Abstract
Examinations of observed within-subject variation in the energy intake of 29 adults participating in the Beltsville One-Year Dietary Intake Study suggest that individuals possess characteristic patterns of variability in total food intake (expressed as energy intake). Although the day-to-day variation appears to contain a sizable random component, significant nonrandom components were detected in the observed variation of all but one subject. Up to 37% of the total variance observed for a subject could be explained by the long- and short-term patterns identified in food intake. Both the shape and the amplitude of these patterns were unique to the individual subject, suggesting that observed within-subject variance is a function of the particular combination of environmental and biological pressures on the individual's total food intake at any one time and of the methodological errors inherent in the estimation of this intake.