Validation of Food Security Instruments in Hispanic Households

Year: 2001

Research Center: Department of Nutrition at the University of California, Davis

Investigator: Kaiser, Lucia, and Hugo Melgar-Quiñonez

Institution: Department of Nutrition, University of California, Davis

Project Contact:
Lucia Kaiser, Nutrition Specialist
University of California, Davis
Department of Nutrition
3113 Meyer Hall
One Shields Avenue
Davis, CA 95616-8669
Phone: 530-754-9063


Validation studies of food security instruments have reported strong relationships between food insecurity and (1) declines in household food supplies, (2) infrequent fruit and vegetable consumption, (3) unemployment and participation in food assistance programs, and (4) disordered eating behaviors. Validity testing of the Federal 18-item food security instrument has supported its usefulness for monitoring food insecurity and hunger in the general U.S. population. However, researchers conducting studies among Hawaiian and Pacific Islanders have questioned the validity of applying the categorical measure of food insecurity to that population. Similar research has not been conducted among Latinos.

The main goal of this study was to validate the 18-item food security instrument in a Latino population. The authors also developed and tested a cultural framework that links food insecurity to nutritional outcomes in Latino families with young children, primarily of Mexican descent. This research may contribute to more effective monitoring of food insecurity and hunger in the United States and for the design of nutrition education programs in diverse cultural groups.

The authors used data from a cross-sectional survey, carried out between February and May 2001, of approximately 250 low-income Latino households in six California counties. Prior to the survey, 4 focus groups were conducted to examine cultural interpretation of the 18 food security questions. The survey included the following instruments: (1) the 18-item food security instrument, (2) a 171-item self-reported household food inventory, (3) a 66-item food frequency questionnaire, and (4) a 16-item family demographic record form. All families included in the study self-identified as Latino, Mexican, or Mexican- American and had at least one healthy child, between 3 and 5 years old. Trained bilingual interviewers recruited subjects from a variety of community-based agencies, including those that administer the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) and Head Start, migrant camps, the local public health department, local health centers, and family resource centers. Subjects were interviewed in a private clinic room or in their homes. Statistical procedures included Pearson correlations, the Kruskal-Wallis test (for nonparametric data), and the Mantel Haenzel chi-square. About 80 percent of the survey respondents were primarily Spanish-speaking and of Mexican descent. Seventy-nine percent were enrolled in the WIC program, and 25 percent received Food Stamps. Forty-four percent of the families (n=105) reported food insecurity without hunger, 13 percent reported moderate hunger (n=30), and 3 percent reported severe hunger (n=8).

Across the four levels of food security (food secure, food insecure with no hunger, food insecure with moderate hunger, and food insecure with severe hunger), the frequency of affirmative responses to each of 18 food security items increased as the level of food insecurity became more severe. However, within a given level of food insecurity, the frequency of affirmative responses did not always decline as expected as the severity of the items increased. In particular, subjects tended to respond positively more often to some of the child hunger items than to some of the adult hunger items.

The food insecurity scale measure was negatively associated with all categories of household food supplies: dairy, fruit, grains, meat, snack foods, and vegetables. Similarly, the categorical measure of food insecurity was significantly associated with lower household food stores. Neither the scale nor categorical measure of food security was correlated with daily servings of fruits or vegetables among preschool children. However, child fruit and vegetable intakes were significantly correlated with household supplies of those foods. Food insecurity was associated with declines in household supplies of many nutritious foods (carrots, tomatoes, whole wheat bread, apples, and oranges), as well as several less nutritious foods (soda, cookies, and chocolate powder). Household supplies of traditional Mexican foods, including beans, corn tortillas, and chili, tended to remain stable, as did supplies of several relatively high-fat or inexpensive food items (hot dogs, ice cream, Kool-Aid, and instant soup).

The authors conclude that their research findings indicate that the 18-item instrument used to monitor food insecurity and hunger in the United States is valid for use in the Latino population. However, they note that the Latino subjects responded more sensitively than expected to some of the child hunger items in comparison to the adult items, suggesting that the tool may be unable to detect the subtle differences between the quantity of foods available in these households and the nutritional quality of the food available. This research may contribute to more effective monitoring of food insecurity and hunger in the United States and to the design of nutrition education programs for diverse cultural groups.