| The purpose of the study was to examine the effects of cognitive complexity, dietary patterns, age and education on inter-individual differences in the reproducibility of a food frequency questionnaire (FFQ) administered to women 60 years of age and older.; Subjects were divided into three groups according to time interval between FFQ administrations: group 1 = next day, group 2 = 7 days, and group 3 = 6-8 weeks. Demographic and cognitive data were obtained by in-person interviews; dietary intake data were obtained by telephone interviews (i.e., two FFQs and three random 24-hour recalls were collected by telephone). Cognitive measures assessed memory and organizational abilities. Dietary intake data were obtained to assess dietary complexity indicators such as the total number of different foods consumed and the total number of mixed dishes consumed. Analysis of covariance was used to study time interval (i.e., group) differences and the separate effect of each predictor factor on reproducibility of nutrient reports, and stepwise regression was used to study the effect of the various factors examined concurrently on the reproducibility of the FFQ.; Pearson correlation coefficients varied from 0.67 (vitamin A) to 0.88 (calcium) and Intraclass correlation coefficients varied from 0.49 (calcium) to 0.86 (niacin). About one-administrations. Group effect was significant for energy, fat, saturated fat, cholesterol, sodium, vitamin A, vitamin C, and vitamin E: women in group 2 had the lowest reproducibility; there were no significant differences in reproducibility between women in groups 1 and 3.; Age and education explained very little of the variance in reproducibility. The cognitive factors studied, also, had little or no effect. There was a consistent dietary pattern effect on reproducibility; however, R-squares were small (R{dollar}sp2 leq{dollar} 0.08) and of little practical significance. The causes of individual differences in reliability of FFQs remain poorly understood. |
