Estimating the variations and autocorrelations in dietary intakes on weekdays and weekends

The high incidence of breast cancer in the U.S. and the possible link with dietary fat has led to the development of educational programmes for reducing women's fat intakes by agencies such as the National Cancer Institute and the U.S. Department of Agriculture. In this paper, we analyse the effects of an intervention on the intakes of 12 nutrients by 37 women in the Houston area. We estimate a dynamic random effects model by maximum likelihood to estimate the between and the within variations and the autocorrelations using 7 consecutive food records before and after the intervention programme. The main findings are that the pattern of within variations differs during weekdays and weekends. Secondly, the mean intakes of nutrients such as beta-carotene and ascorbic acid tend to be lower on weekends. Lastly, the intervention programme reduced the overall fat intakes and also increased the variation in the consumption of foods high in fats during weekdays. We discuss the implications of the results for the design of further studies.