Dietary adherence in the women's health initiative dietary modification trial

This article describes adherence to a low-fat dietary pattern (less than 20% energy from fat, five or more fruit/vegetable and six or more grain servings daily) in Years 1 and 5 of the Women's Health Initiative Dietary Modification Trial, which was designed to examine the effects of a low-fat dietary pattern on risk of breast and colorectal cancers and other chronic diseases in postmenopausal women. Participants were randomly assigned to a low-fat dietary intervention arm (40%, n=19,542) or a usual diet control arm (60%, n=29,294). Women in the intervention arm completed 18 group sessions during the first year, followed by quarterly annual maintenance sessions. Adherence was assessed as control minus intervention (C-I) group differences in percent total energy from fat as estimated by a food frequency questionnaire. Based on these self-reported dietary data, mean C-I was 10.9 percentage points of energy from fat at Year 1, decreasing to 9.0 at Year 5. Factors associated with poorer adherence were being older, being African American or Hispanic (compared with white), having low income, and being obese. Group session attendance was strongly associated with better dietary adherence. There are many limitations of self-reported dietary data, particularly related to social desirability and intervention-associated bias. Nonetheless, these data indicate that long-term dietary change was achieved in this clinical trial setting and reinforce the potential of the ongoing trial to answer questions of public health importance.