researchers to believe that the caloric deficit, rather than the state of hydration, was responsible for the majority of the weight lost in the hot environment.

A number of military studies conducted by the U.S. Department of Defense have looked at garrison feeding, food choices, and food waste; they have also conducted tests of the rations that have been developed. In each case these studies have been conducted during only one season, usually fall or spring; thus comparative information regarding summer food choices is not available.

A few studies have investigated the relationship between seasonal changes in body composition and seasonal changes in caloric intake and body weight. Some of these studies have also evaluated nutrient intake in adults by season of the year in hot environments. Decreased intake of several nutrients, such as vitamins A and C (Aldashev et al., 1986) and protein, vitamin C, and total energy (Mommadov and Grafova, 1983), has been reported. However, these studies did not evaluate changes in food preferences or appetite.

Empirical data, based on observations and practices in food service in both the military and the commercial sector, indicate a change in food preferences during seasons associated with elevated mean environmental temperatures. Few basic studies have attempted to specifically address food patterns that change according to season in self-selected diets. In a study of seasonal variations in self-selected lunches in a large employee cafeteria in Maryland, Zifferblatt and colleagues (1980) found a decreased selection of starches and cooked vegetables with increased purchases of fruits, salads, yogurt, and cottage cheese as the noontime temperature rose. As the temperature increased, average caloric purchases also tended to decrease. The workplace cafeteria, along with the work areas of most of the employees, was kept at 72°F; and thus, the environmental temperature of the location at which the food was ingested may have only moderately influenced workers' appetites.

National surveys have investigated the food consumption patterns of Americans, but they have not gathered recent data on the same individuals or on individuals in similar geographic areas at different times of the year. Such data would help determine whether changes of season, and thus changes in environmental temperatures, affect appetite (resulting in changes in food intake) or food selection patterns. The 1977–1978 Nationwide Food Consumption Survey reported three-day food intake information for about 36,100 individuals from a sample of households in 48 states that included four seasonal samples. The seasonal differences in average intakes of 10 major food groups were low—11 percent or less. For the major food groups, average intakes were typically higher in one season than in the three others with intakes of vegetables, fruits, and beverages increasing in the summer. Intake