RESEARCH

The Community Childhood Hunger Identification Project: A Model of Domestic Hunger-Demonstration Project

in Seattle, Washington

CHERYL A. WEHLER/ RICHARD IRA Scorr,2 AND JENNIFER J. ANDERSON3

lCommunity Childhood Hunger Identification Project, Framingham, Massachusetts 01701; 2Honors College and Department of Sociology, University of Central Arkansas, Conway, Arkansas 72032; and

Arthritis Center, Boston University, Boston, Massachusetts 02215

ABSTRACT The article presents a general description of the Community Childhood Hunger Identification Project (CCHIP), whose goal was to construct a measure of hunger appropriate for the socioeconomic conditions of the United States. The measure is part of a survey instrument developed to document the preva­lence of hunger among low income families (with incomes at or below 185% of the federal poverty level) having at least one child under the age of 12. A conceptual model of domestic hunger upon which the survey instrument was based is also described. The hunger index is an additive measure of various aspects of food insufficiency due to constrained resources. The results of a demonstration project survey of 377 low income families, con­ducted in Seattle, Washington are reported. Both the full and collapsed version of the index are strongly associated in the ex­pected direction with economic and sociodemographic variables, with reliance on strategies to cope with food shortages and with health problems of the children. Judging from these results, the hunger index appears to meet internal and external validity crite­ria, cohering in expected ways with a theoretical model of domes­tic hunger. (jNE 24:29S-35S, 1992)

INTRODUCTION

Hunger in the U.S. context. The Community Child­hood Hunger Identification Project (CCHIP) represents an attempt to broaden the conceptualization of hunger, in a way that is most appropriate to the socioeconomic context of the United States. In less developed countries, hunger has histOrically been understood in terms of its physical manifestations. Given the extreme level of material depri­vation in these countries, hunger becomes synonymous with acute undernutrition. Thus, measures of severe nutri­tional insult, such as hematologic, anthropometric, and clinical indicators, can be used to assess hunger.

While severe primary malnutrition has been reported

Address for correspondence: Cheryl A. Wehler, M.S., National Director, Community Childhood Hunger Identification Project, 10 Wilmont Road, Framingham, MA 01701; (508) 872-4448 0022-318219212401-029S$03.00/0 © 1992 SOCIETY FOR NUTRITION EDUCATION

29S

for children in the United States (1), this condition, and the absolute material deprivation that begets it, are rare in industrialized societies. Relative material deprivation exists, however, and it, too, begets food problems. Based on the relatively high standard of living in the U.S., and because food assistance programs are in place to help the needy, starvation rarely occurs and phYSiological measures are not sensitive enough to detect the subtle nature of chronic but sub-clinical hunger among the poor of the United States. Therefore, while hunger must be measured as a distinct phenomenon in industrialized settings, the measure cannot be an indicator of phYSiological changes due to acute, primary malnutrition, but, instead, must as­sess food insufficiency and inadequate food resources in an effort to capture the effects of chronic, sub-clinical undernutrition among poor families in the United States.

General description of CCHIP. The Community Childhood Hunger Identification Project (CCHIP) is a research project employing a point-prevalence survey to document the extent of hunger among low income families (with incomes at or below 185% of the federal poverty level) with at least one child under the age of twelve.

The need to accurately document the extent of child­hood hunger followed reports in the early 1980s of an increasing demand for emergency foods by families with young children. CCHIP was deSigned as a systematic ap­proach to studying the problem of hunger, particularly among families with children.

The project's aims have been to 1) develop a systemic model of domestic hunger, including a definition of hunger that allows for empirical research in industrialized nations; 2) operationalize the model, testing its indicators, by 3) implementing surveys around the nation; 4) disseminate the survey results to provide sound data for policy planning and service provision; and 5) disseminate the methodology of the survey, which could be used in future studies by

30S Wehler et aI. / COMMUNITY CHILDHOOD HUNGER

community organizations and public health professionals or researchers.

CCHIP has three phases: Phase I-Development; Phase II-National Replication & Evaluation; and Phase III- Dissemination. CCHIP has been gUided by a Techni­cal Advisory Committee (TAC) chaired by Dr. Victor Sidel, Distinguished University Professor of Social Medicine at Montefiore Medical Center and the Albert Einstein Col­lege of Medicine. The TAC consists of professionals with a broad range of expertise, including medicine, public health research and administration, child health policy, nutrition, and social science research. Corporate and private founda­tions have provided funding for the project.

Phase I-Development. Phase I was sponsored by the Connecticut Association for Human Services, a statewide research and education organization. During Phase I, the CCHIP Technical AdviSOry Committee and staff devel­oped definitions, modifying a conceptual model to guide both the construction of a survey instrument and a hunger measure. Two pretests on various versions of the survey questionnaire were undertaken and a pilot study was con­ducted in New Haven, Connecticut. The data from this pilot was used to finalize the hunger index and standardize the hunger scale questions.

Phase II-Replication & Evaluation. In Phases II and III the work of the CCHIP staff is being sponsored by the Food Research and Action Center, a non-profit, non­partisan research, public policy and legal action center located in Washington, D.C.

During Phase II, the CCHIP staff, in conjunction with local organizations, has implemented CCHIP surveys in Alabama, California, Connecticut, Florida, Michigan, Min­nesota, and New York. A multi-site comparative analysis was conducted and a report was released in March 1991. Throughout, an on-going process and outcome evaluation has been conducted that will be used to develop Phase III plans.

Prior to the replication of CCHIP surveys in the seven sites previously mentioned, a demonstration project was conducted in two sites in Washington state. This paper describes and provides results from the CCHIP survey conducted in Seattle, Washington.

METHODS

Definition of hunger. CCHIP defines hunger as the mental and physical condition that comes from not eating enough food, due to insufficient economic, family or com­munity resources. Since there are currently no reliable measures of the subtle mental or physical changes that may occur from chronic, sub-clinical food intake deficits, CCHIP's measurement concern centers instead on food insufficiency due to constrained resources.

Measure of hunger. The CCHIP hunger index is a scale composed of eight questions that indicate whether adults or children in the household are affected by food insecurity, food shortages, perceived food insufficiency or altered food intake due to resource constraints (see Col­umn 1 of Table 1 for the eight hunger questions). In addition, for each aspect of hunger, three questions are asked to determine the extent of that aspect over a twelve­month period. These questions provide information on the temporal severity and periodicity of the hunger problem (see Table 1, Note).

These questions were chosen because they ascertain the extent of sustained food insufficiency due to constrained resources. They were pretested and used in a pilot study conducted in New Haven, Connecticut.

A score of five or more on the scale of zero to eight (that is, five affirmative responses out of eight) indicates a food shortage problem affecting everyone in the household. Therefore, families answering affirmatively five or more of the eight hunger questions are considered "hungry." A score of five or more: • indicates that five or more different signs of hunger are

present in the household; and • indicates that at least one of these signs of hunger di-

rectly affects the children in the household. A score of one to four indicates that the family is "at risk of hunger," because it shows at least one sign of a food shortage problem.

In addition to the hunger measure, a conceptual model based on an understanding of the interrelated factors asso­ciated with hunger was developed. The model focuses on hunger as the principal outcome rather than on the physi­cal manifestations of severe nutritional deprivation. To op­erationalize the model, a survey instrument was developed to measure those aspects of hunger that are amenable to investigation at the household level.

Conceptual model. The conceptual framework is a modified version of a model that was originally developed by the senior author to guide the questionnaire develop­ment and multi-variate analysis of the 1983 Massachusetts Nutrition Survey (2). It is intended to facilitate the study of hunger by providing a perspective that puts the most salient factors associated with hunger in a coherent frame­work. The model, shown in Figure 1, does not depict all the possible factors associated with poverty and hunger, nor does it exhaust all possible interrelationships; however, it offers a graphiC representation of the sequential nature of the determinants of hunger and some of their attendant outcomes.

As illustrated in Figure 1, food expenditures (Food­Related Behaviors, found in the center of the model) are determined at the household level by the amount of eco­nomic resources available minus the amount of expenses for other basic needs. Apart from food purchasing power (Access to Food), community characteristics such as the

J. of Nutr. Educ. Vol. 24, No.1 January/February Supplement 1992 31S

Table 1. Hunger items in the CCHIP questionnaire, prevalence of positive responses and factor analysis of hunger items, Demonstration Project, Seattle, WA (n = 266).

Column 1 Questionnaire Items About Hunger

HOUSEHOLD FOCUS

Does your household ever run out of money to buy food?

Do you ever rely on a limited number of foods to feed your children because you are running out of money to buy food for a meal?

ADULT FOCUS

Do you or adult members of your household ever eat less than you feel you should because there is not enough money for food?

Do you or adult members of your household ever cut the size of meals or skip meals because there is not enough money for food?

CHILD FOCUS

Do your children ever say they are hungry because there is not enough food in the house?

Do your children ever eat less than you feel they should be­cause there is not enough money for food?

Do any of your children ever go to bed hungry because there is not enough money to buy food?

Do you ever cut the size of your children's meals or do they ever skip meals because there is not enough money to buy food?

Column 2 Percent> 0

62%

69%

50%

50%

36%

38%

14%

34%

Eigenvalue

Percent of variance explained

Reliability coefficient (Cronbach's alpha)

Column 3 Factor Loadings

0.70

0.45

0.78

0.81

0.70

0.83

0.61

0.83

4.1 9

52%

0.86

NOTES: Each of the hunger concepts provided in Table 1 was asked in a series of four questions similar to the example that follows:

1) Thinking about the past 12 months, did your household ever run out of money to buy food to make a meal? 2) Thinking about the past 30 days, how many days was your household out of money to buy food to make a meal? 3) When your household runs out of money to make a meal, how many days per month are you usually without food? 4) Thinking about the past 12 months, during about how many months did your household run out of money to buy food to

make a meal?

number and type of stores, the aVailability of transportation systems and the availability and acceptability of public (Federal) assistance programs can also affect a household's access to food. Low-income households can enhance food availability (over and above food purchasing power) by relying on friends, relatives and/or private charity. Re­sources, education or knowledge, food accessibility and various household features determine the quality and quantity of household food availability. Household food availability determines, to a large extent, individual food intake, especially among young children. A household hun­ger problem is a deficit of food aVailability relative to food requirements. Not all members of a household are neces­sarily affected by a household hunger problem. Indeed, an individual's experience of involuntary food intake deficits constitutes (individual) hunger. Recurrent, inadequate

food intake may compromise nutritional status, depending on the chronicity of the shortage, the specific nutrient deficit and the magnitude of the deprivation. Other nega­tive emotional, psychological , functional, behavioral or health outcomes may result from hunger, even if overt clinical signs of malnutrition are not presented or detected.

The questionnaire. Following scalability analyses and assessment of the reliability and validity of the index (sub­mitted for publication), standardized versions of the ques­tions about hunger were incorporated into the demonstra­tion project survey instrument. The questionnaire also contained questions on the follOwing topics: household composition, socio-economic information, shopping and eating patterns, strategies used to cope with food emergen­cies, participation in various publicly-funded programs,

328 Wehler et al. / COMMUNITY CHILDHOOD HUNGER

ECONOMIC RESOURCES

Variables in squar ... pivotal 'acton in the oonc:eptua/ mode .. of detenninal1tl of hunger and hung.r's oonsaquancea

Circled variablts • factors whiCh influence hun"" I nd Its conl equancts directly or indlreclly

H.avy boundaries • variablea in this study

o 8

NUTRITIONAL STATUS

FUNCTIONAl/

BEHAVIORAl EFFECTS

Framework adapted from 1983 Massachusetts Nutrition Survey Multivariate Analysis by Charyl A. Wohler.

Figure 1. CCHIP Conceptual Model. Factors associated with hunger and its outcomes.

barriers to participation, household financial information, and the health status and school attendance of the children.

The sample. Data for the demonstration project were collected in Seattle, Washington from December 1987 to January 1988 (3). The area surveyed centered on four census tracts where at least 33% of the families had chil­dren and at least 33% of those families were living below the poverty level. To qualifY for the survey, a household must have had at least one child under age 12. After signing confidentiality agreements with public school districts, a list of students under 12 years who were eligible for free or reduced-price school lunches (income eligibility is ~ 185% Poverty) was provided. After duplicate addresses were removed, a total of 993 households were included in the sampling frame.

A systematic sample of 554 households was drawn using a random digit start and a predetermined interval.

Data collection. Community residents with no prior re­search experience were employed as interviewers in Seat­tle. The national CCHIP staff and Task Force staff pro­vided extensive training for both the field supervisors and the interviewers. Interviewers contacted the sampled households directly; four call-backs to an address were required before a household was classified as non-respon-

sive. From a population of 987 households, 554 were se­lected into the initial sample. Of the 554 households se­lected into the initial sample, 425 (76.7%) were occupied and able to be contacted. Of the 425 households contacted, 407 (95.8%) were eligible for the survey (had at least one child under the age of twelve and had an income at or below 185% of the federal poverty level). Of the 407 eligi­ble households, 377 (92.6%) agreed to participate and were interviewed.

A one-hour, face-to-face, in-home interview was con­ducted. Of the respondents, 76% were the mother of the children in the household, while 24% were either the fa­ther, guardian or related in some other way.

RESULTS

This analysis was confined to those households whose in­comes were at or below 185% of the Federal Poverty level. Summary statistics of characteristics of the sample are pre­sented in Table 2. Eighty percent of the sample households had incomes below 100% of the Federal Poverty level. The sample was predominantly African American (43%); there were lower proportions of Asians (23%), Whites (non-His­panics) (18%) and Hispanics (12%). About three-fifths of the households were headed by Single parents (nearly all

J. of Nutr. Educ. Vol. 24, No.1

Table 2. Characteristics of the CCHIP demonstration survey sample, Seattle, WA.

PERCENTAGE

PERCENT POVERTY LEVELa o - 50% poverty

51 - 75% poverty 76 -100% poverty

101 -185% poverty RACE AND ETHNICITyb

Black Asian White (non-Hispanic) Hispanic (descent) Native American

HOUSEHOLD COMPOSITIONb Single parent Two parents Mean household size Mean number of children in household

an = 266 bn = 377

12.0% 57.5% 10.5% 19.9%

43% 23% 18% 12% 4%

58% 28%

4.2 2.6

of whom were females). The average household had 4.2 members of which 2.6 were children.

The hunger indicator. A more complete description of the formation and analysis of the hunger indicator has been presented elsewhere (submitted for publication). Column 2 of Table 1 shows the percentages of positive responses to the eight prevalence items. Each variable was coded zero for a negative response (indicating no experience of that aspect of food insufficiency) and one for a positive response (having experienced that aspect of food insuffi­ciency).

As can be seen in Column 3 of Table 1, principal compo­nents factor analysis showed a strong, even loading of these eight items onto a single factor, accounting for 52% of the variance (4). The reliability coefficient (Cronbach's alpha) for the eight items exceeded 0.80 (5). Since the eigenvalue of a second factor was 0.99, these eight questions would seem adequately scalable. As an additive scale, values range from zero (experience of no aspect of food insufficiency) to eight (experience of all aspects).

Table 3 shows associations for a three-category version of the scale, where the categories are No Hunger (scale score of 0), At Risk of Hunger (scale score of 1-4), and Hunger (scale score of 5-8). Notice that since four of the items in the eight-item scale concern children's experience of food insuffiCiency, every household in the Hunger cate­gory necessarily reported at least one hunger aspect per­taining to its children. The distribution of the scale scores and the strong associations observed in these correlational assessments, as well as in those conducted using pilot study data (6), support this categorization. Based on this opera­tional definition, it was estimated that 42.2% ± 3.9% of families below 185% poverty were hungry, another 37.4% ± 3.8% were at risk of hunger, while the remaining 20.3%

January/February Supplement 1992 33S

± 3.2% did not experience these aspects of food insuffi­ciency. (Standard errors for prevalence estimates have been calculated as if the initial sampling were random and nonresponses among contacted eligible households were also random.)

Note that in Table 3 the category of hunger scores be­tween 1 and 4 behaves as an intermediate category. Judg­ing from these bivariate analyses, it is sometimes not signif­icantly different from the hunger category; at other times it is not Significantly different from the non-hungry cate­gory; and at still other times, it stands as a Significantly different category between them.

Antecedents and consequents. As can be seen in Ta­ble 3, the three-category version of the hunger scale and the hunger score are both associated, as one would expect, with many of the antecedents and consequents of hunger. Households with more monthly gross income per capita and households who are less poor (those at higher poverty levels) are less likely to be hungry. If the respondent is employed, the household has a lower hunger score. The higher the hunger score the greater the number of reliance strategies used, both reliance on friends and family as well as on emergency programs. The higher the hunger score, the greater the number of health problems the child expe­rienced in the past six months, and the greater the percent­age of children experiencing increased school absences.

The only result that runs counter to a priori expectations is the positive association between hunger score and high school graduate status. The many low income, yet well­educated, recent Asian immigrants may account for this unexpected positive association.

DISCUSSION

This paper provides an overview of the Community Child­hood Hunger Identification Project, presents a model of the interrelated factors affecting and affected by domestic hunger, and reports the results of a demonstration project in Seattle, Washington.

The definition of hunger offered by CCHIP focuses on food insuffiCiency due to constrained resources. The authors contend that this is appropriate to the socioeco­nomic context of the United States. The CCHIP hunger index, measured by an additive scale, is sensitive enough to identify chronic, subclinical undernutrition among poor families in the United States. Remaining components of the model that are amenable to examination at the house­hold level (represented by the bolded symbols in Figure 1) have also been operationalized into a survey instrument.

Results from pretesting suggested that the CCHIP hun­ger questions possess face validity in that they were under­stood as intended. Results from the pilot study indicated that the hunger index exhibits internal content validity in that each item is highly correlated with the overall scale.

34S Wehler et al. / COMMUNITY CHILDHOOD HUNGER

Table 3. Association between hunger categories and various household factors, CCHIP Demonstration Project, Seattle, WA.

No Hunger At-Risk Hungry (score = 0) (score = 1-4) (score = 5-8) (n = 47) (n = 117) (n = 102)

(18%) (44%) (38%)

ECONOMIC FACTORS Income as % of

poverty level 90.1% 77.5% 66.1%

Household monthly gross per capita income $217.10 $186.20 $160. 10

SOCIODEMOGRAPHIC FACTORS Respondent employed 44.7% 28.6% 21.7% High school graduate 44.7% 61.5% 68.6% AFDC recipient 44.7% 62.6% 77.5% # of children per adult

in household 1.88 2.14 2.00 STRATEGIES Amt of reliance on

friends & relatives" 0.11 0.91 1.35 Amt of reliance on

emergency food programs" 0.13 0.85 1.77 Total number of reliance

strategies usedb 0.30 2.44 4 .20 CHILD OUTCOMES Number of child's health

problems in last 6 monthsC 0.80 1.05 2.22 Number of child's health prob-

lems plus increased school absences in last six months 0.87 1.13 2.45

Increased school absences 7.0% 8.4% 25.6%

Note: Means connected by an underline do not differ significantly at the .05 level. All others differ significantly at .05 or less. "Mean number of strategies; range 0 to 3 bMean number of strategies; range 0 to 9 cMean number of strategies; range 0 to 9

FollOwing the standardization of the wording of the ques­tions in the hunger scale, a demonstration project was conducted in Seattle, Washington to retest content validity and to assess construct validity.

As shown here, the results from the demonstration sur­vey in Seattle corroborate earlier pilot study results . Indi­vidual hunger scale items have high factor loads onto a single factor, suggesting that the index exhibits strong in­ternal content validity. Further, these data suggest that the CCHIP hunger measure also possesses construct validity, cohering in an expected manner with variables in a theoret­ical model of domestic hunger.

Further testing is needed to verify that the CCHIP hun­ger index meets all validity criteria. In addition, further replications of the CCHIP survey will permit an examina­tion of the reliability of this measure. Nonetheless, results to date appear positive.

CONCLUSION

CCHIP seeks to join the vital dialogue about the return of

hunger to America, and in so dOing, to punctuate that discussion with reliable information about the prevalence, determinants, and outcomes of hunger. The authors hope that such information will not only contribute to the sci­ence of domestic hunger but will also serve in developing a sound national policy to mitigate the need for CCHIP in the years to come.

ACKNOWLEDGMENTS

The National CCHIP Project was supported in Phase I and II by grants from The Prime rica Company Foundation, The Pillsbury Company, The Burlington Northe rn Foundation, The Ford Foundation, The W.T. Grant Foundation and The Sara Lee Foundation. Phase III is being supported by the Kraft General Foods Foundation.

The Washington Demonstration Project was a collaborative effort between the National CCHIP Project and The Washington State Governor's Task Force on Hunger. Primary financial sup­port for the Washington CCHIP survey was provided by the Burlington Northern Foundation.

J. of Nutr. Educ. Vol. 24, No.1

REFERENCES

1 Listernick, R., K. Christoffel, J. Pace, and J. Chiaramonte. Severe primary malnutrition in U.S. children. American Journal of Diseases of Children 139 (ll ): 1157-60, 1985.

2 Guyer, B., C. Wehler, M. Anderka, A. Friede, W. Bithoney, D. Frank, and S. Fogerty. Anthropometric evidence of malnutrition among low­income children in Massachusetts in 1983. Massachusetts Journal of Community Health FalllWinter 1985-86, pp. 3-9.

3 Governor's Task Force on Hunger. Hunger in Washington state, Oct 1988. Appendix: Profile of hunger in WA state, pp. 6-11.

4 Harman, H.H. Modem factor analysis. Chicago: University of Chicago Press, 1967.

5 Chronbach, 1.J. Coefficient alpha and the internal structure of tests. Psychometrika 16:297-334, 1951.

6 Connecticut Association for Human Services. Community Childhood Hunger Identification Project- New Haven Risk Factor Study, by Cheryl A. Wehler, 1987, appendix 6.

RESUME Nous presenterons une deSCription generale du Community Childhood Hunger Identification Project (CCHIP), dont Ie but etait de construire une echelle de famine appropriee pour les conditions socio-economiques des Etats-U nis. L' echelle fait partie de l' enquete developpe pour evaluer les risques de famine chez les familIes a faibles revenus (revenus situe a, ou sous 185 % du seuil de pauvrete federal ) ayant au moins un enfant de moins de 12 ans. Nous allons aussi decrire un modele de famine "domestique" selon lequell'enquete a ete formule. L'indice de famine est une mesure cumulative d'accessibilite alimentaire associee a des ressources limitees. Les resultats d'un enquete pilote effectue aupres de 377 familles a faibles revenus de Seattle (Washington), indiquent q'une etroite correlation co-

JanuarylFebruary Supplement 1992 35S

existe entre l'indice de famine et les variables economiques et socio-demographiques, Ie recours a des strategies d'accessibilite aux aliments ainsi que les problemes de sante des enfants. D'apres ces resultats, I'indice de famine semble rencontrer les criteres de validite internes et externes, agissant de fat;on at­ten due avec Ie modele de famine domestique.

Translated by Anick Gauvin-Fleurant, R.P., Dt.

RESUMEN Presentamos una descripcion general del Proy­ecto Comunitario para Identificar el Hambre en la Ninez [CCHIP], cuyo objetivo is disenar una medida especffica para medir el hambre que sea apropriada para las condiciones socio­economicas de los Estados Unidos. Esta medida es parte de un instrumento de encuestas que documente la prevalencia del hambre entre las familias pobres [con ingresos en, 0 mas bajos que, el 185% del nivel de pobreza federal) , y que tengan por 10 menos un nino menor de 12 anos. Tambien describimos un mod­elo conceptual del hambre domestica en la cual se baso el instru­mento. El indice de hambre es una medida aditiva de los varios aspectos de la insuficiencia de com ida debido a recursos lim­itados. Los resultados del proyecto de desmostracion con 377 familias de bajos ingresos, realizado en Seattle, Washington, son aqui reportados. Tanto la version completa del indice, como la colapsada, estan fuertemente asociadas, en la direccion esperada, con las variables economicas y sociodemograficas, con apoyo de las estrategias usadas en caso de escasez de alimentos, 0 de problemas de salud de los ninos. Juzgando por estos resultados, el fndice de hambre parece satisfacer el criterio de validez interna y externa, consistente, en la forma esperada, con el modelo teor­ico del hambre domestica.

Translated by Diva Sajur, Ph.D., M.P.H.

ABSTRACT OF INTEREST

Food security: A nutritional outcome or a predictor variable? Campbell, C.A. Journal of Nutrition 121:408-415, 1991.

The phenomenon loosely labelled hunger in the 1980s is now being discussed as food security or insecurity. Food security is defined as access by all people at all times to enough food for an active, healthy life, and at a minimum includes the follOwing: (1) the ready availability of nutri­tionally adequate and safe foods, and (2) the assured ability to acquire personally acceptable foods in a SOcially acceptable way. Food insecurity exists whenever food security is limited or uncertain. The measurement of food insecurity at the household or individual level involves the measurement of those quantitative, qualitative, psycholOgical and social or normative constructs that are central to the experience of food insecurity, qualified by their involuntariness and periodicity. Risk factors for food insecurity include any factors that affect household resources and the proportion of those resources available for food acquisition. Potential consequences of food insecurity include hunger, malnutrition and (either directly or indirectly) negative effects on health and quality of life. The precise relationships between food insecurity and its risk factors and potential consequences need much more research now that there is an emerging consensus on the definition and measurement of food insecurity. Indicators of food security or insecurity are proposed as a necessary component of the core measures of the nutritional state of individuals, communities or nations.