Evaluation of four methods for determining energy intake in young and older women: comparison with doubly labeled water measurements of total energy expenditure.

The accuracy and precision of four different food intake assessment methods were evaluated in young and older women by comparing reported energy intakes with doubly labeled water measurements total energy expenditure (TEE). A study lasting 8 d was conducted in 10 young women aged 25.2+/-1.1 y (-x+/-SEM) and in 10 older women aged 74.0+/-1.4 y. Free-living TEE was measured over 7 d and food consumption was determined from weighed food intake data (7 d), a 24-h food recall (in duplicate), and two different food-frequency questionnaires [Fred Hutchinson Cancer Research Center (FHCRC)/Block and Willett, both in duplicate]. In addition, body composition was determined by using hydrodensitometry, and strenuous physical activity and the extent of dietary restraint were determined by questionnaire. In young women, 24-h recall gave mean energy intakes that were closest to measures of TEE (-0.34+/-3.71 MJ/d compared with TEE, P=0.178), and energy intakes by food-frequency questionnaires were the only intake data that correlated significantly with individual values for TEE (P<0.05). In older women, food-frequency questionnaires gave mean energy intakes that were closest to measured TEE (+0.53+/-2.95 MJ/d with the Willett questionnaire and -1.19+/-3.02 MJ/d with FHCRC/Block questionnaire). No energy intake data from this group correlated significantly with values for TEE. The 7-d weighed dietary intakes were significantly lower than measured TEE in both young and older women (-2.0 MJ/d in young and older women combined, P<0.001), and did not correlate significantly with values for TEE, although they did most closely mirror the mean difference in TEE between the young and older women (2.30 MJ/d for TEE and 2.11 MJ/d for 7-d weighed intake). These data suggest that none of the methods studied gave accurate estimates of the usual energy requirements of individual subjects. In addition, the results suggest that for some types of studies, simple methods for assessing group mean dietary intake may actually give more accurate information than weighed dietary intakes.