survey—the shorter recall period used and the exclusion
of away-from-home fish and shellfish consumption—may have an important effect on the
results. This section discusses the likely effect of these differences.
Differences in fish and shellfish consumption between states depend in part on the
different recall periods used in the surveys. Respondents were asked how frequently they ate fish
and shellfish (such as number of meals per week or month) and how much was eaten on each
occasion. An individual's fish consumption is based on the product of the frequency (number of
meals or servings per unit time) and the typical quantity per meal, summed over all species.
Between-state comparisons of the per capita (or population average for both consumers
and non-consumers) fish and shellfish consumption are affected by the respondent's ability to
recall their fish and shellfish consumption and by differences in reporting in response to different
questionnaires. The respondent's ability to recall their fish and shellfish consumption may be
affected by the length of the recall period (Medlin and Skinner, 1988). Even when the
respondents have perfect recollection and reporting of their fish and shellfish consumption, the
estimates of the proportion of respondents that report eating fish or shellfish and the consumption
rate for those that report consuming fish and shellfish will be affected by the length of the recall
period. Therefore, these estimates are not directly comparable between surveys.
The effect of recall period on the calculations is illustrated by the following simplified
example. Assume for now that respondents have perfect recollection and reporting of their fish
and shellfish consumption. If a respondent eats 112 g of salmon once every 4 weeks on a regular
schedule and the recall period is 1 week, there is a 25% chance that the recall period will fall in
the week when salmon is consumed. If so, the estimated consumption rate is the weight of
salmon consumed divided by the length of the recall period in days, i.e., 112 g per 7 days = 16
g/day. If the recall period does not fall on a week when salmon is consumed, the salmon
consumption rate is 0 g/day. For a group of respondents with similar salmon consumption
patterns, 25% of respondents will be classified as eating salmon, for those that eat salmon the
consumption rate is 16 g/day. The consumption rate is 0 for the 75% of respondents that did not
eat salmon in the recall period. Across the targeted population, the standard deviation of the
consumption rate is 6.9 g/day. The average consumption across the entire targeted population is
25% x 16 = 4 g/day.
If the recall period is 4 weeks, every respondent in the population will report eating one
serving of salmon. Across the population of similar respondents, 100% of respondents will be
classified as eating salmon with an average consumption rate of 4 g/day. In this example, the
estimated average consumption rate is the same, regardless of the recall period. However, the
3-20
-------
percent of respondents that reported eating salmon and the consumption rate for those that
consume salmon depend on the recall period. Table 3-16 summarizes the results for this
illustrative example. Although the numbers would change somewhat, the relationships would be
the same if the population also included some respondents that never consume salmon.
Table 3-16. Illustrative results for two recall periods
Result
Percentage of respondents eating salmon
Salmon consumption rate for consumers (those
that ate salmon during the recall period)
Per capita salmon consumption
Standard deviation of salmon consumption rate
per capita
Recall
1 Week
25%
16 g/day
4 g/day
6.9 g/day
Period
4 Weeks
100%
4 g/day
4 g/day
0 g/day
Although the arguments are somewhat more complicated, still assuming that respondents
have perfect recollection and reporting of their fish and shellfish consumption, similar
conclusions apply to the summary statistics calculated from the all of surveys. In general:
Estimates of the percentage of the population that consume fish and shellfish depend on
the recall period. Longer recall periods are associated with higher percentages of
respondents consuming fish and shellfish during the recall period.
Estimates of the fish and shellfish consumption rate (in g/day) for those that consume fish
also depend on the length of the recall period. Longer recall periods are associated with
lower estimates offish and shellfish consumption rates by those that consume fish and
shellfish.
Per capita estimates of the fish and shellfish consumption rate averaged across the
population do not depend on the length of the recall period. However, the precision of
the per capita estimate does. Longer recall periods are associated with more precise
estimates of the population average consumption.
Thus, still assuming that respondents have perfect recollection and reporting of their fish
and shellfish consumption, the per capita fish consumption estimates may be compared among
states with less uncertainty than the consumers only estimates. However, any comparison of (1)
3-21
-------
the percentage of respondents that consume fish and shellfish during the recall period, or (2) the
consumption rate for those that eat fish and shellfish must take into account the length of the
recall period (Tran et al., 2004; Stern et al., 1996). Since the same recall period is used for all
respondents within a state, comparison of groups within states has no similar problem.
The patterns cited in the bullets above can be seen in the data, although other factors also
affect the differences. The estimated per capita fish consumption is similar in Florida and
Connecticut, 26.9 and 27.2 g/day respectively. However, half of the Florida respondent's
reported eating fish or shellfish in the 7 day recall period; whereas, approximately 88% of the
Connecticut respondents reported eating fish or shellfish over a longer 1 year recall period. For
those that ate fish or shellfish, the consumption was 31 g/day in Connecticut and 55 g/day in
Florida. As a result, the estimates of the percentage of respondents that consumed fish and
shellfish and the consumption rate for those that consumed fish or shellfish during the recall
period are not directly comparable between surveys.
The simplifying assumption in the previous paragraphs may not be realistic. In particular,
the following factors may also affect comparison of the fish and shellfish consumption between
states:
• The surveys used different operational definitions of a serving or meal;
• The surveys used different questions to assess the quantity offish and shellfish consumed
in an average serving;
• Beside differences in the recall period, the surveys used different questions to assess the
frequency offish and shellfish consumption;
• The surveys used different methods to summarize the data and convert between uncooked
and as-consumed weights (see Section 2.1);
• The methods used to aid recall, such as providing a list of possible fish species, differed
among surveys; and
• With longer recall periods, respondents will have more difficulty remembering the fish
and shellfish that they consumed, how often it was consumed, and quantity consumed.
As a result, respondents may compensate by overestimating or underestimating their fish
and shellfish consumption.
One important difference between the survey methods in Florida and those in the other
states is that the Florida questionnaire procedures did not collect away-from-home consumption
for household members other than the randomly selected adult respondent (RSA). As a result the
fish consumption in Florida is underestimated. Using fish and shellfish consumption in raw
3-22
-------
g/day, the mean fish and shellfish consumption for the RSA is 2.12 times the at-home
consumption. Assuming this ratio holds for the as-consumed consumption rate and holds for
children also, the fish and shellfish consumption for children, including the away-from-home
consumption, would be 2.12 times the values shown in the tables. Since away-from-home
proportion offish and shellfish consumption for children is likely to be lower than for adults, the
fish and shellfish consumption for children is likely to be between 1 and 2.12 times the values in
the tables. When combining the RSA, for whom away-from-home consumption is available, and
the other adults in the home, for whom away-from-home consumption was not available, the fish
consumption for adults would be higher than the tabled values by a factor of 1.46.
3.7. COMPARISONS AMONG STATES
The procedural differences described in Section 3.7 make comparisons among states
difficult to interpret. Therefore, the tables and discussions emphasize comparison among
subgroups within states rather than comparisons among states. However, some observations can
be made.
The percentages of state residents that consumed fish and shellfish during the recall
period were 50% for Florida, 90% for Connecticut, 94% for Minnesota, and 95% for North
Dakota (Table 3-17). The proportion of state residents that consumed fish and shellfish, as
estimated from the survey data, was much lower for Florida than for the other states. This
difference is due in part to the shorter recall period in the Florida survey. The Florida survey
used a seven day recall period. The Minnesota and North Dakota surveys used a 1 year recall
period. The recall period for the Connecticut survey was not specified, however the
questionnaire response category "# times per year" for the question "How often do you eat each
type of seafood?" allows for recall periods of up to a year. The proportion of respondents eating
fish and shellfish in Connecticut is more comparable to the proportion in Minnesota and North
Dakota. However, differences may still be due to differences in how the questions were asked.
3-23
-------
Table 3-17. Percent eating fish and shellfish, per capita, by state and
subpopulation (with 95% CIs)
State
CT
FL
MN
ND
Population
General
Angler
Aquaculture Student
Asians
Commercial Fishermen
EFNEP Participants
WIC Participants
General
General
American Indians
Anglers
New Mothers
General
American Indians
Anglers
Weighted
N N/1000
431 3388
266
25
402
178
71
703
17,181 15,952
841 4,900
221
1,171
415
602 639
133
871
Percent
0.88
0.97
0.76
0.99
0.97
0.85
0.79
0.50
0.94
0.89
0.96
0.85
0.95
0.59
0.95
Lower
Conf.
Limit
0.83
0.94
0.56
0.97
0.94
0.71
0.76
0.48
0.91
0.82
0.95
0.79
0.93
0.43
0.93
Upper
Conf.
Limit
0.92
0.99
0.96
1.00
0.99
0.98
0.82
0.51
0.98
0.96
0.98
0.90
0.97
0.74
0.97
CT = Connecticut, FL = Florida, MN = Minnesota, ND = North Dakota, N = population.
For those that consumed fish and shellfish, the quantity consumed depends on the recall
period. The average fish consumption for those that reported any consumption is 31.41 g/day in
Connecticut, 54.65 g/day in Florida, 19.15 g/day in Minnesota, and 20.01 g/day in North Dakota
(Table 3-12). These consumption rates are for the fish and shellfish as-consumed. The higher
value for Florida reflects, in part, the shorter recall period.
The per capita fish and shellfish consumption rate across the entire population (those that
eat fish and those that do not) is not directly affected by the length of the recall period and thus is
more comparable between states. However uncertainties arise because there might be
differences in the survey estimates due to how the fish consumption data were collected by the
different surveys and how well the respondents remembered the fish and shellfish they ate. Each
survey obtained information on the frequency and quantity using different questions. Some
3-24
-------
questions may elicit better recall or different responses than others, resulting in differences in the
per capita fish and shellfish consumption between states. In general, it is difficult to say how the
different data collection methodologies might affect the per capita estimates.
In Florida, the survey procedures are likely to underestimate the total fish and shellfish
consumption. The Florida survey collected away-from-home fish consumption for the randomly
selected adult respondent, but not for other adults in the household and not for children. As a
result, the estimated per capita fish and shellfish consumption in Florida is likely to
underestimate the true amount. The estimated average per capita consumption for the general
population was 27.10 g/day in Florida, 26.46 g/day in Connecticut, 18.06 g/day in Minnesota,
and 18.99 g/day in North Dakota (Table 3-13). The highest estimated consumption was in
Florida. Given that the Florida consumption is likely to be underestimated, one might conclude
that Florida residents consume as much or more fish and shellfish per capita than residents of the
other three states. At the same time possible differences in how well respondents can estimate
consumption over different recall periods and other differences between the surveys make any
comparison more speculative.
The average per capita consumption offish and shellfish for the U.S. population, based
on the CSFII survey, is 12.83 g/day (U.S. EPA, 2002). This consumption rate is somewhat lower
than the per capita values for the general population in Florida, Connecticut, Minnesota, and
North Dakota shown above. Although fish and shellfish consumption data from National Health
and Nutrition Examination Survey (NHANES) 2003-2006 are reported in the Exposure Factors
Handbook: 2011 Edition (U.S. EPA, 2011), they are presented in g/kg-day and therefore not
directly comparable. The average per capita fish and shellfish consumption from NHANES
2003-2006 for all ages combined was estimated to be 0.22 g/kg-day. The average per capita fish
and shellfish intake from Florida, Connecticut, Minnesota, and North Dakota in g/kg-day can be
estimated from Table 3-8 by calculating a weighted average from all age groups and genders
combined. These are calculated to be 0.47 g/kg-day, 0.42 g/kg-day, 0.29 g/kg-day, and 0.33
g/kg-day for Florida, Connecticut, Minnesota, and North Dakota, respectively. These are higher
than the value obtained from NHANES 2003-2006. The National Health and Nutrition
Examination Survey (NHANES; CDC, 2002) survey has also collected fish consumption data.
United States regional estimates based on NHANES 1999-2002 have been reported (Mahaffey et
al., 2005). The average per capita 24-hour consumption estimates reported are: 20.8 g/day for
the Atlantic Coast region (which includes most of Connecticut and almost half of Florida; 13.5
g/day for the Gulf Coast region (which includes the rest of Florida) and 11.5 g/day for the
Midwest (which includes Minnesota and North Dakota).
Differences between states may also be associated with seasonal or year-to-year changes
in fish consumption. The Minnesota and North Dakota data were collected in September and
3-25
-------
October. The data from the other surveys were collected over a year-long period. Consumption
of self-caught fish is higher in summer months and lower in the winter. Although respondents in
Minnesota and North Dakota were asked about fish and shellfish consumption in the last year,
the timing of the survey may have affected the responses. There may also have been changes in
fish and shellfish consumption across years that would affect the comparisons, since the three
surveys were completed in different years. The effect of changes over time has not been
assessed.
3.8. LIMITATIONS OF THE ANALYSIS
The analysis reported here consists primary of tabulated means by subgroups within
states. This represents a first attempt to find factors that might identify populations that consume
more fish and shellfish than the mean for the general population within each state and thus may
be at greater risk from contaminants in fish and shellfish. More complicated statistical models
may be used in the future.
Although the data from the three surveys discussed in this report provide a useful source
of information on fish consumption, any analysis results are limited by the fact that the original
data files for Connecticut and Florida had some missing and inconsistent values such that EPA
could not replicate numerous statistics derived from the original study reports. Conclusions that
might apply to other states outside of the four states covered by the surveys are limited due to
differences among the surveys and to unknown differences between the four states and other
regions of the United States.
3-26
-------
4. REFERENCES
Anderson, S.A. (1988) Guidelines for use of dietary intake data. J. Amer. Diet. Assoc. 88(10): 1258-1259.
Balcom, N.C., Capacchione, C.M., Hirsch, D.W. (1999) Quantification offish and seafood consumption rates for
Connecticut, submitted to the CT Dept. of Environmental Protection, Contract No. CWF-332-R. CT Sea
Grant Publication No. CT-SG-99-02. 187pp.
Barnett, V., Lewis, T. (1996) Outliers in statistical data. 3rd Ed. reprinted with corrections. Chichester, NY: John
Wiley and Sons.
Benson, S. A., Crocker, C.R., Erjavec, J., et al. (2001) Fish consumption survey: Minnesota and North Dakota,
prepared by the Energy and Research Center, University of North Dakota, Grand Forks, ND.
Blake, A.J., Guthrie, H.A., Smiciklas-Wright, H. (1989) Accuracy of food portion estimation by overweight and
normal-weight subjects. J. Amer. Diet. Assoc. 89(7):962-966.
Block, G., Hartman, C.M., Dresserr, C.M., et al. (1986) A data-based approach to diet questionnaire design and
testing. Amer. J. Epidemiol. 124(3):453-469.
Block, G. (1982) A review of validations of dietary assessment methods. Amer. J. Epidemiol. 115(6):492-504.
Centers for Disease Control (CDC). (2002) National Health and Nutrition Examination Survey (NHANES)
National Center of Health Statistics (NCHS). NHANES' web site
http://www.cdc.gov/nchs/nhanes/nhanes2001-2002/nhanesO l_02.htm.
Centers for Disease Control and Prevention (CDC). (2006) National Center for Health Statistics (NCHS). National
Health and Nutrition Examination Survey Data 2005-2006. US Department of Health and Human Services,
Centers for Disease Controland Prevention: Hyattsville, MD 2006, Available at
http://www.cdc.gov/nchs/nhanes/nhanes2005-2006/nhanes05_06.htm.
Degner, R.L., Adams, C.M., Moss, S.D., et al. (1994) Per capita fish and shellfish consumption in Florida,
University of Florida, Gainsville, FL.
Dillman, D.A. (1978) Mail and telephone surveys: The total design method. New York, NY: John Wiley and Sons,
Inc. 325 pp.
Dillman, D.A. (2000) Mail and internet surveys: The tailored design method. 2nd Ed. New York, NY: John Wiley
and Sons, Inc.
Dubois, S., Boivin, J.F. (1990) Accuracy of telephone dietary recalls in elderly subjects. J. Amer. Diet. Assoc.
90(12): 1680-1687.
Hyndman, R.J., Fan, Y. (1996) Sample quantiles in statistical packages, Am. Statistician 50(4): 361-365.
Lohr, S.L. (1999) Sampling: design and analysis. Pacific Grove, CA: Duxbury Press.
Mahaffey, K.R., Clickner, R.P., Phillips, L.J., Jeffries, R.A. (2005) Patterns offish and shellfish consumption in the
U.S. population. Presented to the International Society of Exposure Analysis. November 1, 2005.
Medlin, C., Skinner, J.D. (1988) Individual dietary intake methodology: a 50-year review of progress. J. Am. Diet
Assoc. 88(10): 1250-1257.
Peters, E., Houseknect, C. (1992) A review and analysis of fish/shellfish consumption survey methods and
approaches. Tetra Tech. Inc., EPA Contract No. 68-C9-0013.
4-1
-------
Portier, K.M., Urn, Y., Degner, R.L., Mack, S.K., Adams, C.M. (1995) Statistical analysis of Florida per capita fish
and shellfish consumption data, Florida Department of Environmental Protection in fulfillment of contract
SP332. Florida Agricultural Research Market Center Food and Resource Economic Department, Institute
of Food and Agricultural Sciences, Univeristy of Florida, Gainsville, Fl. Available at
http://www.fredifas.ufl.edu/agmarketing/pubs/1990s/Fish%20&%20Shellfish%20statistical%20analysis.pd
f
Rogers, J.W. (2003) Estimating the variance of percentiles using replicate weights. ASA Proceedings of the Joint
Statistical Meetings, 3525-3532. Alexandria, VA: American Statistical Association.
Stern, A.H., Korn, L.R., Ruppel, B.E. (1996) Estimation offish consumption and methylmercury intake in the New
Jersey population. J. Expo Anal. Environ. Epidemiol. 6:503-525.
Tran, N.L., Barraj, L., Smith K., Javier, A., Burke, T.A. (2004) Combining food frequency and survey data to
quantify long-term dietary exposure: a methyl mercury case study. Risk Anal. 24(1): 19-30.
USDA Agricultural Research Service. (1987) Composition of foods: finfish and shellfish products: Raw, processed,
prepared. Agriculture Handbook: 8-15.
USDA. (1998) Continuing survey of food intakes by individuals: 1994-96, 1998. U.S. Department of Agriculture,
Agricultural Research Service.
U.S. EPA. (1992) Consumption surveys for fish and shellfish: A review and analysis of survey methods. Office of
Water, Washington, D.C. EPA/822/R-92/001.
U.S. EPA. (1997) Mercury study report to Congress. U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards, Office of Research and Development, Washington, D.C. EPA-452/R-97-003.
U.S. EPA. (1998) Guidance for conducting fish and wildlife consumption surveys. U.S. Environmental Protection
Agency, Office of Water, Washington, D.C. EPA-823-B-98-007.
U.S. EPA. (2002) Estimated per capita fish consumption in the United States. Chapter 4.1.1.1 Table 3.
(http://water.epa.gov/scitech/swguidance/standards/criteria/health/upload/consumption_report.pdf).
U.S. EPA. (2008) Child-Specific Exposure Factors Handbook (Final Report) U.S. Environmental Protection
Agency, Washington, DC, EPA/600/R-06/096F, 2008. Available at
http://cfpub.epa. gov/ncea/risk/recordisplay.cfm?deid=199243
U.S. EPA. (2011) Exposure Factors Handbook 2011 Edition (Final). U.S. Environmental Protection Agency,
Washington, DC, EPA/600/R-09/052F, 2011. Available at
http://cfpub.epa. gov/ncea/risk/recordisplay.cfm?deid=236252.
West, P.C., Fly, J.M., Marans, R., Larkin, F., Rosenblatt, D. (1993) 1991-1992 Michigan sport anglers fish
consumption study. Ann Arbor, MI: Michigan Department of Natural Resources
4-2
-------
APPENDIX A. DATABASE DEVELOPMENT
A.l. INTRODUCTION
The following sections described the processing of the data from the three studies to
convert them to a common format in a combined file. The first step was to process each study's
data to create SAS files with the same file organization across studies. Sections A.2, A.3, and
A.4 describe the processing of the data from individual studies. Section A.5 describes how the
files were combined to create one file for further analysis. Section A.6 describes how
approximate sampling weights were constructed.
File names are shown in capital letters (such as HOMERH), generally without the
extension. In the text, variable names are shown in Courier font (such as CNTYNAME).
Abbreviations used throughout the appendix are shown in Table A-l. The general
approach was to reorganize the files to create one file with household information, a second file
with information about individuals within households (excluding the household information
common to all individuals in the same household), and a third file with fish consumption
information for each individual (excluding information in the other files). For analysis, these
files are combined to create a file for analysis. From the combined files, fish consumption
quantities are totaled across species to calculate a total for each individual. These totals are then
merged with the individual and household data to create a file with one record per individual for
analysis. When the files are merged, individuals with no fish consumption in the fish
consumption file are given a value of 0 for total fish consumption.
Table A-l. Abbreviations used throughout the data documentation report
FL
CT
MN
ND
HH
FFQ
RSA
PMP
MNND
Florida
Connecticut
Minnesota
North Dakota
Household
Food frequency questionnaire
Randomly selected adult (in the Florida survey)
Primary meal preparer (in the Florida survey)
Minnesota/North Dakota study
A-l
-------
A.2. FLORIDA DATA
A.2.1. Description of the Survey and the Original Data Files
A.2.1.1. Survey methodology
The Florida data were collected by telephone from a random sample of households
stratified by county. The telephone survey approach was chosen because of its relatively low
cost and the ease of probability sampling (Peters and Houseknecht, 1992). The telephone
surveys targeted two populations, residents statewide and residents of five counties where
industrial pollution from bleached kraft paper mills could result in chemically contaminated fish
and shellfish. Households from all counties in Florida were sampled at the same sampling rate
for the "State" sample. An additional supplemental sample of 740 households was selected in
the five counties that had paper mills (Escambia, Gulf, Nassau, Putnam, and Taylor). In total,
1,000 households were interviewed in the five paper mill counties (260 from the State sample
plus 740 from the supplemental sample). In the study report, the "Paper Mill" sample refers to
the 1,000 households from counties with paper mills. Data collection took place from March 15,
1993 to March 13, 1994.
In each household, one randomly selected adult (RSA) was selected. The RSA was asked
about his/her away-from-home fish and shellfish consumption (referred to here as just fish
consumption). Then the household primary meal preparer (PMP), if available, was asked about
at-home fish consumption by all family members (referred to as householders in the study
report). If the PMP was not available or the RSA was the only household member, the RSA was
asked about their at-home fish consumption. According to the published questionnaire, if the
PMP was not available and the RSA was not the only household member, fish consumption for
non-RSA household members was not collected. Therefore, the data may underestimate the in-
home fish consumption for non-RSA household members and the data do not have away-from-
home fish consumption for non-RSA household members. The analysis weights adjust for the
lack of information for non-RSA household members in households where the PMP was not
available. However the weights do not adjust for missing away-from-home fish and shellfish
consumption. According to the survey report, the primary meal preparer was available in 6,617
of the 8,000 households in the general population.
A.2.1.2. Survey data files
The Florida data were received in seven files. The data in the files overlap. The files and
their contents are:
A-2
-------
HOMERH has at-home fish consumption for the RSA for the State sample of 8,000
respondents;
AWAYRH has away-from-home fish consumption for the RSA for the State sample of
8,000 respondents;
HOMERC has at-home fish consumption for the RSA for the Paper Mill sample, with
1,000 respondents. The Paper Mill sample includes respondents from the State sample
that lived in the counties with paper mills;
AWAYRC has away-from-home fish consumption for the RSA for the Paper Mill
sample, with 1,000 respondents;
STHHOLD has at-home fish consumption for all household members in the State sample.
This file is essentially the concatenation of HOMERH and the at-home fish consumption
data for the non-RSA household members;
COHHOLD has at-home fish consumption for all household members for the Paper Mill
sample. This file is essentially the concatenation of HOMERC and at-home fish
consumption data from the non-RSA household members in the paper mill counties; and
Finally the TOTALRH file has the total fish consumption by species for each RSA in the
State sample. This total is essentially the sum of the values in AWAYRH and
HOMERH.
The study reports show 15,672 individuals in the State sample. However, two RSAs in
the HOMERH file were not in the STHHOLD file. Thus, after combining the data from all of
the files, the processed data files have 15,674 individuals in the State sample. The study report
also shows 2,099 household members in the Paper Mill sample; the processed files have 2,100
household members in the Paper Mill sample.
A.2.2. Processing of the Data Files
A.2.2.1. Demographic data
Each of the files mentioned in Section A.2.1.2 include demographic data. However, for
some individuals, the demographic data differed among files or was present in some files and not
in others. Also, most of the bodyweight values for the RSA's were missing. In response to
questions, Professor Ken Portier from the University of Florida, who was one of the authors of
the study report and published some additional analysis of the data, provided an additional file
(WTOT96.TXT) with the RSA bodyweights for the State sample. However, many of the
bodyweight values for RSAs in the Paper Mill supplemental sample were still missing.
A-3
-------
Various data files were used to attempt to replicate some of the numbers in the study
report. Use of the TOTALRH file provided the closest approximation to the values of total
consumption and recreationally caught consumption by species for the RSA State sample. The
demographic data in the WTOT96.TXT file agreed with the corresponding data in the
TOTALRH file (when the data in TOTALRH was not missing). In addition, the WTOT96.TXT
demographic variables were also apparently used in the analysis performed by Professor Portier
(Portier et al., 1995). As a result, the values in the WTOT96.TXT file were assumed to be the
correct demographic variables for the State RSA sample.
Cases where demographic data differed between files were printed and reviewed.
Differences were resolved as follows:
Values in WTOT96.TXT were used, if present;
Otherwise, the demographic variables were selected by taking the first non-missing value
found when going through the files in the order: HOMERH, AWAYRH, HOMWRC,
AWAYRC, STHHOLD, COHHOLD, and TOTALRH.
This was a somewhat arbitrary approach. However, no information was found to suggest
that one value was more likely than another to be correct. Records were flagged to indicate if the
demographic variables were not consistent across all files (see the variable FAMPROB).
The data from these seven files were reorganized into three files:
• FLFAMDAT with data for households (e.g., county and household income);
• FLINDDAT with data for individuals within households (e.g., sex, age, bodyweight); and
• FLFISHDAT with data on fish consumption.
In the questionnaire, sex and age were obtained for all household members from the
initial questions (used to identify the RSA). Bodyweight was requested from the RSA and, for
other household members, from the PMP. Bodyweight from non-RSA household members was
apparently not obtained if the RSA was not the PMP and the PMP was not available. These
variables were in the FLINDDAT file.
In the questionnaire, race-ethnicity was obtained from two questions. The first asked for
race (White, Black, American Indian, and Asian). The second question was about Hispanic or
Spanish origin or descent. The results of these two questions were combined to create a race-
A-4
-------
ethnicity variable (White non-Hispanic, Black non-Hispanic, Hispanic, American Indian, and
Asian). In the data files, the race-ethnicity variable is named Race. To minimize confusion, the
race-ethnicity variable was renamed to RSARaceEthn in the processed data files. The
responses to the first race question were in a variable named RSARaceResp.
Education (as years of schooling), race-ethnicity, and household income were obtained
for the PMP or, if not available, for the RSA. Thus, there was some inconsistency as to who
provided this information. It is reasonable to assume the household income was essentially the
same regardless of who answers the question. However, the education and race-ethnicity was
asked for the respondent and may be different between the RSA and PMP. The data files
provide no indication of whether the race and education refers to the RSA or the PMP. In the
processed data files, these values were associated with the household. This is consistent with
how the results were presented in the study report. Since these variables were asked once for the
household, they are considered household variables and were in the FLFAMDAT file.
In the study report, the summary tables used race-ethnicity rather than reported race
(RSARaceResp). Therefore, although the RSARaceResp variable is retained in the final files,
this variable was not used. In addition, there are a few inconsistencies between the
RSARaceEthn and RSARaceResp variables. No attempt was made to resolve these
inconsistencies.
In the original files, household variables appeared on other records for some children. In
a few cases the values on the non-RSA records differed from those on the corresponding RSA
records. In these cases the values for the RSA record was used.
Finally, if the sex was coded as unknown (5) it was set to female (2). This was done to
(1) obtain values that were consistent with the values in the WTOT96.TXT file and conform to
procedures in a program provided by Professor Portier, and (2) to get the distribution of
demographic data to agree, or agree much more closely, with the values in the published report.
This change affected non-RSA household members. There were no non-RSA records coded as
female in the data files. The number of non-RSA records receded from code 5 (unknown) to
code 2 (female) was approximately equal to the number of non-RSA females, consistent with the
assumption that these records are for females. After receding the sex variables, there were still
812 missing values for sex (coded as -8 or -9 in the original files).
Tables A-2 through A-4 show the number of household records, individuals records, and
fish consumption records for which the demographic variables differed among files and the
percent of missing values in the processed files. In general, individual demographic information
was more likely to be missing for the non-RSA household members. However, bodyweight was
missing on 735 of 740 RSA records in the Paper Mill supplemental sample.
A-5
-------
Table A-2. RSA/Household records with inconsistent and missing values (out
of 8,740 records)
Records with inconsistencies
Variable
County
Income
Race-Ethnicity
Education
RaceResp
Any HH demographic variable
Number of records
13
39
200
8
2
227
Percent
0.15
0.45
2.29
0.09
0.02
2.60
Missing
Number of
records
0
1,532
175
99
170
values
Percent
0.00
17.53
2.00
1.13
1.95
Table A-3. Individual records with inconsistent and missing values (out of
17,213 records)
Records with inconsistencies
Variable
Weight
Age
Sex
Any individual
demographic variable
Number of records
16
30
16
43
Percent
0.09
0.17
0.09
0.25
Missing values
Number of
records
1,814
827
812
Percent
10.54
4.80
4.72
Table A-4. Fish consumption records with inconsistent and missing values
(out of 16,099 records)
Records with inconsistencies
Variable
Cooking Method
Skinned before cooking
Number of Occasions
Uncooked Grams
Percent recreationally caught
Inconsistencies in any variable
Number of records
1
0
1
1
16
17
Percent
0.01
0.00
0.01
0.01
0.10
0.11
Missing values
Number of
records
6,868
12,952
5,422
0
4,338
Percent
42.7
80.5
33.7
0.0
26.9
A-6
-------
A.2.2.2. Fish consumption data
The data files received from Florida had one record per individual with separate variables
for each of 62 species offish or shellfish. The fish consumption data were processed to create a
file with one record per combination of individual, species, and location (at-home or away-from-
home). The fish consumption file had 16,099 records after eliminating records for which no fish
consumption was reported. As with the demographic variables, the fish consumption variables
for some individuals were replicated in several files and the values in those files differed for a
few records; the inconsistencies were resolved by arbitrarily selecting the first non-missing value
for the individual when searching through the files in the order: HOMERH, AWAYRH,
HOMWRC, AWAYRC, STHHOLD, COHHOLD, and TOTALRH. In some cases the values
were missing because they were not appropriate (e.g., the survey files did not record if canned
tuna was skinned before cooking). In a few cases a variable in one file was not in another. For
example, the cooking method for king mackerel was not in the STHHOLD and COHHOLD files,
but was in other files. As a result, the cooking method for king mackerel is missing for non-RSA
household members. Variables that were in some files and missing in other files generally
recorded if the fish was skinned before cooking. These missing variables were treated as if they
had been present, but had contained no information.
A.2.2.3. Data file comparisons
Table A-5 illustrates how well the values in the study report can be replicated using the
processed data files. This table shows the annual seafood consumption for the State and Paper
Mill samples. The top portion of the table shows the consumption as calculated from the
processed files. The lower half of the table shows the corresponding values from the study
report. The totals generally agree within a few percent. However, the totals for shellfish are
higher using the processed file. Since the subtotal for all fish or shellfish is about the same from
both sources, the classification of species into finfish or shellfish may be different. The
classification of species into finfish and shellfish was based on the tables in Appendix B of the
study report. There were many missing values for demographic variables in the data file that
were apparently not missing when the original report was compiled. When broken down by
demographic variables, most annual fish consumption estimates from the processed files were
within 5% of the corresponding values in the study report. However, there were some notable
exceptions. The estimates derived from the processed file for fish and shellfish consumption for
Hispanics, American Indians, and Asians are much lower in the Paper Mill sample than reported
in the study report.
A-7
-------
Table A-5. Annual seafood consumption estimates for the two samples
(kg/yr)
State Sample
RSAs
Householders
Paper Mill Sample
RSAs
Householders
Values from Processed Files
Number
Away
Away
Away
Home
Home
Home
All
All
All
Finfish
Shellfish
Subtotal
Finfish
Shellfish
Subtotal
Finfish
Shellfish
Subtotal
8,000
6.35
2.49
8.85
6.73
1.22
7.95
13.09
3.71
16.80
15,674
6.88
1.21
8.09
1,000
6.73
2.98
9.71
7.86
1.62
9.48
14.58
4.60
19.19
2,099
7.87
1.83
9.70
Values from Study Report
Number
Away
Away
Away
Home
Home
Home
All
All
All
Finfish
Shellfish
Subtotal
Finfish
Shellfish
Subtotal
Finfish
Shellfish
Subtotal
8,000
6.38
2.48
8.85
6.82
1.13
7.95
13.20
3.60
16.80
15,672
6.98
1.12
8.10
1,000
6.60
2.97
9.57
7.93
1.54
9.47
14.53
4.51
19.04
2,099
8.14
1.55
9.69
The appendix to the data documentation report has tables of demographic variable
frequencies by populations that can be compared to the values in the appendix of the study
report. For many variables the distribution is very close (often within several respondents).
However there are some notable differences. For the State RSA population:
• The processed files appeared to have classified 12 missing codes as female. These 12
individuals were classified as female in the WTOT96.TXT file, but not in other files.
A-8
-------
Rather than showing a missing category, the processed files had more respondents in the
65+ category.
For the Paper Mill RSA population:
The number of American Indian and Asian respondents differed somewhat from the
report. This may be because many of the inconsistencies in race-ethnicity were either
American Indian or Asian classifications.
The bodyweight distributions were very different because there were only 44 weight
measurements for the Paper Mills RSAs in the files.
In general, the differences between values from the processed files and values in
Appendix A of the study report are greater for all household members (Householders) than for
just the RSAs. For the State householder population:
• The processed files had many more values for bodyweight than were reported in the
study report.
• Differences in the number of respondents by income category were greater than for other
variables.
For the Paper Mill householder population:
• The male and female rows appeared to be reversed in the report.
• The processed files had more bodyweights than were shown in the study report.
The total consumption by species for the State RSAs is shown in the Table B-l of the
study report. The numbers in this table can be calculated from the TOTALRH file. However,
there were some differences between the totals from the processed files and the TOTALRH file.
The processed file was derived from files other than TOTALRH and had separate information on
at-home and away-from-home fish consumption. The total fish consumption (away-from-home
+ at-home) for each individual from the processed files differed from the TOTALRH values in
the following ways:
• There were 13 fish consumption records that were in the processed files, but not in the
TOTALRH file.
• There were five fish consumption amounts that were in both files, but were attributed to
different respondents in different files. In all five cases, the IDs of the different
A-9
-------
respondents differ only in the first digit (either a ' 1' or '9'). This pattern might be due to
data entry or editing errors.
There were five fish consumption records for which the total consumption differed.
A.2.2.4. Calculation of uncooked and as-consumed fish consumption
The Florida survey questionnaire asked for the quantity offish eaten and the cooking
method used. The reported as-consumed quantity was converted to uncooked weight (in g/week)
in the original data files. The conversions used were documented in the study report and
depended on type of food (several types of shellfish, other shellfish, or any fmfish) and the
cooking method. Assumptions were made when the cooking method was not provided. The
conversions that were reported in the study report were used to convert the uncooked quantities
to as-consumed quantities. Those conversions are shown in Table A-6.
Table A-6. Factors for converting the Florida data from uncooked to as-
consumed weights
Fish Category
Cooking
Method
Fried
Broiled
Steamed
Boiled or
poached
Raw
Microwave
baked
Conventional
oven
Grilled or
smoked
Other
Salad
shrimp Clams
0.75 0.57
0.78 0.6
0.79 0.62
0.8 0.63
1 1
0.8 0.66
0.85 0.64
0.78 0.6
0.77 0.61
Blue
crab
0.75
0.78
0.79
0.8
1
0.8
0.85
0.78
0.77
Crab
meat
0.57
0.6
0.62
0.63
1
0.66
0.64
0.6
0.61
Imitation
crab meat
0.75
0.78
0.78
0.78
1
0.87
0.79
0.78
0.82
Conch
0.57
0.6
0.62
0.63
1
0.66
0.64
0.6
0.61
Crayfish
0.75
0.78
0.79
0.8
1
0.8
0.85
0.78
0.77
Other
misc
shell
fish
0.67
0.7
0.72
0.72
1
0.75
0.75
0.7
0.71
Finfish
0.75
0.78
0.78
0.78
1
0.87
0.79
0.78
0.82
Other includes: Other cooking methods, Don't Know, and missing data values.
A-10
-------
The FLFISHDAT data file had two variables, GDayRaw and GDayCooked, with the
fish and shellfish consumption in g/day as uncooked weight or cooked (as-consumed) weight,
respectively. GDayRaw was calculated by dividing the quantity in the survey data files by seven
to convert from weekly to daily consumption. GDayCooked is GDayRaw multiplied by the
appropriate conversion factor from Table A-6. The conversion factors used were in the
FLFISHDAT data file in the Conversion variable.
A.3. CONNECTICUT DATA
A.3.1. Description of the Survey and the Original Data Files
A.3.1.1. Survey methodology
The Connecticut survey data were collected from July 1996 through May 1997. Data
were collected from several different populations using a combination of a mail questionnaire
and personal interviews (for the Southeast Asians, food stamp recipients, and some anglers).
The survey asked for the rate offish consumption for all family members and any fish or
shellfish dish. The fish consumption was reported as as-consumed (generally cooked) fish
consumption (in g/day). The survey results were presented in the Connecticut study report
(Balcom et al., 1999).
A.3.1.2. Data files
The data and documentation for the Connecticut data included:
• An IBM format tape with SAS data sets in SAS transport format;
• A notebook with some programs that were used for processing the files; and
• A box of SAS output.
According to the file contents and the programs in the notebook, (1) the programs in the
notebook comprised a series of programs used to process the data, (2) the data files from the last
program in the series (GDAY and INDIV) were not on the tape, and (3) corrections were made
to the data in essentially all of the data processing steps. Based on the output in the box, (1)
there were apparently other programs which were not available, and (2) the files used to create
the output that went into the study report were derived from files other than INDIV and GDAY,
and (3) at least one additional variable (Complete) that was used to create the reports was not
in the data files (it indicated which records to keep when summarizing the data). The algorithm
to create Complete was not been found in any of the programs. Table A-7 shows the variables
with differences between the INDIV and GDAY files for the same household.
A-ll
-------
Table A-7. Variables with differences between the INDIV and GDAY files
for the same household (out of 827 households)
Variable
EATING
PEOPLE
Multiple variables
COUNTY
WHYNOT
APPLYFIS
Question
Do you know fish is good for you?
Number of people in the household
Multiple questions
County
Why do you not follow the advisories?
Do the advisories apply to the fish you eat?
Number of differences
1
9
10
2
2
1
The printed program that created INDIV and GDAY was input and run using the data
files from the tape. The INDIV and GDAY files were then processed to eliminate
inconsistencies and, to the extent possible, to create a file from which the reported values could
be derived. The INDIV file contains fish information on fish dish consumption, with one record
per combination of dish with individual. The GDAY file contains total fish consumption by
person. Each file also contains additional information, such as demographic information and
responses to questions about fish advisories.
The variables in the data files were divided into three files:
• CTFAMDAT with household demographic data (such as household income);
• CTINDDAT with demographic data for individuals within households (such as
bodyweight); and
• CTFISHDAT with fish consumption data for individuals by species.
In these files, individuals were identified by a combination of the family ID (FamID) and
the individual ID (IndID).
A-12
-------
A.3.2. Processing the Data Files
A.3.2.1. Demographic data
Both the INDIV and GDAY files have information on 827 households. However, the
demographic information differed between files for some households. Cases where the family
data differed between the files were printed and reviewed. Differences were resolved as follows:
If information was missing on one record and present on another, the non-missing data
was assumed to be correct.
If the number of people in the family (as indicated by the PEOPLE variable) was
inconsistent between files, the number of people in the family was set to equal the
number of household members in the data files (in all cases where there were inconsistent
values, at least one file had a value for PEOPLE that equaled the number of household
members in the data files). Note that there might have been more people in the family
than were reported because some respondents apparently lost interest in filling out the
questionnaire for all family members.
Otherwise, for three individuals, the data from the first record in the INDIV was
arbitrarily selected as the correct record, assuming this record corresponded to the person
filling out the questionnaire. This arbitrary choice affected the PEOPLE and COUNTY
variables for two families and affected most variables for the third family.
For the family data, 25 records with inconsistent data were found.
The LIMI TED variable indicated limited income families. As a result of changing the
PEOPLE variable, there were two cases where the LIMITED variable (derived from PEOPLE
and INCOME) was inconsistent with the PEOPLE variable. As a result, the LIMITED variable
was recalculated using the algorithm found in the notebook.
The files had information on 2,133 individuals. There were discrepancies in the
individual data for 12 individuals (four from the same family). Six records had differences in
one variable only. In these cases the non-missing value was used. Six records had differences in
multiple variables, including AGE, BFEED, fishing frequency, and the TRIM variable (Does the
person trim skin and fatty meat from the fish they catch?). In all cases, the values were different
on only one of several records in the INDIV file. The most common value in the INDIV file was
used.
The following additional changes were made to the individual data:
• For most records, respondents who never fish have FISHYR set to 0. Therefore, some
records with missing FISHYR were set to 0.
A-13
-------
Respondent birth date and age were inconsistent for respondents born before 1919
(possibly due to how SAS handles two digit years). The respondent age as of 9/1/1997
was recalculated from their birthday. There were 59 inconsistencies between SEX, AGE,
and AGECAT, so AGECAT was recalculated also based on SEX and AGE.
A.3.2.2. Fish consumption data
The fish consumption data had many variables that could be combined and were unlikely
to be used for the EPA analysis. The primary revisions to the fish consumption data were to
combine multiple variables into one. The following changes were made:
Convert months in which fish were eaten (variables: JAN FEB MAR APR MAY JUN
JUL AUG SEP OCT NOV DEC) to a text string. A ' J' in the first position of the text
string indicates that fish were eaten in January. An 'F' in the second position of the text
string indicates that fish were eaten in February, etc. Each month was indicated by one
character. The order of the months and the letters used to indicate fish consumption are:
J January
F February
M March
A April
M Ma
J June
J July
A August
S September
O October
N November
D December
A value of" MAM S " indicates that the fish was eaten in March, April,
May, and September. Spaces indicate that the fish was not eaten in the
corresponding months.
Convert parts eaten (variables: PART A PARTB PARTC PARTD PARTE PARTF
PARTG PARTH) to a text string. As above the parts that were eaten are noted by the
position in the text string and by the letters used. The character codes, in order from left
to right are:
A-14
-------
A Entire fish
M Meat
S Skin/Fin
H Head
B Bones
E Eyes
R Roe/eggs
T Lobster Tomally
M Crab Muster
O Organs
F Fat/oil
Convert cook methods (variables: COOKA COOKB COOKC COOKD COOKE COOKF
COOKG) to a text string. As above the cooking methods used are noted by the position
in the text string and by the letters used. In this case, each cooking method used two
character positions. The character codes, in order from left to right are:
Ba Baked
Br Broiled
Po Poached/boiled/steamed
SF Sauteed/stir fried
DF Deep fried
Gr Grilled
Sm Smoked
Mi Micro waved
SS Soup stock
SI Salad
SC Soup/chowder/stew/casserole/bisque
Ra Raw
Su Sauce
Convert parts cooked (variables: PICA PTCB PTCC PTCD PTCE PTCF PTCG PTCH)
to a text string. As above the parts cooked are noted by the position in the text string and
by the letters used. The character codes, in order from left to right are:
A All
M Meat
S Skin/Fin
A-15
-------
H Head
B Bones
E Eyes
R Roe/eggs
T Lobster Tomally
M Crab Muster
O Organs
F Fat/oil
In the data files, the dishes were identified by a numeric code. A file with the codes and
corresponding text descriptions was created and merged the text descriptions into the file.
The fish data file has one record for each dish eaten by each individual. One record for a
respondent who reported not eating fish or seafood was removed from the file So that there are
would be no records in the file for respondents that do not eat fish.
A.3.2.3. Removing incomplete cases
The reported analysis of the Connecticut data excluded records for surveys that were
incomplete. In the programs received from Connecticut, the incomplete surveys were identified
by records with COMPLETE = 3 (based on discussions with Nancy Balcom at the University of
Connecticut and review of the available programs). However, the COMPLETE variable used by
Connecticut to process the files was not on the INDIV or GDAY files. There was also no
information on how the COMPLETE variable was calculated. Data that were likely to be from
incomplete surveys were identified and removed as follows. Based on the box of computer
output, there were 53 records for which COMPLETE is 3 and for which values for numerous
variables were all missing (LOBSTER, LOBROLL, CRAB, CRABCAKE, CLAMS,
CHOWDER, STRIPS, OYSTERS, OYSTSTEW, MUSSELS, BLUEFISH, STRIPER,
EEL, PORGY, BLACKFSH, TUNACAN, TUNASTK, FSHSTICK, FAKECRAB).
Therefore, the records where these variables were all missing were removed. No other
algorithms for identifying the cases to be removed for analysis were identified.
After processing and removing incomplete survey data, the resulting files have 810
families, 2,080 individuals, and 15,367 fish items eaten. For 18 of the fish items eaten, either the
meals per year or grams per meal are missing, so the grams per year consumption cannot be
calculated. For analysis, the fish consumption for these 18 individuals was set to zero. Finally,
internal consistency of variables was not checked for many variables that were not used for the
analysis.
A-16
-------
Two variables, EATYR and EATWK, were unique within individual and did not differ by
fish item. These variables were thus inconsistent with the documentation. Another variable,
ANGTYPE (Angler type) was a character variable that had been truncated in INDIV and GDAY
so as to be useless. All these variables had been left off the individual file.
A.3.2.4. Data file comparison
The following tables illustrate how well the values in the study report can be replicated
using the data files.
Table A-8 compares the reported number of respondents in each population to the
number calculated from the processed files. Population sizes from the processed files are within
4% of the reported values. The response frequencies for demographic variables are generally
within 5%.
Table A-8. Sample sizes by population from the processed files and Table 5
of the study report
Number of Households
Population
General
Angler
Commercial
Minority
Limited Income
Women
Children
Southeast Asian
Non- South East Asian
Processed data
206
341
74
245
274
406
303
89
156
Report Table 5
207
341
73
245
276
420
305
89
156
Number of Individuals
Processed data
433
524
178
866
935
480
554
334
532
Report Table 5
434
504
178
860
937
493
559
329
531
The study report showed the population mean and standard deviation of total fish
consumption (total across all fish dishes, in g/day). The values calculated from the processed
files are within 5% of the reported values. Table A-9 compares the reported mean, standard
deviation, and maximum total fish consumption (in g/day) from the study report with the values
calculated from the processed files.
A-17
-------
Table A-9. Total fish consumption (g/day) by population from the processed
files and Table 11 of the Study report
Calculated from the processed files
Population
General
Angler
Commercial
Minority
Southeast Asian
Non Southeast Asian
Limited Income
Women
Children
N
433
524
178
866
334
532
935
480
554
Mean
27.7
49.5
46.9
49.7
57.9
44.6
43.0
47.2
18.4
Std Dev
42.6
64.2
57.3
57.1
48.9
61.2
60.3
57.8
29.9
Maximum
493.6
570.5
490.2
429.0
245.0
429.0
570.5
493.6
324.1
From Table 11
N
437
502
178
861
329
532
937
497
559
Mean
27.7
51.1
47.4
50.3
59.2
44.8
43.1
46.5
18.3
of the Study report
Std Dev
42.7
66.1
58.5
57.5
49.3
61.5
60.4
57.4
29.8
Maximum
494.8
586.0
504.3
430.0
245.6
430.0
571.9
494.8
324.8
A.3.2.5. Calculation of uncooked and as-consumed fish consumption
The Connecticut data files provided the quantity offish as "cooked weight of edible
fish." However, for the few items that were clearly eaten uncooked (e.g., oysters, sushi), the
quantity appeared to be the uncooked weight. This weight is referred to as the as-consumed
weight. The Connecticut respondents assessed the quantity of a food item by selecting a shape
of the food item and a corresponding size. Connecticut collected cooking method information
for only a subset of the respondents. That information was apparently not used in the conversion
to as-consumed consumption. When converting the reported shape and size to the as-consumed
weights, some assumptions were made, such as that certain dishes were eaten uncooked or
cooked or that the reported shape corresponded to an uncooked or cooked food item.
The conversion factors used were provided in an appendix to the study report. If the
respondent reported the as-consumed quantity, then no conversion was needed. Table A-10 lists
the factors used to convert the Connecticut data from as-consumed to uncooked weights. If
provided, the conversion factors in the report were used. In a few cases judgment was used to
decide which conversion factor to use. Otherwise, the conversion factor from the EPAMercwry
Study Report to Congress (U.S. EPA, 1997) was used. If the conversion factor did not come
directly from the Connecticut study report, a comment column explained the source of the
conversion factor. In some cases the conversion factors were based on a USDA publication
(USDA, 1987). The conversion factor for dried shark fin was somewhat arbitrary because no
basis for a conversion factor was found. The CTFISHDAT file had variables with the uncooked
A-18
-------
and as-consumed fish consumption in g/day (GDayRaw and GDayCooked) as well as the
conversion factor for that fish dish (Conversion).
Table A-10. Factors for converting the Connecticut data from as-consumed
to uncooked weights
Species
caviar
combination platter
frozen sticks
gefilte fish
lox
sardines
seafood salad
sushi
sushimi
frozen filets
clams, stuffed
mackerel, canned
herring, canned
salmon croquettes
salmon, canned
fish cakes
shrimp roll/salad
fish chowder, canned
anchovies, canned
shrimp soup
Ig & jumbo stuffed
shrimp
undefined
largemouth bass
smallmouth bass
Code
2
4
6
7
9
10
11
13
14
16
17
19
20
21
22
23
24
25
26
27
28
29
101
102
Conversion (divide by this
to get uncooked weight)
0.75
0.75
0.78
0.75
0.75
0.75
0.82
1
1
0.78
0.5
0.75
0.75
0.705
0.705
0.75
0.79
0.75
0.75
0.79
0.79
0.75
0.78
0.78
Notes*
EPA
EPA
(assuming pollock)
EPA
EPA
EPA
(used by FL for unknown
cooking method for imitation
crab)
Assume uncooked
Assume uncooked
(assuming pollock)
EPA
EPA
EPA
EPA
EPA
EPA
A-19
-------
Table A-10. Factors for converting the Connecticut data from as-consumed
to uncooked weights (continued)
Species
striped bass
unspecified bass
bluefish
tautog
bonito
bream
buffalo fish
bullhead
carp
catfish, freshwater
catfish, saltwater
catfish, farm raised
catfish, unspecified
cod
crappie
mahi-mahi, dolphin fish
drum
eel
eel, unspecified
flounder
grouper
haddock
halibut
hake
herring
mackerel
lobster whole
blue mussels
octopus
oysters, raw/cooked,
unspecified
scallops, unspecified
scungilli/whelk
Code
103
104
105
106
108
109
110
111
112
113
114
115
116
118
119
121
123
124
126
127
128
129
130
131
132
135
205
206
207
208
210
211
Conversion (divide by this
to get uncooked weight)
0.78
0.78
0.78
0.78
0.78
0.75
0.75
0.78
0.75
0.788
0.788
0.788
0.788
0.788
0.75
0.78
0.75
0.677
0.677
0.721
0.78
0.722
0.717
0.78
0.75
0.659
0.75
0.75
0.75
0.75
0.769
0.485
Notes*
EPA
EPA
EPA
EPA
EPA
EPA
EPA
EPA
EPA
A-20
-------
Table A-10. Factors for converting the Connecticut data from as-consumed
to uncooked weights (continued)
Species
shrimp, unspecified
squid, cooked
shellfish, other
crab, imitation
crab, Alaskan king legs
crab, stone
crab, dungeness
crab, canned
rock crab
squid, dried
undefined
spot
whitesucker
sea urchin roe
blowfish
bass, calico
salmon, smoked
shad, smoked
fish chowder, bluefish
fish chowder, scup
fish chowder, tautog
milkfish
covina, yellow
cod, dried
eel, raw
salmon, raw
squid, raw
Code
212
213
214
215
216
217
218
219
220
222
223
224
225
226
228
230
231
232
233
234
235
237
238
239
241
242
243
Conversion (divide by this
to get uncooked weight)
0.79
0.822
0.75
0.75
0.78
0.78
0.78
0.75
0.78
0.215
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.75
0.78
0.75
0.78
0.75
0.75
0.187
1
1
1
Notes*
EPA
EPA
(based on conversion for crab
legs)
(based on conversion for crab
legs)
EPA
(based on conversion for crab
legs)
78.55 g water in 100 g raw
squid (USD A)
EPA
EPA
EPA
EPA
EPA
EPA
EPA
EPA
(based on bluefish, probably
higher as cooked in water)
EPA
(based on tautog, probably
higher as cooked in water)
EPA
EPA
81.28g water in 100 g raw
Pacific cod & 81.22 g water in
100 g raw Atlantic cod (USD A)
A-21
-------
Table A-10. Factors for converting the Connecticut data from as-consumed
to uncooked weights (continued)
Species
grunt
doctorfish
snow crab
sheepshead
sharkfin, dried
herring, dried
snapper, yellowtail
snapper, pink
undefined
clams, raw
clam chowder
clam sauce
clam strips
clams, raw/cooked
unspecified
clams, undefined
clams, quohogs cooked
clam chowder/sauce
unspecified
crab cakes
crab salad
crab cake/salad
crab, undefined
lobster roll
lobster salad
Conversion (divide by this
Code to get uncooked weight) Notes*
244
245
246
247
248
249
250
251
300
20101
20102
20103
20104
20105
20106
20107
20109
20201
20202
20203
20204
20501
20502
0.75
0.75
0.78
0.75
0.25
0.284
0.78
0.78
0.75
1
0.5
0.5
0.425
0.75
0.5
0.5
0.5
0.78
0.78
0.78
0.78
0.75
0.75
EPA
EPA
(based on conversion for crab
legs)
EPA
Bold assumption
71.52 g water in 100 g raw
Pacific herring & 72.05 g water
in 100 g raw Atlantic herring
(USDA)
(based on conversion for
snapper blues)
(based on conversion for
snapper blues)
EPA
(maybe higher as cooked in
water)
(maybe higher as cooked in
water)
(maybe higher as cooked in
water)
(based on conversion for crab
legs)
(based on conversion for crab
legs)
(based on conversion for crab
legs)
(based on conversion for crab
legs)
EPA
EPA
A-22
-------
Table A-10. Factors for converting the Connecticut data from as-consumed
to uncooked weights (continued)
Species
lobster tail
lobster unspecified
oysters, raw
oysters, cooked
oyster stew
oysters, rockefeller
sea scallops
bay scallops
popcorn shrimp
small shrimp
med shrimp
large shrimp
jumbo shrimp
Code
20503
20504
20801
20802
20803
20805
21001
21002
21201
21202
21203
21204
21205
Conversion (divide by this
to get uncooked weight) Notes*
0.75 EPA
0.75 EPA
1
0.5
0.5 (maybe higher as cooked in
water)
0.5
0.769
0.769
0.79
0.79
0.79
0.79
0.79
* EPA = EPA Mercury Study Report to Congress (U.S. EPA, 1997).
A.4. MINNESOTA AND NORTH DAKOTA DATA
A.4.1. Description of the Survey and the Original Data Files
The Minnesota and North Dakota data were distributed (1) by mail from a stratified
random sample of the general population, and (2) in person for tribal members. For mailed
surveys, sampling strata were defined by lists of (1) the general populations of each state, (2)
anglers, and (3) new mothers. A portion of the total sample was allocated to each stratum, with
60% of surveys going to Minnesota and 40% to North Dakota. The Minnesota portion was
allocated equally between the nine county region in the northeast portion of the state and the rest
of the state. The general population list came from a mailing services provider. The lists of
licensed anglers were provided by the Minnesota Department of Natural Resources and the
North Dakota Department of Game and Fish. The list of new mothers in Minnesota was
provided by the Minnesota Department of Health. No list of new mothers was available within
the survey period in North Dakota. The Indian tribes distributed questionnaires to tribal
members following a non-random procedure. Finally, there were five additional surveys in the
"Add-on by project personnel" category. The data were provided in a Microsoft Access file
A-23
-------
(MN-ND_fish consumption_97.mdb). Data collection took place in October and November of
2000.
A.4.1.1. Reformatting the data
The survey respondent was asked to provide fish consumption information for up to five
household members. If there were more than five household members, data was to be provided
for one or two adults and the oldest, middle, and youngest children.
The questionnaire requested demographic information and information about fish
consumption for 12 different classes offish species, divided into two groups, purchased fish and
non-purchased or self-caught fish. For self-caught fish the respondent was asked for the
minimum and maximum length of the fish and the locations where the fish were caught. To
define location the questionnaire divided Minnesota and North Dakota into 17 areas. Different
locations could be selected for each class offish. The file included coefficients for predicting the
mercury concentration in the fish from the length of the fish, the location the fish was caught,
and the class offish.
Several types of information were requested in multiple different ways. Question 3
allowed the respondent to check-off all purchased (restaurant or store-bought) fish consumed by
any family member in the last year—listing 37 different species offish. The respondent was not
asked how much of these fish were eaten. Fish consumption was requested for the 12 classes of
fish. Consumption was defined both in terms of general frequency of consumption (such as 1-5
times per year or 1-2 times per week) and in the number of meals in the last month. The
Microsoft Access file included programmed procedures to calculate the fish consumption in
grams per unit time. The frequency of consumption was used for those calculations rather than
the number of meals per month. As a result, the frequency of consumption was used to calculate
fish consumption in g/day.
Except for seven surveys, the survey ID (household ID) was used to identify the
population from which the household was sampled and the state of residence. Charlene Crocker
from the University of North Dakota, one of the authors of the Minnesota/North Dakota study
report (Benson et al., 2001), provided recommendations for resolving the status of the final seven
surveys. The State variable had the state of residence for each survey. In addition, for the mail
questionnaires, the database had information on the state to which the survey was mailed
(MailState). The respondent was also asked their state of residence (QlSState). In some
cases the respondent reported state of residence was missing or reported as "Other", i.e., than
Minnesota or North Dakota.
The mailing addresses included zip code. The address and zip code were used to get the
county name (CntyName) from geographic databases. The county name was undefined for
some respondents, primarily surveys that were distributed by the Indian tribes and not mailed.
A-24
-------
An additional variable (Strat 9) flagged the nine northeast counties in Minnesota. This
variable was defined based on the available information on sample population and address. All
members of the Bois Forte Tribe were assumed to live in the nine northeast counties because the
tribal reservations were located there. As a result, the variable Strat 9 had no missing values.
The data files have information on 1,572 households and 4,272 individuals. The data
from many Microsoft Access tables were separated into five files:
• MNNDHHDat with data for the households (e.g., household income);
• MNNDIndDat with data for individuals within households (sex, age, bodyweight);
• MNNDFishDat with data on fish consumption;
• MNNDQ3 with data on fish purchased by any family member in the last year; and
• MNNDFISHLOC with information on where fish were caught.
A.4.1.2. Data file comparison
Because of how the data files were organized there were no inconsistencies between files
(as was found in the Florida and Connecticut data). Demographic information in the study report
could be replicated from the processed files with the exception of Income and Angler. Charlene
Crocker from the University of North Dakota provided a revised summary of Income that agreed
with the numbers derived from the processed files. In a few cases, the values in the tables in the
report could only be replicated by assuming the table was mislabeled.
A.4.1.3. Calculation of uncooked and as-consumed fish consumption
The Minnesota and North Dakota questionnaire asked for the "usual" portion size in
ounces. Pictures of the food items with associated weights were provided as a guide. The
pictures showed the food on a plate, suggesting that the weight provided with each picture was
the as-consumed quantity. The respondent was asked the weight of the fish, and if that weight
was the cooked or raw weight. The questionnaire did not ask if the item was cooked before
being eaten. The MN/ND Microsoft Access data file had programming code to calculate fish
consumption. However, that code ignores the uncooked versus cooked responses. Since the
Microsoft Access file had the reported values, EPA chose how to convert the reported values to
uncooked or as-consumed consumption. For the purposes of converting from uncooked to as-
consumed or visa versa, EPA assumed that all reported consumption was eaten cooked because
the questionnaire was not specific enough to identify food items that might have been consumed
uncooked. All food items were converted from uncooked to as-consumed weight by multiplying
A-25
-------
the uncooked weight by 0.75, the value used in the EPA Mercury Study Report to Congress (U.S.
EPA, 1997).
A.5. COMBINED DATA
A.5.1. Description of the Combined Data Files
The data from the three survey databases were combined into one database for analysis.
The combined file contained variables that were common to all three databases. For example,
fish consumption was available in all three databases and was also in the combined database.
The variable indicating whether the respondent catches fish in fresh or salt water locations was in
the Connecticut database, but not in the other files and was not included in the combined
database. Although the variables may be present in all databases, in some cases the values were
measured differently. As a result, some transformations were made to transform the data into
similar units. For categorical data the categories may differ among databases—such as for race-
ethnicity. The combined file contains the categories as recorded in the individual files. To
represent income, education, and race-ethnicity, three variables (CHHIncome, CEducation,
CRaceEthn) were created for the categories that were common across all three studies. In a
few cases, the combined files had variables that were of interest to this analysis even if those
variables were not available in all of the individual databases. For example, the Connecticut and
Minnesota/North Dakota questionnaire had information about whether a household member was
pregnant. The combined file has this information; however, it is missing for the Florida data.
There were three combined files, one for household variables, one for individual
variables, and the third for fish consumption data. In general, these files correspond to the
organization of the files in the individual state databases. In a few cases some transformations
were needed. For example, Connecticut had the education for the head of household, Florida had
education for the randomly selected adult, and Minnesota and North Dakota had education for
the first two listed members, generally the adults. The individual data for the respondent
(member 1) in the Minnesota and North Dakota data was judged to be the closest approximation
to the education data in the other two files. In this case, the education data was moved from the
individual file to the household file.
Table A-l 1 lists the variables in the combined files and the corresponding variables in the
original files for each survey. If a cell was blank, then the individual survey file had no
information for that variable. When the files were combined, the "Add-on" surveys from the
Minnesota/North Dakota survey were excluded from the combined files. Table A-12 shows the
number of household, individual, and fish species records coming from each of the individual
files and in the combined file.
A-26
-------
Table A-ll. Crosswalk of variables in original and combined data files
Combined File
FL
CT
MNND
Identifiers
Study
HHID
PersonID
"FL"
ID
PERSON
"CT"
FAMID
INDID
"MNND"
Survey ID
HMID
Household (HH) variables
State
CntyName
HHIncome
Education
RaceEthn
Population
HHSize
HHEatsCaughtFish
AnyCurPregNurs
"FL"
CntyName
HHIncome
RSAEduc
RSARaceEthn
"General"
"CT"
CntyName
HHIncome
HHHEduc
HHRace
Group [See text]
People
MailState or State
CntyName
HHIncome
Education[Member 1]
RaceEthn[Member 1]
Population
NumPeople
CaughtFish
AnyCurPregNurs
Consumption data by person and species
Location [Away, Home or
Both]
Species
MealPerYr
CookGDayBought
CookGDay Caught
RawGDayBought
RawGDay Caught
GramsPerMeal
Location
Name
Species
NumOccasions
RawGrams, PctRecCaught,
and raw to cooked
conversion
"Both"
FishName
Species
MealsYr
Fsource and Gday
Gmeal
"Both"
HMFish
MealsPerYear
Species, RawCook,
MealPerYear, and
GramsPerMeal
GramsPerMeal
Person variables
Age
Sex
Weight
CurPregNurs
Age
Sex
Weight
Age
SexC
Wt
BFeedC and PregC
Age
Sex
Weight
CurPregNurs
A-27
-------
Table A-12. Number of records in the processed data files
Households
Individuals
Fish dishes or meals
Combined
11,118
23,555
51,653
FL
8,740
17,213
16,099
CT
810
2,080
15,367
MNND
1,568
4,262
20,187
FL = Florida, CT - Connecticut, MNND = Minnesota/North Dakota.
The variable Study identifies the study from which the data was derived. The HHID
numeric variable was used to identify households. HHID was unique within a study, but may not
be unique between studies. The Study and HHID variables were in all files. The PersonID
was a character variable and was used to identify individuals within households. PersonID
was unique within households. These variables were used when merging files.
A.5.1.1. Demographic data
The household variables are variables that were obtained once for each household or
apply to all members of the household. Location variables included the state of residence of the
household (State) and the county name (CntyName).
All studies asked questions about total household income, each using a different set of
questions. The income responses were in the HHIncome variable as descriptive text strings.
Table A-13 shows how the income categories from the different studies were combined into a
common set of income categories.
In the Minnesota/North Dakota study education was requested for the first two listed
household members, the respondent and the second listed member. The combined files had the
education of the respondent. The Connecticut study asked for the education of the head of
household. The Florida study asked for the education of the randomly selected adult. Since the
randomly selected adult may not be the head of household and the respondent (the person that
filled out the questionnaire) in the Minnesota/North Dakota study was not randomly selected and
might not be an adult, the measures of education level associated with the household were not
completely equivalent between studies. Table A-14 shows how the education categories from
the different studies were combined into a common set of education categories.
Race-ethnicity was obtained in all studies, each using a different set of questions. Since
the race-ethnicity of all household members was often the same, uncertainty in defining a
household race-ethnicity using the respondent in the Minnesota/North Dakota study is probably
A-28
-------
less of a problem than with education. Table A-15 shows how the race-ethnicity categories from
the different studies were combined into a common set of race-ethnicity categories.
Table A-13. Common income categories
Common income category
(CHHIncome)
$0-20000
$0-20000
$0-20000
$0-20000
$0-20000
$0-20000
$0-20000
$0-20000
$20000-50000
$20000-50000
$20000-50000
$20000-50000
$20000-50000
$20000-50000
$20000-50000
$20000-50000
$20000-50000
$50000-
$50000-
$50000-
$50000-
$50000-
$50000-
No Response
No Response
No Response
No Response
Household income
(HHIncome)
Less than $10,000
$0-4,999
$5,000-9,999
$10,000 to <$ 15,000
$10,000-14,999
$15,000 to < $20,000
$15,000-19,999
$20,000 and under
$20,000 to < $25,000
$20,000-24,999
$20,000-35,000
$25,000 to < $35,000
$25,000-29,999
$30,000-39,999
$35,000 to < $50,000
$35,000 to $50,000
$40,000-49,999
$50,000 to < $75,000
$50,000-59,999
$60,000-69,999
$70,000-
$75,000 and over
Over $50,000
Missing
No response
Retired
A-29
-------
Table A-14. Common education categories
Common education category (CEducation)
Education
Unknown
Unknown
Unknown
Unknown
l)Some High School or less
1)0-11 years
1)0-11 years
1)0-11 years
1)0-11 years
1)0-11 years
1)0-11 years
1)0-11 years
1)0-11 years
1)0-11 years
1)0-11 years
1)0-11 years
1)0-11 years
1)0-11 years
1)0-11 years
1)0-11 years
2)High School
2)High School
2)High School
3)Some College
3)Some College
3)Some College
3)Some College
3)Some College
4)College graduate
4)College graduate
4)College graduate
4)College graduate
4)College graduate
4)College graduate
No data recorded
No response
Unknown
0 years
1 years
2 years
3 years
4 years
5 years
6 years
7 years
8 years
9 years
10 years
11 years
11 years or less
Grade 1 to 5
Grade 6 to 8
Some High School
12 years
High School/GED
High school (12 years)
13 years
14 years
15 years
Some college
Some college (13-15 years)
16 years
17 years
18 years
2-4 year degree
College grad. (>15 years)
Post Graduate Degree
A-30
-------
Table A-15. Common race-ethnicity categories
Common race-ethnicity categories
(CRaceEthn)
American Indian
American Indian
Asian
Asian
Asian
Asian
Asian
Asian
Asian
Asian
Asian
Black, Non-Hispanic
Black, Non-Hispanic
Black, Non-Hispanic
Hispanic
Hispanic
Hispanic
Hispanic
Hispanic
Hispanic
Unknown
Unknown
Unknown
Unknown
White, Non-Hispanic
White, Non-Hispanic
White, Non-Hispanic
Race-ethnicity
(RaceEthn)
Amer Indian
American Indian
Asian
Asian Indian
Cambodian
Chinese
Filipino
Hmong
Korean
Laotian
Vietnamese
African American
Black non-H
Black, Non-Hispanic
Central American
Dominican
Hispanic
Mexican
Puerto Rican
South American
No response
Other
Unknown
White
White non-H
White, Non-Hispanic
A-31
-------
The Population variable indicated the population (in a statistical sampling sense)
from which the respondents were selected. "General" refers to a random sample of residents
from the state (note that the Minnesota/North Dakota study report uses this term differently in
some tables). The Florida study report discussed the state sample and the Paper Mill sample.
However, from a statistical point of view, these were samples of the general population with
different rates of selection in different counties.
Household size (HHS i z e) included the number of household members. In the
Connecticut study, the respondent provided information for all household members. However,
sometimes the fish consumption information was not reported for some members. For this study,
HHS i z e was the reported number of household members, not the number in the files. Similar
information was not provided in the other studies. If family size is of interest, it would be
possible to calculate the number of records for individuals each household as an indication of the
family size.
It was also possible to identify families that ate self-caught fish from the data files. Since
this variable was of interest in the EPA analysis and because it was provided in the
Minnesota/North Dakota study, it was included in the combined files. This variable was missing
for the Florida and Connecticut data.
The Florida study did not collect information on whether any household members were
pregnant or breastfeeding, while the Connecticut study and Minnesota/North Dakota study did
collect this information. In the Minnesota/North Dakota study, the question asked if any
household member was pregnant or nursing. If so, the survey then asked which household
member was pregnant or nursing. In most cases, when there was someone in the home that was
pregnant or nursing, the specific household member was not identified. The
AnyCurPregNurs variable contained the response to the question about any currently
pregnant or nursing women in the household from the Minnesota/North Dakota study.
A.5.1.2. Fish consumption data
The Florida study asked separate questions regarding fish consumption at-home and
away-from-home. This distinction was retained in the data files using the location variable
(Location). For the Florida study, the at-home and away-from-home fish consumption were
on separate records in the fish consumption data file, Location is set to either "Away" or
"Home". For the other studies, Location = "Both".
The surveys collected information on fish consumption of individual fish species or
classes of species. Each survey defined species differently. The surveys generally reported fish
consumption by species (or a classification very similar to common names for biological species)
A-32
-------
or by classifications similar to a dish name (e.g., crab cake). The Minnesota/North Dakota study
asked about classes of species. In the combined file, the Species variable described the fish
species or fish class.
Each survey asked in some way about grams offish per meal and the number offish
meals per unit time. This information was used to calculate the fish consumption. The Florida
data provided the fish consumption in grams of uncooked fish per week and the meals per week
(NumOccasions). The Florida study did not keep the grams per meal data in the data file.
The Connecticut data had the grams per meal, meals per year, and the grams of as-consumed fish
per day. The Minnesota/North Dakota study provided the meals per year and the grams offish
per meal. The fish consumption was calculated from these variables. Note that a "meal" in
general refers to one type of fish at a meal. If a respondent eats two types of fish at the same
meal, these will be recorded as separate meals. Thus "meals" does not have the usual
interpretation.
The information on fish consumption was used to calculate fish consumption for bought
and caught fish. Each of these was also provided as uncooked weight and cooked (as-consumed)
weight. The following paragraphs describe how these values were calculated.
For the Florida study, the data file provided fish consumption in grams of uncooked fish
per week and the percent of the fish that were recreationally caught (the percentage was almost
always 0% or 100%). The percent recreationally caught was used to divide the uncooked grams
into self-caught and store-bought portions. Dividing the value by 7 (days) converted the weekly
consumption to g/day. The study report also reported conversions for translating from uncooked
to as-consumed weight that were used. These same conversions were used to calculate the as-
consumed weights.
For the Connecticut study, the data files provided the as-consumed fish consumption in
grams per day. The respondents were also asked if the fish was bought at a restaurant, store, or
self-caught. Any combinations of sources could be checked. If the respondent selected two
sources (such as store and caught), half of fish was assumed to come from each source. If all
three sources were selected, one-third was assumed to come from each source. The Study report
also provided some information on how weights were converted from uncooked to as-consumed
(Balcom et al, 1999). This information was used to convert from the as-consumed weight to the
uncooked weight.
The Minnesota/North Dakota survey asked for the frequency of eating fish in each class
as well as the usual serving size. Another check box indicated if the weight was uncooked
weight or as-consumed weight. These categories were associated with numeric values for the
number of meals per year and the grams offish per meal. All fish was assumed to be eaten
cooked. A generic value from the Mercury Study Report to Congress (U.S. EPA, 1997) was
A-33
-------
used to convert weights from uncooked to as-consumed. The as-consumed weight was assumed
to be 75% of the uncooked weight for all fish. The survey divided the fish classes into store-
bought and self-caught fish. These classes were used to define the consumption of store-bought
and self-caught fish.
All surveys collected data on the age, sex, and bodyweight of the respondents. The
Minnesota/North Dakota study collected information on individuals that were pregnant or
nursing. However, as noted above, this information may not be reliable. Nonetheless, this
information was provided in the individual data file in the CurPregNurs variable. The
Connecticut study asked if the household member was pregnant or breast feeding. If either of
these questions was answered positively, the CurPregNurs variable was "Yes". If either was
missing the CurPregNurs variable was DK, otherwise the CurPregNurs variable was "No".
The CurPregNurs variable was "NA" for men.
A.6. CALCULATION OF WEIGHTS
In general, a statistical sampling weight can the thought of as the number of units
(generally, people or households) in the target population "statistically represented" by a unit in
the sample. In a formal sense, the weight is the inverse of the product of the probability of
selection into the sample and the probability of responding to the survey. If weighting strata are
defined such that the probability of selection and the probability of responding are believed to be
the same for all individuals in the strata, the weights for individuals in the strata can be
calculated as the population in the stratum divided by the sample size in the stratum. Typically,
additional weight adjustments are then made to force agreement with certain population counts
(Lohr, 1999).
Sampling weights were calculated to extrapolate the reported fish consumption for the
"General" population respondents to the population in the four states for which there was data:
Florida, Connecticut, Minnesota, and North Dakota.
The weights were calculated in three steps. In the first step weighting strata were defined
for calculation of the initial weights. In each survey, the sample of households to be contacted
was a random sample stratified by county. In most states the probability of selecting a household
was the same in each county. However, in Minnesota, the probability of selecting a household
was higher in the nine northeast counties and in Florida the probability of selection was higher in
counties with paper mills. Without additional information on which segments of the population
were more likely to respond to the survey once selected, it was assumed that every household
within a county had the same probability of responding to the survey. In some counties the
number of completed surveys was very small, making the sampling weight sensitive to the
precise number of respondents if the county was used as the weighting strata. Having more
A-34
-------
respondents in a weighting strata helps to reduce any random variation in the weights. Having
fewer respondents in a strata helps to create weights that reflect different probabilities of
selection in different counties, if differences exist. However, having more variable weights will
increase the variance of the estimates.
The following procedures were used to minimize variation in the weights due to a small
number of respondents in some counties: In each state the counties were sorted from smallest to
largest. Counties were grouped into weighting strata, starting with the smallest counties, until
the number of respondents in the strata was roughly 70 or greater. The cut-off of 70 was
considered reasonable to balance these two kinds of errors. The exact choice of the cutoff is not
expected to have much effect on the estimates. In Minnesota, a different sampling rate was used
for the nine northeast counties compared to the other counties. As a result, the counties were
divided into two groups, the nine northeast counties and other counties. The weighting strata
were then defined separately within these two groups of counties. In all states, the weighting
strata are either individual counties or groups of small counties.
The initial weight was calculated in the second step as:
w = NPop
initial -. T
(A-l)
where Npop is the 2000 census population in the weighting strata in which the respondent lives
and NResp is the number of respondents from the weighting strata with completed surveys. This
calculation assumes that households in the weighting strata are equally likely to respond to the
survey regardless of other characteristics of the household, such as whether they have an
answering machine. A more detailed adjustment for non-response was not possible because data
on non-responding households was not available.
In the last step, the weights were adjusted, or raked, so that the total weights for selected
populations equals known totals from the 2000 Census. Raking the data required classifying all
respondents within each state by characteristics, such as child versus adult. Using population
totals for these groups, raking adjusts the initial weights to agree with the population totals.
When the weights are raked to agree with more that one set of totals (in this case, the total
population in each weighting strata and the total state population for adults and children) the
weights are adjusted sequentially for each set of totals until the weights converge. The raking
procedure can be represented as follows:
A-35
-------
T
Subgroup TT7
r J — l
Subgroup
where Wj is the weight after raking step J, Wj_\ is the weight after the previous raking step
(^/-i = ^initial f°r trie first calculation), the sum is over the subgroup to which the respondent
belongs, and TSuhgroup is the population total for that subgroup. All weights were raked to agree
with the Census 2000 total population for the weighting strata and for the number of adults and
children in the state.
In Connecticut, Minnesota, and North Dakota, the raking had a minimal effect on the
weights and the weighted estimates. In Florida, responses for children and adults other than the
RSA were not included in the data unless the primary meal preparer (PMP) was present.
Because the PMP was not present in 1,383 of the 8,000 households, the data under-represented
the children and some adults. The raked weights corrected for the under representation of the
number children and of some adults by increasing their weights relative to the RSA. As a result,
the weighted survey estimates for Florida should fairly represent the characteristics of at-home
consumption offish and shellfish for all Florida residents. However, the weights cannot adjust
the estimates to account for the fact that out-of-home fish and shellfish consumption was not
obtained household members other than the RSA.
The weights estimate the number of individuals in the population represented by each
respondent (individual with data). When calculating estimates for a state, counties with more
people (and a higher weight total) contribute more to the state estimate than counties with few
people. Thus the estimates do not weight the counties equally. The same weights can be used
for estimates offish and shellfish consumption by the general population within the state as for
the population offish and shellfish consumers. The weight for a consumer offish and shellfish
estimates the number offish and shellfish consumers in the population represented by the
respondent. When calculating estimates for consumers in a state, counties with more consumers
contribute more to the state estimate than counties with few consumers. If counties have a
different proportions of the population that eat fish and shellfish, the relative influence of each
county on the state total will be different than for estimates for all of the population.
Standard statistical procedures do not calculate the correct variance (from which the
standard errors and confidence intervals are derived) for weighted estimates because (1) the
weights are not equal, and (2) household, not individuals, were selected using a random sample.
Special procedures must be used to calculate the correct variances. To implement those
procedures defined "Variance units" and replicate weights. Using these variables, the variance
A-36
-------
of weighted estimates can be calculated using the WesVar or SUDAAN programs or other
software, such as the SAS PROC SURVEYMEANS procedure. WesVar uses replicate weights
to calculate the variance. The replicate weights are related to the variance units. The data files
had replicate weights that can be used in WesVar or other programs. Other programs, using a
method called Taylor series linearization, use the full sample weight and the variance units to
approximate the variance calculated by WesVar. SUDAAN uses Taylor series linearization and
can use the replicate weight for some procedures. PROC SURVEYMEANS also uses Taylor
series linearization.
The variance units were defined by randomly ordering the households within each state
and targeted population. Households were then sequentially numbered starting from 1 up to 50.
After 50, the next household was numbered 1, etc. The numbers defined the variance units, for
example, all households (and individuals in those households) that had the number 1 formed the
first variance unit. Together, these variance units were used to calculate a jackknife estimate of
variance. The variance estimate will have about 49 degrees of freedom.
A-37
-------
APPENDIX B. QA/QC PROCESS
The data files collected and originally processed by others were analyzed. The
documentation for the original files was sometimes inadequate or was not consistent with the
contents of the data files. The reports on the original surveys did not include discussions of the
QA/QC process used in the implementation and data processing of the surveys.
These files were processed as described in Appendix A. When processing the files,
various QC checks were performed, including:
Cross-tabulations of original and receded variables to verify that the receding was
completed correctly;
Calculation of summary statistics to compare results from different steps in the
processing of the files;
Comparing summary statistics from the processed files to published summary statistics;
Using SAS macros for preparation of the tables to simplify some repetitive operations
and minimize some coding problems;
Recreate a subset of the tables using a completely unrelated program to check for
possible errors in the SAS macros; and
The calculations of derived variables were checked by exporting to a Microsoft Excel file
the derived variables and the variables from which they were derived. The calculation of
the derived variables was then checked for a stratified random subset of cases.
Any errors that were found were corrected.
B-l
-------
APPENDIX C. CALCULATION OF PERCENTILES
Different programs use different formulas for estimating percentiles for data from a
random sample. Estimating percentiles for weighted survey data based on a complex probability
sample design requires selection of a formula for calculating the percentiles. The choice can
affect the bias of the percentile estimate.
Hyndman and Fan (1996) provide a summary of commonly used formulas used for
calculating percentiles. The various formulas fall into two general categories, discontinuous and
piecewise continuous functions to approximate the population quantile function. For a
continuous distribution, extreme percentiles can be outside the range of the observed data. Using
a discontinuous quantile function, estimates of extreme percentiles can be very biased. Extreme
percentiles cannot be estimated when using a piecewise continuous quantile function. However,
for percentiles within the range of the data, the piecewise continuous functions interpolate
between the observed values and can provide more reasonable estimates for intermediate
percentiles.
The following formula (Equation C-l, presented by Hyndman and Fan, 1996) defines
plotting points. Linear interpolation between the plotting points is used to define the piecewise
continuous sample quantile function. The location of the plotting points depends on two
parameters, a and (3. Assume a continuous underlying variable and a piecewise-linear sample
quantile function/? = f(x) defined by plotting points (xk,pk), where Xk is the kth ordered
observation,
k-a
Pk =
\-a-p
and a and J3 are constants between 0 and 1. The percentile corresponding to the desired
percentage P is obtained by interpolation between neighboring plotting points.
Different software programs use different a and Rvalues. For symmetric distributions,
Hyndman and Fan argue that a and (3 should be equal. For simple random samples they
recommended setting a and (3 between % and Vs. Unless a and (3 equal 1.0, there will be
percentages pn for which percentiles cannot be calculated. Some software uses the
maximum (xn) or minimum (x\) for these percentiles. However the maximum and minimum may
provide very biased estimates of the desired percentiles. Many programs set these percentiles to
missing.
C-l
-------
There are no standard formulas for calculating percentiles from weighted data. Hyndman
and Fan (1996) implicitly assume that there are no ties and do not discuss weighted data.
Equation (C-2) is a reasonable analog of Equation (C-l) for weighted data and data with tied
values. Note that if all the weights are equal, Equation (C-2) is the same as Equation (C-l).
xk =k'h ordered unique value, k = 1... n,
k
Sk =^ W , Wk = weight for observations equal to xk,
Sn+Wk(l-a-p}
(C-2)
Based on simulated weighted data, Rogers (2003) recommends setting a= /3= 0.5 to
minimize bias, unless adequate data are available for selecting other values.
The only procedure in SAS for calculating weighted percentiles uses a discrete quantile
function. The SUDAAN and WesVar programs that are designed for weighted data use a
calculation procedure equivalent to setting a = 0.0 and P = 1.0. This choice can results in biased
quantile estimates for many distributions.
A SAS macro was written to calculate percentiles using Equation (C-2) with a = /3= 0.5.
The following example illustrates the calculation of weights and compares the results to
SAS PROC Univariate. The characteristics of the calculations are easiest to illustrate using a
small data set. The data used are in Table C-l
Table C-l. Example data
Data value
1
2
3
4
5
Weight
1
2
7
4
16
Table C-2 shows the percentiles for the example data calculated using the macro and SAS
PROC Univariate.
C-2
-------
Table C-2. Percentile estimates from the macro and from PROC Univariate
for the example data
Percentage for the
desired percentile
1%
5%
10%
25%
50%
75%
90%
95%
99%
Percentile estimate calculated from
PROC
Macro Univariate
1
1.667 2
2.222 2.5
3.182 3
4.30 5
5
5
5
5
Figure C-l shows the assumed inverse cumulative distribution function defined by the
macro and by PROC Univariate. On the horizontal axis is the percentile to be estimated
(expressed as a fraction) and on the vertical axis is the estimated percentile value. For example,
going up from 0.50 on the horizontal axis until hitting the line for the macro estimate and then
going left from that point to the vertical axis gives an estimate of the median (50th percentile) of
4.3 using the macro. Based on the line for PROC Univariate, the estimate of the median is 5.0.
The curves in Figure C-l approximate the cumulative distribution function for the population
values. Note that the macro uses a piecewise continuous function to approximate the cumulative
distribution function and PROC Univariate uses a step function where the width of the step is
proportional to the weight of the observation. If the percentage for the desired percentile falls
exactly on the rise between steps, the estimate is the midpoint between the two steps. If a = |3 =
0.5 then the linear interpolation from Equation (C-2) goes through the middle of each step. If the
percentage for the desired percentile is close to 0% or 100%, PROC Univariate returns the
minimum or maximum respectively and the macro returns a missing value.
-------
Example percentages
•Univariate
•Macro
0.2 0.4 0.6 0.8 1
Percentage for desired percentile
Figure C-l. Comparison of the cumulative inverse cumulative distribution
function estimates from the macro and PROC Univariate.
If the values in the population (in this report, fish consumption rates) can be considered
measures on a continuous scale, then (1) the true percentiles can take values other than the
sampled values, (2) the extreme percentiles of the distribution can be less than the sampled
minimum or greater than the sampled maximum value, and (3) the extent to which the extreme
percentiles are outside the range of the observed minimum and maximum cannot be determined
from the data (in this case the macro returns a missing value). On the assumption that the fish
consumption rate can be considered a continuous measure, the characteristics of the estimates
from the macro more closely match the characteristics of the true percentiles than do the
estimates from PROC Univariate. Note that the difference between the estimates from the macro
and PROC Univariate will always be less than the difference between the two closest
observations that bracket the estimate from the macro. For a fixed population, larger sample
sizes will be associated with smaller differences between the percentiles from the macro and
PROC Univariate.
Even with a small number of observations, the macro will provide estimates of extreme
percentiles if the weight for the smallest or largest observation is small. In the example data, the
weight for the smallest observation is 3.33% of the total weight. In this case, the macro will
C-4
-------
estimate percentiles for percentages as low as 1.67% (half of 3.33%). However, the weight for
the highest observation is 46.67% of the total weight. As a result, the macro will not estimate
percentiles for percentages greater than 73.3% (100%-half of (100%-46.67%)). The maximum
observed value returned by PROC Univariate is likely to be a biased estimate for extreme upper
percentiles. Likewise, the minimum observed value is likely to be a biased estimate for extreme
lower percentiles.
The macro uses linear interpolation between the observed values to calculate percentiles.
Linear interpolation is expected to work best when the cumulative distribution function is close
to linear in the region where interpolation is used. If the cumulative distribution function is not
close to linear, the percentile estimates may be slightly improved by transforming the data so that
the cumulative distribution function in the transformed scale is approximately linear, calculating
percentiles in the transformed scale, and then transforming the percentiles back to the original
measurement scale. Percentiles were calculated using the log-transformed values because the
cumulative distribution function for the log-transformed values is likely to be more linear than
for the untransformed values when the data are skewed. The differences in the percentiles
between using the measured data and the log-transformed data are small relative to the
differences between using the macro with interpolation and using PROC Univariate. Table C-3
compares the example-data percentile estimates using different methods.
Table C-3. Percentile estimates from the macro and from PROC Univariate
for the example data
Percentile estimate calculated from
Percentage for the desired
percentile
5%
10%
25%
50%
Macro, original data
1.667
2.222
3.182
4.300
Macro, log-
transformed data
1.587
2.189
3.161
4.277
PROC Univariate
2
2.5
o
J
5
C-5
-------
APPENDIX D. GLOSSARY
Table D-l details abbreviations used in this report.
Table D-l. Abbreviations used throughout the data documentation report
FL
CT
MN
ND
HH
FFQ
RSA
PMP
MNND
CI
Florida
Connecticut
Minnesota
North Dakota
Household
Food frequency questionnaire
Randomly selected adult (in the Florida survey)
Primary meal preparer (in the Florida survey)
Minnesota/North Dakota study
95% confidence interval
Consumption—Unless otherwise quantified, the fish and shellfish consumption rate is given in
g/day.
Consumption per kilogram bodyweight—Fish and shellfish consumption rate (g/day) divided by
the respondents reported bodyweight in grams per day to calculate grams per kilogram
bodyweight per day.
Consumption rate—The fish or shellfish consumption during the recall period divided by the
length of the recall period. The consumption rate can also be calculated as the typical weight of
fish consumed per meal divided by the number of meals per unit time (such as three meals per
month). For this report, the consumption rate was converted to g/day.
General population—The general population refers to a set of respondents that were randomly
selected to represent the state population. The term "General" distinguishes those respondents
that are used to calculate population estimates from targeted populations.
D-l
-------
Population—All residents within a state. In this document, "Population" associated with survey
results refers to estimates of values for a state population derived from the weighted survey data.
See also "General population."
Subgroup—A subset of the data defined by an analysis variable, such as a subset of respondents
in an income, race-ethnicity, or gender group. Confidence intervals are calculated for
respondents within subgroups. Significance tests test if the means or geometric means are the
same across subgroups defined by an analysis variable.
Targeted population—A subset of the general population within a state (e.g., children,
recreational anglers).
D-2
-------
APPENDIX E. TABLES OF FISH AND SHELLFISH CONSUMPTION
Appendix E has fish and shellfish consumption tables for each state's general population
using the as-consumed weight offish and shellfish, additional tables summarizing consumption
as raw weight, and tables for targeted populations. Section E. 1 has a detailed overview of the
tables. For the state's general population, Sections E.2 through E.8 have tables presenting mean
and geometric mean fish and shellfish consumption and the percentage of respondents that
reported eating fish or shellfish. Each table provides separate estimates for each state.
Additional tables provide percentile estimates, a more detailed breakdown of the general
population, and consumption in raw weight and raw weight per kilogram bodyweight. Section
E.9 has plots showing the mean and confidence intervals for fish and shellfish consumption and
the percentage of respondents that reported eating fish or shellfish. Section E. 10 has tables
showing the mean and geometric mean fish consumption for each state's general population and
targeted populations. Section E.I 1 lists the species offish consumed by the general population.
Tables for the targeted populations with additional details are in Section E-12 along with tables
listing the species consumed by the targeted populations.
E.I. ORGANIZATION OF THE TABLES
There are tables for each summary statistic organized into the following sections:
• Section E.2 Mean fish consumption;
• Section E.3 Mean fish consumption per kilogram bodyweight;
• Section E.4 Percentage of respondents that reported eating fish and shellfish;
• Section E.5 Mean fish consumption for respondents that consumed fish or shellfish
(consumers only);
• Section E.6 Mean fish consumption per kilogram bodyweight for respondents that
consumed fish or shellfish (consumers only);
• Section E.7 Geometric mean fish consumption for respondents that consumed fish or
shellfish (consumers only);
• Section E.8 Geometric mean fish consumption per kilogram bodyweight for respondents
that consumed fish or shellfish (consumers only);
• Section E-9 Plots of confidence intervals for the state's general population;
• Section E-10 Fish and shellfish consumption for targeted populations;
E-l
-------
• Section E-l 1 Species eaten and caught; and
• Section E-12 Additional details for targeted populations.
Within each of these sections are separate tables, each using a different independent
variable to define the subgroups for analysis. Each table presents summary statistics, confidence
intervals, and ap-va\ue to assess the significance of differences among subgroups within states.
In the discussion of the results in Section 3, differences with ^-values < 0.05 are described as
statistically significant.
The tables reporting the data on fish and shellfish consumption show the state, the levels
of the independent variable (the column heading is Subgroup), the number of respondents (N),
the weighted population size (in thousands) represented by the respondents, (weighted N/1,000),
the mean or average fish and shellfish consumption rate (as-consumed in g/day or g/kg-day), the
upper and lower 95% confidence interval for the mean, and ap-va\ue for assessing the statistical
significance of within-state differences between the subgroup means. The confidence intervals
are shown for each level of the independent variable except for race-ethnicity for which the
Asian and American Indian categories were combined as "Other." Missing values of the
independent variable are shown as "Unknown." The/>-value calculations excluded the
"Unknown" category.
The calculations for the confidence interval and ^-values assume the estimates of the
mean have a normal distribution. If the sample size (N) is smaller than about 100, this
assumption is uncertain and the confidence interval andp-va\ue are approximate. If there are
relatively few respondents in a subgroup category, the confidence intervals can be imprecise.
Results for these categories are left in the tables for completeness, although the results may not
be useful due to the small sample sizes. If there is only one respondent, the confidence interval
cannot be calculated. In a few cases with small sample sizes the normality assumption is not
correct and the calculated lower confidence interval is <0 even though fish and shellfish
consumption rates can never be negative. In these cases, the calculated values are presented
even though they are unrealistic. A plus sign (+) between the columns for the upper confidence
limit and the/?-value indicates categories that were combined for the calculation of the/?-value.
E-2
-------
E.2. MEAN FISH AND SHELLFISH CONSUMPTION FOR THE STATE'S GENERAL
POPULATION, AS-CONSUMED WEIGHT PER DAY
Tables described in this section (Tables E-l-E-5) present the estimated mean
consumption rate offish and shellfish (as-consumed g/day) for the general population in
Connecticut, Florida, Minnesota, and North Dakota. The estimates are broken out by state and
categorical level of the independent variable, in separate tables defined by the independent
variables: adult versus child, two classifications of age and gender combinations, education,
income, race-ethnicity, gender, and type offish consumed.
Table E-l. Mean consumption, general population, per capita, by state and
adult/child (as-consumed g/day with 95% CIs)
State
CT
FL
MN
ND
Subgroup
Adult
Child
Unknown
Adult
Child
Adult
Child
Unknown
Adult
Child
Unknown
N
337
83
11
13589
3592
650
185
6
430
165
7
Weighted
N/1000
2502
786
99
12339
3613
3614
1224
62
474
157
7
Mean
32
9
6
31
13
21
9
0
20
15
23
.442
.897
.869
.052
.617
.252
.573
.016
.215
.028
.980
Lower
Conf .
Limit
25
6
-2
29
12
15
7
-0
16
9
1
.853
.091
.802
.253
.100
.786
.342
.021
.538
.565
.703
Upper
Conf.
Limit
39
13
16
32
15
26
11
0
23
20
46
.032
.704
.539
.851
.134
.718
.803
.053
.893
.490
.257
p-value
<.0001
<.0001
0.0001
0.0387
E-3
-------
Table E-2. Mean consumption, general population, per capita, by state and
age/gender (5 categories) (as-consumed g/day with 95% CIs)
Weighted
State Subgroup N N/1000
CT Child 1-14 77 726
Female 15-44 91 689
Female 45+ 94 694
Male 15-44 14 119
Male 45+ 144 1061
Unknown 11 99
FL Child 1-14 2751 2787
Female 15-44 3799 3486
Female 45+ 2833 2553
Male 15-44 1783 1646
Male 45+ 5020 4539
Unknown 995 941
MN Child 1-14 146 1017
Female 15-44 147 968
Female 45+ 203 978
Male 15-44 63 292
Male 45+ 276 1583
Unknown 6 62
ND Child 1-14 121 116
Female 15-44 124 129
Female 45+ 128 144
Male 15-44 45 47
Male 45+ 177 196
Unknown 7 7
Lower
Conf .
Mean
9.
34.
26.
11.
36.
6.
13.
30.
27.
32.
33.
16.
9.
22.
22.
8.
20.
0.
15.
15.
22.
17.
20.
23.
600
496
670
932
109
869
005
470
750
497
076
388
556
499
370
648
602
016
634
468
323
781
924
980
Limit
6
20
20
6
28
-2
11
27
25
26
30
12
7
6
16
5
16
-0
9
11
16
12
16
1
.265
.707
.539
.369
.098
.802
.487
.289
.710
.412
.915
.419
.052
.050
.996
.523
.736
.021
.745
.810
.237
.024
.509
.703
Upper
Conf.
Limit
12.
48.
32.
17.
44.
16.
14.
33.
29.
38.
35.
20.
12.
38.
27.
11.
24.
0.
21.
19.
28.
23.
25.
46.
936
284
801
495
120
539
523
652
790
582
236
356
059
948
745
774
468
053
524
125
409
537
339
257
p-value
<.0001
<.0001
<.0001
0.0728
E-4
-------
Table E-3. Mean consumption, general population, per capita, by state and
education (as-consumed g/day with 95% CIs)
State Subgroup
CT 0-11 years
High School
Some College
College grad
FL 0-11 years
High School
Some College
College grad
Unknown
MN 0-11 years
High School
Some College
College grad
Unknown
ND 0-11 years
High School
Some College
College grad
Unknown
N
13
89
66
263
1744
5677
5261
4367
132
46
236
260
256
43
31
143
195
196
37
Weighted
N/1000
97
682
504
2105
1523
5118
4948
4240
123
214
1332
1330
1808
215
35
144
212
206
42
Lower
Conf .
Mean
25.
23.
34.
25.
23.
26.
28.
27.
26.
23.
16.
22.
14.
16.
13.
23.
17.
18.
19.
237
778
481
468
775
277
233
998
400
954
890
761
963
509
946
007
185
738
614
Limit
7
17
18
19
18
24
25
26
16
6
11
10
10
12
6
16
12
12
7
.646
.407
.457
.295
.914
.179
.340
.156
.310
.975
.833
.248
.690
.264
.303
.089
.559
.270
.342
Upper
Conf.
Limit
42.
30.
50.
31.
28.
28.
31.
29.
36.
40.
21.
35.
19.
20.
21.
29.
21.
25.
31.
828
150
505
641
635
375
127
839
490
934
947
274
236
754
589
924
812
206
886
p-value
0.6465
0.1296
0.5431
0.2235
E-5
-------
Table E-4. Mean consumption, general population, per capita, by state and
income (as-consumed g/day with 95% CIs)
State Subgroup N
CT $ 0-20000 41
$20000-50000 155
$50000- 219
Unknown 16
Weighted
N/1000
312
1179
1778
119
Mean
23.532
32.019
23.551
22.525
Lower
Conf.
Limit
13.970
18.294
18.869
11.253
Upper
Conf.
Limit
33.094
45.744
28.234
33.797
FL $ 0-20000 3746 3408 26.004 22.007 30.000
$20000-50000 7353 6814 28.114 25.880 30.349
$50000- 3417 3250 30.601 28.187 33.015
Unknown 2665 2480 21.251 19.083 23.419
MN $ 0-20000 89 373 28.456 15.506 41.406
$20000-50000 328 1802 19.268 9.906 28.630
$50000- 327 2155 16.354 12.093 20.614
Unknown 97 570 13.932 10.095 17.769
ND $ 0-20000 53 56 24.527 9.301 39.753
$20000-50000 252 268 17.542 13.218 21.866
$50000- 239 251 19.215 15.597 22.834
Unknown 58 63 19.253 10.808 27.699
p-value
0.5246
0.1249
0.2516
0.4371
E-6
-------
Table E-5. Mean consumption, general population, per capita, by state and
race-ethnicity (as-consumed g/day with 95% CIs)
State Subgroup
CT White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
FL White, Non-Hispanic 12957
Black, Non-Hispanic
Hispanic
Other
Unknown
MN White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
ND White, Non-Hispanic
Black, Non-Hispanic
Other
Unknown
N
380
9
20
20
2
957
842
673
382
327
779
1
3
19
39
551
2
17
32
Weighted
N/1000
2968
66
178
155
21
11887
1690
1719
330
325
4473
1
50
173
204
585
2
16
36
Mean
Lower
Conf.
Limit
Upper
Conf.
Limit
16.111 13.389
0.000
45.011 -10.06
56.041 -45.60
22.254 12.125
p-value
26.984 22.218 31.751 <.0001
4.362 -3.422 12.145
29.688 8.433 50.942
25.709 -8.090 59.508
0.427 0.427 0.427
27.077 25.660 28.494 0.6051
26.552 22.572 30.532
26.604 20.661 32.547
35.779 24.189 47.369
24.751 15.478 34.024
18.832 <.0001
100.08
157.68
32.383
19.111 14.972 23.250 0.1451
15.717 15.028 16.405
20.887 14.317 27.457
16.239 6.998 25.479
E.3. MEAN FISH AND SHELLFISH CONSUMPTION FOR THE GENERAL
POPULATION, AS-CONSUMED WEIGHT PER KILOGRAM BODYWEIGHT
PER DAY
Tables described in this section (Tables E-6-E-11) present the estimated mean
consumption rate per kilogram bodyweight (as-consumed g/kg-day) for the general population in
Connecticut, Florida, Minnesota, and North Dakota. The estimates are broken out by state and
categorical level of the independent variable, in separate tables defined by the independent
variables: adult versus child, two classifications of age and gender combinations, education,
income, race-ethnicity, gender, and type offish consumed.
E-7
-------
Table E-6. Mean consumption per bodyweight, general population, per
capita, by state and adult/child (as-consumed g/kg-day with 95% CIs)
State
CT
Subgroup
Adult
Child
Unknown
N
331
78
11
Weighted
N/1000
2460
737
99
Mean
0.437
0.365
0.089
Lower
Conf.
Limit
0.357
0.237
-0.032
Upper
Conf.
Limit
0.517
0.494
0.210
p-value
0.2807
FL
Adult
Child
12078
3289
11517
3310
0.442
0.557
0.415
0.497
0.468
0.617
0.0003
MN
Adult
Child
Unknown
648
184
5
3612
1224
61
0.297
0.361
0.003
0.202
0.241
-0.004
0.392
0.481
0.009
0.3830
ND
Adult
Child
Unknown
414
158
3
456
151
3
0.283
0.443
0.107
0.229
0.259
-0.076
0.337
0.627
0.290
0.0577
E-S
-------
Table E-7. Mean consumption per bodyweight, general population, per
capita, by state and age/gender (9 categories) (as-consumed g/kg-day with
95% CIs)
State Subgroup
CT Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
MN Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
ND Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
N
26
26
21
17
85
77
14
80
63
11
1102
938
864
1537
2264
2080
1638
2540
2206
198
47
46
68
47
132
162
55
120
155
5
30
44
55
42
95
99
36
90
81
3
Weighted
N/1000
253
239
193
141
634
563
119
594
461
99
1134
956
848
1477
2178
2025
1551
2383
2090
185
437
298
337
331
722
854
275
731
851
61
30
42
52
43
101
112
38
97
92
3
Lower
Conf .
Mean
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.317
.508
.270
.671
.463
.434
.162
.470
.345
.089
.885
.435
.365
.436
.529
.412
.441
.428
.384
.352
.568
.333
.219
.665
.240
.342
.099
.237
.243
.003
.665
.513
.397
.180
.281
.377
.217
.215
.288
.107
Limit
0
0
0
0
0
0
0
0
0
-0
0
0
0
0
0
0
0
0
0
0
0
0
0
-0
0
0
0
0
0
-0
0
0
0
0
0
0
0
0
0
-0
.133
.325
.129
.061
.316
.332
.079
.372
.270
.032
.752
.370
.290
.349
.489
.382
.358
.386
.354
.256
.304
.168
.147
.200
.195
.251
.060
.161
.197
.004
.251
.279
.217
.121
.212
.254
.128
.169
.187
.076
Upper
Conf.
Limit
0.
0.
0.
1.
0.
0.
0.
0.
0.
0.
1.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
1.
0.
0.
0.
0.
0.
0.
1.
0.
0.
0.
0.
0.
0.
0.
0.
0.
501
692
410
281
611
536
246
569
420
210
019
500
440
523
570
442
524
469
414
447
832
498
292
530
285
433
138
313
290
009
080
746
576
238
350
500
306
261
389
290
p-value
0.0006
<.0001
<.0001
0.0617
E-9
-------
Table E-8. Mean consumption per bodyweight, general population, per
capita, by state and age/gender (5 categories) (as-consumed g/kg-day with
95% CIs)
State Subgroup
CT Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
FL Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
MN Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
ND Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
N
72
88
92
14
143
11
2740
3477
2487
1719
4746
198
145
146
203
63
275
5
115
118
124
44
171
3
Weighted
N/1000
676
668
679
119
1055
99
2776
3350
2413
1629
4473
185
1016
968
978
292
1582
61
111
123
139
46
189
3
Mean
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.374
.539
.403
.162
.415
.089
.607
.479
.437
.429
.407
.352
.405
.376
.335
.106
.240
.003
.523
.253
.347
.249
.250
.107
Lower
Conf .
Limit
0
0
0
0
0
-0
0
0
0
0
0
0
0
0
0
0
0
-0
0
0
0
0
0
-0
.252
.343
.311
.079
.342
.032
.540
.428
.402
.350
.381
.256
.262
.065
.255
.067
.195
.004
.304
.195
.246
.160
.194
.076
Upper
Conf.
Limit
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
496
736
494
246
489
210
675
529
473
509
434
447
549
687
414
145
286
009
743
310
448
339
307
290
p-value
0.0002
<.0001
<.0001
0.0117
E-10
-------
Table E-9. Mean consumption per bodyweight, general population, per
capita, by state and education (as-consumed g/kg-day with 95% CIs)
State Subgroup
CT 0-11 years
High School
Some College
College grad
FL 0-11 years
High School
Some College
College grad
Unknown
MN 0-11 years
High School
Some College
College grad
Unknown
ND 0-11 years
High School
Some College
College grad
Unknown
N
13
87
62
258
1481
4992
4791
4012
91
46
234
259
255
43
29
138
183
188
37
Weighted
N/1000
97
667
477
2055
1387
4722
4650
3979
89
214
1331
1329
1808
215
32
139
200
197
42
Mean
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.325
.378
.410
.425
.400
.464
.488
.471
.464
.340
.290
.407
.256
.242
.225
.420
.275
.311
.345
Lower
Conf .
Limit
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.073
.264
.271
.324
.322
.422
.431
.439
.275
.101
.158
.180
.200
.177
.091
.222
.197
.208
.107
Upper
Conf.
Limit
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
576
491
549
526
478
506
546
502
654
579
421
634
312
307
360
617
352
413
583
p-value
0.8676
0.2720
0.5706
0.4066
E-ll
-------
Table E-10. Mean consumption per bodyweight, general population, per
capita, by state and income (as-consumed g/kg-day with 95% CIs)
State Subgroup N
CT $ 0-20000 40
$20000-50000 150
$50000- 214
Unknown 16
Weighted
N/1000
303
1137
1737
119
Mean
0.389
0.472
0.380
0.323
Lower
Conf.
Limit
0.230
0.298
0.292
0.148
Upper
Conf.
Limit
0.547
0.646
0.467
0.497
p-value
0.6479
FL $ 0-20000 3314 3158 0.473 0.401 0.545
$20000-50000 6678 6430 0.480 0.438 0.523
$50000- 3136 3066 0.514 0.471 0.558
Unknown 2239 2172 0.354 0.315 0.393
0.4704
MN $ 0-20000 87 371 0.401 0.204 0.597
$20000-50000 326 1801 0.337 0.166 0.507
$50000- 327 2155 0.288 0.197 0.379
Unknown 97 570 0.244 0.191 0.297
0.6062
ND $ 0-20000 51 54 0.517 0.035 0.999
$20000-50000 235 251 0.272 0.200 0.344
$50000- 233 245 0.307 0.252 0.361
Unknown 56 60 0.415 0.189 0.642
0.3507
E-12
-------
Table E-ll. Mean consumption per bodyweight, general population, per
capita, by state and race-ethnicity (as-consumed g/kg-day with 95% CIs)
State Subgroup
CT White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
FL White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
MN White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
ND White, Non-Hispanic
Black, Non-Hispanic
Other
Unknown
Lower Upper
Weighted Conf. Conf.
N N/1000 Mean Limit Limit p-value
0.342 0.473 <.0001
-0.037 0.136
0.140 0.826
-0.115 1.332
0.006 0.006
0.431 0.481 0.2214
0.448 0.623
0.352 0.574
0.367 0.817
0.292 0.571
0.218 0.320 <.0001
-0.152 1.442
-0.590 3.060
0.179 0.466
0.239 0.410 0.2590
0.216 0.284
0.181 0.370
0.080 0.512
370
9
20
19
2
607
603
556
327
274
775
1
3
19
39
528
2
13
32
2888
66
178
143
21
11113
1522
1619
297
276
4469
1
50
173
204
559
2
13
36
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
.407
.050
.483
.609
.006
.456
.535
.463
.592
.431
.269
.000
.645
.235
.322
.325
.250
.275
.296
E.4. PERCENTAGE OF THE GENERAL POPULATION RESPONDENTS THAT
REPORTED EATING FISH AND SHELLFISH
Tables described in this section (Tables E-12 and E-13) present the estimated percentage
of respondents in the general population that reported eating fish or shellfish in Connecticut,
Florida, Minnesota, and North Dakota. The estimates are broken out by state and categorical
level of the independent variable, in separate tables defined by the independent variables: adult
versus child, two classifications of age and gender combinations, education, income, race-
ethnicity, gender, and type offish consumed. The expected percentage of respondents eating
fish or shellfish depends in part on the recall period used by the surveys. In Florida the recall
period was 7 days. In Minnesota and North Dakota the recall period was 1 year. In Connecticut
the recall period was roughly a year based on the question wording, but was not specified. As a
result, the percentage of respondents reporting that they ate fish or shellfish during the recall
period is expected to be lower for Florida than for the other states.
E-13
-------
Table E-12. Percent eating fish and shellfish, per capita, by state and
adult/child (with 95% CIs)
State
Subgroup
Weighted
N N/1000
Percent
Lower
Conf.
Limit
Upper
Conf.
Limit
p-value
CT
Adult
Child
Unknown
337
83
11
2502
786
99
89.4
68.9
76.1
85.2
54.7
48.4
93.6
83.1
103.9
<.0001
FL
Adult
Child
13589
3592
12339
3613
52.5
39.8
51.4
37.1
53.5
42.5
<.0001
MN
Adult
Child
Unknown
650
185
6
3614
1224
62
96.6
92.4
1.5
94.3
84.8
-2.0
98.9
99.9
5.0
0.0492
ND
Adult
Child
Unknown
430
165
7
474
157
7
95.5
93.9
75.2
93.5
88.5
39.5
97.5
99.3
111.0
0.5587
E-14
-------
Table E-13. Percent eating fish and shellfish, per capita, by state and
age/gender (5 categories) (with 95% CIs)
Weighted
State Subgroup N N/1000
CT Child 1-14 77 726
Female 15-44 91 689
Female 45+ 94 694
Male 15-44 14 119
Male 45+ 144 1061
Unknown 11 99
FL Child 1-14 2751 2787
Female 15-44 3799 3486
Female 45+ 2833 2553
Male 15-44 1783 1646
Male 45+ 5020 4539
Unknown 995 941
MN Child 1-14 146 1017
Female 15-44 147 968
Female 45+ 203 978
Male 15-44 63 292
Male 45+ 276 1583
Unknown 6 62
ND Child 1-14 121 116
Female 15-44 124 129
Female 45+ 128 144
Male 15-44 45 47
Male 45+ 177 196
Unknown 7 7
Percent
Lower
Conf.
Limit
Upper
Conf.
Limit
69.2
89.1
87.0
70.5
91.9
76.1
54.3
81.0
79.9
48.3
87.1
48.4
84.2
97.2
94.0
92.7
96.6
103.9
39.9
52.3
55.9
45.4
53.6
39.2
36.9
50.6
54.0
42.6
52.1
34.3
42.9
54.0
57.9
48.1
55.0
44.1
93.3
94.9
95.4
92.4
98.1
1.5
86.4
89.2
91.6
78.8
96.1
-2.0
100.1
100.7
99.1
106.0
100.1
5.0
93.5
93.5
94.9
100.0
96.1
75.2
88.0
88.3
91.2
100.0
93.3
39.5
99.1
98.7
98.5
100.0
98.8
111.0
p-value
<.0001
<.0001
0.0435
0.5223
E.5. MEAN FISH AND SHELLFISH CONSUMPTION FOR THE GENERAL
POPULATION, CONSUMERS ONLY, AS-CONSUMED WEIGHT PER DAY
Tables described in this section (Tables E-14-E-18) present the estimated mean
consumption rate offish and shellfish (as-consumed g/day) for those that reported consuming
fish and shellfish. The estimates are broken out by state (Connecticut, Florida, Minnesota, and
North Dakota) and categorical level of the independent variable, in separate tables defined by the
independent variables: adult versus child, two classifications of age and gender combinations,
education, income, race-ethnicity, gender, and type offish consumed. The expected
consumption rate for those that eat fish or shellfish depends in part on the recall period used by
the surveys. In Florida the recall period was 7 days. In Minnesota and North Dakota the recall
period was 1 year. In Connecticut the recall period was roughly a year based on the question
E-15
-------
wording, but was not specified. As a result, the calculated consumption rate for consumers of
fish and shellfish is expected to be higher for Florida than for the other states.
Table E-14. Mean consumption, consumers only, by state and adult/child
(as-consumed g/day with 95% CIs)
State
CT
Subgroup
Adult
Child
Unknown
N
302
58
9
Weighted
N/1000
2237
542
75
Mean
36.290
14.369
9.022
Lower
Conf.
Limit
29.214
10.231
-0.715
Upper
Conf.
Limit
43.365
18.508
18.760
p-value
<.0001
FL
Adult
Child
7131
1435
6473
1438
59.191
34.204
56.201
31.663
62.181
36.745
<.0001
MN
Adult
Child
Unknown
623
172
1
3492
1131
1
21.994
10.365
1.047
16.371
8.024
27.618
12.706
0.0002
ND
Adult
Child
Unknown
410
155
5
453
147
6
21.172
16.001
31.:
17.323
10.303
5.107
25.022
21.699
58.654
0.0498
E-16
-------
Table E-15. Mean consumption, consumers only, by state and age/gender (5
categories) (as-consumed g/day with 95% CIs)
State Subgroup
CT Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
FL Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
MN Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
ND Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
N
54
82
82
10
132
9
1102
1996
1588
813
2678
389
138
140
189
59
269
1
113
116
121
45
170
5
Weighted
N/1000
503
614
604
84
974
75
1113
1825
1428
747
2431
369
948
920
932
270
1552
1
108
121
137
47
188
6
Lower
Conf .
Mean
13.
38.
30.
16.
39.
9.
32.
58.
49.
71.
61.
41.
10.
23.
23.
9.
21.
1.
16.
16.
23.
17.
21.
31.
865
702
671
924
302
022
579
221
624
620
748
825
247
697
460
358
007
047
715
547
533
781
781
880
Limit
10
23
24
11
30
-0
30
52
46
59
58
34
7
6
17
6
17
10
12
16
12
17
5
.367
.801
.209
.800
.770
.715
.108
.826
.652
.401
.206
.216
.583
.438
.786
.197
.154
•
.514
.747
.916
.024
.263
.107
Upper
Conf.
Limit
17.
53.
37.
22.
47.
18.
35.
63.
52.
83.
65.
49.
12.
40.
29.
12.
24.
•
22.
20.
30.
23.
26.
58.
364
603
133
048
834
760
050
616
595
838
290
434
911
957
135
520
859
917
346
150
537
298
654
p-value
<.0001
<.0001
<.0001
0.0879
E-17
-------
Table E-16. Mean consumption, consumers only, by state and education (as-
consumed g/day with 95% CIs)
State Subgroup
CT 0-11 years
High School
Some College
College grad
FL 0-11 years
High School
Some College
College grad
Unknown
MN 0-11 years
High School
Some College
College grad
Unknown
ND 0-11 years
High School
Some College
College grad
Unknown
N
13
78
60
218
716
2683
2739
2376
52
41
220
250
243
42
27
139
183
189
32
Weighted
N/1000
97
584
450
1723
620
2419
2550
2274
48
185
1237
1268
1719
215
31
141
199
199
36
Lower
Conf .
Mean
25.
27.
38.
31.
58.
55.
54.
52.
67.
27.
18.
23.
15.
16.
15.
23.
18.
19.
22.
237
778
610
107
377
605
783
198
073
774
186
877
743
559
916
624
295
350
504
Limit
7
21
21
24
48
51
49
49
46
8
13
10
11
12
7
16
13
12
8
.646
.950
.327
.497
.154
.627
.992
.400
.072
.233
.065
.956
.295
.302
.527
.534
.611
.759
.620
Upper
Conf.
Limit
42.
33.
55.
37.
68.
59.
59.
54.
88.
47.
23.
36.
20.
20.
24.
30.
22.
25.
36.
828
606
893
717
600
583
573
997
073
315
307
799
190
816
305
714
980
942
388
p-value
0.6934
0.2546
0.4751
0.3684
E-18
-------
Table E-17. Mean consumption, consumers only, by state and income (as-
consumed g/day with 95% CIs)
State Subgroup
CT $ 0-20000
$20000-50000
$50000-
Unknown
FL $ 0-20000
$20000-50000
$50000-
Unknown
MN $ 0-20000
$20000-50000
$50000-
Unknown
ND $ 0-20000
$20000-50000
$50000-
Unknown
N
36
135
186
12
1707
3709
1960
1190
79
302
321
94
50
235
231
54
Weighted
N/1000
270
1006
1490
87
1537
3404
1844
1126
339
1645
2109
530
53
250
244
59
Lower
Conf .
Mean
27.
37.
28.
30.
57.
56.
53.
46.
31.
21.
16.
14.
26.
18.
19.
20.
124
523
100
695
651
274
931
809
261
111
707
994
081
797
783
685
Limit
16
22
23
20
51
52
50
43
17
11
12
11
10
14
16
11
.491
.601
.137
.059
.154
.279
.346
.284
.285
.055
.417
.327
.031
.404
.124
.667
Upper
Conf.
Limit
37
52
33
41
64
60
57
50
45
31
20
18
42
23
23
29
.756
.444
.063
.330
.147
.269
.516
.333
.238
.166
.998
.660
.132
.189
.441
.704
p-value
0.5102
0.5398
0.1734
0.5251
Table E-18. Mean consumption, consumers only, by state and gender (as-
consumed g/day with 95% CIs)
State
CT
FL
MN
ND
Subgroup
Male
Female
Male
Female
Unknown
Male
Female
Male
Female
N
177
192
4066
4206
294
403
393
277
293
Weighted
N/1000
1377
1478
3758
3874
279
2379
2244
295
311
Mean
32.
30.
59.
51.
40.
17.
21.
20.
19.
234
641
424
Oil
828
159
252
980
095
Lower
Conf.
Limit
25.
24.
56.
47.
31.
14.
13.
16.
15.
921
115
160
944
163
273
085
718
132
Upper
Conf.
Limit
38
37
62
54
50
20
29
25
23
.547
.167
.687
.077
.494
.045
.419
.243
.059
p-value
0.6219
<.0001
0.3173
0.2808
E-19
-------
E.6. MEAN FISH AND SHELLFISH CONSUMPTION FOR THE GENERAL
POPULATION, CONSUMERS ONLY, AS-CONSUMED WEIGHT PER
KILOGRAM BODYWEIGHT PER DAY
Tables described in this section (Tables E-19-E-26) present the estimated mean
consumption rate offish and shellfish per kilogram bodyweight (as-consumed g/kg-day) for
those that reported consuming fish and shellfish. The estimates are broken out by state
(Connecticut, Florida, Minnesota, and North Dakota) and categorical level of the independent
variable, in separate tables defined by the independent variables: adult versus child, two
classifications of age and gender combinations, education, income, race-ethnicity, gender, and
type offish consumed. The expected consumption rate for those that eat fish or shellfish
depends in part on the recall period used by the surveys. In Florida the recall period was 7 days.
In Minnesota and North Dakota the recall period was 1 year. In Connecticut the recall period
was roughly a year based on the question wording, but was not specified. As a result, the
calculated consumption rate for fish and shellfish consumers is expected to be higher for Florida
than for the other states.
Table E-19. Mean consumption per bodyweight, general population,
consumers only, by state and adult/child (as-consumed g/kg-day with 95%
CIs)
State
CT
FL
MN
ND
Subgroup
Adult
Child
Unknown
Adult
Child
Adult
Child
Unknown
Adult
Child
Unknown
N
296
57
9
6425
1332
621
171
1
396
149
1
Weighted
N/1000
2195
534
75
6155
1335
3490
1130
1
437
142
1
Mean
0
0
0
0
1
0
0
0
0
0
0
.490
.504
.117
.827
.380
.307
.391
.178
.295
.470
.341
Lower
Conf .
Limit
0
0
-0
0
1
0
0
0
0
.404
.362
.003
.782
.264
.210
.264
.238
.276
m
Upper
Conf.
Limit
0.
0.
0.
0.
1.
0.
0.
•
0.
0.
m
575
647
238
871
495
405
519
352
664
p-value
0.8510
<.0001
0.2862
0.0501
E-20
-------
Table E-20. Mean consumption per bodyweight, general population,
consumers only, by state and age/gender (9 categories) (as-consumed g/kg-
day with 95% CIs)
State Subgroup
CT Child I- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
MN Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
ND Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
N
14
22
18
14
74
70
10
74
57
9
420
375
365
753
1287
1171
754
1334
1192
106
46
42
63
44
127
150
52
115
153
1
28
41
53
38
93
92
36
88
76
1
Weighted
N/1000
131
207
165
113
550
511
84
551
417
75
428
377
364
725
1232
1145
714
1264
1139
102
425
264
313
318
686
810
254
702
850
1
28
39
50
39
99
104
38
95
86
1
Lower
Conf .
Mean
0
0
0
0
0
0
0
0
0
0
2
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.613
.587
.315
.839
.534
.479
.230
.507
.381
.117
.343
.103
.852
.887
.936
.728
.958
.806
.704
.643
.583
.377
.236
.693
.253
.361
.107
.247
.244
.178
.705
.557
.409
.200
.286
.404
.217
.220
.306
.341
Limit
0
0
0
0
0
0
0
0
0
-0
2
0
0
0
0
0
0
0
0
0
0
0
0
-0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.416
.399
.169
.098
.369
.371
.142
.404
.300
.003
.053
.980
.728
.735
.871
.686
.796
.738
.655
.496
.315
.209
.161
.205
.208
.263
.067
.170
.197
.244
.309
.226
.139
.216
.268
.128
.173
.200
Upper
Conf.
Limit
0.
0.
0.
1.
0.
0.
0.
0.
0.
0.
2.
1.
0.
1.
1.
0.
1.
0.
0.
0.
0.
0.
0.
1.
0.
0.
0.
0.
0.
•
1.
0.
0.
0.
0.
0.
0.
0.
0.
.
811
774
461
580
699
587
319
609
462
238
634
225
975
039
001
771
120
875
753
789
852
544
312
590
298
458
147
323
290
166
805
591
260
355
540
306
267
412
p-value
0.0004
<.0001
<.0001
0.0810
E-21
-------
Table E-21. Mean consumption per bodyweight, general population,
consumers only, by state and age/gender (5 categories) (as-consumed g/kg-
day with 95% CIs)
State Subgroup
CT Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
FL Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
MN Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
ND Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
N
53
79
80
10
131
9
1100
1835
1408
782
2526
106
137
139
189
59
268
1
108
112
117
44
164
1
Weighted
N/1000
495
593
589
84
968
75
1112
1764
1368
742
2403
102
947
919
932
270
1552
1
104
117
131
46
181
1
Mean
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.511
.608
.464
.230
.452
.117
.517
.908
.771
.943
.758
.643
.435
.396
.351
.115
.245
.178
.556
.266
.367
.249
.261
.341
Lower
Conf .
Limit
0
0
0
0
0
-0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.377
.397
.368
.142
.375
.003
.385
.825
.720
.787
.714
.496
.282
.070
.267
.075
.199
•
.323
.205
.257
.160
.203
m
Upper
Conf.
Limit
0.
0.
0.
0.
0.
0.
1.
0.
0.
1.
0.
0.
0.
0.
0.
0.
0.
•
0.
0.
0.
0.
0.
m
646
818
561
319
530
238
649
992
823
099
802
789
587
723
435
154
291
790
326
477
339
319
p-value
<.0001
<.0001
<.0001
0.0108
E-22
-------
Table E-22. Mean consumption per bodyweight, general population,
consumers only, by state and education (as-consumed g/kg-day with 95%
CIs)
State Subgroup
CT 0-11 years
High School
Some College
College grad
FL 0-11 years
High School
Some College
College grad
Unknown
MN 0-11 years
High School
Some College
College grad
Unknown
ND 0-11 years
High School
Some College
College grad
Unknown
N
13
76
56
217
613
2405
2511
2190
38
41
219
249
242
42
25
134
174
181
32
Weighted
N/1000
97
569
424
1714
576
2291
2430
2157
37
185
1237
1267
1718
215
28
135
190
190
36
Mean
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
.325
.443
.462
.510
.963
.956
.934
.868
.127
.394
.312
.427
.269
.243
.260
.431
.289
.321
.396
Lower
Conf .
Limit
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
073
331
311
405
805
885
838
815
770
121
175
193
211
178
110
229
210
217
117
Upper
Conf.
Limit
0
0
0
0
1
1
1
0
1
0
0
0
0
0
0
0
0
0
0
.576
.554
.613
.614
.121
.028
.030
.921
.483
.668
.449
.661
.327
.308
.410
.633
.367
.425
.675
p-value
0.5364
0.1936
0.5026
0.5174
E-23
-------
Table E-23. Mean consumption per bodyweight, general population,
consumers only, by state and income (as-consumed g/kg-day with 95% CIs)
State Subgroup N
CT $ 0-20000 35
$20000-50000 133
$50000- 182
Unknown 12
Weighted
N/1000
261
994
1461
87
Mean
0.450
0.540
0.451
0.440
Lower
Conf.
Limit
0.274
0.359
0.360
0.270
Upper
Conf.
Limit
0.626
0.722
0.543
0.609
p-value
0.6951
FL $ 0-20000 1534 1451 1.030 0.914 1.146
$20000-50000 3370 3241 0.953 0.876 1.030
$50000- 1806 1763 0.895 0.827 0.963
Unknown 1047 1035 0.743 0.685 0.801
0.1342
MN $ 0-20000 77 337 0.440 0.227 0.653
$20000-50000 301 1644 0.369 0.185 0.552
$50000- 321 2109 0.294 0.202 0.386
Unknown 94 530 0.263 0.201 0.325
0.4633
ND $ 0-20000 48 50 0.551 0.042 1.060
$20000-50000 221 236 0.289 0.215 0.363
$50000- 225 238 0.316 0.260 0.372
Unknown 52 56 0.448 0.202 0.694
0.4064
E-24
-------
Table E-24. Mean consumption per bodyweight, general population,
consumers only, by state and race-ethnicity (as-consumed g/kg-day with 95%
CIs)
State Subgroup N
CT White, Non-Hispanic 331
Black, Non-Hispanic 3
Hispanic 15
Other 12
Unknown 1
Weighted
N/1000
2562
22
126
85
9
Mean
0.459
0.149
0.681
1.028
0.015
Lower
Conf.
Limit
0.393
0.119
0.308
0.126
Upper
Conf.
Limit
0.525
0.179
1.053
1.930
p-value
<.0001
FL White, Non-Hispanic 5957 5734 0.884 0.842 0.927
Black, Non-Hispanic 785 736 1.107 0.968 1.246
Hispanic 721 742 1.009 0.810 1.209
Other 167 151 1.165 0.775 1.555
Unknown 127 127 0.940 0.697 1.183
0.0025
MN White, Non-Hispanic 732 4194 0.287 0.235 0.339
Hispanic 3 50 0.645 -0.152 1.442
Other 19 173 1.235 -0.590 3.060
Unknown 39 204 0.322 0.179 0.466
0.5923
ND White, Non-Hispanic 501 532 0.341 0.252 0.430
Black, Non-Hispanic 2 2 0.250 0.216 0.284
Other 13 13 0.275 0.181 0.370
Unknown 30 33 0.317 0.097 0.537
0.1532
E-25
-------
Table E-25. Mean consumption per bodyweight, general population,
consumers only, by state and gender (as-consumed g/kg-day with 95% CIs)
State
CT
Subgroup
Male
Female
N
175
187
Weighted
N/1000
1362
1441
Mean
0.447
0.517
Lower
Conf.
Limit
0.367
0.416
Upper
Conf.
Limit
0.527
0.617
p-value
0.1277
FL
Male
Female
Unknown
3880
3861
16
3723
3753
14
0.896
0.955
0.852
0
841
0.896
0.297
0.950
1.013
1.407
0.0623
MN
Male
Female
401
392
2378
2243
0.277
0.381
0.232
0.231
0.323
0.531
0.1627
ND
Male
Female
265
281
282
298
0.332
0.344
0.234
0.264
0.430
0.424
0.7750
Table E-26. Mean consumption per bodyweight, general population,
consumers only, by state and angler status (as-consumed g/kg-day with 95%
CIs)
Lower Upper
Weighted Conf. Conf.
State Subgroup N N/1000 Mean Limit Limit p-value
CT Eats Caught Only 1 9 0.015 . . <.0001
Eats Caught&Bought 70 530 0.486 0.360 0.612
Eats Bought Only 291 2265 0.484 0.399 0.568
FL Eats Caught Only 511 454 0.761 0.661 0.860 <.0001
Eats Caught&Bought 701 636 1.811 1.567 2.056
Eats Bought Only 6545 6400 0.849 0.806 0.892
MN Eats Caught Only 38 221 0.156 0.053 0.259 0.0103
Eats Caught&Bought 555 2746 0.399 0.274 0.524
Eats Bought Only 200 1653 0.232 0.184 0.280
ND Eats Caught Only 30 32 0.208 0.094 0.322 0.0312
Eats Caught&Bought 359 384 0.388 0.288 0.488
Eats Bought Only 157 164 0.247 0.133 0.362
E-26
-------
E.7. GEOMETRIC MEAN FISH AND SHELLFISH CONSUMPTION FOR THE
GENERAL POPULATION, CONSUMERS ONLY, AS-CONSUMED WEIGHT
PER DAY
Tables described in this section (Tables E-27-E-33) present the estimated geometric
mean consumption rate offish and shellfish (as-consumed g/day) for those that reported
consuming fish and shellfish. The estimates are broken out by state (Connecticut, Florida,
Minnesota, and North Dakota) and categorical level of the independent variable, in separate
tables defined by the independent variables: adult versus child, two classifications of age and
gender combinations, education, income, race-ethnicity, gender, and type offish consumed. The
expected consumption rate for those that eat fish or shellfish depends in part on the recall period
used by the surveys. In Florida the recall period was 7 days. In Minnesota and North Dakota the
recall period was 1 year. In Connecticut the recall period was roughly a year based on the
question wording, but was not specified. As a result, the calculated consumption rate for fish
and shellfish consumers is expected to be higher for Florida than for the other states.
Table E-27. Geometric mean consumption, general population, consumers
only, by state and adult/child (as-consumed g/day with 95% CIs)
State
CT
FL
MN
ND
Subgroup
Adult
Child
Unknown
Adult
Child
Adult
Child
Unknown
Adult
Child
Unknown
N
302
58
9
7131
1435
623
172
1
410
155
5
Weighted
N/1000
2237
542
75
6473
1438
3492
1131
1
453
147
6
Geometric
Mean
21
9
3
37
25
12
6
1
12
8
24
.274
.839
.143
.780
.093
.789
.239
.047
.572
.939
.819
Lower
Conf .
Limit
18.
7.
1.
36.
23.
10.
4.
•
10.
6.
10.
248
121
318
517
459
741
838
762
837
163
Upper
Conf.
Limit
24
13
7
39
26
15
8
14
11
60
.803
.593
.492
.087
.840
.229
.045
.687
.688
.612
p-value
<.0001
<.0001
<.0001
0.0101
E-27
-------
Table E-28. Geometric mean consumption, general population, consumers
only, by state and age/gender (9 categories) (as-consumed g/day with 95%
CIs)
State Subgroup
CT Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
MN Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
N
14
22
19
14
77
72
10
75
57
9
421
376
365
791
1446
1315
785
1406
1272
389
46
43
63
44
128
150
52
115
154
Weighted
N/1000
131
207
173
113
571
526
84
557
417
75
428
378
364
739
1292
1192
719
1275
1156
369
425
265
313
318
686
810
254
702
851
Geometric
Mean
6
10
10
24
17
18
14
28
22
3
22
23
29
33
38
33
40
42
38
30
5
5
8
10
11
14
5
15
14
.635
.440
.828
.849
.869
.294
.888
.423
.789
.143
.043
.165
.486
.011
.500
.672
.315
.527
.282
.154
.104
.743
.324
.056
.674
.195
.361
.045
.156
Lower
Conf .
Limit
4.
7.
6.
13.
13.
13.
10.
22.
18.
1.
20.
20.
26.
30.
36.
32.
37.
40.
36.
26.
2.
3.
5.
6.
9.
11.
3.
11.
11.
183
164
772
213
360
705
779
689
326
318
157
748
519
565
550
098
066
339
021
559
860
579
326
363
449
003
203
805
565
Upper
Conf.
Limit p-value
10
15
17
46
23
24
20
35
28
7
24
25
32
35
40
35
43
44
40
34
9
9
13
15
14
18
8
19
17
.522 <.0001
.215
.314
.733
.898
.420
.563
.607
.338
.492
.106 <.0001
.863
.785
.651
.553
.325
.848
.833
.684
.235
.107 <.0001
.216
.010
.891
.423
.314
.972
.174
.327
1.047
ND Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
28
43
56
39
97
95
37
90
80
5
28
41
53
40
103
108
39
97
91
6
4
9
12
7
11
13
9
13
14
24
.185
.873
.564
.880
.594
.748
.847
.316
.294
.819
2.
7.
9.
5.
9.
10.
6.
10.
10.
10.
456
142
044
798
333
534
901
811
613
163
7
13
17
10
14
17
14
16
19
60
.133
.649
.453
.709
.402
.943
.051
.402
.251
.612
0.0004
E-28
-------
Table E-29. Geometric mean consumption, general population, consumers
only, by state and education (as-consumed g/day with 95% CIs)
Weighted Geometric
State Subgroup N N/1000 Mean
CT 0-11 years 13 97 13.944
High School 78 584 15.425
Some College 60 450 23.309
College grad 218 1723 17.121
FL 0-11 years 716 620 37.689
High School 2683 2419 35.233
Some College 2739 2550 34.348
College grad 2376 2274 34.758
Unknown 52 48 50.745
Lower
Conf.
Limit
6.!
10.726
16.390
14.162
33.549
33.416
32.680
33.126
38.708
Upper
Conf.
Limit
28.260
22.184
33.148
20.698
42.340
37.149
36.101
36.470
66.524
p-value
0.3066
0.3940
MN 0-11 years 41 185 16.164 9.938 26.292
High School 220 1237 11.554 8.802 15.166
Some College 250 1268 11.066 8.051 15.209
College grad 243 1719 9.237 6.913 12.342
Unknown 42 215 13.472 9.926 18.286
0.2394
ND 0-11 years 27 31 10.797 7.334 15.894
High School 139 141 13.564 10.147 18.132
Some College 183 199 11.188 8.655 14.461
College grad 189 199 11.001 8.093 14.955
Unknown 32 36 11.676 7.019 19.424
0.5899
E-29
-------
Table E-30. Geometric mean consumption, general population, consumers
only, by state and income (as-consumed g/day with 95% CIs)
State Subgroup N
CT $ 0-20000 36
$20000-50000 135
$50000- 186
Unknown 12
Weighted
N/1000
270
1006
1490
87
Geometric
Mean
16.716
18.587
16.538
25.127
Lower
Conf.
Limit
11.067
13.904
13.017
16.802
Upper
Conf.
Limit
25.250
24.847
21.012
37.579
p-value
0.8365
FL $ 0-20000 1707 1537 35.746 33.297 38.375
$20000-50000 3709 3404 35.950 34.169 37.824
$50000- 1960 1844 34.825 32.981 36.772
Unknown 1190 1126 32.078 29.831 34.494
0.6722
MN $ 0-20000 79 339 19.157 13.388 27.412
$20000-50000 302 1645 10.490 7.792 14.122
$50000- 321 2109 9.880 7.578 12.880
Unknown 94 530 10.983 8.077 14.935
0.0118
ND $ 0-20000 50 53 13.705 8.438 22.258
$20000-50000 235 250 10.038 8.137 12.383
$50000- 231 244 12.933 10.719 15.604
Unknown 54 59 12.235 8.071 18.547
0.0517
E-30
-------
Table E-31. Geometric mean consumption, general population, consumers
only, by state and race-ethnicity (as-consumed g/day with 95% CIs)
State Subgroup
CT White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
FL White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
MN White, Non-Hispanic
Hispanic
Other
Unknown
ND White, Non-Hispanic
Black, Non-Hispanic
Other
Unknown
Lower Upper
Weighted Geometric Conf. Conf.
N N/1000 Mean Limit Limit
p-value
338
3
15
12
1
607
867
762
191
139
735
3
19
39
521
2
17
30
2612
22
126
85
9
6053
780
773
166
139
4197
50
173
204
555
2
16
33
17.458
12.500
20.972
20.267
0.985
34.492
38.820
34.600
42.005
35.615
10.263
30.013
13.301
17.168
11.576
15.709
13.933
11.568
14.907 20.446 <.0001
8.391 18.619
11.899 36.966
9.586 42.848
33.325 35.701 0.0082
35.598 42.334
31.337 38.203
35.562 49.615
29.410 43.129
8.596 12.253 0.0319
9.260 97.272
5.454 32.436
11.802 24.974
9.775 13.709 0.0036
15.032 16.417
9.008 21.552
6.589 20.308
Table E-32. Geometric mean consumption, general population, consumers
only, by state and gender (as-consumed g/day with 95% CIs)
State
CT
FL
MN
ND
Subgroup
Male
Female
Male
Female
Unknown
Male
Female
Male
Female
N
177
192
4066
4206
294
403
393
277
293
Weighted
N/1000
1377
1478
3758
3874
279
2379
2244
295
311
Geometric
Mean
19
16
37
33
29
10
10
12
10
.202
.002
.609
.187
.474
.494
.974
.826
.624
Lower
Conf.
Limit
15.
13.
36.
31.
25.
8.
8.
10.
9.
832
392
250
872
206
801
996
706
129
Upper
Conf.
Limit
23
19
39
34
34
12
13
15
12
.291
.120
.018
.556
.465
.513
.389
.367
.363
p-value
0.0771
<.0001
0.6470
0.0103
E-31
-------
Table E-33. Geometric mean consumption, general population, consumers
only, by state and angler status (as-consumed g/day with 95% CIs)
State Subgroup
CT Eats Caught Only
Eats Caught&Bought
Eats Bought Only
FL Eats Caught Only
Eats Caught&Bought
Eats Bought Only
MN Eats Caught Only
Eats Caught&Bought
Eats Bought Only
ND Eats Caught Only
Eats Caught&Bought
Eats Bought Only
Lower
Weighted Geometric Conf.
N N/1000 Mean Limit
1
74
294
9
559
2286
0.985
23.413
16.453
600
802
164
38
556
202
33
376
161
493
667
6752
221
2747
1655
36
403
167
30.
77.
32.
3.
14.
7.
8.
15.
6.
625
649
743
996
987
021
437
345
414
17.
13.853
Upper
Conf.
Limit
30.646
19.540
p-value
<.0001
27.371 34.267 <.0001
72.359 83.326
31.584 33.946
1.759 9.077 <.0001
12.827 17.511
5.267 9.359
5.757 12.364 <.0001
13.176 17.872
4.784 8.599
E.8. GEOMETRIC MEAN FISH AND SHELLFISH CONSUMPTION FOR THE
GENERAL POPULATION, CONSUMERS ONLY, AS-CONSUMED WEIGHT
PER KILOGRAM BODYWEIGHT PER DAY
Tables described in this section (Tables E-34-E-41) present the estimated geometric
mean consumption rate offish and shellfish per kilogram bodyweight (as-consumed g/kg-day)
for those that reported consuming fish and shellfish. The estimates are broken out by state
(Connecticut, Florida, Minnesota, and North Dakota) and categorical level of the independent
variable, in separate tables defined by the independent variables: adult versus child, two
classifications of age and gender combinations, education, income, race-ethnicity, gender, and
type offish consumed. The expected consumption rate for those that eat fish or shellfish
depends in part on the recall period used by the surveys. In Florida the recall period was 7 days.
In Minnesota and North Dakota the recall period was 1 year. In Connecticut the recall period
was roughly a year based on the question wording, but was not specified. As a result, the
calculated consumption rate for those that consumed fish or shellfish is expected to be higher for
Florida than for the other states.
E-32
-------
Table E-34. Geometric mean consumption per bodyweight, general
population, consumers only, by state and adult/child (as-consumed g/kg-day
with 95% CIs)
State
CT
Subgroup
Adult
Child
Unknown
N
296
57
9
Weighted
N/1000
2195
534
75
Geometric
Mean
0.292
0.348
0.047
Lower
Conf.
Limit
0.250
0.244
0.024
Upper
Conf.
Limit
0.340
0.496
0.092
p-value
0.3222
FL
Adult
Child
6425
1332
6155
1335
0.530
0.870
0.513
0.810
0.547
0.935
<.0001
MN
Adult
Child
Unknown
621
171
1
3490
1130
1
0.168
0.211
0.178
0.141
0.157
0.200
0.284
0.1801
ND
Adult
Child
Unknown
396
149
1
437
142
1
0.169
0.253
0.341
0.144
0.198
0.198
0.323
0.0013
E-33
-------
Table E-35. Geometric mean consumption per bodyweight, general
population, consumers only, by State and age/gender (9 categories) (as-
consumed g/kg-day with 95% CIs)
State Subgroup
CT Child I- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
MN Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
ND Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
N
14
22
18
14
74
70
10
74
57
9
420
375
365
753
1287
1171
754
1334
1192
106
46
42
63
44
127
150
52
115
153
1
28
41
53
38
93
92
36
88
76
1
Weighted
N/1000
131
207
165
113
550
511
84
551
417
75
428
377
364
725
1232
1145
714
1264
1139
102
425
264
313
318
686
810
254
702
850
1
28
39
50
39
99
104
38
95
86
1
Geometric
Mean
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.450
.423
.221
.437
.281
.279
.191
.333
.270
.047
.625
.808
.594
.542
.601
.517
.530
.519
.475
.472
.344
.202
.153
.158
.177
.205
.063
.171
.166
.178
.272
.353
.243
.135
.183
.203
.129
.154
.166
.341
Lower
Conf .
Limit
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.288
.283
.139
.226
.209
.209
.130
.267
.216
.024
.453
.721
.536
.502
.570
.492
.488
.492
.448
.394
.186
.126
.098
.097
.142
.158
.035
.133
.134
.161
.256
.173
.099
.146
.153
.089
.125
.121
m
Upper
Conf.
Limit
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.703
.633
.351
.845
.377
.372
.280
.415
.337
.092
.818
.905
.657
.585
.634
.544
.575
.547
.503
.565
.638
.323
.238
.258
.222
.266
.113
.219
.205
.461
.487
.342
.184
.229
.270
.186
.191
.228
m
p-value
0.0233
<.0001
0.0101
0.0001
E-34
-------
Table E-36. Geometric mean consumption per bodyweight, general
population, consumers only, by state and age/gender (5 categories) (as-
consumed g/kg-day with 95% CIs)
State Subgroup
CT Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
FL Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
MN Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
ND Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
N
53
79
80
10
131
9
1100
1835
1408
782
2526
106
137
139
189
59
268
1
108
112
117
44
164
1
Weighted
N/1000
495
593
589
84
968
75
1112
1764
1368
742
2403
102
947
919
932
270
1552
1
104
117
131
46
181
1
Geometric
Mean
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.357
.318
.263
.191
.304
.047
.974
.576
.533
.525
.498
.472
.240
.164
.204
.068
.168
.178
.301
.172
.189
.139
.160
.341
Lower
Conf .
Limit
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.254
.240
.199
.130
.258
.024
.902
.547
.506
.485
.475
.394
.165
.131
.162
.038
.142
•
.225
.140
.150
.101
.131
m
Upper
Conf.
Limit
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.501
.421
.349
.280
.359
.092
.053
.606
.562
.569
.521
.565
.350
.206
.257
.120
.199
•
.404
.210
.240
.191
.195
m
p-value
0.1133
<.0001
0.0017
<.0001
E-35
-------
Table E-37. Geometric mean consumption per bodyweight, general
population, consumers only, by state and education (as-consumed g/kg-day
with 95% CIs)
State Subgroup
CT 0-11 years
High School
Some College
College grad
FL 0-11 years
High School
Some College
College grad
Unknown
MN 0-11 years
High School
Some College
College grad
Unknown
ND 0-11 years
High School
Some College
College grad
Unknown
N
13
76
56
217
613
2405
2511
2190
38
41
219
249
242
42
25
134
174
181
32
Weighted
N/1000
97
569
424
1714
576
2291
2430
2157
37
185
1237
1267
1718
215
28
135
190
190
36
Geometric
Mean
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.175
.259
.313
.299
.611
.585
.574
.566
.852
.229
.173
.193
.164
.189
.160
.225
.176
.181
.172
Lower
Conf .
Limit
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.087
.182
.227
.238
.545
.551
.543
.539
.621
.142
.124
.142
.129
.134
.102
.165
.135
.138
.091
Upper
Conf.
Limit
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
.349
.369
.431
.377
.685
.621
.607
.594
.168
.369
.241
.263
.209
.268
.251
.307
.230
.237
.325
p-value
0.4268
0.6419
0.6016
0.4063
E-36
-------
Table E-38. Geometric mean consumption per bodyweight, general
population, consumers only, by state and income (as-consumed g/kg-day with
95% CIs)
State Subgroup N
CT $ 0-20000 35
$20000-50000 133
$50000- 182
Unknown 12
Weighted
N/1000
261
994
1461
87
Geometric
Mean
0.264
0.296
0.283
0.347
Lower
Conf.
Limit
0.166
0.218
0.217
0.221
Upper
Conf.
Limit
0.418
0.401
0.368
0.547
p-value
0.9251
FL $ 0-20000 1534 1451
$20000-50000 3370 3241
$50000- 1806 1763
Unknown 1047 1035
0.608
0.596
0.569
0.509
0.562
0.563
0.533
0.470
0.657
0.630
0.607
0.551
0.3966
MN $ 0-20000 77 337
$20000-50000 301 1644
$50000- 321 2109
Unknown 94 530
0.243
0.183
0.163
0.186
0.166
0.140
0.126
0.137
0.356
0.240
0.211
0.254
0.2493
ND $ 0-20000 48 50
$20000-50000 221 236
$50000- 225 238
Unknown 52 56
0.213
0.157
0.212
0.201
0.114
0.126
0.175
0.124
0.399
0.196
0.257
0.325
0.0563
E-37
-------
Table E-39. Geometric mean consumption per bodyweight, general
population, consumers only, by state and race-ethnicity (as-consumed g/kg-
day with 95% CIs)
State Subgroup
Lower Upper
Weighted Geometric Conf. Conf.
N N/1000 Mean Limit Limit p-value
CT White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Other
Unknown
FL White, Non-Hispanic 5957
Black, Non-Hispanic
Hispanic
Other
Unknown
MN White, Non-Hispanic
Hispanic
Other
Unknown
ND White, Non-Hispanic
Black, Non-Hispanic
Other
Unknown
331
3
15
12
1
957
785
721
167
127
732
3
19
39
501
2
13
30
2562
22
126
85
9
5734
736
742
151
127
4194
50
173
204
532
2
13
33
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
283
147
393
472
015
559
712
592
691
607
168
421
376
240
186
249
202
186
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.236
.118
.217
.177
•
.541
.644
.530
.584
.490
.139
.127
.159
.160
.156
.217
.126
.097
0
0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
.339
.182
.710
.255
•
.578
.787
.660
.816
.752
.202
.400
.888
.361
.223
.284
.324
.357
<.0001
<.0001
0.1082
0.0204
E-38
-------
Table E-40. Geometric mean consumption per bodyweight, general
population, consumers only, by state and gender (as-consumed g/kg-day with
95% CIs)
State
CT
Subgroup
Male
Female
N
175
187
Weighted
N/1000
1362
1441
Geometric
Mean
0.286
0.289
Lower
Conf.
Limit
0.231
0.237
Upper
Conf.
Limit
0.354
0.352
p-value
0.9195
FL
Male
Female
Unknown
3880
3861
16
3723
3753
14
0.555
0.604
0.503
0.534
0.578
0.224
0.577
0.631
1.128
MN
Male
Female
401
392
2378
2243
0.164
0.194
0.140
0.160
0.192
0.233
0.0632
ND
Male
Female
265
281
282
298
0.184
0.189
0.153
0.162
0.222
0.222
0.7060
E-39
-------
Table E-41. Geometric mean consumption per bodyweight, general
population, consumers only, by state and angler status (as-consumed g/kg-
day with 95% CIs)
State Subgroup
CT Eats Caught Only
Eats Caught&Bought
Eats Bought Only
FL Eats Caught Only
Eats Caught&Bought
Eats Bought Only
MN Eats Caught Only
Eats Caught&Bought
Eats Bought Only
ND Eats Caught Only
Eats Caught&Bought
Eats Bought Only
ND Eats Caught Only
Eats Caught&Bought
Eats Bought Only
Lower Upper
Weighted Geometric Conf. Conf.
N N/1000 Mean Limit Limit p-value
1
70
291
511
701
6545
38
555
200
30
359
157
30
359
157
9
530
2265
454
636
6400
221
2746
1653
32
384
164
.
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
.015
.330
.282
.501
.249
.542
.086
.236
.122
.132
.242
.110
.132
.242
.110
0
0
0
1
0
0
0
0
0
0
0
0
0
0
.
.252
.229
.443
.155
.521
.034
.201
.098
.086
.207
.080
.086
.207
.080
0
0
0
1
0
0
0
0
0
0
0
0
0
0
.
.430
.345
.566
.351
.563
.214
.278
.151
.201
.283
.150
.201
.283
.150
<.0001
<.0001
<.0001
<.0001
<.0001
E.9. PLOTS OF CONFIDENCE INTERVALS FOR THE GENERAL POPULATION
Figures E-l through E-20 show confidence intervals for selected estimates for each
state's general population. There are separate figures for each combination of state and
independent variable (per capita consumption, per capita consumption per bodyweight, percent
of respondents eating fish or shellfish, consumption for consumers only, and consumption per
bodyweight for consumers only).
Each figure shows the independent variable on the vertical axis and the categorical
dependent variables on the horizontal axis. For each level of each variable, the figure shows the
mean as a dot and the 95% confidence interval using a line going vertically through the dot. In a
few cases there was only one observation in a category and no confidence interval could be
calculated. These cases are indicated by a dot with an open center.
The plots using the same summary statistic all have the same range on the vertical axis to
facilitate rough comparisons among states. However, the expected percentage of respondents
that report eating fish or shellfish during the recall period and the consumption rate for those that
eat fish or shellfish depends in part on the length of the recall period used by the surveys. In
E-40
-------
Florida the recall period was 7 days. In Minnesota and North Dakota the recall period was 1
year. In Connecticut the recall period was roughly a year based on the question wording, but
was not specified. As a result, in Florida the percentage of respondents that consume fish and
shellfish is expected to be lower and the consumption rate for them is expected to be higher than
for the other states.
The confidence intervals give the reader a visual indication of the precision of the means
and facilitate comparisons of the means between levels of the variable. The values plotted can
be found in the tables in Sections E.2 through E.6 along with ap-va\ue for evaluating the
significance of differences between levels. In general, longer confidence intervals are associated
with more variable intake data or fewer respondents. In a few cases there is only one respondent
for a level. In this case, an open dot is used and no confidence interval is shown. The
confidence interval calculations assume the mean estimates have a normal distribution. If the
sample size (N) is smaller than about 100, this assumption is uncertain and the confidence
interval and/?-value are approximate. If there are relatively few respondents in a subgroup
category, the confidence intervals can be imprecise. Results for these categories are shown in
the plots for completeness, although the results may not be useful due to the small sample sizes.
In a few cases with small sample sizes the normality assumption is not correct and the calculated
lower confidence interval for mean consumption is <0 or the confidence limits for the percentage
of respondents consuming fish and shellfish go outside the range from 0% to 100%. In these
cases, the line connecting the confidence limits may be truncated at the edge of the plot.
Figures E-l through E-5 show means and confidence intervals for Connecticut. Figures
E-6 through E-10 show means and confidence intervals for Florida. Figures E-l 1 through E-15
show means and confidence intervals for Minnesota. Figures E-l6 through E-20 show means
and confidence intervals for North Dakota.
E-41
-------
W
to
140
120
100
«s
"S
80
•< 60
a
o
••a
cs
t
o
a.
40
20
i
,
ii
Adult/Child
Age/Gender
o
Education
Income
Race-Ethnicity
Gender
Figure E-l. Mean consumption, Connecticut general population, per capita, with 95% confidence intervals.
-------
2.5
1.5
w
s
c.
£
0.5
II
Adult/Child
o
t
Age/Gender
Education
Income
ffi
ffi
Race-Ethnicity
Gender
Figure E-2. Mean consumption per bodyweight, Connecticut general population, per capita, with 95%
confidence intervals.
-------
w
1UU/0 •
90% •
ono/ .
x 70% •
1
33 60% -
•a
a
ss
u.
8
•-C
« 40% •
^ 30% •
ft.
20% -
no/
T
i
4
4
•L
••
^ s i
•o J3 o
cj
g
0}
ft
CTs
0
a
->
g
a>
ft
O
c
3
ft
CTs
(N
u
1
0^
0
a
c
^
f£
}
3
H
>•
+
O
in
a>
cj
S
Age/
nown
^4
6
4>
1 4
f I
II
|
•x
f
T3
O
•x
•x
03
u
ft
^r
^i-
in
g
03
S
+
in
^r
a>
03
S
Gender
1 1
1 ill1 ill!
"c g^QM) C^lln-^
S 0 CJ MO
M s
Education Income
_o
03
OH
-H
S
c
0
u
1^
^
<
»
1
4
4
<: •
-'-
0
_o
'g
03
OH
.2
1
0
^
^
03
m
•g I s i
03 t; o "5
OH O C ^
a
Race-Ethnicity G(
> _a>
3 cj
3 ^
ft
snder
Figure E-3. Percent eating fish and shellfish during a year period, Connecticut general population, per capita,
with 95% confidence intervals.
-------
o
•a
g 60
V
40
20
w
11
A
C
Adult/Child
O
o
O
Age/Gender
Education
S
Income
ffi
ffi
o
3
Race-Ethnicity
00
O
J3
00
O
m
Caught Fish
Jendei
Figure E-4. Mean consumption, Connecticut general population, consumers only, with 95% confidence
intervals.
-------
2.5
cs
Q
O
1.5
W
0.5
H
il
T
Adult/Child
Age/Gender
Education
Income
Race-Ethnicity
Caught Fish
jende:
Figure E-5. Mean consumption per bodyweight, Connecticut general population, consumers only, with 95%
confidence intervals.
-------
W
140
120
100
«s
"S
80
•< 60
a
o
••a
cs
t
(£
40
20
T
HIT
\dult/Chilc
Age/Gender
Education
Income
m
Race-Ethnicity
Gender
Figure E-6. Mean consumption, Florida general population, per capita, with 95% confidence intervals.
-------
2.5
«S
~
W
oo
§.
o
a.
•f
A.dult/Child
Age/Gender
Education
Income
ffi
Race-Ethnicity
Gender
Figure E-7. Mean consumption per bodyweight, Florida general population, per capita, with 95% confidence
intervals.
-------
W
-U
vo
100%
90%
80%
70%
K 60%
•a
a
SJ
* 50%
u.
8
1 40%
30%
10%
f 1
T T
r
ii -0
3 r3
-0 J3
-------
120
100
cs
Q 80
O
•a
-con
c^
O
40
20
1
1—f
i
\
F1".!'
w
I
o
^.dult/Chik
Age/Gender
Education
S
Income
m
Race-Ethnicity
•§)
w
Q
W
Caught Fish
§
Gender
Figure E-9. Mean consumption, Florida general population, consumers only, with 95% confidence intervals.
-------
w
Figure E-10. Mean consumption per bodyweight, Florida general population, consumers only, with 95%
confidence intervals.
-------
140
120
100
«s
"S 80
1 J
* J { I
\ \ 1
'iT I I
T H 4' 1 J '' T ° T
rl|i ii T^^f
{ *il 1
0 "Q ^ C
G ^ ^ PH ^ K o U
PH PH PH C/3
Adult/Child Age/Gender Education
1 1
1 ''T
§ 1 § i
8 § § §
6 o &3 ^
8 ^
65 O
Income
.0
rt
OH
c/5
S
c
0
-S"
,0
rt
OH
c/5
S
c
0
o
rt
5
Rac
4
_o
'S
rt
OH
c/5
3
e-E
T
< >
1
f^ ^ a
6 § >
1
hnicity G(
> J£
3 cj
H ^H
' 5
ft
sndei
60
40
20
a
o
••a
cs
I
(£
w
to
Figure E-ll. Mean consumption, Minnesota general population, per capita, with 95% confidence intervals.
-------
2.5
«S
~M
&
$ 1-5
•a
§.
o
a.
w
0.5
\i
J
i
O
Adult/Child
O
O
Age/Gender
Education
U
Income
Race-Ethnicity
Gender
Figure E-12. Mean consumption per bodyweight, Minnesota general population, per capita, with 95%
confidence intervals.
-------
w
1UU/0 •
90% •
ono/ .
_i 7O.O/ .
33 60% -
•a
a
ss
u.
8
•C
« 40% •
"3
s« ^n% -
0.
ono/ .
no/
J
••
1 FTTT rr * j r " T * T j
1 i <. 1 I <. •• i i 4. "• T
4 l """ """ """
1 .1
1
j j j|
•ass
•a -a o
O ^/5 G '^-M ^ h
O|D ^T i h
MO 0 ,0
o" -^
^ 03
^ m
Age/Gender Education Income Race
III -
a 6 § S
« a
-Ethnicity G
U JD
•3 cj
3 ^
ft
endei
Figure E-13. Percent eating fish and shellfish during a roughly year-long period, Minnesota general population,
per capita, with 95% confidence intervals.
-------
w
1 9O, .
i no,
o
•a
a 60 -
O
V
VI
on .
*. h
"3 — i
Adult/Child
1
{
1
^
1
T
T n
1 * M*' }i, '{
ggftSiSi gftSi g
ft ft ft
Age/Gender EC
"^3 C O ^
00 rt < O O
^ 00 0 § §
u ^ a ° §
!/J <^
ucation I
nrh
§ $50000-
0 Unknown
o
cj
OH
M
S
c
0
-S"
Ra
4
o
'S
S
ce-
( (
1 i '
u S -^ S ^>
6 1 HI
£ 00 Q^ 00
03 £ O
O 00 pq
c/5 ^ c/5
03 Q 03
w » w
03
w
Ethnicity Caught Fish
r
rt rt
ft
3ende:
Figure E-14. Mean consumption, Minnesota general population, consumers only, with 95% confidence
intervals.
-------
W
2.5
O
1.5
0.5
u
O
Adult/Child
O
O
O
n
n
Age/Gender
H
ft
o
Education
Income
Race-Ethnicity
'
o
w
J3
°O
o
W
w
Caught Fish
jende:
Figure E-15. Mean consumption per bodyweight, Minnesota general population, consumers only, with 95%
confidence intervals.
-------
w
140
120
100
«s
"S
a
o
••a
cs
t
(£
80
60
40
20
m
-------
«s
~M
M
<3 1.5
•a
§.
o
a.
w
oo
0.5
Adult/Child
f
• T
| H \ *
1 !
i M
1.1
Age/Gender
ffi
o
Education
Income
o
K
m
Race-Ethnicity
Gender
Figure E-17. Mean consumption per bodyweight, North Dakota general population, per capita, with 95%
confidence intervals.
-------
w
lUUyo '
ono/ .
_, 7no/ .
"O
a
A
JS
E
a
!0
UJ
(J
1 30/0 '
ono/ .
no/
i
1
1
<
4
T3
Ac
O
lull/
P Unknown
. <
Tj
4
i
1
i
i
1
<
»
1
1
0
i
ON
(N
ft
ON
0
03
1
ft
o
ON
(N
03
0
cj
Ag
^ Male 50+
o
c
M
a
ender
4
1 3J p
i ±
<• i
• ll
1
^r^f + ^f+c ^-^u-rsc oo
uft Sou MO
M ««
Education Ir
6 | .2 .2
0 ? C C
O O 03 03
m -^ .2-2
M a ^ ^
II
QJ ^^
IS rt
^ 5
come Race-
^ |> o3 rt
a
Ethnicity Gender
Figure E-18. Percent eating fish and shellfish during a roughly year-long period, North Dakota general
population, per capita, with 95% confidence intervals.
-------
i nn
SJ
O
•a
a 60 -
O
V
VI
Af\ .
W
1
O
on .
1 1
I
• T
*l| (Hi*
J 2 I ^ 3 2 3 2
<; u a ^ ^ ^ m
^ T3 -^3
'-' u u
Adult/Child
Inl' i hi" ll-
^l||Ht}I||i:
D ft
ft c
Age/Gender Edi
.
r T «•
n ^
1
J T3 C O ^
50 03 J O O
| ™ § 88
J I1 "c 2 2
2 — D o
= ° *£, °
jcation Ir
d J 1
i { i.l
6 | .a .a fc
O ? C C js
O O 03 03 X
0 C tt tt O
m -^ 2-2
M ^ ffi ffi
P c c
0 ,0
_u" j^
^ 3
iconic Race-Eth
I i
J T
l|i'i "
S ^ £ ^
? c oo c
§ 2 o 9
-^ £ m £
JD g1 =a g1
03 ^ O
o oo pq
c/5 ^ c/5
03 rj 03
w » w
03
w
nicity Caught Fish
03 03
ft
3ende:
Figure E-19. Mean consumption, North Dakota general population, consumers only, with 95% confidence
intervals.
-------
2.5
cs
Q
O
1.5
W
ON
0.5
F?
Adult/Child
Age/Gender
Education
Income
Race-Ethnicity
-a
o
w
Caught Fish
jende:
Figure E-20. Mean consumption per bodyweight, North Dakota general population, consumers only, with 95%
confidence intervals.
-------
E.10. FISH AND SHELLFISH CONSUMPTION FOR TARGETED POPULATIONS
Tables E-42 through E-45 summarize fish and shellfish consumption for the general
population and for targeted populations within each state, tabulating values for the following
summary statistics:
• Mean consumption offish and shellfish per capita (as-consumed g/day);
• Mean consumption per kilogram bodyweight per capita (as-consumed g/kg-day);
• Percentage of respondents that reported eating fish and shellfish;
• Mean consumption for those than consume fish and shellfish (as-consumed g/day);
• Mean consumption per kilogram bodyweight for those than consume fish and shellfish
(as-consumed g/kg-day);
• Geometric mean consumption for those than consume fish and shellfish (as-consumed
g/day);
• Geometric mean consumption per kilogram bodyweight for those than consume fish and
shellfish (as-consumed g/kg-day);
The expected percentage of respondents that report eating fish or shellfish during the
recall period and the consumption rate for those that eat fish or shellfish depends in part on the
length of the recall period used by the surveys. In Florida the recall period was 7 days. In
Minnesota and North Dakota the recall period was 1 year. In Connecticut the recall period was
roughly a year based on the question wording, but was not specified. As a result, for Florida the
percentage of respondents that consume fish and shellfish is expected to be lower and the
consumption rate for them is expected to be higher than for the other states.
The tables show the state, the targeted population, the number of respondents (N), the
weighted population size (in thousands) represented by the respondents in the general population
(Weighted N/1,000), the mean (either the arithmetic or geometric mean fish and shellfish
consumption rate or the percentage consuming fish and shellfish), and the upper and lower 95%
confidence limit. Because the respondents for the general population in each state were selected
using a stratified random sample of all households in the state, the calculations for the general
population are weighted to estimate the mean for the each states' general population.
Unweighted means are calculated for the targeted populations because either, (1) there is
inadequate information to determine appropriate weights, or (2) the respondents were selected
E-62
-------
using a simple random sample, in which case weights (if they could be calculated) would not
make a difference.
The tables include 95% confidence intervals. In general longer confidence intervals are
associated with more variable intake data or fewer respondents. The confidence interval
calculations assume the mean estimates have a normal distribution. If the sample size (N) is
smaller than about 100, this assumption is uncertain and the confidence interval and/>-value are
approximate. If there are relatively few respondents in a subgroup category, the confidence
intervals can be imprecise.
Table E-42. Mean consumption per bodyweight, per capita, by state and
subpopulation (as-consumed g/kg-day with 95% CIs)
State Subpopulation
CT General
Angler
Aquaculture Student
Asians
Commercial Fisherme
EFNEP Participants
WIC Participants
FL General
MN General
American Indians
Anglers
New Mothers
ND General
American Indians
Anglers
Weighted
N N/1000
420 3296
250
25
396
173
67
699
15367 14827
837 4897
216
1152
401
575 610
106
854
Mean
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
.410
.640
.224
.153
.650
.995
.801
.467
.309
.212
.306
.325
.322
.352
.323
Lower
Conf .
Limit
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.335
.532
.035
.974
.506
.253
.655
.440
.232
.147
.254
.239
.248
.089
.280
Upper
Conf.
Limit
0
0
0
1
0
1
0
0
0
0
0
0
0
0
0
.486
.747
.414
.331
.795
.738
.947
.495
.387
.276
.358
.410
.396
.615
.365
E-63
-------
Table E-43. Mean consumption per bodyweight, consumers only, by state
and subpopulation (as-consumed g/kg-day with 95% CIs)
State Subpopulation N
CT General 362
Angler 244
Aquaculture Student 19
Asians 393
Commercial Fisherme 166
EFNEP Participants 58
WIC Participants 553
FL General 7757
MN General 793
American Indians 192
Anglers 1109
New Mothers 341
Weighted
N/1000
2804
7490
4621
Mean
0.483
0.655
0.295
1.161
0.678
1.149
1.013
0.925
0.328
0.238
0.317
0.382
Lower
Conf.
Limit
0.404
0.548
0.067
0.981
0.532
0.309
0.839
0.878
0.247
0.167
0.264
0.286
Upper
Conf.
Limit
0.561
0.763
0.524
1.342
0.824
1.989
1.186
0.972
0.409
0.309
0.370
0.477
ND General
American Indians
Anglers
546
64
580
0.338
0.583
0.341
0.260
0.192
0.297
0.416
0.974
0.385
E-64
-------
Table E-44. Geometric mean consumption, consumers only, by state and
subpopulation (as-consumed g/day with 95% CIs)
State Subpopulation N
CT General 369
Angler 257
Aquaculture Student 19
Asians 396
Commercial Fisherme 171
EFNEP Participants 60
WIC Participants 557
FL General 8566
MN General 796
American Indians 196
Anglers 1127
New Mothers 352
Weighted
N/1000
2854
7912
4623
Geometric
Mean
17.472
27.364
7.563
38.441
27.687
26.291
24.012
35.071
10.724
7.192
12.472
8.246
Lower
Conf.
Limit
14.952
22.598
2.136
30.593
21.322
13.760
20.126
9.138
5.369
11.282
6.813
Upper
Conf.
Limit
20.418
33.134
26.772
48.302
35.952
50.233
28.648
36.260
12.586
9.634
13.789
9.979
ND General
American Indians
Anglers
570
78
825
606
11.644
9.773
12.641
10.004
5.105
11.591
13.553
18.709
13.787
E-65
-------
Table E-45. Geometric mean consumption per bodyweight, consumers only,
by state and subpopulation (as-consumed g/kg-day with 95% CIs)
State Subpopulation
CT General
Angler
Aquaculture Student
Asians
Commercial Fisherme
EFNEP Participants
WIC Participants
FL General
MN General
American Indians
Anglers
New Mothers
ND General
American Indians
Anglers
N
362
244
19
393
166
58
553
7757
793
192
1109
341
546
64
Weighted
N/1000
2804
7490
4621
Geometric
Mean
0.287
0.385
0.127
0.832
0.416
0.479
0.552
0.579
Lower
Conf.
Limit
0.240
0.321
0.042
0.663
0.325
0.246
0.463
0.559
Upper
Conf.
Limit
0.344
0.462
0.389
1.045
0.533
0.935
0.658
0.600
580
0.
0.
0.
0.
0.
0.
0.
178
127
188
192
187
196
197
0.
0.
0.
0.
0.
0.
0.
153
095
170
159
160
100
180
0
0
0
0
0
0
0
.206
.169
.207
.232
.219
.382
.217
E.11. SPECIES EATEN AND CAUGHT
Tables E-46-E-49 present consumption statistics by species categories using as-
consumed grams offish or shellfish per day. Each survey categorized the species differently.
The categories in the Connecticut data files (corresponding roughly to seafood dishes) were
receded to species categories. The categories in the Florida data were left unchanged. However,
note that there is some repetition in the Florida data in that there are separate records for tuna and
tuna salad. In the Minnesota/North Dakota survey respondents were asked about five
classifications of purchased fish and shellfish and seven classifications of self-caught fish. The
data are summarized using these classifications.
The species tables list the habitat of the species (freshwater versus marine or estuarine),
the species name or category, the number of records (N), the percent of total consumption that is
in that species category, the average fish or shellfish consumption in as-consumed in g/day for
those that consumed the species, the number of records reporting caught fish quantities, the
percent of total caught fish consumption in that species category, and the mean consumption of
caught fish or shellfish for those that caught and consumed the species.
E-66
-------
Table E-46. Total and caught fish consumption for the Connecticut general population (weighted, as-consumed
g/day)
Habitat
Species or class of fish
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Marine
Marine
Marine
i Marine
-O Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Catfish
Bass
Trout
Sunf ish
Tilapia
Bullhead
Walleye
Snails
Tuna
Shrimp
Clams
Scallops
Lobster
Crab
Flounder
Unspecified Fish
Swordf ish
Cod
Salmon
Sardines
Shad
Sole
Haddock
Scrod
Bluefish
Oysters
Perch
Porgy
Snapper
Mussels
Percent
N
32
50
48
5
8
2
2
1
417
210
506
127
365
236
100
170
108
85
109
14
6
56
35
28
76
88
16
17
15
56
of Total
0.
0.
0.
0.
0.
0.
0.
0.
34.
14.
6.
6.
4.
4.
3.
3.
3.
3.
3.
1.
1.
1.
1.
0.
0.
0.
0.
0.
0.
0.
8
8
7
2
2
0
0
0
0
0
9
0
3
0
6
4
4
1
1
6
5
2
1
8
7
5
4
4
4
4
Mean
g/day
2.
1.
1.
6.
2.
1.
0.
0.
9.
7.
1.
5.
1.
1.
4.
2.
3.
4.
3.
14.
30.
2.
3.
3.
1.
0.
3.
2.
2.
0.
89
68
74
30
18
45
88
48
28
63
56
45
37
98
09
27
51
34
34
64
16
21
72
55
09
64
37
61
92
74
N Eating
Caught
Fish
1
26
15
3
0
2
2
0
12
2
11
0
6
6
8
0
0
0
1
0
0
1
0
0
20
1
4
5
0
1
Percent
of Total
Caught
0.
14.
12.
8.
.
0.
0.
.
7.
3.
2.
.
3.
1.
3.
.
.
.
0.
.
.
0.
.
.
6.
0.
11.
3.
.
1.
4
2
8
4
8
6
4
1
4
2
6
4
1
9
9
3
0
6
9
Mean g/day
of Caught
Fish
1
1
2
10
1
0
1
5
0
1
0
1
0
3
1
0
9
2
4
.41
.57
.64
.20
.
.45
.88
.
.83
.58
.68
.
.91
.70
.16
.
.
.
.22
.
.
.30
.
.
.05
.86
.41
.09
.
.61
-------
Table E-46. Total and caught fish consumption for the Connecticut general population (weighted, as-consumed
g/day) (continued)
w
oo
Habitat Species or class of fish
Marine Pollock
Marine Whiting
Marine Tautog
Marine Turbot
Marine Grouper
Marine Herring
Marine Halibut
Marine Shark
Marine Eel
Marine Sea Trout
Marine Orange Roughy
Marine Milkfish
Marine Mackerel
Marine Monkfish
Marine Sea bass
Marine Anchovies
Marine Scungilli
Marine Whitefish
Marine Dolphinfish
Marine Bonito
Marine Butterfirsh
Marine Bream
Marine Smelt
Marine Caviar
N
13
12
19
3
8
8
21
15
8
2
7
4
7
7
2
17
3
6
6
1
1
1
2
5
Percent
of Total
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.4
.3
.3
.2
.2
.2
.2
.1
.1
.1
.1
.1
.1
.1
.1
.0
.0
.0
.0
.0
.0
.0
.0
.0
Mean
g/day
3
3
1
8
3
3
1
0
1
5
1
1
1
0
4
0
1
0
0
2
2
0
0
0
.16
.04
.73
.53
.02
.10
.04
.98
.34
.07
.08
.95
.05
.91
.05
.30
.64
.58
.56
.91
.57
.59
.29
.05
N Eating Percent
Caught of Total
Fish Caught
0
2 0.7
14 10.1
0
0
0
0
1 1.5
2 3.0
0
0
0
2 0.6
0
0
0
0
4 1.0
0
0
0
0
0
0
Mean g/day
of Caught
Fish
.
1.28
2.14
.
.
.
.
4.40
4.40
.
.
.
1.17
.
.
.
.
0.66
.
.
.
.
.
m
-------
Table E-47. Total and caught fish consumption for the Florida general population (weighted, as-consumed
g/day)
W
ON
VO
Habitat Species or class of fish
Freshwater Freshwater catfish
Freshwater Panfish
Freshwater Largemouth bass
Freshwater Other freshwater finfish
Freshwater Freshwater crayfish
Freshwater Sunshine bass
Freshwater Panfish roe
Freshwater Largemouth bass roe
Marine Canned tuna
Marine Shrimp
Marine Flounder
Marine Grouper
Marine Snapper
Marine Breaded fish fillets
Marine Fish sticks
Marine Salmon
Marine Mullet
Marine Dolphin
Marine Fresh tuna
Marine Clams
Marine Seatrout
Marine Stone crab claws
Marine Oysters
Marine Other marine finfish
Marine Sardines
Marine Mackerel
Marine Cod
Marine Crab meat
Marine Immitation crab meat
Marine Shark
Marine Unknown finfish
Percent
N
597
273
107
39
25
13
3
1
2888
2895
819
832
663
485
391
618
402
360
219
289
221
148
366
301
241
156
230
349
218
110
111
of Total
4.
1.
0.
0.
0.
0.
0.
0.
23.
7.
6.
6.
5.
3.
3.
3.
3.
3.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
1.
0.
0.
9
8
8
2
1
1
0
0
9
2
6
5
8
9
7
5
3
0
8
6
6
5
5
5
4
3
2
1
0
9
8
Mean
g/day
42.
34.
39.
26.
24.
29.
34.
21.
37.
11.
38.
35.
38.
38.
42.
25.
50.
35.
38.
25.
36.
47.
23.
22.
26.
42.
22.
15.
19.
38.
34.
85
42
16
54
12
33
29
65
80
88
46
53
76
59
13
34
37
63
38
37
88
00
58
08
61
21
41
30
76
55
04
N Eating
Caught
Fish
126
69
107
2
0
5
0
1
0
271
139
176
158
0
0
13
154
82
33
10
85
29
48
25
0
50
8
20
0
54
19
Percent
of Total
Caught
7.
4.
5.
0.
.
0.
.
0.
.
4.
6.
9.
13.
.
.
0.
9.
5.
4.
0.
5.
3.
1.
1.
.
3.
0.
0.
.
3.
0.
0
1
5
0
2
0
2
9
2
0
5
1
7
3
4
0
4
6
3
2
2
4
0
8
Mean g/day
of Caught
Fish
53.
49.
39.
19.
.
38.
.
21.
.
12.
35.
34.
51.
.
.
21.
52.
42.
87.
25.
40.
81.
29.
33.
.
47.
16.
13.
.
41.
33.
81
77
16
12
80
65
54
29
80
58
27
54
72
44
62
41
33
01
20
19
13
16
71
48
-------
Table E-47. Total and caught fish consumption for the Florida general population (weighted, as-consumed
g/day) (continued)
W
Habitat
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Species or class of fish
Swordfish
Red drum
Scallops
Blue crab
Whole lobster
Lobster tails
Orange Roughy
King Mackerel
Whitefish
Conch
Mussels
Other fresh frozen shellfish
Salad shrimp
Sheepshead
Snook
Mullet roe
Haddock
Bluefish
Pompano
Saltwater catfish
Amberjack
Sea bass
Halibut
Processed shellfish
Unknown shellfish
Seatrout roe
Other fresh/frozen finfish
Other processed finfish
Percent
N
112
115
398
126
152
155
93
55
80
50
44
38
69
33
30
23
33
16
11
14
25
17
16
2
10
6
3
3
of Total
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.8
.8
.7
.6
.6
.6
.4
.4
.3
.3
.3
.3
.2
.2
.2
.2
.2
.1
.1
.1
.1
.1
.1
.0
.0
.0
.0
.0
Mean
g/day
30.
33.
8.
24.
17.
15.
19.
27.
21.
27.
29.
29.
15.
26.
26.
41.
19.
24.
33.
27.
20.
23.
16.
57.
17.
26.
22.
17.
04
14
51
19
60
94
97
37
19
94
35
89
07
36
51
10
75
03
22
44
27
04
17
96
47
01
48
76
N Eating
Caught
Fish
8
115
28
0
28
29
0
4
5
5
0
0
2
6
7
10
3
6
2
3
19
5
0
0
0
3
0
0
Percent
of Total
Caught
0
5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.4
.5
.7
.
.8
.6
.
.1
.1
.4
•
.0
.2
.3
.8
.1
.2
.1
.1
.3
.1
.
.
.
.1
.
.
Mean g/day
of Caught
Fish
31.
33.
16.
.
18.
12.
.
21.
18.
48.
•
23.
25.
29.
49.
12.
24.
21.
20.
16.
16.
.
.
.
26.
.
.
67
14
46
48
68
17
14
36
51
96
52
87
94
61
65
84
11
27
14
-------
Table E-48. Total and caught fish consumption for the Minnesota general population (weighted, as-consumed
g/day)
W
Habitat Species or class of fish
Freshwater Panfish
Freshwater Walleye or Sauger
Freshwater Northern pike or Muskie
Freshwater Salmon or Lake trout
Freshwater Bass
Freshwater Other non-purchased fish
Freshwater Stream trout
Marine Canned tuna
Marine Other purchased fish
Marine Breaded fish, fish sticks
Marine Shellfish
Marine Swordfish & Shark
N
447
498
275
206
125
71
53
657
559
524
560
102
Percent
of Total
13.6
8.2
4.2
4.1
1.9
1.9
1.4
21.3
14.5
13.3
13.2
2.4
Mean
g/day
4.84
3.21
3.05
4.76
2.48
3.87
5.30
5.10
3.93
3.99
3.54
2.82
N Eating
Caught
Fish
447
498
275
206
125
71
53
0
0
0
0
0
Percent
of Total
Caught
38.5
23.3
11.9
11.6
5.4
5.4
3.9
.
.
.
.
Mean g/day
of Caught
Fish
4.84
3.21
3.05
4.76
2.48
3.87
5.30
.
.
.
.
-------
Table E-49. Total and caught fish consumption for the North Dakota general population (weighted, as-
consumed g/day)
Percent
Habitat
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
V^ Marine
t° Marine
Marine
Marine
Marine
Species or class of fish
Walleye or Sauger
Panfish
Northern pike or Muskie
Salmon or Lake trout
Other non-purchased fish
Bass
Stream trout
Canned tuna
Other purchased fish
Breaded fish, fish sticks . . .
Shellfish
Swordfish & Shark
N
369
274
230
105
36
62
14
472
389
375
423
58
of Total
13.
10.
5.
1.
1.
0.
0.
20.
16.
15.
12.
2.
2
1
5
5
0
8
2
8
5
5
2
6
Mean
g/day
4
4
2
1
3
1
1
5
4
4
3
5
.08
.05
.69
.68
.17
.39
.72
.07
.82
.68
.28
.01
N Eating
Caught
Fish
369
274
230
105
36
62
14
0
0
0
0
0
Percent
of Total
Caught
40.
31.
17.
4.
3.
2.
0.
.
.
.
.
.
8
2
0
8
2
4
6
Mean g/day
of Caught
Fish
4.
4.
2.
1.
3.
1.
1.
.
.
.
.
.
08
05
69
68
17
39
72
-------
E.12. ADDITIONAL DETAILS FOR TARGETED POPULATIONS
Tables E-50 through E-181 summarize fish and shellfish consumption by various
dependent variables. Tables E-182 to E-192 have lists of species consumed, tabled by targeted
populations.
There are separate tables for fish and shellfish consumption per capita and for those
individuals that reported consuming fish or shellfish (consumers only). The tables tabulate the
fish and shellfish consumption rate (either as g/day or g/kg-day bodyweight). There are separate
tables for fish and shellfish consumption in as-consumed weight per day and raw or uncooked
weight per day. In the table titles, the word "Consumption" is used to stand for the fish and
shellfish consumption rate. Within each table the fish and shellfish consumption is broken out
by the independent variables.
Tables with summaries using per capita, or estimates for the sampled populations, show
the following statistics (column headers are in parentheses):
• State abbreviation (State);
• Independent variables;
• Sample size (SampN);
• Weighted sample size for the general population in thousands (PopN/1,000);
• Arithmetic mean (Pop Arith Mean);
• Geometric mean (Pop Geom Mean), this is missing in all cases;
• Percent of respondents in the row that reported eating fish or shellfish (Percent Eating
Fish);
• Minimum (Min);
• Percentiles, the 5th, 25th, 50th, 75th, 90th, 95th, and 99th percentiles (Pop Q5, Pop Q10, Pop
Q25, Pop Q50, Pop Q75, Pop Q90, Pop Q95, and Pop Q99); and
• Maximum (Max).
Since some respondents did not report fish or shellfish consumption during the recall
period, the consumption rate is zero for some respondents. As a result, the minimum
consumption rate is generally 0. Because the geometric mean cannot be calculated for the
population when individuals do not report consuming fish or shellfish, the population geometric
mean is missing in all cases. This column is retained so the tables per capita and consumers only
E-73
-------
have the same format. Tables with summaries for those that consumed fish or shellfish in the
recall period, consumers only, show the following statistics (column headers are in parentheses):
• State abbreviation (State);
• Independent variables;
• Sample size for consumers (SampNC);
• Weighted number offish and shellfish consumers in the general population in thousands
(WtdNC/1,000);
• Arithmetic mean (Arith Mean);
• Geometric mean (Geom Mean);
• Percent of respondents in the row that reported eating fish or shellfish (Percent Eating
Fish), this value is 100% in the tables for consumers only;
• Minimum (Min);
• Percentiles, the 5th, 25th, 50th, 75th, 90th, 95th, and 99th percentiles (Q5, Q10, Q25, Q50,
Q75, Q90, Q95, and Q99); and
• Maximum (Max).
In the tables for consumers only, for the respondents in each row the percentage eating
fish is 100%. This column is retained so the tables per capita and consumers only have the same
format.
Tables E-182 to E-192 have lists of species eaten, classified by the following targeted
populations:
• Table 182, Connecticut recreational anglers;
• Table 183, Connecticut aquaculture students;
• Table 184, Connecticut Asian students;
• Table 185, Connecticut Commercial fishermen
• Table 186, Connecticut EFNEP participants;
• Table 187, Connecticut WIC participants;
• Table 188, Minnesota American Indians;
• Table 189, Minnesota recreational anglers;
E-74
-------
• Table 190, Minnesota families with new mothers;
• Table 191, North Dakota American Indians; and
• Table 192, North Dakota recreational anglers.
These tables present consumption statistics by species categories using as-consumed
grams offish or shellfish per day. Each survey categorized the species differently. The
categories in the Connecticut data files (corresponding roughly to seafood dishes) were receded
to species categories. The categories in the Florida data were left unchanged. However, note
that there is some repetition in the Florida data in that there are separate records for tuna and tuna
salad. I n the Minnesota/North Dakota survey respondents were asked about five classifications
of purchased fish and shellfish and seven classifications of self-caught fish. The data are
summarized using these classifications.
The species tables list the habitat of the species (freshwater versus marine or estuarine),
the species name or category, the number of records (N), the percent of total consumption that is
in that species category (Percent of Total), the average fish or shellfish consumption in as-
consumed g/ day for those that consumed the species (Mean g/day), the number of records
reporting caught fish quantities (N Eating Caught Fish), the percent of total caught fish
consumption in that species category (Percent of Total Caught), and the mean consumption of
caught fish or shellfish for those that caught and consumed the species (Mean g/day of Caught
Fish).
E-75
-------
Table E-50. Consumption, per capita, by state and income (as-consumed g/day)
Household
State Income
Pop Pop
PopN/ Arith Geom
SampN 1000 Mean Mean
Percent
Eating Pop
Fish Min
Pop Pop Pop
Q5 Q10 Q25
Pop
Q50
Pop Pop
Q75 Q90
Pop
Q95
Pop
Q99
Max
CT
FL
w
ND
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-51. Consumption, consumers only, by state and income (as-consumed g/day)
Arith
Mean
Percent
Eating
Fish
Min
CT
FL
w
ND
10
6
18
12
11
13
12
12
10
13
12
12
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-52. Consumption, per capita, by state and race-ethnicity (as-consumed g/day)
State Race Ethnicity
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
w
-!j
oo
CT White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
All
FL White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
MN White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
0
0
4
0
0
11
10
White, Non-Hispanic
Black, Non-Hispanic
Asian
American Indian
Unknown
All
70
124
15
3
14
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-53. Consumption, consumers only, by state and race-ethnicity (as-consumed g/day)
State Race Ethnicity
Arith
Mean
Min
w
CT White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
All
FL White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
MN White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
ND White, Non-Hispanic
Black, Non-Hispanic
Asian
American Indian
Unknown
All
11
8
10
11
21
22
18
22
10
6
12
21
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-54. Consumption, per capita, by state, adult/child and race-ethnicity (as-consumed g/day)
Adult
Child
Race Ethnicity
Percent
Eating
Fish
Pop
Min
CT Adult White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
CT Child White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
3
11
15
12
16
18
21
w
oo
o
CT
FL
11
FL
Adult White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
Child White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
70
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-54. Consumption, per capita, by state, adult/child and race-ethnicity (as-consumed g/day) (continued)
Adult
State Child Race Ethnicity
Pop Percent
Geom Eating Pop
Mean Fish Min
w
oo
ND
Adult White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
Child White, Non-Hispanic
Asian
American Indian
Unknown
Adult White, Non-Hispanic
Black, Non-Hispanic
Asian
American Indian
Unknown
Child White, Non-Hispanic
American Indian
Unknown
13
16
18
41
100 167
13
ND
18
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-55. Consumption, consumers only, by state, adult/child, and race-ethnicity (as-consumed g/day)
Adult
Child
Race Ethnicity
Arith
Mean
Percent
Eating
Fish
Min
CT Adult White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
CT Child White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
51
41
24
11
9
14
18
16
18
168
18
21
w
oo
to
CT
FL
100
11
FL
Adult White, Non-Hispanic 5620
Black, Non-Hispanic 625
Hispanic 605
Asian 114
American Indian 51
Unknown 116
Child White, Non-Hispanic 987
Black, Non-Hispanic 242
Hispanic 157
Asian 14
American Indian 12
Unknown 23
12
17
11
15
14
17
16
6
22
22
18
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-55. Consumption, consumers only, by state, adult/child, and race-ethnicity (as-consumed g/day)
(continued)
Adult
Child
Race Ethnicity
Arith
Mean
Min
Adult
Child
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
19. 4
13
18
14
18
w
oo
ND
Adult
Child
1
White, Non-Hispanic 371
Black, Non-Hispanic 2
Asian 4
American Indian 8
Unknown 25
100
411
2
13
4
15
100
14
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-56. Consumption, per capita, by state and gender (as-consumed g/day)
Gender
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
Male
Female
All
17
14
15
66
311
494
494
FL
Male
Female
Unknown
All
w
oo
ND
11
12
11
13
10
11
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-57. Consumption, consumers only, by state and gender (as-consumed g/day)
Samp
Gender NC
Arith
Mean
Min
CT
Male
Female
All
11
8
9
FL
12
11
16
12
w
oo
ND
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-58. Consumption, per capita, by state, adult/child and gender (as-consumed g/day)
Adult
State Child
Gender
Percent
Eating Pop
Fish Min
CT
Adult
Male
Female
156
181
1158
1344
34.2
30. 9
90 . 4
88. 6
0 . 00
0.00
o
0
2
0
11
6
22
20
44
40
75
66
110
93
161
278
311
494
CT Child
Male
Female
15
14
66
44
4
12
w
oo
FL Adult
FL Child
Male
Female
0
0
0
1
1
0
0
0 0
0 0
0 0
3 8
2 6
1 2
0 2
13
14
113
111
131
41
ND
ND
ND
Adult
Child
Unknown
Male
Female
Male
Female
Male
Female
200
230
83
82
5
2
222
253
7 9
78
5
2
20.
19.
19.
11.
21.
30.
, 6
, 9
,0
, 0
, 4
,0
96.
94 .
97 .
89.
64.
,5
, 6
9
9
,5
100
0 . 00
0.00
0.00
0 . 00
0.00
8. 91
1
0
1
0
0
3
2
3
0
0
7 14
6 11
5 11
2 6
0 19
30
25
23
24
12
31
55
43 84
46 66
55 67
26 33
126
165
96
130
240
97
97
60
60
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-59. Consumption, consumers only, by state, adult/child and gender (as-consumed g/day)
Adult
State Child
CT Adult
CT Child
Gender
Male
Female
Male
Female
Samp
NC
141
161
31
27
WtdNC/
1000
1047
1190
286
256
Arith
Mean
37.
34.
15.
13.
Geom
Mean
8
9
5
1
25.
18.
9 .
9 .
,0
,5
, 9
7
Percent
Eating
Fish
100
100
100
100
Min
1.72
0.26
0. 63
1.34
Q5
6
1
2
3
Q10
9
4
3
3
Q25
14
9
5
6
Q50
23
23
10
10
Q75
46
45
19
17
Q90
81
67
35
2 6
Q95
114
96
53
34
Q99 Max
170 311
298 494
66
51
w
oo
FL Adult
FL Child
Male
Female
13.1
12.5
14
11
16
18
17
18
14
71
13
14
55
41
ND Adult
Male
Female
14
12
Male
Female
11
24
13
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-60. Consumption, per capita, by state and education (as-consumed g/day)
Respondent
Education
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
FL
w
oo
oo
14
14
22
14
13
11
11
11
14
13
11
11
111
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-61. Consumption, consumers only, by state and education (as-consumed g/day)
Respondent
Education
Arith
Mean
Percent
Eating
Fish
Min
CT
FL
13
w
oo
VO
2
2
6
2
11
16
11
12
11
19
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-62. Consumption, per capita, by state and age-gender category (as-consumed g/day)
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL
w
JD
o
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-62. Consumption, per capita, by state and age-gender category (as-consumed g/day) (continued)
State
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
11
w
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
18
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-63. Consumption, consumers only, by state and age-gender category (as-consumed g/day)
Arith
Mean
Percent
Eating
Fish
Min
CT
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
3.1
FL
w
JD
to
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
8
8
11
10
12
12
12
14
13
16
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-63. Consumption, consumers only, by state and age-gender category (as-consumed g/day) (continued)
Arith
Mean
Percent
Eating
Fish
Min
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
31
41
ND
w
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
126
14
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-64. Consumption, per capita, by state and age-gender category (as-consumed g/day)
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
6
20
18
11
FL
w
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
12
0
11
0
14
0
13
19
24
12
ND
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
9
10
11
11
15
18
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-65. Consumption, consumers only, by state and age-gender category (as-consumed g/day)
Arith
Mean
Percent
Eating
Fish
Min
CT
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
6
11
8
11
14
1
44
FL
w
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
11
12
12
14
16
1
2
4
1
18
22
22
ND
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
14
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-66. Consumption, per capita, by state and acquisition method (as-consumed g/day)
State
Bought or
Caught
Acquisition
Method
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
Pop
Q10
CT
Bought
Caught
All
14
0
15
FL
Bought
Caught
All
w
ND
Bought
Caught
All
Bought
Caught
All
841
841
841
11.7
6.4
18.1
6.1
1
11
6
2
11
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-67. Consumption, consumers only, by state and acquisition method (as-consumed g/day)
State
Bought or
Caught
Acquisition
Method
Arith
Mean
Percent
Eating
Fish
Q10
CT
Bought
Caught
All
17.1
1. 4
17.5
FL
Bought
Caught
All
142
147
157
w
ND
Bought
Caught
All
Bought
Caught
All
4.7
11. 6
4
12
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-68. Consumption, per capita, by state acquisition method, and income (as-consumed g/day)
Bought or
Caught
Acquisition Household
State Method Income
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
Pop
Q10
CT
w
JD
oo
FL
FL
Bought
"aught
Bought
"aught
21
11
37
16
1
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-68. Consumption, per capita, by state acquisition method, and income (as-consumed g/day) (continued)
Bought or
Caught
Acquisition Household
State Method Income
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
Bought
w
ND
Bought
Caught
11
17
12
14
13
11
14
14
15
19
14
14
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-69. Consumption, consumers only, by state, acquisition method, and income (as-consumed g/day)
Bought or
Caught
Acquisition Household
State Method Income
Arith
Mean
Percent
Eating
Fish
CT
Bought
14
"aught $ 0-20000 5 42
$20000-50000 30 206
$50000- 39 312
Unknown 1 7
11
16
10
11
37
16
1
o
o
FL Bought $ 0-20000 1572 1422
$20000-50000 3441 3191
$50000- 1830 1737
Unknown 1123 1069
FL Caught $ 0-20000 289 224
$20000-50000 656 534
$50000- 301 268
Unknown 156 134
11
12
11
10
12
11
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-69. Consumption, consumers only, by state, acquisition method, and income (as-consumed g/day)
(continued)
Bought or
Caught
Acquisition Household
State Method Income
Arith
Mean
Percent
Geom Eating
Mean Fish
Q5 Q10 Q25 Q50 Q7
MN
w
o
ND
ND
Bought
"aught
Bought
"aught
14.7
13.1
11.8
10
18
13
15
14
15
9
11
47
17
18
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-70. Consumption, per capita, by state and habitat (as-consumed g/day)
State Habitat
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
o
to
CT
FL
ND
Freshwater
Estuarine
Marine
All
431
431
431
431
17181
17181
17181
17181
11
11
14
44
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-71. Consumption, consumers only, by state and habitat (as-consumed g/day)
State Habitat
Arith
Mean
Min
CT
FL
ND
Freshwater
Estuarine
Marine
All
8
18
22
12
10
10
14
24
18
4
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-72. Consumption, per capita, by state and fish/shellfish type (as-consumed g/day)
State
Finfish or
Shellfish
Type
SampN
PopN/
1000
Pop
Arith
Mean
Pop
Geom
Mean
Percent
Eating
Fish
Pop
Min
Pop
Q5
Pop
Q10
Pop
Q25
Pop
Q50
Pop
Q75
Pop
Q90
Pop
Q95
Pop
Q99
Max
CT
Shellfish
Finfish
All
FL
Shellfish
Finfish
All
15
71
MN
Shellfish
Finfish
All
1
9
11
Shellfish
Finfish
All
1
10
11
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-73. Consumption, consumers only, by state and fish/shellfish type (as-consumed g/day)
Finfish or
Shellfish
State Type
Arith
Mean
Min
CT
Shellfish
Finfish
All
FL
Shellfish
Finfish
All
MN
Shellfish
Finfish
All
1
9
12
Shellfish
Finfish
All
10
12
41
46
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-74. Consumption, consumers only, by state and type of fish consumed (as-consumed g/day)
Type of
Fish/Shellfish
Eaten
Arith
Mean
Min
CT Eats Caught Only
Eats Caught&Bought
Eats Bought Only
All Fish Consumers
FL Eats Caught Only
Eats Caught&Bought
Eats Bought Only
All Fish Consumers
MN Eats Caught Only
Eats Caught&Bought
Eats Bought Only
All Fish Consumers
ND Eats Caught Only
Eats Caught&Bought
Eats Bought Only
All Fish Consumers
30
11
12
4
14
10
12
15
14
24
108
157
1
311
494
494
317 2339
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-75. Consumption, consumers only, by state and fresh/estuarine fish consumption (as-consumed g/day)
Arith
Mean
Percent
Eating
Fish
CT
FL
ND
Exclusively
Sometimes
Never
All Fish Consumers
Exclusively
Sometimes
Never
All Fish Consumers
Exclusively
Sometimes
Never
All Fish Consumers
113
494
494
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-76. Fish consumption, per capita, by state and income (uncooked g/day)
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
FL
o
oo
ND
18
14
14
17
15
15
12
17
15
15
144
129
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-77. Fish consumption, consumers only, by state and income (uncooked g/day)
Household
State Income
Percent
Samp WtdNC/ Arith Geom Eating
NC 1000 Mean Mean Fish Min
Q5 Q10 Q25
Q50
Q75 Q90
Q95
Q99
Max
CT
FL
ND
13
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-78. Fish consumption, per capita, by state and race-ethnicity (uncooked g/day)
State Race Ethnicity
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
w
o
CT White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
All
FL White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
MN White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
ND White, Non-Hispanic
Black, Non-Hispanic
Asian
American Indian
Unknown
All
11
10
n
11
11
2
11
15
222
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-79. Fish consumption, consumers only, by state and race-ethnicity (uncooked g/day)
Percent
Geom Eating
Mean Fish
FL
|Tl
MN
ND
White, Non-Hispanic 338
Black, Non-Hispanic 3
Hispanic 15
Asian 12
Unknown 1
All 369
White, Non-Hispanic 6607
Black, Non-Hispanic 867
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
White, Non-Hispanic
Black, Non-Hispanic
Asian
American Indian
Unknown
All
14
11
14
11
10
2
15
21
13
17
18
19
16
11
11
2
13
13
10
14
13
22
28
28
14
16
16
15
21
11
2 9
11
16
166
222
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-80. Fish consumption, per capita, by state, adult/child and race-ethnicity (uncooked g/day)
Adult
Child
Race Ethnicity
Percent
Eating Pop
Fish Min
CT Adult White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
CT Child White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
4
15
16
22
w
to
CT
FL
11
41
44
FL
Adult White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
Child White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-80. Fish consumption, per capita, by state, adult/child and race-ethnicity (uncooked g/day) (continued)
Percent
Eating
Fish
Pop
Min
Max
Adult
MN
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
White, Non-Hispanic
Black, Non-Hispanic
Asian
American Indian
Unknown
222
11
8
11
12
0
18
4
18
78
66
ND
ND
Child White, Non-Hispanic
American Indian
Unknown
Unknown White, Non-Hispanic
155
5
5
7
147 19.
5 18.
5 25.
7 32 .
, 9
,0
,4
, 0
93.5
100
100
75.2
0 .
0.
5 .
0 .
,00
,78
,31
, 00
025
9
12
007
11
2 1
3 6
23
22 44 79
29
38
51
130
2 9
38
7 9
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-81. Fish consumption, consumers only, by state, adult/child, and race-ethnicity (uncooked g/day)
Adult
State Child Race Ethnicity
Percent
Samp WtdNC/ Arith Geom Eating
NC 1000 Mean Mean Fish Min Q5 Q10 Q25 Q50 Q75 Q90 Q95 Q99 Max
w
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
FL
Adult White, Non-Hispanic 5620
Black, Non-Hispanic 625
Hispanic 605
Asian 114
American Indian 51
Unknown 116
Child White, Non-Hispanic 987
Black, Non-Hispanic 242
Hispanic 157
Asian 14
American Indian 12
Unknown 23
4.3
100
11
16
11
16
g
14
11
8
10
6
17
14
11
21
11
13
20
17
13
14
13
6
15
651
44
-------
Table E-81. Fish consumption, consumers only, by state, adult/child, and race-ethnicity (uncooked g/day)
(continued)
Adult
State Child Race Ethnicity
WtdNC/
1000
Arith
Mean
Percent
Geom Eating
Mean Fish
Q5 Q10 Q25 Q50 Q7
ND
Adult White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
Child White, Non-Hispanic
Asian
American Indian
Unknown
Unknown White, Non-Hispanic
Adult White, Non-Hispanic
Black, Non-Hispanic
Asian
American Indian
Unknown
Child
145
1.4
1.4
11
11
1
11
8
11
11
18
18
17
21
11
32
11
18
78
62
78
111
ND
Unknown White, Non-Hispanic
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-82. Fish consumption, per capita, by state and gender (uncooked g/day)
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
Male
Female
All
FL
w
ND
24.1
17
13
15
26
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-83. Fish consumption, consumers only, by state and gender (uncooked g/day)
Samp
Gender NC
Arith
Mean
Min
CT
FL
Male
Female
All
11
17
14
21
16
15
11
13
44
w
14.3
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
ND
Male
Female
All
277
293
570
295
311
606
28.0
25.5
26.7
17.1
14.2
15.5
100
100
100
1.16
0.78
0.78
3
2
2
5
3
4
9
T
8
18
14
16
34
29
32
64
59
61
104
84
91
159
178
164
174
320
320
-------
Table E-84. Fish consumption, per capita, by state, adult/child and gender (uncooked g/day)
Adult
State Child
Gender
Percent
Eating
Fish
Pop
Min
CT Adult
Male
Female
1158
1344
411
651
CT Child
Male
Female
44
FL Adult
oo
FL Child
Male
Female
11
8
10
9
17
18
145
145
ND Adult
Male
Female
61
111
88
Male
Female
2
2
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-85. Fish consumption, consumers only, by state, adult/child and gender (uncooked g/day)
Adult
State Child
CT Adult
CT Child
Gender
Male
Female
Male
Female
Samp
NC
141
161
31
27
WtdNC/
1000
1047
1190
286
256
Arith
Mean
51.
47.
21.
18.
Geom
Mean
8
3
1
1
34.
25.
13.
13.
, 6
,8
, 6
Percent
Eating
Fish
100
100
100
100
Min
2.58
0.46
0.83
1.77
Q5
9
2
3
3
Q10
12
5
4
5
Q25
20
13
7
9
Q50
34
30
15
16
Q75
61
60
27
23
Q90
110
91
46
35
Q95
164
123
71
47
Q99
233
385
Max
411
651
90
68
w
FL Adult
FL Child
1.71
1.43
4.57
12
10
14
12
11
11
20
22
21
22
111
104
104
Male
Female
18
18
10
13
17
19
145
145
1.4
1.4
ND Adult
Male
Female
Male
Female
168
226
26
FL consumption
FL consumption
Statistics are
on a 7-day recall, CT, MN, ND consumtpion is based on rate of
away-from-home consumption by children < 18.
to represent the general population in the states.
-------
Table E-86. Fish consumption, per capita, by state and education (uncooked g/day)
Respondent
Education
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
FL
to
o
2
1
10
10
4
12
4
173
117
154
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-87. Fish consumption, consumers only, by state and education (uncooked g/day)
Respondent
State Education
Percent
Samp WtdNC/ Arith Geom Eating
NC 1000 Mean Mean Fish Min
Q5 Q10 Q25
Q50
Q75 Q90
Q95
Q99
Max
CT
FL
w
to
13
2
2
10
14
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-88. Fish consumption, per capita, by state and age-gender category (uncooked g/day)
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL
to
to
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
113
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-88. Fish consumption, per capita, by state and age-gender category (uncooked g/day) (continued)
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
Max
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
15
17
21
22
26
ND
to
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-89. Fish consumption, consumers only, by state and age-gender category (uncooked g/day)
Arith
Mean
Percent
Eating
Fish
Min
CT
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
10
n
10
13
12
1
FL
to
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
9
9
11
10
11
12
12
13
11
14
14
16
16
17
20
18
21
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-89. Fish consumption, consumers only, by state and age-gender category (uncooked g/day) (continued)
Arith
Mean
Percent
Eating
Fish
Min
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
ND
to
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
11. 9
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-90. Fish consumption, per capita, by state and age-gender category (uncooked g/day)
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
FL
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
to
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
ND
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-91. Fish consumption, consumers only, by state and age-gender category (uncooked g/day)
Arith
Mean
Percent
Eating
Fish
Min
CT
FL
to
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
10
11
11
12
13
1
2
11
14
16
17
18
21
2
2
11
13
131
ND
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
31
104
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-92. Fish consumption, per capita, by state and acquisition method (uncooked g/day)
State
Bought or
Caught
Acquisition
Method
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
Pop
Q10
CT
Bought
Caught
All
431
431
431
110
3
113
FL
Bought
Caught
All
121
28
137
Bought
Caught
All
841
841
841
15
18
8
to
oo
ND
Bought
Caught
All
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-93. Fish consumption, consumers only, by state and acquisition method (uncooked g/day)
State
Bought or
Caught
Acquisition
Method
Arith
Mean
Percent
Eating
Fish
Q10
CT
Bought
Caught
All
12
1
13
FL
Bought
Caught
All
14
14
16
26
43
42
44
Bought
Caught
All
78
41
24
to
VO
ND
Bought
Caught
All
10
6
16
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-94. Fish consumption, per capita, by state acquisition method, and income (uncooked g/day)
Bought or
Caught
Acquisition Household
State Method Income
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
Pop
Q10
CT
w
OJ
o
FL
FL
Bought
"aught
Bought
"aught
6
2
1
116
121
132
110
37
17
12
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-94. Fish consumption, per capita, by state acquisition method, and income (uncooked g/day)
(continued)
Bought or
Caught
Acquisition Household
State Method Income
Percent
Eating Pop
Fish Min
w
ND
Bought
Caught
13
2
1
11
8
8
10
14
17
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-95. Fish consumption, consumers only, by state, acquisition method, and income (uncooked g/day)
Bought or
Caught
Acquisition Household
State Method Income
Arith
Mean
Percent
Eating
Fish
CT
Bought
22
28
26
44
w
OJ
to
CT Caught $ 0-20000 5 42
$20000-50000 30 206
$50000- 39 312
Unknown 1 7
FL Bought $ 0-20000 1572 1422
$20000-50000 3441 3191
$50000- 1830 1737
Unknown 1123 1069
FL Caught $ 0-20000 289 224
$20000-50000 656 534
$50000- 301 268
Unknown 156 134
11
8
11
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-95. Fish consumption, consumers only, by state, acquisition method, and income (uncooked g/day)
(continued)
Bought or
Caught
Acquisition Household
State Method Income
Percent
Geom Eating
Mean Fish
Q5 Q10 Q25 Q50 Q7
w
MN
ND
ND
Bought
"aught
Bought
"aught
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-96. Fish consumption, per capita, by state and habitat (uncooked g/day)
State Habitat
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
FL
w
ND
Freshwater
Estuarine
Marine
All
431
431
431
431
17181
17181
17181
17181
1
8
15
20
15
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-97. Fish consumption, consumers only, by state and habitat (uncooked g/day)
State Habitat
Arith
Mean
Min
CT
FL
10
12
16
83
64
111
137
w
Freshwater
594
2968
14.0
100
0.78
Estuarine
Marine
All
ND Freshwater
Estuarine
Marine
All
560
758
796
409
422
537
570
3309
4402
4623
439
451
570
606
2 .
15.
25.
11.
2 .
17.
2 6 .
,4
, 6
,5
, 9
, 2
,5
7
1.3
8.8
14.3
6.3
1 . 2
8.7
15.5
100
100
100
100
100
100
100
0
0
0
0
0
0
0
.29 0
.39 1
.78 2
.78 1
.29 0
.39 1
.78 2
0
2
2
2
0
2
4
1
4
8
3
1
4
8
1
9
16
6
1
9
16
3
17
27
14
3
19
32
5
39
52
30
6
38
61
8
62
74
43
8
72
91
22
102
140
71
15
133
164
32
181
651
170
42
229
320
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-98. Fish consumption, per capita, by state and fish/shellfish type (uncooked g/day)
State
Finfish or
Shellfish
Type
SampN
PopN/
1000
Pop
Arith
Mean
Pop
Geom
Mean
Percent
Eating
Fish
Pop
Min
Pop
Q5
Pop
Q10
Pop
Q25
Pop
Q50
Pop
Q75
Pop
Q90
Pop
Q95
Pop
Q99
Max
CT
Shellfish
Finfish
All
431
431
431
11
21
55
90
FL
Shellfish
Finfish
All
MN
Shellfish
Finfish
All
1
12
15
w
Shellfish
Finfish
All
1
13
15
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-99. Fish consumption, consumers only, by state and fish/shellfish type (uncooked g/day)
Finfish or
Shellfish
State Type
Arith
Mean
Min
CT
Shellfish
Finfish
All
61
FL
Shellfish
Finfish
All
11
MN
w
Shellfish
Finfish
All
Shellfish
Finfish
All
14
16
11
50
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-100. Fish consumption, consumers only, by state and type offish consumed (uncooked g/day)
Type of
Fish/Shellfish
Eaten
Arith
Mean
Min
CT
FL
°°
ND
Eats Caught Only 1
Eats Caught&Bought 74
Eats Bought Only 294
All Fish Consumers 369
Eats Caught Only
Eats Caught&Bought
Eats Bought Only
All Fish Consumers
Eats Caught Only 33
Eats Caught&Bought 376
Eats Bought Only 161
All Fish Consumers 570
33
9
10
14
13
13
11
4
11
43
91
43
44
60 120 179 280 780
166 250 396 849 2605
72 123 173 333 1679
79 137 199 407 2605
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-101. Fish consumption, consumers only, by state and fresh/estuarine fish consumption (uncooked
g/day)
Eats
Freshwater/
Estuarine Caught
State Fish
Arith
Mean
Fish
Min
w
OJ
vo
CT
FL
ND
Sometimes
Never
All Fish Consumers
Exclusively
Sometimes
Never
All Fish Consumers
Exclusively
Sometimes
Never
All Fish Consumers
11
4
8
11
43
44
58 121 174 289 289
156 257 402 680 1770
76 129 185 366 2605
79 137 199 407 2605
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-102. Fish consumption per kg, per capita, by state and income (as-consumed g/kg-day)
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
FL
ND
40 303 0.389
150 1137 0.472
214 1737 0.380
16 119 0.323
420 3296 0.410
87 371 0.401
326 1801 0.337
327 2155 0.288
97 570 0.244
837 4897 0.309
0.00 0.00 0.00 0.08 0.26 0.49
0.00 0.00 0.00 0.10 0.28 0.58
0.00 0.00 0.00 0.07 0.24 0.52
0.00 0.00 0.00 0.00 0.30 0.50
0.00 0.00 0.00 0.08 0.25 0.55
0.00 0.00 0.00 0.00 0.06 0.60
0.00 0.00 0.00 0.00 0.21 0.65
0.00 0.00 0.00 0.00 0.00 0.48
0.00 0.00 0.00 0.00 0.06 0.59
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-103. Fish consumption per kg, consumers only, by state and income (as-consumed g/kg-day)
Arith
Mean
Percent
Eating
Fish
Min
CT
FL
w
ND
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-104. Fish consumption per kg, per capita, by state and race-ethnicity (as-consumed g/kg-day)
State Race
CT Whit
Ethnicity
e. Non-Hisoanic
PopN/
SampN 1000
370 2888
Pop
Arith
Mean
0407
Pop
Geom
Mean
Percent
Eating
Fish
88.7
Pop
Min
0 00
Pop
Q5
0 00
Pop
Q10
0 00
Pop
Q25
0 10
Pop
Q50
0 27
Pop
Q75
0.57
Pop
Q90
0. 98
Pop
Q95
1.27
Pop
Q99
2.01
Max
4.53
FL
to
MN
ND
Black, Non-Hispanic
Hispanic
Asian
Unknown
All
Hispanic
Asian
American Indian
Unknown
All
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
White, Non-Hispanic
Black, Non-Hispanic
Asian
American Indian
Unknown
All
0.00 0.00 0.00 0.00 0.00 0.13
0.00 0.00 0.00 0.00 0.21 0.50
0.00 0.00 0.00 0.00 0.14 0.53
0.00 0.00 0.00 0.00 0.00 0.01
0.00 0.00 0.00 0.08 0.25 0.55
0.00 0.00 0.00 0.00 0.09 0.58
0.00 0.00 0.00 0.00 0.00 0.70
0.00 0.00 0.00 0.00 0.00 0.54
0.00 0.00 0.00 0.00 0.00 0.67
0.00 0.00 0.00 0.00 0.15 0.70
0.00 0.00 0.00 0.00 0.00 0.58
0.00 0.00 0.00 0.00 0.06 0.59
0.19 . . 0.21 0.27 1.18
0.09 0.11 0.13 0.19 0.47 0.84
0.07 . 0.09 0.11 0.16 1.38
0.02 0.09 0.10 0.12 0.24 0.34
0.00 0.00 0.02 0.08 0.18 0.33
-------
Table E-105. Fish consumption per kg, consumers only, by state and race-ethnicity (as-consumed g/kg-day)
State Race Ethnicity
Arith
Mean
Percent
Eating
Fish
Min
CT White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
All
FL White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
MN White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
ND White, Non-Hispanic
Black, Non-Hispanic
Asian
American Indian
Unknown
All
1.48
1.09
9.21
1.01 1.37 1.94
1.08 1.84 9.21
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-106. Fish consumption per kg, per capita, by state, adult/child and race-ethnicity (as-consumed g/kg-
day)
Percent
Eating
Fish
Pop
Min
Pop
Q10
CT
CT
Adult White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
Child White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
CT
FL Adult White, Non-Hispanic 9325
Black, Non-Hispanic 1072
Hispanic 1189
Asian 177
American Indian 89
Unknown 226
FL Child White, Non-Hispanic 2282
Black, Non-Hispanic 531
Hispanic 367
0.00 0.00 0.02 0.13 0.29 0.58
0.00 0.00 0.00 0.00 0.12 0.16
0.00 0.00 0.00 0.10 0.31 1.12
0.00 0.00 0.00 0.00 0.05 0.32
0.01 .....
0 .
0.94 1.29 2.15
4.
76.1
0.00 0.00 0.00 0.00 0.15 0.57
0.00 0.00 0.00 0.00 0.11 0.62
0.00 0.00 0.00 0.00 0.00 0.53
0.00 0.00 0.00 0.00 0.26 0.71
0.00 0.00 0.00 0.00 0.12 0.69
0.00 0.00 0.00 0.00 0.00 0.56
0.00 0.00 0.00 0.00 0.00 0.62
0.00 0.00 0.00 0.00 0.00 0.95
0.00 0.00 0.00 0.00 0.00 0.57
0.00 0.00 0.00 0.00 0.00 0.33
0.00 0.00 0.00 0.00 0.37 2.38
0.00 0.00 0.00 0.00 0.00 0.86
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-106. Fish consumption per kg, per capita, by state, adult/child and race-ethnicity (as-consumed g/kg-
day) (continued)
Adult
State Child Race Ethnicity
ND
Adult White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
Child White, Non-Hispanic
Asian
American Indian
Unknown
Unknown White, Non-Hispanic
Adult White, Non-Hispanic
Black, Non-Hispanic
Asian
American Indian
Unknown
0.00 0.01 0.02 0.09 0.17 0.31 0.55 0.81 1.48 1.88
.000.
0.
0.00 0.00 0.00 0.00 0.00 0.28
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-107. Fish consumption per kg, consumers only, by state, adult/child, and race-ethnicity (as-consumed
g/kg-day)
Percent
Arith Geom Eating
Mean Mean Fish
Q5 Q10 Q25 Q50 Q7
CT Adult
White, Non-Hispanic 272
Black, Non-Hispanic 3
Hispanic 11
Asian 9
CT
Child
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
FL Adult White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
FL Child White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
18
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-107. Fish consumption per kg, consumers only, by state, adult/child, and race-ethnicity (as-consumed
g/kg-day) (continued)
Adult
State Child Race Ethnicity
MN Adult White, Non-Hispanic 576 3175 0.26 0.16 100
Black, Non-Hispanic . . . . 100 ......
Hispanic 3 50 0.65 0.42 100 0.19 . . 0.21 0.27 0.88
Asian 4 35 0.16 0.16 100 0.09 0.09 0.11 0.13 0.16 0.20
American Indian 10 57 2.82 0.50 100 0.07 . 0.08 0.12 0.24 1.46
Unknown 28 173 0.32 0.23 100 0.02 0.09 0.10 0.11 0.22 0.35 0.84 1.01
w
tL^ MN Unknown White, Non-Hispanic 1 1 0.18 0.18
ND Adult White, Non-Hispanic 359 397 0.30 0.17 100 0.01 0.03 0.04 0.09 0.17 0.33 0.64 1.14 1.99 4.23
Black, Non-Hispanic 2 2 0.25 0.25 100 0.23 . . . 0.25 0.27 . . . 0.28
Asian 4 3 0.20 0.13 100 0.04 . . 0.05 0.14 0.34 . . . 0.38
American Indian 6 6 0.35 0.29 100 0.14 . 0.14 0.19 0.24 0.40 . . . 0.80
Unknown 25 28 0.23 0.15 100 0.04 0.05 0.05 0.07 0.10 0.40 0.55 0.65 . 0.69
ND Child
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-108. Fish consumption per kg, per capita, by state and gender (as-consumed g/kg-day)
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
oo
CT
FL
ND
Male
Female
All
1.22
1.32
1.41
1.27
0 .
1.51
1.87
1.81
1.18
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-109. Fish consumption per kg, consumers only, by state and gender (as-consumed g/kg-day)
Samp
Gender NC
Arith
Mean
Percent
Eating
Fish
Min
CT
Male
Female
All
FL
ND
1.21
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-110. Fish consumption per kg, per capita, by state, adult/child and gender (as-consumed g/kg-day)
Adult
State Child
Gender
Percent
Eating Pop
Fish Min
CT Adult
Male
Female
1.31
1.34
CT Child
Male
Female
1.41
1.13
1.59
1.51
44
FL Adult
FL Child
Male
Female
1.13
1.10
ND Adult
Male
Female
Male
Female
1.57
1.31
ND Unknown Male
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-lll. Fish consumption per kg, consumers only, by state, adult/child and gender (as-consumed g/kg-
day)
Adult
State Child
Q5 Q10 Q25 Q50 Q7
CT
Adult
1041
1154
:hild
CT
Male
Female
w
FL
Male
Female
FL
Child
Male
Female
Unknown
1.71
1.74
1.44
:hild
1.17
1.31
ND
ND
ND
Adult
Child
Unknown
Male
Female
Male
Female
Male
187
209
7 7
7 2
i
207
230
74
68
1
0.
0.
0.
0 .
0 .
, 2 6
,33
,54
,39
,34
0.15
0.18
0.30
0.21
0.34
100
100
100
100
100
0.
0.
0.
0 .
0 .
,01
,01
,02
,04
,34
0.
0.
0.
0 .
, 02
,03
,05
, 05
0.04
0.05
0.08
0.06
0.08
0.10
0.16
0. 08
0.17
0.17
0.29
0.21
0.30
0.37
0. 60
0 . 33
0.54
0.73
1.22
0 . 75
0.
1.
1.
1.
,88
,18
,59
,45
1.78
2.15
4 . 07
1.
4 .
4 .
6 .
0 .
92
,23
,29
, 75
,34
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-112. Fish consumption per kg, per capita, by state and education (as-consumed g/kg-day)
Respondent
Education
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
FL
to
0.03 0.03 0.05 0.08 0.15 0.39
0.00 0.00 0.00 0.05 0.22 0.58
0.00 0.00 0.00 0.14 0.30 0.51
0.00 0.00 0.00 0.08 0.25 0.53
0.00 0.00 0.00 0.00 0.00 0.48
0.00 0.00 0.00 0.00 0.00 0.56
0.00 0.00 0.00 0.00 0.11 0.61
0.00 0.00 0.00 0.00 0.15 0.63
0.00 0.00 0.00 0.00 0.00 0.65
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-113. Fish consumption per kg, consumers only, by state and education (as-consumed g/kg-day)
Respondent
State Education
Arith
Mean
Percent
Eating
Fish
Min
CT
FL
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-114. Fish consumption per kg, per capita, by state and age-gender category (as-consumed g/kg-day)
Percent
Eating Pop
Fish Min
CT
FL
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-114. Fish consumption per kg, per capita, by state and age-gender category (as-consumed g/kg-day)
(continued)
State
Pop Pop
Arith Geom
Mean Mean
Percent
Eating
Fish
Pop
Min
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-115. Fish consumption per kg, consumers only, by state and age-gender category (as-consumed g/kg-
day)
State
Arith
Mean
Percent
Eating
Fish
Q5 Q10 Q25 Q50 Q7
CT
FL
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
14
4 .54
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-115. Fish consumption per kg, consumers only, by state and age-gender category (as-consumed g/kg-
day) (continued)
State
Age-Gender
Category
Samp
NC
WtdNC/ Arith
1000 Mean
Geom
Mean
Percent
Eating
Fish
Min
Q5
Q10
Q25
Q50
Q75
Q90
Q95
Q99
Max
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-116. Fish consumption per kg, per capita, by state and age-gender category (as-consumed g/kg-day)
Percent
Eating
Fish
Pop
Min
CT
-------
Table E-117. Fish consumption per kg, consumers only, by state and age-gender category (as-consumed g/kg-
day)
State
Arith
Mean
Percent
Eating
Fish
Q5 Q10 Q25 Q50 Q7
CT
FL
MN
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-118. Fish consumption per kg, per capita, by state and acquisition method (as-consumed g/kg-day)
Bought or
Caught
Acquisition
State Method
Percent
Eating
Fish
Pop
Min
Pop
Q10
CT
Bought
Caught
All
FL
Bought
Caught
All
Bought
Caught
All
ND
Bought
Caught
All
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-119. Fish consumption per kg, consumers only, by state and acquisition method (as-consumed g/kg-
day)
Bought or
Caught
Acquisition
Method
Arith
Mean
Fish
Min
Bought
Caught
All
FL
Bought
Caught
All
w
ND
Bought
Caught
All
1.44
1.26
1.84
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-120. Fish consumption per kg, per capita, by state acquisition method, and income (as-consumed g/kg-
day)
to
State
Bought or
Caught
Acquisition Household
Method Income
Samp PopN/
N 1000
Pop Pop
Arith Geom
Mean Mean
Percent
Eating
Fish
Pop Pop
Min Q5
Pop
Q10
Pop
Q25
Pop
Q50
Pop
Q75
Pop
Q90
Pop
Q95
Pop
Q99
Max
CT
FL
FL
Bought
Caught
11 0
18.1
16.8
6 . 2
0 00
0.00
0 . 00
0.00
n
0.
0 .
0.
00
,00
,00
,00
0 00
0.00
0 . 00
0.00
0 00
0.00
0 . 00
0.00
0 00
0.00
0 . 00
0.00
n
0.
0 .
0.
00
,00
,00
,00
n
0.
0 .
0.
00
,02
,01
,00
0 0
0.0
0.0
0.0
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-120. Fish consumption per kg, per capita, by state acquisition method, and income (as-consumed g/kg-
day) (continued)
Bought or
Caught
Acquisition Household
State Method Income
Percent
Eating Pop
Fish Min
ND
Bought
Caught
0.00 0.00 0.00 0.00 0.03 0.13
0.00 0.00 0.00 0.00 0.04 0.10
0.00 0.00 0.00 0.00 0.02 0.09
0.00 0.00 0.00 0.00 0.02 0.03
1.23
7.47
1. 84
1.41
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-121. Fish consumption per kg, consumers only, by state, acquisition method, and income (as-consumed
g/kg-day)
Bought or
Caught
Acquisition Household
State Method Income
Arith
Mean
Percent
Eating
Fish
Min
CT
FL
FL
Bought
Caught
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-121. Fish consumption per kg, consumers only, by state, acquisition method, and income (as-consumed
g/kg-day) (continued)
Bought or
Caught
Acquisition Household
State Method Income
Arith
Mean
Percent
Eating
Fish
Min
ND
Bought
Caught
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-122. Fish consumption per kg, per capita, by state and habitat (as-consumed g/kg-day)
State Habitat
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
FL
ND
Freshwater
Estuarine
Marine
All
0.00 0.00 0.00 0.00 0.00 0.01
0.00 0.00 0.00 0.00 0.04 0.11
0.00 0.00 0.00 0.05 0.17 0.40
0.00 0.00 0.00 0.08 0.25 0.55
0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.06
0.00 0.00 0.00 0.00 0.00 0.37
0.00 0.00 0.00 0.00 0.06 0.59
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-123. Fish consumption per kg, consumers only, by state and habitat (as-consumed g/kg-day)
State Habitat
Arith
Mean
Percent
Eating
Fish
Min
CT
FL
ND
Freshwater
Estuarine
Marine
All
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-124. Fish consumption per kg, per capita, by state and fish/shellfish type (as-consumed g/kg-day)
Finfish or
Shellfish
State Type
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
Shellfish
Finfish
All
FL
Shellfish
Finfish
All
1.27
MN
Shellfish
Finfish
All
-------
Table E-125. Fish consumption per kg, consumers only, by state and fish/shellfish type (as-consumed g/kg-day)
Finfish or
Shellfish
State Type
Arith
Mean
Percent
Eating
Fish
Min
CT
Shellfish
Finfish
All
FL
Shellfish
Finfish
All
MN
Shellfish
Finfish
All
ND
Shellfish
Finfish
All
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-126. Fish consumption per kg, consumers only, by state and type offish consumed (as-consumed g/kg-
day)
Type of
Fish/Shellfish
State Eaten
Percent
Samp WtdNC/ Arith Geom Eating
NC 1000 Mean Mean Fish Min Q5 Q10 Q25 Q50 Q75
Q90 Q95
Q99
Max
CT
FL
ND
Eats Caught Only 1
Eats Caught&Bought 70
Eats Bought Only 291
All Fish Consumers 362
Eats Caught Only 511
Eats Caught&Bought 701
Eats Bought Only 6545
All Fish Consumers 7757
Eats Caught Only 38
Eats Caught&Bought 555
Eats Bought Only 200
All Fish Consumers 793
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-127. Fish consumption per kg, consumers only, by state and fresh/estuarine fish consumption (as-
consumed g/kg-day)
Eats
Freshwater/
Estuarine Caught
State Fish
Arith
Mean
Percent
Eating
Fish
Min
w
FL
ND
Exclusively
Sometimes
Never
All Fish Consumers
Exclusively
Sometimes
Never
All Fish Consumers
1.25
1.41
1.37
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-128. Fish consumption per kg, per capita, by state and income (uncooked g/kg-day)
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
FL
to
ND
0.00 0.00 0.00 0.13 0.34 0.65
0.00 0.00 0.00 0.15 0.39 0.78
0.00 0.00 0.00 0.09 0.34 0.72
0.00 0.00 0.00 0.00 0.42 0.69
0.00 0.00 0.00 0.12 0.35 0.76
0.00 0.00 0.00 0.00 0.08 0.76
0.00 0.00 0.00 0.00 0.27 0.82
0.00 0.00 0.00 0.00 0.00 0.60
0.00 0.00 0.00 0.00 0.08 0.74
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-129. Fish consumption per kg, consumers only, by state and income (uncooked g/kg-day)
Arith
Mean
Percent
Eating
Fish
Min
CT
FL
ND
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-130. Fish consumption per kg, per capita, by state and race-ethnicity (uncooked g/kg-day)
State Race
CT Whit
Ethnicity
e. Non-Hisoanic
PopN/
SampN 1000
370 2888
Pop
Arith
Mean
0.556
Pop
Geom
Mean
Percent
Eating
Fish
88.7
Pop
Min
0 00
Pop
Q5
0 00
Pop
Q10
0 00
Pop
Q25
0.15
Pop
Q50
0.38
Pop
Q75
0.78
Pop
Q90
1.32
Pop
Q95
1. 69
Pop
Q99
2. 64
Max
5. 98
Black, Non-Hispanic
Hispanic
Asian
Unknown
All
FL White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
MN White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
ND White, Non-Hispanic
Black, Non-Hispanic
Asian
American Indian
Unknown
All
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-131. Fish consumption per kg, consumers only, by state and race-ethnicity (uncooked g/kg-day)
Percent
Eating
Fish
CT White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
All
FL White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
MN White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
All
ND White, Non-Hispanic
Black, Non-Hispanic
Asian
American Indian
Unknown
All
1.41
1.31
1.34
1.44
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-132. Fish consumption per kg, per capita, by state, adult/child and race-ethnicity (uncooked g/kg-day)
Percent
Eating
Fish
Pop
Min
CT
FL
Adult White, Non-Hispanic 296 2194 0.579
Black, Non-Hispanic 5 34 0.131
Hispanic 13 103 1.023
Asian 16 119 0.703
Unknown 1 9 0.029
Child White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
11
Adult White, Non-Hispanic 9325
Black, Non-Hispanic 1072
Hispanic 1189
Asian 177
American Indian 89
Unknown 226
FL
92.0 0.00 0.00 0.03 0.18 0.40 0.79
64.5 0.00 0.00 0.00 0.00 0.16 0.22
82.5 0.00 0.00 0.00 0.13 0.41 1.61 3.35
51.0 0.00 0.00 0.00 0.00 0.06 0.45 1.21
100 0.03.
1.71 2.7"
0.00 0.00 0.00 0.00 0.20 0.72
0.00 0.00 0.00 0.00 0.15 0.76
0.00 0.00 0.00 0.00 0.00 0.68
0.00 0.00 0.00 0.00 0.32 0.91
0.00 0.00 0.00 0.00 0.16 0.91
0.00 0.00 0.00 0.00 0.00 0.71
40.1 0.00 0.00 0.00 0.00 0.00 0.79 2.04
42.0 0.00 0.00 0.00 0.00 0.00 1.16
40.3 0.00 0.00 0.00 0.00 0.00 0.70
30.2 0.00 0.00 0.00 0.00 0.00 0.41
58.1 0.00 0.00 0.00 0.00 0.46 2.90
37.6 0.00 0.00 0.00 0.00 0.00 1.05
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-132. Fish consumption per kg, per capita, by state, adult/child and race-ethnicity (uncooked g/kg-day)
(continued)
Adult
State Child Race Ethnicity
Adult White, Non-Hispanic 602 3296 0.331
Black, Non-Hispanic 1 1 0.000
Hispanic
Asian
American Indian
Unknown
Child White, Non-Hispanic 168 1112 0.461
Asian 3 59 1.002
American Indian 2 22 0.172
Unknown 11 31 0.452
Unknown White, Non-Hispanic 5
Adult White, Non-Hispanic 375 414 0.384
Black, Non-Hispanic 2 2 0.333
Asian 4 3 0.264
American Indian 6
Unknown 27
0.25 . . 0.28 0.36 1.57 . . . 1.97
0.12 0.12 0.14 0.17 0.22 0.26 . . . 0.26
0.09 . 0.10 0.16 0.32 6.82 . . . 12.27
0.03 0.12 0.13 0.15 0.29 0.47 1.13 1.37 . 1.43
0.00 0.00 0.02 0.08 0.28 0.62 1.00 1.54 3.09 10.67
0.58 . . 0.68 0.97 1.33 . . . 1.46
0.15 . . 0.15 0.17 0.20 . . . 0.20
0.05 0.08 0.31 0.37 0.40 0.42 0.52 0.95 . 2.59
0.00 0.02 0.04 0.11 0.22 0.42 0.83 1.47 2.63 5.64
0.31 . . . 0.33 0.37 . . . 0.37
0.06 . . 0.07 0.24 0.46 . . . 0.51
0.18 . 0.19 0.25 0.32 0.58 . . . 1.06
0.00 0.00 0.06 0.08 0.13 0.51 0.73 0.87 . 0.92
ND
ND
Unknown White, Non-Hispanic
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-133. Fish consumption per kg, consumers only, by state, adult/child, and race-ethnicity (uncooked g/kg-
day)
Percent
Arith Geom Eating
Mean Mean Fish
Q5 Q10 Q25 Q50 Q7
CT Adult White, Non-Hispanic 272
Black, Non-Hispanic 3
Hispanic 11
Asian 9
Unknown 1
CT
Child White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
Unknown
61
41
24
0.05 0.08 0.14 0.26 0.46 1.02 1.54 1.75 2.10
oo
CT
FL
75 0.16 0.06
100
Adult
FL
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
White, Non-Hispanic
Black, Non-Hispanic
Hispanic
Asian
American Indian
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-133. Fish consumption per kg, consumers only, by state, adult/child, and race-ethnicity (uncooked g/kg-
day) (continued)
Adult
State Child Race Ethnicity
MN Adult White, Non-Hispanic 576 3175 0.34 0.21 100
Black, Non-Hispanic . . . . 100 ......
Hispanic 3 50 0.86 0.56 100 0.25 . . 0.28 0.36 1.18
Asian 4 35 0.22 0.21 100 0.12 0.13 0.14 0.18 0.21 0.26
American Indian 10 57 3.76 0.67 100 0.09 . 0.10 0.16 0.32 1.94
Unknown 28 173 0.43 0.31 100 0.03 0.12 0.13 0.15 0.29 0.47 1.12 1.35
w
KJ MN Unknown White, Non-Hispanic
-------
Table E-134. Fish consumption per kg, per capita, by state and gender (uncooked g/kg-day)
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
oo
o
CT
FL
ND
Male
Female
All
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-135. Fish consumption per kg, consumers only, by state and gender (uncooked g/kg-day)
Samp
Gender NC
Arith
Mean
Min
w
oo
CT
FL
ND
Male
Female
All
51
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-136. Fish consumption per kg, per capita, by state, adult/child and gender (uncooked g/kg-day)
Adult
State Child
Gender
Percent
Eating
Fish
Pop
Min
CT Adult
Male
Female
CT Child
Male
Female
2.11
2. 01
44
FL Adult
oo
to
FL Child
Male
Female
1.51
1.47
ND Adult
Male
Female
1.13
1.54
Male
Female
9. 00
ND Unknown Male
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-137. Fish consumption per kg, consumers only, by state, adult/child and gender (uncooked g/kg-day)
Adult
State Child
Gender
Min
CT Adult
Male
Female
1041
1154
CT Child
Male
Female
2.11
2. 01
FL Adult
oo
FL Child
662
662
11
Male
Female
2.51
10.
ND Adult
Male
Female
Male
Female
ND Unknown Male
100
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-138. Fish consumption per kg, per capita, by state and education (uncooked g/kg-day)
Respondent
Education
Percent
Eating
Fish
Pop
Min
CT
FL
oo
0.00 0.00 0.00 0.00 0.00 0.60
0.00 0.00 0.00 0.00 0.00 0.70
0.00 0.00 0.00 0.00 0.15 0.76
0.00 0.00 0.00 0.00 0.20 0.79
0.00 0.00 0.00 0.00 0.00 0.85
1.34 1.74 . 1.76
1.40 1.55 2.07 2.11
1.09 1.87 3.44 3.62
1.40 1.78 3.53 9.08
1.45 2.16 4.80 24.22
1.59 2.45 5.29 38.29
1.59 2.47 4.95 30.27
1.64 2.34 5.03 18.97
2.04 3.05 3.21 3.22
1.64 2.08 . 2.18
0.86 1.48 2.69 3.20
0.86 1.27 11.37 12.27
0.76 1.40 1.81 2.24
0.55 0.68 0.99 2.59
0.86 1.15
1.19 2.08 5.22 5.72
0.84 1.32 2.12 2.74
0.92 1.69 2.49 9.00
0.98 1.76 . 5.64
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-139. Fish consumption per kg, consumers only, by state and education (uncooked g/kg-day)
Respondent
State Education
Percent
Samp WtdNC/ Arith Geom Eating
NC 1000 Mean Mean Fish Min
Q5 Q10 Q25
Q50 Q75
Q90
Q95
Q99
Max
CT
FL
oo
3.53 8.21 24.22
3.58 7.82 38.29
3.39 6.87 30.27
3.17 6.19 18.97
3.17 . 3.22
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-140. Fish consumption per kg, per capita, by state and age-gender category (uncooked g/kg-day)
Percent
Eating Pop
Fish Min
CT
FL
oo
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-140. Fish consumption per kg, per capita, by state and age-gender category (uncooked g/kg-day)
(continued)
State
Pop Pop
Arith Geom
Mean Mean
Percent
Eating
Fish
Pop
Min
oo
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
0.00 0.00 0.00 0.00 0.00 0
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-141. Fish consumption per kg, consumers only, by state and age-gender category (uncooked g/kg-day)
Arith
Mean
Percent
Eating
Fish
Min
CT
FL
oo
oo
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-141. Fish consumption per kg, consumers only, by state and age-gender category (uncooked g/kg-day)
(continued)
State
Age-Gender
Category
Samp
NC
WtdNC/ Arith
1000 Mean
Geom
Mean
Percent
Eating
Fish
Min
Q5
Q10
Q25
Q50
Q75
Q90
Q95
Q99
Max
oo
VO
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
Child 1- 5
Child 6-10
Child 11-15
Female 16-29
Female 30-49
Female 50+
Male 16-29
Male 30-49
Male 50+
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-142. Fish consumption per kg, per capita, by state and age-gender category (uncooked g/kg-day)
Percent
Eating
Fish
Pop
Min
CT
FL
ND
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
0.00 0.00 0.00 0.00 0.30 0.81
0.00 0.00 0.00 0.15 0.41 0.88
0.00 0.00 0.00 0.11 0.37 0.82
0.00 0.00 0.00 0.00 0.21 0.32
0.00 0.00 0.04 0.20 0.37 0.73
0.00 0.00 0.00 0.01 0.03 0.11
0.00 0.00 0.07 0.17 0.36 0.81
0.00 0.03 0.05 0.10 0.23 0.42
0.00 0.00 0.03 0.11 0.20 0.51
0.02 0.03 0.06 0.10 0.18 0.31
0.00 0.02 0.05 0.10 0.23 0.39
0.00 0.00 0.00 0.00 0.00 0.37
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-143. Fish consumption per kg, consumers only, by state and age-gender category (uncooked g/kg-day)
Arith
Mean
Percent
Eating
Fish
Min
CT
FL
w
ND
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
Child 1-14
Female 15-44
Female 45+
Male 15-44
Male 45+
Unknown
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-144. Fish consumption per kg, per capita, by state and acquisition method (uncooked g/kg-day)
Bought or
Caught
Acquisition
State Method
Percent
Eating
Fish
Pop
Min
Pop
Q10
CT
Bought
Caught
All
FL
Bought
Caught
All
Bought
Caught
All
to
ND
Bought
Caught
All
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-145. Fish consumption per kg, consumers only, by state and acquisition method (uncooked g/kg-day)
State
Bought or
Caught
Acquisition
Method
Arith
Mean
Percent
Eating
Fish
Q10
CT
Bought
Caught
All
1.43
0.17
1.51
FL
Bought
Caught
All
Bought
Caught
All
ND
Bought
Caught
All
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-146. Fish consumption per kg, per capita, by state acquisition method, and income (uncooked g/kg-day)
State
Bought or
Caught
Acquisition
Method
Household
Income
Samp
N
PopN/
1000
Pop
Arith
Mean
Pop
Geom
Mean
Percent
Eating
Fish
Pop
Min
Pop
Q5
Pop
Q10
Pop
Q25
Pop
Q50
Pop
Q75
Pop
Q90
Pop
Q95
Pop
Q99
Max
CT
FL
FL
Bought
"aught
Bought
"aught
0.00 0.00 0.00 0.13 0.34 0.65
0.00 0.00 0.00 0.14 0.37 0.
0.00 0.00 0.00 0.09 0.33 0.72
0.00 0.00 0.00 0.00 0.42 0.68
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-146. Fish consumption per kg, per capita, by state acquisition method, and income (uncooked g/kg-day)
(continued)
Bought or
Caught
Acquisition Household
State Method Income
Percent
Eating Pop
Fish Min
ND
Bought
Caught
0.00 0.00 0.00 0.00 0.04 0.18
0.00 0.00 0.00 0.00 0.06 0.14
0.00 0.00 0.00 0.00 0.03 0.12
0.00 0.00 0.00 0.00 0.02 0.04
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-147. Fish consumption per kg, consumers only, by state, acquisition method, and income (uncooked
g/kg-day)
Bought or
Caught
Acquisition Household
State Method Income
Arith
Mean
Percent
Eating
Fish
Min
CT
CT
FL
FL
Bought
Caught
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-147. Fish consumption per kg, consumers only, by state, acquisition method, and income (uncooked
g/kg-day) (continued)
Bought or
Caught
Acquisition Household
State Method Income
Arith
Mean
Percent
Eating
Fish
Min
ND
Bought
Caught
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-148. Fish consumption per kg, per capita, by state and habitat (uncooked g/kg-day)
State Habitat
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
FL
oo
ND
Freshwater
Estuarine
Marine
All
0.00 0.00 0.00 0.00 0.00 0.01
0.00 0.00 0.00 0.00 0.06 0.15
0.00 0.00 0.00 0.07 0.23 0.53
0.00 0.00 0.00 0.12 0.35 0.76
0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00 0.00 0.08
0.00 0.00 0.00 0.00 0.00 0.46
0.00 0.00 0.00 0.00 0.08 0.74
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-149. Fish consumption per kg, consumers only, by state and habitat (uncooked g/kg-day)
State
CT
Habitat
Freshwater
Samp
NC
157
WtdNC/
1000
1201
Arith
Mean
0. 05
Geom
Mean
0 02
Percent
Eating
Fish
100
Min
0 00
Q5
0 00
Q10
0 00
Q25
0 01
Q50 Q75
0 03 0 05
Q90
0 10
Q95
0.21
Q99
049
Max
0. 67
FL
VO
ND
Freshwater
Estuarine
Marine
All
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-150. Fish consumption per kg, per capita, by state and fish/shellfish type (uncooked g/kg-day)
Finfish or
Shellfish
State Type
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
Shellfish
Finfish
All
FL
Shellfish
Finfish
All
MN
Shellfish
Finfish
All
0.24
1. 11
1.43
w
to
o
o
Shellfish
Finfish
All
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-151. Fish consumption per kg, consumers only, by state and fish/shellfish type (uncooked g/kg-day)
Finfish or
Shellfish
State Type
Arith
Mean
Percent
Eating
Fish
Min
CT
Shellfish
Finfish
All
FL
Shellfish
Finfish
All
1.42
MN
Shellfish
Finfish
All
1.44
w
to
o
Shellfish
Finfish
All
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
A respondent can be represented in more than one row.
-------
Table E-152. Fish consumption per kg, consumers only, by state and type offish consumed (uncooked g/kg-day)
Type of
Fish/Shellfish
State Eaten
Percent
Samp WtdNC/ Arith Geom Eating
NC 1000 Mean Mean Fish Min
Q5 Q10 Q25
Q50 Q75
Q90
Q95
Q99
Max
w
to
o
to
CT Eats Caught Only
Eats Caught&Bought
Eats Bought Only
All Fish Consumers
FL Eats Caught Only
Eats Caught&Bought
Eats Bought Only
All Fish Consumers
MN Eats Caught Only
Eats Caught&Bought
Eats Bought Only
All Fish Consumers
ND Eats Caught Only
Eats Caught&Bought
Eats Bought Only
All Fish Consumers
1.54
1.50
1.51
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-153. Fish consumption per kg, consumers only, by state and fresh/estuarine fish consumption
(uncooked g/kg-day)
Eats
Freshwater/
Estuarine Caught
State Fish
Arith
Mean
Percent
Eating
Fish
Min
w
to
o
FL
ND
Exclusively
Sometimes
Never
All Fish Consumers
Exclusively
Sometimes
Never
All Fish Consumers
1.53
1.50
1.51
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general population in the states.
-------
Table E-154. Fish consumption, per capita, by state and subpopulation (as-consumed g/day)
Population for
sample selection
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
FL
Angler
Aquaculture Students
Asians
Commercial Fishermen
EFNEP Participants
General
WIG Participants
17181
108
228
w
to
o
ND
221
1171
841
415
16
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-155. Fish consumption, consumers only, by state and subpopulation (as-consumed g/day)
State
Population for
sample selection
Percent
Geom Eating
Mean Fish
Angler
Aquaculture Students
Asians
Commercial Fishermen
EFNEP Participants
General
WIG Participants
1
12
14
2 6
15
11
9
12
FL
w
to
o
MN
ND
American Indians
Anglers
General
New Mothers
American Indians
Anglers
General
196
1127
796
352
78
825
570
13.2
21.7
19.1
16.8
26.0
20. 6
20.0
T
12.
10.
8.
9 .
12.
11.
,2
,5
7
2
,8
, 6
, 6
100
100
100
100
100
100
100
0.
0 .
0.
0 .
0.
0 .
0.
,58
, 44
,58
,44
, 44
, 44
,58
1
2
1
1
1
2
2
2
3
2
2
1
4
3
3
T
6
3
3
T
6
9
12
12
9
12
13
12
14
23
20
19
31
23
24
27
45
39
35
93
44
46
43
65
56
58
107
69
68
119
174
105
171
119
123
140
434
489
227
144
172
240
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-156. Fish consumption, per capita, by state, subpopulation, and adult/child (as-consumed g/day)
State
CT
CT
Population for
sample selection
Angler
Aquaculture Students
Adult
Child
Adult
Child
Unknown
Adult
Child
SampN
227
20
19
19
6
Pop Pop
Arith Geom
Mean Mean
49.3
36.1
38.2
14. 9
14.1
Percent
Eating
Fish
97 . 8
100
78.9
84.2
50.0
Pop
Min
0 . 00
0.77
0 . 00
0 . 00
0.00
Pop
Q5
4
1
0
0
0
Pop
Q10
8
1
0
0
0
Pop
Q25
15
3
1
1
0
Pop
Q50
30
17
11
7
0
Pop
Q75
57
75
42
29
3
Pop
Q90
100
85
82
39
74
Pop
Q95
164
102
218
48
Pop
Q99 Max
343 550
117
322
54
82
CT
Adult
Child
10
88
w
to
o
EFNEP Participants Adult
Child
Unknown
CT
20
6
1
43
14
186
61
WIG Participants
FL
Adult
Child
84
44
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-156. Fish consumption, per capita, by state, subpopulation, and adult/child (as-consumed g/day)
(continued)
State
MN
Population for
sample selection
American Indians
Adult
Child
Adult
Child
SampN
140
79
Pop Pop
Arith Geom
Mean Mean
14.3
7.5
Percent
Eating
Fish
92.1
84.8
Pop
Min
0 . 00
0.00
Pop
Q5
0
0
Pop
Q10
1
0
Pop
Q25
2
2
Pop
Q50
9
8
Pop
Q75
16
11
Pop
Q90
31
15
Pop
Q95
46
16
Pop
Q99
127
48
Max
140
49
MN General
w
to
o
ND
American Indians
1
0
0
0
0
0
0
0
2
1
0
1
0
o
0
0
T
2
0
4
1
0
0
0
112
48
1
144
144
ND
13
11
0
71
48
119
101
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-157. Fish consumption, consumers only, state, subpopulation, and adult/child (as-consumed g/day)
Population for Adult Samp Arith
State sample selection Child NC Mean
Min
CT
Aquaculture Students Adult
Child
10
CT
w
to
o
oo
CT
EFNEP Participants
10
177
124
CT
General
45
18
11
WIG Participants
FL
Adult
Child
7131
1435
0.17
12
66
44
118
171
88
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-157. Fish consumption, consumers only, state, subpopulation, and adult/child (as-consumed g/day)
(continued)
Population for
State sample selection
Min
17
11
34
15
44
48
MN General
w
to
o
VO
ND American Indians
1. 6
5
3
1
3
2
11
6
2
11
13
22
16
2
22
38
44
28
105
88
144
144
ND
Anglers
14
12
ND
General
13
4
14
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-158. Fish consumption, per capita, by state, subpopulation, and gender (as-consumed g/day)
Population for
State sample selection
Gender
Pop
Samp Arith
N Mean
Percent
Eating Pop
Fish Min
CT
67
41
1
Aquaculture Students Male
Female
44
41
CT
Male
Female
6
12
47
45
82
68
113
104
w
to
o
CT EFNEP Participants Male
Female
Unknown
11
4
Male
Female
74
311
494
CT
WIG Participants Male
Female
17
18
FL
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-158. Fish consumption, per capita, by state, subpopulation, and gender (as-consumed g/day)
(continued)
Population for
State sample selection Gender
Pop
Arith
Mean
14
12
MN
Male
Female
13
11
434
w
to
14 . 1
14.4
11
12
17
16
14
17
144
144
ND
Male
Female
13
12
ND
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-159. Fish consumption, consumers only, state, subpopulation, and gender (as-consumed g/day)
Population for
State sample selection
Gender
Samp Arith
NC Mean
Min
CT
16
10
CT
Aquaculture Students Male
Female
Male
Female
10
0.13
1.14
12
113
105
w
to
to
CT EFNEP Participants Male
Female
Unknown
21
14
15
110
43
Male
Female
11
8
40
311
494
CT
WIG Participants Male
Female
12
12
FL
10
12
11
16
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-159. Fish consumption, consumers only, state, subpopulation, and gender (as-consumed g/day)
(continued)
Population for Samp Arith
State sample selection Gender NC Mean
Min
126
118
MN
Male
Female
14
11
434
w
to
100
100
12
13
37
41
144
144
ND
Male
Female
13.5
11.7
14
12
ND
14
11
22
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-160. Fish consumption, consumers only, by state, subpopulation, and fresh/estuarine fish consumption
(as-consumed g/day)
Population for
State sample selection
Eats
Freshwater/
Estuarine
Caught Fish
Arith
Mean
w
to
CT
CT
Angler
38.4
11. 4
100
100
1
13
14
1
2
26
15
121
131
105
68
CT
EFNEP Participants
11
305 413
CT
WIG Participants
113
191
113
494
FL
4
20
28
12
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-160. Fish consumption, consumers only, by state, subpopulation, and fresh/estuarine fish consumption
(as-consumed g/day) (continued)
Eats
Freshwater/
Population for Estuarine
State sample selection Caught Fish
Fish
Min
Anglers
w
to
1.7
13.1
1. 60
11
14
6
4
14
10
1
14
71
42
ND
10
1
ND
General
,16
, 60
15
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-161. Fish consumption, per capita, by state and subpopulation (as-consumed g/day)
Population for
sample selection
SampN
Pop
Arith
Mean
Percent
Eating
Fish
Pop
Min
CT
FL
Angler
Aquaculture Students
Asians
Commercial Fishermen
EFNEP Participants
General
WIG Participants
17181
34.1
17
1
34
19
137
288
w
to
ND
221
1171
841
415
11
16
15
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-162. Fish consumption, consumers only, by state and subpopulation (as-consumed g/day)
Population for
sample selection
Samp Arith
NC Mean
Min
CT
Angler
Aquaculture Students
Asians
Commercial Fishermen
EFNEP Participants
General
WIG Participants
FL
100
10
10
1
15
16
26
44
145
148
142
137
w
to
ND
12
17
16
11
26
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-163. Fish consumption, per capita, by state, subpopulation, and adult/child (as-consumed g/day)
Population for Adult
State sample selection Child
Pop
Arith
SampN Mean
Percent
Eating Pop
Fish Min
CT
12
2
0
148
114
111
Aquaculture Students Adult
Child
CT
44
11
13
w
to
oo
CT
EFNEP Participants
13
31
14
37
185
115
CT
General
12
0
0
20
FL
WIG Participants
Adult
Child
44
12
21
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-163. Fish consumption, per capita, by state, subpopulation, and adult/child (as-consumed g/day)
(continued)
State
MN
Population for
sample selection
American Indians
Adult
Child
Adult
Child
SampN
140
79
Pop Pop
Arith Geom
Mean Mean
19.0
10.0
Percent
Eating
Fish
92.1
84.8
Pop
Min
0.00
0 . 00
Pop
Q5
0
o
Pop
Q10
1
0
Pop
Q25
3
2
Pop
Q50
11
10
Pop
Q75
21
14
Pop
Q90
41
19
Pop
Q95
61
2 1
Pop
Q99
169
64
Max
186
65
Unknown
0.0
w
to
VO
MN
General
142
121
ND
Anglers
ND
168
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-164. Fish consumption, consumers only, state, subpopulation, and adult/child (as-consumed g/day)
Population for Adult
State sample selection Child
Arith
Mean
Min
CT
149
114
127
Aquaculture Students Adult
Child
15
4
CT
0.17
1.31
28
6
2
44
14
49
114
w
to
to
o
CT
EFNEP Participants
13
16
CT
General
WIG Participants
12
48
FL
Adult
Child
7131
1435
11
17
11
86
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-164. Fish consumption, consumers only, state, subpopulation, and adult/child (as-consumed g/day)
(continued)
Population for Adult Samp Arith
State sample selection Child NC Mean
Min
13
11
Anglers
w
to
to
MN
17.1
8.3
1.4
18
12
ND
American Indians
ND
10
18
16
61
44
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-165. Fish consumption, per capita, by state, subpopulation, and gender (as-consumed g/day)
Population for
State sample selection
Gender
Pop
Samp Arith
N Mean
Percent
Eating Pop
Fish Min
CT
Aquaculture Students Male
Female
17.1
23.1
CT
Male
Female
16
150
144
w
to
to
to
CT EFNEP Participants Male
Female
Unknown
14
Male
Female
78
411
651
CT
WIG Participants Male
Female
FL
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-165. Fish consumption, per capita, by state, subpopulation, and gender (as-consumed g/day)
(continued)
Population for
State sample selection Gender
Pop
Samp Arith
N Mean
16.7
14 . 4
11
10
MN
Male
Female
18
14
w
to
to
ND
Male
Female
18
22
62
188
181
171
151
ND
17
13
82
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-166. Fish consumption, consumers only, state, subpopulation, and gender (as-consumed g/day)
Population for
State sample selection
Gender
Samp Arith
NC Mean
Min
CT Angler
12
44
31
Aquaculture Students
10
CT Asians
Male
Female
0.17
1.45
111
95
151
147
w
to
to
CT EFNEP Participants
Male
Female
12
20
9
145
Male
Female
15
11
141
114
411
651
CT WIG Participants
16
16
FL General
11
14
17
14
21
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-166. Fish consumption, consumers only, state, subpopulation, and gender (as-consumed g/day)
(continued)
Percent
Population for Samp Arith Geom Eating
State sample selection Gender NC Mean Mean Fish Min
2 2 6 13
1 1 4 11
MN Anglers Male 595 31.3 18.3 100 0.78 3 5 10 18
Female 532 26.3 14.9 100 0.58 3 4 8 15
2 2 9 16
2 3 7 17 27 54 81
1 2 4 12
2 2 5 11
W ND American Indians Male 36 36.2 14.2 100 0.78 1 2
tO Female 42 33.4 12.1 100 0.58 1 1
to
ND Anglers Male 450 28.8 18.0 100 0.58 4 5 10
Female 375 25.7 15.6 100 0.58 348
ND General Male 277 28.0 17.1 100 1.16 3 5 9 18
14.2 100 0.78 2 3 7 14
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-167. Fish consumption, consumers only, by state, subpopulation, and fresh/estuarine fish consumption
(as-consumed g/day)
Population for
State sample selection
Eats
Freshwater/
Estuarine
Caught Fish
Arith
Mean
w
to
to
CT
CT
CT
Angler
EFNEP Participants
100
100
18
10
10
2
20
31
124
171
141
194
134
171
CT
WIG Participants
47 .2
42.1
12
13
FL
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-167. Fish consumption, consumers only, by state, subpopulation, and fresh/estuarine fish consumption
(as-consumed g/day) (continued)
State
Population for
sample selection
Eats
Freshwater/
Estuarine
Caught Fish
Arith
Mean
Percent
Eating
Fish
MN
American Indians
166
w
to
to
MN
General
4
11
4
2
11
4
16
11
1
11
1
14
1
141
142
33
ND
ND
Anglers
11
3
16
15
31
33
111
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-168. Fish consumption per kg, per capita, by state and subpopulation (as-consumed g/kg-day)
Population for
sample selection
Pop
Arith
SampN Mean
Percent
Eating
Fish
Pop
Min
CT Angler
Aquaculture Students
Asians
Commercial Fishermen
EFNEP Participants
General
WIG Participants
18.11
FL
1. 91
w
to
to
oo
ND
0.52 0.64 1.62 2.38
0.66 0.97 2.21 4.63
0.62 1.07 1.81 9.21
0.80 1.21 2.65 8.26
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-169. Fish consumption per kg, consumers only, by state and subpopulation (as-consumed g/kg-day)
State
Population for
sample selection
Percent
Samp Arith Geom Eating
NC Mean Mean Fish Min
Q5 Q10 Q25
Q50 Q75
Q90
Q95
Q99
Max
CT
Angler
Aquaculture Students
Asians
Commercial Fishermen
EFNEP Participants
General
WIG Participants
18.11
FL
100
w
to
to
VO
ND
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-170. Fish consumption per kg, per capita, by state, subpopulation, and adult/child (as-consumed g/kg-
day)
State
Population for
sample selection
Adult
Child
Pop
Arith
SampN Mean
Pop
Geom
Mean
Percent
Eating
Fish
Pop
Min
Pop
Q5
Pop
Q10
Pop
Q25
Pop
Q50
Pop
Q75
Pop
Q90
Pop
Q95
Pop
Q99
Max
CT
Angler
CT
w
to
OJ
o
CT
EFNEP Participants Adult
Child
Unknown
WIG Participants
FL
Adult
Child
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
w
to
Table E-170. Fish consumption per kg, per capita, by state, subpopulation, and adult/child (as-consumed g/kg-
day) (continued)
State
Population for
sample selection
Adult
Child
Pop
Arith
SampN Mean
Pop
Geom
Mean
Percent
Eating
Fish
Pop
Min
Pop
Q5
Pop
Q10
Pop
Q25
Pop
Q50
Pop
Q75
Pop
Q90
Pop
Q95
Pop
Q99
Max
MN American Indians
MN Anglers
MN Anglers
ND American Indians
1.70
3.11
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-171. Fish consumption per kg, consumers only, state, subpopulation, and adult/child (as-consumed
g/kg-day)
State
Population for
sample selection
Adult
Child
Samp
NC
Arith
Mean
Geom
Mean
Percent
Eating
Fish
Min
Q5
Q10
Q25
Q50
Q75
Q90
Q95
Q99
Max
CT
216
20
8
3.77
Aquaculture Students Adult
Child
CT
w
to
OJ
to
CT
EFNEP Participants Adult
Child
Unknown
CT
General
CT
WIG Participants
FL
Adult
Child
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-171. Fish consumption per kg, consumers only, state, subpopulation, and adult/child (as-consumed
g/kg-day) (continued)
Population for Adult Samp Arith
State sample selection Child NC Mean
Min
w
to
MN General
ND
American Indians
1.30
4 .11
ND
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-172. Fish consumption per kg, per capita, by state, subpopulation, and gender (as-consumed g/kg-day)
Population for
State sample selection
Pop
Samp Arith
Gender N Mean
Percent
Eating
Fish
Pop
Min
CT
Male
Female
w
to
CT
CT
CT
FL
Aquaculture Students Male
Female
Male
Female
EFNEP Participants
WIG Participants
Male
Female
1.44
1.57
1.34
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-172. Fish consumption per kg, per capita, by state, subpopulation, and gender (as-consumed g/kg-day)
(continued)
Pop Pop Percent
Population for Samp Arith Geom Eating Pop Pop Pop Pop Pop Pop Pop Pop
State sample selection Gender N Mean Mean Fish Min Q5 Q10 Q25 Q50 Q75 Q90 Q95
MN
Male
Female
w
to
1.74
3.44
ND
Male
Female
1. 14
1.18
ND
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-173. Fish consumption per kg, consumers only, state, subpopulation, and gender (as-consumed g/kg-
day)
Population for
State sample selection
Percent
Arith Geom Eating
Mean Mean Fish
w
to
CT
CT
CT
CT
FL
Angler
Aquaculture Students Male
Female
Male
Female
EFNEP Participants
WIG Participants
Male
Female
10
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-173. Fish consumption per kg, consumers only, state, subpopulation, and gender (as-consumed g/kg-
day) (continued)
Percent
Population for Samp Arith Geom Eating
State sample selection Gender NC Mean Mean Fish Min Q5 Q10 Q25 Q50 Q75 Q90 Q95 Q99 Max
MN Anglers Male 587 0.31 0.19 100 0.01 0.04 0.05 0.10 0.18 0.34 0.63 0.93 2.18 4.34
Female 522 0.33
W ND American Indians Male 29 0.60 0.18 100
K> Female 35 0.57 0.21 100
ND Anglers Male 445 0.33 0.19 100 0.01 0.04 0.05 0.10 0.20 0.38 0.78 1.14
Female 363 0.35 0.20 100 0.01 0.04 0.05 0.11 0.21 0.42 0.83 1.29
ND
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-174. Fish consumption per kg, consumers only, by state, subpopulation, and fresh/estuarine fish
consumption (as-consumed g/kg-day)
Population for
State sample selection
Eats
Freshwater/
Estuarine
Caught Fish
Percent
Samp Arith Geom Eating
NC Mean Mean Fish Min
Angler
w
to
OJ
oo
CT
CT
EFNEP Participants
100
100
CT
WIG Participants
FL
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-174. Fish consumption per kg, consumers only, by state, subpopulation, and fresh/estuarine fish
consumption (as-consumed g/kg-day) (continued)
State
Population for
sample selection
Eats
Freshwater/
Estuarine
Caught Fish
Arith
Mean
Percent
Eating
Fish
Q5 Q10 Q25 Q50 Q7
MN
American Indians
w
to
OJ
VO
MN
General
ND
Anglers
ND
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-175. Fish consumption per kg, per capita, by state and subpopulation (as-consumed g/kg-day)
Population for
State sample selection
Pop
Arith
SampN Mean
Pop
Geom
Mean
Percent
Eating
Fish
Pop
Min
Pop
Q5
Pop
Q10
Pop
Q25
Pop
Q50
Pop
Q75
Pop
Q90
Pop
Q95
Pop
Q99
Max
CT
FL
w
to
j^.
o
ND
Angler
Aquaculture Students
Asians
Commercial Fishermen
EFNEP Participants
General
WIG Participants
0.70 0.86 2.16 3.17
0.88 1.30 2.95 6.17
0.83 1.43 2.41 12.27
1.06 1.61 3.54 11.01
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-176. Fish consumption per kg, consumers only, by state and subpopulation (as-consumed g/kg-day)
Population for
sample selection
Samp Arith
NC Mean
Min
CT
FL
w
to
ND
Angler
Aquaculture Students
Asians
Commercial Fishermen
EFNEP Participants
General
WIG Participants
100
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-177. Fish consumption per kg, per capita, by state, subpopulation, and adult/child (as-consumed g/kg-
day)
State
Population for
sample selection
Adult
Child
Pop
Arith
SampN Mean
Pop
Geom
Mean
Percent
Eating
Fish
Pop
Min
Pop
Q5
Pop
Q10
Pop
Q25
Pop
Q50
Pop
Q75
Pop
Q90
Pop
Q95
Pop
Q99
Max
CT
CT
w
to
j^.
to
CT
FL
EFNEP Participants Adult
Child
Unknown
General
331
11
1.14
3.14
4.74
1.75
6.44 11.07
7.41 25.54
1.79
-------
w
to
MN
ND
ND
ND
Table E-177. Fish consumption per kg, per capita, by state, subpopulation, and adult/child (as-consumed g/kg-
day) (continued)
State
Population for
sample selection
Adult
Child
Pop
Arith
SampN Mean
Pop
Geom
Mean
Percent
Eating
Fish
Pop
Min
Pop
Q5
Pop
Q10
Pop
Q25
Pop
Q50
Pop
Q75
Pop
Q90
Pop
Q95
Pop
Q99
Max
General
4.15
-------
Table E-178. Fish consumption per kg, consumers only, state, subpopulation, and adult/child (as-consumed
g/kg-day)
Population for
State sample selection
Percent
Adult Samp Arith Geom Eating
Child NC Mean Mean Fish Min Q5 Q10 Q25 Q50 Q75
Q90 Q95
Q99
Max
CT
Angler
1.16
1.34
w
to
CT
100
100
EFNEP Participants
100
CT
General
CT
WIG Participants
FL
Adult
Child
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-178. Fish consumption per kg, consumers only, state, subpopulation, and adult/child (as-consumed
g/kg-day) (continued)
Population for Adult Samp Arith
State sample selection Child NC Mean
Min
w
to
MN General
ND
American Indians
ND
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-179. Fish consumption per kg, per capita, by state, subpopulation, and gender (as-consumed g/kg-day)
Population for
State sample selection
CT Angler Male
Female
CT Aquaculture Students Male
Female
CT
CT EFNEP Participants Male
Female
CT General Male 201 0.527 . 86.2 0.00 0.00 0.00 0.14 0.34 0.72 1.48 1.78
Female 219 0.591 . 84.0 0.00 0.00 0.00 0.10 0.39 0.82 1.29 1.73
CT WIG Participants
FL General
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-179. Fish consumption per kg, per capita, by state, subpopulation, and gender (as-consumed g/kg-day)
(continued)
Pop Pop Percent
Population for Samp Arith Geom Eating Pop Pop Pop Pop Pop Pop Pop Pop
State sample selection Gender N Mean Mean Fish Min Q5 Q10 Q25 Q50 Q75 Q90 Q95
MN
Male
Female
1.41
1.46
w
to
ND
Male
Female
ND
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-180. Fish consumption per kg, consumers only, state, subpopulation, and gender (as-consumed g/kg-
day)
Population for
State sample selection
Percent
Arith Geom Eating
Mean Mean Fish
w
to
j^.
oo
CT
CT
CT
CT
FL
Angler
Male
Female
Aquaculture Students Male
Female
Male
Female
EFNEP Participants
WIG Participants
Male
Female
10
1.34
1.16
1.34
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-180. Fish consumption per kg, consumers only, state, subpopulation, and gender (as-consumed g/kg-
day) (continued)
Percent
Samp Arith Geom Eating
NC Mean Mean Fish
MN Anglers Male 587 0.41 0.25 100 0.01 0.05 0.07 0.14 0.24 0.46 0.84 1.24
Female 522 0.44 0.25 100 0.01 0.05 0.07 0.13 0.24 0.47 0.96 1.40
MN General Male 401 0.37 0.22 100 0.01 0.03 0.05 0.12 0.23 0.44 0.82 1.43
Female 392 0.51 0.26 100 0.01 0.03 0.06 0.13 0.29 0.50 0.93 1.62
MN New Mothers Male 177 0.41 0.24 100 0.01 0.04 0.05 0.11 0.26 0.49 1.01 1.41
Female 164 0.62 0.28 100 0.01 0.03 0.07 0.12 0.28 0.58 1.39 2.44
w
|\j ND American Indians Male
^ Female
ND Anglers Male
Female
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-181. Fish consumption per kg, consumers only, by state, subpopulation, and fresh/estuarine fish
consumption (as-consumed g/kg-day)
Eats
Freshwater/ Percent
Population for Estuarine Samp Arith Geom Eating
State sample selection Caught Fish NC Mean Mean Fish Min Q5 Q10 Q25 Q50 Q75 Q90 Q95
CT Angler
2.31
1.17
CT Asians Sometimes 199 1.65 1.18 100 0.02 0.34 0.43 0.71 1.25 2.11 3.88 4.81
Never 194 1.47 1.09 100 0.01 0.25 0.47 0.77 1.17 1.92 2.74 3.14
W CT Commercial Fishermen Sometimes 120 1.06 0.73 100 0.05 0.14 0.24 0.41 0.76 1.33 2.13 2.69
tO Never 46 0.56 0.30 100 0.01 0.01 0.04 0.14 0.41 0.78 1.19 1.87
O
CT EFNEP Participants
CT WIG Participants
FL
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
MN
Table E-181. Fish consumption per kg, consumers only, by state, subpopulation, and fresh/estuarine fish
consumption (as-consumed g/kg-day) (continued)
Eats
Freshwater/ Percent
Population for Estuarine Samp Arith Geom Eating
State sample selection Caught Fish NC Mean Mean Fish
Min Q5 Q10 Q25 Q50 Q75
Q90 Q95
Q99
Max
American Indians
w
MN
General
1.21
ND
ND
Anglers
71
1. 66
FL consumption is based on a 7-day recall, CT, MN, ND consumtpion is based on rate of consumption.
FL consumption excludes away-from-home consumption by children < 18.
Statistics are weighted to represent the general populations. Subpopulations statistics are unweighted.
-------
Table E-182. Total and caught fish consumption for the Connecticut angler population (as-consumed g/day)
Habitat
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
W Freshwater
to
i^rt Freshwater
to
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Species or class of fish
Bass
Trout
Crappie
Northern Pike
Pickerel
Sunfish
Catfish
Snails
Tilapia
Bullhead
Walleye
Tuna
Flounder
Bluefish
Clams
Lobster
Shrimp
Tautog
Scallops
Crab
Perch
Salmon
Cod
Mussels
Porgy
Unspecified Fish
Swordfish
Whiting
Dolphinfish
N
157
109
5
4
5
10
13
1
1
2
2
256
147
157
309
242
99
91
57
166
36
64
56
66
42
59
45
2
16
Percent
of Total
8.
4.
0.
0.
0.
0.
0.
0.
0.
0.
0.
23.
8.
7.
6.
5.
5.
3.
2.
2.
2.
2.
2.
1.
1.
1.
0.
0.
0.
9
1
9
6
5
5
5
1
0
0
0
8
1
0
8
9
3
8
9
6
4
3
0
1
0
0
9
8
7
Mean
g/day
7.
4.
22.
18.
13.
6.
4.
8.
2.
0.
0.
11.
6.
5.
2.
3.
6.
5.
6.
1.
8.
4.
4.
2.
2.
2.
2.
49.
5.
,13
,73
,98
,58
,19
,54
,38
,79
,57
,57
,21
, 66
,89
, 62
,75
,07
, 68
,21
,42
,95
,24
,45
,58
,09
,90
,03
, 60
,68
,79
N Eating
Caught
Fish
151
92
5
4
5
10
11
1
0
2
2
14
111
149
66
33
1
85
1
59
36
4
10
13
39
0
0
2
1
Percent
of Total
Caught
19,
8,
2,
1,
1,
1,
1,
0,
0,
0,
2,
14,
15,
3,
3,
0,
8,
0,
2,
5,
0,
1,
0,
2,
1,
0,
.8
.3
.0
.3
.2
.2
.0
.2
.0
.0
.3
.2
.4
.5
.4
.0
.4
.6
.5
.3
.1
.4
.6
.0
.8
.0
Mean g/day
of Caught
Fish
7,
5,
22,
18,
13,
6,
4,
8,
0,
0,
9,
7,
5,
3,
5,
2,
5,
36,
2,
8,
1,
8,
2,
2,
49,
1,
.38
.08
.98
.58
.19
.54
.96
.79
.57
.21
.42
.20
.81
.03
.76
.20
.55
.55
.37
.24
.44
.06
.81
.95
.68
.65
-------
Table E-182. Total and caught fish consumption for the Connecticut angler population (as-consumed g/day)
(continued)
w
Habitat
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Species or class of fish N
Pollock
Scrod
Sardines
Oysters
Eel
Shark
Sole
Halibut
Sea Trout
Haddock
Mackerel
Other fish
Snapper
Sea bass
Weakf ish
Grouper
Herring
Dorado
Monkf ish
Shad
Anchovies
Drum
Smelt
Sheepshead
5
11
7
31
18
12
17
8
1
18
2
1
6
5
2
1
3
1
1
1
4
1
1
1
Percent
of Total
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.7
.6
. 6
.5
.5
.4
.4
.3
.3
.3
.2
.2
.1
.1
.1
.1
.1
.0
.0
.0
.0
.0
.0
.0
Mean
g/day
17.
7.
10.
2.
3.
4.
2.
5.
34.
1.
14.
24.
2.
2.
4.
7.
2.
3.
1.
1.
0.
0.
0.
0.
,48
,27
,56
,22
, 64
,31
,96
,00
,61
,75
,40
,62
,04
,18
,51
,72
,12
,49
,98
,76
,20
,73
,44
,31
N Eating
Caught
Fish
0
0
0
5
14
1
0
2
1
0
2
1
3
4
2
0
0
0
0
1
0
1
0
1
Percent
of Total
Caught
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
.5
.6
.3
.0
.6
.5
.4
.1
.2
.2
.0
.0
.0
Mean g/day
of Caught
Fish
5,
2,
18,
0,
34,
14,
24,
1,
2,
4,
1,
0,
0,
.95
.22
.47
.22
.61
.40
.62
.17
.46
.51
.76
.73
.31
-------
Table E-183. Total and caught fish consumption for the Connecticut aquaculture students population (as-
consumed g/day)
Habitat
Freshwater
Freshwater
Freshwater
Marine
Marine
Marine
Marine
|Tl Marine
i
^ Marine
r\
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Species or class of fish
Trout
Catfish
Tilapia
Tuna
Scallops
Clams
Lobster
Crab
Shrimp
Anchovies
Salmon
Bluefish
Mussels
Swordfish
Oysters
Unspecified Fish
Halibut
Flounder
Sole
Grouper
Sardines
N
2
1
1
20
6
17
14
7
7
4
5
4
3
6
2
5
1
1
1
1
1
Percent
of Total
1.3
0.6
0.4
43.1
13.0
12.7
8.0
7.0
4.7
1.9
1.4
1.4
1.1
1.0
0.9
0.8
0.2
0.2
0.2
0.1
0.1
N Eating Percent
Mean Caught of Total
g/day Fish Caught
2.
2.
1.
7.
7.
2.
2.
3.
2.
1.
1.
1.
1.
0.
1.
0.
0.
0.
0.
0.
0.
,27
,20
,47
,77
,83
,70
,05
,59
,43
,67
,05
,22
,30
,60
,58
,59
,88
,88
,73
,37
,33
0
0
0
2 32.5
0
1 10.0
0
0
0
0
0
4 33.7
0
0
1 23.9
0
0
0
0
0
0
Mean g/day
of Caught
Fish
1.92
1.18
1.00
2.82
-------
Table E-184. Total and caught fish consumption for the Connecticut Asian population (as-consumed g/day)
to
Habitat
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Species or class of fish N
Bass
Trout
Catfish
Sunfish
Carp
Grapple
Northern Pike
Covina
Bullhead
Whitesucker
Snails
Shrimp
Flounder
Tuna
Mackerel
Bluefish
Salmon
Porgy
Butterf irsh
Crab
Tautog
Whitefish
Perch
Mussels
Whiting
Lobster
Clams
Oysters
218
122
164
130
63
14
20
4
4
2
1
355
225
184
187
124
260
101
83
200
25
52
47
43
39
89
120
65
Percent
of Total
7.
4.
2.
1.
1.
0.
0.
0.
0.
0.
0.
28.
9.
9.
7.
6.
5.
2.
1.
1.
1.
1.
1.
1.
1.
0.
0.
0.
5
2
3
7
4
3
2
1
0
0
0
4
5
3
5
2
5
2
6
5
2
0
0
0
0
9
8
6
Mean
g/day
7.
7.
3.
2.
4.
4.
2.
5.
1.
2.
3.
17.
9.
10.
8.
10.
4.
4.
4.
1.
10.
4.
4.
4.
5.
2.
1.
1.
,40
,31
,05
,84
,82
,97
,26
,57
,31
,54
,33
,18
,11
,85
, 65
,70
,57
,62
,03
, 66
,62
,16
,57
,99
,45
,06
,46
,91
N Eating
Caught
Fish
93
68
113
95
36
1
5
0
0
2
0
0
17
0
0
74
0
18
0
9
21
24
23
0
2
0
10
0
Percent
of Total
Caught
19,
13,
9,
8,
6,
0,
0,
0,
3,
22,
1,
0,
5,
4,
4,
0,
0,
.0
.3
.0
.9
.4
.1
.4
.2
.2
.4
.3
.3
.9
.7
.0
.4
.3
Mean g/day
of Caught
Fish
6,
6,
2,
3,
5,
1,
2,
2,
6,
9,
2,
0,
9,
6,
5,
6,
0,
.75
.45
.64
.08
.87
.65
.49
.54
.17
.98
.30
.94
.32
.46
.71
.87
.92
-------
Table E-184. Total and caught fish consumption for the Connecticut Asian population (as-consumed g/day)
(continued)
Habitat
Marine
Marine
Marine
Marine
Marine
Marine
Marine
|Tl Marine
i
^ Marine
Marine
Marine
Marine
Marine
Marine
Marine
Species or class of fish
Mullet
Swordfish
Skate
Sea bass
Pollock
Sardines
Shad
Scallops
Eel
Shark
Spot
Cod
Halibut
Unspecified Fish
Herring
N
54
18
33
23
4
27
22
14
24
3
6
4
2
2
1
Percent
of Total
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.5
.5
.5
.3
.3
.2
.2
.2
.1
.1
.1
.0
.0
.0
.0
N Eating
Mean Caught
g/day Fish
2.
5.
2.
3.
14.
1.
2.
2.
1.
7.
2.
2.
1.
0.
0.
,10
,72
,93
,10
,47
,76
,16
,52
,08
,85
,93
,41
,09
,51
,86
0
0
0
4
0
0
0
0
3
0
0
0
0
1
0
Percent Mean g/day
of Total of Caught
Caught Fish
0.1 0.58
0.2 2.70
0.0 0.18
-------
Table E-185. Total and caught fish consumption for the Connecticut commercial fishermen population (as-
consumed g/day)
Habitat
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Marine
|Tl Marine
i
^ Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Species or class of fish
Bass
Trout
Catfish
Crawfish
Bullhead
Snails
Tuna
Lobster
Clams
Shrimp
Cod
Scallops
Flounder
Tautog
Crab
Salmon
Bluefish
Swordfish
Unspecified Fish
Snapper
Porgy
Sole
Mussels
Oysters
Eel
Haddock
Dolphinfish
N
86
19
14
2
1
2
217
217
266
83
68
57
75
86
171
42
75
55
65
4
25
13
56
59
23
18
8
Percent
of Total
4.
0.
0.
0.
0.
0.
20.
16.
8.
6.
5.
5.
4.
4.
4.
3.
2.
2.
1.
1.
1.
1.
0.
0.
0.
0.
0.
0
9
3
0
0
0
3
7
1
7
1
0
8
3
3
8
9
8
6
1
0
0
8
6
5
5
4
Mean
g/day
3.
3.
1.
0.
0.
0.
7.
6.
2.
6.
6.
7.
5.
4.
2.
7.
3.
4.
1.
23.
3.
6.
1.
0.
1.
2.
4.
,77
,63
,58
,94
,88
,32
,57
,25
,47
,56
,06
,17
,18
,02
,02
,28
,12
,06
,97
,24
,35
,07
,17
,77
,77
,25
,12
N Eating
Caught
Fish
83
15
3
0
1
2
36
140
95
0
18
5
57
82
65
0
75
0
1
0
24
0
24
13
17
2
2
Percent
of Total
Caught
10,
2,
0,
0,
0,
3,
35,
8,
4,
0,
8,
10,
3,
7,
0,
1,
0,
0,
0,
0,
0,
.2
.1
.1
.0
.0
.5
.5
.2
.0
.9
.8
.6
.1
.5
.0
.8
.4
.3
.3
.2
.0
Mean g/day
of Caught
Fish
3,
4,
0,
0,
0,
2,
7,
2,
6,
5,
4,
3,
1,
3,
0,
2,
0,
0,
0,
3,
0,
.77
.38
.88
.88
.32
.97
.79
.66
.83
.42
.76
.96
.47
.07
.12
.33
.55
.61
.60
.47
.73
-------
Table E-185. Total and caught fish consumption for the Connecticut commercial fishermen population (as-
consumed g/day) (continued)
Habitat
Marine
Marine
Marine
Marine
Marine
Marine
Marine
|Tl Marine
i
^ Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Species or class of fish N
Shark
Scrod
Milkfish
Whitefish
Smelt
Shad
Scungilli
Sardines
Halibut
Pollock
Herring
Caviar
Anchovies
Mackerel
Grouper
Sea bass
Blowf ish
Other fish
Weakf ish
Sea Urchin
Perch
Mullet
20
4
2
2
3
10
11
11
8
6
4
5
5
4
4
3
1
2
2
2
2
1
Percent
of Total
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.4
.3
.3
.2
.2
.2
.2
.2
.2
.1
.1
.1
.1
.0
.0
.0
.0
.0
.0
.0
.0
.0
Mean
g/day
1.
5.
10.
9.
5.
1.
1.
1.
1.
1.
1.
1.
0.
0.
0.
0.
1.
0.
0.
0.
0.
0.
, 62
,43
,45
,85
,89
, 68
,33
,32
,80
,66
,78
,34
,83
,90
,87
, 66
,86
,54
,48
,47
,33
,22
N Eating
Caught
Fish
14
0
0
2
3
0
4
0
0
3
0
2
0
1
0
1
1
1
2
2
2
0
Percent
of Total
Caught
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
.8
.4
.5
.2
.1
.1
.0
.0
.1
.0
.0
.0
.0
Mean g/day
of Caught
Fish
1,
6,
4,
1,
1,
1,
0,
0,
1,
0,
0,
0,
0,
.86
.57
.78
.44
.32
.38
.82
.88
.86
.73
.48
.47
.33
-------
Table E-186. Total and caught fish consumption for the Connecticut EFNEP participant population (as-
consumed g/day)
Habitat
Freshwater
Freshwater
Freshwater
Marine
Marine
Marine
Marine
|Tl Marine
i
^ Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Species or class of fish
Catfish
Bass
Trout
Tuna
Shrimp
Perch
Crab
Salmon
Bluefish
Tautog
Unspecified Fish
Scallops
Lobster
Clams
Pollock
Swordfish
Porgy
Cod
Flounder
Oysters
Haddock
Sardines
Mussels
Sole
Mackerel
N
15
8
1
55
30
18
27
15
22
8
20
13
27
35
1
4
12
13
11
5
2
4
2
5
3
Percent
of Total
2.5
0.6
0.0
36.8
16.8
10.0
7.6
5.5
3.1
3.0
2. 6
2.5
2.1
1.8
1.2
1.2
0.9
0. 6
0.3
0.3
0.3
0.2
0.1
0.1
0.0
Mean
g/day
5.
2.
0.
23.
19.
19.
9.
13.
4.
13.
4.
6.
2.
1.
43.
10.
2.
1.
1.
2.
4.
2.
1.
0.
0.
,94
,67
,32
,75
,86
,79
,99
,02
,95
,18
,53
,80
,75
,84
,53
,35
,80
,56
,04
,20
, 64
,21
,40
,44
,43
N Eating Percent
Caught of Total
Fish Caught
0
8 8.1
0
0
0
0
0
0
18 37.3
8 39.9
0
0
0
0
0
0
12 12.7
4 1.6
3 0.5
0
0
0
0
0
0
Mean g/day
of Caught
Fish
2.67
5.48
13.18
2.80
1.06
0.41
-------
Table E-187. Total and caught fish consumption for the Connecticut WIC participant population (as-consumed
g/day)
w
to
ON
O
Habitat
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Species or class of fish
Bass
Trout
Catfish
Crawfish
Crappie
Whitesucker
Tuna
Unspecified Fish
Shrimp
Crab
Clams
Porgy
Lobster
Whiting
Bluefish
Cod
Haddock
Salmon
Snapper
Scallops
Swordfish
Sole
Oysters
Flounder
Tautog
Mussels
N
62
34
10
2
4
2
522
330
257
259
589
88
418
58
116
68
35
23
17
87
43
26
52
25
12
41
Percent
of Total
1.5
0.5
0.3
0.0
0.0
0.0
34.2
14.3
9.0
5.9
5.9
4.9
4.8
3.0
2.8
2.3
1.5
1.4
1.2
1.2
0.8
0.7
0. 6
0.5
0.3
0.3
Mean
g/day
5.
3.
7.
5.
1.
2.
16.
10.
8.
5.
2.
13.
2.
12.
5.
8.
10.
14.
16.
3.
4.
6.
3.
5.
6.
1.
,97
,28
,32
,45
,92
,08
,14
,67
,65
, 64
,47
,76
,84
,73
,98
,44
, 68
,93
,69
,26
,39
,62
,01
,23
,45
,75
N Eating
Caught
Fish
45
12
5
0
4
2
0
0
0
6
12
16
10
5
60
0
0
0
0
1
0
1
0
6
5
0
Percent
of Total
Caught
27,
3,
2,
0,
0,
0,
0,
11,
3,
8,
28,
0,
0,
0,
2,
.6
.6
.0
.7
.4
.6
.8
.5
.6
.6
.3
.1
.0
.5
.9
Mean g/day
of Caught
Fish
6,
3,
4,
1,
2,
1,
0,
7,
3,
18,
5,
0,
0,
0,
6,
.55
.20
.28
.92
.08
.01
.67
.71
.85
.30
.04
.64
.43
.83
.16
-------
Table E-187. Total and caught fish consumption for the Connecticut WIC participant population (as-consumed
g/day) (continued)
Habitat
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
y"' Marine
to
O\ Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Species or class of fish N
Perch
Scrod
Eel
Bream
Herring
Butterf irsh
Shark
Whitefish
Mackerel
Other fish
Halibut
Doctorf ish
Scungilli
Grunt
Anchovies
Grouper
Smelt
Sea Urchin
Caviar
10
23
5
2
3
4
2
4
2
3
8
1
1
1
2
1
1
1
3
Percent
of Total
0.3
0.3
0.2
0.2
0.2
0.2
0.2
0.1
0.1
0.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
N Eating Percent Mean g/day
Mean Caught of Total of Caught
g/day Fish Caught Fish
7.
3.
11.
26.
15.
11.
19.
6.
10.
6.
1.
10.
3.
0.
0.
0.
0.
0.
0.
,10
,04
,07
,06
,79
,23
,89
,76
,41
,37
,54
,90
,78
,93
,36
, 66
,64
,14
,02
4 4.0 10.59
0
3 5.0 17.70
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1 0.0 0.14
0
-------
Table E-188. Total and caught fish consumption for the Minnesota American Indian population (as-consumed
g/day)
w
to
ON
to
Habitat
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Marine
Marine
Marine
Marine
Marine
Species or class of fish
Walleye or Sauger
Panfish
Northern pike or Muskie
Salmon or Lake trout
Other non-purchased fish
Stream trout
Bass
Canned tuna
Breaded fish, fish sticks . . .
Shellfish
Other purchased fish
Swordfish & Shark
N
169
81
73
29
32
7
24
133
130
106
57
12
Percent
of Total
26.2
14.7
8.2
6.8
1.4
0.9
0.8
16.0
12.0
8.3
3.7
1.0
Mean
g/day
4.
4.
2.
6.
1.
3.
0.
3.
2.
2.
1.
2.
,02
,69
,89
,09
,11
,39
,88
,12
,38
,02
,70
,11
N Eating
Caught
Fish
169
81
73
29
32
7
24
0
0
0
0
0
Percent
of Total
Caught
44,
24,
13,
11,
2,
1,
1,
.5
.9
.8
.6
.3
.6
.4
Mean g/day
of Caught
Fish
4,
4,
2,
6,
1,
3,
0,
.02
.69
.89
.09
.11
.39
.88
-------
Table E-189. Total and caught fish consumption for the Minnesota angler population (as-consumed g/day)
w
K>
(Ts
Habitat
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Marine
Marine
Marine
Marine
Marine
Species or class of fish
Panfish
Walleye or Sauger
Northern pike or Muskie
Salmon or Lake trout
Bass
Other non-purchased fish
Stream trout
Canned tuna
Breaded fish, fish sticks . . .
Other purchased fish
Shellfish
Swordfish & Shark
N
743
779
484
287
255
148
90
927
745
763
808
109
Percent
of Total
16.7
13. 6
8.5
2.2
1.9
1.2
0.7
17.9
12.0
11.1
11.1
3.2
Mean
g/day
5.
4.
4.
1.
1.
1.
1.
4.
3.
3.
3.
7.
,49
,28
,28
,90
,78
,98
,83
,72
,95
,57
,35
,08
N Eating
Caught
Fish
743
779
484
287
255
148
90
0
0
0
0
0
Percent
of Total
Caught
37,
30,
18,
5,
4,
2,
1,
.3
.5
.9
.0
.2
.7
.5
Mean g/day
of Caught
Fish
5,
4,
4,
1,
1,
1,
1,
.49
.28
.28
.90
.78
.98
.83
Table E-190. Total and caught fish consumption for the Minnesota families with new mothers (as-consumed
g/day)
Percent
Habitat
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Marine
Marine
Marine
Marine
Marine
Species or class of fish
Panfish
Walleye or Sauger
Other non-purchased fish
Northern pike or Muskie
Bass
Salmon or Lake trout
Stream trout
Canned tuna
Breaded fish, fish sticks . . .
Other purchased fish
Shellfish
Swordfish & Shark
N
175
168
33
116
69
54
15
282
256
189
203
32
of Total
11.
9.
3.
2.
1.
1.
0.
25.
14.
13.
12.
2.
3
4
8
7
6
2
3
5
9
7
8
9
Mean
g/day
3.
3.
6.
1.
1.
1.
1.
5.
3.
4.
3.
5.
,84
,30
,77
,37
,39
,31
,04
,36
,44
,30
,74
,30
N Eating
Caught
Fish
175
168
33
116
69
54
15
0
0
0
0
0
Percent
of Total
Caught
37,
31,
12,
8,
5,
3,
0,
.5
.0
.5
.8
.3
.9
.9
Mean g/day
of Caught
Fish
3,
3,
6,
1,
1,
1,
1,
.84
.30
.77
.37
.39
.31
.04
-------
Table E-191. Total and caught fish consumption for the North Dakota American Indian population (as-
consumed g/day)
w
1
to
o\
Habitat
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Marine
Marine
Marine
Marine
Marine
Species or class of fish
Panfish
Walleye or Sauger
Northern pike or Muskie
Bass
Other non-purchased fish
Salmon or Lake trout
Canned tuna
Breaded fish, fish sticks . . .
Shellfish
Other purchased fish
Swordfish & Shark
N
10
36
13
8
5
2
56
61
45
17
2
Percent
of Total
14.5
12.4
2.2
0.3
0.2
0.1
31.0
22.5
11.5
2.8
2.5
Mean
g/day
29.
7.
3.
0.
0.
0.
11.
7.
5.
3.
25.
,47
,00
,40
,79
,77
,58
,23
,51
,21
,30
,73
N Eating
Caught
Fish
10
36
13
8
5
2
0
0
0
0
0
Percent
of Total
Caught
48,
41,
7,
1,
0,
0,
.9
.8
.3
.0
.6
.2
Mean g/day
of Caught
Fish
29,
7,
3,
0,
0,
0,
.47
.00
.40
.79
.77
.58
Table E-192. Total and caught fish consumption for the North Dakota angler population (as-consumed g/day)
Percent
Habitat
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Freshwater
Marine
Marine
Marine
Marine
Marine
Species or class of fish
Walleye or Sauger
Panfish
Northern pike or Muskie
Bass
Salmon or Lake trout
Other non-purchased fish
Stream trout
Canned tuna
Breaded fish, fish sticks . . .
Shellfish
Other purchased fish
Swordfish & Shark
N
633
512
407
196
136
87
34
672
575
585
550
78
of Total
19.
12.
8.
1.
1.
1.
0.
17.
14.
11.
11.
1.
0
0
8
7
4
3
3
7
4
0
0
5
Mean
g/day
5.
3.
3.
1.
1.
2.
1.
4.
4.
3.
3.
3.
,10
,99
, 66
,47
,79
, 61
,42
,46
,25
,18
,38
,16
N Eating
Caught
Fish
633
512
407
196
136
87
34
0
0
0
0
0
Percent
of Total
Caught
42,
27,
19,
3,
3,
3,
0,
.6
.0
.7
.8
.2
.0
.6
Mean g/day
of Caught
Fish
5,
3,
3,
1,
1,
2,
1,
.10
.99
.66
.47
.79
.61
.42
-------
United States
Environmental Protection
Agency
------- |