LONG-TERM STUDY IN THE HOME; DIET AND RESULTS

Abstract

A long-term metabolic balance study of healthy infants was performed in their homes to determine food consumption and the intake and retention of calcium, phosphorus, stable strontium, and strontium-90. Balances were measured in consecutive 28-day periods. The diets were selected by the parents on the advice of their pediatricians, and consisted of milk, commercially available premodified milk formulas, and packaged infant foods. The only strontium-90 fed to the infants was that which was present in foods and water due to fallout. Aliquots of all foods were analyzed for the substances of interest. Excreta were collected in paper diapers for analysis. Calcium, phosphorus, and strontium-90 balances are available for 30 infants in a total of 214 periods; stable strontium balances are available for 10 infants in 56 periods. The average age of the 30 infants was 164 days and the average weight was 7.4 kg. The average daily intake of milk and formula decreased from 654 ml at age 31 to 60 days to 590 ml at age 271 to 300 days; in the same interval, the average daily consumption of all other infant food increased from 80 to 413 gm. The diet in this study differed from uncontrolled diets mainly in the use of formula instead of cows' milk for the older infants, and the substitution of a packaged cereal mix for the more common dry cereal that is mixed with milk just before feeding. Use of the dry cereal would have added approximately 80 ml to the measured average consumption of 615 ml of milk per day. A major problem in performing an accurate metabolic balance study of strontium lies in the small fractional retention of the substance. As a result, small systematic errors in determining intake and collecting excreta lead to fractionally large systematic retention errors, and small random errors in intake and excreta analysis lead to large random retention errors. Fortunately, the large number of retention values reduces the random error of the mean. Typical individual retentions of stable strontium and strontium-90 have standard deviations of 38 and 24%, respectively; standard deviations of the mean of 56 stable strontium and 200 strontium-90 retentions are 5 and 2%, respectively. Based on numerical estimates of the recognized systematic errors, the respective true values of stable strontium and strontium-90 retention are believed to be 20 and 13% lower than the measured averages. The average retentions have been corrected by these percentages. The probable error of the systematic correction is undoubtedly large, but is smaller than the correction factor, according to three comparisons with independently derived values. This study demonstrates the feasibility of long-term metabolic balance studies in the home; it also indicates the need for internal controls in the samples, highly precise analyses, and methods for reducing systematic errors. For information on intake, this type of study complements surveys by questionnaires and interviews by providing more accurate values for a much smaller population. For retention, the study may be less accurate than one performed in a metabolic ward, but can be extended over a long time with less stress on the infant. Results of this type of study may be tested for accuracy by comparison with studies in metabolic wards, by tissue analyses for accumulated retention, and by in vivo measurements of the retention of radionuclides, although each of these has its own inherent inaccuracies that must be considered.