A longitudinal study of food intake patterns and obesity in adult Danish men and women.

OBJECTIVE: The aim of this study was to test the hypothesis that specific food intake patterns or changes in food intake patterns were related to future changes in body mass index (BMI).
DESIGN: Longitudinal observational study, with clinical and questionnaire examinations at baseline and two follow-up surveys, after 5 and 11 years.
SUBJECTS: In all, 3785 men and women attended at baseline, of which 2436 aged 30-60 y attended all three examinations. MEASUREMENTS: A 26-item food frequency questionnaire, standardised measurements of height and weight and a lifestyle questionnaire. Food intake patterns were identified by factor analysis. Regression models including: scores on each factor, BMI, smoking, leisure time physical activity, education, parity, age; and as outcomes: baseline BMI, BMI change between baseline, 5- and 11-y follow-up and obesity at 11-y follow-up, respectively.
RESULTS: For men, three factors labelled 'Green', 'Sweet' and 'Traditional', and for women, two factors labelled 'Green' and 'Sweet-Traditional' were identified. Scores on the 'Sweet' and 'Sweet-Traditional' factors were inversely associated with baseline BMI. For men, baseline 'Traditional' factor score and, for women, baseline 'Sweet-Traditional' factor score was inversely associated with subsequent 11- and 5-y BMI change, respectively. Using the three examinations, a more advanced longitudinal model, which included preceding changes in BMI and factor scores, was tested but no significant associations between factor scores, changes in factor scores and subsequent BMI changes or obesity were found.
CONCLUSION: In this longitudinal study of a Danish population, food intake factors could not consistently predict changes in BMI or obesity development.