vvEPA
     Estimated Fish Consumption Rates for the U.S.
       Population and Selected Subpopulations
               (NHANES 2003-2010)

                    Final Report
                     April 2014
                   EPA-820-R-14-002

-------
                                                       Acknowledgments
The U.S. Environmental Protection Agency (EPA) Project Manager for this project was Jeffrey D.
Bigler, who provided management and technical direction for the project. EPA was supported in
this effort by Westat under EPA Contract No. EP-C-10-023. The Westat project manager was
Rebecca Jeffries Birch who was responsible for the overall technical accuracy and quality of the
project. John Rogers served as the senior statistician. The methodology developed for the project
was externally peer-reviewed by Versar under EPA Contract No. EP-C-13-010.

-------
                                                                 Table of Contents



Chapter                                                                          Page

   1          Background	     1

   2          National Health and Nutrition Examination Survey	     4

             2.1    Survey Description	     4
             2.2    Survey Data	     5

                    2.2.1    24-Hour Recall	     5
                    2.2.2    30-Day Fish Consumption Frequency	     6

             2.3    Regions	     7

   3          Data Processing Methodology	     9

             3.1    Habitat Apportionment	     9

                    3.1.1    NHANES Fish Groupings	     11
                    3.1.2    Use of NOAA Landings Data	     12
                    3.1.3    Imported Fish and Farmed Fish	     13

             3.2    Trophic Level Assignments	     15
             3.3    Extracting Reported Amounts of Fish Consumed	     17

   4          Statistical Methodology	     21

             4.1    Overview of the NCI Method	     21
             4.2    Calculation Steps for the NCI and EPA Models	     24

                    4.2.1    NCI Method	     24
                    4.2.2    EPA Method	     27

             4.3    Simulation of the Usual Fish Consumption	     31
             4.4    Calculation of Standard Errors and  Confidence Intervals	     33
             4.5    Application of EPA Method	     35
             46    Comparison of Results from the EPA and NCI Methods	     38

                    4.6.1    Analysis of Simulated Data	     38
                    4.6.2    Confidence Intervals for Percentiles of Fish
                            Consumption	     39
                    4.6.3    Analysis of NHANES Fish Data Using Various
                            Models	     42
                                           in

-------
                                                                   Contents (continued)



Chapter                                                                          Page

   5          Results	     46

              5.1    Sample Size	     46
              5.2    Usual Fish Consumption Rates	     48
              5.3    Uncertainty	     93

                    5.3.1     Seasonally	     94
                    5.3.2     Bias in the Reported Fish Consumption	     94
                    5.3.3     Use of Standard Recipes	     94
                    5.3.4     Habitat Assignments	     95
                    5.3.5     Estimation of Usual Fish Consumption	     95
                    5.3.6     Weights for Coastal Versus Non-Coastal
                             Regions	     97

   6          Discussion	     98

   7          References	     99


Appendixes

   A          Final Habitat Apportionment	     A-l

   B          FNDDS Processing and Fish-Containing Food Codes	     B-l

   C          Supplemental Statistical Methodology	     C-l

   D          EPA Method SAS Code	     D-l

   E          UFCR Raw Weight, Edible Portion	     E-l

   F          UFCR as Prepared Weight	     F-l

   G          Unweighted Sample Sizes	     G-l

   H          Response to Peer Review Comments	     H-l
                                           IV

-------
                                                                     Contents (continued)



Tables                                                                              Page

   1          Habitat assignments of NHANES fish groups	     10

   2          NOAA landings data, clam apportionment	     14

   3          Trophic level assignments	     16

   4          Estimated moisture loss due to cooking or processing	     19

   5          A *values used for each combination of dependent variable and
              data set	     29

   6          Comparison of NCI and EPA Methods using NHANES fish
              data	     43

   7          Sample size and number reporting fish consumption, by fish type	     47

   8a         UFCR estimates (g/day raw weight, edible portion): Total fish,
              adults, 21 years and older, by demographic characteristics	     49

   8b         UFCR estimates (g/day raw weight, edible portion): Total fish,
              adults, 21 years and older, by geographic area	     50

   9a         UFCR estimates (g/day raw weight, edible portion): Freshwater
              + estuarine fish, adults, 21  years and older, by demographic
              characteristics	     51

   9b         UFCR estimates (g/day raw weight, edible portion): Freshwater
              + estuarine fish, adults, 21  years and older, by geographic area	     52

   lOa        UFCR estimates (g/day raw weight, edible portion): Marine fish,
              adults, 21 years and older, by demographic characteristics	     53
   lOb        UFCR estimates (g/day raw weight, edible portion): Marine fish,
              adults, 21 years and older, by geographic area	
   lla        UFCR estimates (g/day raw weight, edible portion): Total finfish,
              adults, 21 years and older, by demographic characteristics	     55

   lib        UFCR estimates (g/day raw weight, edible portion): Total finfish,
              adults, 21 years and older, by geographic area	     56

-------
                                                                      Contents (continued)
Tables (continued)                                                                   Page

   12a        UFCR estimates (g/day raw weight, edible portion): Total
              shellfish, adults, 21 years and older, by demographic
              characteristics	     57
   12b        UFCR estimates (g/day raw weight, edible portion): Total
              shellfish, adults, 21 years and older, by geographic area	     58

   13a        UFCR estimates (g/day raw weight, edible portion): Total trophic
              level 2 fish, adults, 21 years and older, by demographic
              characteristics	     59
   13b        UFCR estimates (g/day raw weight, edible portion): Total trophic
              level 2 fish, adults, 21 years and older, by geographic area	     60

   14a        UFCR estimates (g/day raw weight, edible portion): Total trophic
              level 3 fish, adults, 21 years and older, by demographic
              characteristics	     61

   14b        UFCR estimates (g/day raw weight, edible portion): Total trophic
              level 3 fish, adults, 21 years and older, by geographic area	     62

   15a        UFCR estimates (g/day raw weight, edible portion): Total trophic
              level 4 fish, adults, 21 years and older, by demographic
              characteristics	     63

   15b        UFCR estimates (g/day raw weight, edible portion): Total trophic
              level 4 fish, adults, 21 years and older, by geographic area	     64

   16a        UFCR estimates (g/day raw weight, edible portion): Total
              freshwater + estuarine trophic level 2 fish, adults, 21 years and
              older, by demographic characteristics	     65

   16b        UFCR estimates (g/day raw weight, edible portion): Total
              freshwater + estuarine trophic level 2 fish, adults, 21 years and
              older, by geographic area	     66

   17a        UFCR estimates (g/day raw weight, edible portion): Total
              freshwater + estuarine trophic level 3 fish, adults, 21 years and
              older, by demographic characteristics	     67

-------
                                                                     Contents (continued)
Tables (continued)                                                                   Page

   17b        UFCR estimates (g/day raw weight, edible portion): Total
              freshwater + estuarine trophic level 3 fish, adults, 21 years  and
              older, by geographic area	     68

   18a        UFCR estimates (g/day raw weight, edible portion): Total
              freshwater + estuarine trophic level 4 fish, adults, 21 years  and
              older, by demographic characteristics	     69

   18b        UFCR estimates (g/day raw weight, edible portion): Total
              freshwater + estuarine trophic level 4 fish, adults, 21 years  and
              older, by geographic area	     70

   19a        UFCR estimates (g/day raw weight, edible portion): Total fish,
              youth, <21 years, by demographic characteristics	     71

   19b        UFCR estimates (g/day raw weight, edible portion): Total fish,
              youth, <21 years, by geographic area	     72

   20a        UFCR estimates (g/day raw weight, edible portion): Freshwater
              + estuarine fish, youth, <21 years, by demographic characteristics	     73

   20b        UFCR estimates (g/day raw weight, edible portion): Freshwater
              + estuarine fish, youth, <21 years, by geographic area	     74

   21a        UFCR estimates (g/day raw weight, edible portion): Marine fish,
              youth, <21 years, by demographic characteristics	     75

   21b        UFCR estimates (g/day raw weight, edible portion): Marine fish,
              youth, <21 years, by geographic area	     76

   22a        UFCR estimates (g/day raw weight, edible portion): Total finfish,
              youth, <21 years, by demographic characteristics	     77

   22b        UFCR estimates (g/day raw weight, edible portion): Total finfish,
              youth, <21 years, by geographic area	     78

   23a        UFCR estimates (g/day raw weight, edible portion): Total
              shellfish, youth, <21 years, by demographic characteristics	     79
                                             Vll

-------
                                                                     Contents (continued)
Tables (continued)                                                                   Page

   23b        UFCR estimates (g/day raw weight, edible portion): Total
              shellfish, youth, <21 years, by geographic area	     80

   24a        UFCR estimates (g/day raw weight, edible portion): Total trophic
              level 2 fish, youth, <21 years, by demographic characteristics	     81

   24b        UFCR estimates (g/day raw weight, edible portion): Total trophic
              level 2 fish, youth, <21 years, by geographic area	     82

   25a        UFCR estimates (g/day raw weight, edible portion): Total trophic
              level 3 fish, youth, <21 years, by demographic characteristics	     83

   25b        UFCR estimates (g/day raw weight, edible portion): Total trophic
              level 3 fish, youth, <21 years, by geographic area	     84

   26a        UFCR estimates (g/day raw weight, edible portion): Total trophic
              level 4 fish, youth, <21 years, by demographic characteristics	     85

   26b        UFCR estimates (g/day raw weight, edible portion): Total trophic
              level 4 fish, youth, <21 years, by geographic area	     86

   27a        UFCR estimates (g/day raw weight, edible portion): Total
              freshwater + estuarine trophic level 2 fish, youth, <21 years, by
              demographic characteristics	     87

   27b        UFCR estimates (g/day raw weight, edible portion): Total
              freshwater + estuarine trophic level 2 fish, youth, <21 years, by
              geographic area	     88

   28a        UFCR estimates (g/day raw weight, edible portion): Total
              freshwater + estuarine trophic level 3 fish, youth, <21 years, by
              demographic characteristics	     89

   28b        UFCR estimates (g/day raw weight, edible portion): Total
              freshwater + estuarine trophic level 3 fish youth, <21 years, by
              geographic area	     90
                                            Vlll

-------
                                                                    Contents (continued)
Tables (continued)                                                                 Page

   29a        UFCR estimates (g/day raw weight, edible portion): Total
              freshwater + estuarine trophic level 4 fish, youth, <21 years, by
              demographic characteristics	     91

   29b        UFCR estimates (g/day raw weight, edible portion): Total
              freshwater + estuarine trophic level 4 fish, youth, <21 years, by
              geographic area	     92
Figures
              Comparison of NCI, EPA, and true percentiles using simulated
              data and different transformations	     40

              Comparison of confidence intervals using the NCI and EPA
              Methods	     41

              Comparison of NCI and EPA percentiles offish consumption
              rates	     45
                                            IX

-------
                                                           Background
The U.S. population is exposed to environmental contaminants through the consumption of
contaminated finfish and shellfish (Thompson and Boekelheide, 2013; National Research Council,
2000; Ahmed, Hattis, Wo Ike, and Steinman, 1993). The analysis presented here provides EPA's
recommended methodology for developing a national-level fish consumption rate (FCR) for use in
developing ambient water quality criteria as required under Section 304(a) of the Clean Water Act.

As more current data are available and new analytical methodologies have been developed, the
Office of Water has conducted a new analysis of FCR. These new FCRs were estimated using data
from the National Health and Nutrition Examination Survey (NHANES) 2003-2010. NHANES is a
continuous survey designed to collect data on the health and nutritional status of the U.S.
population. Each 2-year cycle is designed to be representative of the general U.S. population.

An individual's FCR is the expected quantity of fish consumed per unit time. For a population, there
is a distribution of FCR; some individuals consume more fish per unit time and some less. With
adequate data, we can calculate the average FCR across the population or percentiles, such as the
90th percentile (10 percent of the population has an individual FCR greater than the 90th
percentile).

Different time units can be used  to express the same rate, e.g., per day or per week.  The FCR is a
theoretical quantity and is often estimated using statistical analysis.  It may change over time, for
example, be higher in the summer than the winter. Thus, the FCR depends on the time frame (e.g.,
summer, winter, annual).

Due to the infrequent consumption of fish, the estimated FCR may be variable or imprecise. If a
person eats fish  for dinner every Friday and not at other times, the FCR is one fish meal per week
and the estimated FCR is likely to be relatively constant. If a fish meal is consumed  on average once
every 7 days but sometimes 3 days in a week and other times not for several weeks,  the estimated
FCR over a short time frame can be quite variable, even though the true FCR is constant and is the
same as in the first example. As the time  frame covered by the data gets longer, the  estimated FCR
becomes less variable. Assuming the true long-term FCR is constant over time, if the time frame

-------
covered by the data is very long, the estimated FCR becomes a relatively precise estimate of the true
long-term or usual FCR.

Assuming the FCR is constant over time, methodologies can be designed to estimate the distribution
of the true, long-term, FCR even though the data are collected over a limited time frame. We can
add the term "usual" to "fish consumption rate" (UFCR) to imply that the resulting estimates are
those that correspond to long-term averages, rather than short-term estimates and to avoid a
distinction between the true rate and the estimated rate.

In the mid-2000s, the National Cancer Institute (NCI) developed a statistical methodology to
estimate usual intake of episodically consumed foods. This method, known as the NCI Method, has
been published and statistical programs are available on NCI's web site. There are other methods
that have been developed to estimate the distribution of usual intake of episodically consumed
foods. However, the NCI Method is preferred because it accounts for days without consumption;
distinguishes within-person from between-person variation; allows  for the correlation between the
probability of consumption and the consumption-day amount; and can use covariate data to better
predict usual intake.

The NCI Method provides estimates of UFCR representing the long-term average grams of fish
consumed per day. Due to the episodic nature of fish consumption, the NCI Method models both
the probability of consumption on a given day and the amount consumed on days when some fish is
consumed. These two predicted values are then multiplied together to get a usual intake value. The
calculations using the NCI Method are very time consuming. To get estimates in a reasonable time,
EPA created a program, hereinafter referred to as the EPA Method, which approximates the results
from the NCI Method. Details of the NCI Method, the EPA Method, and how they compare are
provided in Section 4, Statistical Methods.

UFCRs were estimated for the general U.S. population, the youth population under 21 years of age,
and the adult population 21 years and older. UFCR estimates were calculated for various
subpopulations, e.g., by age, gender, race/ethnicity, income, U.S. Census region, and coastal and
noncoastal populations. We estimated UFCR for 18 different categories offish, both raw weight of
edible portion and as-prepared weight. These fish types were chosen as they represent various
categories of interest to states and tribes. For example, a coastal state  may be interested in knowing
the UFCRs of total fish and of marine and freshwater + estuarine, separately. An inland state may
only be interested in freshwater fish UFCRs. Additionally, as fish bioaccumulate toxins at different

-------
rates depending on their trophic level, UFCR were also calculated for fish by trophic level. The fish
types are the following:

      •    Total fish;
      •    Total finfish;
      •    Total shellfish;
      •    Marine fish;
      •    Freshwater fish;
      •    Estuarine fish;
      •    Freshwater +  estuarine fish;
      •    Freshwater +  marine fish;
      •    Estuarine + marine fish;
      •    Trophic level 2 fish;
      •    Trophic level 3 fish;
      •    Trophic level 4 fish;
      •    Marine trophic level 2 fish;
      •    Marine trophic level 3 fish;
      •    Marine trophic level 4 fish;
      •    Freshwater +  estuarine trophic level 2 fish;
      •    Freshwater +  estuarine trophic level 3 fish; and
      •    Freshwater +  estuarine trophic level 4 fish.

This report presents the methodologies used to extract fish consumption data from the NHANES
data sets, the habitat apportionment methodology, the trophic level assignment methodology, the
statistical methodology, and the UFCR estimates and 95 percent confidence intervals (95% CI)  of
the mean and the 25th, 50th, 75th, 90th, 95th, 97th, and 99th percentiles.

-------
         National  Health and Nutrition Examination
                                                                  Survey


2.1       Survey Description

NHANES is designed to assess the health and nutritional status of adults and children in the United
States. It is conducted by the National Center for Health Statistics (NCHS, 2013), part of the
Centers  for Disease Control and Prevention (CDC) that is responsible for producing vital and health
statistics for the United States. NHANES began in the 1960s. In 1999, the survey became a
continuous  program that examines a nationally representative sample of about 5,000 persons located
in 15 counties across the country each year.

The NHANES interview includes demographic, socioeconomic, dietary, and health-related
questions. The examination component consists of medical, dental, and physiological measurements,
as well as laboratory tests.

NHANES collects 2 days of dietary data from all participants. The first day, the data are collected in
person at the examination  portion of the survey. The second day's data are collected by telephone
interview 3  to 10 days after the in-person interview. Both interviews include a 24-hour dietary recall
section. The primary goal of the 24-hour recall is to collect a detailed list of all the foods and
beverages consumed within a 24-hour period. Food models are used to help participants estimate
the amount consumed. The in-person interview also includes a section on the frequency of
consumption offish and shellfish in the past 30 days (NCHS, 2009). Survey participants are not
asked to provide detailed recipes  for mixed  dishes. For those, standard default recipes are used.

A complex, multistage probability sampling design is used to select participants representative of the
civilian, noninstitutionalized U.S. population.

     •     Stage 1: Primary sampling units (PSUs) are selected with probability proportional to a
           measure of size (PPS). These are mostly single counties or, in a few cases, groups of
           contiguous counties.
     •     Stage 2: The PSUs are divided up into segments (generally city blocks or their
           equivalent).  As with each PSU, sample segments are selected with PPS.

-------
      •    Stage 3: Households within each segment are listed, and a household sample is
           randomly drawn. In geographic areas where the proportion of age, ethnic, or income
           groups selected for oversampling is high, the probability of selection for those groups is
           greater than in other areas.

      •    Stage 4: Individuals are chosen to participate in NHANES from a list of all persons
           residing in selected households. Individuals are drawn at random within designated age-
           sex-race/ethnicity  screening subdomains.  On average, 1.6 persons are selected per
           household. Oversampling of certain population subgroups is done to increase the
           reliability and precision of health status indicator estimates for these groups.

The NHANES data files include analysis weights to account for the complex survey design
(including oversampling), survey nonresponse, and poststratification. Weighted NHANES results
describe the U.S. Census civilian noninstitutionalized population. A person's analysis weight is a
measure of the number of people in the population represented by that sampled person.
2.2       Survey Data

2.2.1     24-Hour Recall

The 24-hour dietary recall interview data provide (1) what food items the participants ate and (2)
how much of each food item they ate. All NHANES participants are eligible for the dietary
interview component that occurs during the examination portion of the survey. The first interview is
conducted in person via a computer-assisted dietary interview software program that was developed
for NHANES. The interviewer uses a standard set of measuring guides to help the participant report
the volume and dimensions of the foods consumed. The second dietary interview is conducted via
telephone. It occurs 3 to 10 days after the first dietary interview. The participants are given a set of
measuring guides to take home and use  during the telephone interview.

The 24-hour recall data are collected using the USDA Automated Multiple-Pass Method (AMPM).
Detailed information on the method can be found on USDA's web site at
http:/7www.ars.usda.gov/Semces/docs.htm?docid=7710. The method is computerized and
research based. It uses five steps designed to assist participants with complete and accurate food
recall and  reduce respondent burden.

-------
The five steps follow:

      1.    Collect a list of foods and beverages consumed the previous day.
      2.    Probe for foods forgotten during step 1.
      3.    Collect the time and the name of the eating occasion for each food.
      4.    For each food, collect detailed description, amount, and additions (i.e., anything that
           may have been added to the food). Review 24-hour day.
      5.    Final probe for anything else consumed.

We assume that the reports of 24-hour consumption are unbiased estimates of each respondent's
true consumption.


2.2.2     30-Day Fish Consumption Frequency

The 30-day fish consumption frequency data are derived from questionnaire data that ask
participants how often in the past 30 days they consumed different fish species. These species are
clams, crabs, crayfish, lobster, mussels, oysters, scallops, shrimp, other shellfish, unknown shellfish,
breaded fish products, tuna, bass, catfish, cod, flatfish, haddock, mackerel, perch, pike, pollock,
porgy, salmon, sardines, sea bass, shark, swordfish, trout, walleye, other fish, and unknown fish.
Using these data, we can derive a variable for the number of times fish was consumed in the past 30
days by summing up the values for all 31 variables. This information improves intake estimates for
episodically consumed foods like fish, as even people who consumed fish frequently do not do so
every day; therefore, it is not always reported in 24-hour recall data. This derived frequency of
consumption can then be used as a predictor in statistical models of the probability of fish
consumption and fish consumption amount.

In 2003-2004, only children less than 6 years of age and women 16 to 49 years old were asked these
questions. As frequency of fish consumption is an important predictor in the  statistical models, we
only included these age and gender groups from NHANES 2003-2004 in the  analysis. The analysis
weights of male participants in 2005-2010 and females not in these age groups were adjusted to
account for this difference. Since they are only in three of the four cycles of NHANES their weights
were multiplied by a factor of 4/3.

-------
2.3       Regions

Patterns of fish and shellfish consumption may vary by geography, such as between U.S. residents
who live on or near the coast and those who live inland, or among regions of the United States as
defined by the U.S. Census Bureau (Mahaffey, Clickner, and Jeffries, 2009). Fish consumption
patterns may also vary by specific coast (e.g., residents near the Atlantic coast may have different
fish consumption patterns than those on the Gulf of Mexico coast). To estimate FCRs by region
and coast, we assigned NHANES respondents to U.S. Census Bureau regions and coastal  or
noncoastal status, which when combined created the following: Atlantic Coast, Northeast, Great
Lakes, Midwest, South, Gulf of Mexico, West, and Pacific Coast. The geography data were obtained
from the NCHS Research Data Center through its restricted-use data access procedures.

The geographic unit used by NHANES is a county or county equivalent; therefore our definitions of
coastal and noncoastal were limited to county  boundaries. All counties  that bordered the Pacific or
Atlantic Oceans, the Gulf of Mexico or any of the Great Lakes were defined as coastal.  Additionally,
counties that bordered estuaries and bays were defined as coastal as were counties whose centroid
was within approximately 25 miles of any coast even if not directly bordering a coast. The four
coastal regions were then defined based on nearest body of water. The  following provides
definitions of each region:
      •    U.S. Census Regions
                Midwest = OH, MI, IN, WI, IL, MO, IA, MN, SD,  ND, NE, and KS
                Northeast = PA, NY, NJ, CT, RI, MA, NH, VT, and ME
                South = DE, MD, DC, VA, WV, KY, TN, NC, SC,  GA, AL, MS, FL, LA, AR,
                OK, and TX
                West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, and HI
      •    Coastal and Inland Regions
           —   Pacific Coast = coastal counties in CA, OR,  WA, AK, and HI
           —   Atlantic Coast = coastal counties in CT, DE, DC, FL (bordering Atlantic Ocean),
                GA, ME, MD, MA, NH, NJ, NY, NC, PA, RI, SC, and VA
           —   Gulf of Mexico Coast = coastal counties in AL, FL (bordering Gulf of  Mexico),
                LA, MS, and TX
           —   Great Lakes Coast = counties bordering the  Great Lakes in MI, WI, OH, NY,
                MN, IN, IL, and PA

-------
Inland West - remaining counties in CA, OR, WA, AK, and HI and all of NM,
CO, WY, MT, ID, UT, AZ, and NV

Inland South = remaining non-coastal counties in DE, MD, DC, VA, NC, SC,
GA, AL, MS, FL, LA, and TX and all of WV, KY, TN, AR, and OK

Inland Northeast = remaining counties in PA, NY, NJ, CT, RI, MA, NH, and
ME and all of VT.

Inland Midwest = remaining counties in OH, MI, IN, WI, IL, and MN and all of
MO, IA, SD, ND, NE, and KS.

-------
                              Data Processing Methodology
3.1       Habitat Apportionment


To make estimates of FCRs for marine fish, estuarine fish, freshwater fish, and various
combinations of these types, the fish species reported as consumed by NHANES participants were
apportioned to habitats. The assignments of species were completed by a fisheries biologist.
Appendix A contains the detailed documentation of the assignments for each species.


The fish were apportioned to align with EPA's long-standing interpretation of section 303(c) (2) (A)
of the Clean Water Act that state and tribal waters should support safe consumption of fish and
shellfish and that the standards need to be set to enable residents to safely consume from local
waters the amount of fish they would normally consume from all fresh and estuarine  (including near
coastal) waters. Thus marine species that are harvested in near coastal waters were assigned to the
estuarine habitat in order to be included in the freshwater + estuarine FCR. The following decisions
concerning habitat assignments were made:

     •     Estuarine fish and shellfish include  estuarine species harvested in near-coastal areas
           (clams, mussels, crabs, lobster, shrimp) and single species that live in both marine and
           estuarine habitats (e.g., specific clam and octopus species or the single jellyfish species
           that constitutes the U.S. jellyfish fishery).

     •     Tilapia was assigned 50 percent freshwater and 50 percent estuarine, even though it is
           rare in U.S. waters, to be consistent with EPA's long-standing interpretation of section
           303(c) (2)  (A) of the Clean Water Act, as mentioned above, that the standards need to
           be set to enable residents to safely consume from local waters the amount of fish they
           would normally  consume from all fresh and estuarine (including near coastal) waters.

     •     Shrimp was assigned 17.6 percent marine and 82.4 percent estuarine. National Oceanic
           and Atmospheric Administration (NOAA) landings data show that 17.6 percent of
           shrimp harvested in 2009-2010 were "Ocean Shrimp (Oregon Pink Shrimp)," "Rock
           Shrimp," "Royal Red Shrimp," and "Marine Shrimp, Other."

     •     Salmon was assigned 96 percent marine, 0.5 percent freshwater, and 3.5 percent
           estuarine. The freshwater percent is landlocked sockeye salmon (Kokanee)  found
           natively in Alaska, Washington, and Oregon, but they have also been introduced to
           many other states for recreational fishing. The estuarine percent includes saltwater trout,

-------
           which are included in the NHANES salmon group, and the small proportion of salmon
           that are harvested in estuaries. Note that farmed Atlantic salmon were assigned to the
           marine habitat as they are produced outside of the United States in marine waters.

Table 1 presents the final proportion of each NHANES fish group that is assigned to marine,
freshwater, and estuarine habitats. Note that unspecified fish consumed was assigned the overall
average habitat apportionment of all species reported consumed. The remainder of Section 3.1
describes the habitat apportionment methodology.
Table 1.
Habitat assignments of NHANES fish groups
Proportion
Species/group
Abalone
Anchovy
Barracuda
Breaded Fish Products (e.g., fish sticks)
Carp
Catfish
Clam
Cod
Conch
Crab
Crayfish
Croaker
Eel
Fish, not specified
Flatfish
Haddock
Halibut
Herring
Jellyfish
Lobster
Mackerel
Mullet
Mussel
Octopus
Oyster
Perch
Pike
Pompano
Rockfish/Ocean Perch
Roe
Salmon
Sardine
Scallop
Scup/Porgy
Sea Bass
Shad
Shark
Shrimp
Snail
Marine
1.000
0.000
1.000
1.000
0.000
0.000
0.840
1.000
1.000
0.273
0.000
0.071
0.000
0.520
0.870
0.945
0.780
0.304
0.000
0.044
0.411
0.000
0.000
0.620
0.000
0.000
0.000
0.661
0.925
0.085
0.960
0.900
0.000
0.981
0.925
0.304
0.866
0.176
0.450
Freshwater
0.000
0.000
0.000
0.000
1.000
0.900
0.000
0.000
0.000
0.000
1.000
0.050
1.000
0.160
0.000
0.050
0.000
0.010
0.000
0.000
0.000
0.000
0.000
0.000
0.000
1.000
1.000
0.002
0.000
0.235
0.005
0.000
0.000
0.000
0.025
0.010
0.000
0.000
0.100
Estuarine
0.000
1.000
0.000
0.000
0.000
0.100
0.160
0.000
0.000
0.727
0.000
0.879
0.000
0.320
0.130
0.006
0.220
0.686
1.000
0.956
0.589
1.000
1.000
0.380
1.000
0.000
0.000
0.338
0.075
0.680
0.035
0.100
1.000
0.019
0.050
0.686
0.134
0.824
0.450
                                            10

-------
Table 1.
Habitat assignments of NHANES fish groups (continued)
Species/group
Snapper
Squid
Sturgeon
Swordfish
Tilapia
Trout
Tuna
Whelk
Whitefish
Whiting

Marine
0.981
0.800
0.000
1.000
0.000
0.106
1.000
0.000
0.877
1.000
Proportion
Freshwater
0.000
0.000
0.420
0.000
0.500
0.869
0.000
0.000
0.000
0.000

Estuarine
0.019
0.200
0.580
0.000
0.500
0.025
0.000
1.000
0.123
0.000
3.1.1     NHANES Fish Groupings
When the raw 24-hour recall data are processed by NHANES, fish species reported consumed are
grouped, and foods (e.g., Pompano, baked or broiled) are assigned food codes. The list below
presents the species of fish that are specified in the USDA Food and Nutrient Database for Dietary
Studies (FNDDS) and the additional species that are included in each group.
           Abalone
           Anchovy
           Barracuda
           Carp (bream; buffalofish; and
           sucker)
           Catfish (bullhead)
           Clams
           Cod
           Conch
           Crab
           Crayfish
           Croaker (angelfish; butterflyfish;
           drumfish; goatfish; kingfish; sea
           trout; freshwater sheepshead;
           spadefish; spot; surgeonfish;
           weakfish; weke; goo; and
           gaspergou)
           Eel
           Fish stick, patty, or fillet, not
           specified as to type (commercial
           products such as Mrs. Paul's,
           Gorton's, Van de Kamp's)
           Fish, not specified as to type
                                             Flounder (dab; fluke; halibut;
                                             sole; and turbot)
                                             Haddock (blowfish; burbot; cusk;
                                             hake; ling; monkfish; pollock; and
                                             scrod)
                                             Halibut
                                             Herring (alewife; milkfish; and
                                             shad)
                                             Jellyfish
                                             Lobster
                                             Mackerel (garfish; ono;
                                             needlefish; and wahoo)
                                             Mullet
                                             Mussels
                                             Ocean perch (bocaccio;
                                             menpachi; orange roughy; redfish;
                                             and rockfish)
                                             Octopus
                                             Oysters
                                             Perch (freshwater bass; bluegill;
                                             crappie; sunfish; and walleye)
                                             Pike (muskellunge; and pickerel)
                                            11

-------
           Pompano (akule; blackfish;
           bluefish; butterfish; dolphinfish;
           jack; mahimahi; paplo; parrot
           fish; sablefish; scad; tilefish; ulva;
           and yellowtail)
           Porgy (scup; sea bream; marine
           sheepshead; and snapper)
           Ray (skate) [not reported ever
           consumed]
           Roe
           Roe, sturgeon (caviar)
           Salmon (saltwater trout)
           Sardines
           Scallops
           Sea bass (grouper; striped bass;
           wreakfish; and bass)
Shark (dogfish and grayfish)
Shrimp
Smelt [not reported ever
consumed in the 2003-2010 data]
Snails
Snapper
Squid (cuttlefish)
Sturgeon
Swordfish (marlin)
Tilapia
Trout (cisco; lake herring;
steelhead; and whitefish)
Tuna (ahi; aku; and bonito)
Whelk
Whitefish
Whiting
This grouping of species complicates the assignment of habitat because in many cases, the grouped
fish inhabit different habitats. For example, burbot, a freshwater fish, is part of the haddock group,
which is defined by the Order Gadiformes (excluding cod). All of the other species in this group are
marine and estuarine. For these groups, we used raw (uncoded) 24-hour recall files from NHANES
from 2007-2008 (which are not publically available, and the only cycle made available to us) and
counted the number of times a species was reported. Using the haddock group as an example, in
2007-2008 blowfish, burbot, cusk, hake, ling, and monkfish were reported 0 times, pollock was
reported 10 times, scrod was reported 2 times, and haddock was reported 4 times. These counts
were then used  to assign proportions of each species in the group to the total group. No species in a
group was assigned 0 percent based on a 0 count in the files, because it may be reported in another
NHANES cycle. These species were assigned between 1 and 5 percent depending on how many
species are included in the group and how many times  other species in their group were reported
consumed. Appendix A provides  the percentages assigned to each species. The assigned proportions
were then multiplied by the habitats and summed to get the total habitat proportions for the fish
group.
3.1.2     Use of NOAA Landings Data


Other assignments were complicated by the fact that a species lives in multiple habitat types, either
at different life stages or because different species occupy different habitats. For these species,
habitat apportionment was aided by using the NOAA landings data
(http: / /www.st.nmfs.noaa.gov/commercial-fisheries /).
                                            12

-------
Table 2 is an example of the NOAA landings data for clams for 2010. To apportion the total
consumption of clams to estuarine and marine, we first assigned a habitat to each clam species listed.
According to these data, excluding the catch-all category, 84 percent of all clams landed in 2010 were
from the marine environment and 16 percent were from the estuarine environment (multiplying the
proportion of total without catch-all by the habitat proportion for each species and then summing
for each habitat). These proportions excluding the catch-all category were then applied to the catch-
all category, and the overall proportions were re-calculated.

This methodology was used to assist the apportionment of the following species: catfish, clam, crab,
flatfish, flounder, sole, halibut, lobster, mackerel, porgy, shrimp, and whiting and species in the
following food code groups: croaker, pompano, sardine, and trout.
3.1.3     Imported Fish and Farmed Fish

It is known that the United States imports a large proportion of the fish consumed from overseas.
According to NOAA Fish Watch, 86 percent of the fish consumed in the United States are
imported (http://www.fishwatch.gov/wild seafood/outside the  us.htm). The top imported species
are shrimp, freshwater fish (mainly tilapia and catfish), tuna, salmon, groundfish (e.g., cod, haddock,
flounder), crab, and squid. As marine fish are not harvested from U.S. waters for which states would
be developing water quality standards, the issue of importation for these species is not relevant.
However, shrimp is the most commonly consumed fish by U.S. consumers. It is unknown whether
the proportion consumed that was harvested in non-U.S. waters is distributed equally across the
distribution offish consumers. For example, it is possible that high fish consumers eat  more locally
caught fish as they may be more likely to be recreational or subsistence  fishers. For the purposes of
developing UFCR, we assumed that all estuarine, freshwater, and near coastal fish that were
consumed were from U.S. waters. The reason for this is that standards need to be set to enable
residents to safely consume from local waters the amount of fish they would normally consume
from all fresh and estuarine (including near coastal) waters.
                                            13

-------
Table 2. NOAA landings data, clam apportionment

Clam, Arc, Blood
Clam, Atlantic Jackknife
Clam, Atlantic Surf
Clam, Butter
Clam, Manila
Clam, Northern Quahog
Clam, Ocean Quahog
Clam, Pacific Geoduck
Clam Pacific Littleneck
Clam, Pacific Razor
Clam Pacific Gaper
Clam, Quahog
Clam, Softshell
Clams or Bivalves
Total Pounds
Total Pounds without catch-all
Without catch-all
Total
Pounds
landed, 2010
23,738
67,334
37,465,740
15,133
937,915
4,406,313
31,704,091
2,777,529
26,811
138,826
6,061
634,131
4,278,356
6,980,468
89,462,446
82,481,978
Proportion
Estuarine
Proportion
Marine
Proportion
Estuarine
Proportion
Marine
Proportion of total
Proportion of (without catch-all
total category) Habitat
0.0003
0.0008
0.4188
0.0002
0.0105
0.0493
0.3544
0.0310
0.0003
0.0016
0.0001
0.0071
0.0478
0.0780


0.15971
0.84029
0.15973
0.84027
0.0003
0.0008
0.4542
0.0002
0.0114
0.0534
0.3844
0.0337
0.0003
0.0017
0.0001
0.0077
0.0519





Estuarine & marine harvested near coast
Estuarine
Marine
Estuarine & marine harvested near coast
Estuarine
Estuarine
Marine
Estuarine & marine harvested near coast
Estuarine & marine harvested near coast
Marine
Estuarine & marine harvested near coast
Estuarine
Estuarine & marine harvested near coast
estuarine & marine (catch-all category)




Habitat
percent
100E
100E
100M
100E
100E
100E
100M
100E
100E
100M
100E
100E
100E
16E/84M





-------
There are similar issues with farmed freshwater fish. Freshwater fish can be farmed in man-made
ponds or tanks for which the states will not be developing water quality standards. However, as
noted above in the discussion concerning imported fish, the proportion of freshwater fish
consumed that is farmed may not be evenly distributed across the distribution of consumption.
Again, it is possible that high fish consumers are eating locally caught fish through recreational or
subsistence fishing and thus eating a smaller proportion of farmed fish than those at the middle and
low end of the consumption distribution. Therefore farmed species were assumed to be wild caught.
This allows residents to safely consume from local waters the amount of fish they would normally
consume from fish farms.
3.2       Trophic Level Assignments

The trophic level of an organism is the place it occupies in the food web. Organisms with higher
trophic levels have higher exposures to environmental contaminants.

      •    Trophic level 1 organisms are primary producers (plants and algae).
      •    Trophic level 2 organisms are herbivores, also called primary consumers.
      •    Trophic level 3 organisms are carnivores that consume herbivores.
      •    Trophic level 4 organisms are carnivores that consume other carnivores.
      •    Trophic level 5 organisms are the apex predators.

Trophic level assignments were made using the data provided in the following documents: (1)
Methodology for Deriving Ambient Water Quality Criteria for the Protection of Human Health (2000), Table 6-4
(U.S. Environmental Protection Agency, 2003) and (2) Trophic Level and Exposure Analyses for Selected
Piscivorous Birds and Mammals:  Volume III: Appendices (U.S Environmental Protection Agency, 2002b).

For species that were not in those documents, we performed a search of literature available on the
Internet and applied the same rules that were described in the December 2003 document:

      •    For game fish, data were used for edible size ranges (about 20 cm [8 inches] or larger).
      •    For species where multiple size ranges were available, preference was given to the larger
           specimens in determining the species trophic level.
                                            15

-------
           Trophic level 2 was assigned to a species if appropriate trophic level data ranged
           between 1.6 and 2.4; trophic level 3 if trophic level data ranged from 2.5 to 3.4; and
           trophic level 4 if trophic level data were 3.5 or higher. This is consistent with the
           approach taken in the Great Lakes Water Quality Initiative guidance (U.S.
           Environmental Protection Agency,  1995).

           In determining NHANES fish grouping trophic level assignments, best professional
           judgment was used. If the vast majority of the species in a group are within one trophic
           level, then that trophic level is assigned. If species span two levels  it is split 50-50. For
           example, the NHANES grouping for catfish includes four species that are assigned to
           trophic level 3 and three species assigned to trophic level 4. Thus,  it is assumed that half
           (50 percent) of consumption in the  catfish NHANES grouping is  from TL3 and half
           from TL4. Other fish this rule applies to are croaker, flatfish, and shrimp.
Table 3 presents the final trophic level assignments.

Table 3.       Trophic level assignments
Fish species/group
ABALONE
ANCHOVY
BARRACUDA
BREADED FISH PRODUCTS (e.g., fish sticks)
CARP
CATFISH
CLAM
COD
CONCH
CRAB
CRAYFISH
CROAKER
EEL
FISH NOT SPECIFIED
FLATFISH
HADDOCK
HALIBUT
HERRING
JELLYFISH
LOBSTER
MACKEREL
MULLET
MUSSEL
ROCKFISH/OCEAN PERCH
OCTOPUS
OYSTER
PERCH
PIKE
POMPANO
PORGY/SCUP
ROE
SALMON
Proportion
Trophic level 2
1
0.5
0
0
0
0
1
0
1
0
0
0
0
0
0
0
0
0
1
0
0
1
1
0
0
1
0
0
0
0
0
0
of Assigned to Trophic
Trophic level 3
0
0.5
0
0.5
1
0.5
0
0
0
1
1
0.5
0
0.5
0.5
0
0
1
0
1
0
0
0
0
0.5
0
0
0
0
0
0
0
Level
Trophic level 4
0
0
1
0.5
0
0.5
0
1
0
0
0
0.5
1
0.5
0.5
1
1
0
0
0
1
0
0
1
0.5
0
1
1
1
1
0
1
                                              16

-------
Table 3.       Trophic level assignments (continued)
Fish species/group
SARDINE
SCALLOP
SEA BASS
SHAD
SHARK
SHRIMP
SNAIL
SNAPPER
SQUID
STURGEON
SWORDFISH
TILAPIA
TROUT
TUNA
WHELK
WHITEFISH
WHITING
Proportion
Trophic level 2
0
1
0
0
0
0.5
1
0
0
0
0
1
0
0
1
0
0
of Assigned
to Trophic Level
Trophic level 3 Trophic level 4
1
0
0
1
0
0.5
0
0
0.5
0
0
0
0
0
0
1
1
0
0
1
0
1
0
0
1
0.5
1
1
0
1
1
0
0
0
3.3       Extracting Reported Amounts of Fish Consumed

The FNDDS is the underlying database used to code dietary intakes for NHANES. It is a database
of foods, their nutrient values, and gram weight equivalents for various ingredients in the foods. For
each new version of FNDDS, foods, gram weights, and nutrient values are reviewed and updated to
reflect the U.S. food supply by incorporating new foods based on what is reported in the survey and
updating existing entries.

In FNDDS, each food is given an 8-digit food code. The first digit identifies one of nine major food
groups. The second, third, and fourth digits identify increasingly more specific subgroups. Most fish-
containing foods are found under "26 — Fish and Shellfish," "27 — Meat, Poultry, Fish with nonmeat
items," and 28, which includes soups and frozen meals. Other fish-containing foods  are found under
"5 — Grains" such as seafood pizza and pasta dishes and "7 — Vegetables" for dishes that are mainly
vegetables but also contain fish and/or shellfish.

The NHANES 24-hour recall data include these same food codes for each reported  food consumed;
therefore the reported foods can be merged to the FNDDS files to obtain recipe information. The
FNDDS files are available from  the Agriculture Research Service of the USDA (USDA, 2006;
USDA, 2008; USDA 2010; Ahuja et al., 2012). FNDDS includes several files (or tables), including a
file that is linked to the USDA National Nutrient Database for Standard Reference (SR) that
                                           17

-------
provides recipes for reported foods. For example, the standard recipe for "Perch, baked or broiled,"
consists of the ingredients (1) fish, perch, mixed species, raw; (2) margarine, stick, salted; (3) lemon
juice, raw; and (4) salt, table. The FNDDS-SR link file provides weights in grams for each ingredient
in each recipe. In the above example, these amounts are 907.2 grams of fish, 28.2 grams of
margarine, 30.5 grams of lemon juice, and 6 grams of table salt. From these amounts, the fraction by
weight of the recipe that is fish can be calculated. In the example, 907.2 / (907.2+28.2+30.5+6) =
0.933 grams of prepared fish per gram of recipe.

The FNDDS files were searched to find all food codes that contain finfish and/or shellfish. These
records were then processed to determine the weight of each fish ingredient as a fraction of the
weight across all ingredients in the recipe. The recipe ingredients may be raw, canned (cooked), or
otherwise processed before being put into the recipes. The FNDDS description of each ingredient
generally includes the processing before the ingredient is added to the recipe. After the dish is
prepared from the ingredients, the food dish may have additional cooking or processing, such as
baking. This processing is often described in the FNDDS food description.

As NHANES participants report the amount consumed "as prepared" (which is converted to a
weight, in grams, in the NHANES file), it is relatively easy to estimate the grams of prepared fish
that is consumed. However, because cooking can change the moisture content of the fish,
calculating the grams of raw fish consumed requires to a weight conversion based on the likely
moisture loss due to cooking. The  calculation of the weight of as-prepared and raw fish consumed
are based on the following:

      •     Estimates of the moisture loss associated with various cooking methods.
      •     Assuming the weight of fish as a proportion of the weight of the food is the same for
            the recipe in the FNDDS files as in the final as-prepared dish. In effect, we assume the
            proportional weight loss due to cooking of the prepared recipe as the same for the fish
            and non-fish ingredients.
      •     If the recipe specifies two cooking steps, one  for the fish used in the recipe (for
            example, using canned ingredients) and one for the prepared recipe (for example,
            baking before serving), assuming a moisture loss associated with the cooking method
            with the most moisture loss.

The uncooked amount of fish was determined using the recipe databases, which list the amount of
each ingredient in the food code. The weight of each ingredient as a fraction of the weight of the
recipe was calculated, as above. During this data processing, each fish ingredient in the recipe was
apportioned to marine, estuarine, and freshwater habitat and to trophic levels 2, 3, and 4, as
                                             18

-------
discussed in Sections 3.1 and 3.2. As many food codes comprise multiple fish species, each of these
values was summed, along with total fish percent, across all fish-containing ingredients to get total
values for each habitat, trophic level, and total fish for each fish-containing food code.

The adjustment factors for cooking by dry heat, moist heat, and frying and the adjustment factors
for canning and restructured fish are also used in the analysis of the CSFII data published in 2002
(EPA, 2002a) and in the Mercury Study Report to Congress (EPA, 1997, Volume 4). These cooking and
processing methods represent 90 percent of all reported fish consumed. The percent moisture  loss
for the remaining cooking and processing methods (dried, kippered, smoked, salted, and pickled) are
estimated using the FNDDS "MoistNFatAdjust" file. This file provides the percent moisture and fat
loss or gain due to cooking, by food code; there is a file specific to each NHANES release. These
adjustments are used in the  calculations of nutrient intake (e.g., calcium, protein) for NHANES
participants. However, for many food codes they are set to zero because the FNDDS recipe uses a
cooked or processed fish  as the ingredient, and no further adjustments were needed for nutrient
intake calculations. We calculated the mean value of moisture loss for the remaining cooking
methods for those fish food codes that did not have a 0 value, using this file. Table 4 provides  the
adjustments applied by cooking and processing method. For unspecified cooking method,
approximately 5 percent of all reported fish consumed, an average adjustment across all reported
fish food codes was applied (22 percent moisture loss).
Table 4.       Estimated moisture loss due to cooking or processing

  Cooking/Processing method    Percent moisture loss
 Dried                                57
 Kippered                             46
 Smoked, (other than salmon)            36
 Salted                               33
 Canned                              25
 Cooked, dry heat                      25
 Restructured                          25
 Cooked, moist heat                    21
 Smoked salmon                       17
 Pickled                              16
 Fried                                12
 Raw                                  0

There is uncertainty associated with these  values. They are average values of moisture loss given the
various processing and cooking methods.  If participants cooked their fish a bit longer, then the
                                             19

-------
moisture loss would be a bit greater than average, and if they cooked it a bit less, the moisture loss
would be a bit less than average.

Appendix B provides a detailed description of how the fish foods were abstracted and processed
from FNDDS and it provides the final number of grams of raw weight, of the edible portion fish
per 1 gram of the final prepared recipe in each fish-containing food code reported in the NHANES
data 2003-2010. It contains the values for total, marine, estuarine, freshwater, and trophic levels 2
through 4 fish.

As an example calculation, the standard recipe for food code 27250400 "shrimp cake or patty"
contains  0.475 grams of shrimp per gram of total recipe. The shrimp ingredient in the recipe is
canned; therefore moisture loss is estimated to be 25 percent. We divide .475 by 0.75 to get the
grams of raw fish in 1 gram of the final prepared recipe, which is .633 g. Shrimp was apportioned to
the habitats as 17.6 percent marine and 82.4 percent estuarine. We then multiply these percentages
by the grams of raw shrimp in 1  gram of the final prepared recipe, 0.176*.633 = .111 grams raw
marine fish in 1 gram of the final prepared recipe, and 0.824*.633 = .522 grams raw estuarine fish in
1 gram of the final prepared recipe. Similar calculations  are made to determine grams of raw fish by
trophic levels in 1 gram of the prepared recipe. These amounts are then multiplied by the reported
grams of food code 27250400 consumed by the participants and summed across all fish-containing
food codes reported by each respondent to get the reported 24-hour intake.

Using this example, we can see the uncertainty added to the estimates by using standard recipes.
Recipes used for shrimp cakes could vary from the assumed 47.5 percent fish by weight composition
by using  more or less eggs or bread crumbs. Or the shrimp cake could have been prepared using raw
shrimp that was fried, instead  of canned shrimp,  which would change the weight loss estimate to 12
percent. Thus a participant who reported consuming a shrimp cake probably consumed somewhat
less or somewhat more than is estimated through the calculations. Nevertheless, these data are the
best data available on a nationally representative sample.

An additional complication is  that recipes may include two steps of processing; for example, a
salmon loaf may list canned salmon as  an ingredient, but it is then mixed with other ingredients and
baked. Canning and baking have different moisture losses in Table 4. It was decided to use the
adjustment that indicates the greatest moisture loss and apply that to the estimation of raw weight.
In some  recipes the second processing step is not categorized. We reviewed these and were able to
impute the logical unreported  process (e.g., pizza is baked, soup is wet cooked in moist heat) for
many recipes; those that remained uncategorized were assumed to  have the average moisture
reduction described above.
                                             20

-------
                                       Statistical Methodology
The NCI Method (Tooze et al., 2006; Tooze et al., 2010) is the preferred approach for estimating
usual dietary intake, such as usual fish consumption. NHANES has data for many individuals,
allowing fitting models with many parameters. With many individuals and many parameters, the
computation time to implement the NCI Method was unacceptable. Therefore, EPA developed an
alternate approach to estimate the usual fish consumption that requires relatively little computation
time and provides a good approximation to the results from the NCI Method. The following
sections describe both the NCI and the EPA Methods and compare the two approaches. Appendix
C provides some further discussion and Appendix D provides the macro code for estimating
parameters and simulating the UFCR.
4.1       Overview of the NCI Method

The NCI Method can be used to estimate the distribution of usual intake for a population or
subpopulation. Two steps are required to estimate usual intake:

      1.    Fit the NCI model to the reported consumption data.
      2.    Calculate the usual intake from the model parameters.

The premise of the NCI model (step 1 above)  is that usual fish intake is equal to the probability of
consumption on a given day times the average amount consumed on a day when some fish is
consumed, i.e., a "consumption day." For episodically consumed foods, such as fish, the NCI model
consists of two parts, or sub-models. The first sub-model estimates the probability of consumption
using logistic regression with a person-specific random effect. The second sub-model uses linear
regression on a transformed scale to estimate the consumption-day amount, also with a person-
specific random effect. The two sub-models are linked by allowing the person-specific effects to be
correlated and by including common predictors in both sub-models. Data from one or more non-
consecutive 24-hour recalls provide the values  for the dependent variable. At least a subset of the
population (generally 50 or more individuals) needs to have reported fish consumption from two or
more 24-hour recalls. Predictors related to either the probability of consumption or consumption
amount, such as gender, age, race, and income can be included in the modeling. In most cases, the
                                           21

-------
most important predictor is a measure of frequency of consumption of the food of interest (in this
case, fish) obtained from a food frequency questionnaire.

In the second step, the parameters from the NCI model are used to estimate population and
subpopulation distributions of usual fish intake. The NCI Method calculates the distribution using
simulated values for the probability of fish consumption and the mean consumption amount. The
usual fish consumption (or usual fish intake) is the product of the probability offish consumption
and the mean amount of fish consumed, when it is consumed.

Evidence for the validity of the NCI Method has  been published in a series of papers in the Journal of
the American Dietetic Association, Statistics in Medicine, and Biometrics (Dodd et al., 2006; Tooze et al.,
2006; Tooze et al., 2010; Kipms et al., 2009).

The NCI Method is an improvement over other methods designed to estimate usual intake of
episodically consumed foods because it:

     •    Accounts for reported days without consumption or for consumption-day amounts that
           are positively skewed;
     •    Distinguishes within-person from between-person variation;
     •    Allows for the correlation between the probability of consumption and the
           consumption-day amount; and
     •    Relates covariate information to usual intake.

The sub-model predicting the probability of fish consumption in a 24-hour period is a logistic
regression model. The logistic regression model is commonly used to model the probability of an
event, such as consuming fish. The model assumes the logit-transformed probability is a linear
function  of various continuous and discrete predictor variables. The logit transformation is
commonly used as the link between the continuous predictors and the probability of a discrete
outcome, as in logistic regression. The sub-model has two variance components, person-specific
random effects for an individual's long-term probability of consuming fish and within-individual
binomial  variation between days when fish was or was not consumed. The logit-transformed person-
specific random effects are assumed to be normally distributed.

The amount sub-model involves a Box-Cox transformation such that the transformed amount of
fish consumed in a 24-hour recall is reasonably normally distributed. The Box-Cox transformation is
                                            22

-------
a power transformation, such as raising the amount to the 'A power (taking the fourth root),
followed by rescaling to keep the variance relatively constant. In the transformed units, the amount
sub-model has two variance components, person-specific random effects for an individual's long-
term mean fish consumption and within-individual differences in the amount of fish consumed on
different days. In the transformed units, the person-specific mean fish consumption and the within-
individual daily fish consumption are assumed to have normal distributions.

The person-specific random effects in the two sub-models may be correlated, for example, those
with a higher probability of consuming fish in a 24-hour period may also tend to consume larger
daily amounts of fish. The assumption that the random effects are normally distributed is a
characteristic of the model which is not directly testable. However, the distribution of the Box-Cox-
transformed reported consumption amounts is roughly normally distributed, suggesting that the
assumption is at least reasonable.

Both sub-models can have additional predictors, such as person-specific demographic characteristics
and reported frequency of fish consumption.  In addition, the model can incorporate the following
within-person predictors: (1) differences between weekends (Friday to Sunday) and weekdays
(Monday to Thursday),1 and (2) consistent differences between the first 24-hour recall and the
second 24-hour recall in NHANES (the first was completed in person and the second was
completed by phone).

We consider the NCI Method as the preferred method for estimating fish consumption rates and
believe the results to have minimal bias. However, with large sample sizes and many predictors, the
computation time required to run the NCI Method and calculate confidence intervals was
unacceptable given the schedule and budget. Additionally, our preferred model has more predictors
than the NCI Method is set up to handle. The EPA Method was developed to provide acceptably
unbiased estimates within a  reasonable computation time. We are using non-publically available data
from NHANES that can only be accessed on site at NCHS. This precludes our use of alternative
computing scenarios that might reduce the computation time.

The following illustrates the time savings. We ran a simplified model with 4 main effects (age,
race/ethnicity, income, and  frequency offish  consumption). The NCI Method took 9.5 hours for
1 The NCI Method includes Friday as part of the weekend. A study of CSFII data showed that intake on Friday was
 more similar to Saturday and Sunday than to the rest of the weekdays, Monday through Thursday (Haines, Kama,
 Guilkey, and Popkin, 2003).
                                             23

-------
one run of one fish type. To obtain an estimate of the precision of the estimates, we need to run the
model 65 times, one for each replicate weight. This gives us an estimated time of over 25 days of
continuous computer time for each fish type. There are 18 fish types. Therefore, to obtain estimates
for all fish types would take 450 days. The EPA Method took 1.5 minutes to run the same model,
approximately 1.5 hours for each fish type.

The NCI Method can be implemented using two SAS macros (programs) available from the NCI
web site (the MIXTRAN and DISTRIB macros). The equations fit using the NCI macros are
presented below. More information concerning the NCI Method can be found in Tooze et al., 2006,
Tooze et al., 2010, and on NCI's web site at:
http: / /appliedresearch.cancer.gov/diet/usualintakes /method.html.

EPA created a SAS macro to approximate the results from the NCI Method while taking
considerably less time for the calculations, referred to as the EPA Method in this report. The
following sections describe both the NCI and EPA Methods. The macro code for estimating
parameters and simulating the usual fish consumption is in Appendix D.
4.2       Calculation Steps for the NCI and EPA Models

4.2.1     NCI Method

In the NHANES data, each individual (indicated by i) has results from one or two 24-hour dietary
recalls (indicated byyjy = 1 or 2). In the data, most individuals have two 24-hour recalls; only a
portion of individuals have 24-hour recalls reporting fish consumption, and a smaller portion have
two 24-hour recalls with reported fish consumption. Using the i andy subscripts, the following
describes the statistical model fit using the NCI MIXTRAN macro. The parameters and variables
are described below. In these equations, the parameters for the probability model are represented by
7T, the parameters for the amount model are represented by a, the standard deviations of the
variance components are represented by a.

Data for each individual and 24-hour recall, extracted from the NHANES data files follow:

     •     AIJ is the reported grams of fish consumed (zero if no fish consumption was reported
           in the 24-hour recall).
                                           24

-------
            Wij is an indicator of whether the 24-hour recall was for a weekday (Wjy = 0) or
            weekend (W^ = 1).

            Sij indicates if the 24-hour recall was the first (in-person, 5^ = 0) or the second (by
            phone, 5j2 =  1) dietary recall.

            Xik represent k individual level covariates (demographic variables; see Section 4.5 for
            additional details).
Transformed data:
            GIJ is an indicator for reported fish consumption in a 24-hour recall (1 indicates fish
            consumption, otherwise 0). This is the dependent variable in the logistic probability
            model.
                                            °
                                                AtJ>0
      •    TIJ is the amount of fish consumed after being transformed using a Box-Cox power
           transformation. This is the dependent variable for the linear regression model predicting
           the consumption-day amount of fish consumed.

Parameters for the probability sub-model:

      •    n0 is the intercept parameter for the probability model.
                is a vector of regression parameters for the k person-specific covariates in the
           probability model.

           nw is the regression parameter for the difference between weekend and weekday days.

           ns is the regression parameter for the difference between the second 24-hour recall and
           the first 24-hour recall.

           TTj is the person-specific random effect for the probability model, assumed to be
           normally distributed on the logit scale. This value has theoretical meaning but is not
           observed.

           PIJ is the probability of fish consumption for a 24-hour recall. This value has theoretical
           meaning but is not observed.
                                             25

-------
Parameters for the amount sub-model:

      •    A is the power used for the Box-Cox transformation.

      •    a0 is the intercept parameter for the amount model.

      •    CiXk is a vector of regression parameters for the k person-specific covariates in the
           amount model.

      •    aw is the regression parameter for the difference between weekend and weekday days.

      •    as is the regression parameter for the difference between the second 24-hour recall and
           the first 24-hour recall.

      •    tfj is the person-specific random effect for the amount model, assumed to be  normally
           distributed. This value has theoretical meaning but is not observed.

      •    (%ij is the within-person random effect for the amount model representing different
           amounts for fish consumed on different days, assumed to be normally distributed. This
           value has theoretical meaning but is not observed.

Variance and correlation parameters:

      •    (?i is the variance of the person-specific random effect in the probability model (TTj).

      •    <72 is the variance of the person-specific random effect in the amount model (tfj).

      •    p is the correlation between TTj and a^.

      •    <73 is the variance of the within-person  random effect in the amount model (<*Ģ_/).

The NCI macro fits some preliminary models to obtain approximate parameter estimates to use as
starting values for the NLMIXED procedure (a SAS procedure that fits non-linear mixed models),
which fits the following set of equations simultaneously, using maximum likelihood. The following
equations describe the model fit using the NCI Method. The tilde (~) symbol can be read as "is
distributed as."
                               /  p. .  \
             Logit(Pij) = log    _t;    = n0 + Xiknxk + nt
                               V1   M/V
                                   Cij~Binomial(l,
                              A^ - 1
         // Aij>0 then Ttj = - - - = a0 + Xikaxk + Wtjaw + S^-cis + atj + at
                        a^-BivariateNormaliO  0],
                                              \
                                            26

-------
The NCI model can be fit assuming the probability of consuming fish and the amount of fish
consumed, if consumed, are uncorrelated, i.e., p = 0. However, we assume these values may be
correlated and thus specified a correlated model when using the NCI Method to compare the NCI
and EPA Method results.
4.2.2     EPA Method


The EPA and NCI Methods differ as follows:

      •    As part of fitting the model, the NCI Method finds the power transformation (A) that
           best fits the data and is consistent with the assumption that the variance components
           are normally distributed, as judged by maximum likelihood. The EPA Method finds the
           power transformation  that makes the transformed fish consumption amounts (T^y)
           roughly normally distributed as judged by the correlation between the transformed
           amounts and the expected values for a normal distribution. The power with the highest
           correlation is then used when fitting the amount sub-model.

      •    The  EPA Method assumes the person-specific random effects in the probability and
           amount sub-models are uncorrelated. The assumption of zero correlation is for
           computational convenience, not because these values should be uncorrelated. When
           using the NCI Method, we assumed these random effects may be correlated and let the
           NCI algorithm estimate the correlation. The correlation estimates were generally close
           to zero.

      •    When assuming the person-specific random effects are uncorrelated, the two sub-
           models can be fit separately rather than simultaneously. The EPA Method fits the two
           models separately.

      •    For the probability sub-model, the NCI model fits the parameter values and random
           effects in one non-linear mixed model. The EPA Method approximates that approach
           using two steps: (1) using logistic regression without random effects to estimate the
           parameter values and calculate predicted values (in the logit scale); and (2) using a non-
           linear mixed model to  fit the random effects using only the predicted values from the
           previous logistic regression as a predictor.

      •    For the amount sub-model, the NCI Method fits the parameter values, transformation,
           and random effects in  one non-linear mixed model. The EPA Method approximates
           that approach by (1) selecting the transformation as described above; (2) using linear
           regression without random effects to estimate the parameter values and calculate
           predicted values; and (3) using a non-linear mixed model to fit the random  effects using
           the predicted values from linear regression as the only predictor.
                                            27

-------
The equations for the EPA Method are described below using the same notation as above for the
NCI Method. The EPA Method has the following steps:

      1.    Estimate A.

      2.    Use logistic regression to predict the probability of fish consumption as a function of
           various predictors; save the predicted values.

      3.    Use a non-linear mixed model to estimate the person-specific variance component using
           the predicted values from logistic regression as the predictor.

      4.    Use linear regression to predict the Box-Cox-transformed amount of fish consumed,
           when consumed, as a function of various predictors; save the predicted values.

      5.    Use a non-linear mixed model to estimate the variance components for the amount
           model using the  predicted values from linear regression as the predictor.

In the NCI Method, the maximum likelihood procedure finds the best transformation, defined byA,
consistent with the data and the assumption that the random effects are normally distributed. In the
EPA model, A = A* is set prior to fitting the amount sub-model, using the following steps:

      1.    Calculate normal scores associated with each observation by first, ignoring the
           distinction between the first and second recall; second, for amounts greater than zero,
           summing the weights across tied values (values with the same reported amounts) to get
           one record for each unique amount (Ar~) and the associated weight (Wr~); third, sorting
           the R unique amounts from smallest to largest; fourth, calculating cumulative weight for
           the each unique value, Sr = Ģ^=1Wm; and fifth, calculating the normal scores as
      2.    Using values of A* which are multiples of 0.01 between -0.20 and 0.30, find the A* value,
           which maximizes the Pearson correlation between Zr and

                                       ^log(Ar}G  A* = 0

                                             	   i* _/. n
                                             „,. „   A.  f U
           where G is the geometric mean of Ar. This form of the Box-Cox transformation allows
           A* =  0, corresponding to using a log transformation.

      3.    If A*  =  0  then set A* = 0.005 (this case was encountered for marine tropic level 2 fish).
                                            28

-------
 Table 5 shows the A* values used for each of the dependent variables.

 Table 5.       A* values used for each combination of dependent variable and data set

_ Dependent variable _ A^
 All fish                                                                               0.21
 Marine fish                                                                            0.24
 Estuarine fish                                                                          0.13
 Freshwater fish                                                                        -0.04
 Finfish                                                                               0.25
 Shellfish                                                                              0.11
 Freshwater + estuarine fish                                                              0.11
 Marine + estuarine fish                                                                  0.21
 Marine + freshwater fish                                                                0.21
 Trophic level 2 fish                                                                     0.11
 Trophic level 3 fish                                                                     0.16
 Trophic level 4 fish                                                                     0.20
 Marine trophic level 2 fish                                                               0.005
 Marine trophic level 3 fish                                                               0.09
 Marine trophic level 4 fish                                                               0.23
 Freshwater + estuarine trophic level 2 fish                                                 0.08
 Freshwater + estuarine trophic level 3 fish                                                 0.12
 Freshwater + estuarine trophic level 4 fish _ -0.05


 For two types offish consumption (freshwater and freshwater plus estuarine trophic level 4), A* was
 less than zero. The NCI Method constrains A to be greater than 0.01. As a result, the results from
 the NCI Method and the EPA Method differ somewhat when A*  < 0.01.

 The transformed consumption amounts for 24-hour recalls with reported fish consumption are
 shown in the  following equation:
                                        T.. — —. _
                                          11 ~     A*

 To estimate the probability of consumption, the following logistic regression model was fit using the
 SAS SURVEYLOGISTIC procedure and the NHANES survey weights, strata, and PSU variables.
 This logistic regression model predicts the probability of consuming fish in a 24-hour recall without
 considering a person-specific random effect; B[; is the linear predictor of the logit transformed
 probability. The apostrophes indicate values from the logistic model that has no random effects.
                                               29

-------
                        = log  —^—  = n'0 + Xikn'xk + Wtjn^ + S^ = B'tj
                              V1   My/

The person-specific random effect is included by assuming the predicted logit (5jy) when including
the random effect is proportional to the predicted logit when excluding the random effect (5t'y).
This approximation is justified in Appendix C. The following non-linear mixed model is fit to
estimate the variance of the person-specific random effect and the inflation factor (/?) for scaling the
parameter estimates from the model above.

                                          f  P--   \
                        Logit(Pij) = log  —^— } = p*B'ij+ nt
                                          V1   My/
                                   7Tj~JVorma/(0, of)
                                   Cij~Binomial(l, Pjy)

The SAS SURVEYREG procedure is used to fit the amount sub-model assuming no person-specific
random effect. The variance of the regression error  (174) is the combination of the variance of the
person-specific random effect and the within-person variation. DIJ is the linear predictor of the
transformed amount of fish consumed.

     // AtJ > 0 then Ttj = -ąjf - = a0 + Xikaxk + Wi}aw + Si}as + ofy = Dtj + ofy
                                   a-y~JVorma/(0,  0 then TtJ = -    - = DtJ + atj + at
Because of different estimation methods, the parameters calculated using the NCI Method are
slightly different than those from the EPA Method.
                                            30

-------
4.3       Simulation of the Usual  Fish Consumption

The distribution of usual fish consumption can be calculated from the model parameters. Due to the
complexity of the model, the direct calculation of the distribution of usual fish consumption
involves numerical integration and is relatively complex. The integration is simplified by (1)
simulating values of usual fish consumption and (2) calculating mean and percentiles offish
consumption rates from the simulated values. When using simulations, the estimated fish
consumption rates have a small random component that can be reduced by increasing the number
of simulations. The default number of simulations in the NCI  DISTRIB macro is 100. Analysis of
preliminary results showed that the precision of the parameter estimates increased as the number of
simulations increased; however, the precision was similar when using either 50 or 100 simulations.
The final analysis used 100 simulated fish consumption values  for each NHANES respondent. The
NHANES data set provides the population distribution of the independent predictors in the
probability and amount sub-models. For each NHANES respondent, the simulated values represent
possible fish consumption rates for a respondent with the same independent predictors as the
NHANES respondent.

Because usual fish consumption is different from reported fish consumption, the equation used to
simulate usual fish consumption is slightly different from the equation fit to the data from the 24-
hour recalls.  The equation fit to the data was modified as follows to simulate usual fish
consumption:

      •    The simulated values reflect a standard week (3 weekend days and 4 weekday days)
           rather than the distribution of weekday and weekend recalls in the data. Given Friday is
           part of the weekend, the average for the standard week would be 4/7 x (weekday
           average) + 3/7 x (weekend average). Since the weekend parameter models the
           difference between the three weekend days and the four weekday days,  this average can
           be obtained by setting the parameter for the weekend variable to 3/7.
      •    The simulated values assume the first (in-person) 24-hour recall is unbiased by ignoring
           the difference between the first and second recall, i.e., as = TTS = 0.
      •    The simulated values do not include the within-person variation, i.e., binomial variation
           within persons in the probability model and the within-person variation in the amount
           model.

The equations  for simulating usual fish consumption use the parameters estimated from the models
predicting probability of reported  fish consumption and the amount for fish consumed, when
consumed.
                                            31

-------
In the equations below, the V subscript represents the simulation number (V - 1 to 100). The
following equations are used by the NCI DISTRIB macro to simulate an individual's long-term
probability of fish consumption (Qvt) and transformed long-term mean fish consumption when fish
is consumed (Tyi)', the logistic function is the inverse of the Logit function.
                       Qvi = Logistic \n0 + Xiknxk + nvi + -nwj
                             Tvi = a0+ Xikaxk + avi + -nw
                 [nvi   avi]~BivariateNormal[[Q  0],
                                               \       Lp0i02     02
A slightly modified version of these equations is used in the EPA procedure because the EPA
approach uses a two-step procedure with the inflation factor (/?) to fit the probability model and
assumes the random effects are uncorrelated. The EPA equations are as follows:

                                                     ,-)             \
                                      n'0 + Xikn'xk + -Ti'wP + nvi
                                                       3
                               i — aO + %ikaXk + aVi +
                                                  /        T/T2    f) 1\
                    [TT^J  aVi]~BivariateNormal\[0  0],  x     J)
                                                  \        L 0   0|J/
Finally, the simulated transformed consumption amounts are untransformed using the following
equation (see Tooze et al., 2010):
                   Bvi =
This equation includes an adjustment involving the within person variance in the fish consumption
amount (o"|). The NCI Method assumes the reported fish consumption amounts in the 24-hour
recalls are unbiased. This adjustment makes the untransformed usual fish consumption essentially
unbiased.
                                           32

-------
Although when AIJ - 0, the transformed fish consumption is defined (T^y = — ), it is possible to
simulate a value of TVi such that TVi < —  for which the untransformed value is not defined. In the
                                    A.
NCI macro, these small simulated values in the transformed scale are set to half of the minimum
reported fish consumption for any 24-hour recall. The same procedure is used in the EPA
calculations. The probability that TVi < — depends on several factors. The expected probability is
                                     A,
less than or equal to 1/N, where N is the number of respondents with reported fish consumption.
In preliminary analysis, no values were set to half the minimum reported value.

The usual  fish consumption for a simulated person is then:
The following summary statistics for the usual fish consumption were calculated using the simulated
values (Uw) and the NHANES analysis weights: mean and 25th, 50th, 75th, 90th, 95th, 97th, and
99th percentiles.
4.4       Calculation of Standard Errors and Confidence Intervals

Both the NCI Method and the EPA Method use the NHANES survey weights for all the
calculations (i.e., weighted regressions and weighted estimates of the variance components). The
NHANES survey weights are inversely proportional to the probability of selection for each
respondent. The survey weights are adjusted for nonresponse and allow for calculation of national
estimates.

The SURVEYLOGISTIC and SURVEYREG procedures calculate standard errors for only the
linear parameters in the models, using a Taylor series linearization approach. These standard errors
were used to select the independent predictors for the probability and amount models. Standard
errors for all EPA and NCI Method parameters, including the random effects and percentiles of the
distribution of usual fish consumption, can be calculated by (1) preparing replicate weights
consistent with the NHANES survey design and strata and PSU variables; (2) running the macros
using the full-sample weight and each replicate weight; and (3) combining the results using each
weight to estimate the standard errors. EPA constructed replicate weights for calculating the
standard errors and confidence intervals for percentiles of usual fish consumption. See Wolter, 1985,
for a discussion of variance estimation procedures for complex survey designs such as NHANES.
                                            33

-------
In general, the variance of the weighted estimates from the NCI Method or the EPA Method can be
calculated by repeatedly dividing the sampling PSUs into subgroups (or replicates) and comparing
the estimates from each subgroup to the estimate for the entire sample. Several approaches have
been developed to efficiently estimate the variance with a minimum number of carefully constructed
subgroups. In general, dividing the respondents into subgroups can be achieved by creating an
analysis weight for each subgroup, i.e., a replicate weight. One such approach is Balanced Repeated
Replication (BRR) which divides the PSUs into two equal size groups on each division. A
modification of the basic BRR method due to Fay compares a weighted estimate from one half of
the PSUs to a different weighted estimate for the other half of PSUs. The Fay approach was selected
because it has advantages when  estimating percentiles. The replicate weights for the BRR method
using the Fay factor adjustment (Fay factor K = .3) were created using standard procedures (see
Judkins, 1990) and the strata and PSU variables in the NHANES files provided for variance
estimation. The basic BRR procedure assumes two PSU values in each stratum. However, a few of
the NHANES strata have three  PSU values, which required slightly modified calculations for
creating the weights (Wolter, 1985; Rust, 1986). We created 64 replicate weights. Parameter estimates
were calculated using the NHANES (full sample) weight and each of the replicate weights.

Given the replicate BRR weights, the variance of an estimate of 6 can be calculated using the steps
below; 6 might be a regression parameter, an estimated percentile of usual fish consumption, or the
log-transformed percentile of usual fish consumption.

      1.     Calculate 6 using the  full sample NHANES weight and using each of the 64 replicate
           weights (6g, g = 1 to G, G = 64), and

      2.     Calculate the variance of 6 as
      3.
A 95 percent confidence interval for 9 is 9 - l.96^/Var(9) to 9 + l.96^/Var(9).
Various summary statistics (means and percentiles) are calculated using the simulated usual fish
consumption values. Since the usual fish consumption estimates are generally skewed with a roughly
lognormal distribution, calculating the confidence intervals on the log scale appears reasonable and
has the beneficial effect that confidence limits cannot be negative. As a result, the confidence
intervals for the summary statistics are calculated by (1) fitting the EPA Method using the full
sample weight and each replicate weight; (2) log-transforming the estimates; (3) calculating the
                                             34

-------
confidence intervals for the estimates assuming a normal distribution using the equations above; and
(4) un-transforming the confidence interval bounds.

See Gilbert (1987) for additional comments on calculating confidence intervals for log-normally
distributed values.
4.5       Application of EPA Method

The EPA Method was used to model and predict usual consumption for the following types of fish
and shellfish:

      •    Total finfish and shellfish;
      •    Finfish;
      •    Shellfish;
      •    Marine fish;
      •    Estuarine fish;
      •    Freshwater fish;
      •    Freshwater + estuarine fish;
      •    Marine + estuarine fish;
      •    Marine + freshwater fish;
      •    Trophic level 2 fish;
      •    Trophic level 3 fish;
      •    Trophic level 4 fish;
      •    Freshwater + estuarine trophic level 2 fish;
      •    Freshwater + estuarine trophic level 3 fish;
      •    Freshwater + estuarine trophic level 4 fish;
      •    Marine trophic level 2  fish;
      •    Marine trophic level 3  fish; and
      •    Marine trophic level 4  fish.
                                            35

-------
All models included the weekend indicator and the indicator of the first or second recall for both the
amount and probability sub-models.


Other candidate variables for inclusion into the models included the following:

      •    Age group: 1 to  < 3, 3 to <6, 6 to < 11, 11 to < 16, 16 to <18, 18 to < 21, 21 to <35,
           35 to <50, 50 to <65, and 65 years and older;

      •    Income: $0 to <$20K, $20 to <$45K, $45 to <$75K, $75K+, >$20k, Refused/DK
           Income, Income Missing;

      •    Male: an indicator, 1 =  male 0 = female;

      •    Race/Ethnicity:  Mexican American, other Hispanic, non-Hispanic white, non-Hispanic
           black, and other race;

      •    Region: U.S.  Census regions (Northeast, Midwest, South, and West);

      •    Coastal Status: Coastal  counties and non-coastal counties;

      •    Bodyweight,  log-transformed; and

      •    Reported  frequency offish  consumption in 30 days (Fts/l30), transformed.

Based on preliminary analysis, the following transformations were used: for the  probability sub-
model: Ln(Fish3Q + 0.1) and for the amount sub-model: FtS/l300i45.


The following process was used to select the final predictors for the models:

      1.    For each dependent variable, start with all main effects, sequentially drop the least
           significant main  effect until all remaining main effects are significant at the 5 percent
           level.

      2.    Use any main effects that were selected when predicting any dependent variable.

      3.    For each dependent variable, include the selected main effects and all two-way
           interactions of the selected main effects, sequentially dropping the least significant
           interaction until  all remaining interactions are significant at the 1 percent level.

      4.    Use any two-way interaction selected for predicting at least three of the dependent
           variables.

The selected main effects and two-way interactions were then used as independent predictors in the
final models predicting all the dependent variables.
                                             36

-------
The procedure above was performed separately for the amount model and the probability model.
The significance was based on the SURVEYLOGISTIC and SURVEYREG output.


The final lists of independent variables for the probability and amount sub-models follow:

      •    Probability Sub-Model: reported frequency offish consumption (transformed),
           bodyweight (log-transformed), race/ethnicity, income, age group, region, coastal status,
           race/ethnicity*income, race/ethnicity*age group, income*age group, income*region,
           income* coastal status, and age group* region

      •    Amount Sub-model: reported frequency offish consumption (transformed),
           bodyweight (log-transformed), race/ethnicity, male, age group, region, coastal status,
           race/ethnicity*age group, race/ethnicity*region, male*age group, age group*region, and
           age group* coastal status

The simulated usual fish consumption values were summarized by the following demographic
categories:

      •    Gender;

      •    Age group;

      •    Women of childbearing age (13 to 49 years);

      •    Race;

      •    Income;

      •    Region;

      •    Coastal status;

      •    Four coastal regions (Atlantic, Gulf of Mexico, Pacific, and Great Lakes);

      •    Four inland regions (Inland Northeast, Inland Midwest, Inland South, Inland West);

      •    Youth (<21 years of age) by gender, race, income, region, coastal status, coastal regions,
           and inland regions; and

      •    Adults (>21 years of age) by gender, race, income, region, coastal status, coastal regions,
           and inland regions.

When fitting the NCI and EPA Methods, the distribution can be sensitive to the magnitude of the
variance components. With small sample sizes, the number of respondents with reported fish
consumption on two 24-hour recalls can be small resulting in imprecise variance estimates and
                                            37

-------
possible convergence problems. The authors of the NCI Method (Kipnis et al., 2009) generally
recommend having at least 50 respondents with at least two 24-hour recalls with reported
consumption offish (or whatever dietary component is being assessed).

The NCI and EPA Methods can be applied to all NHANES respondents or to subsets, such as
subsets defined by demographic characteristics.  Fitting one model using all NHANES respondents
implicitly assumes that the magnitude of the variance components are the same for all individuals
and do not vary by, say, demographic characteristics. Fitting all respondents has the advantage that
there are plenty of individuals from which to estimate the variance components. Fitting the models
separately by subset allows the variance components to be different for different subsets of
individuals. However, small subsets may not have adequate numbers of individuals with two recalls
with fish consumption. After considering the trade-offs, the EPA Method was applied to all
NHANES respondents. For each fish type, the subgroup estimates were derived from simulated
usual fish consumption values and the associated demographic covariates and sampling weights for
the NHANES  respondents.
4.6       Comparison of Results from the EPA and NCI Methods

In order to evaluate how estimates from the EPA Method compared to estimates from the NCI
Method, we ran both methods using different dependent variables, different sets of independent
predictors, and different numbers of simulated values. The procedures used to evaluate the EPA
Method and to compare the two methods are described in this section.


4.6.1     Analysis of Simulated Data

Fish consumption data (both usual intake and reported intake) were  simulated consistent with the
model assumed by the NCI Method. Ideally, when analyzing the simulated data, parameter estimates
from the NCI and EPA Methods will agree with the parameters used to simulate the data and the
estimated percentiles of usual fish consumption will agree with the corresponding percentiles in the
simulated data. Differences can indicate programming errors or possible bias associated with to the
estimation method. Different scenarios were used to evaluate the EPA Method, with good
agreement between the parameter estimates and percentiles compared  to the values used to simulate
the data.
                                            38

-------
As an illustration, the following plots (Figure 1) show how the parameters and percentiles from the
NCI and EPA Methods compare to the simulated values for data with six different Box-Cox
transformations (Lambda = -0.12, -0.06, 0.06, 0.12, 0.18, and 0.30). For these simulations, the other
parameters were set to values similar to those found when analyzing freshwater fish for adults (for
which Lambda = -0.06), with the exception that the intercept for the probability was increased
somewhat to raise the number of simulated individuals with two 24-hour recalls with fish
consumption. The plots show the percentile estimates derived from the NCI and EPA models
compared to the values from the simulated data. The  plots and analysis suggest that the EPA
Method provides a good approximation to the NCI Method and the true values when lambda is
greater than zero; and, for negative lambda, the EPA Method appears to provide better estimates
than the NCI Method when compared to the true values. It should be noted that (1) for positive
lambda, whether the NCI or EPA estimates are closer to the true value is different for different
simulated data sets using the same simulation parameters; (2) for negative lambda, the NCI and EPA
Methods provide more similar results when the magnitudes of the variance components are smaller;
and (3) the NCI Method could be modified to  allow for negative lambda values.
4.6.2     Confidence Intervals for Percentiles of Fish Consumption

Figure 2 shows confidence intervals for parameters and percentiles of fish consumption, calculated
using both the NCI and EPA Methods. Due to the long computation time required for the NCI
Method, only simple models with few predictors were used. The confidence intervals were
calculated for freshwater and estuarine fish consumption by all respondents, marine fish
consumption by all respondents, and shellfish consumption by adults. The predictors are
transformed frequency offish consumption in the past 30 days, an indicator of weekend versus
weekday, and the difference between the first and second recall. The first column in Figure 2
compares parameter estimates and confidence intervals. The second column compares percentile
estimates and confidence intervals for various percentages and demographic groups. In general,
there is good agreement between both the NCI and EPA estimates and the width of the confidence
intervals.
                                            39

-------
Figure 1.        Comparison of NCI, EPA, and true percentiles using simulated data and different
                   transformations
                     Comparison of EPA and NCI Percentiles
                              Lambda = 0.30
Comparison of EPA and NCI Percentiles
         Lambda = 0.18
                                                                                             True Percent!!
                                                                                    |      True o EPA Method o MCI M'-th.
                     Comparison of EPA and NCI Percentiles
                              Lambda = 0.12
Comparison of EPA and NCI Percentiles
         Lambda = 0.06
                              True Percentile
                          -True o EPA Method
                                                                                             True P ere entile

                                                                                    |      Till* o EPA Method o MCI Melh'-"l]
                     Comparison of EPA and NCI Percentiles
                              Lambda = -0.06
Comparison of EPA and NCI Percentiles
         Lambda = -0.12
                                                              40

-------
Figure 2.         Comparison  of confidence intervals using the NCI  and EPA Methods
                  NCI versus EPA Parameter Estimates with 95% CIs
                                 All FWRt
                                    t  '•' Estimate .3111:1 '.I ]
NCI versus EPA Percentile Estimates with 95% CIs
               All rwEst
                                                                               | Percentile
                                                                                              a.ijIiMmenl o 25  + 50  x 75 fl 80 q 95 t 97 •:.• 99 |
                  NCI versus EPA Parameter Estimates with 95% CIs
                                  All Marine
NCI versus EPA Percentile Estimates with 95% CIs
               All Marine
                  NCI versus EPA Parameter Estimates with 95% CIs
                                All Shellfish
NCI versus EPA Percentile Estimates with 95% CIs
               All Shellfish
                                                                                                     NCI Estimata
                                    i  . Estimate .arn:l i..l
                                                                                Pii;?nlile
                                                                                              Agreement o 25  + 50  x 75 a 90 g 95 * 97 ... as]
                                                                   41

-------
4.6.3     Analysis of NHANES Fish Data Using Various Models


NHANES fish consumption data were analyzed using the NCI and EPA Methods with different
sets of independent predictors and different numbers of simulations. These comparisons between
the EPA and NCI Methods were selected to represent a range offish types. Each dependent
variable was predicted by different independent variables to assess the effect of the choice of
predictors and the number of parameters on the results, all using 50 simulations. Comparisons were
also run using different numbers of simulations to assess how many simulations to use. Table 6
summarizes the results followed by example plots comparing the percentile estimates from the two
methods.


For each of the comparisons, Table 6 shows the following values:

      •    Pred Vars: the independent predictors, F = transformed frequency of fish consumption,
           A = Age groups, I = Income groups, R = Race groups, and M = Male indicator.

      •    Sim Num: the number of simulated values of usual fish consumption generated for each
           individual.

      •    Geo Mean Ratio (EPA/NCI): The geometric mean ratio  of the EPA percentile to the
           NCI percentile across  multiple percentiles and population subgroups. A ratio of 1.00
           corresponds to no difference between the geometric means, on average.

      •    RMSE (percent): The RMSE difference between the log-transformed EPA and NCI
           percentiles estimated across multiple percentiles and population subgroups, converted
           to a percentage difference. This can be thought of as the average absolute percent
           difference between the NCI and EPA percentiles. Smaller values are better. Larger
           values are generally associated with fish types that are consumed less often.

      •    90th percentile (Adults>21): the 90th percentile offish consumption (a value of
           particular interest to EPA)  calculated using the NCI and EPA Methods.

      •    Rel. Time (NCI/EPA): the computation time for the NCI Method relative to the EPA
           Method. These values are not precise and depend on what other programs were running
           at the same  time.

      •    Num Parms (NCI): the number of parameters in each model.

      •    NCI Lambda: the power for the Box-Cox transformation estimated using the NCI
           Method.

      •    EPA lambda: the power for the Box-Cox transformation used in the EPA Method.
                                           42

-------
      •    NCI Rho: The correlation between the person-specific random effects in the probability
           and amount sub-models, as estimated by the NCI Method.

Note that the geometric mean ratio, RMSE, and percentiles in the table are subject to random
variation associated with the simulation process. As a result, somewhat different values would be
obtained if the calculations were repeated.
Table 6.
Comparison of NCI and EPA methods using NHANES fish data
Fish type
All Fish
All Fish
All Fish
All Fish
All Fish
All Fish
All Fish
Finfish
Finfish
Finfish
Fresh
Fresh
Fresh
Fresh
FWEst
FWEst
FWEst
FWEst
FWEst
FWEst
FWEst
Marine
Marine
Marine
Marine
Marine
Marine
Marine
Shellfish
Shellfish
Shellfish
Shellfish
Shellfish
Shellfish
Shellfish
Pred.
vars
F
F
F
F
F
FA
FAIRM
F
FA
FAIRM
F
F
FA
FAIRM
F
F
F
F
F
FA
FAIRM
F
F
F
F
F
FA
FAIRM
F
F
F
F
F
FA
FAIRM
Sim
num
5
10
20
50
100
50
50
50
50
50
50
50
50
50
5
10
20
50
100
50
50
5
10
20
50
100
50
50
5
10
20
50
100
50
50
Geo mean
ratio
(EPA/ NCI)
1.008
1.008
1.006
1.004
1.003
1.002
1.001
1.006
1.016
1.010
1.030
0.872
1.049
1.022
0.987
0.992
0.992
0.995
0.991
0.991
0.995
1.010
1.003
1.013
1.006
1.005
1.008
1.004
0.980
0.988
0.990
0.988
0.987
0.988
0.991
RMSE
(%)
5.258
4.690
4.367
4.084
4.230
4.005
4.361
2.803
14.544
12.936
36.877
21.225
35.486
36.953
4.746
4.092
2.781
2.500
2.361
2.998
2.465
8.605
8.017
7.336
7.596
7.641
7.410
7.747
4.629
4.262
3.560
3.059
3.194
3.508
3.542
90th percentile
(AdultsSSl)
EPA
48.77
49.64
49.13
49.08
49.14
51.62
51.59
36.98
38.37
38.51
6.70
5.92
7.43
7.14
20.04
20.35
20.19
20.27
20.23
21.81
21.91
31.16
30.96
31.09
30.96
31.11
32.39
32.17
14.00
14.17
14.07
14.06
14.06
15.34
15.40
NCI
49.89
49.36
49.92
49.73
49.82
52.51
52.47
36.13
41.00
40.85
6.47
6.49
7.12
6.83
20.03
20.41
20.40
20.38
20.38
22.12
22.07
32.10
31.76
32.01
31.84
31.93
33.24
33.10
14.20
13.98
14.05
14.03
14.03
15.23
15.28
Rel. time
(NCI/
EPA)
13.5
12.4
10.1
5.7
2.5
44.7
170.2
8.7
26.0
118.2
8.3
9.57
77.3
303.7
17.2
13.7
6.6
8.3
2.4
64.6
290.1
11.9
8.5
3.6
5.5
1.9
22.1
126.9
21.5
19.6
10.5
10.1
3.9
48.6
228.1
Num
parms
13
13
13
13
13
31
53
13
31
53
13
13
31
53
13
13
13
13
13
31
53
13
13
13
13
13
31
53
13
13
13
13
13
31
53
NCI
lambda
0.210
0.210
0.210
0.210
0.210
0.208
0.211
0.255
0.252
0.256
0.010
0.010
0.010
0.010
0.105
0.105
0.105
0.105
0.105
0.104
0.106
0.218
0.218
0.218
0.218
0.218
0.218
0.220
0.112
0.112
0.112
0.112
0.112
0.112
0.111
EPA
lambda
0.21
0.21
0.21
0.21
0.21
0.21
0.21
0.26
0.26
0.26
-0.04
0.01
-0.04
-0.04
0.11
0.11
0.11
0.11
0.11
0.11
0.11
0.23
0.23
0.23
0.23
0.23
0.23
0.23
0.11
0.11
0.11
0.11
0.11
0.11
0.11
NCI
Rho
0.17
0.17
0.17
0.17
0.17
0.18
0.25
0.03
-0.01
-0.07
-0.21
-0.21
-0.19
-0.26
0.15
0.15
0.15
0.15
0.15
0.16
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.18
0.50
0.50
0.50
0.50
0.50
0.50
0.50
                                            43

-------
For freshwater and freshwater and estuarine trophic level 2, the lambda estimated using the EPA
Method was less than zero (only freshwater is shown in Table 6). The NCI Method constrains
lambda to be greater than 0.01. For freshwater fish consumption, the model with frequency offish
consumption as the only predictor was run using the transformation selected for the EPA Method,
A* = —0.04, and with A* = 0.01, the transformation used in the NCI Method (see the yellow
shaded cell). When the preferred transformation corresponds to a negative lambda, the EPA and
NCI Methods give different results (see the cells with the dark border with high RMSE), although
the 90th percentiles are relatively close. Based on the simulations in Figure 1, the EPA percentiles
appear to be preferred to the NCI percentiles when lambda is less than 0.01.


Analysis of the data in Table  6 suggests the following:

      •    Relative to the NCI estimates, the percentile estimates from the EPA Method are
           essentially unbiased (the average of the geometric mean ratio is 0.9923, using only rows
           with 50 simulations and excluding the cases with negative lambda).
      •    The results from the EPA Method are acceptably close to those from the NCI Method:
           the average RMSE is 6.5 percent (using only rows with 50 simulations and excluding the
           cases with negative lambda).
      •    There is little additional reduction in RMSE when using more than 50 simulations. The
           final runs used 100 simulations.
      •    Uncertainty in the estimated percentiles is associated with (1) the NHANES sampling
           error (represented by the confidence intervals); (2) selection of independent predictors
           (which depends in part on what variables are available); (3) which analysis method is
           used (EPA versus NCI); and (4) the number of simulations. These sources of
           uncertainty are ordered roughly  from most to least important.
      •    The transformation (X) estimated for the EPA model is very close to the transformation
           estimated in the NCI model. Differences in lambda do not explain differences in the
           90th percentiles from the two procedures.
      •    The estimated correlation between the variance components (Rho) is not significantly
           correlated with the difference between the 90th percentiles from the NCI and EPA
           Methods.
      •    Computation time for the NCI Method relative to the EPA Method increases
           significantly with increasing numbers of parameters.

The shaded rows in Table 6 correspond to the example plots shown in Figure 3. These plots were
selected to show the comparison with the best RMSE, the median RMSE, and the worst RMSE
among models with positive lambda and 50 simulations, and the comparison for fresh water fish, for
which the lambda estimated using the EPA Method is negative. Figure 3 shows EPA versus NCI
means and percentiles for different subpopulations.
                                             44

-------
Figure 3.
Comparison of NCI and EPA percentiles offish  consumption  rates
                                      AII_FWEst_F
                Comparison of EPA and Mixtran percentiles for various subpopulations
                 GMRatio(EPA/Mixtran] - 0.995, CV(Ratio] - 2.442%, RMSE - 2.5001'.
                                     Mmtian F'eicenitle
                  Equality   • Mean     o 25th Pctile   EOth Pctile  o 75th Pctile    90th Pctile I
                  95th Pctile    99th Pctile
                                                                                            AII_Shellfish_F
                                                                      Comparison of EPA and Mixtran percentiles for various subpopulations
                                                                        GMRatiofEPA/Mixtran) = 0.988, CV(Ratio) = 2.793%, RMSE = 3.059%
                                                                                                                                            10   15  20   30 40
                                                                                                                    Ivlmtian F'eicemtle
                                                                   	Equality   • Mean     o 251h Pctile   50th Pclile o 75th Pctile    90th Pctile
                                                                    o   95th Pctile    99th Pctile
                                     AII_FinFish_FA
                Comparison of EPA and Mixtran percentiles for various subpopulations
                 GMRatio(EPA/Mixtran) = 1.016, CV(Ratio) = 14.458%, RMSE = 14.544%
                                                                                            AII_Fresh_FA
                                                                      Comparison of EPA and Mixtran percentiles for various subpopulations
                                                                       GMRatiofEPA/Mixtran) = 1.049, CV(Ratio) = 35.117%, RMSE = 35.486%
                                     Mixtran Percenitle
                                                                                                                     Mmtian F'eicenitle
                - Equality   • Mean     o 25th Pctile   SOth Pctile  o 75th Pctile    90th Pctile
                  95th Pctile   99th Pctile
O
- Equality
95th Pctile
Mean
99th Pctile
o ">5th Pctile
50th Pctile
o 75th Pctile
90th Pctile
                                                                               45

-------
                                                                   Results
This section presents the sample sizes and the estimated UFCR (raw weight, edible portion) for all
fish, freshwater + estuarine fish, marine fish, all finfish, all shellfish, trophic level 2 fish, trophic level
3 fish, trophic level 4 fish, trophic level 2  freshwater + estuarine fish, trophic level 3 freshwater +
estuarine fish, and trophic level 4 freshwater + estuarine fish, for adults and youth, by demographic
characteristics and geographic area. Full tables including UFCR for the total population (youth and
adults combined), adults only, and youth only are in Appendix E. Appendix F contains the UFCR
for as prepared weights. The fish types selected to be presented in the body of the report represent
those that are of most interest to EPA, states, and tribes.

Note that the adult population is defined  as people  aged 21 years and over. The US EPA Exposure
Factors Handbook classifies those aged 21 years and over as adults. Children are grouped as follows:
1 to  <3 years, 3 to <6 years, 6 to <11 years, 11 to <16 years, 16 to <18 years, and 18 to <21 years.
Note that children 1 to <2 and 2 to <3 were combined due to  small sample sizes for these age
groups.
5.1       Sample Size

Table 7 presents the sample sizes for each subpopulation that reported fish consumption on at least
one 24-hour recall. An expanded table that includes the other fish types for which rates were
calculated can be found in Appendix G.
                                            46

-------
Table 7.
Sample size and number reporting fish consumption, by fish type

Total
Gender
Female
Male
Age, years
Ito <3
3 to <6
6 to <11
11 to <16
16 to <18
18to<21
21to<35
35 to <50
50 to <65
65 and older
WCA" (13 to 49 years)
Income
<$20k
$20kto<$45k
$45k to <$75k
$75k and over
>$20k
Ref/DK income"
Income missing
Race/Ethnicity
Mexican American
Other Hispanic
Non-Hispanic white
Non-Hispanic black
Other race
U.S. Region
Midwest
Northeast
South
West
Coastal Status
Noncoastal
Coastal
US Coastal/Inland
Region
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
N
29,463

15,694
13,769

2,325
2,185
2,705
2,806
1,417
1,662
4,381
4,522
3,730
3,730
7,870

6,679
8,955
5,561
6,288
825
808
347

6,868
2,405
11,980
6,734
1,476

6,445
4,475
11,036
7,507

17,251
12,212


3,802
4,646
1,370
2,394
2,584
4,137
6,825
3,705
Any
fish
6,891

3,807
3,084

345
350
454
445
252
311
1,070
1,332
1,216
1,116
1,919

1,374
1,969
1,334
1,768
203
164
79

1,350
532
2,678
1,818
513

1,235
1,202
2,688
1,766

3,719
3,172


976
1,320
361
515
628
741
1,560
790
FW+
Est
4,868

2,667
2,201

198
196
264
310
177
208
801
997
901
816
1,421

897
1,382
959
1,308
151
118
53

949
351
1,854
1,308
406

821
805
1,925
1,317

2,532
2,336


739
938
292
367
409
463
1,082
578
Marine
6,286

3,495
2,791

305
322
416
402
237
294
992
1,221
1,101
996
1,785

1,256
1,775
1,211
1,634
182
153
75

1,212
490
2,509
1,603
472

1,070
1,154
2,416
1,646

3,377
2,909


900
1,247
316
446
600
645
1,386
746
Fin
fish
5,095

2,792
2,303

269
272
351
301
164
209
723
962
938
906
1,409

1,043
1,442
1,002
1,285
144
122
57

937
390
2,000
1,376
392

938
867
2,003
1,287

2,813
2,282


720
939
255
368
454
588
1,204
567
Shell
fish
2,612

1,448
1,164

101
106
137
179
104
132
509
546
454
344
768

470
732
498
739
81
61
31

535
187
1,023
654
213

400
484
1,036
692

1,287
1,325


385
553
196
191
248
213
519
307
Trophic
level 2
2,706

1,495
1,211

111
118
143
180
98
131
531
566
468
360
839

491
792
511
740
86
57
29

618
202
1,006
669
211

431
445
1,087
743

1,345
1,361


425
524
203
209
234
226
567
318
Trophic
level 3
4464

2,435
2,029

209
225
286
296
173
209
745
883
775
663
1,300

911
1,286
856
1,108
140
111
52

886
329
1,573
1,291
385

773
733
1,828
1,130

2,363
2,101


621
865
269
346
364
437
1,053
509
Trophic
level 4
4,578

2,521
2,057

243
246
315
273
155
199
651
848
842
806
1,179

936
1,263
904
1,176
126
117
56

828
350
1,855
1,188
357

853
812
1,739
1,174

2,566
2,012


646
840
202
324
420
547
1,071
528
a Women of childbearing age.b Income refused or don't know.
                                             47

-------
5.2       Usual Fish Consumption Rates

Tables 8 through 18 present the UFCR estimates of raw weight, edible portion for adults 21 years
and older for total fish, freshwater + estuarine fish, marine fish, all finfish, all shellfish, trophic level
2 fish, trophic level 3 fish, trophic level 4 fish, trophic level 2 freshwater + estuarine fish, trophic
level 3 freshwater + estuarine fish, and trophic level 4 freshwater + estuarine fish. The tables
provide the 50th, 75th, 90th, 95th, 97th, and 99th percentiles, along with their respective 95 percent
confidence intervals. Tables 19 through 29 present the UFCR estimates of raw weight, edible
portion for youth less than 21 years old.

The tables show percentiles for total fish consumed and for various fish types that make up the
total. The mean consumption for all fish  should be equal, not counting random errors, to the sum of
the mean consumption across different types of fish,  e.g., marine, estuarine, and freshwater or
trophic levels 2, 3, and 4. The same cannot be said about percentiles. At the extreme, the sum of the
maximum fish consumption across fish types will not equal the maximum fish consumption for all
fish except in the very unusual  case where one individual is the  largest consumer in all fish type
categories. For a selected percentile, the difference between the sum of the percentiles across fish
types  and the percentile for the sum across all fish types will increase as the percentile of interest
increases from the 50th percentile to 90th percentile,  99th percentile, and the maximum. The 90th
percentile for all fish will be greater than  or equal to the 90th percentile for any one type of fish and
will usually be less than the sum of the 90th percentiles across all types.

There are two tables for each fish type, a and b. Tables  8a—29a present the UFCR by demographic
characteristics (gender, age, income, and race/ethnicity) and Tables 8b—29b present the UFCR by
geographic area.
                                             48

-------
Table 8a.
UFCR estimates (g/day raw weight, edible portion): Total fish, adults, 21 years and older, by demographic characteristics
All Finfish and Shellfish
Adults (>21 yrs)
Age
21 to <35 yrs
35 to <50 yrs
50 to <65 yrs
65+ yrs
Women of childbearing
age (13 to 49 yrs)
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know
Income
Income Missing
Percentiles(95%CI)
50th
17.6 (15.8,19.7)

13.1 (11.1,15.4)
18.3 (16.0,20.9)
22.4 (19.1,26.2)
16.9 (14.4,20.0)
11.6 (10.2,13.2)

15.3 (13.7,17.1)
20.6 (18.2,23.3)

16.7 (13.8,20.1)
16.6 (13.3,20.7)
16.7 (14.7,18.9)
19.6 (16.9,22.7)
32.3 (25.8,40.4)

13.6 (11.7,15.8)
15.4 (13.4,17.7)
16.5 (14.2,19.2)
23.1 (20.5,26.1)
17.1 (12.6,23.0)
16.9 (10.9,26.2)
8.8 (4.1,18.8)
75th
32.8 (30.1,35.7)

26.8 (23.6,30.4)
33.1 (29.4,37.2)
38.8 (34.2,44.1)
31.1 (27.1,35.6)
23.6 (21.5,25.9)

28.4 (26.0,31.0)
38.0 (34.5,41.8)

31.3 (27.0,36.3)
31.4 (25.8,38.4)
31.0 (28.2,34.2)
35.3 (31.5,39.6)
54.0 (44.5,65.5)

27.0 (24.1,30.2)
28.8 (26.0,31.9)
30.7 (27.2,34.7)
40.1 (36.1,44.6)
31.3 (24.6,39.8)
36.6 (25.2,53.3)
22.0 (12.0,40.4)
90th
52.8 (48.0,58.1)

46.6 (40.5,53.6)
52.7 (46.1,60.2)
59.3 (52.3,67.2)
49.5 (43.2,56.9)
39.4 (35.4,43.8)

45.2 (40.9,50.0)
60.6 (54.6,67.2)

50.8 (43.8,59.0)
50.7 (42.1,61.1)
49.8 (44.8,55.4)
55.7 (49.9,62.2)
81.1 (66.3,99.1)

45.2 (40.1,51.1)
46.7 (42.1,51.8)
49.6 (43.5,56.5)
61.3 (54.8,68.7)
48.4 (38.9,60.3)
64.5 (46.9,88.8)
46.6 (27.0,80.3)
95th
68.1 (61.2,75.8)

63.5 (54.1,74.5)
67.5 (58.4,78.0)
74.4 (65.3,84.7)
63.8 (55.1,73.9)
51.7 (45.9,58.2)

57.8 (51.7,64.6)
77.6 (69.2,87.2)

66.1 (56.4,77.4)
65.0 (54.2,77.9)
64.3 (57.1,72.4)
71.1 (63.4,79.8)
102.7 (82.7,127.4)

59.9 (52.4,68.4)
61.1 (54.5,68.4)
64.0 (55.6,73.8)
77.1 (68.2,87.2)
61.7 (49.5,77.0)
83.5 (62.8,111.2)
65.0 (40.2,105.0)
97th
79.7 (71.0,89.5)

77.2 (64.6,92.4)
78.5 (67.4,91.5)
85.5 (74.7,97.9)
74.0 (63.6,86.1)
61.0 (53.7,69.2)

67.1 (59.7,75.5)
90.5 (79.8,102.5)

77.5 (65.7,91.5)
75.7 (63.3,90.7)
75.1 (66.1,85.4)
82.7 (73.2,93.3)
117.6 (93.4,148.1)

71.7 (61.9,82.9)
71.9 (63.9,81.0)
75.0 (64.6,87.1)
88.6 (77.6,101.1)
72.0 (57.7,89.7)
96.4 (73.3,126.7)
76.6 (49.6,118.2)

99th
105.1 (92.0,120.2)

109.7 (87.5,137.4)
102.4 (86.8,120.7)
109.1 (94.1,126.5)
96.2 (81.6,113.5)
81.5 (70.6,94.1)

87.2 (76.4,99.5)
118.1 (102.3,136.3)

103.8 (86.2,125.1)
99.5 (82.7,119.7)
98.1 (85.0,113.2)
107.2 (93.5,122.8)
153.0 (117.1,200.0)

99.6 (82.8,119.9)
96.7 (84.6,110.5)
99.8 (84.4,118.1)
113.5 (97.7,131.7)
93.2 (74.7,116.4)
124.8 (95.5,163.3)
99.5 (68.3,145.0)
  Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
  identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
  black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
          Table 8b.
UFCR estimates (g/day raw weight, edible portion): Total fish, adults, 21 years and older, by geographic area
c_n
O
Percentiles(95%CI)
All Finfish and Shellfish
Adults (>21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
17.6 (15.8,19.7)

23.9 (20.0,28.7)
12.9 (10.6,15.6)
17.6 (15.1,20.4)
20.0 (17.1,23.4)

15.9 (13.7,18.5)
20.9 (18.4,23.7)

22.1 (18.2,26.7)
24.5 (20.7,28.9)
19.0 (15.2,23.8)
14.6 (12.1,17.5)
22.1 (17.5,28.0)
12.4 (10.1,15.1)
15.6 (13.1,18.4)
18.4 (15.1,22.5)
75th
32.8 (30.1,35.7)

42.5 (36.3,49.8)
24.1 (20.6,28.3)
32.4 (28.8,36.4)
35.6 (30.7,41.2)

30.0 (26.4,34.1)
37.9 (34.1,42.1)

39.3 (33.2,46.4)
43.4 (37.6,50.2)
34.5 (29.4,40.5)
26.5 (22.6,31.1)
39.6 (32.6,48.1)
23.3 (19.7,27.6)
29.0 (25.6,32.9)
32.6 (27.0,39.4)
90th
52.8 (48.0,58.1)

65.2 (55.9,76.1)
39.2 (33.4,46.0)
52.1 (46.3,58.7)
55.7 (47.7,65.0)

48.3 (42.3,55.3)
59.9 (53.7,66.9)

61.2 (51.3,72.9)
67.2 (58.8,76.9)
55.0 (47.1,64.3)
41.8 (35.7,49.0)
60.7 (50.3,73.2)
38.3 (32.3,45.4)
46.9 (41.4,53.1)
50.6 (41.5,61.5)
95th
68.1 (61.2,75.8)

82.0 (70.0,96.1)
50.9 (43.1,60.2)
67.4 (59.1,76.8)
71.1 (60.3,83.9)

62.4 (54.1,72.1)
76.7 (68.3,86.2)

78.2 (64.5,94.7)
84.8 (74.0,97.2)
70.6 (59.5,83.7)
53.5 (45.6,62.9)
76.1 (62.9,92.0)
49.9 (41.7,59.7)
60.7 (53.0,69.5)
64.2 (52.7,78.3)
97th
79.7 (71.0,89.5)

93.7 (79.9,110.0)
60.0 (50.3,71.6)
79.0 (68.6,90.9)
82.6 (69.5,98.2)

73.0 (62.8,85.0)
89.3 (78.8,101.2)

91.0 (74.3,111.4)
97.7 (85.0,112.4)
82.4 (68.6,98.9)
62.2 (52.8,73.2)
87.1 (71.9,105.4)
59.1 (49.0,71.3)
71.2 (61.4,82.6)
74.5 (61.0,91.0)
99th
105.1 (92.0,120.2)

119.6 (100.9,141.6)
79.7 (66.5,95.6)
105.1 (89.0,124.1)
108.4 (89.8,130.9)

96.2 (81.4,113.7)
115.9 (100.6,133.6)

118.7 (94.8,148.5)
124.6 (106.7,145.4)
106.8 (87.9,129.8)
80.5 (68.3,94.9)
111.7 (91.4,136.6)
79.5 (64.9,97.4)
95.6 (79.2,115.5)
96.3 (78.6,118.1)
          1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

          2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
Table 9a.       UFCR estimates (g/day raw weight, edible portion): Freshwater + estuarine fish, adults, 21 years and older, by
                 demographic characteristics
Freshwater + Estuarine
Finfish and Shellfish
Adults (>21 yrs)
Age
21 to <35 yrs
35 to <50 yrs
50 to <65 yrs
65+ yrs
Women of childbearing age (13 to 49 yrs)
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
5.0 (4.1,6.0)

3.8 (3.1,4.8)
5.2 (4.1,6.6)
6.3 (5.0,7.9)
4.5 (3.3,6.1)
2.9 (2.3,3.6)

4.1 (3.4,5.0)
6.2 (5.0,7.6)

6.8 (5.3,8.6)
6.1 (4.4,8.6)
4.2 (3.4,5.2)
7.2 (5.8,8.9)
12.6 (9.4,16.9)

3.5 (2.8,4.4)
4.3 (3.5,5.4)
4.8 (3.8,6.2)
6.6 (5.4,8.1)
5.5 (3.6,8.3)
5.4 (3.2,9.1)
1.9 (0.8,4.5)
75th
11.4 (9.9,13.1)

9.9 (8.3,11.7)
11.9 (9.6,14.8)
13.2 (10.9,15.9)
9.9 (7.9,12.4)
7.6 (6.4,9.1)

9.3 (8.1,10.8)
13.8 (11.8,16.2)

15.3 (12.4,18.9)
14.1 (10.3,19.3)
9.4 (8.0,11.1)
15.4 (13.0,18.1)
25.1 (19.2,32.9)

9.1 (7.7,10.7)
9.9 (8.5,11.5)
10.8 (9.1,12.9)
13.9 (11.7,16.6)
12.1 (8.5,17.1)
13.8 (9.2,20.8)
7.1 (3.6,13.9)
90th
22.0 (19.1,25.4)

21.1 (17.6,25.1)
23.0 (18.4,28.7)
23.8 (19.7,28.9)
18.7 (15.4,22.7)
15.8 (13.2,19.0)

18.0 (15.4,21.0)
26.3 (22.4,30.9)

28.7 (23.1,35.8)
27.2 (19.5,37.9)
17.9 (15.1,21.1)
28.2 (23.8,33.4)
44.5 (33.3,59.6)

19.2 (16.3,22.6)
19.4 (16.6,22.7)
20.6 (17.6,24.3)
25.6 (21.3,30.8)
22.3 (16.5,30.2)
29.0 (19.7,42.6)
18.9 (10.6,33.7)
95th
31.8 (26.9,37.6)

32.2 (26.2,39.7)
33.0 (26.0,41.8)
33.3 (26.9,41.3)
26.5 (21.9,32.2)
23.5 (19.2,28.7)

25.7 (21.5,30.7)
38.0 (31.6,45.6)

40.9 (32.2,51.9)
38.7 (27.5,54.5)
25.5 (21.2,30.8)
39.6 (32.7,48.0)
62.3 (45.2,86.1)

28.9 (24.2,34.6)
28.4 (23.7,33.9)
29.6 (24.6,35.5)
36.2 (29.4,44.5)
30.9 (23.3,40.9)
43.1 (28.6,65.0)
31.7 (18.4,54.5)
97th
40.2 (33.3,48.5)

42.3 (33.3,53.7)
41.4 (32.2,53.1)
41.4 (32.7,52.4)
33.1 (26.9,40.6)
29.9 (24.1,37.0)

32.1 (26.4,39.1)
47.7 (39.0,58.4)

51.0 (39.6,65.6)
47.8 (33.7,67.6)
31.9 (26.0,39.0)
48.8 (39.4,60.3)
78.3 (55.0,111.5)

37.4 (30.9,45.4)
35.9 (29.6,43.6)
37.3 (30.4,45.9)
45.0 (35.9,56.5)
38.7 (29.2,51.1)
56.6 (36.3,88.3)
41.6 (24.5,70.7)
99th
61.1 (48.7,76.6)

68.1 (50.1,92.5)
62.5 (46.8,83.4)
60.4 (45.9,79.4)
48.8 (38.4,62.1)
46.6 (36.4,59.6)

48.2 (38.0,61.2)
71.9 (56.4,91.8)

75.7 (56.8,100.8)
69.7 (48.3,100.6)
47.9 (37.2,61.6)
70.8 (55.0,91.3)
114.7 (76.4,172.1)

59.3 (47.5,74.0)
55.4 (43.8,70.0)
56.8 (43.4,74.4)
66.2 (51.1,85.9)
56.1 (41.8,75.4)
88.6 (54.5,144.1)
65.9 (39.3,110.5)
 Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
 identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
 black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
Table 9b.       UFCR estimates (g/day raw weight, edible portion): Freshwater + estuarine fish, adults, 21 years and older, by geographic
                 area
Freshwater + Estuarine
Finfish and Shellfish
Adults (>21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
Percentiles(95%CI)
50th
5.0 (4.1,6.0)

5.8 (4.4,7.6)
3.2 (2.5,4.2)
6.4 (4.7,8.5)
5.1 (3.9,6.6)

4.2 (3.4,5.2)
6.6 (5.1,8.4)

6.3 (4.4,9.0)
8.3 (6.4,10.7)
7.3 (4.8,11.1)
4.0 (3.1,5.1)
5.0 (3.5,7.3)
3.0 (2.3,4.0)
5.3 (4.0,7.1)
4.3 (3.3,5.4)
75th
11.4 (9.9,13.1)

12.6 (9.9,16.0)
7.4 (6.0,9.0)
14.0 (11.3,17.4)
11.4 (8.8,14.8)

9.8 (8.2,11.6)
14.4 (11.8,17.5)

14.0 (10.1,19.5)
17.0 (13.9,20.8)
15.7 (11.7,21.1)
8.7 (7.1,10.7)
11.3 (8.0,16.0)
6.9 (5.5,8.7)
12.0 (9.7,14.9)
9.4 (7.4,12.1)
90th
22.0 (19.1,25.4)

23.1 (18.3,29.2)
14.3 (11.8,17.4)
26.3 (21.6,32.0)
22.4 (16.8,29.8)

19.0 (15.8,22.9)
27.1 (22.4,32.8)

27.3 (19.3,38.6)
30.8 (25.3,37.5)
28.6 (22.5,36.4)
16.5 (13.5,20.2)
21.0 (14.8,29.7)
13.5 (10.8,17.0)
22.8 (18.6,27.9)
18.2 (13.7,24.3)
95th
31.8 (26.9,37.6)

32.3 (25.4,41.0)
20.8 (16.9,25.7)
37.5 (30.5,46.1)
32.7 (23.9,44.9)

27.4 (22.3,33.8)
38.6 (31.4,47.6)

39.7 (27.4,57.7)
42.8 (34.5,53.0)
40.1 (31.8,50.6)
23.6 (19.1,29.1)
29.5 (20.6,42.2)
19.8 (15.5,25.2)
32.7 (26.2,40.7)
26.3 (19.1,36.1)
97th
40.2 (33.3,48.5)

39.9 (31.0,51.5)
26.3 (21.0,33.0)
46.7 (37.6,58.1)
42.0 (30.0,58.8)

34.6 (27.7,43.3)
48.4 (38.6,60.6)

51.2 (34.3,76.3)
52.3 (41.8,65.5)
50.3 (39.3,64.4)
29.4 (23.5,36.8)
36.5 (25.3,52.8)
25.1 (19.4,32.6)
40.9 (32.3,51.7)
33.3 (23.8,46.7)
99th
61.1 (48.7,76.6)

58.5 (44.2,77.5)
41.1 (31.3,54.0)
69.0 (54.3,87.7)
66.9 (45.4,98.5)

52.8 (40.7,68.4)
72.7 (55.6,95.0)

81.2 (51.6,127.8)
75.8 (58.8,97.7)
73.8 (55.6,97.8)
44.5 (34.1,57.9)
54.4 (36.7,80.6)
39.5 (29.1,53.5)
61.0 (46.7,79.7)
51.6 (35.5,74.9)
1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
 TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
 border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
Table lOa.     UFCR estimates (g/day raw weight, edible portion): Marine fish, adults, 21 years and older, by demographic characteristics

Marine Finfish and Shellfish
Adults (>21 yrs)
Age
21 to <35 yrs
35 to <50 yrs
50 to <65 yrs
65+ yrs
Women of childbearing
age (13 to 49 yrs)
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
9.9 (8.5,11.5)

7.3 (6.0,9.0)
10.2 (8.7,11.8)
13.0 (10.6,16.1)
9.5 (7.5,12.1)
6.8 (5.8,8.0)

8.9 (7.5,10.5)
11.2 (9.6,13.1)

7.9 (6.4,9.8)
8.2 (6.6,10.2)
9.9 (8.3,11.9)
9.3 (7.8,11.1)
17.3 (13.6,22.2)

7.4 (6.0,9.0)
8.4 (7.0,9.9)
9.5 (8.0,11.4)
13.4 (11.4,15.8)
9.4 (6.6,13.4)
9.6 (6.2,14.9)
5.2 (2.5,10.8)
75th
19.4 (17.4,21.7)

15.4 (13.2,18.0)
19.2 (17.0,21.8)
24.0 (20.4,28.1)
18.9 (15.8,22.5)
14.5 (13.0,16.1)

17.5 (15.5,19.7)
21.8 (19.5,24.3)

15.7 (13.4,18.5)
16.4 (13.6,19.8)
19.4 (17.0,22.2)
18.1 (15.8,20.8)
30.8 (25.5,37.2)

15.5 (13.4,17.9)
16.5 (14.5,18.7)
18.6 (16.0,21.6)
24.3 (21.7,27.3)
18.0 (13.8,23.4)
22.1 (15.0,32.6)
13.8 (7.5,25.3)
90th
32.8 (29.6,36.3)

27.4 (23.4,32.1)
31.8 (27.8,36.5)
38.3 (33.3,44.0)
32.3 (27.6,37.8)
25.3 (22.8,28.1)

29.3 (26.1,32.9)
36.4 (32.7,40.4)

26.6 (22.8,31.0)
28.4 (23.5,34.3)
32.5 (28.7,36.9)
30.6 (26.6,35.2)
47.7 (39.7,57.4)

27.3 (23.8,31.3)
28.4 (25.2,31.9)
31.3 (27.0,36.4)
38.5 (34.4,43.1)
29.8 (23.3,38.1)
40.5 (28.9,56.8)
27.9 (16.2,48.2)
95th
43.3 (38.8,48.4)

37.4 (31.5,44.5)
41.7 (35.8,48.5)
49.4 (42.9,56.9)
42.9 (36.3,50.7)
34.0 (30.2,38.2)

38.6 (34.1,43.6)
47.9 (42.6,53.8)

35.5 (30.2,41.8)
38.1 (31.5,46.2)
42.9 (37.6,49.0)
40.7 (34.9,47.5)
60.4 (49.0,74.3)

36.9 (31.9,42.6)
37.9 (33.4,43.0)
41.5 (35.6,48.3)
49.6 (43.8,56.3)
39.2 (30.6,50.2)
54.8 (40.1,74.9)
39.4 (24.1,64.5)
97th
51.5 (45.5,58.1)

45.5 (37.7,54.9)
49.3 (41.8,58.0)
57.8 (49.8,67.0)
51.0 (42.8,60.6)
40.5 (35.7,45.9)

45.7 (40.2,52.1)
56.7 (49.8,64.5)

42.2 (35.6,50.0)
45.5 (37.5,55.2)
50.9 (44.1,58.7)
48.6 (41.1,57.4)
70.2 (56.2,87.7)

44.6 (38.2,52.2)
45.5 (39.8,52.1)
49.1 (42.0,57.3)
58.1 (50.7,66.6)
46.1 (35.9,59.1)
64.6 (47.6,87.6)
48.4 (30.4,77.2)
99th
69.4 (60.1,80.2)

64.7 (51.7,81.0)
66.4 (54.7,80.5)
75.6 (64.5,88.5)
68.8 (56.8,83.5)
55.5 (47.8,64.4)

61.8 (53.0,72.2)
75.7 (65.1,88.1)

57.9 (47.6,70.4)
62.2 (50.3,76.8)
68.6 (58.3,80.7)
66.0 (54.4,80.0)
92.4 (71.3,119.7)

62.7 (51.5,76.3)
62.6 (53.4,73.4)
66.4 (55.9,78.9)
76.2 (65.2,89.2)
62.7 (48.4,81.1)
85.5 (63.6,114.9)
67.4 (43.6,104.1)
 Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
 identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
 black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
Table lOb.     UFCR estimates (g/day raw weight, edible portion): Marine fish, adults, 21 years and older, by geographic area
Percentiles(95%CI)
Marine Finfish and Shellfish
Adults (>21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
9.9 (8.5,11.5)

15.0 (12.1,18.6)
7.4 (5.7,9.6)
8.8 (7.6,10.1)
12.2 (9.8,15.2)

9.1 (7.4,11.2)
11.4 (9.6,13.6)

13.0 (10.4,16.1)
13.1 (10.2,16.7)
9.5 (7.9,11.3)
8.0 (6.3,10.2)
14.0 (11.1,17.6)
7.2 (5.4,9.6)
7.9 (6.5,9.5)
11.6 (8.8,15.4)
75th
19.4 (17.4,21.7)

28.1 (23.5,33.4)
14.6 (11.5,18.3)
16.9 (15.3,18.8)
22.5 (18.8,26.9)

18.0 (15.3,21.3)
22.0 (19.1,25.3)

24.1 (20.3,28.6)
24.9 (20.0,30.9)
18.1 (15.7,20.8)
15.4 (12.6,18.9)
26.4 (21.9,31.8)
14.3 (11.1,18.4)
15.2 (13.2,17.6)
21.2 (16.6,27.2)
90th
32.8 (29.6,36.3)

44.4 (37.5,52.6)
24.6 (19.7,30.8)
28.4 (25.2,31.9)
36.1 (30.6,42.7)

30.7 (26.1,35.9)
36.3 (31.8,41.4)

38.6 (33.0,45.2)
40.6 (33.1,50.0)
29.7 (25.4,34.8)
25.5 (20.9,31.1)
41.7 (34.8,50.0)
24.4 (19.2,31.0)
25.7 (22.1,29.9)
33.9 (26.8,43.0)
95th
43.3 (38.8,48.4)

56.7 (47.8,67.3)
32.9 (26.2,41.4)
37.6 (32.7,43.2)
46.7 (39.5,55.2)

40.5 (34.4,47.7)
47.6 (41.5,54.8)

49.9 (42.2,58.9)
52.8 (43.2,64.5)
39.7 (32.8,48.1)
33.5 (27.2,41.2)
53.6 (44.4,64.8)
32.8 (25.6,41.9)
34.1 (28.8,40.3)
43.6 (34.6,55.0)
97th
51.5 (45.5,58.1)

65.8 (55.5,78.1)
39.2 (31.0,49.5)
44.9 (38.4,52.4)
54.9 (46.0,65.5)

48.2 (40.7,57.1)
56.2 (48.6,65.1)

58.5 (49.1,69.8)
61.8 (50.6,75.5)
47.1 (38.5,57.6)
39.8 (32.2,49.1)
62.1 (51.2,75.2)
39.1 (30.4,50.2)
40.7 (33.8,48.9)
51.3 (40.6,64.9)
99th
69.4 (60.1,80.2)

85.4 (70.9,102.7)
53.5 (42.1,68.1)
61.5 (50.6,74.6)
72.2 (59.6,87.6)

65.5 (54.2,79.1)
74.9 (63.7,88.1)

77.0 (63.2,93.7)
81.1 (66.3,99.3)
64.3 (50.6,81.7)
54.5 (43.0,69.0)
80.7 (65.8,98.8)
53.4 (41.4,68.8)
56.6 (45.0,71.3)
68.2 (53.0,87.7)
1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
 TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
 border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
          Table lla.     UFCR estimates (g/day raw weight, edible portion): Total finfish, adults, 21 years and older, by demographic
                           characteristics
c_n
c_n
Percentiles(95%CI)
All Finfish
Adults (>21 yrs)
Age
21 to <35 yrs
35 to <50 yrs
50 to <65 yrs
65+ yrs
Women of childbearing
age (13 to 49 yrs)
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income, finer detail
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
50th
12.8 (11.3,14.6)

9.2 (7.6,11.1)
12.4 (10.6,14.5)
17.1 (14.2,20.6)
13.8 (11.6,16.4)
8.2 (7.1,9.4)

11.4 (10.1,12.9)
14.6 (12.6,16.9)

11.3 (9.1,13.9)
10.7 (8.4,13.5)
12.3 (10.6,14.2)
14.6 (12.3,17.4)
24.2 (18.8,31.3)

10.4 (8.9,12.3)
11.9 (10.1,13.9)
12.2 (10.3,14.5)
15.9 (13.6,18.5)
12.1 (8.3,17.6)
10.7 (6.5,17.4)
4.9 (2.2,11.2)
75th
24.1 (21.9,26.5)

18.7 (16.2,21.6)
22.8 (20.2,25.7)
29.9 (25.7,34.8)
24.9 (21.6,28.9)
16.7 (15.2,18.3)

21.5 (19.5,23.6)
27.2 (24.4,30.3)

21.5 (18.3,25.4)
21.5 (17.1,27.1)
22.9 (20.5,25.5)
27.1 (23.6,31.0)
41.0 (32.7,51.4)

20.8 (18.7,23.3)
22.1 (19.6,24.8)
23.1 (19.8,26.9)
28.2 (24.9,32.0)
23.3 (17.0,32.0)
24.3 (17.0,34.8)
13.0 (6.7,25.2)
90th
39.3 (35.5,43.5)

32.9 (28.3,38.4)
37.1 (32.6,42.1)
46.0 (39.7,53.2)
39.5 (33.9,46.0)
28.6 (25.8,31.7)

34.7 (31.2,38.6)
44.1 (39.4,49.5)

35.4 (30.0,41.7)
36.3 (28.8,45.8)
36.9 (32.8,41.5)
43.8 (38.5,49.8)
62.2 (50.3,76.8)

35.4 (31.5,39.8)
36.1 (32.2,40.4)
37.7 (32.0,44.3)
44.1 (38.5,50.5)
38.9 (29.0,52.2)
43.6 (32.9,57.8)
28.8 (14.8,55.9)
95th
51.2 (45.7,57.3)

44.9 (37.9,53.2)
48.6 (42.0,56.2)
57.9 (49.7,67.5)
50.6 (43.1,59.4)
38.1 (33.9,42.8)

44.9 (39.9,50.6)
57.2 (50.7,64.6)

46.7 (39.3,55.7)
48.4 (38.5,60.8)
47.8 (42.0,54.4)
56.7 (49.7,64.7)
78.8 (63.6,97.5)

47.3 (41.7,53.7)
47.3 (41.9,53.4)
49.2 (41.5,58.3)
56.1 (48.6,64.8)
51.5 (38.2,69.4)
57.7 (44.0,75.6)
41.9 (22.9,76.8)
97th
60.1 (53.2,67.9)

54.6 (45.5,65.5)
57.3 (49.1,66.7)
66.8 (57.1,78.2)
59.0 (50.0,69.6)
45.6 (40.1,51.9)

52.5 (46.2,59.6)
67.1 (59.1,76.1)

55.5 (46.3,66.6)
57.5 (45.7,72.2)
55.9 (48.6,64.2)
66.5 (57.9,76.4)
90.0 (72.4,112.0)

56.4 (49.5,64.4)
55.8 (49.2,63.3)
57.9 (48.7,69.0)
64.8 (55.8,75.3)
61.3 (45.2,83.2)
67.6 (51.5,88.6)
53.0 (30.8,91.1)
99th
80.1 (69.8,92.0)

77.5 (62.5,95.9)
76.7 (64.6,91.0)
85.6 (72.4,101.4)
77.9 (65.0,93.5)
62.2 (53.8,71.8)

69.1 (59.7,80.1)
89.0 (77.4,102.3)

75.6 (61.9,92.3)
78.3 (61.8,99.3)
73.5 (62.9,85.8)
88.1 (75.0,103.6)
114.9 (91.6,144.0)

78.0 (66.9,91.0)
75.0 (65.3,86.2)
78.3 (64.9,94.5)
84.4 (71.2,100.0)
81.4 (59.9,110.7)
87.4 (63.6,120.1)
75.7 (48.9,117.2)
           Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-

           identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
           black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
          Table lib.      UFCR estimates (g/day raw weight, edible portion): Total finfish, adults, 21 years and older, by geographic area
c_n
CS
Percentiles(95%CI)
All Finfish
Adults (>21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
12.8 (11.3,14.6)

15.4 (12.6,18.7)
10.3 (8.1,13.2)
12.5 (10.8,14.4)
14.9 (12.6,17.7)

12.1 (10.3,14.3)
14.2 (12.4,16.2)

15.5 (12.9,18.7)
15.8 (13.3,18.7)
12.5 (10.0,15.5)
10.7 (8.4,13.7)
14.7 (11.8,18.3)
10.2 (7.9,13.2)
11.5 (9.7,13.7)
14.5 (11.5,18.3)
75th
24.1 (21.9,26.5)

28.5 (23.9,33.9)
19.4 (15.9,23.7)
23.4 (21.2,26.0)
27.0 (22.7,32.1)

22.9 (20.0,26.1)
26.4 (23.7,29.4)

28.5 (23.9,34.1)
29.1 (25.0,33.7)
23.2 (19.7,27.5)
19.7 (16.0,24.2)
27.4 (22.7,32.9)
19.3 (15.6,23.8)
21.8 (19.2,24.7)
25.8 (20.4,32.7)
90th
39.3 (35.5,43.5)

45.0 (37.7,53.6)
32.0 (26.4,38.7)
38.4 (34.4,42.9)
43.0 (35.8,51.5)

37.2 (32.5,42.6)
42.9 (38.3,48.0)

45.5 (37.9,54.6)
46.5 (40.2,53.7)
38.2 (31.8,45.9)
32.1 (26.4,39.0)
42.8 (35.9,51.1)
31.9 (26.0,39.1)
35.7 (31.7,40.2)
40.8 (32.1,51.9)
95th
51.2 (45.7,57.3)

57.2 (47.9,68.5)
42.0 (34.5,51.0)
50.1 (44.2,56.8)
55.4 (46.1,66.6)

48.4 (41.9,56.0)
55.7 (49.4,62.9)

59.0 (48.8,71.4)
59.6 (51.4,69.1)
49.6 (40.6,60.6)
42.1 (34.6,51.2)
54.5 (45.7,64.9)
41.8 (33.9,51.5)
46.9 (40.9,53.7)
52.5 (41.3,66.7)
97th
60.1 (53.2,67.9)

66.1 (55.1,79.4)
49.9 (40.7,61.2)
59.0 (51.6,67.6)
64.9 (53.8,78.3)

56.8 (48.8,66.1)
65.2 (57.4,74.0)

69.2 (56.9,84.3)
68.9 (59.1,80.3)
58.7 (47.3,73.0)
49.6 (40.7,60.5)
63.0 (52.8,75.3)
49.9 (40.0,62.2)
55.2 (47.8,63.8)
61.2 (48.2,77.7)
99th
80.1 (69.8,92.0)

85.6 (70.2,104.3)
67.7 (54.5,84.1)
79.0 (67.8,92.2)
86.0 (70.8,104.3)

76.0 (64.3,89.9)
85.5 (74.2,98.5)

90.6 (73.7,111.2)
88.9 (74.8,105.7)
77.4 (62.1,96.5)
67.0 (54.8,81.8)
82.0 (68.4,98.3)
67.8 (53.4,86.1)
75.1 (62.9,89.6)
81.0 (64.2,102.4)
          1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

          2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
Table 12a.      UFCR estimates (g/day raw weight, edible portion): Total shellfish, adults, 21 years and older, by demographic
                  characteristics
Percentiles(95%CI)
All Shellfish
Adults (>21 yrs)
Age
21 to <35 yrs
35 to <50 yrs
50 to <65 yrs
65+ yrs
Women of childbearing age (13 to 49 yrs)
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income, finer detail
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
50th
3.1 (2.4,3.9)

2.7 (2.0,3.7)
3.7 (2.8,4.8)
3.7 (2.8,4.7)
1.9 (1.4,2.7)
2.0 (1.5,2.7)

2.5 (2.0,3.2)
3.8 (3.0,4.9)

3.6 (2.6,4.9)
3.6 (2.2,5.8)
2.9 (2.2,3.7)
2.8 (2.1,3.6)
5.9 (3.9,9.0)

1.9 (1.3,2.6)
2.4 (1.8,3.2)
2.8 (2.1,3.8)
4.9 (3.9,6.0)
2.7 (1.4,5.2)
3.6 (1.7,7.9)
0.9 (0.2,4.9)
75th
7.6 (6.4,9.0)

7.2 (5.7,9.1)
8.8 (6.9,11.2)
8.3 (6.7,10.4)
5.1 (3.8,6.8)
5.4 (4.4,6.6)

6.2 (5.2,7.3)
9.4 (7.8,11.4)

8.9 (6.9,11.6)
9.1 (6.1,13.6)
7.1 (5.9,8.6)
6.9 (5.6,8.4)
13.5 (9.3,19.6)

5.2 (4.1,6.7)
5.9 (4.7,7.4)
6.8 (5.3,8.6)
10.8 (9.1,13.0)
6.6 (4.2,10.3)
10.9 (5.4,22.0)
4.7 (1.4,16.0)
90th
15.6 (13.2,18.5)

15.6 (12.4,19.6)
17.6 (13.6,22.8)
16.0 (12.7,20.1)
11.1 (8.6,14.5)
11.4 (9.5,13.8)

12.5 (10.6,14.8)
19.0 (15.7,23.0)

18.2 (13.8,24.0)
18.2 (12.9,25.7)
14.5 (12.0,17.6)
14.0 (11.4,17.3)
27.0 (18.6,39.3)

11.7 (9.4,14.5)
12.2 (10.0,14.9)
13.8 (10.9,17.5)
20.3 (16.6,24.8)
13.5 (9.6,19.1)
24.6 (13.3,45.3)
17.4 (6.5,46.4)
95th
23.1 (19.2,27.8)

24.1 (18.8,30.7)
25.6 (19.3,33.9)
22.7 (17.8,28.9)
16.9 (13.2,21.6)
17.1 (13.9,20.9)

18.2 (15.1,22.0)
27.9 (22.7,34.3)

26.7 (19.9,35.9)
26.4 (19.0,36.7)
21.4 (17.4,26.3)
20.7 (16.3,26.1)
39.7 (26.9,58.7)

18.1 (14.5,22.6)
18.2 (14.9,22.2)
20.6 (16.0,26.4)
28.3 (22.8,35.3)
19.8 (14.3,27.5)
36.7 (20.6,65.2)
30.1 (12.2,74.3)
97th
29.1 (23.9,35.4)

31.3 (24.2,40.5)
31.8 (23.7,42.6)
28.0 (21.9,35.9)
21.8 (17.0,27.9)
21.8 (17.5,27.0)

22.9 (18.7,28.0)
35.0 (28.2,43.5)

33.6 (24.7,45.8)
32.9 (23.8,45.4)
27.0 (21.7,33.4)
26.0 (20.3,33.2)
50.2 (33.6,74.9)

23.7 (18.7,30.1)
23.5 (19.1,28.9)
26.2 (20.1,34.1)
34.7 (27.6,43.7)
25.1 (18.0,34.8)
45.7 (26.6,78.8)
40.4 (17.8,91.6)
99th
43.7 (35.2,54.2)

49.4 (36.7,66.6)
46.9 (33.9,64.8)
39.9 (31.1,51.3)
33.3 (25.8,43.1)
32.7 (25.5,41.8)

33.7 (27.1,42.0)
51.5 (40.5,65.6)

49.5 (34.7,70.8)
48.1 (34.4,67.3)
40.1 (31.9,50.5)
38.3 (29.5,49.9)
73.4 (48.7,110.6)

37.4 (28.4,49.2)
36.2 (29.1,45.1)
40.3 (29.8,54.4)
49.3 (38.4,63.3)
37.5 (26.4,53.4)
63.7 (37.8,107.5)
60.8 (31.8,116.2)
 Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
 identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
 black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
          Table 12b.      UFCR estimates (g/day raw weight, edible portion): Total shellfish, adults, 21 years and older, by geographic area
c_n
oo
Percentiles(95%CI)
All Shellfish
Adults (>21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
3.1 (2.4,3.9)

5.9 (4.5,7.7)
1.6 (1.1,2.2)
3.4 (2.4,4.8)
3.5 (2.4,5.0)

2.4 (1.9,3.1)
4.7 (3.5,6.2)

4.6 (3.0,7.0)
6.4 (4.8,8.6)
4.8 (3.2,7.4)
2.3 (1.6,3.3)
4.9 (3.4,6.9)
1.4 (1.0,2.0)
2.6 (1.9,3.7)
2.7 (1.9,3.9)
75th
7.6 (6.4,9.0)

13.3 (10.7,16.5)
3.7 (2.7,5.2)
8.0 (6.0,10.7)
8.0 (5.6,11.3)

6.0 (4.8,7.4)
10.9 (8.6,13.8)

10.5 (7.0,15.7)
13.9 (10.9,17.8)
10.8 (7.7,15.2)
5.4 (3.9,7.5)
11.2 (8.2,15.2)
3.3 (2.3,4.6)
6.1 (4.7,7.9)
6.1 (4.4,8.5)
90th
15.6 (13.2,18.5)

24.6 (19.6,30.8)
7.6 (5.4,10.6)
15.7 (11.9,20.8)
15.8 (11.1,22.6)

12.3 (9.8,15.5)
21.0 (16.6,26.4)

20.3 (13.4,30.7)
25.7 (20.3,32.5)
20.1 (14.7,27.4)
10.5 (7.5,14.6)
20.7 (14.8,28.9)
6.6 (4.7,9.4)
12.1 (9.5,15.3)
11.8 (8.5,16.5)
95th
23.1 (19.2,27.8)

34.2 (27.1,43.2)
11.1 (7.8,15.9)
22.7 (17.2,30.1)
23.2 (16.0,33.5)

18.2 (14.2,23.5)
29.8 (23.5,37.8)

29.0 (18.8,44.5)
35.2 (28.0,44.3)
28.4 (21.0,38.4)
15.0 (10.6,21.2)
28.7 (20.1,41.0)
9.8 (6.8,14.2)
17.6 (13.8,22.4)
17.1 (12.1,24.0)
97th
29.1 (23.9,35.4)

41.7 (32.8,53.0)
14.2 (9.8,20.5)
28.4 (21.5,37.5)
29.1 (19.9,42.5)

23.3 (17.8,30.5)
36.9 (29.0,46.9)

36.1 (23.2,56.2)
42.8 (33.8,54.3)
35.0 (26.1,46.9)
18.8 (13.3,26.7)
35.3 (24.4,51.0)
12.6 (8.6,18.5)
22.2 (17.2,28.7)
21.4 (15.0,30.6)
99th
43.7 (35.2,54.2)

58.3 (44.8,75.8)
21.5 (14.6,31.5)
41.7 (31.8,54.7)
43.8 (29.1,65.8)

35.4 (26.4,47.4)
53.2 (41.6,68.0)

53.4 (34.1,83.5)
59.5 (46.0,76.8)
48.6 (37.0,64.0)
27.5 (19.2,39.2)
49.0 (32.9,73.1)
19.2 (12.9,28.7)
33.6 (25.7,44.0)
31.6 (21.9,45.7)
          1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

          2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
Table 13a.     UFCR estimates (g/day raw weight, edible portion): Total trophic level 2 fish, adults, 21 years and older, by demographic
                 characteristics
Trophic Level 2
Finfish and Shellfish
Adults (>21 yrs)
Age
21 to <35 yts
35 to <50 yrs
50 to <65 yrs
65+ yrs
Women of childbearing age (13 to 49 yrs)
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
1.9 (1.5,2.4)

1.6 (1.2,2.2)
2.1 (1.5,2.8)
2.3 (1.7,3.2)
1.4 (1.0,1.9)
1.2 (0.9,1.6)

1.5 (1.2,2.0)
2.4 (1.9,3.1)

2.9 (2.2,3.9)
2.5 (1.7,3.6)
1.7 (1.3,2.2)
2.1 (1.6,2.9)
3.5 (2.2,5.5)

1.2 (0.9,1.7)
1.6 (1.2,2.2)
1.8 (1.3,2.4)
2.7 (2.2,3.4)
1.9 (1.1,3.2)
2.1 (0.8,5.4)
0.6 (0.1,2.8)
75th
4.7 (4.0,5.5)

4.2 (3.5,5.1)
5.3 (4.1,6.8)
5.1 (3.9,6.6)
3.7 (2.8,4.8)
3.1 (2.6,3.9)

3.7 (3.2,4.4)
5.8 (4.9,6.9)

7.2 (5.7,9.0)
6.3 (4.5,8.6)
4.1 (3.4,4.9)
5.1 (4.0,6.4)
7.9 (5.3,11.6)

3.4 (2.6,4.4)
4.0 (3.2,5.0)
4.3 (3.5,5.3)
6.0 (5.1,7.1)
4.5 (3.1,6.6)
6.1 (2.9,12.9)
2.6 (0.8,8.3)
90th
9.6 (8.3,11.0)

9.0 (7.7,10.6)
11.2 (8.5,14.7)
9.6 (7.6,12.2)
7.7 (6.0,9.8)
6.7 (5.6,8.1)

7.5 (6.5,8.7)
11.9 (10.1,14.0)

14.5 (11.5,18.4)
12.7 (9.3,17.5)
8.3 (7.0,9.8)
10.0 (8.0,12.6)
15.1 (10.4,21.9)

7.6 (6.1,9.5)
8.3 (7.0,10.0)
8.8 (7.4,10.5)
11.6 (9.6,14.0)
9.3 (6.8,12.6)
13.5 (7.3,25.2)
9.1 (3.5,23.5)
95th
14.2 (12.1,16.8)

13.7 (11.4,16.5)
16.7 (12.2,22.9)
13.8 (10.9,17.4)
11.4 (8.9,14.5)
10.1 (8.2,12.4)

11.0 (9.3,13.0)
17.5 (14.5,21.1)

21.3 (16.4,27.7)
18.5 (13.1,26.1)
12.3 (10.2,14.9)
14.5 (11.4,18.4)
21.8 (14.8,32.1)

11.9 (9.5,14.8)
12.6 (10.5,15.0)
13.1 (10.7,16.0)
16.5 (13.2,20.7)
13.6 (9.9,18.5)
20.5 (11.5,36.4)
15.6 (6.5,37.5)
97th
18.1 (15.1,21.8)

17.7 (14.4,21.6)
21.4 (15.2,30.3)
17.1 (13.5,21.7)
14.5 (11.2,18.6)
12.9 (10.2,16.3)

13.8 (11.5,16.7)
22.2 (17.9,27.5)

26.9 (20.3,35.5)
23.4 (16.3,33.7)
15.6 (12.7,19.2)
18.3 (14.2,23.6)
27.3 (18.0,41.3)

15.7 (12.4,19.8)
16.2 (13.5,19.4)
16.6 (13.4,20.6)
20.7 (16.0,26.9)
17.6 (12.8,24.3)
26.2 (14.8,46.6)
21.5 (9.2,50.4)
99th
27.7 (21.8,35.2)

27.9 (21.5,36.1)
32.6 (21.9,48.6)
24.9 (19.3,32.2)
21.5 (16.4,28.3)
19.7 (14.6,26.5)

20.6 (16.3,26.1)
33.3 (25.6,43.4)

40.5 (29.0,56.7)
34.2 (22.5,52.1)
23.7 (18.3,30.8)
26.9 (20.4,35.6)
39.8 (24.8,63.9)

25.1 (19.1,32.9)
25.2 (20.2,31.4)
25.2 (19.6,32.5)
30.4 (21.9,42.2)
27.6 (19.3,39.5)
39.0 (19.5,78.3)
36.0 (16.2,79.7)
 Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
 identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
 black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
          Table 13b.      UFCR estimates (g/day raw weight, edible portion): Total trophic level 2 fish, adults, 21 years and older, by geographic
                           area
CS
o
Trophic Level 2
Finfish and Shellfish
Adults (>21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
Percentiles(95%CI)
50th
1.9 (1.5,2.4)

3.1 (2.4,3.9)
1.0 (0.7,1.4)
2.3 (1.6,3.3)
2.0 (1.4,2.8)

1.5 (1.2,2.0)
2.7 (2.0,3.7)

2.5 (1.7,3.8)
3.8 (2.8,5.1)
2.9 (1.9,4.6)
1.4 (0.9,2.1)
2.6 (1.8,3.7)
0.9 (0.6,1.3)
1.9 (1.3,2.6)
1.6 (1.2,2.3)
75th
4.7 (4.0,5.5)

7.1 (5.6,8.9)
2.4 (1.7,3.4)
5.4 (4.2,7.0)
4.7 (3.5,6.2)

3.8 (3.1,4.7)
6.3 (5.1,7.9)

5.8 (4.2,8.1)
8.1 (6.6,10.1)
6.6 (4.8,9.1)
3.3 (2.3,4.8)
6.1 (4.1,8.9)
2.1 (1.5,3.1)
4.4 (3.4,5.6)
3.7 (2.9,4.9)
90th
9.6 (8.3,11.0)

13.6 (10.3,18.0)
4.9 (3.4,7.2)
10.7 (8.7,13.1)
9.3 (7.3,11.9)

7.9 (6.3,10.0)
12.3 (10.2,14.8)

11.3 (8.4,15.3)
15.1 (12.3,18.4)
12.5 (9.7,16.1)
6.6 (4.6,9.6)
11.8 (7.5,18.8)
4.4 (2.9,6.7)
8.7 (7.0,10.9)
7.4 (5.7,9.7)
95th
14.2 (12.1,16.8)

19.6 (14.0,27.3)
7.4 (4.9,11.1)
15.4 (12.6,18.9)
13.7 (10.7,17.7)

11.9 (9.0,15.7)
17.5 (14.5,21.0)

16.2 (12.0,21.9)
20.9 (16.9,25.8)
17.4 (13.7,22.2)
9.6 (6.5,14.2)
17.3 (10.3,29.0)
6.6 (4.3,10.2)
12.9 (10.1,16.5)
11.0 (8.4,14.6)
97th
18.1 (15.1,21.8)

24.3 (16.8,35.1)
9.5 (6.2,14.6)
19.4 (15.7,24.0)
17.4 (13.4,22.6)

15.3 (11.3,20.8)
21.8 (17.9,26.5)

20.4 (15.0,27.7)
25.3 (20.1,31.7)
21.4 (16.8,27.1)
12.0 (8.0,17.9)
21.9 (12.6,38.1)
8.6 (5.4,13.6)
16.4 (12.7,21.3)
14.2 (10.6,19.0)
99th
27.7 (21.8,35.2)

35.9 (23.0,56.1)
14.7 (9.2,23.4)
28.8 (22.7,36.6)
26.8 (20.1,35.6)

24.4 (16.9,35.2)
31.8 (25.3,40.0)

30.2 (21.9,41.6)
35.8 (26.5,48.3)
31.2 (23.7,41.0)
17.9 (11.4,28.0)
33.4 (18.2,61.3)
13.5 (8.2,22.2)
25.5 (18.5,35.1)
22.4 (15.9,31.5)
          1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

          2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
          Table 14a.      UFCR estimates (g/day raw weight, edible portion): Total trophic level 3 fish, adults, 21 years and older, by demographic
                           characteristics
CS
Trophic Level 3
Finfish and Shellfish
Adults (>21 yrs)
Age
21 to <35 yts
35 to <50 yrs
50 to <65 yrs
65+ yrs
Women of childbearing age (13 to 49 yrs)
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
4.7 (3.9,5.7)

3.7 (2.9,4.8)
5.0 (4.1,6.1)
6.0 (4.9,7.4)
4.0 (3.0,5.3)
3.1 (2.5,3.8)

4.0 (3.3,4.9)
5.6 (4.6,6.8)

5.2 (4.0,6.6)
4.4 (3.2,6.1)
4.2 (3.4,5.2)
6.1 (5.1,7.4)
11.9 (9.2,15.5)

3.8 (3.0,4.9)
4.1 (3.3,5.2)
4.4 (3.6,5.5)
6.0 (5.0,7.3)
4.6 (3.0,7.2)
5.1 (3.0,8.4)
1.7 (0.6,4.7)
75th
9.6 (8.5,10.8)

8.2 (6.8,9.8)
9.8 (8.3,11.6)
11.3 (9.5,13.3)
8.2 (6.8,10.0)
6.7 (5.9,7.8)

8.2 (7.1,9.3)
11.2 (9.8,12.8)

10.4 (8.6,12.5)
9.1 (6.8,12.1)
8.4 (7.3,9.7)
11.8 (10.2,13.5)
21.2 (16.8,26.7)

8.4 (7.1,9.9)
8.5 (7.2,10.0)
8.8 (7.5,10.3)
11.5 (9.9,13.3)
9.1 (6.8,12.2)
11.7 (7.7,17.9)
5.4 (2.4,12.4)
90th
16.6 (14.7,18.8)

15.4 (12.9,18.3)
16.8 (14.1,20.0)
18.4 (15.4,21.9)
14.6 (12.7,16.9)
12.4 (10.8,14.2)

14.1 (12.4,16.1)
19.2 (16.6,22.2)

17.7 (14.6,21.3)
15.6 (11.9,20.4)
14.5 (12.5,16.7)
19.5 (16.8,22.5)
33.2 (26.0,42.3)

15.3 (13.2,17.8)
14.8 (12.8,17.2)
15.3 (13.2,17.9)
19.1 (16.2,22.5)
15.2 (11.8,19.5)
21.6 (14.8,31.7)
14.7 (6.6,32.6)
95th
22.4 (19.4,25.9)

22.0 (17.9,27.1)
22.5 (18.7,27.1)
23.9 (19.7,28.9)
20.1 (17.4,23.2)
17.1 (14.6,19.9)

19.0 (16.4,22.0)
25.7 (21.8,30.4)

23.5 (19.2,28.9)
21.0 (16.0,27.5)
19.3 (16.5,22.7)
25.6 (21.7,30.2)
42.4 (31.9,56.2)

21.1 (18.0,24.8)
20.2 (17.4,23.5)
20.9 (17.5,24.8)
25.0 (20.8,30.2)
20.2 (15.9,25.7)
29.1 (19.1,44.2)
22.8 (10.9,47.5)
97th
27.0 (23.0,31.6)

27.6 (21.6,35.3)
26.9 (22.0,32.7)
28.1 (22.9,34.4)
24.3 (20.8,28.4)
20.8 (17.6,24.6)

22.7 (19.4,26.6)
30.8 (25.7,36.9)

28.1 (22.5,35.2)
25.2 (19.0,33.3)
23.1 (19.3,27.5)
30.1 (25.2,35.9)
49.1 (35.9,67.1)

25.8 (21.7,30.6)
24.5 (20.8,28.7)
25.3 (20.8,30.7)
29.5 (24.3,35.9)
23.9 (18.6,30.6)
35.0 (22.4,54.7)
29.4 (15.1,57.1)
99th
37.4 (30.7,45.7)

41.3 (29.1,58.5)
36.5 (29.3,45.5)
37.4 (29.3,47.7)
33.8 (27.9,40.8)
29.6 (24.4,35.9)

31.1 (25.8,37.4)
42.4 (33.9,53.1)

38.2 (29.5,49.5)
34.2 (25.0,46.9)
31.4 (25.4,38.9)
40.0 (32.6,49.3)
62.0 (43.1,89.3)

37.0 (29.8,46.0)
34.7 (28.6,42.2)
35.4 (27.7,45.3)
39.6 (31.6,49.8)
32.2 (24.7,42.1)
47.7 (29.3,77.5)
42.1 (24.6,72.1)
           Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
           identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
           black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
          Table 14b.      UFCR estimates (g/day raw weight, edible portion): Total trophic level 3 fish, adults, 21 years and older, by geographic
                           area
CS
Trophic Level 3
Finfish and Shellfish
Adults (>21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
Percentiles(95%CI)
50th
4.7 (3.9,5.7)

5.9 (4.5,7.7)
3.0 (2.2,3.9)
5.4 (4.3,6.9)
5.6 (4.3,7.1)

4.1 (3.3,5.1)
6.0 (4.9,7.4)

6.4 (4.9,8.4)
7.2 (5.7,9.1)
6.3 (4.5,8.7)
3.7 (2.9,4.7)
5.3 (3.9,7.1)
2.7 (2.0,3.7)
4.7 (3.7,6.0)
4.9 (3.7,6.6)
75th
9.6 (8.5,10.8)

11.7 (9.3,14.6)
5.8 (4.7,7.2)
10.6 (8.9,12.6)
10.8 (8.7,13.4)

8.3 (7.1,9.8)
11.8 (10.1,13.7)

12.7 (9.8,16.5)
13.2 (11.1,15.8)
12.2 (9.7,15.2)
7.0 (5.7,8.6)
10.6 (8.3,13.6)
5.4 (4.3,6.9)
9.3 (7.8,11.1)
9.4 (7.3,12.0)
90th
16.6 (14.7,18.8)

19.4 (15.6,24.1)
10.1 (8.2,12.5)
17.9 (15.1,21.1)
18.6 (14.7,23.4)

14.6 (12.5,17.1)
19.9 (17.0,23.2)

21.7 (16.5,28.6)
21.6 (18.3,25.4)
20.2 (16.7,24.5)
11.7 (9.4,14.5)
17.7 (13.9,22.7)
9.5 (7.6,11.9)
15.9 (13.5,18.8)
15.8 (12.3,20.2)
95th
22.4 (19.4,25.9)

25.3 (20.3,31.5)
13.6 (11.0,17.0)
23.8 (19.9,28.5)
25.1 (19.5,32.3)

19.7 (16.6,23.5)
26.5 (22.3,31.4)

29.4 (21.9,39.6)
27.9 (23.6,33.1)
26.8 (21.9,32.8)
15.4 (12.2,19.4)
23.4 (18.1,30.2)
13.0 (10.2,16.4)
21.3 (17.8,25.4)
21.0 (16.3,27.1)
97th
27.0 (23.0,31.6)

29.7 (23.7,37.3)
16.5 (13.1,20.8)
28.4 (23.6,34.1)
30.3 (23.2,39.6)

23.7 (19.7,28.6)
31.7 (26.2,38.3)

35.3 (25.7,48.5)
32.8 (27.3,39.4)
31.9 (25.9,39.4)
18.3 (14.4,23.4)
27.5 (21.1,35.8)
15.7 (12.3,20.0)
25.5 (21.0,30.9)
25.2 (19.2,33.0)
99th
37.4 (30.7,45.7)

39.7 (30.9,51.0)
23.0 (17.9,29.6)
38.8 (31.5,47.9)
42.1 (30.6,57.9)

32.8 (26.4,40.7)
43.3 (34.0,54.9)

48.2 (32.8,70.8)
43.7 (34.9,54.5)
43.4 (34.0,55.5)
24.9 (19.2,32.4)
36.9 (27.8,48.9)
22.2 (17.0,29.0)
35.0 (28.1,43.6)
34.2 (25.6,45.7)
          1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

          2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
          Table 15a.      UFCR estimates (g/day raw weight, edible portion): Total trophic level 4 fish, adults, 21 years and older, by demographic
                           characteristics
CS
Trophic Level 4
Finfish and Shellfish
Adults (>21 yrs)
Age
21 to <35 yts
35 to <50 yrs
50 to <65 yrs
65+ yrs
Women of childbearing age (13 to 49 yrs)
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
8.6 (7.5,9.9)

6.0 (4.8,7.4)
8.3 (7.1,9.8)
11.8 (9.7,14.3)
9.5 (7.7,11.6)
5.4 (4.6,6.3)

7.8 (6.8,8.9)
9.7 (8.3,11.3)

6.4 (5.2,8.0)
6.8 (5.3,8.8)
8.8 (7.5,10.2)
8.4 (6.9,10.1)
14.2 (10.6,19.0)

6.9 (5.7,8.2)
7.7 (6.5,9.2)
8.3 (6.9,9.9)
11.1 (9.5,13.0)
7.9 (5.3,11.8)
7.3 (4.4,12.1)
3.7 (1.7,8.4)
75th
17.1 (15.5,18.8)

12.9 (10.9,15.3)
16.0 (14.1,18.2)
21.5 (18.3,25.3)
17.8 (15.1,21.0)
11.7 (10.5,13.0)

15.4 (14.1,16.9)
19.0 (17.0,21.2)

13.1 (11.1,15.4)
14.2 (11.1,18.1)
17.1 (15.4,19.1)
16.4 (14.1,19.2)
25.6 (20.0,32.7)

14.5 (12.9,16.3)
15.1 (13.4,17.1)
16.4 (14.0,19.1)
20.5 (18.1,23.1)
16.0 (11.3,22.6)
18.0 (12.2,26.6)
10.9 (5.3,22.2)
90th
28.8 (26.1,31.9)

23.8 (19.9,28.6)
26.7 (23.2,30.8)
34.4 (29.3,40.3)
29.3 (24.7,34.7)
20.8 (18.7,23.1)

25.9 (23.3,28.7)
32.0 (28.6,35.9)

22.8 (19.3,26.8)
25.2 (19.9,31.9)
28.7 (25.7,32.1)
28.0 (24.0,32.8)
39.6 (31.3,50.3)

25.6 (22.6,29.0)
25.7 (22.7,29.1)
27.6 (23.4,32.5)
33.0 (29.1,37.4)
27.7 (19.8,38.7)
34.6 (25.4,47.1)
23.5 (11.8,46.5)
95th
38.3 (34.1,42.9)

33.2 (27.2,40.5)
35.4 (30.3,41.5)
44.4 (37.6,52.4)
38.5 (32.1,46.2)
28.2 (25.1,31.6)

34.1 (30.3,38.3)
42.6 (37.5,48.5)

30.7 (25.7,36.7)
34.1 (27.1,43.0)
38.0 (33.5,43.1)
37.3 (31.8,43.9)
51.1 (40.3,64.9)

34.9 (30.3,40.2)
34.4 (30.1,39.3)
36.7 (30.8,43.8)
42.7 (37.3,48.9)
36.8 (26.2,51.7)
47.4 (35.8,62.8)
33.8 (17.8,64.3)
97th
45.5 (40.1,51.7)

40.9 (33.0,50.5)
42.2 (35.6,50.1)
51.8 (43.5,61.7)
45.3 (37.5,54.9)
33.9 (29.9,38.4)

40.3 (35.4,45.9)
50.5 (44.1,57.8)

37.3 (30.7,45.3)
41.1 (32.9,51.5)
45.1 (39.3,51.8)
44.5 (37.6,52.7)
59.5 (46.4,76.4)

42.4 (36.1,49.7)
41.1 (35.7,47.4)
43.8 (36.4,52.7)
50.0 (43.2,57.9)
44.2 (31.3,62.4)
56.4 (43.1,73.9)
41.6 (23.2,74.6)
99th
61.6 (53.2,71.2)

58.4 (45.9,74.3)
57.0 (47.1,69.0)
68.1 (56.5,82.0)
60.8 (49.2,75.0)
46.7 (40.4,54.0)

54.2 (46.3,63.3)
68.4 (58.4,80.1)

52.9 (42.0,66.6)
57.2 (45.7,71.6)
60.8 (52.0,71.1)
60.7 (50.4,73.1)
80.0 (59.9,107.0)

59.3 (49.0,71.8)
56.8 (47.9,67.3)
59.2 (48.5,72.1)
66.1 (56.2,77.8)
60.5 (43.2,84.7)
74.9 (57.1,98.2)
60.7 (37.9,97.3)
           Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
           identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
           black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
          Table 15b.      UFCR estimates (g/day raw weight, edible portion): Total trophic level 4 fish, adults, 21 years and older, by geographic
                           area
CS
Trophic Level 4
Finfish and Shellfish
Adults (>21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
Percentiles(95%CI)
50th
8.6 (7.5,9.9)

11.5 (9.4,14.1)
7.2 (5.4,9.6)
7.6 (6.6,8.9)
10.2 (8.5,12.3)

8.3 (7.0,9.8)
9.3 (7.9,11.0)

10.5 (8.5,12.9)
10.4 (8.4,12.9)
7.6 (6.0,9.5)
7.0 (5.2,9.5)
11.1 (8.8,13.8)
7.3 (5.4,9.9)
7.1 (5.9,8.5)
10.0 (7.8,12.9)
75th
17.1 (15.5,18.8)

22.0 (18.4,26.2)
14.5 (11.5,18.3)
15.1 (13.4,17.0)
19.1 (16.1,22.8)

16.3 (14.3,18.7)
18.4 (16.3,20.9)

19.9 (16.5,24.1)
20.4 (17.0,24.5)
14.9 (12.3,18.1)
14.1 (11.0,18.2)
21.1 (17.6,25.3)
14.6 (11.4,18.7)
13.9 (12.0,16.2)
18.5 (14.5,23.5)
90th
28.8 (26.1,31.9)

36.0 (30.1,43.1)
24.8 (19.9,30.8)
25.8 (22.6,29.4)
31.2 (25.9,37.6)

27.5 (23.9,31.6)
31.2 (27.5,35.4)

32.8 (27.0,39.8)
34.5 (28.9,41.2)
25.7 (21.0,31.6)
24.4 (19.2,30.9)
34.2 (28.7,40.7)
24.8 (19.7,31.2)
23.7 (20.3,27.7)
30.0 (23.4,38.4)
95th
38.3 (34.1,42.9)

46.7 (38.8,56.2)
33.1 (26.6,41.3)
34.2 (29.6,39.6)
40.8 (33.5,49.7)

36.4 (31.3,42.3)
41.4 (36.2,47.3)

42.7 (34.8,52.3)
45.7 (38.1,54.8)
33.8 (27.5,41.6)
32.6 (25.8,41.1)
44.0 (37.0,52.3)
33.2 (26.3,42.0)
31.6 (26.6,37.5)
39.0 (30.4,50.1)
97th
45.5 (40.1,51.7)

55.0 (45.5,66.5)
39.8 (31.6,50.0)
40.9 (34.9,48.0)
48.0 (39.2,58.8)

43.4 (36.9,50.9)
49.1 (42.7,56.5)

50.4 (40.7,62.5)
53.9 (44.9,64.7)
40.3 (32.6,49.8)
39.0 (30.9,49.2)
51.6 (43.1,62.0)
40.0 (31.3,51.1)
37.5 (31.3,44.8)
45.9 (35.8,58.9)
99th
61.6 (53.2,71.2)

73.1 (60.1,89.0)
54.5 (42.7,69.6)
56.0 (46.5,67.3)
63.5 (51.4,78.4)

58.7 (49.1,70.2)
65.9 (56.3,77.1)

66.8 (53.0,84.1)
72.1 (59.0,88.0)
53.9 (43.0,67.5)
53.7 (42.0,68.8)
68.7 (56.6,83.4)
54.7 (42.2,71.0)
52.2 (41.8,65.1)
60.8 (47.6,77.7)
          1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

          2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
         Table 16a.      UFCR estimates (g/day raw weight, edible portion): Total freshwater + estuarine trophic level 2 fish, adults, 21 years and
                           older, by demographic characteristics
CS
c_n
Freshwater + Estuarine Trophic Level 2
Finfish and Shellfish
Adults (>21 yrs)
Age
21 to <35 yrs
35 to <50 yrs
50 to <65 yrs
65+ yrs
Women of childbearing age (13 to 49 yrs)
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
1.5 (1.1,1.9)

1.4 (1.0,1.9)
1.6 (1.2,2.3)
1.7 (1.2,2.4)
1.0 (0.7,1.4)
0.9 (0.7,1.3)

1.2 (0.9,1.6)
1.8 (1.4,2.3)

2.5 (1.8,3.3)
2.0 (1.3,3.0)
1.2 (0.9,1.6)
1.8 (1.3,2.4)
3.0 (1.9,4.8)

1.0 (0.7,1.3)
1.2 (0.9,1.7)
1.4 (1.0,1.9)
2.1 (1.6,2.7)
1.4 (0.8,2.5)
1.6 (0.6,4.3)
0.5 (0.1,2.3)
75th
3.6 (3.0,4.4)

3.6 (2.9,4.3)
4.2 (3.1,5.6)
3.8 (2.9,5.0)
2.7 (2.0,3.6)
2.6 (2.0,3.2)

3.0 (2.5,3.7)
4.4 (3.6,5.4)

6.2 (4.9,7.9)
5.1 (3.5,7.3)
3.1 (2.5,3.8)
4.3 (3.4,5.5)
6.9 (4.6,10.3)

2.7 (2.1,3.5)
3.1 (2.5,4.0)
3.4 (2.8,4.2)
4.6 (3.7,5.7)
3.5 (2.3,5.2)
4.7 (2.1,10.1)
2.1 (0.7,6.4)
90th
7.6 (6.4,9.1)

7.8 (6.6,9.3)
8.9 (6.5,12.1)
7.4 (5.7,9.5)
5.7 (4.4,7.5)
5.6 (4.5,7.0)

6.2 (5.1,7.5)
9.3 (7.6,11.2)

12.9 (9.9,16.6)
10.5 (7.3,15.1)
6.3 (5.1,7.7)
8.6 (6.7,11.1)
13.3 (9.0,19.4)

6.4 (5.1,7.9)
6.7 (5.5,8.3)
7.0 (5.8,8.5)
9.0 (7.1,11.4)
7.4 (5.4,10.2)
10.5 (5.6,19.8)
7.4 (2.9,19.2)
95th
11.5 (9.4,14.0)

12.0 (9.8,14.6)
13.4 (9.5,18.7)
10.6 (8.2,13.8)
8.7 (6.6,11.4)
8.5 (6.6,10.9)

9.2 (7.4,11.3)
13.8 (11.1,17.2)

19.1 (14.3,25.5)
15.3 (10.3,22.7)
9.3 (7.5,11.6)
12.6 (9.6,16.5)
19.0 (12.6,28.4)

10.0 (8.0,12.7)
10.4 (8.5,12.7)
10.5 (8.5,12.9)
13.0 (9.9,17.0)
11.1 (8.0,15.3)
16.3 (9.0,29.5)
13.1 (5.5,31.3)
97th
14.7 (11.8,18.3)

15.5 (12.5,19.2)
17.3 (12.1,24.7)
13.3 (10.1,17.4)
11.1 (8.3,14.8)
11.0 (8.4,14.4)

11.6 (9.2,14.7)
17.8 (14.0,22.5)

24.5 (18.0,33.5)
19.6 (12.9,29.6)
11.9 (9.3,15.1)
15.8 (11.9,21.1)
23.6 (15.4,36.3)

13.3 (10.4,17.0)
13.6 (10.9,16.8)
13.4 (10.7,16.9)
16.3 (12.1,21.9)
14.5 (10.4,20.2)
21.6 (12.0,38.6)
17.8 (7.8,40.4)
99th
23.0 (17.6,30.2)

24.7 (18.7,32.5)
27.1 (18.1,40.6)
20.0 (14.6,27.2)
16.8 (12.3,23.0)
17.1 (12.4,23.7)

17.7 (13.3,23.5)
27.5 (20.6,36.6)

37.6 (26.3,53.9)
29.7 (18.7,47.3)
18.2 (13.6,24.3)
23.6 (17.2,32.6)
35.0 (21.2,57.8)

21.9 (16.1,29.8)
21.6 (16.8,27.7)
21.3 (16.0,28.3)
24.5 (17.1,35.1)
23.1 (16.0,33.5)
34.3 (17.1,68.7)
30.4 (14.0,65.8)
           Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
           identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
           black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
CS
CS
         Table 16b.      UFCR estimates (g/day raw weight, edible portion): Total freshwater + estuarine trophic level 2 fish, adults, 21 years and
                           older, by geographic area
Freshwater + Estuarine Trophic Level 2
Finfish and Shellfish
Adults (>21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
1.5 (1.1,1.9)

2.1 (1.5,2.8)
0.8 (0.5,1.1)
1.9 (1.3,2.8)
1.5 (1.0,2.3)

1.2 (0.9,1.5)
2.1 (1.5,2.9)

1.9 (1.2,3.1)
2.8 (2.0,3.9)
2.3 (1.4,3.7)
1.1 (0.7,1.7)
1.7 (1.1,2.7)
0.7 (0.5,1.0)
1.6 (1.1,2.2)
1.3 (0.9,1.8)
75th
3.6 (3.0,4.4)

4.8 (3.5,6.6)
1.9 (1.3,2.8)
4.5 (3.4,5.9)
3.7 (2.7,5.0)

3.0 (2.4,3.8)
4.9 (3.9,6.3)

4.6 (3.1,6.8)
6.2 (4.8,8.0)
5.3 (3.8,7.3)
2.7 (1.8,3.9)
4.1 (2.6,6.6)
1.7 (1.1,2.5)
3.7 (2.8,4.9)
2.9 (2.2,4.0)
Percentiles(95%CI)
90th
7.6 (6.4,9.1)

9.5 (6.6,13.6)
4.1 (2.7,6.1)
9.0 (7.2,11.3)
7.7 (5.7,10.2)

6.3 (4.8,8.3)
9.8 (7.9,12.1)

9.2 (6.5,13.2)
11.6 (9.1,14.8)
10.4 (8.0,13.5)
5.4 (3.7,8.0)
8.2 (4.8,14.1)
3.6 (2.3,5.6)
7.6 (5.8,9.8)
6.2 (4.6,8.3)
95th
11.5 (9.4,14.0)

13.9 (9.3,20.7)
6.2 (4.0,9.5)
13.2 (10.6,16.4)
11.5 (8.6,15.4)

9.6 (7.1,13.0)
14.1 (11.4,17.5)

13.5 (9.5,19.0)
16.4 (12.7,21.2)
14.6 (11.5,18.6)
7.9 (5.3,12.0)
12.1 (6.7,21.9)
5.5 (3.5,8.9)
11.3 (8.6,15.0)
9.3 (6.8,12.9)
97th
14.7 (11.8,18.3)

17.6 (11.5,27.1)
8.0 (5.1,12.7)
16.7 (13.3,21.0)
14.7 (10.9,19.9)

12.5 (9.0,17.4)
17.7 (14.1,22.2)

16.9 (12.0,23.9)
20.4 (15.4,27.1)
18.4 (14.2,23.8)
10.1 (6.6,15.5)
15.5 (8.3,28.8)
7.2 (4.4,11.9)
14.6 (10.8,19.6)
12.3 (8.7,17.3)
99th
23.0 (17.6,30.2)

27.0 (16.4,44.4)
12.8 (7.7,21.2)
25.4 (19.4,33.2)
23.2 (16.7,32.1)

20.2 (13.9,29.3)
26.5 (20.3,34.8)

25.9 (18.1,37.2)
29.6 (20.9,42.1)
27.1 (20.3,36.2)
15.6 (9.7,25.1)
24.3 (12.4,47.5)
11.6 (6.7,20.0)
23.1 (16.3,32.9)
20.0 (13.6,29.4)
         1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

         2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
         Table 17a.      UFCR estimates (g/day raw weight, edible portion): Total freshwater + estuarine trophic level 3 fish, adults, 21 years and
                           older, by demographic characteristics
CS
Freshwater + Estuarine Trophic Level 3
Finfish and Shellfish
Adults (>21 yrs)
Age
21 to <35 yrs
35 to <50 yrs
50 to <65 yrs
65+ yrs
Women of childbearing age (13 to 49 yrs)
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
2.0 (1.6,2.5)

1.6 (1.2,2.1)
2.1 (1.6,2.8)
2.6 (2.0,3.4)
1.5 (1.2,2.1)
1.2 (0.9,1.5)

1.6 (1.3,2.1)
2.5 (2.0,3.2)

2.6 (2.0,3.4)
2.3 (1.6,3.3)
1.7 (1.3,2.2)
2.7 (2.2,3.3)
5.3 (4.0,7.0)

1.5 (1.1,2.1)
1.7 (1.3,2.2)
1.9 (1.5,2.4)
2.7 (2.1,3.4)
1.8 (1.1,2.9)
2.4 (1.3,4.2)
0.9 (0.3,2.8)
75th
4.5 (3.9,5.3)

3.9 (3.2,4.9)
4.8 (3.8,6.0)
5.3 (4.3,6.5)
3.6 (2.8,4.7)
3.0 (2.5,3.5)

3.7 (3.1,4.3)
5.5 (4.7,6.6)

5.7 (4.6,7.1)
5.2 (3.7,7.4)
3.8 (3.2,4.6)
5.7 (4.8,6.7)
10.3 (7.8,13.6)

3.9 (3.1,4.8)
3.9 (3.2,4.7)
4.1 (3.4,5.1)
5.6 (4.7,6.8)
4.0 (2.9,5.5)
6.0 (3.8,9.5)
3.2 (1.3,7.7)
90th
8.6 (7.2,10.2)

8.1 (6.4,10.3)
9.0 (7.2,11.2)
9.4 (7.5,11.8)
7.0 (5.6,8.9)
6.0 (4.9,7.2)

6.9 (5.8,8.3)
10.3 (8.5,12.5)

10.5 (8.2,13.4)
9.7 (6.8,13.7)
7.2 (5.9,8.7)
10.3 (8.5,12.5)
17.7 (12.4,25.4)

7.9 (6.6,9.5)
7.5 (6.2,9.0)
7.7 (6.2,9.6)
10.1 (8.2,12.4)
7.4 (5.5,10.1)
12.2 (8.0,18.6)
8.9 (4.1,19.6)
95th
12.2 (9.9,15.0)

12.4 (9.3,16.5)
12.7 (9.9,16.2)
12.9 (9.9,16.7)
10.0 (7.9,12.7)
8.7 (7.0,10.9)

9.7 (7.9,12.0)
14.6 (11.6,18.4)

14.6 (11.1,19.2)
13.7 (9.5,19.7)
10.1 (8.1,12.7)
14.3 (11.4,17.9)
24.2 (15.5,37.6)

11.7 (9.7,14.2)
10.7 (8.6,13.3)
11.0 (8.5,14.2)
13.9 (10.9,17.7)
10.5 (7.6,14.5)
17.6 (10.7,29.1)
14.8 (7.2,30.6)
97th
15.2 (12.0,19.3)

16.1 (11.6,22.4)
15.6 (12.0,20.4)
15.7 (11.7,20.9)
12.4 (9.6,15.9)
11.1 (8.7,14.1)

12.0 (9.4,15.3)
18.1 (14.0,23.5)

18.0 (13.3,24.2)
16.9 (11.5,24.8)
12.5 (9.7,16.0)
17.5 (13.6,22.6)
29.5 (17.9,48.5)

15.0 (12.2,18.4)
13.4 (10.5,17.1)
13.7 (10.2,18.3)
17.0 (12.9,22.3)
12.9 (9.1,18.4)
22.2 (12.8,38.5)
19.8 (9.9,39.8)
99th
22.5 (16.6,30.6)

25.8 (16.5,40.5)
22.6 (16.5,31.1)
22.2 (15.6,31.7)
17.7 (13.0,24.0)
16.7 (12.4,22.7)

17.5 (12.9,23.7)
26.5 (19.0,37.1)

25.6 (18.1,36.2)
24.4 (16.0,37.3)
18.2 (13.3,24.7)
25.3 (18.4,34.9)
41.9 (23.5,74.7)

23.4 (18.1,30.1)
20.1 (14.8,27.2)
20.3 (13.7,30.2)
24.2 (17.3,34.0)
18.4 (12.1,28.0)
32.7 (17.3,61.7)
30.2 (16.1,56.4)
           Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
           identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
           black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
CS
oo
         Table 17b.      UFCR estimates (g/day raw weight, edible portion): Total freshwater + estuarine trophic level 3 fish, adults, 21 years and
                           older, by geographic area
Freshwater + Estuarine Trophic Level 3
Finfish and Shellfish
Adults (>21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
2.0 (1.6,2.5)

2.6 (2.0,3.4)
1.0 (0.8,1.4)
2.7 (1.9,3.7)
2.3 (1.7,3.0)

1.7 (1.3,2.1)
2.8 (2.1,3.7)

2.8 (1.9,4.1)
3.6 (2.7,4.7)
3.2 (2.0,5.0)
1.4 (1.0,2.0)
2.3 (1.6,3.3)
0.9 (0.7,1.3)
2.2 (1.7,3.0)
1.9 (1.4,2.5)
75th
4.5 (3.9,5.3)

5.5 (4.4,6.8)
2.3 (1.8,2.9)
5.6 (4.3,7.3)
4.9 (3.7,6.5)

3.8 (3.2,4.4)
5.9 (4.7,7.5)

6.0 (4.2,8.7)
7.1 (5.8,8.8)
6.6 (4.8,9.2)
3.0 (2.3,3.8)
4.9 (3.7,6.6)
2.1 (1.6,2.7)
4.7 (3.8,5.9)
3.9 (3.1,5.0)
Percentiles(95%CI)
90th
8.6 (7.2,10.2)

9.7 (7.7,12.1)
4.3 (3.4,5.5)
10.2 (7.9,13.0)
9.2 (6.6,12.8)

7.2 (5.9,8.7)
10.9 (8.7,13.7)

11.4 (7.7,16.9)
12.3 (10.0,15.2)
11.9 (8.9,15.7)
5.4 (4.3,6.9)
8.8 (6.5,11.8)
3.9 (3.0,5.2)
8.6 (6.7,10.9)
7.3 (5.4,9.9)
95th
12.2 (9.9,15.0)

13.1 (10.2,16.8)
6.2 (4.8,8.1)
14.2 (10.9,18.4)
13.3 (9.1,19.4)

10.2 (8.1,12.7)
15.3 (11.8,19.7)

16.3 (10.5,25.5)
16.6 (13.1,21.0)
16.4 (12.5,21.7)
7.6 (5.9,9.7)
11.9 (8.7,16.4)
5.7 (4.2,7.7)
11.9 (9.1,15.5)
10.3 (7.2,14.6)
97th
15.2 (12.0,19.3)

15.7 (12.0,20.6)
7.8 (5.9,10.3)
17.5 (13.2,23.1)
16.7 (11.0,25.3)

12.6 (9.8,16.3)
18.9 (14.2,25.1)

20.5 (12.5,33.7)
20.1 (15.4,26.2)
20.2 (15.1,27.0)
9.3 (7.1,12.1)
14.3 (10.2,20.1)
7.2 (5.2,10.0)
14.7 (11.0,19.6)
12.9 (8.8,18.9)
99th
22.5 (16.6,30.6)

22.2 (15.8,31.3)
12.0 (8.7,16.5)
25.2 (18.5,34.4)
25.2 (15.1,42.3)

18.6 (13.6,25.4)
27.7 (19.1,40.1)

31.0 (16.6,57.7)
28.5 (20.4,39.7)
28.9 (20.9,40.0)
13.7 (10.1,18.7)
20.3 (13.7,30.2)
11.2 (7.7,16.3)
21.4 (15.2,30.0)
19.0 (12.2,29.4)
         1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

         2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
         Table 18a.      UFCR estimates (g/day raw weight, edible portion): Total freshwater + estuarine trophic level 4 fish, adults, 21 years and
                           older, by demographic characteristics
CS
Freshwater + Estuarine Trophic Level 4
Finfish and Shellfish
Adults (>21 yrs)
Age
21 to <35 yrs
35 to <50 yrs
50 to <65 yrs
65+ yrs
Women of childbearing age (13 to 49 yrs)
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
0.6 (0.4,0.9)

0.3 (0.2,0.5)
0.6 (0.4,0.9)
1.0 (0.7,1.5)
0.8 (0.5,1.3)
0.3 (0.2,0.5)

0.5 (0.4,0.7)
0.7 (0.5,1.1)

0.6 (0.4,0.9)
0.5 (0.3,0.9)
0.5 (0.4,0.8)
1.1 (0.7,1.6)
2.4 (1.3,4.3)

0.5 (0.3,0.7)
0.5 (0.4,0.8)
0.6 (0.4,0.9)
0.8 (0.5,1.1)
0.6 (0.3,1.2)
0.6 (0.3,1.2)
0.1 (0.0,0.5)
75th
1.9 (1.5,2.5)

1.0 (0.7,1.4)
1.7 (1.2,2.3)
2.9 (2.1,4.0)
2.4 (1.5,3.7)
1.0 (0.7,1.5)

1.6 (1.2,2.1)
2.3 (1.7,3.0)

1.9 (1.3,2.7)
1.7 (0.9,3.0)
1.5 (1.2,2.0)
3.2 (2.3,4.5)
6.6 (3.8,11.3)

1.6 (1.2,2.2)
1.7 (1.2,2.3)
1.9 (1.4,2.5)
2.2 (1.7,2.9)
2.0 (1.2,3.6)
2.2 (1.3,3.5)
0.9 (0.3,2.4)
90th
5.1 (4.0,6.4)

3.0 (2.2,4.1)
4.2 (3.1,5.8)
7.4 (5.4,10.2)
6.0 (3.8,9.5)
2.9 (2.1,3.9)

4.2 (3.3,5.4)
6.0 (4.6,7.8)

5.1 (3.6,7.3)
4.6 (2.5,8.7)
3.9 (3.0,5.1)
8.4 (6.2,11.4)
16.1 (9.3,27.7)

4.7 (3.6,6.0)
4.5 (3.4,6.1)
5.2 (4.0,6.7)
5.5 (4.2,7.2)
5.5 (3.2,9.3)
6.7 (4.3,10.5)
3.7 (1.5,9.0)
95th
9.1 (7.0,11.7)

5.8 (4.2,8.1)
7.3 (5.2,10.2)
12.8 (9.0,18.2)
10.6 (6.7,16.8)
5.3 (3.8,7.4)

7.5 (5.7,9.7)
10.8 (8.2,14.3)

9.2 (6.3,13.6)
8.2 (4.1,16.5)
6.8 (5.1,9.0)
14.7 (10.7,20.2)
28.2 (16.0,49.7)

8.6 (6.5,11.2)
8.3 (6.1,11.2)
9.4 (7.0,12.6)
9.4 (7.0,12.7)
9.5 (5.8,15.5)
12.8 (7.9,20.7)
8.2 (3.7,18.3)
97th
13.2 (10.0,17.4)

9.1 (6.2,13.3)
10.4 (7.2,15.0)
18.2 (12.6,26.4)
15.1 (9.5,24.0)
7.7 (5.4,11.0)

10.8 (8.1,14.3)
15.8 (11.7,21.3)

13.5 (8.8,20.6)
11.7 (5.5,25.1)
9.6 (7.2,13.0)
21.1 (15.0,29.6)
40.2 (22.5,72.0)

12.6 (9.5,16.7)
12.1 (8.8,16.7)
13.7 (10.0,18.8)
13.5 (9.8,18.6)
13.7 (8.5,22.2)
19.6 (11.5,33.4)
13.0 (6.1,27.8)
99th
27.1 (19.4,38.0)

20.9 (12.1,36.0)
20.4 (13.2,31.7)
37.0 (23.4,58.4)
29.7 (18.1,48.7)
16.1 (10.4,24.8)

21.4 (15.2,30.0)
32.8 (22.7,47.6)

28.8 (16.9,49.0)
23.4 (9.8,55.9)
18.7 (13.1,26.7)
41.2 (28.0,60.8)
77.9 (41.4,146.5)

26.4 (18.8,37.2)
25.1 (17.4,36.1)
28.5 (19.6,41.4)
26.9 (17.9,40.4)
27.2 (16.8,44.2)
43.7 (23.6,80.6)
28.3 (13.5,59.3)
           Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
           identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
           black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
Table 18b.     UFCR estimates (g/day raw weight, edible portion): Total freshwater + estuarine trophic level 4 fish, adults, 21 years and
                 older, by geographic area
Freshwater + Estuarine Trophic Level 4
Finfish and Shellfish
Adults (>21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
0.6 (0.4,0.9)

0.4 (0.3,0.6)
0.5 (0.3,1.1)
0.8 (0.5,1.1)
0.7 (0.5,0.9)

0.6 (0.4,0.9)
0.7 (0.5,1.0)

0.7 (0.5,1.1)
0.8 (0.5,1.1)
0.7 (0.5,1.2)
0.5 (0.2,0.9)
0.4 (0.3,0.6)
0.6 (0.3,1.2)
0.7 (0.5,1.1)
0.6 (0.5,0.9)
75th
1.9 (1.5,2.5)

1.3 (0.9,1.8)
1.8 (1.0,3.1)
2.4 (1.8,3.2)
1.9 (1.5,2.5)

1.8 (1.3,2.4)
2.1 (1.6,2.7)

2.2 (1.5,3.1)
2.2 (1.7,3.0)
2.1 (1.5,3.0)
1.5 (0.9,2.5)
1.2 (0.8,1.7)
1.8 (1.0,3.4)
2.3 (1.7,3.1)
1.7 (1.3,2.4)
Percentiles(95%CI)
90th
5.1 (4.0,6.4)

3.2 (2.3,4.4)
4.9 (2.9,8.1)
6.3 (4.9,8.2)
5.0 (3.6,6.9)

4.9 (3.6,6.5)
5.4 (4.3,7.0)

5.9 (3.9,8.7)
5.8 (4.3,7.8)
5.4 (3.8,7.7)
4.0 (2.5,6.5)
3.1 (2.2,4.4)
5.1 (3.0,8.8)
6.1 (4.6,8.1)
4.3 (3.0,6.3)
95th
9.1 (7.0,11.7)

5.5 (4.0,7.6)
8.9 (5.4,14.5)
11.2 (8.5,14.9)
8.9 (5.9,13.3)

8.7 (6.4,11.8)
9.7 (7.4,12.7)

10.5 (6.6,16.7)
10.2 (7.4,13.9)
9.7 (6.7,14.2)
7.3 (4.5,11.9)
5.3 (3.7,7.6)
9.3 (5.5,15.7)
10.9 (8.1,14.8)
7.4 (4.7,11.7)
97th
13.2 (10.0,17.4)

7.8 (5.5,10.9)
13.0 (8.0,21.2)
16.2 (12.1,21.8)
12.8 (8.0,20.5)

12.6 (9.2,17.5)
14.2 (10.5,19.1)

15.6 (9.1,26.5)
14.7 (10.5,20.7)
13.8 (9.1,20.9)
10.9 (6.6,18.3)
7.5 (5.1,10.9)
13.5 (8.1,22.5)
15.8 (11.5,21.9)
10.7 (6.4,17.8)
99th
27.1 (19.4,38.0)

15.1 (10.3,22.1)
26.6 (16.1,44.0)
32.7 (22.8,46.8)
27.1 (14.6,50.0)

25.8 (17.8,37.5)
29.0 (20.2,41.8)

33.8 (17.1,66.9)
28.8 (19.6,42.5)
28.1 (17.1,46.1)
22.9 (12.8,41.0)
14.6 (9.7,22.1)
27.8 (16.4,46.9)
32.5 (22.0,48.0)
20.9 (10.9,40.3)
1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, Ni, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
 TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
 border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
Table 19a.      UFCR estimates (g/day raw weight, edible portion): Total fish, youth, <21 years, by demographic characteristics
All Finfish and Shellfish
Youth (<21 yrs)
Age
1 to <3 yrs
3 to <6 yrs
6to$20K
Refused/Don't Know
Income
Income Missing
Percentiles(95%CI)
50th
4.9 (4.0,6.1)

2.7 (1.9,3.8)
3.6 (2.6,5.1)
5.1 (3.6,7.3)
5.0 (3.6,7.0)
6.1 (4.3,8.8)
9.1 (6.3,13.0)

4.5 (3.6,5.6)
5.5 (4.4,6.8)

4.4 (3.4,5.8)
4.2 (2.7,6.5)
4.2 (3.3,5.5)
7.8 (6.0,10.1)
10.3 (7.4,14.2)

5.0 (3.8,6.7)
4.9 (3.9,6.2)
5.0 (3.7,6.6)
5.1 (3.9,6.6)
3.8 (1.8,7.9)
4.2 (0.9,18.7)
6.4 (2.6,15.7)
75th
12.5 (10.6,14.7)

6.7 (5.0,9.0)
8.8 (6.9,11.2)
12.8 (9.2,17.7)
12.3 (9.5,15.9)
14.5 (11.5,18.3)
20.9 (15.5,28.2)

11.4 (9.7,13.5)
13.5 (11.2,16.3)

10.9 (9.0,13.2)
10.9 (7.5,15.9)
10.9 (8.7,13.7)
16.8 (13.7,20.7)
23.8 (18.6,30.3)

13.0 (10.4,16.3)
12.1 (10.2,14.4)
13.0 (9.9,16.9)
12.5 (10.0,15.7)
8.2 (4.5,15.0)
10.7 (2.4,48.1)
17.0 (7.4,39.0)
90th
24.0 (20.5,28.3)

12.7 (9.8,16.4)
16.3 (13.3,19.8)
24.2 (16.7,35.1)
22.6 (17.9,28.5)
26.9 (22.0,32.7)
38.5 (28.2,52.5)

21.9 (19.0,25.4)
26.1 (21.2,32.1)

20.3 (17.1,24.1)
20.8 (14.8,29.1)
21.4 (16.7,27.5)
28.9 (23.9,35.0)
40.9 (31.9,52.4)

24.5 (20.2,29.8)
23.2 (19.8,27.2)
26.0 (19.3,35.1)
23.5 (18.7,29.4)
15.4 (8.8,27.1)
27.5 (6.8,111.8)
39.2 (15.6,98.4)
95th
34.2 (28.6,40.8)

17.9 (13.8,23.2)
22.5 (18.4,27.7)
34.1 (22.6,51.4)
30.9 (24.5,39.0)
36.8 (30.4,44.7)
53.4 (37.9,75.4)

30.8 (26.7,35.5)
37.6 (29.7,47.6)

28.9 (24.2,34.6)
28.7 (20.5,40.1)
30.8 (23.3,40.6)
38.6 (32.1,46.3)
53.4 (40.3,70.8)

33.7 (27.8,41.0)
33.3 (27.9,39.8)
38.0 (27.6,52.5)
32.4 (25.4,41.4)
21.2 (12.0,37.3)
43.8 (15.4,124.3)
55.9 (23.4,133.5)
97th
42.4 (35.1,51.3)

22.3 (17.0,29.2)
27.5 (22.2,34.1)
41.9 (26.9,65.0)
37.4 (29.8,47.1)
44.3 (36.4,54.0)
65.0 (45.3,93.3)

37.7 (32.6,43.5)
47.1 (36.4,60.9)

36.2 (30.0,43.8)
35.0 (24.9,49.2)
38.9 (28.9,52.4)
45.8 (38.2,55.0)
62.7 (46.6,84.2)

41.0 (33.9,49.8)
41.7 (33.8,51.4)
46.9 (33.4,65.9)
39.7 (30.4,51.7)
25.9 (14.4,46.7)
54.8 (25.4,118.3)
67.9 (30.1,152.8)

99th
61.7 (49.1,77.5)

32.8 (24.1,44.4)
39.6 (31.4,50.1)
58.6 (36.6,93.8)
53.2 (42.7,66.2)
61.0 (49.7,74.8)
88.1 (60.6,128.0)

52.9 (44.4,63.1)
69.2 (51.8,92.4)

54.3 (43.4,67.8)
49.4 (34.2,71.3)
58.9 (41.0,84.6)
62.3 (51.7,74.9)
83.4 (62.2,111.8)

57.3 (46.8,70.2)
61.9 (44.3,86.5)
69.1 (47.6,100.5)
56.5 (41.5,76.9)
38.0 (20.8,69.6)
76.2 (44.7,130.1)
87.2 (44.2,172.2)
  Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
  identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
  black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
Table 19b.     UFCR estimates (g/day raw weight, edible portion): Total fish, youth, <21 years,  by geographic area
Percentiles(95%CI)
All Finfish and Shellfish
Youth (<21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
4.9 (4.0,6.1)

5.7 (4.1,7.8)
3.3 (2.5,4.3)
5.7 (4.2,7.7)
5.9 (4.1,8.7)

4.5 (3.5,5.7)
5.9 (4.7,7.4)

5.9 (4.3,8.1)
7.2 (5.4,9.6)
7.0 (4.3,11.5)
3.9 (2.9,5.2)
5.1 (3.6,7.2)
3.1 (2.3,4.1)
4.9 (3.7,6.4)
6.0 (3.5,10.1)
75th
12.5 (10.6,14.7)

14.1 (11.2,17.7)
8.8 (7.4,10.5)
13.5 (10.7,17.0)
14.6 (10.3,20.6)

11.5 (9.3,14.2)
14.5 (12.2,17.1)

15.1 (11.3,20.0)
16.4 (13.3,20.2)
16.0 (11.0,23.3)
10.2 (8.2,12.8)
12.9 (9.9,16.8)
8.3 (6.8,10.1)
12.1 (9.6,15.2)
14.2 (8.9,22.8)
90th
24.0 (20.5,28.3)

27.6 (21.1,36.1)
17.5 (15.1,20.3)
24.6 (19.9,30.3)
27.8 (19.5,39.7)

22.2 (17.6,28.1)
27.2 (23.4,31.8)

28.8 (22.2,37.5)
29.3 (24.4,35.3)
29.8 (20.6,43.3)
20.1 (15.9,25.5)
25.9 (17.8,37.6)
16.4 (14.0,19.3)
22.0 (17.7,27.3)
27.1 (16.1,45.5)
95th
34.2 (28.6,40.8)

40.4 (27.4,59.6)
25.5 (21.9,29.6)
33.9 (27.4,41.9)
39.2 (27.3,56.3)

31.7 (24.3,41.4)
38.0 (32.4,44.6)

40.2 (31.1,52.0)
40.3 (33.6,48.3)
41.5 (27.4,62.9)
28.4 (21.7,37.1)
39.1 (21.7,70.4)
24.0 (20.4,28.2)
30.3 (24.3,37.6)
38.4 (22.2,66.5)
97th
42.4 (35.1,51.3)

50.6 (31.3,82.0)
31.7 (27.1,37.1)
41.1 (33.0,51.3)
47.6 (33.1,68.4)

39.8 (29.8,53.2)
46.4 (39.1,55.1)

48.9 (37.7,63.4)
48.6 (40.4,58.5)
51.6 (33.1,80.5)
35.5 (26.9,46.7)
50.5 (24.6,103.6)
29.9 (25.4,35.3)
36.6 (29.4,45.5)
46.6 (26.9,80.8)
99th
61.7 (49.1,77.5)

75.1 (43.6,129.5)
47.7 (40.2,56.6)
58.5 (45.6,75.1)
67.0 (46.4,96.8)

58.8 (41.5,83.2)
65.7 (54.1,79.7)

68.6 (52.1,90.1)
68.2 (55.8,83.4)
72.6 (45.2,116.6)
50.4 (36.9,68.9)
76.6 (35.3,166.1)
45.6 (38.1,54.6)
51.3 (41.2,63.8)
65.8 (38.5,112.5)
1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, Ni, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
 TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
 border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
Table 20a.     UFCR estimates (g/day raw weight, edible portion): Freshwater + estuarine fish, youth, <21 years, by demographic
                 characteristics
Freshwater + Estuarine
Finfish and Shellfish
Youth (<21 yrs)
Age
1 to <3 yrs
3 to <6 yrs
6 to <11 yrs
11 to <16 yrs
16 to <18 yrs
18to<21yrs
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
1.1 (0.8,1.4)

0.6 (0.3,1.2)
0.7 (0.4,1.3)
1.1 (0.8,1.5)
1.1 (0.7,1.7)
1.4 (0.7,2.8)
1.7 (1.1,2.7)

0.9 (0.7,1.3)
1.2 (0.9,1.6)

1.3 (1.0,1.9)
1.1 (0.6,2.2)
0.7 (0.5,1.0)
2.6 (1.8,3.9)
2.7 (1.7,4.2)

1.1 (0.7,1.7)
1.2 (0.8,1.7)
0.9 (0.6,1.4)
1.2 (0.9,1.6)
0.6 (0.2,2.4)
0.3 (0.1,0.9)
1.0 (0.3,3.3)
75th
3.3 (2.7,4.2)

1.9 (1.2,3.2)
2.4 (1.6,3.6)
3.3 (2.5,4.4)
3.4 (2.4,4.9)
4.2 (2.5,6.9)
5.0 (3.5,7.2)

3.0 (2.3,4.0)
3.6 (2.9,4.5)

3.9 (3.0,5.1)
3.6 (1.9,6.9)
2.2 (1.7,3.0)
7.2 (5.3,9.9)
7.4 (4.9,11.1)

3.6 (2.6,5.0)
3.4 (2.5,4.7)
2.9 (2.0,4.1)
3.6 (3.0,4.4)
2.2 (0.9,5.5)
1.3 (0.6,3.0)
4.3 (1.0,18.6)
90th
8.0 (6.6,9.8)

4.7 (3.1,7.0)
5.8 (4.1,8.3)
7.7 (5.7,10.5)
8.3 (6.4,10.7)
9.5 (6.5,13.8)
11.6 (8.3,16.2)

7.4 (5.8,9.5)
8.6 (7.0,10.5)

8.7 (6.9,11.0)
8.7 (4.5,16.9)
5.1 (3.9,6.7)
15.1 (11.2,20.1)
15.3 (9.9,23.6)

8.7 (6.6,11.6)
8.2 (6.2,10.7)
7.0 (5.0,9.7)
8.3 (6.8,10.1)
6.1 (3.1,12.0)
4.4 (1.8,11.0)
15.0 (2.7,84.4)
95th
12.9 (10.5,15.9)

7.5 (5.1,10.9)
9.5 (6.7,13.5)
12.3 (8.8,17.3)
13.2 (10.4,16.7)
14.9 (10.6,21.0)
18.2 (12.8,25.8)

12.1 (9.5,15.4)
13.7 (11.1,16.9)

13.6 (10.8,17.2)
13.9 (6.8,28.4)
8.1 (6.1,10.8)
22.4 (16.7,30.0)
22.5 (14.1,35.6)

14.0 (10.7,18.4)
13.0 (9.9,17.1)
11.4 (8.3,15.6)
13.1 (10.5,16.3)
10.2 (5.3,19.6)
8.8 (3.2,24.6)
25.6 (5.2,126.3)
97th
17.3 (13.9,21.6)

10.1 (6.9,14.7)
12.9 (8.9,18.7)
16.3 (11.3,23.5)
17.7 (14.0,22.4)
19.3 (13.7,27.2)
23.8 (16.4,34.4)

16.2 (12.5,20.9)
18.4 (14.7,23.0)

18.0 (14.1,23.0)
18.6 (8.8,39.6)
10.7 (7.9,14.5)
28.5 (21.2,38.2)
28.2 (17.3,45.9)

18.7 (14.2,24.6)
17.2 (13.0,22.9)
15.4 (11.3,21.1)
17.3 (13.6,22.0)
14.4 (7.6,27.4)
14.6 (4.8,45.0)
35.6 (8.2,155.0)
99th
28.9 (22.2,37.7)

17.1 (11.3,25.9)
22.3 (13.8,36.1)
27.3 (18.3,40.7)
29.6 (22.9,38.2)
32.2 (22.6,45.9)
37.5 (24.3,57.8)

27.2 (20.4,36.3)
30.5 (23.1,40.2)

29.9 (22.2,40.2)
30.3 (13.4,68.6)
18.0 (12.6,25.7)
44.5 (32.8,60.4)
43.0 (25.7,71.8)

30.2 (22.4,40.9)
28.4 (21.0,38.5)
26.4 (19.1,36.4)
29.0 (21.5,39.1)
24.7 (12.5,48.8)
32.6 (8.9,119.7)
55.4 (15.5,198.1)
 Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
 identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
 black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
Table 20b.     UFCR estimates (g/day raw weight, edible portion): Freshwater + estuarine fish, youth, <21 years, by geographic area
Freshwater + Estuarine
Finfish and Shellfish
Youth (<21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
Percentiles(95%CI)
50th
1.1 (0.8,1.4)

0.9 (0.5,1.4)
0.7 (0.5,1.0)
1.6 (1.0,2.6)
1.2 (0.8,1.7)

0.9 (0.6,1.2)
1.6 (1.1,2.3)

1.5 (0.9,2.3)
1.9 (1.3,2.9)
2.3 (1.1,4.9)
1.0 (0.6,1.7)
0.7 (0.4,1.2)
0.6 (0.4,0.9)
1.2 (0.8,1.9)
1.0 (0.7,1.5)
75th
3.3 (2.7,4.2)

2.6 (1.7,4.0)
2.3 (1.6,3.2)
4.6 (3.2,6.8)
3.5 (2.5,5.0)

2.7 (2.1,3.4)
4.8 (3.5,6.5)

4.5 (3.0,6.8)
5.4 (3.8,7.7)
6.3 (3.3,11.8)
3.3 (2.1,5.3)
2.2 (1.3,3.6)
1.9 (1.3,2.7)
3.6 (2.6,5.2)
2.9 (2.0,4.2)
90th
8.0 (6.6,9.8)

6.0 (3.8,9.5)
5.8 (4.2,8.1)
10.4 (7.5,14.4)
8.3 (5.9,11.8)

6.3 (5.1,7.8)
11.1 (8.4,14.8)

10.6 (7.2,15.6)
12.0 (8.6,16.7)
13.8 (7.8,24.4)
8.6 (5.3,14.0)
5.1 (3.0,8.5)
4.7 (3.4,6.5)
8.2 (6.2,11.0)
6.7 (4.8,9.6)
95th
12.9 (10.5,15.9)

9.5 (5.9,15.4)
10.0 (7.0,14.4)
16.2 (11.8,22.2)
13.3 (9.3,19.1)

10.1 (8.1,12.6)
17.5 (13.1,23.4)

16.7 (11.2,25.0)
18.4 (13.2,25.4)
21.3 (11.9,38.0)
14.3 (8.5,24.0)
7.8 (4.5,13.6)
8.0 (5.7,11.1)
12.8 (9.7,16.9)
10.6 (7.4,15.2)
97th
17.3 (13.9,21.6)

12.6 (7.7,20.5)
14.0 (9.6,20.5)
21.3 (15.4,29.5)
17.7 (12.1,25.8)

13.7 (10.9,17.1)
22.9 (16.9,31.0)

22.0 (14.5,33.3)
23.8 (17.1,33.1)
27.9 (15.3,50.9)
19.4 (11.3,33.3)
10.1 (5.7,17.9)
11.2 (7.8,15.9)
16.8 (12.7,22.2)
14.1 (9.7,20.6)
99th
28.9 (22.2,37.7)

20.7 (12.4,34.6)
25.2 (16.6,38.4)
34.7 (24.7,48.7)
28.6 (18.5,44.1)

22.9 (17.7,29.8)
37.0 (26.3,52.0)

34.6 (21.8,54.8)
37.2 (25.9,53.3)
46.4 (25.4,84.7)
32.9 (18.7,58.1)
16.2 (8.8,29.8)
19.9 (13.1,30.3)
27.5 (20.6,36.8)
23.0 (15.1,35.2)
1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, Ni, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
 TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
 border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
Table 21a.      UFCR estimates (g/day raw weight, edible portion): Marine fish, youth, <21 years, by demographic characteristics

Marine Finfish and Shellfish
Youth (<21 yrs)
Age
1 to <3 yrs
3 to <6 yrs
6to$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
3.1 (2.4,4.1)

1.7 (1.2,2.5)
2.3 (1.6,3.4)
3.2 (2.1,5.0)
3.1 (2.1,4.5)
3.9 (2.8,5.5)
5.6 (3.7,8.5)

2.7 (2.1,3.6)
3.5 (2.6,4.7)

2.4 (1.7,3.3)
2.3 (1.4,3.6)
3.0 (2.2,4.2)
3.6 (2.7,4.8)
6.3 (4.4,9.0)

2.9 (2.2,4.0)
2.9 (2.3,3.7)
3.2 (2.3,4.5)
3.4 (2.4,4.8)
2.5 (1.1,6.0)
2.5 (0.6,11.4)
3.9 (1.5,9.9)
75th
8.1 (6.5,10.0)

4.4 (3.2,6.0)
5.7 (4.3,7.6)
8.3 (5.5,12.5)
7.7 (5.6,10.6)
9.6 (7.5,12.3)
14.3 (9.8,20.9)

7.1 (5.8,8.8)
9.1 (7.1,11.7)

6.1 (4.7,8.0)
6.2 (4.2,9.2)
8.0 (5.9,10.8)
8.4 (6.8,10.4)
15.8 (11.6,21.4)

7.8 (6.2,9.9)
7.5 (6.1,9.1)
8.5 (6.1,11.8)
8.8 (6.6,11.6)
5.5 (2.8,10.7)
7.1 (1.5,34.2)
11.0 (4.6,26.5)
90th
16.4 (13.3,20.3)

8.8 (6.6,11.9)
11.0 (8.6,14.2)
16.7 (10.3,27.0)
14.7 (10.9,19.8)
18.3 (14.6,23.0)
29.4 (19.3,44.8)

14.2 (11.8,17.1)
18.7 (14.4,24.4)

12.2 (9.8,15.2)
12.6 (9.1,17.5)
16.4 (12.1,22.2)
15.4 (12.8,18.6)
29.8 (22.8,38.9)

15.7 (12.5,19.8)
15.2 (12.5,18.4)
18.1 (12.4,26.4)
17.0 (13.0,22.2)
10.5 (5.4,20.3)
20.0 (4.3,93.7)
25.9 (11.1,60.7)
95th
24.3 (19.3,30.5)

12.9 (9.6,17.4)
15.8 (12.4,20.3)
24.2 (14.0,41.9)
20.6 (15.2,27.9)
25.7 (20.4,32.4)
43.5 (26.7,70.8)

20.5 (17.0,24.7)
28.1 (21.0,37.7)

18.0 (14.6,22.3)
18.1 (13.2,24.7)
24.5 (17.7,33.8)
21.3 (17.7,25.5)
40.7 (30.6,54.3)

22.7 (17.8,29.0)
22.6 (18.1,28.3)
27.4 (18.1,41.7)
24.2 (18.3,32.2)
15.1 (7.7,29.3)
35.3 (9.6,129.5)
39.4 (17.4,89.3)
97th
31.0 (24.2,39.6)

16.5 (12.1,22.6)
19.7 (15.3,25.4)
30.2 (16.9,54.1)
25.3 (18.6,34.5)
31.7 (25.0,40.2)
54.1 (31.7,92.4)

25.6 (21.1,31.1)
36.2 (26.3,49.7)

22.9 (18.4,28.5)
22.2 (16.0,30.8)
31.4 (22.3,44.3)
25.9 (21.6,31.1)
48.6 (35.1,67.4)

28.5 (22.2,36.6)
29.3 (22.2,38.6)
35.5 (22.7,55.6)
30.1 (22.4,40.3)
18.5 (9.3,36.8)
47.1 (16.5,134.6)
49.2 (22.8,106.3)
99th
48.2 (35.6,65.2)

25.2 (17.7,35.7)
29.3 (22.4,38.4)
44.5 (24.1,82.4)
36.8 (26.8,50.5)
45.6 (35.5,58.4)
80.3 (45.3,142.2)

37.9 (30.9,46.5)
56.7 (37.7,85.4)

36.1 (28.3,45.9)
31.7 (21.3,47.3)
49.9 (32.1,77.6)
37.0 (30.6,44.8)
67.8 (45.2,101.9)

43.1 (33.1,56.1)
47.2 (28.0,79.6)
56.5 (33.1,96.4)
44.3 (32.3,60.8)
26.5 (12.9,54.5)
70.9 (36.2,138.6)
69.5 (35.8,134.6)
 Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
 identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
 black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
          Table 21b.      UFCR estimates (g/day raw weight, edible portion): Marine fish, youth, <21 years, by geographic area
CS
Percentiles(95%CI)
Marine Finfish and Shellfish
Youth (<21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
3.1 (2.4,4.1)

4.0 (2.6,6.0)
2.2 (1.6,3.0)
3.3 (2.4,4.4)
3.7 (2.3,6.1)

2.9 (2.1,4.1)
3.5 (2.7,4.4)

3.5 (2.4,5.0)
4.3 (3.2,5.9)
3.8 (2.6,5.7)
2.2 (1.6,3.1)
3.7 (2.4,5.8)
2.1 (1.5,3.1)
2.9 (2.1,3.9)
3.9 (2.0,7.6)
75th
8.1 (6.5,10.0)

10.3 (7.3,14.6)
5.8 (4.5,7.5)
8.0 (6.2,10.3)
9.6 (6.0,15.3)

7.7 (5.7,10.3)
8.8 (7.4,10.5)

9.3 (6.9,12.6)
10.2 (7.9,13.1)
9.3 (6.6,13.2)
6.1 (4.9,7.7)
9.8 (6.7,14.4)
5.8 (4.3,7.8)
7.3 (5.5,9.7)
9.8 (5.2,18.6)
90th
16.4 (13.3,20.3)

21.4 (14.2,32.0)
12.0 (9.8,14.6)
15.6 (12.2,19.9)
19.5 (12.0,31.8)

15.8 (11.5,21.6)
17.8 (15.1,20.9)

19.2 (14.5,25.4)
19.5 (15.3,24.9)
18.7 (12.8,27.3)
12.6 (10.1,15.6)
20.7 (12.0,35.8)
11.8 (9.2,15.1)
14.3 (10.7,19.1)
19.7 (9.9,39.4)
95th
24.3 (19.3,30.5)

32.7 (18.8,56.9)
17.6 (14.7,21.1)
22.3 (17.2,28.9)
28.6 (17.3,47.0)

23.2 (16.3,33.0)
26.0 (21.9,30.7)

28.6 (21.5,38.0)
27.9 (21.7,35.8)
28.0 (18.3,42.9)
18.5 (14.4,23.7)
32.8 (14.4,74.5)
17.3 (13.9,21.5)
20.3 (15.0,27.4)
28.5 (13.9,58.3)
97th
31.0 (24.2,39.6)

43.0 (21.9,84.5)
22.1 (18.5,26.4)
27.9 (21.2,36.6)
35.5 (21.9,57.8)

29.8 (20.4,43.7)
32.9 (27.6,39.2)

35.7 (27.1,47.2)
35.0 (27.0,45.4)
36.2 (22.8,57.5)
23.3 (17.7,30.7)
44.5 (16.9,117.4)
21.8 (17.8,26.8)
25.2 (18.5,34.3)
35.4 (17.4,71.8)
99th
48.2 (35.6,65.2)

68.5 (31.1,150.9)
33.8 (28.2,40.6)
42.4 (31.1,57.8)
51.0 (32.2,80.8)

47.2 (29.4,75.8)
49.5 (40.0,61.4)

50.8 (37.4,69.0)
53.5 (39.2,73.2)
56.6 (33.7,95.2)
34.4 (24.8,47.7)
73.6 (25.7,211.4)
33.6 (27.7,40.8)
37.2 (27.2,50.9)
51.1 (26.1,100.0)
          1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, Ni, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

          2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
Table 22a.      UFCR estimates (g/day raw weight, edible portion): Total finfish, youth, <21 years, by demographic characteristics
Percentiles(95%CI)
All Finfish
Youth (<21 yrs)
Age
1 to <3 yrs
3 to <6 yrs
6 to <11 yrs
11 to <16 yrs
16 to <18 yrs
18to<21yrs
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income, finer detail
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
50th
3.8 (3.0,4.8)

2.1 (1.5,2.9)
2.7 (1.9,4.0)
4.2 (2.9,6.1)
3.8 (2.5,5.6)
4.5 (3.1,6.5)
6.3 (4.3,9.1)

3.4 (2.6,4.3)
4.2 (3.2,5.4)

3.3 (2.5,4.5)
3.2 (1.9,5.4)
3.4 (2.5,4.5)
5.6 (4.2,7.3)
6.8 (4.5,10.1)

4.1 (3.0,5.6)
3.6 (2.7,4.7)
4.1 (3.0,5.7)
3.7 (2.7,5.0)
2.6 (1.3,5.3)
2.9 (0.6,14.2)
4.9 (1.7,13.5)
75th
9.4 (7.9,11.3)

5.2 (4.1,6.8)
6.6 (5.1,8.7)
10.4 (7.3,14.7)
9.1 (6.6,12.4)
10.5 (8.0,14.0)
15.2 (10.9,21.2)

8.6 (7.1,10.3)
10.3 (8.5,12.6)

8.2 (6.6,10.2)
8.4 (5.7,12.4)
8.5 (6.6,10.9)
12.0 (9.5,15.1)
17.1 (12.5,23.3)

10.5 (8.2,13.4)
8.8 (7.0,11.1)
10.7 (7.8,14.8)
8.9 (7.1,11.3)
6.2 (3.2,12.1)
8.0 (1.4,46.8)
13.2 (6.4,27.2)
90th
18.5 (15.4,22.2)

10.1 (8.1,12.7)
12.8 (10.2,16.0)
20.1 (13.3,30.4)
16.6 (12.7,21.8)
19.7 (15.1,25.6)
30.1 (20.5,44.4)

16.8 (14.1,19.9)
20.3 (16.4,25.2)

15.6 (12.9,18.9)
16.8 (12.2,23.2)
17.0 (12.8,22.4)
21.0 (16.7,26.4)
31.4 (23.3,42.3)

20.0 (15.8,25.3)
17.1 (13.9,21.1)
21.9 (15.2,31.5)
16.8 (13.3,21.3)
12.7 (5.8,27.8)
23.1 (3.8,141.6)
26.4 (14.9,46.8)
95th
26.8 (21.8,32.9)

14.6 (11.6,18.4)
18.2 (14.5,22.8)
28.1 (17.6,45.0)
23.1 (17.7,30.2)
27.5 (20.9,36.1)
43.8 (27.7,69.2)

23.8 (19.9,28.6)
29.9 (23.1,38.7)

22.3 (18.3,27.1)
24.3 (17.7,33.5)
24.8 (18.1,34.0)
28.6 (22.7,36.0)
42.7 (30.4,60.0)

28.0 (22.0,35.8)
24.7 (19.7,30.9)
32.6 (21.8,48.7)
23.7 (18.2,31.0)
18.2 (7.7,43.1)
38.0 (8.8,163.8)
37.3 (22.9,60.7)
97th
33.8 (26.9,42.5)

18.4 (14.5,23.5)
22.6 (17.8,28.6)
34.6 (20.8,57.6)
28.7 (22.0,37.5)
34.2 (26.0,44.9)
54.9 (33.2,90.7)

29.6 (24.6,35.6)
38.1 (28.3,51.4)

28.2 (23.0,34.6)
30.1 (21.5,42.1)
31.8 (22.5,45.0)
34.4 (27.3,43.5)
51.6 (35.9,74.1)

34.4 (26.7,44.4)
31.4 (24.3,40.6)
41.8 (27.1,64.5)
29.5 (22.1,39.4)
22.5 (9.3,54.5)
50.5 (15.4,165.8)
45.2 (28.3,72.2)

99th
51.7 (39.1,68.4)

27.2 (20.4,36.2)
33.1 (25.4,43.2)
49.9 (28.4,87.7)
42.7 (32.2,56.5)
50.5 (38.7,65.9)
78.5 (46.7,131.9)

43.3 (35.9,52.2)
59.2 (40.6,86.5)

44.0 (34.8,55.7)
43.9 (30.4,63.3)
50.4 (32.5,78.1)
48.4 (38.2,61.2)
69.4 (45.7,105.4)

50.9 (39.0,66.4)
48.5 (30.2,77.9)
64.9 (40.2,104.7)
43.7 (30.7,62.3)
32.7 (13.1,81.4)
73.3 (32.4,165.7)
65.3 (40.5,105.4)
 Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
 identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
 black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
          Table 22b.      UFCR estimates (g/day raw weight, edible portion): Total finfish, youth, <21 years, by geographic area
oo
Percentiles(95%CI)
All Finfish
Youth (<21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
3.8 (3.0,4.8)

4.1 (2.7,6.3)
2.6 (1.9,3.8)
4.2 (3.2,5.5)
4.5 (2.9,7.1)

3.5 (2.7,4.7)
4.2 (3.3,5.4)

4.2 (2.9,6.0)
4.9 (3.6,6.9)
4.8 (3.2,7.0)
3.0 (2.0,4.3)
3.9 (2.4,6.1)
2.5 (1.8,3.7)
3.8 (2.9,4.9)
4.8 (2.6,8.7)
75th
9.4 (7.9,11.3)

10.0 (7.1,13.9)
7.0 (5.4,8.9)
10.0 (7.9,12.7)
11.5 (7.6,17.5)

9.0 (7.1,11.4)
10.3 (8.5,12.4)

11.1 (7.9,15.6)
11.1 (8.5,14.4)
11.3 (8.0,15.8)
7.7 (5.6,10.5)
9.5 (6.6,13.7)
6.7 (5.2,8.7)
9.3 (7.3,11.9)
11.8 (6.8,20.6)
90th
18.5 (15.4,22.2)

20.2 (13.2,31.1)
13.8 (11.3,16.8)
18.5 (14.7,23.4)
22.8 (14.7,35.5)

17.7 (13.6,23.0)
20.0 (16.7,23.9)

22.5 (16.1,31.4)
20.8 (16.4,26.4)
21.2 (15.2,29.8)
15.1 (11.0,20.8)
19.6 (11.0,34.8)
13.3 (10.9,16.1)
17.3 (13.5,22.3)
23.0 (12.5,42.6)
95th
26.8 (21.8,32.9)

31.4 (16.9,58.4)
20.1 (16.6,24.3)
25.7 (20.2,32.7)
32.6 (20.7,51.6)

25.7 (18.8,35.0)
28.7 (23.9,34.5)

32.6 (23.4,45.5)
29.4 (23.0,37.6)
29.8 (21.4,41.4)
21.8 (15.3,31.0)
31.3 (12.9,76.0)
19.2 (16.1,22.8)
23.9 (18.3,31.2)
32.6 (17.0,62.7)
97th
33.8 (26.9,42.5)

42.0 (19.8,88.7)
25.1 (20.6,30.6)
31.4 (24.6,40.2)
40.8 (25.6,65.1)

32.5 (22.9,46.2)
35.9 (29.8,43.2)

40.8 (29.2,57.1)
36.6 (28.2,47.4)
36.7 (26.4,50.9)
27.4 (18.9,39.8)
43.1 (15.3,121.4)
24.2 (20.5,28.5)
29.3 (22.2,38.6)
40.9 (20.8,80.4)
99th
51.7 (39.1,68.4)

70.1 (31.2,157.3)
38.2 (30.9,47.3)
45.1 (34.8,58.4)
58.6 (37.2,92.5)

50.9 (33.0,78.5)
52.8 (43.1,64.6)

57.9 (41.0,81.6)
55.5 (41.2,74.7)
50.3 (35.7,70.9)
41.5 (28.0,61.6)
73.0 (24.9,213.8)
36.3 (30.9,42.8)
41.9 (31.8,55.3)
59.1 (30.5,114.6)
          1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, Ni, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

          2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
Table 23a.      UFCR estimates (g/day raw weight, edible portion): Total shellfish, youth, <21 years, by demographic characteristics
Percentiles(95%CI)
All Shellfish
Youth (<21 yrs)
Age
1 to <3 yrs
3 to <6 yrs
6 to <11 yrs
11 to <16 yrs
16 to <18 yrs
18to<21yrs
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income, finer detail
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
50th
0.6 (0.4,0.9)

0.3 (0.2,0.6)
0.5 (0.3,1.0)
0.5 (0.3,0.9)
0.7 (0.4,1.2)
0.9 (0.5,1.7)
1.1 (0.5,2.4)

0.6 (0.4,0.9)
0.6 (0.4,0.9)

0.7 (0.4,1.0)
0.5 (0.2,1.5)
0.5 (0.3,0.7)
1.2 (0.8,1.8)
1.8 (1.0,3.3)

0.5 (0.3,0.8)
0.8 (0.5,1.1)
0.5 (0.3,0.8)
0.8 (0.5,1.2)
0.3 (0.1,1.6)
0.2 (0.0,1.5)
0.6 (0.1,2.2)
75th
1.9 (1.5,2.5)

0.9 (0.5,1.8)
1.6 (0.9,2.7)
1.5 (0.9,2.2)
2.2 (1.5,3.4)
2.8 (1.7,4.6)
3.5 (2.0,6.3)

1.9 (1.4,2.5)
2.0 (1.5,2.6)

2.0 (1.4,2.7)
1.8 (0.7,4.2)
1.4 (1.0,2.1)
3.4 (2.4,4.8)
5.0 (3.0,8.3)

1.6 (1.1,2.4)
2.2 (1.6,3.1)
1.5 (1.0,2.3)
2.3 (1.5,3.3)
0.9 (0.3,3.1)
1.3 (0.3,5.8)
3.4 (0.6,18.7)
90th
4.8 (3.9,6.1)

2.1 (1.1,4.0)
3.5 (2.2,5.6)
3.4 (2.4,4.8)
5.5 (3.7,8.2)
6.8 (4.5,10.2)
8.5 (5.3,13.6)

4.8 (3.7,6.1)
4.9 (3.8,6.4)

4.8 (3.5,6.4)
4.8 (2.2,10.5)
3.4 (2.4,5.0)
7.7 (5.6,10.7)
10.5 (6.3,17.6)

4.0 (2.8,5.6)
5.3 (4.0,7.0)
3.9 (2.7,5.7)
5.6 (4.0,7.9)
2.2 (0.8,6.1)
4.5 (1.2,16.7)
15.0 (1.8,126.0)
95th
8.1 (6.4,10.2)

3.2 (1.6,6.5)
5.5 (3.5,8.6)
5.5 (3.9,7.9)
9.0 (6.1,13.3)
10.9 (7.5,15.9)
13.6 (8.4,22.0)

8.0 (6.2,10.3)
8.1 (6.1,10.7)

7.9 (5.7,10.9)
8.3 (3.8,17.8)
5.7 (3.9,8.4)
12.0 (8.7,16.7)
15.7 (9.1,27.2)

6.6 (4.7,9.2)
8.5 (6.4,11.4)
6.7 (4.5,9.8)
9.2 (6.5,13.2)
3.7 (1.4,10.2)
8.3 (2.4,28.7)
27.9 (4.8,163.5)
97th
11.0 (8.4,14.4)

4.4 (2.1,9.0)
7.2 (4.6,11.4)
7.5 (5.0,11.1)
12.0 (8.1,17.8)
14.6 (10.1,21.2)
18.1 (10.9,30.0)

10.9 (8.3,14.5)
11.1 (8.1,15.1)

10.9 (7.6,15.6)
11.5 (5.4,24.5)
7.8 (5.2,11.7)
15.8 (11.3,22.1)
19.8 (10.8,36.4)

8.9 (6.3,12.6)
11.5 (8.6,15.5)
9.2 (6.1,13.8)
12.4 (8.5,18.0)
5.2 (1.9,14.1)
11.6 (3.3,40.4)
37.4 (9.6,146.3)
99th
19.2 (13.9,26.7)

7.3 (3.2,16.5)
12.0 (7.2,20.0)
13.0 (7.4,22.8)
19.8 (13.3,29.7)
23.2 (15.5,34.6)
28.7 (16.3,50.8)

19.2 (13.7,26.9)
19.3 (13.2,28.4)

19.8 (12.5,31.3)
19.4 (9.3,40.6)
14.0 (8.9,22.1)
25.0 (17.5,35.8)
29.9 (15.1,59.5)

15.1 (10.4,21.9)
19.4 (13.9,27.2)
15.9 (10.0,25.4)
21.1 (13.6,32.6)
9.2 (3.2,26.8)
22.9 (5.9,89.1)
42.2 (23.6,75.5)
 Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
 identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
 black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
          Table 23b.      UFCR estimates (g/day raw weight, edible portion): Total shellfish, youth, <21 years, by geographic area
oo
o
Percentiles(95%CI)
All Shellfish
Youth (<21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
0.6 (0.4,0.9)

0.7 (0.4,1.5)
0.4 (0.2,0.7)
0.8 (0.5,1.2)
0.7 (0.4,1.1)

0.5 (0.3,0.7)
0.9 (0.6,1.3)

0.9 (0.6,1.4)
1.2 (0.7,1.8)
1.1 (0.5,2.3)
0.6 (0.3,1.0)
0.6 (0.3,1.3)
0.3 (0.2,0.6)
0.6 (0.4,0.9)
0.6 (0.3,1.0)
75th
1.9 (1.5,2.5)

2.3 (1.3,4.2)
1.2 (0.7,2.0)
2.2 (1.5,3.3)
2.2 (1.5,3.2)

1.5 (1.1,2.1)
2.8 (2.1,3.8)

2.9 (2.0,4.3)
3.5 (2.4,5.0)
3.3 (1.6,6.6)
1.8 (1.1,2.8)
2.0 (1.0,3.8)
1.0 (0.6,1.8)
1.7 (1.2,2.4)
1.8 (1.1,2.9)
90th
4.8 (3.9,6.1)

5.6 (3.1,10.1)
3.1 (1.9,5.0)
5.4 (3.7,7.7)
5.6 (3.9,8.0)

3.7 (2.8,4.9)
7.1 (5.4,9.4)

7.5 (5.0,11.3)
8.2 (5.8,11.7)
8.1 (4.1,16.1)
4.5 (2.8,7.2)
4.6 (2.5,8.3)
2.5 (1.5,4.4)
3.9 (2.9,5.2)
4.3 (2.6,7.0)
95th
8.1 (6.4,10.2)

9.1 (5.1,16.3)
5.3 (3.3,8.5)
8.7 (6.1,12.6)
9.3 (6.2,13.9)

6.1 (4.6,8.1)
11.7 (8.6,15.9)

12.4 (7.4,20.7)
12.8 (9.0,18.3)
13.9 (6.7,29.1)
7.5 (4.4,12.8)
7.2 (4.1,12.6)
4.3 (2.4,7.5)
6.2 (4.7,8.2)
7.0 (4.2,11.5)
97th
11.0 (8.4,14.4)

12.3 (7.0,21.7)
7.4 (4.5,12.0)
11.8 (8.1,17.3)
12.6 (7.9,20.0)

8.1 (6.0,11.0)
15.8 (11.3,22.0)

16.9 (9.2,30.9)
16.7 (11.8,23.8)
20.2 (8.9,45.9)
10.2 (5.7,18.4)
9.4 (5.4,16.4)
6.0 (3.4,10.7)
8.3 (6.2,11.0)
9.4 (5.5,15.8)
99th
19.2 (13.9,26.7)

20.6 (11.9,35.6)
13.5 (8.0,22.9)
20.6 (13.3,32.0)
21.2 (11.4,39.3)

13.9 (9.8,19.8)
26.4 (17.4,39.9)

27.8 (13.6,56.7)
26.1 (18.0,37.9)
37.4 (19.2,73.1)
17.7 (8.9,35.0)
15.6 (9.1,26.7)
10.9 (5.8,20.3)
13.7 (9.9,19.0)
15.5 (8.6,28.0)
          1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

          2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
          Table 24a.      UFCR estimates (g/day raw weight, edible portion): Total trophic level 2 fish, youth, <21 years, by demographic
                           characteristics
oo
Trophic Level 2
Finfish and Shellfish
Youth (<21 yrs)
Age
1 to <3 yrs
3 to <6 yrs
6 to <11 yrs
11 to <16 yrs
16 to <18 yrs
18to<21yrs
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
0.4 (0.3,0.6)

0.2 (0.1,0.4)
0.3 (0.2,0.6)
0.3 (0.1,0.5)
0.5 (0.3,1.0)
0.5 (0.2,1.0)
0.6 (0.3,1.1)

0.4 (0.2,0.6)
0.4 (0.3,0.6)

0.6 (0.4,0.9)
0.4 (0.1,0.9)
0.3 (0.2,0.4)
0.6 (0.4,0.9)
1.1 (0.5,2.4)

0.3 (0.2,0.6)
0.5 (0.3,0.7)
0.3 (0.2,0.5)
0.4 (0.3,0.7)
0.2 (0.1,0.8)
0.1 (0.0,0.8)
0.3 (0.1,1.4)
75th
1.2 (0.9,1.6)

0.6 (0.4,1.0)
0.9 (0.6,1.5)
0.8 (0.5,1.3)
1.7 (1.0,2.7)
1.5 (0.9,2.5)
1.8 (1.1,3.1)

1.2 (0.9,1.6)
1.2 (0.9,1.7)

1.7 (1.2,2.4)
1.2 (0.5,2.6)
0.8 (0.5,1.3)
1.8 (1.3,2.3)
3.2 (1.7,6.3)

1.1 (0.7,1.6)
1.5 (1.1,2.0)
0.9 (0.6,1.5)
1.3 (0.9,2.1)
0.6 (0.2,1.9)
0.6 (0.1,3.3)
1.6 (0.3,7.8)
90th
3.1 (2.4,4.0)

1.5 (1.0,2.2)
2.3 (1.5,3.5)
2.2 (1.5,3.3)
4.0 (2.5,6.3)
3.6 (2.3,5.5)
4.4 (2.7,7.1)

3.0 (2.3,4.0)
3.2 (2.3,4.3)

4.1 (3.0,5.4)
2.8 (1.4,5.8)
2.2 (1.4,3.3)
4.1 (3.0,5.4)
6.9 (3.6,12.9)

2.7 (1.9,3.8)
3.5 (2.7,4.7)
2.5 (1.6,3.8)
3.4 (2.2,5.2)
1.5 (0.6,4.1)
2.6 (0.7,10.2)
7.2 (1.1,45.2)
95th
5.2 (4.0,6.7)

2.4 (1.7,3.5)
3.8 (2.6,5.6)
3.9 (2.6,6.1)
6.4 (4.1,10.1)
5.6 (3.7,8.4)
7.1 (4.3,11.6)

5.1 (3.9,6.7)
5.2 (3.8,7.2)

6.4 (4.8,8.6)
4.6 (2.2,9.3)
3.7 (2.4,5.7)
6.5 (4.8,8.8)
10.1 (5.3,19.2)

4.4 (3.2,6.2)
5.7 (4.3,7.6)
4.2 (2.7,6.5)
5.7 (3.7,8.7)
2.6 (1.0,6.7)
4.5 (1.3,15.5)
12.3 (3.1,47.7)
97th
7.1 (5.3,9.3)

3.3 (2.2,4.9)
5.2 (3.6,7.6)
5.6 (3.5,9.0)
8.6 (5.4,13.5)
7.3 (4.8,11.0)
9.3 (5.5,15.7)

7.0 (5.3,9.2)
7.1 (5.1,9.9)

8.4 (6.2,11.5)
6.0 (2.9,12.3)
5.1 (3.3,8.1)
8.6 (6.2,11.9)
12.6 (6.7,23.9)

6.0 (4.3,8.3)
7.7 (5.7,10.4)
5.8 (3.7,9.2)
7.7 (5.0,11.9)
3.6 (1.4,9.2)
6.3 (1.8,21.8)
15.4 (5.3,44.6)
99th
12.0 (8.6,16.6)

5.8 (3.7,9.2)
8.8 (6.0,12.9)
10.5 (6.1,18.1)
14.0 (8.8,22.3)
11.6 (7.4,18.4)
14.9 (8.7,25.7)

11.8 (8.5,16.4)
12.1 (8.3,17.7)

13.8 (9.5,20.0)
10.1 (4.9,20.6)
9.2 (5.7,14.8)
14.0 (9.8,20.0)
18.5 (9.7,35.0)

9.8 (6.9,14.0)
13.0 (9.4,17.9)
10.0 (6.0,16.6)
12.9 (8.2,20.3)
6.3 (2.4,16.6)
11.7 (3.2,42.1)
20.0 (10.2,39.3)
           Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
           identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
           black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
          Table 24b.      UFCR estimates (g/day raw weight, edible portion): Total trophic level 2 fish, youth, <21 years, by geographic area
oo
Trophic Level 2
Finfish and Shellfish
Youth (<21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
Percentiles(95%CI)
50th
0.4 (0.3,0.6)

0.4 (0.2,0.8)
0.2 (0.1,0.4)
0.4 (0.3,0.7)
0.5 (0.3,0.8)

0.3 (0.2,0.5)
0.6 (0.4,0.8)

0.6 (0.4,1.1)
0.6 (0.4,0.9)
0.7 (0.4,1.2)
0.3 (0.2,0.6)
0.4 (0.2,0.7)
0.2 (0.1,0.4)
0.4 (0.2,0.6)
0.4 (0.2,0.7)
75th
1.2 (0.9,1.6)

1.3 (0.8,2.3)
0.8 (0.5,1.4)
1.3 (0.9,2.0)
1.5 (1.0,2.4)

1.0 (0.7,1.4)
1.8 (1.3,2.5)

2.1 (1.3,3.3)
1.9 (1.4,2.8)
2.1 (1.2,3.8)
1.1 (0.7,2.0)
1.1 (0.6,2.0)
0.7 (0.4,1.1)
1.0 (0.7,1.6)
1.2 (0.7,2.0)
90th
3.1 (2.4,4.0)

3.3 (1.8,6.3)
2.1 (1.2,3.6)
3.2 (2.3,4.6)
3.9 (2.5,6.1)

2.4 (1.8,3.2)
4.5 (3.4,6.1)

5.0 (3.0,8.5)
4.8 (3.2,7.1)
5.3 (3.0,9.3)
3.1 (1.9,5.1)
2.7 (1.5,4.8)
1.7 (1.0,2.9)
2.4 (1.7,3.4)
3.0 (1.8,5.0)
95th
5.2 (4.0,6.7)

5.5 (2.8,10.8)
3.7 (2.2,6.1)
5.3 (3.7,7.6)
6.3 (3.9,10.3)

3.9 (2.9,5.3)
7.3 (5.4,9.9)

7.9 (4.5,13.9)
7.6 (4.9,11.7)
8.7 (5.2,14.6)
5.1 (3.2,8.2)
4.1 (2.3,7.6)
3.0 (1.7,5.0)
3.8 (2.7,5.4)
5.0 (3.0,8.3)
97th
7.1 (5.3,9.3)

7.5 (3.7,15.0)
5.1 (3.1,8.3)
7.2 (5.1,10.3)
8.4 (5.0,14.0)

5.3 (3.9,7.2)
9.7 (7.0,13.4)

10.2 (5.7,18.5)
10.1 (6.4,15.8)
11.5 (7.2,18.4)
6.9 (4.3,11.0)
5.6 (3.1,10.1)
4.2 (2.4,7.1)
5.1 (3.6,7.2)
6.7 (4.0,11.3)
99th
12.0 (8.6,16.6)

12.9 (6.5,25.7)
8.9 (5.5,14.5)
12.2 (8.6,17.3)
13.4 (7.6,23.7)

9.1 (6.4,12.9)
15.5 (10.9,22.0)

16.1 (8.9,29.2)
16.2 (10.0,26.3)
17.3 (12.0,25.1)
11.6 (7.2,18.6)
9.3 (5.1,17.1)
7.4 (4.1,13.3)
8.5 (6.0,12.1)
11.1 (6.3,19.5)
          1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

          2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
          Table 25a.      UFCR estimates (g/day raw weight, edible portion): Total trophic level 3 fish, adults, youth, <21 years, by demographic
                           characteristics
oo
Trophic Level 3
Finfish and Shellfish
Youth (<21 yrs)
Age
1 to <3 yrs
3 to <6 yrs
6 to <11 yrs
11 to <16 yrs
16 to <18 yrs
18to<21yrs
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
1.4 (1.0,1.8)

0.8 (0.5,1.3)
1.1 (0.8,1.7)
1.6 (1.1,2.4)
1.3 (0.9,2.0)
1.2 (0.7,2.1)
2.2 (1.4,3.3)

1.2 (0.9,1.6)
1.5 (1.1,2.1)

1.1 (0.8,1.5)
0.9 (0.6,1.5)
1.1 (0.8,1.6)
3.1 (2.3,4.2)
3.6 (2.4,5.6)

1.6 (1.2,2.1)
1.4 (1.0,1.8)
1.3 (0.9,2.0)
1.3 (0.9,1.9)
1.1 (0.4,2.6)
1.8 (0.4,9.3)
1.3 (0.4,4.2)
75th
3.5 (2.8,4.5)

2.1 (1.4,3.2)
2.9 (2.1,4.0)
4.2 (2.7,6.3)
3.3 (2.3,4.7)
3.1 (2.1,4.8)
5.3 (3.8,7.5)

3.2 (2.5,4.0)
3.9 (3.0,5.1)

2.7 (2.1,3.5)
2.5 (1.6,4.0)
2.8 (2.1,3.8)
7.0 (5.4,9.2)
8.3 (5.7,12.0)

4.3 (3.2,5.7)
3.5 (2.8,4.3)
3.4 (2.3,5.1)
3.3 (2.4,4.6)
2.4 (1.2,4.9)
4.3 (1.0,17.9)
5.3 (1.3,21.8)
90th
7.4 (6.0,9.2)

4.2 (2.7,6.3)
5.9 (4.5,7.6)
8.8 (5.7,13.7)
6.8 (5.1,9.1)
7.2 (5.3,9.8)
10.0 (7.4,13.5)

6.8 (5.5,8.5)
8.0 (6.2,10.3)

5.5 (4.4,6.9)
5.2 (3.2,8.5)
5.6 (4.1,7.6)
12.6 (9.9,16.2)
15.2 (10.1,22.9)

8.9 (6.9,11.5)
7.1 (5.7,8.8)
7.3 (4.7,11.3)
6.9 (5.2,9.2)
4.7 (2.7,8.2)
8.4 (2.9,24.8)
16.1 (3.0,86.1)
95th
11.1 (8.9,13.8)

6.0 (4.0,9.0)
8.5 (6.6,10.9)
13.1 (8.2,20.9)
10.0 (7.5,13.2)
11.6 (8.5,15.6)
14.0 (10.2,19.1)

10.2 (8.2,12.8)
11.8 (9.1,15.3)

8.2 (6.4,10.4)
7.6 (4.4,13.2)
8.1 (5.9,11.2)
17.1 (13.4,21.9)
20.3 (12.9,32.0)

12.8 (9.9,16.7)
10.5 (8.4,13.1)
10.7 (6.8,16.9)
10.3 (7.7,14.0)
6.7 (3.9,11.4)
12.4 (5.1,29.9)
29.6 (6.3,139.9)
97th
14.1 (11.2,17.7)

7.5 (5.0,11.4)
10.7 (8.3,13.8)
16.6 (10.2,27.1)
12.7 (9.6,16.8)
15.1 (11.0,20.7)
17.0 (12.1,23.8)

13.1 (10.4,16.5)
14.9 (11.4,19.5)

10.4 (8.0,13.5)
9.7 (5.5,17.0)
10.3 (7.3,14.4)
20.7 (16.2,26.4)
24.1 (14.9,38.9)

16.1 (12.3,21.0)
13.4 (10.7,16.8)
13.5 (8.5,21.4)
13.3 (9.6,18.4)
8.4 (4.9,14.4)
15.4 (6.9,34.4)
38.2 (10.2,143.0)
99th
21.5 (16.5,27.9)

11.4 (7.4,17.8)
15.7 (11.8,20.8)
25.4 (15.0,42.9)
19.4 (14.8,25.5)
23.6 (16.6,33.5)
23.9 (16.1,35.7)

20.3 (15.6,26.5)
22.4 (16.6,30.1)

16.0 (11.5,22.2)
14.7 (8.0,26.9)
15.8 (10.7,23.1)
28.6 (22.2,36.9)
32.1 (19.5,52.9)

23.6 (17.8,31.3)
20.3 (16.1,25.5)
19.8 (12.4,31.6)
20.5 (13.7,30.5)
12.4 (6.9,22.3)
24.4 (10.7,56.0)
51.8 (19.7,136.7)
           Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
           identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
           black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
          Table 25b.      UFCR estimates (g/day raw weight, edible portion): Total trophic level 3 fish, youth, <21 years, by geographic area
oo
Trophic Level 3
Finfish and Shellfish
Youth (<21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
Percentiles(95%CI)
50th
1.4 (1.0,1.8)

1.4 (0.8,2.3)
0.9 (0.6,1.3)
1.9 (1.3,2.7)
1.4 (0.9,2.0)

1.2 (0.9,1.6)
1.8 (1.3,2.4)

1.5 (1.0,2.2)
2.3 (1.5,3.4)
2.3 (1.2,4.3)
1.2 (0.9,1.7)
1.2 (0.8,1.9)
0.8 (0.6,1.2)
1.6 (1.2,2.2)
1.3 (0.8,2.2)
75th
3.5 (2.8,4.5)

3.4 (2.2,5.2)
2.4 (1.7,3.3)
4.7 (3.4,6.4)
3.6 (2.5,5.2)

3.1 (2.4,4.0)
4.5 (3.4,5.9)

4.0 (2.8,5.6)
5.4 (3.8,7.7)
5.7 (3.3,9.7)
3.2 (2.4,4.2)
2.9 (2.0,4.2)
2.1 (1.5,3.0)
4.0 (3.1,5.2)
3.4 (2.0,5.6)
90th
7.4 (6.0,9.2)

6.5 (4.2,10.1)
5.2 (3.9,7.0)
9.3 (7.0,12.4)
7.8 (5.2,11.7)

6.4 (4.9,8.3)
9.2 (7.1,12.0)

8.6 (6.0,12.4)
10.4 (7.7,14.1)
11.2 (6.8,18.4)
6.7 (4.7,9.5)
5.5 (3.9,7.8)
4.5 (3.3,6.2)
8.1 (6.3,10.3)
7.2 (4.0,12.9)
95th
11.1 (8.9,13.8)

9.3 (5.8,14.7)
8.0 (5.9,10.9)
13.5 (10.2,18.0)
11.7 (7.5,18.3)

9.6 (7.3,12.6)
13.5 (10.3,17.7)

13.0 (8.7,19.5)
14.7 (10.9,19.7)
16.8 (9.5,30.0)
9.9 (6.4,15.4)
7.7 (5.4,11.0)
6.9 (4.9,9.8)
11.7 (9.1,15.0)
10.7 (5.5,20.5)
97th
14.1 (11.2,17.7)

11.5 (7.2,18.3)
10.4 (7.4,14.5)
16.9 (12.7,22.7)
14.9 (9.2,24.1)

12.2 (9.1,16.2)
17.0 (12.8,22.6)

16.4 (10.5,25.5)
18.0 (13.4,24.1)
22.3 (11.4,43.4)
12.8 (7.9,20.6)
9.6 (6.6,13.9)
9.1 (6.3,13.0)
14.6 (11.3,18.8)
13.6 (6.7,27.9)
99th
21.5 (16.5,27.9)

17.0 (10.6,27.1)
16.6 (11.4,24.2)
25.4 (18.4,35.1)
22.3 (12.9,38.6)

18.6 (13.4,25.8)
25.6 (18.5,35.3)

23.6 (14.5,38.6)
25.8 (19.3,34.4)
37.6 (17.5,80.9)
19.4 (11.2,33.6)
14.0 (9.5,20.8)
14.8 (10.2,21.6)
21.4 (16.6,27.7)
21.1 (9.3,47.8)
          1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

          2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
          Table 26a.      UFCR estimates (g/day raw weight, edible portion): Total trophic level 4 fish, youth, <21 years, by demographic

                           characteristics
oo
c_n
Trophic Level 4
Finfish and Shellfish
Youth (<21 yrs)
Age
1 to <3 yrs
3 to <6 yrs
6 to <11 yrs
11 to <16 yrs
16 to <18 yrs
18to<21yrs
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
2.5 (1.9,3.1)

1.3 (0.9,1.8)
1.8 (1.2,2.7)
2.7 (1.9,4.0)
2.4 (1.6,3.5)
3.2 (2.1,4.9)
4.4 (2.9,6.5)

2.2 (1.7,2.8)
2.7 (2.1,3.5)

2.0 (1.4,2.9)
2.2 (1.2,3.7)
2.3 (1.7,3.2)
3.0 (2.3,4.1)
4.2 (2.7,6.5)

2.5 (1.8,3.4)
2.3 (1.7,3.0)
2.8 (2.0,3.8)
2.5 (1.8,3.5)
1.8 (0.8,3.9)
2.0 (0.4,9.8)
2.9 (0.9,9.7)
75th
6.4 (5.3,7.7)

3.3 (2.5,4.2)
4.4 (3.3,6.1)
6.9 (4.8,10.0)
5.9 (4.3,8.1)
7.8 (5.6,10.9)
11.6 (7.9,17.0)

5.8 (4.8,7.0)
7.0 (5.7,8.6)

5.2 (4.0,6.8)
6.1 (3.8,9.7)
6.2 (4.7,8.0)
6.9 (5.4,8.8)
11.1 (7.8,15.7)

6.5 (5.2,8.2)
5.8 (4.6,7.3)
7.4 (5.3,10.4)
6.4 (5.0,8.2)
4.4 (2.2,8.9)
5.9 (1.0,35.5)
8.3 (3.3,21.0)
90th
13.1 (10.8,15.9)

6.6 (5.2,8.4)
8.6 (6.6,11.3)
13.4 (8.6,21.0)
11.3 (8.4,15.3)
15.2 (11.0,21.0)
24.8 (15.6,39.5)

11.8 (9.7,14.2)
14.5 (11.6,18.2)

10.7 (8.4,13.7)
12.6 (8.4,19.0)
12.8 (9.6,17.1)
12.8 (10.2,16.0)
22.5 (16.7,30.4)

13.2 (10.6,16.3)
12.0 (9.7,14.9)
15.7 (10.4,23.5)
12.4 (9.7,15.9)
9.6 (4.1,22.4)
17.4 (2.7,114.0)
18.0 (9.2,34.9)
95th
19.6 (15.8,24.4)

9.7 (7.5,12.4)
12.4 (9.5,16.1)
19.2 (11.6,32.0)
16.2 (11.9,22.2)
21.4 (15.3,30.1)
37.3 (21.3,65.4)

17.2 (14.2,20.9)
22.1 (16.8,29.2)

16.0 (12.5,20.6)
18.4 (12.2,27.7)
19.3 (13.8,26.9)
17.8 (14.2,22.3)
31.7 (22.2,45.3)

19.3 (15.1,24.5)
18.4 (14.5,23.3)
24.2 (15.5,37.9)
17.9 (13.7,23.3)
14.8 (5.7,38.0)
31.0 (6.4,150.2)
26.3 (15.2,45.4)
97th
25.3 (19.9,32.3)

12.5 (9.7,16.3)
15.5 (11.8,20.4)
23.8 (13.8,41.0)
20.2 (14.5,28.3)
26.6 (18.6,38.2)
48.4 (26.3,89.1)

21.6 (17.7,26.5)
29.3 (21.1,40.7)

20.8 (16.0,27.2)
23.1 (15.2,35.1)
25.1 (17.2,36.5)
21.9 (17.4,27.4)
39.4 (25.6,60.7)

24.3 (18.6,31.8)
24.0 (17.8,32.3)
31.9 (19.6,51.9)
22.2 (16.7,29.5)
18.6 (6.9,50.2)
41.0 (11.2,150.2)
32.7 (19.5,54.8)
99th
40.5 (28.8,56.9)

19.3 (14.3,26.1)
23.2 (16.9,31.9)
34.1 (19.1,60.9)
30.4 (20.2,45.7)
39.3 (26.7,57.7)
72.4 (38.6,135.9)

31.9 (25.6,39.8)
49.5 (30.7,80.0)

34.0 (25.0,46.4)
33.4 (20.5,54.4)
41.3 (24.2,70.5)
31.2 (24.7,39.5)
57.3 (32.3,101.7)

38.1 (27.9,52.1)
39.5 (22.0,70.9)
52.8 (29.4,94.8)
32.9 (23.9,45.3)
26.1 (9.2,74.2)
63.9 (25.5,159.9)
50.0 (29.5,84.7)
           Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-

           identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic

           black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
          Table 26b.      UFCR estimates (g/day raw weight, edible portion): Total trophic level 4 fish, youth, <21 years, by geographic area
oo
CS
Trophic Level 4
Finfish and Shellfish
Youth (<21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
Percentiles(95%CI)
50th
2.5 (1.9,3.1)

3.0 (2.0,4.4)
1.8 (1.2,2.7)
2.5 (1.9,3.4)
3.1 (1.9,5.0)

2.4 (1.8,3.2)
2.5 (1.9,3.3)

2.6 (1.7,3.9)
3.0 (2.2,4.2)
2.6 (1.7,4.0)
1.8 (1.1,2.8)
2.9 (1.9,4.4)
1.8 (1.2,2.7)
2.4 (1.8,3.1)
3.4 (1.8,6.6)
75th
6.4 (5.3,7.7)

7.6 (5.4,10.5)
4.8 (3.5,6.4)
6.2 (4.9,8.0)
8.2 (5.0,13.2)

6.4 (5.0,8.2)
6.4 (5.2,7.7)

7.2 (5.0,10.2)
6.9 (5.3,9.0)
6.4 (4.5,9.3)
4.7 (3.4,6.6)
7.7 (5.3,11.2)
4.8 (3.5,6.5)
6.1 (4.7,7.9)
8.9 (4.7,16.6)
90th
13.1 (10.8,15.9)

16.4 (10.3,26.2)
9.7 (7.7,12.1)
12.1 (9.5,15.5)
16.6 (10.2,27.1)

13.2 (10.0,17.5)
13.0 (10.9,15.5)

15.4 (11.0,21.6)
13.6 (10.6,17.4)
12.9 (8.8,18.9)
9.6 (7.0,13.2)
17.1 (9.3,31.5)
9.7 (7.8,12.2)
11.8 (9.0,15.6)
17.4 (9.1,33.5)
95th
19.6 (15.8,24.4)

26.7 (13.0,54.9)
14.2 (11.6,17.4)
17.5 (13.5,22.7)
24.1 (14.9,39.2)

19.7 (14.2,27.3)
19.5 (16.3,23.4)

23.0 (16.4,32.4)
20.2 (15.2,27.0)
19.1 (13.0,28.2)
14.2 (10.1,19.9)
28.1 (10.8,73.4)
14.3 (11.8,17.3)
17.0 (12.6,22.9)
24.8 (12.8,48.2)
97th
25.3 (19.9,32.3)

37.0 (15.6,88.0)
18.1 (14.8,22.1)
21.8 (16.6,28.6)
30.3 (19.0,48.3)

25.3 (17.6,36.4)
25.4 (20.9,30.8)

29.5 (20.8,41.9)
26.6 (19.0,37.2)
24.0 (15.9,36.1)
18.1 (12.6,26.1)
39.8 (13.2,119.5)
18.1 (15.1,21.7)
21.2 (15.5,29.0)
30.7 (15.9,59.1)
99th
40.5 (28.8,56.9)

63.5 (25.0,161.7)
27.8 (22.3,34.6)
32.9 (24.4,44.4)
44.1 (28.5,68.1)

40.6 (24.9,66.2)
40.3 (31.3,52.0)

44.5 (29.1,68.1)
45.4 (28.2,73.2)
36.6 (22.3,60.1)
28.1 (18.2,43.2)
67.2 (21.6,209.3)
27.5 (22.9,33.1)
31.2 (22.5,43.2)
43.8 (23.9,80.4)
          1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

          2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
         Table 27a.      UFCR estimates (g/day raw weight, edible portion): Total freshwater + estuarine trophic level 2 fish, youth, <21 years, by
                           demographic characteristics
oo
Freshwater + Estuarine Trophic Level 2
Finfish and Shellfish
Youth (<21 yrs)
Age
1 to <3 yrs
3 to <6 yrs
6 to <11 yrs
11 to <16 yrs
16 to <18 yrs
18to<21yrs
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
0.3 (0.2,0.4)

0.1 (0.1,0.3)
0.2 (0.1,0.4)
0.2 (0.1,0.4)
0.3 (0.2,0.6)
0.3 (0.2,0.7)
0.4 (0.2,0.8)

0.3 (0.2,0.4)
0.3 (0.2,0.4)

0.5 (0.3,0.7)
0.3 (0.1,0.8)
0.2 (0.1,0.3)
0.5 (0.4,0.7)
0.8 (0.4,1.8)

0.3 (0.2,0.4)
0.4 (0.2,0.5)
0.2 (0.1,0.3)
0.3 (0.2,0.4)
0.1 (0.0,0.5)
0.1 (0.0,0.5)
0.2 (0.1,0.8)
75th
0.9 (0.7,1.2)

0.5 (0.3,0.8)
0.6 (0.4,1.0)
0.7 (0.4,1.1)
1.1 (0.7,1.7)
1.0 (0.6,1.8)
1.4 (0.8,2.4)

0.9 (0.6,1.2)
0.9 (0.6,1.2)

1.5 (1.0,2.1)
1.0 (0.4,2.2)
0.5 (0.4,0.8)
1.5 (1.2,2.0)
2.5 (1.2,5.1)

0.9 (0.6,1.3)
1.1 (0.8,1.5)
0.7 (0.4,1.0)
0.9 (0.6,1.3)
0.4 (0.2,1.2)
0.4 (0.1,2.2)
1.2 (0.2,6.0)
90th
2.3 (1.8,3.0)

1.2 (0.8,1.8)
1.7 (1.1,2.7)
1.9 (1.3,2.8)
2.9 (2.0,4.1)
2.5 (1.5,3.9)
3.4 (2.2,5.5)

2.4 (1.8,3.1)
2.3 (1.7,3.1)

3.5 (2.6,4.8)
2.5 (1.2,5.2)
1.4 (0.9,2.0)
3.5 (2.7,4.5)
5.5 (2.8,11.0)

2.3 (1.6,3.3)
2.9 (2.1,3.8)
1.8 (1.2,2.7)
2.3 (1.6,3.4)
1.2 (0.5,3.0)
1.8 (0.5,7.0)
5.7 (0.8,41.5)
95th
4.0 (3.1,5.2)

2.1 (1.4,3.0)
3.0 (2.0,4.6)
3.5 (2.2,5.5)
4.7 (3.4,6.6)
3.9 (2.6,6.1)
5.6 (3.5,8.9)

4.1 (3.2,5.3)
3.9 (2.9,5.4)

5.6 (4.1,7.7)
4.1 (2.0,8.4)
2.4 (1.6,3.5)
5.6 (4.2,7.4)
8.3 (4.2,16.4)

3.8 (2.7,5.4)
4.8 (3.5,6.5)
3.2 (2.1,4.8)
4.0 (2.8,5.9)
2.1 (0.8,5.4)
3.2 (0.9,11.6)
10.1 (2.4,43.2)
97th
5.5 (4.2,7.3)

2.9 (2.0,4.3)
4.2 (2.8,6.4)
5.0 (3.1,8.3)
6.4 (4.6,8.9)
5.2 (3.4,8.1)
7.4 (4.5,12.1)

5.6 (4.3,7.4)
5.4 (3.9,7.6)

7.5 (5.4,10.4)
5.5 (2.7,11.3)
3.3 (2.2,5.0)
7.4 (5.5,10.0)
10.7 (5.4,21.2)

5.1 (3.6,7.3)
6.5 (4.7,8.9)
4.4 (2.9,6.8)
5.5 (3.7,8.2)
2.9 (1.1,7.5)
4.6 (1.2,16.9)
13.0 (4.3,39.9)
99th
9.8 (7.1,13.4)

5.2 (3.3,8.1)
7.5 (5.0,11.2)
9.3 (5.3,16.5)
11.0 (7.8,15.5)
8.7 (5.5,13.8)
12.3 (7.4,20.6)

9.9 (7.2,13.5)
9.6 (6.6,14.0)

12.4 (8.4,18.2)
9.0 (4.3,19.0)
6.2 (4.0,9.4)
12.3 (8.8,17.3)
15.5 (7.6,31.8)

8.5 (5.7,12.7)
11.1 (7.8,15.8)
7.9 (4.9,12.6)
9.8 (6.4,15.1)
5.4 (2.0,14.3)
9.0 (2.2,37.0)
18.2 (8.6,38.2)
           Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
           identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
           black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
oo
oo
         Table 27b.      UFCR estimates (g/day raw weight, edible portion): Total freshwater + estuarine trophic level 2 fish, youth, <21 years, by
                           geographic area
Freshwater + Estuarine Trophic Level 2
Finfish and Shellfish
Youth (<21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
0.3 (0.2,0.4)

0.3 (0.1,0.5)
0.1 (0.1,0.2)
0.4 (0.2,0.6)
0.3 (0.2,0.6)

0.2 (0.1,0.3)
0.4 (0.3,0.6)

0.4 (0.3,0.7)
0.4 (0.3,0.7)
0.5 (0.3,1.0)
0.2 (0.1,0.4)
0.2 (0.1,0.4)
0.1 (0.1,0.2)
0.3 (0.2,0.5)
0.3 (0.2,0.5)
75th
0.9 (0.7,1.2)

0.8 (0.5,1.5)
0.5 (0.3,0.8)
1.1 (0.7,1.7)
1.1 (0.7,1.7)

0.7 (0.5,1.0)
1.3 (0.9,1.8)

1.5 (0.9,2.4)
1.4 (1.0,2.0)
1.7 (0.9,3.0)
0.7 (0.5,1.2)
0.7 (0.4,1.3)
0.4 (0.3,0.7)
0.9 (0.6,1.4)
0.9 (0.5,1.5)
Percentiles(95%CI)
90th
2.3 (1.8,3.0)

2.1 (1.1,3.7)
1.4 (0.9,2.3)
2.8 (2.0,4.1)
2.9 (1.8,4.6)

1.8 (1.3,2.5)
3.4 (2.5,4.5)

3.8 (2.3,6.4)
3.5 (2.5,4.8)
4.3 (2.4,7.6)
2.1 (1.3,3.4)
1.7 (0.9,3.2)
1.2 (0.7,1.9)
2.2 (1.5,3.1)
2.3 (1.4,3.8)
95th
4.0 (3.1,5.2)

3.4 (1.9,6.2)
2.7 (1.7,4.2)
4.7 (3.3,6.7)
4.9 (3.0,8.1)

3.1 (2.3,4.2)
5.5 (4.1,7.5)

6.1 (3.4,10.9)
5.5 (4.0,7.7)
7.1 (4.1,12.3)
3.6 (2.2,6.2)
2.7 (1.4,5.3)
2.1 (1.3,3.7)
3.6 (2.5,5.1)
3.9 (2.3,6.6)
97th
5.5 (4.2,7.3)

4.7 (2.6,8.4)
3.8 (2.4,6.1)
6.4 (4.5,9.2)
6.6 (3.9,11.3)

4.3 (3.2,5.9)
7.4 (5.3,10.3)

8.1 (4.4,15.0)
7.3 (5.1,10.3)
9.6 (5.7,16.2)
5.1 (3.0,8.8)
3.7 (1.9,7.2)
3.1 (1.8,5.4)
4.9 (3.5,6.9)
5.4 (3.2,9.1)
99th
9.8 (7.1,13.4)

8.0 (4.5,14.3)
7.1 (4.2,11.7)
11.0 (7.6,16.0)
11.2 (6.3,20.0)

7.7 (5.5,10.9)
12.4 (8.6,17.7)

13.5 (7.3,25.0)
11.7 (7.7,17.8)
15.2 (9.9,23.3)
9.0 (5.1,15.8)
6.5 (3.3,12.9)
5.9 (3.2,10.7)
8.6 (6.0,12.1)
9.2 (5.2,16.3)
         1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
           TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

         2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
           border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
         Table 28a.      UFCR estimates (g/day raw weight, edible portion): Total freshwater + estuarine trophic level 3 fish, youth, <21 years, by
                           demographic characteristics
oo
Freshwater + Estuarine Trophic Level 3
Finfish and Shellfish
Youth (<21 yrs)
Age
1 to <3 yrs
3 to <6 yrs
6 to <11 yrs
11 to <16 yrs
16 to <18 yrs
18to<21yrs
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
0.4 (0.3,0.6)

0.2 (0.1,0.4)
0.4 (0.2,0.8)
0.5 (0.3,0.7)
0.3 (0.2,0.5)
0.6 (0.3,1.2)
0.7 (0.4,1.4)

0.4 (0.3,0.6)
0.5 (0.3,0.7)

0.5 (0.3,0.7)
0.4 (0.2,0.9)
0.3 (0.2,0.5)
1.3 (0.9,1.9)
1.3 (0.8,2.1)

0.5 (0.3,0.7)
0.5 (0.3,0.7)
0.4 (0.2,0.6)
0.4 (0.3,0.7)
0.2 (0.1,0.9)
0.1 (0.0,0.5)
0.4 (0.1,1.2)
75th
1.3 (1.0,1.7)

0.6 (0.3,1.2)
1.2 (0.7,1.9)
1.3 (1.0,1.9)
1.1 (0.8,1.6)
1.6 (1.0,2.8)
2.0 (1.2,3.4)

1.2 (0.9,1.6)
1.3 (1.0,1.8)

1.3 (1.0,1.8)
1.3 (0.7,2.4)
0.8 (0.6,1.2)
3.3 (2.4,4.7)
3.4 (2.3,4.9)

1.5 (1.1,2.1)
1.4 (1.0,1.9)
1.1 (0.7,1.7)
1.3 (0.9,1.8)
0.7 (0.2,2.3)
0.7 (0.3,2.0)
1.9 (0.4,9.3)
90th
3.1 (2.4,3.9)

1.4 (0.8,2.6)
2.5 (1.6,3.9)
3.1 (2.2,4.5)
2.8 (2.1,3.7)
3.8 (2.6,5.6)
4.6 (3.1,6.8)

3.0 (2.3,3.9)
3.1 (2.4,4.0)

3.0 (2.2,3.9)
2.8 (1.5,5.4)
1.8 (1.3,2.5)
6.7 (4.7,9.3)
6.8 (4.5,10.3)

3.7 (2.8,5.0)
3.2 (2.3,4.3)
2.7 (1.9,4.0)
2.9 (2.1,3.9)
1.8 (0.7,4.8)
2.3 (0.9,5.5)
7.7 (0.8,71.2)
95th
5.0 (3.9,6.4)

2.2 (1.2,4.1)
3.8 (2.5,5.9)
5.0 (3.3,7.7)
4.8 (3.6,6.4)
6.2 (4.2,9.2)
7.1 (4.6,11.1)

4.9 (3.7,6.5)
5.1 (3.9,6.6)

4.7 (3.4,6.4)
4.3 (2.2,8.4)
2.7 (1.9,3.9)
9.6 (6.7,13.5)
9.8 (5.9,16.3)

6.0 (4.4,8.0)
5.0 (3.7,6.9)
4.5 (3.1,6.6)
4.6 (3.4,6.4)
2.9 (1.1,7.7)
4.5 (1.7,12.1)
14.8 (2.1,104.0)
97th
6.7 (5.1,8.9)

3.0 (1.6,5.5)
4.9 (3.2,7.6)
6.8 (4.3,10.7)
6.5 (4.8,8.9)
8.4 (5.5,12.9)
9.3 (5.7,15.1)

6.7 (4.9,9.1)
6.8 (5.1,9.0)

6.3 (4.5,9.0)
5.6 (2.8,11.0)
3.5 (2.4,5.1)
11.9 (8.3,17.0)
12.4 (7.2,21.4)

7.8 (5.7,10.9)
6.6 (4.7,9.3)
6.1 (4.2,9.0)
6.2 (4.4,8.9)
3.8 (1.4,10.5)
7.0 (2.3,21.0)
19.8 (3.7,105.9)
99th
11.5 (8.1,16.2)

5.2 (2.7,10.0)
7.8 (4.8,12.7)
11.3 (6.5,19.5)
11.2 (7.9,15.8)
14.2 (8.6,23.4)
15.0 (8.7,26.0)

11.5 (8.0,16.5)
11.5 (8.0,16.5)

11.2 (7.3,17.1)
8.9 (4.5,17.6)
5.8 (3.6,9.3)
17.7 (12.2,25.9)
17.3 (9.0,33.4)

12.7 (8.6,18.7)
10.8 (7.4,15.7)
10.2 (6.8,15.4)
10.9 (7.0,17.1)
6.6 (2.3,18.7)
18.8 (5.1,68.7)
30.9 (7.7,123.0)
           Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
           identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
           black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
Table 28b.     UFCR estimates (g/day raw weight, edible portion): Total freshwater + estuarine trophic level 3 fish, youth, <21 years, by
                 geographic area
Freshwater + Estuarine Trophic Level 3
Finfish and Shellfish
Youth (<21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
0.4 (0.3,0.6)

0.4 (0.2,0.7)
0.3 (0.1,0.5)
0.7 (0.4,1.1)
0.5 (0.3,0.7)

0.4 (0.2,0.5)
0.6 (0.4,0.9)

0.5 (0.3,0.9)
0.8 (0.5,1.2)
0.8 (0.3,1.9)
0.4 (0.2,0.6)
0.3 (0.2,0.6)
0.2 (0.1,0.4)
0.6 (0.4,0.8)
0.4 (0.3,0.7)
75th
1.3 (1.0,1.7)

1.1 (0.6,1.8)
0.8 (0.4,1.4)
1.8 (1.2,2.8)
1.4 (1.0,1.9)

1.1 (0.8,1.5)
1.7 (1.2,2.5)

1.6 (1.1,2.5)
2.1 (1.3,3.2)
2.1 (1.0,4.5)
1.1 (0.6,2.0)
0.9 (0.5,1.6)
0.7 (0.4,1.2)
1.6 (1.1,2.2)
1.2 (0.8,1.8)
Percentiles(95%CI)
90th
3.1 (2.4,3.9)

2.3 (1.3,4.0)
2.0 (1.2,3.4)
4.2 (2.9,6.0)
3.2 (2.4,4.4)

2.5 (1.9,3.2)
4.2 (2.9,5.9)

4.1 (2.6,6.6)
4.5 (3.1,6.7)
5.1 (2.6,10.2)
3.0 (1.6,5.5)
2.0 (1.1,3.4)
1.6 (0.9,3.0)
3.5 (2.6,4.8)
2.7 (1.8,3.8)
95th
5.0 (3.9,6.4)

3.5 (2.0,6.3)
3.4 (1.9,5.9)
6.5 (4.5,9.5)
5.3 (3.7,7.6)

4.0 (3.1,5.2)
6.7 (4.6,9.7)

6.8 (4.0,11.7)
6.9 (4.7,10.1)
8.6 (3.9,18.9)
5.1 (2.5,10.3)
2.9 (1.6,5.2)
2.6 (1.4,4.8)
5.5 (4.1,7.5)
4.2 (2.9,6.0)
97th
6.7 (5.1,8.9)

4.6 (2.6,8.4)
4.8 (2.7,8.4)
8.6 (5.9,12.7)
7.1 (4.7,10.8)

5.4 (4.1,7.1)
9.0 (6.0,13.4)

9.2 (5.0,16.9)
8.9 (6.0,13.1)
12.1 (5.0,29.4)
7.2 (3.4,15.0)
3.8 (2.1,6.8)
3.7 (2.0,6.7)
7.2 (5.2,10.0)
5.5 (3.8,8.0)
99th
11.5 (8.1,16.2)

7.5 (4.0,14.2)
8.9 (5.0,15.7)
14.0 (9.0,22.0)
12.0 (7.0,20.4)

9.0 (6.6,12.4)
15.0 (9.5,23.5)

15.0 (7.6,29.6)
13.9 (9.2,20.9)
21.2 (9.0,50.2)
12.6 (5.9,26.7)
5.9 (3.1,11.4)
6.7 (3.8,11.8)
11.5 (7.9,16.6)
9.0 (5.9,13.8)
1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
 TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
 border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
Table 29a.     UFCR estimates (g/day raw weight, edible portion): Total freshwater + estuarine trophic level 4 fish, youth, <21 years, by
                 demographic characteristics
Freshwater + Estuarine Trophic Level 4
Finfish and Shellfish
Youth (<21 yrs)
Age
1 to <3 yrs
3 to <6 yrs
6 to <11 yrs
11 to <16 yrs
16 to <18 yrs
18 to <21 yrs
Gender
Female
Male
Race/Ethnicity1
Mexican American
Other Hispanic
Non-Hispanic White
Non-Hispanic Black
Other Race
Income
$0 to <$20K
$20 to <$45K
$40 to <$75K
$75+K
>$20K
Refused/Don't Know Income
Income Missing
Percentiles(95%CI)
50th
0.2 (0.1,0.3)

0.1 (0.0,0.3)
0.1 (0.0,0.2)
0.2 (0.1,0.3)
0.1 (0.1,0.3)
0.3 (0.1,1.1)
0.2 (0.1,0.5)

0.1 (0.1,0.2)
0.2 (0.1,0.4)

0.1 (0.1,0.2)
0.1 (0.0,0.4)
0.1 (0.1,0.3)
0.4 (0.2,0.8)
0.4 (0.2,0.7)

0.2 (0.1,0.3)
0.1 (0.1,0.3)
0.2 (0.1,0.3)
0.2 (0.1,0.4)
0.0 (0.0,0.3)
0.0 (0.0,0.1)
0.1 (0.0,0.4)
75th
0.6 (0.4,1.0)

0.4 (0.2,0.9)
0.3 (0.2,0.6)
0.7 (0.4,1.2)
0.6 (0.3,1.1)
1.1 (0.4,3.3)
0.8 (0.4,1.7)

0.5 (0.3,0.8)
0.8 (0.5,1.4)

0.4 (0.3,0.7)
0.6 (0.2,1.9)
0.5 (0.2,0.9)
1.4 (0.7,2.6)
1.4 (0.7,2.6)

0.7 (0.4,1.1)
0.5 (0.3,0.9)
0.6 (0.4,1.0)
0.8 (0.4,1.4)
0.3 (0.1,1.6)
0.1 (0.0,0.5)
0.4 (0.1,1.9)
90th
2.0 (1.2,3.2)

1.2 (0.6,2.3)
1.1 (0.6,2.1)
2.2 (1.3,3.7)
1.9 (0.9,3.7)
3.2 (1.2,8.7)
2.7 (1.3,5.4)

1.5 (0.9,2.3)
2.5 (1.5,4.3)

1.4 (0.9,2.2)
2.0 (0.4,8.6)
1.5 (0.8,3.0)
3.9 (2.1,7.5)
4.4 (2.2,8.7)

2.1 (1.3,3.5)
1.6 (0.9,2.9)
2.0 (1.2,3.4)
2.3 (1.3,4.1)
1.7 (0.5,5.7)
0.5 (0.1,2.1)
1.6 (0.3,7.8)
95th
3.9 (2.4,6.3)

2.2 (1.2,4.0)
2.3 (1.0,5.0)
4.1 (2.3,7.5)
3.7 (1.8,7.6)
5.8 (2.1,15.9)
5.3 (2.5,11.1)

2.9 (1.8,4.5)
4.9 (2.9,8.4)

2.7 (1.7,4.5)
4.1 (0.7,22.7)
2.9 (1.4,5.7)
7.3 (3.8,14.1)
8.8 (4.2,18.3)

4.1 (2.4,6.9)
3.2 (1.8,5.8)
4.0 (2.3,7.0)
4.5 (2.6,7.7)
3.7 (1.2,10.9)
1.3 (0.3,5.5)
3.6 (0.8,15.5)
97th
6.0 (3.6,9.9)

3.3 (1.9,5.8)
3.6 (1.4,9.5)
6.2 (3.3,11.7)
5.6 (2.7,11.8)
8.7 (3.2,23.5)
8.3 (3.8,17.9)

4.4 (2.8,7.1)
7.5 (4.2,13.1)

4.3 (2.5,7.2)
6.2 (0.9,43.6)
4.3 (2.1,8.8)
10.8 (5.5,20.9)
13.3 (6.0,29.7)

6.2 (3.6,10.9)
5.0 (2.7,9.2)
6.2 (3.4,11.3)
6.6 (3.7,11.6)
5.8 (2.0,16.2)
2.4 (0.5,11.5)
5.5 (1.4,21.4)
99th
13.2 (7.4,23.5)

7.1 (4.1,12.3)
8.5 (1.8,39.4)
13.7 (6.7,28.0)
12.1 (5.4,27.3)
18.4 (6.7,50.0)
18.9 (7.3,48.8)

9.8 (5.8,16.7)
16.5 (8.8,30.8)

9.8 (5.1,18.9)
13.9 (1.3,145.9)
9.1 (4.2,19.9)
22.5 (11.1,45.3)
30.7 (10.6,88.8)

13.3 (6.9,25.3)
11.3 (5.6,22.9)
14.6 (7.2,29.7)
13.9 (7.5,25.8)
14.1 (5.3,37.2)
9.5 (1.4,64.2)
12.1 (3.5,41.6)
 Race/ethnicity is as defined by NHANES. Respondents who self-identified as "Mexican American" were coded as such regardless of their other race-ethnicity identities. Otherwise, self-
 identified "Hispanic" ethnicity was coded as "Other Hispanic." All other non-Hispanic participants were then categorized based on their self-reported races: non-Hispanic white, non-Hispanic
 black, and other non-Hispanic race including non-Hispanic multiracial (other race).

-------
Table 29b.     UFCR estimates (g/day raw weight, edible portion): Total freshwater + estuarine trophic level 4 fish, youth, <21 years, by
                 geographic area
Freshwater + Estuarine Trophic Level 4
Finfish and Shellfish
Youth (<21 yrs)
Region1
Northeast
Midwest
South
West
Coastal Status2
Noncoastal
Coastal
Coastal/Inland Region12
Pacific
Atlantic
Gulf of Mexico
Great Lakes
Inland Northeast
Inland Midwest
Inland South
Inland West
50th
0.2 (0.1,0.3)

0.1 (0.0,0.2)
0.2 (0.0,0.6)
0.2 (0.1,0.4)
0.2 (0.1,0.3)

0.1 (0.1,0.3)
0.2 (0.1,0.4)

0.2 (0.1,0.3)
0.2 (0.1,0.4)
0.3 (0.1,0.5)
0.2 (0.1,0.5)
0.1 (0.0,0.2)
0.2 (0.0,0.6)
0.2 (0.1,0.3)
0.2 (0.1,0.3)
75th
0.6 (0.4,1.0)

0.3 (0.1,0.7)
0.7 (0.2,2.3)
0.8 (0.5,1.3)
0.6 (0.4,0.9)

0.6 (0.3,1.0)
0.7 (0.4,1.3)

0.6 (0.3,1.1)
0.8 (0.4,1.4)
0.9 (0.5,1.8)
0.8 (0.3,2.2)
0.2 (0.1,0.6)
0.7 (0.2,2.4)
0.7 (0.4,1.1)
0.5 (0.3,0.9)
Percentiles(95%CI)
90th
2.0 (1.2,3.2)

1.0 (0.4,2.3)
2.4 (0.8,7.0)
2.4 (1.6,3.7)
1.7 (1.1,2.7)

1.8 (1.0,3.1)
2.4 (1.5,4.0)

1.9 (1.1,3.2)
2.4 (1.3,4.3)
2.8 (1.5,5.2)
2.8 (1.0,7.8)
0.8 (0.3,2.1)
2.3 (0.7,6.9)
2.1 (1.3,3.3)
1.6 (0.9,2.7)
95th
3.9 (2.4,6.3)

2.0 (0.8,5.0)
4.9 (1.8,13.3)
4.6 (2.9,7.3)
3.3 (2.0,5.4)

3.5 (2.0,6.0)
4.7 (2.8,7.8)

3.7 (2.1,6.2)
4.6 (2.5,8.5)
5.4 (3.0,9.8)
5.7 (2.0,16.7)
1.7 (0.6,4.9)
4.5 (1.6,12.5)
3.9 (2.4,6.4)
3.0 (1.7,5.2)
97th
6.0 (3.6,9.9)

3.1 (1.1,8.5)
7.6 (2.9,20.0)
6.9 (4.2,11.3)
4.9 (2.9,8.2)

5.3 (3.0,9.3)
7.2 (4.2,12.3)

5.5 (3.1,9.7)
7.0 (3.7,13.5)
8.1 (4.5,14.5)
9.1 (3.0,27.9)
2.7 (0.8,8.8)
6.8 (2.6,17.9)
5.9 (3.5,9.9)
4.4 (2.5,7.8)
99th
13.2 (7.4,23.5)

6.9 (1.9,24.9)
17.9 (7.0,45.8)
14.7 (8.3,25.8)
10.2 (5.4,19.3)

11.7 (6.4,21.6)
15.9 (8.3,30.5)

11.8 (6.0,23.2)
15.3 (6.5,35.8)
15.7 (7.7,32.0)
22.6 (6.5,78.6)
6.0 (1.3,28.0)
15.7 (6.5,37.9)
12.7 (7.2,22.4)
9.1 (4.6,18.0)
1 U.S. regions are the U.S. Census Bureau regions. Midwest = OH, Ml, IN, Wl, IL, MO, IA, MN, SD, ND, NE, KS. Northeast = PA, NY, NJ, CT, Rl, MA, NH, VT, ME. South = DE, MD, DC, VA, WV, KY,
 TN, NC, SC, GA, AL, MS, FL, LA, AR, OK, TX. West = NM, CO, WY, MT, ID, UT, AZ, NV, CA, OR, WA, AK, HI.

2 Coastal regions include counties bordering the 3 coasts (Pacific, Atlantic, and Gulf of Mexico) and the Great Lakes and estuaries and bays. Additionally, any county that did not directly
 border a coast, but the central point was within 25 miles of a coast was defined as coastal. The inland regions are the remaining counties in each of the 4 Census Regions.

-------
5.3       Uncertainty

The estimated fish consumption rates may be uncertain due to either bias or random variation. Bias
results in a consistently high or consistently low fish consumption rate relative to the true or desired
value. Variation results in an uncertain fish consumption rate that might be either higher or lower
than the true value.

The primary sources of random variation are the following:

      •    Sampling error associated with the random selection of NHANES respondents. For
           example, if different counties and individuals had been selected for the NHANES data
           collection, the data and FCRs would be different.
      •    Random differences due to the simulation of usual fish consumption for each
           NHANES respondent. This source of variation can be reduced by increasing the
           number of simulations.

The confidence intervals for the fish consumption rates account for both of these sources of
variation. Estimates for coastal regions will be less precise than national estimates because the
number of respondents in the coastal regions is a fraction of the number of NHANES respondents
nationally. As a result, the confidence intervals for coastal regions are wider than  for national
estimates. Similarly, if there are fewer respondents with reported fish  consumption in two 24-hour
recalls, there is less data to estimate the parameters and particularly the variance components,
resulting in more uncertainty in the fish consumption estimates and wider confidence intervals.

There are multiple sources of bias that can affect the fish consumption rates including:

      •    Seasonally;
      •    Respondent bias;
      •    Use of standard recipes to calculate fish consumption amounts  from the NHANES 24-
           hour recalls;
      •    Classification of the fish consumed into types offish habitats;
      •    Bias associated with the estimation method (either the NCI or EPA Method) and its
           assumptions; and
      •    Use of approximate analysis weights for coastal versus non-coastal comparisons.

Each of these sources of bias is discussed in more detail below.
                                             93

-------
5.3.1     Seasonality

Fish consumption, especially of recreationally or sport-caught fish, is likely to vary by season.
NHANES collects data throughout the year. However, NHANES generally collects data in northern
counties in the summer and southern counties in the winter. Thus the estimates may overestimate
usual intake in the northern regions of the United States and underestimate usual intake in the
southern regions of the United States if summer fish consumption is higher than winter fish
consumption. There is no way to estimate this season effect as there are little or no NHANES data
from northern counties in the winter and southern counties in the summer.
5.3.2     Bias in the Reported Fish Consumption

The reported fish consumption is a combination of the frequency offish consumption and the
amount consumed if fish was consumed. The reported fish consumption may be biased if
NHANES respondents tend to report consistently more or less fish consumption in the 24-hour
recall than actually occurred. Assessing if the reported values are biased requires comparing reported
values to estimates obtained using other data collection approaches, such as analysis of duplicate
meals. Over the years, much research has gone into assessing dietary intake, resulting in the
procedures used by NHANES. As a  result, the estimates from NHANES are generally considered to
have minimal bias. Nonetheless, the estimates may be biased and the bias may be different for
different communities or subpopulations.
5.3.3     Use of Standard Recipes

The FNDDS utilizes standard recipes for foods reported consumed. NHANES participants do not
supply specific recipes of the foods they consumed. They provide details such as whether the fish
was breaded,  cooked in margarine, baked or broiled, etc., but they do not provide exact recipes
(which they are likely not to know anyway). For example, the standard recipe for the food "Scallops
and noodles with cheese sauce" is approximately 35 percent fish. However, the true recipe for the
food consumed by an NHANES respondent may have less fish or more fish than the standard
recipe. Additionally, there is uncertainty associated with the moisture loss values for processing and
cooking methods. They are average values of moisture loss given the various processing and cooking
methods. If participants cooked their fish a bit longer than the moisture loss would be a bit greater
than average, and if they cooked it a bit less, the moisture loss would be a bit less than average.
                                            94

-------
5.3.4     Habitat Assignments

There is some uncertainty associated with the assignment of habitats to reported fish consumption.
When the raw data are processed by NHANES, fish species reported consumed are combined into
groups. Generally, these groupings are based on taxonomic groups. This grouping of species
complicates the assignment of habitat because in some cases, the grouped fish can inhabit different
habitats and there is no way to determine the exact species the participant consumed. For some
species, apportioning relied on NOAA landings data to assign species of fish groups with many
species (e.g., clams) to habitats. Bias in the proportion of each species assigned to each habitat will
directly affect the corresponding fish consumption rate. For example, if more fish are assigned to
the estuarine habitat, then the total amount of estuarine fish consumed and the percentiles of fish
consumption will be higher than if fewer fish are assigned to the habitat. Even if the allocation to
fish habitats is unbiased overall, there may be bias  for local estimates. For example, if residents in
coastal counties each more locally caught estuarine fish and non-coastal residents eat more
commercial non-estuarine fish of the same species, the estimated proportion of estuarine fish will be
biased low for the coastal counties and biased high for  the non-coastal counties.
5.3.5     Estimation of Usual Fish Consumption

Measurements of usual fish consumption are very difficult to obtain. Since usual fish consumption is
a long-term average, we would need many 24-hour recalls over a long time to approximate what
"usual intake" is trying to assess; therefore we rely on a statistical model and associated assumptions
to estimate usual intake. As a result, the estimates of usual fish consumption depend in part on the
statistical assumptions.

The model makes certain assumptions, such as, 24-hour recalls provide unbiased estimates of fish
consumption, all respondents are fish consumers (at least occasionally), and the distribution of fish
consumption among those reporting consumption in a 24-hour recall is normally distributed for
some power transformation. The validity of these assumptions can be discussed and, to some extent
evaluated using data.

The estimates of the frequency of fish consumption depend in part on how non-consumers (those
who never eat fish or don't eat fish for a long time)  are treated. From two 24-hour recalls it is not
possible to separate true non-consumers from those who did not happen to report fish consumption
in either recall. A similar problem relates to  consumption of small amounts for fish. Should a person
                                            95

-------
who never eats an identifiable piece of fish but uses a salad dressing with a small amount of fish in it
be considered a regular consumer of a very small amount or a non-consumer of fish? Whether a
meal is classified as having fish may depend on the procedures used to ask the questions and the
recipes used to estimate fish consumption. Unfortunately we do not have the data needed to identify
non-consumers. Having non-consumers in the data will lower the overall probability offish
consumption (P) but increase the variance of the probability offish consumption among individuals.
The resulting effect on the upper percentiles of the distribution is not clear.

The reported amount of fish consumption will vary from one 24-hour recall to another, in part
because the respondents may be poor at estimating the amount consumed and in part because the
consumption amounts are reported in rounded units, such as a cup or a pint, but not 1.267 cups.
The rounding adds some uncertainty  to the estimates. The within-person variance component
accounts for uncertainty due to poor  estimation by the respondent and rounding that is part of the
process. Because the definition  of usual fish consumption does not include the within-person
variation, this source of error should contribute minimal bias to the estimates of usual fish
consumption.

The statistical models for the NCI and EPA Methods make some assumptions to simplify the
computations, such as an assumption that variance components are normally distributed, additive in
the transformed scale, and linearly correlated. The assumption that the person-specific random
effect in the probability model is normally distributed is difficult to test without many more 24-hour
recalls for each person. The  assumption that the two variance components in the amount model are
normally distributed is generally consistent with the observation that the Box-Cox transformed
consumption amounts are roughly normally distributed. Nevertheless, other assumptions may imply
a similar distribution for the reported amounts while using a somewhat different assumption for the
person-specific variance component and thus somewhat different estimates of fish consumption.
Because the estimated parameters must be consistent with the reported data, the general center and
spread of the predicted distribution will be similar regardless of the distributional assumptions.
Specific percentiles may be either higher or lower using different assumptions or may be relatively
insensitive to the distributional assumptions. Although these assumptions are common in other
statistical applications, it is difficult to assess how the estimates might change using other
assumptions.

If the model assumptions  are accepted as  reasonable, then the question is whether the estimates
from the model are biased. The estimates  are based on maximum likelihood, which can produce
biased estimates, particularly variance estimates, with small sample sizes. However, convergence
                                            96

-------
theory says maximum likelihood is best with large sample sizes. Due to the relatively large sample
sizes, we expect the estimates to have relatively little bias compared to the size of the confidence
intervals.

The fish consumption estimates depend in part on the independent predictors used in the model.
When different predictors are used, the estimates change. It is impossible to know what the best set
of predictors is. A systematic approach was used to selecting the independent predictors from the
available predictors in an effort to minimize any bias. The estimates have unknown bias due to the
decisions that were made.

Relative to the NCI Method, the EPA Method uses approximate methods to estimate  the
parameters and as a result, percentile estimates from the EPA Method may be more biased than
from the NCI Method. The analysis of simulated data and the comparison of the FCR percentiles
between the NCI and EPA Methods suggest that the results from the EPA Method have little bias
overall, although estimates for some percentiles may be more biased than for others. The bias
observed in the comparisons between the NCI Method and the EPA Method are very small
compared to the spread of the overall  distribution, generally small compared to the width of the
confidence interval, and on the same order as  bias due to other sources such as the selection of the
independent predictors.
5.3.6     Weights for Coastal Versus Non-Coastal Regions

The U.S. Census regions are used in the calculation of the NHANES weights. However, the analysis
using coastal versus non-coastal regions is looking at smaller areas than intended when the weights
were constructed. Some of the coastal/noncoastal regions cross these Census regions. As a result,
comparisons among coastal and non-coastal regions may be slightly biased or less precise then
indicated by the confidence intervals. At the same time, the weights also adjust for oversampling of
some populations and survey nonresponse, so we believe it is important to use the weights. While
the estimates may be more imprecise and there may be some uncertainty due to the weighting, they
are still a better representation for each coastal/noncoastal area then using unweighted or national
estimates.
                                            97

-------
                                                              Discussion
Fish consumption is higher among males compared to females and increases with increasing age,
although persons aged 65 and over show decreased consumption. People of races other than Black
and White have the highest fish consumption rates of all race/ethnicity groups, with significant
differences observed across all percentiles and many fish types, excluding freshwater and estuarine
fish, trophic level 2 fish, trophic level 2 freshwater and estuarine fish, and trophic level 2 marine fish.
The  other race category consists of Asian, Native American, Pacific and Caribbean Islander, Alaska
Native, multiracial, and unknown race. There is a general increase in consumption as income
increases.

People in the Northeast have higher total fish consumption rates than those living in the other
Census regions, while people in the Midwest have the lowest rates. Significant differences are
observed between the regions. The inland regions generally have lower fish consumption rates than
the coastal regions  except for the Great Lakes region, which is more similar to  an inland region, and
the inland Northeast, which appears more similar to a coastal region. This pattern is different for
freshwater fish for which the people in the inland south, Great Lakes, and the Gulf of Mexico have
the highest consumption rates.
                                             98

-------
                                                              References
Ahmed, F. E., Hattis, D., Wolke, R.E., and Stemman, D. (1993). Human health risks due to
      consumption of chemically contaminated fishery products. 'EnvironmentalHealth Perspectives, 101
      (Supplement. 3), 297-302.

Ahuja, J. K. A., Montville, J.B., Omolewa-Tomobi, G., Heendeniya, K. Y., Martin, C. L., Steinfeldt,
      L. C., Anand, J., Adler, M. E., La Comb, R. P., and Moshfegh, A. J. (2012). USDA Food and
      Nutrient Database for Dietary Studies, 5.0. Beltsville, MD: U.S. Department of Agriculture,
      Agricultural Research Service, Food Surveys Research Group.

Dodd, K. W., Guenther, P. M., Freedman, L. S., Subar, A. F, Kipnis, V., Midthune, D., Tooze, J.,
      and Krebs-Smith, S. M. 2006. Statistical methods for estimating usual intake of nutrients and
      foods: A review of the theory. Journal of the American Dietetic Association, 106, 1640-1650.

Gilbert, R. O. (1987). Characterizing lognormal populations. In Statistical methods for environmental
      pollution monitoring  (pp.  164-176). New York: Van Norstrand Reinhold.

Hames, P. S., Kama, M. Y., Guilkey, D. K., and Popkm, B. M. (2003). Weekend eating in the United
      States is linked with greater energy, fat, and alcohol intake. Obesity ^search, 11: 945—949.

Judkins, D. (1990). Fay's method for variance estimation. Journal of Official Statistics, 6, 223-240.

Kipnis, V., Midthune, D., Buckman, D. W., Dodd, K. W., Guenther, P. M., Krebs-Smith, S. M.,
      Subar, A. F., Tooze, J. A., Carroll, R. J., and  Freedman, L. S. (2009).  Modeling data with excess
      zeros and measurement error: Application to evaluating relationships between episodically
      consumed foods and health outcomes. Biometrics; 65, 1003-1010.

Mahaffey, K. R., Clickner, R. P., and Jeffries, R. A. (2009). Adult women's  blood mercury
      concentrations vary regionally in the United  States: Association with patterns offish
      consumption (NHANES 1999-2004). Environmental Health Perspectives, 117, 47-53.

National Center for Health Statistics. (2005). Analytic and reporting guidelines: The National Health and
      Nutrition Examination Survey Examination Survey. December 2005. Retrieved March 27, 2014 from
      http://www.cdc.gov/nchs/data/nhanes/nhanes  03 04/nhanes analytic guidelines  dec  200
      5.pdf.

National Center for Health Statistics. (2009). National Health and Nutrition Examination Survey MEC in-
      person dietary interviewers Procedures Manual. Retrieved June 26,  2013 from http: / 7www.cdc.gov/
      nchs/data/nhanes/nhanes 09 10/DietaryInterviewers  Inperson.pdf.
                                             99

-------
National Center for Health Statistics. Continuous NHANES Web Tutorial. Variance Estimation. Key
      concepts about variance estimation within NHANES. Retrieved March 27, 2014 from
      http://www.cdc.gov/nchs/tutorials/NHANES/SurveyDesign/VarianceEstimation/Infol.htm.

National Center for Health Statistics. (2013). About the National Health and Nutrition Examination
      Survey. Retrieved March 27, 2014 from
      http://www.cdc.gov/nchs/nhanes/about nhanes.htm.

National Oceanic and Atmospheric Administration (NOAA). 2014. Commercial Fisheries Statistics.
      Retrieved March 27, 2014 from .http://www.st.nmfs.noaa.gov/commercial-fisheries/.

National Oceanic and Atmospheric Administration (NOAA). 2014. Fishwatch. U.S. Seafood  Facts.
      Retrieved March 27, 2014 from
      http://www.fishwatch.gov/wild seafood/outside  the us.htm.

National Research Council. (2000). Toxicological effects of methjlmercury. Washington, DC: National
      Academies Press.

Rust, K.  (1986). Efficient replicated variance estimation. Proceedings of the Section on Survey Research
      Methods (pp.  81-87). Alexandria, VA: American Statistical Association.

Thompson, M. R., and Boekelheide, K. (2013). Multiple environmental chemical exposures to lead,
      mercury and polychlorinated biphenyls among childbearing-aged women (NHANES 1999-
      2004): Body burden and risk factors. Environmental ^search, 12, 23-30. doi:
      10.1016/j.envres.2012.10.005. Epub 2012 Nov 16.

Tooze, J. A., Kipnis, V, Buckman, D. W., Carroll, R. J., Freedman, L. S., Guenther, P. M., Krebs-
      Smith, S. M., Subar, A. F., and Dodd, K. W. (2010). A mixed-effects model approach for
      estimating the distribution of usual intake of nutrients: The NCI Method. Statistics in Medicine,
      29, 2857-2868.

Tooze, J. A., Midthune, D., Dodd, K. W., Freedman, L. S., Krebs-Smith, S. M., Subar, A. F.,
      Guenther, P. M., Carroll, R. J.,  and Kipnis, V. (2006). A new statistical method for estimating
      the usual intake of episodically consumed foods with application  to their distribution. Journal of
      the American Dietetic Association, 106, 1575-1587.

U.S. Department of Agriculture. (2006). Food and Nutrient Database for Dietary Studies, 2.0.
      Beltsville, MD: Agricultural Research Service, Food Surveys Research Group.

U.S. Department of Agriculture. (2008). Food and Nutrient Database for Dietary Studies, 3.0.
      Beltsville, MD: Agricultural Research Service, Food Surveys Research Group.

U.S. Department of Agriculture. (2010). Food and Nutrient Database for Dietary Studies, 4.1.
      Beltsville, MD: Agricultural Research Service, Food Surveys Research Group.

U.S. Department of Agriculture, Agricultural Research Service. (2013).  USD A. Automated Multiple-
      Pass Method. Retrieved March 27, 2014 from
      http://www.ars.usda.gov/Services/docs.htm?docid=7710.
                                             100

-------
U.S. Environmental Protection Agency. (1995). Great Lakes Water Quality Initiative Technical Support
      Document for the Procedures to Determine 'Rioaccumulation Factors. EPA/820 /B-95/005. Washington,
      DC: U.S. Environmental Protection Agency, Office of Water.

U.S. Environmental Protection Agency. (1997). Mercury study report to Congress: An assessment of exposure
      to mercury in the United States. (Volume 4). EPA-452/R-97-006. Washington, DC: U.S.
      Environmental Protection Agency, Office of Air Quality Planning and Standards and Office
      of Research and Development.

U.S. Environmental Protection Agency. (2002a). Estimated per capita fish consumption in the United States.
      EPA-821- C- 02-003. Washington, DC: U.S. Environmental Protection Agency, Office of
      Science and Technology, Engineering and Analysis Division.

U.S. Environmental Protection Agency. (2002b). Trophic level and exposure analyses for selected piscivorous
      birds and mammals. Volume III: Appendices. Washington, DC: U.S. Environmental Protection
      Agency, Office of Science and Technology, Office of Water.

U.S. Environmental Protection Agency. (2003). Methodology for deriving ambient water quality criteria for the
      protection of human health (2000). Technical Support Document Volume 2: Development of
      national bioaccumulation factors. EPA-822-R-03-030. Washington, DC: U.S. Environmental
      Protection Agency, Office of Science and Technology, Office of Water.

Wolter, K. M. (1985). Introduction to variance estimation. New York: Springer-Verlag.
                                             101

-------