The Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) provides specific supplemental foods, nutrition education, and social service and health care referrals to low-income pregnant, breastfeeding, and postpartum women, infants, and children up to age 5 years who are at nutrition risk. The WIC program is based on the premise that many low-income individuals are at risk of poor nutrition and health outcomes because of insufficient nutrition during the critical growth and development periods of pregnancy, infancy, and early childhood. The WIC program is a supplemental food and nutrition program to help meet the special needs of low-income women, infants, and children during these periods. Income below 185 percent of the poverty level is one of the standards of eligibility for the WIC program. A summary of WIC program components, services, and anticipated outcomes is provided in Figure S-1.

All WIC program participants must be determined to be at nutrition risk on the basis of a medical or nutrition assessment by a physician, nutritionist, dietitian, nurse, or some other competent professional authority. Using nutrition risk as a requirement for certification is a unique feature of the WIC program. Public Law 94-105 broadly defines nutrition risk as ''(a) detrimental or abnormal nutritional conditions detectable by biochemical or anthropometric measures, (b) other documented nutritionally related medical conditions, (c) dietary deficiencies that impair or endanger health, or (d) conditions that predispose persons to inadequate nutritional patterns or nutritionally related medical conditions."

Nutrition risk criteria also provide the basis for a seven-level priority system for eligible women, infants, and children. If a local WIC agency reaches its maximum caseload given its level of funding, the WIC priority system is used to maintain a waiting list of eligible applicants. As program openings become available, they can be filled from the waiting list. In general, priority is given to anthropometric, hematologic, and clinical evidence of medically based nutrition risks over dietary-based nutrition risks; to pregnant and breastfeeding women and infants over children; and to children over postpartum women.

In the summer of 1993, the Food and Nutrition Service of the U.S. Department of Agriculture (now the Food and Consumer Service [FCS]) requested that the Food and Nutrition Board (FNB) of the Institute of Medicine conduct a comprehensive review of the scientific basis for the nutrition risk criteria used in the WIC program. In October 1993, the FNB established the Committee on the Scientific Evaluation of WIC Nutrition Risk Criteria. The committee was charged with conducting a study that included the following tasks:

• Performing a critical review of the literature surrounding the various nutrition risk criteria used by the WIC program.
• Developing scientific consensus (where possible) regarding the nutrition risk criteria used by the WIC program, taking into account the preventive nature of the program.
• Identifying specific segments of the WIC population at risk for each criterion.
• Identifying gaps in the scientific knowledge base for the current nutrition risk criteria used by the WIC program.
• Formulating recommendations regarding appropriate criteria and, if applicable, recommendations of numerical values for determining who is at risk for each criterion.
• Identifying critical areas for future research.
• Identifying the practicality of consensus recommendations for nutrition risk criteria for the variety of WIC program delivery settings.

Over the course of the study, the committee met five times, conducted two public meetings, participated in many conference calls, and made site visits to local WIC program clinics. The committee began its deliberations by reviewing the WIC program. Since the federal WIC program does not have a uniform set of nutrition risk criteria (where a risk criterion is defined as a risk indicator and its cutoff point), the committee obtained a list of nutrition risk criteria used by WIC state agencies in 1992. It categorized these into (1) anthropometric, (2) biochemical and other medical, (3) dietary, and (4) predisposing risks to reflect the definition of nutrition risk in federal WIC program regulations. Using terms

based on this list, the committee conducted bibliographic searches of the scientific literature and compiled and critically reviewed research findings. In reviewing each risk criterion, the committee examined three issues: (1) Is there scientific evidence that the criterion serves as an indicator of nutrition and health risk? (2) Does the criterion serve as an indicator of nutrition and health benefit from participation in the WIC program? (3) What cutoff value, if any, is scientifically justified? The relationship between poverty and nutrition risk is also discussed in the report because it is a separate standard for WIC program eligibility. Poverty is not a WIC nutrition risk criterion and was not reviewed as one.

The two public meetings gathered information from WIC program administrators, staff, and participants as well as from researchers in the fields related to the risk criteria under study. The public meetings and the visits to WIC clinics provided valuable information about the use of nutrition risk criteria in the WIC setting.

Chapter 1 of the committee's report on its study describes the structure and function of the WIC program and provides an overview of the committee's task. Chapter 2 reviews linkages between low-income and risk of inadequate nutrition. Chapter 3 discusses the principles of nutrition risk assessment that guided the committee in conducting its review and provides the framework used to develop the committee's recommendations. Chapters 4 through 7 cover the nutrition risk criteria used by the WIC program: anthropometric, biochemical and medical, dietary, and predisposing risks. Chapter 8 provides conclusions and recommendations regarding nutrition risk criteria and recommendations for research and action.

The concept of nutrition risk assessment is integral to the design and operation of the WIC program. Nutrition risk is a criterion for program eligibility, and nutrition risk criteria are used to assign a priority level to women, infants, and children. By serving those at the highest priority levels first, the WIC priority system is used to allocate limited program resources among eligible individuals. In addition, the nutrition risk assessments are used to tailor the WIC intervention and, in some cases, to monitor the health and nutrition status of program participants.

This report is a scientific assessment of the WIC nutrition risk criteria as they are currently used to establish WIC eligibility and the priority of the WIC eligible individuals. This scientific assessment is the basis for the final chapter's general conclusions, recommendations for specific nutrition risk criteria, and recommendations for future research and action.

The framework that was used in the scientific assessment conducted for this report has two key features. The first is the exposition and utilization of the concept of potential to benefit from the delivery of interventions and services provided by the WIC program. The second is the explicit consideration of the

concepts of yield of risk, yield of benefit, and sensitivity of the nutrition risk criteria used by the WIC program, which are described below.

A nutrition risk assessment is used to determine eligibility for participation in the WIC program. Nutrition risk assessment uses a risk criterion; a risk criterion is defined by a risk indicator and a cutoff point. A risk indicator is any measurable characteristic or circumstance that is associated with an increased likelihood of poor outcomes, such as poor nutrition status, poor health, or death. In some cases (e.g., low hemoglobin level), a risk indicator could also be considered to be an outcome. The cutoff point may be the presence or absence of the condition (e.g., a diagnosis of diabetes mellitus) or a value chosen from many possibilities for a specified population (e.g., a hemoglobin value of 11.0 gm/dl for women in their first trimester of pregnancy).

The committee agreed that nutrition risk criteria used in the WIC program should serve both as indicators of nutrition and health risk and as indicators of nutrition and health benefit. Indicators of nutrition and health risk should select those who have the greatest need for the services provided by the WIC program because they are either more unhealthy or poorly nourished at the time of assessment or are at future risk of ill health, overnutrition, or undernutrition. Indicators of nutrition and health benefit are those that improve the efficacy of participation in the WIC program by selecting those potential participants most likely to benefit from participation over those less likely to benefit from participation.

Once a risk indicator is chosen as a predictor of benefit, a cutoff point for the indicator is set as the level below or above which individuals are eligible for participation in the WIC program. Four important concepts in selecting cutoffs for the nutrition risk indicators used by the WIC program are yield of benefit, yield of risk, sensitivity, and efficacy of WIC interventions:

1. Yield of benefit: the percentage of those truly at risk who will actually experience nutrition and health benefit among all those, with or without the risk, who are selected by a nutrition risk indicator and its cutoff point. Individuals are considered truly at risk if they would have a bad outcome without an intervention.
2. Yield of risk: the percentage of those truly at risk who are identified at risk.
3. Sensitivity: the percentage identified of all those who could benefit from the WIC program.
4. Efficacy of WIC interventions: the percentage of individuals selected for the WIC program whose bad outcomes will be prevented or reduced.

Yield of benefit can be high only if the yield of risk is high and the WIC program can prevent or reduce bad outcomes for those at risk. A perfect yield of risk occurs at the cutoff point at which all those selected for participation in the WIC program are truly at risk. A perfect yield of risk, however, implies that many who could benefit are not selected. Identification of all who could benefit is called perfect sensitivity. In general, there is a trade-off between yield of risk and sensitivity, and it is usually impossible to achieve both maximum yield of risk (serving only those truly at risk) and perfect sensitivity (identification of all those at true risk). Overall yield of benefit is affected both by the yield of risk and the efficacy of the interventions, since it is the product of the yield of risk and the efficacy.

These concepts of yield of risk, yield of benefit, and sensitivity, in conjunction with the concepts of indicators of risk and indicators of benefit, have implications that underlie both the assessments of the nutrition risk criteria used by the WIC program and the development of the report's conclusions and recommendations. Ideally, risk indicators and cutoff points should be chosen such that the highest proportion of those who are truly at risk can be identified and the highest proportion of those identified can benefit from WIC program participation. With limited program resources, cutoff points should be set with less than perfect sensitivity to increase yield, recognizing that as cutoff points become more restrictive, some individuals who could benefit from WIC services will not be served.

The following decision process underlies the committee's recommendations. This process could be used to review other risk criteria that the WIC program may be asked to approve in the future.

• For nutrition risk criteria for which there is good evidence of both nutrition and health risk and benefit from the WIC program, the committee recommends use of these criteria by all state WIC programs.
• For nutrition risk criteria for which the risk indicator is a predictor of both nutrition and health risk and benefit from the WIC program but for which cutoffs have been set so that many individuals selected are not truly at risk, the committee recommends using the risk indicator with more stringent cutoff values.
• For risk criteria for which there is strong evidence of nutrition and health risk but uncertain evidence of benefit, the committee recommends using the nutrition risk criteria and conducting further research on the benefit from the WIC program.
• For risk criteria for which there is good evidence of nutrition and health risk and benefit from the WIC program, but poor ability to identify those at risk with current methods, the committee recommends that action be taken to develop better assessment tools. Pending this result, the committee recommends

• using the best available methods to identify the risk, using scientifically justifiable cutoff values.
• For risk criteria for which there is strong evidence of nutrition and health risk but no direct or indirect evidence of benefit, either theoretical or empirical, the committee recommends discontinuing use of these risk criteria.
• For risk criteria with weak evidence of risk or benefit, the committee recommends discontinuing use of these criteria.

The WIC program is a broad-based and comprehensive food and nutrition program with three main components: (1) supplemental foods, (2) nutrition education, and (3) referrals to health care and social service providers. Thus, evidence of benefit from the WIC program, either theoretical or empirical, could be from any of the three program components. In making its recommendations for each nutrition risk criterion, however, the committee decided that evidence of benefit from the WIC program should reflect the ability of an individual with that risk to benefit (avert bad outcomes) from the WIC food package or, in some cases, from nutrition education.

Benefit from only the referral services of the WIC program was not considered sufficient to justify the use of a nutrition risk criterion. Three main reasons for this decision follow: (1) the provision of supplemental foods and nutrition education account for nearly all the WIC program costs; (2) it is difficult to justify the provision of a monthly food package worth approximately $30 per WIC participant unless there is evidence that the individual can benefit from the food package or the nutrition education that accompanies the provision of food; and (3) the WIC program is designed to be only an adjunct to good health care and is not itself a comprehensive health program. Nonetheless, the committee respects the comprehensive nature of the program.

The committee reached seven general conclusions about the WIC nutrition risk criteria and priority system:

• A body of scientific evidence supports a majority of the nutrition risk criteria used by the WIC program. For some of the risk criteria, however, there are serious gaps in the evidence.
• Nutrition risk criteria used by many states have a high sensitivity and low yield of benefit. This is because the prevalence of many of the risk conditions is low and the cutoffs used are generous, resulting in both the selection of many of those who have the risk condition (high sensitivity) and the selection of many individuals who do not have the risk condition (low yield of risk, which results in low yield of benefit).

• Use of generous cutoff points or loosely defined conditions in categories designated by federal regulation to receive high priority for eligibility may result in denial of services to individuals who are actually at higher nutrition risk. When resources are limited, individuals in lower priority categories may not be served even if their true risk is very high, while those in high priority categories must be served. Very generous cutoff points produce a low yield of benefit without any increase in sensitivity (serving more of those truly at risk). Loosely defined risk conditions are those that encompass a broad range of medical problems with varying degrees of nutrition risk or potential to benefit from WIC participation. Such loosely defined nutrition risk criteria include endocrine disorders, renal disease, chronic and recurrent infections, food allergies, and genetic and congenital disorders.
• There is some inconsistency between the WIC program's goals, design, and implementation. The goal of the WIC program is one of primary prevention—to prevent the occurrence of health problems. Through the use of nutrition risk criteria, the WIC priority system is designed in principle to be a secondary/tertiary prevention program to reduce or cure identified risk. However, through the use of generous cutoff points, loosely defined risk conditions, and a priority system that places pregnant women and infants at the highest priorities, in general, the WIC program operates as a primary prevention program for pregnant women and infants and a secondary and tertiary prevention program for children and postpartum women.
• The WIC priority system should be reexamined. Many individuals now classified in low priority categories have more potential to benefit from WIC services than some individuals placed in higher priority categories. For example, a child of a mentally retarded parent (currently priority VII) or an anemic child age 3 years with a very low hemoglobin (currently priority III) may have a greater potential to benefit than an infant classified as anemic (currently priority I) by a criterion with a too generous cutoff point.
• It is important that the WIC program reevaluate the criteria in use every 5 to 10 years and change cutoffs and incorporate new criteria as necessary. This is because the yield of risk of a criterion increases as the prevalence of the risk in the population increases, and it decreases as the prevalence of the risk in the population decreases. For example, the yield of risk of the nutrition risk criterion for poor growth has decreased over time as the prevalences of wasting and stunting have declined. The addition of homelessness as a nutrition risk criterion by the WIC program reflects, in part, increases in the prevalence of homelessness.
• There is a need to identify or develop additional nutrition risk criteria that select those individuals who are at risk of developing specific health and nutrition problems if they do not receive WIC benefits. Since the WIC program is believed to be a major contributor to the decline in the prevalence of health and nutrition problems (for example, iron deficiency anemia), it is important to

• identify practical indicators of the risk of developing the problem so that the WIC program can maintain its preventive function. Dietary risk criteria or predisposing risk criteria may do this, but data are limited. Setting high cutoff points for anemia or poor growth does not effectively identify those at risk of developing the problem.

In addition, the committee emphasizes the importance of the systematic collection of data about the prevalence of individuals meeting specific WIC nutrition risk criteria.

Table S-1 summarizes the committee's recommendations for use of nutrition risk criteria, cutoff values, and the segments of the population to which they apply. For greater specificity, the name of the criterion used occasionally differs from that used by the WIC program. The recommendations are intended to apply to all states. Exceptions may be made if the meaning of the criterion in a particular context is different or the condition (e.g., pica) is common in one state and uncommon in another. Brief supplementary information about these recommendations follows for each of the categories of nutrition risk criteria. The full report provides the basis for each recommendation.

Anthropometric risk criteria are used in the WIC program to assess individuals for nutrition risk and to monitor their nutrition status or their response to WIC program interventions over time. The committee's review indicated that most of the WIC anthropometric risk indicators are predictors both of nutrition and health risks and of benefit from participation in the WIC program. The cutoff points used for anthropometric risk indicators among WIC programs vary substantially, however, with resulting effects on yield. Therefore, the committee recommends that cutoff points for anthropometric measures be limited to those that are scientifically justified. It further points out that there is no obvious justification for the use of symmetric cutoff points (for example, at the 5^th and 95^th percentiles).

Risk criteria for which there was very little evidence of nutrition risk or benefit from WIC participation include maternal short stature, abnormal postpartum weight change, and infants large for gestational age. Therefore, the committee recommends discontinuing use of these nutrition risk criteria.

In general, the biochemical and other medical risk criteria predict nutrition and health risk, with varying degrees of benefit. The most common concern of the committee was the lack of scientific justification for the generous cutoff points for biochemical and certain other medical risk criteria currently used by many state WIC agencies.

Of the biochemical and other medical risk criteria, anemia is used most frequently in the WIC program to establish the eligibility of women, infants, and children to participate in the program. Cutoff values for anemia vary substantially among state WIC agencies, with little or no scientific justification for variation from standard definitions. The committee recommends that anemia continue to be used as a risk criterion in the WIC program but discourages the use of high cutoff points because of the resulting low yield from increased iron intake. That is, the high cutoff values for anemia used by many state WIC programs result in the inclusion of many who do not have and are not at risk of anemia and, thus, are unlikely to benefit from provision of WIC supplemental food.

Many biochemical and other medical nutrition risks are documented as the result of diagnosis by a medical care provider of an existing condition that affects nutritional needs or may be improved by dietary management. These diagnosed conditions are reported to WIC program staff. The committee recommends that most of these nutrition risk criteria continue to be used in the WIC program, using cutoff points that generally are documentation or diagnosis of the disease or disorder.

Maternal cigarette, alcohol, and illegal drug use among pregnant and lactating women pose significant health risks but uncertain benefit from participation in the WIC program. On an interim basis, the committee recommends that these criteria be used in the WIC program, with a cutoff of ''any use."¹

Risk criteria for which there was risk and benefit only under specific conditions included long-term drug-nutrient interactions and chronic and recurrent infections. The committee feels that these criteria were too vague to be useful in their current form. It recommends that a listing of drugs for which there are clear drug-nutrient interactions or potential for misuse be developed. The use of other medications would not be associated with nutrition risk or benefit, and thus their use would not provide a basis for eligibility. For chronic and recurrent infections, evidence of risk and benefit was available only for

certain chronic infections for which there were documented nutrition deficits, and the committee recommends that states should clearly define "chronic" or "recurrent" in determining cutoff points for these indicators.

Risk criteria for which there was very limited evidence of nutrition risk or benefit from participation in the WIC program included food intolerance other than lactose intolerance, high age at conception, previous placental abnormalities, history of postterm delivery, high parity, preeclampsia and eclampsia, and prematurity as a risk criterion for children ages 1 to 5 years. The committee recommends that these nutrition risk criteria no longer be used in the WIC program.

Three major categories of dietary risk criteria are reviewed: inappropriate dietary patterns, inadequate diets, and food insecurity. Risk criteria classified as inappropriate dietary patterns are listed in Table S-1. The committee found that there are clear health and nutrition risks associated with selected inappropriate dietary patterns and that the potential to benefit from participation in the WIC program is high. For women and for children at least 2 years of age, failure to meet Dietary Guidelines for Americans is a dietary risk criterion that receives increased attention in this report.

As long as the food provided by the supplemental food package is eaten, the WIC program is likely to improve the diets of those WIC participants with inadequate diets. In the WIC setting, however, diet recalls and food frequency questionnaires that compare estimated nutrient intake with Recommended Dietary Allowances have poor ability to ascertain who actually has inadequate diets. Thus, even though the WIC program is likely to improve dietary intake, the committee recommends discontinuing use of inadequate diets as a nutrition risk criterion because it has a very low yield. Nonetheless, diet recalls or food frequency questionnaires are useful in the WIC program for identifying foods commonly consumed and providing a starting point for nutrition education.

Food insecurity is defined as the lack of predictable, sustainable access in socially acceptable ways to enough food of adequate quality to sustain health. Although this risk criterion is just beginning to be used by state WIC agencies, and there is limited evidence to evaluate causal links to nutrition and health risk, the committee believes that there is a fundamental value to addressing the risk to health and nutrition related to a lack of access to food. The benefit of participation in the WIC program for those at risk of food insecurity is high. Therefore, the committee recommends use of food insecurity as a nutrition risk criterion in the WIC program. At present, however, there is insufficient scientific evidence on which to select a cutoff point that would identify those most likely to benefit from the WIC program.

Currently, predisposing nutrition risk criteria receive a low priority within the WIC program. The use of predisposing nutrition risk criteria warrants additional attention. If an individual has a predisposing risk but no other risk, he or she will be placed in a priority category that is usually unserved by the WIC program. This may limit the WIC program's ability to serve as a preventive program. Additional attention to the predisposing nutrition risk criteria is warranted because (1) they have a high yield for risk and a high, but as yet unknown, potential for benefit from WIC services and (2) the prevalence of some of these factors (e.g., homelessness) is increasing, thus increasing the overall yield of these criteria.

The committee supports the use of most of the predisposing risk criteria that have been used in the WIC program (see Table S-1).

The committee recommends that a diagnosis of depression be added as a predisposing risk criterion for women, and that diagnosed maternal depression be added as a predisposing risk criterion for infants and children. Because of the lack of evidence that nutrition will benefit those exposed to passive smoking, the committee recommends that this risk criterion no longer be used in the WIC program.

Regarding the nutrition risk criteria reviewed in the report, the committee recommends the following areas for future research:

• Develop anthropometric standards (including weight change velocity) for pregnant and lactating women, including adolescents. These standards should be suitable to assess the likelihood that these women would benefit from nutrition intervention and to achieve improved reproductive outcomes.
• Evaluate whether the use of a combination of criteria (e.g., an anthropometric risk criterion plus a dietary risk criterion) may be more effective than the use of a single risk criterion in predicting a benefit from participation in the WIC program.
• Evaluate whether overweight or obese mothers and their infants and children benefit from current WIC program interventions. The prevalence of overweight and obesity among low-income women, infants, and children is increasing over time, and the health and nutrition risks of obesity are well-documented.
• Evaluate the yields of benefit for the various cutoff points used for anthropometric risk criteria—recognizing that there is no obvious justification

• for symmetric high and low cutoff points. It is possible that current cutoff points are so generous that the yield of benefit from WIC program interventions is low.
• Examine how the WIC program affects nutrition outcomes for individuals with selected medical risk factors.
• Determine the extent to which women who use cigarettes, alcohol, and/or illegal drugs benefit from the WIC program and the level of use of these substances that should be set as the cutoff point, if applicable.
• Invest in the development and validation of practical dietary assessment instruments that can be used across WIC programs for the identification of inappropriate dietary patterns, inadequate dietary intake, and food insecurity, recognizing that adaptations may be needed for culturally diverse populations.
• Examine the utility of predisposing factors (such as homelessness, migrancy, low level of maternal education, child abuse and neglect, and maternal depression) as predictors of benefit from WIC program services.

In addition to these research recommendations, the committee recommends the following actions be taken by the Food and Consumer Service, U.S. Department of Agriculture, to provide guidance to state WIC agencies in the development of nutrition risk criteria:

• Adopt scientifically justified cutoff values for anemia and for anthropometric criteria among women, infants, and children, realizing that they may be different across populations as prevalences change.
• Define preterm consistently as delivery before the end of the 37^th postmenstrual week for both mothers and their infants.
• Adopt scientifically justified cutoff points for young maternal age (chronological or gynecological, or both), because increased risks associated with births to these women cannot be entirely explained by poverty.
• Distinguish among some of the broadly defined medical and dietary conditions used by the WIC program in order to identify eligible WIC participants truly at high nutrition risk. These broad nutrition risk categories include endocrine disorders, renal disease, chronic and recurrent infections, food allergies, and genetic and congenital disorders. They include a broad range of medical problems with varying degrees of nutrition risk or potential to benefit from participation in the WIC program. Similarly, the category inappropriate diet includes some behaviors for which little nutrition risk is evident. The list in Table S-1 distinguishes among criteria in the broad nutrition risk categories.
• Appoint an expert committee to provide guidance on cutoff points for cigarette, alcohol, and illegal drug use that will identify pregnant and lactating women who are most likely to benefit from the WIC program. Members of the expert committee should have expertise in substance abuse during pregnancy

• and lactation, assessment and treatment of substance abuse, public policy, nutrition, and epidemiology.
• Identify the specific drugs that place individuals at nutrition risk with prolonged use and for which WIC program interventions could provide some benefit. The current nutrition risk criteria drug-nutrient interactions and inappropriate use of medications are too broadly defined and likely to produce very low yield of benefit.
• Disseminate information about risk criteria widely.
• Consider changing the current WIC priority system to give higher priority to those nutrition risk criteria identified in this report as having strong relationships to risk and potential to benefit and lower priority to nutrition risk criteria with weaker relationships to risk and potential to benefit.

1. Risk criteria that merit higher priority: vegan diets, highly restrictive diets, selected aspects of inappropriate infant feeding, food insecurity, homelessness, child of a mentally retarded parent.
2. Risk criteria that merit higher priority among children: nutrient deficiency diseases, failure to thrive, inborn errors of metabolism, gastrointestinal disorders.
3. Risk criteria that merit lower priority: mild nausea and vomiting during pregnancy; lack of prenatal care; cigarette, alcohol, and illegal drug use.²

Such a change in the priority system would require disaggregating the current categories (anthropometric, medical, dietary, and predisposing) that are used for ranking each risk criterion into one of seven priorities. It would also mean that in some cases children could be given priority over pregnant women. Such a change should improve the targeting of the program in terms of both risk and benefit.