Nutrition Services

The Pediatric Nutrition Surveillance System (PedNSS) monitors the prevalence of specific health indicators of nutrition risk in low-income infants, children and adolescents. As surveillance is a continuous process, this allows for the evaluation of health indicators for this population over time. This data is used to supply information for planning health priorities and policies, and guiding, improving and supporting decisions regarding nutrition interventions at the state and local levels.

The Centers for Disease Control and Prevention (CDC) has been collecting PedNSS data since 1973. Currently, forty-two states, seven Indian Reservations, Puerto Rico and the District of Columbia participate in contributing data to PedNSS. Nationwide, the information is collected from a variety of food assistance, public health and nutrition programs, such as Early Periodic Screening, Diagnosis and Treatment (EPSDT), Title V Maternal and Child Health Program (MCH) and the Head Start Program. The majority of the national data (three quarters) is generated by the Special Supplemental Nutrition Program for Women, Infants and Children (WIC).

West Virginia began participating in PedNSS in December, 1992, using data from a WIC database that is now defunct. After 1994, data originated from WIC’s STORC (Storage, Transfer, Organization, and Retrieval of Casefiles) data system. All information in West Virginia PedNSS originates from measurements taken and information collected at WIC clinics. Since WIC is the sole contributor of PedNSS data in West Virginia, data is limited to low income infants and children up to the age of five.

Data items collected by PedNSS include population demographics and data that reflects health and growth problems in children. Nutrition risk indicators that are collected in PedNSS are short stature (low height for age), underweight (low weight for height), overweight (high weight for height), anemia (low hemoglobin) and low birth weight (<2500 grams). The measurements are interpreted by the computer software version of the National Center for Health Statistics (NCHS)/CDC growth reference.1 Additionally, infant feeding practices data is collected for children under two, to assess the prevalence and duration of breastfeeding. National PedNSS information, West Virginia statewide data, regional data and selected local clinic data are presented in this report. Healthy People 2000 goals are presented as appropriate.

A record in PedNSS is generated for each visit by a child when height, weight or hemoglobin is recorded in the STORC system. A child with more than one visit will have multiple records in the system. On average, each child contributes 2.1 records to PedNSS.2 In West Virginia, data collected by local WIC clinic staff and keyed into the STORC data system is processed quarterly by state staff and forwarded to the CDC, which analyzes the data for the completeness and accuracy required to include that data in the data set. The inclusion of as many records as possible is desirable, as the larger the population, the more indicative the measurement is of a true prevalence rate. For analysis purposes, the CDC requires that a group have at least 100 records containing the data under analysis to meet standards of reliability. In this report, those groups having less than 100 records are indicated with an asterisk, not charted, or otherwise noted in text.

A record that is incomplete for a given data item or contains a biologically implausible value will not be included in the prevalence calculation for that item. In order for a record to be considered complete, it must contain the child’s height and weight measurement, birth weight for children under two years of age, and hemoglobin or hematocrit. In 1996, West Virginia had a relatively high percentage of records (22.1%) rejected for missing height and or weight. The cause was found to be a programming error, which has been corrected. West Virginia also has missing data for hemoglobin or hematocrit (31.3%), but this high percentage is consistent with a high national rate (19.6%). This is because the WIC Program does not require blood work at every visit.

The CDC also will not include those items in records which are identified as biologically implausible values, a value which exceeds edits set by the CDC. For example, if a record contains an extremely high weight value that exceeds the CDC’s edits, then this record’s weight is not used in determining the prevalence of high weight for height. This does not prevent the use of this record’s hemoglobin value (if it is within edits) for determining anemia prevalence. The probability that values which exceed the edit are correct is one in a thousand. Nationally, 6.4% of 1996 PedNSS records contained biologically implausible values. West Virginia’s 1996 PedNSS records contained 1.6% rejected for implausible values. While this is well below the national rate, the CDC has stated a rate above 0.1% indicates some of the values are probably in error. 3

This report encompasses West Virginia PedNSS data from 1992 through 1996. Through 1994, the source of data was an essentially unedited database. Beginning with 1995, data originates from the STORC system, which contains field edits on anthropometric, hematologic and other critical values. Alterations in some of the trends in overall state rates beginning in 1995 may reflect the effect of edits within the STORC system.

Caution is warranted when applying information obtained from PedNSS data to the general population. Nationally, PedNSS data represents participants in a variety of public health programs serving low income individuals. In West Virginia, the PedNSS surveillance group is composed entirely of WIC participants, and the WIC Program has as one of its primary eligibility requirements the existence of nutrition risk factors. Therefore, higher than normal prevalence of low birth weight, anemia, and abnormal values for weight and height are to be expected. Keep in mind that PedNSS data is not collected from identical populations between states or even clinics, and data can be affected by the population’s age, ethnicity or income status. In clinics serving smaller numbers, the addition of just a few visits by children with unusually high or low values can skew the clinic prevalence for a given year. An established trend of rising, falling or consistently high or low values is more revealing.

Each county in West Virginia has at least one WIC clinic. For clarity, rates for clinic data are given by county name rather than WIC clinic name. In those counties having more than one WIC clinic, the data is identified by both county and clinic name. Those counties/clinics which exceed Healthy People 2000 targets are listed. Where no Healthy People 2000 target exists, the top ten percent of counties/clinics having the highest or lowest prevalence of a given risk factor are listed in the body of this report, for the purpose of establishing trend data.

The WIC Program in West Virginia is divided administratively into eight regions, or "local agencies", each of which is responsible for multiple counties.

Demographic variables associated with risk for poor infant or child health include race/ethnicity and age. The specific association these variables have with a given nutrition risk factor will be discussed in the section regarding that indicator. Additionally, Healthy People 2000 establishes special target populations based on age and race/ethnicity. It is important to remember that age and racial or ethnic distribution may vary with geography, to avoid inaccurate comparisons.

The age distribution of the 107,141 records accepted from West Virginia PedNSS in 1996 is compared to the national distribution below. National data includes a portion (7%) which is obtained from children between the ages of 5 and 18. Since the PedNSS data in West Virginia is obtained from WIC, West Virginia data represents children up to five years of age. These age distributions are further consolidated here into groups relative to Healthy People 2000 target populations.

The ethnic distribution of West Virginia’s PedNSS population varies greatly from the national population. West Virginia’s PedNSS population consists primarily of two ethnic groups, but is consistent with West Virginia’s general population demographics. The state and national ethnic distributions in 1996 PedNSS is as follows:

All other racial and ethnic categories were of insufficient numbers to generate reliable statistics (* indicates under 100) or were nonspecific.

To view data related to a specific PedNSS Health Indicator, click on the desired indicator listed below:

Low height-for-age, also referred to as shortness or stunting, is defined as a height-for-age value below the 5th percentile of the NCHS/CDC height reference. It is expected that 5% of a normal population will fall in this low height-for-age category.

Stunting reflects the long term health and nutrition history of a child. On an individual level, shortness can reflect the normal variation of growth within a population. In some children, short stature is related to factors such as lower birth weight or short parental stature. Other contributing factors to growth stunting can be frequent infections and long term poor nutrition.4 Short stature has been associated with low developmental and cognitive test scores, and stunting early in life may lead to reduced physical capacity and endurance in adulthood. 5

On a population level, the CDC finds a strong correlation between an increased prevalence of stunting and poor socio-economic conditions. Black, Hispanic and Asian children are more likely to be short than other children.6 The low income population reflected in PedNSS shows a persistent rate of low height-for-age in excess of five percent, but below the Healthy People 2000 goal of ten percent.

Short stature is discussed as an indicator of growth retardation in Healthy People 2000, Objective 2.4, which establishes special target groups for black and Hispanic children under the age of one, using 10% as the goal. The percentage of children reported in West Virginia as being below the 5th percentile of height-for-age in 1996 is 6.8%. The nation as a whole, while reflecting a higher 1996 rate of 8.0%, still does not show short stature to exceed Healthy People 2000 Goals for the general population.

To view the Trend in Prevalence of Short Stature for the United States and West Virginia, click here.

Special target populations mentioned in Healthy People 2000 (Objective 2.4) and present in West Virginia PedNSS data in sufficient numbers to generate reliable detail data are low-income black children under the age of one, having a goal of less than 10% prevalence. The 1996 prevalence of stunting for low-income children in West Virginia by ethnicity and age is:

The change in prevalence at 2 - 4 years of age may be due to a disparity between sample populations used in the NCHS/CDC reference. The sample for children under 2 years of age was composed of "taller than average" children, thus the expected rate of shortness for children in this age group is expected to be higher than 5%.7

While the state average for shortness meets the Healthy People 2000 goal of prevalence below 10%, some individual WIC clinic sites exceeded the target in 1996. A history of low height-for-age prevalence in these counties shows both developing trends in some counties, and others simply an annual fluctuation in the rate. The Kanawha County Cedar Grove clinic had insufficient numbers to report 1994 data, and was not in operation in 1995.

Short Stature Trends for High Ranking Counties, 1996
(Numbers displayed in percentages)

Ranking in 1996	Location	1992	1993	1994	1995	1996
1	Calhoun	7.6	8.4	8.9	12.5	12.0
2	Summers	10.7	11.0	9.3	7.5	11.8
3	Kanawha - Cedar Grove	8.2	8.4	*	*	11.3
4	Taylor	6.4	8.3	9.3	8.6	10.8
5	Ohio	7.3	9.6	8.6	8.9	10.6
6	Roane	7.9	8.5	6.7	9.5	10.2

To view data related to a specific PedNSS Health Indicator, click on the desired indicator listed below:

Children with a weight-for-height value of less than the 5th percentile of NCHS/CDC reference population are considered to be underweight. Low weight-for-height, or thinness, is often associated with recent severe disease, but can also be the result of normal individual variation in a population. In developing countries, thinness indicates acute malnutrition (either the result of insufficient food, infectious or diarrheal disease, or both.) The prevalence of thinness in a population is usually low except during disaster conditions. 8 Low weight-for-height is generally not a problem in the United States, and no Healthy People 2000 goals have been established. The 1996 prevalence of low weight-for-height in West Virginia is 2.4%, compared to the national prevalence of 2.6%.

Trends in prevalence of underweight children shows the rate virtually unchanged since 1994 in West Virginia, with minor decreases reflected in the national rate.

To view the Trend in Prevalence of Underweight Children for the United States and West Virginia, click here.

The rates of underweight in West Virginia are highest in the 1 year to 23 months age range, but show little variation in ethnic prevalence. Interestingly, the national rate is highest in the under 1 year category in all ethnic groups except Asian. The national rate for infants under one year is: White - 3.0, Black - 4.6, Hispanic - 2.6. Insufficient numbers of other ethnic groups in age categories in West Virginia prevent analysis.

To view 1996 Age and Ethnicity data related to Underweight Children in West Virginia, click here.

For the (top ten percent of) clinics with the highest prevalence of underweight in 1996, a history for each clinic’s annual rate in PedNSS is shown below. Some clinics have only recent high prevalence (relatively), others fluctuated, and some display persistent relatively high rates.

To view data related to a specific PedNSS Health Indicator, click on the desired indicator listed below:

High weight-for-height is defined as those children in the 95th percentile and above of the NCHS/CDC reference. The presence of overweight is used as a proxy for obesity, and in the pediatric population has become an important public health issue. One-third to one-half of those children above the 95th percentile will become obese adults. Obesity is associated with long term health consequences such as heart disease, hypertension and diabetes.9 Nationwide, the percentage of children found above the 95th percentile of weight for height continues to climb, with a 1996 prevalence of 10.1% . West Virginia continues to be below the nation’s rate in 1996, having a prevalence of 8.7% high weight for height.

To view the Trend in Prevalence of Overweight Children for the United States and West Virginia, click here.

The highest prevalence of overweight is found in the 1 year category for both black and white groups. Insufficient numbers for Hispanics in under one year and 1 year categories prevent analysis.

To view 1996 Age and Ethnicity data related to Overweight Children in West Virginia, click here.

The (top ten percent of ) clinic sites with the highest prevalence of overweight in 1996 are given below, along with the trend data for those years PedNSS covers. Some clinics show a steady progression in prevalence, but even those which fluctuate seem to indicate a strong presence in the local PedNSS population of children at risk for adult obesity.

Overweight Trends in High Ranking Counties, 1996
(Numbers displayed in percentages)

Ranking in 1996	Location	1992	1993	1994	1995	1996
1	Logan - Amherstdale/Crites	*	*	14.0	14.6	16.1
2	Logan - Logan Clinic	9.7	9.9	11.1	13.9	15.5
3	Doddridge	6.3	7.3	9.7	15.5	13.9
4	Calhoun	9.9	10.5	12.8	16.4	13.4
5	Wyoming	16.1	13.9	13.2	13.2	13.2
6	McDowell - Welch	7.6	7.7	8.5	12.1	12.6

The rates for Logan clinic (number 2) are for all of Logan County up until 1994, when the Amherstdale - Crites clinic began reporting separately.

To view data related to a specific PedNSS Health Indicator, click on the desired indicator listed below:

Low birth weight is defined as a birth weight under 2500 grams (5 pounds, 8 ounces). Low birth weight occurs when an infant is born at less than 37 weeks of age, or when there is intrauterine growth retardation, or as a result of both conditions. Low birth weight reflects maternal health status during pregnancy, and is a strong predictor of growth in early childhood. Premature low birth weight infants have a higher mortality rate, but full-term infants small for their gestational age exhibit slower physical growth, possibly slower mental development, and are more likely to have congenital abnormalities.10

Healthy People 2000, Objective 14.5 seeks to reduce low birth weight to no more than 5% of live births, and establishes a special population target for blacks at 9%.

To view the Trend in Prevalence of Low Birth Weight for the United States and West Virginia, click here.

In contrast to the gradual downward trend of the national rate, an increasing prevalence of low birth weight infants is found in the West Virginia PedNSS population. The increasing rate may be due to the enrollment of less healthy participants in WIC. However, even the general population statistics for West Virginia show a steady annual increase, from 7.2% in 1992 to 8.0% in 1996.11

West Virginia’s PedNSS data, when broken out by ethnicity, shows that black infants exhibit a much higher incidence of low birth weight than white infants, which is consistent with the national rate, having a 1996 prevalence of 13.6% for blacks, and 8.2% for whites. The Hispanic rate is much lower than the national rate of 7.0, but the small sample size in West Virginia (an average of 232 records each year) does not lend certainty to the figure.

To view 1996 Age and Ethnicity data related to low birth weight in West Virginia, click here.

The counties and clinics exceeding 10% prevalence of low birth weight in 1996 are accompanied by their history, to illustrate potential trends.

Low Birth Weight Trends for High Ranking Counties, 1996
(Numbers displayed in percentages)

Ranking in 1996	Location	1992	1993	1994	1995	1996
1	Jefferson	6.5	8.3	8.5	12.2	14.0
2	Mercer	9.4	8.5	9.1	10.2	11.6
3	Logan-Amherstdale/Crites	*	*	10.1	8.0	11.2
4	Raleigh	9.1	9.9	9.8	11.4	11.2
5	Ritchie	5.9	6.4	6.7	7.5	11.0
6	Wayne - Fort Gay	7.3	7.0	10.3	9.7	11.0
7	Brooke	8.1	7.4	7.2	11.0	10.9
8	Hancock	13.5	11.0	9.4	11.6	10.9
9	Kanawha - Charleston	9.1	9.3	9.7	11.0	10.9
10	McDowell - Welch	8.8	9.5	7.9	10.0	10.8
11	Monongalia	9.8	9.5	10.3	11.5	10.8
12	Lincoln	10.8	10.8	8.4	9.5	10.7
13	Pocahontas	11.0	7.4	9.3	13.1	10.7
14	Cabell	9.5	9.4	9.5	10.3	10.5
15	Roane	7.4	7.1	6.4	8.2	10.5
16	Preston	7.2	7.4	9.7	10.2	10.2
17	Ohio	7.6	7.3	7.1	8.3	10.1

Kanawha County data identified as Charleston clinic (number 9) reflects all of Kanawha County (with the exception of Cedar Grove clinic data for 1992, 1993 and 1996.) Only 1996 data is specific to Charleston clinic. Data was combined for past years from a number of now closed clinics in Kanawha County to create a historical frame of reference for data, as Charleston clinic contributed 88.7% of Kanawha County data in 1996, its first full reporting year.

To view data related to a specific PedNSS Health Indicator, click on the desired indicator listed below:

PedNSS accepts either hemoglobin or hematocrit levels as an indicator of anemia. While not all anemia is due to iron deficiency, it is the most common cause of anemia throughout the world. 12 The CDC has established criteria for anemia based on the Second National Health and Nutrition Examination Survey (NHANES II) data. For children under twenty-four months of age, the fifth percentile cutoff is hemoglobin measurement under 11.0 grams per deciliter, and for children aged two to five years, 11.2 grams per deciliter. PedNSS adjusts hematology values for altitude.13

Iron deficiency anemia impairs mental and psychomotor development in infants and children. Although iron deficiency can be reversed with treatment, the reversibility of impairments are not yet clearly understood. In addition, iron deficiency increases a child’s susceptibility to lead toxicity. Iron deficiency anemia is seen most commonly in children six months to three years of age. Those at highest risk are low birth weight infants after two months of age, breastfed infants who receive no supplemental iron after four months of age, and formula fed infants who are not consuming iron-fortified formula. 14

Since hemoglobin is overwhelmingly reported in West Virginia (99.5%), this indicator of anemia was used for the national rate for comparison purposes in this report, rather than the combined "Low Hematology" value which uses either or both hemoglobin and hematocrit.

To view 1996 Anemia by Age data for the United States and West Virginia, click here.

Healthy People 2000, Objective 2.10, is to reduce iron deficiency anemia to less than three percent among children aged 1 through 4 years of age, and specifically targets two age groups. Special population group targets are to reduce iron deficiency among low-income children age 1-2 years to 10%, and children 3-4 years to 5%. Healthy People 2000 further defines iron deficiency as abnormal results from 2 or more of the following tests: Mean corpuscular volume, erythrocyte protoporphyrin, or transferrin saturation.15 While WIC clinics do not perform these tests, the CDC continues to recommend hemoglobin screening for anemia in populations where the risk of anemia due to iron deficiency is high.16

The overall trend in low hemoglobin measurements in PedNSS is one of decrease. West Virginia’s rate is approximately half of the nation’s rate.

To view the Trend in Prevalence of Low Hemoglobin for the United States and West Virginia, click here.

The age categories in PedNSS useful for comparison to Healthy People 2000 goals for special target populations gives an indicator of West Virginia’s progress towards lower prevalence of anemia. In the one-year and two-to-four year age group, a significant proportional decrease occurred in 1993 and 1994, but has not been repeated. West Virginia data for the age category of <1 year only reflects measurements for infants older than six months of age. WIC does not test hemoglobin in children under six months of age, and does not universally test hemoglobin between six months and one year of age.

Black children in West Virginia have had consistently higher prevalence of anemia than other ethnic or racial groups, as is the case nationally as well.

To view 1996 Ethnicity data related to Low Hemoglobin for the United States and West Virginia, click here.

Trend data for those clinics exceeding 10% prevalence of anemia in 1996 show that most have consistently higher anemia rates than the rest of the state over time.

Anemia Trends for High Ranking Counties, 1996
(Numbers displayed in percentages)

Ranking in 1996	Location	1992	1993	1994	1995	1996
1	Morgan	18.0	13.0	14.6	21.0	23.3
2	Greenbrier	12.4	8.1	12.7	22.6	20.6
3	Gilmer	16.1	13.5	15.7	15.8	18.7
4	Berkeley	18.8	15.1	11.5	12.7	17.3
5	Ohio	12.9	6.8	8.8	9.4	16.5
6	Hampshire	31.8	21.8	19.7	23.2	16.3
7	Wirt	25.0	15.4	17.6	14.1	16.0
8	Wyoming	13.5	8.0	5.1	12.2	15.4
9	Pocahontas	8.1	10.5	10.6	13.1	14.7
10	Wetzel/Tyler	9.8	11.2	9.8	15.4	14.5
11	Clay	10.8	6.2	5.7	8.5	14.3
12	Wood	16.5	12.4	14.0	17.7	14.2
13	Pleasants	31.0	21.7	7.3	17.6	13.4
14	Jackson	17.4	10.1	11.1	11.2	13.0
15	Monroe	4.7	4.5	2.4	9.5	12.8
16	Wayne - Fort Gay	14.8	7.6	7.8	7.1	12.2
17	Kanawha-Cedar Grove	12.1	10.6	*	*	12.1
18	Mason	21.3	12.2	9.8	15.2	12.1
19	Calhoun	30.4	21.8	23.8	16.4	11.2
20	Roane	17.0	9.1	10.3	14.7	11.0
21	Braxton	13.7	10.3	11.1	12.7	10.9
22	Ritchie	20.3	17.5	11.4	10.2	10.7
23	Logan - Logan Clinic	5.6	3.4	2.3	28.4	10.5
24	Mineral	24.1	14.9	12.1	16.0	10.5
25	Kanawha - Charleston	13.6	10.3	10.1	10.3	10.4
26	Hancock	21.1	14.7	10.4	15.3	10.4

The Kanawha County data identified as Charleston clinic (number 25) reflects all of Kanawha County (with the exception of Cedar Grove clinic data for 1992, 1993 and 1996.) Only 1996 data is specific to Charleston clinic. Data was combined for past years from a number of now closed clinics to create a historical frame of reference for data, as the Charleston clinic contributed 88.7% of Kanawha County data in 1996, its first full reporting year. Logan clinic (number 23) reflects all of Logan County data in the years 1992 and 1993.

To view data related to a specific PedNSS Health Indicator, click on the desired indicator listed below:

One of the major initiatives of the WIC program is to increase initiation of breastfeeding and to prolong the duration of breastfeeding. Healthy People 2000, Objective 2.11, sets a goal to increase to at least 75% the proportion of mothers who breastfeed their babies in the early postpartum period (initiation), and to at least 50% the proportion who continue breastfeeding until their babies are 5 to 6 months old (duration). Special target populations are low income mothers, and black, Hispanic and American Indian/Alaskan natives.

1996 PedNSS data on breastfeeding rates for infants six to eight months old shows that both the nation and West Virginia fall short of these goals, with West Virginia closer to the national average in initiation, but falling behind in duration. Similar to the national statistics, West Virginia shows the greatest drop off in rates between one week and one month of age.

To view 1996 Breastfeeding Duration Rates data for the United States and West Virginia, click here.

West Virginia’s breastfeeding initiation rates by race/ethnicity compared to the nation show that among white women the initiation rate is lower, but among black women and Hispanics, slightly exceed the national average.

To view 1996 Ethnicity data related to Breastfeeding Initiation for the United States and West Virginia, click here.

Duration of breastfeeding in West Virginia, when compared among ethnic groups, shows a similar proportional decrease among the groups. Caution is warranted in use of these figures, as with the exception of the white racial/ethnic group, numbers are under 100 for most age ranges.

To view 1996 Ethnicity data related to Breastfeeding Duration for West Virginia, click here.

To view data related to a specific PedNSS Health Indicator, click on the desired indicator listed below:

Endnotes
1.) CDC, Division of Nutrition. Enhanced Pediatric Nutrition Surveillance System User Manual. US Department of Health and Human Services, Public Health Service, 1994, pg. 1-6

2.) Ray Yip, et al., "Pediatric Nutrition Surveillance System - United States, 1980 - 1991", Morbidity and Mortality Weekly Report, Vol. 41, SS-7, November 27, 1992, pg. 2

3.) CDC, Division of Nutrition. Enhanced Pediatric Nutrition Surveillance System User Manual. US Department of Health and Human Services, Public Health Service, 1994, pg. 8-1

4.) Ray Yip, et al., "Pediatric Nutrition Surveillance System - United States, 1980 - 1991", Morbidity and Mortality Weekly Report, Vol. 41, SS-7, November 27, 1992, pg. 9

5.) Institute of Medicine, WIC Nutrition Risk Criteria: A Scientific Assessment, National Academy Press, Washington, D.C., 1996. pg. 106-107

6.) CDC, Division of Nutrition. Enhanced Pediatric Nutrition Surveillance System User Manual. US Department of Health and Human Services, Public Health Service, 1994, pg. 1-7

7.) Ray Yip, et al., "Pediatric Nutrition Surveillance System - United States, 1980 - 1991", Morbidity and Mortality Weekly Report, Vol. 41, SS-7, November 27, 1992, pg. 9

8.) I. Pravanta, "Nutrition" in L. Wilcox, J. Marks, eds. From Data to Action: CDC’s Public Health Surveillance for Women, Infants and Children, Centers for Disease Control and Prevention, U.S. Public Health Service, Atlanta, Ga., 1995. pg. 325

9.)CDC, Division of Nutrition. Enhanced Pediatric Nutrition Surveillance System User Manual. US Department of Health and Human Services, Public Health Service, 1994, pg. 1-7

10.) Institute of Medicine, WIC Nutrition Risk Criteria: A Scientific Assessment, National Academy Press, Washington, D.C., 1996. pg. 98-102

11.)WV Vital Statistics West Virginia Bureau for Public Health, Office of Epidemiology and Health Promotion. 1992 - 1995 and Provisional Statistics, 1996.

12.) CDC, Division of Nutrition. Enhanced Pediatric Nutrition Surveillance System User Manual. US Department of Health and Human Services, Public Health Service, 1994, pg. 1-9

15.) U.S. Public Health Service. Healthy People 2000; National Health Promotion and Disease Prevention Objectives, Washington, D.C.: 1991; DHHS publication no. (PHS)91-50212. pg. 122

16.) Centers for Disease Control and Prevention. Recommendations to Prevent and Control Iron Deficiency in the United States. MMWR 1998;47(no. RR-3): April 3, 1998, pg. 21

1996 Pediatric Nutrition Surveillance System Summary Report