880 THOMPSON AND LENGEMANN Table 1— COMPARISON OF APPARENT ANNUAL CONSUMPTION* OF VARIOUS FOODS WHEN A ONE-WEEK SURVEY IS EXPANDED TO AN ANNUAL ESTIMATE, 1961 (PILOT FOOD STAMP PROGRAM) Balancesheet- Rural low income Foods Apr.-May Fresh sweet corn Fresh tomatoes Sept.-Oct. 1.6 62.9 Sept.—Oct. (U.S.) 58.2 35.9 12.7 77,8 325.0 170.5 172.1 214.8 291.2 208.9 251.5 427.5 212.2 383.3 277.5 386.8 146.3 493.1 101.4 All vegetables 146.1 275.1 Meat, fish, and 163.1 eggs Grain products Dairy products 268.1 517.6 126.9 Apr.—May aaetnee 1.6 12.5 All fruits Urban low income 150.8 4.2 TOL 7 176.8 7.9 153.8 *Annual consumption rate in pounds. Sweet corn and tomatoes, give examples of the extremes of annual consumption patterns that can result from expanding a week’s survey results. Major-category groupings show the nature of variations in the complete diet structure. These can be compared with balance- Sheet data that are actually computed on a year’s disappearance of food in the United States.? Grouping foods in this manner minimizes much of the variation found in individual foods, but it is indicative of the changes that occur when estimating procedures vary so widely. Other estimates of food consumption may be obtained from con‘sumer panels or nutrition studies, but such sources are usually lacking in area coverage or are confined to very restricted numbers and Situations. It is obvious that the uses made of most food-consumption data leave much to be desired in terms of accurately estimating radio- nuclide intake. * The typical variations 1960-1963 experienced in radionuclide intake from are shown in Figs. 1 to 3, where various consumption estimates for three geographic areas of the United States are combined with radionuclide data from the Health and Safety Laboratory Tri-City Diet Studies. The estimates for 1963 show the Northeast ranging from 8400 to 16,500 pc of Sr, the North Central area from 5500 to 10,900 pc, and the West from 3800 to 7400 pc. The general trend of increased Sr intake which followed the nuclear tests is well defined in each estimate. In these figures the same “Sr data are used for each estimate of intake, but the food-consumption values are taken from various studies (Refs. 1 to 3, 5, and 6). The range of variations is restricted by the limited number of population groups for which there are estimates of food consumption. However, it is obvious that the teen-age diet had the

Select target paragraph3