890 THOMPSON AND LENGEMANN techniques and general applicability of such work. In addition, the representativeness of the samples from the field studies is questionable, and the conditions inherent in metabolic-ward studies do not reflect normal behavior for most individuals. In an attempt to examine the nature of individual variations as well as to determine the feasibility of this approach under normal conditions, a urine/diet study was conducted during March 1964. Fourteen volunteers (7 males and 7 females) ranging in age from 21 to 33 years were placed on balanced diets for 17 days. The content of each day’s diet was the same throughout the entire 17-day period. All diets were made up in advance from the same coded batches of raw products. They were prepared in frozen-dinner form and distributed to the volunteers on a prescribed schedule. The volunteers followed their normal daily work or recreational patterns, consuming the meals on their normal schedules. Daily urinary collections were started after six days and were subdivided to permit comparison of first morning specimens with “rest-of-day” or total 24-hr samples. Alternate-day samples were analyzed for each person to measure variations over time and between subjects. All volunteers completed the study and adheredstrictly to the schedule. The average Sr level in the diet was 24.8 pe per gram of calcium for the male diet and 22.3 pc per gram of calcium for the female diet. Diet-calcium intakes were approximately 1.0 g in eachinstance. Results of the analyses indicated an overall OR,..../gie, Of 0.97 + 0.08 (1.07 for females and 0.87 for males). This compares with the ORurine /diet Of 0.82 for young adults determined in the Samachson study.’> The Similarity of results would seem to indicate that urinary %Sr/calcium values can be used to predict dietary levels. However, Table 4— RELATIVE DISTRIBUTION OF ORurine/diet VALUES, DIETARY STUDY, 1964 OR urine/diet Percent of observations . ORurine/diet <0.50 0.50 to 0.74 1.9 23.1 1.25 to 1.49 1.50 to 1.74 3.8 1.9 1.00 to 1.24 17.3 >2,.00 3.8 0.75 to 0.99 44.2 1.74 to 1.99 Percent / of observations 3.8 this would not apply to estimates of individual persons since the range of daily OR,.ine/diep Values observed in this study extended from 0.49 to 2.95. Although the difference between high and low values seems rather large, the OR values were concentrated in the range between 0,50 and 1.24 (see Table 4). Since 85% of the OR values were observed within this range, this procedure would seem to be Suitable for esti-

Select target paragraph3