Lack of prospective relationships of the Power of Food Scale with Body Mass Index and dieting over 2 years in U.S. emerging adults.

BACKGROUND

This study investigated prospective relationships of the Power of Food Scale (PFS), a self-report measure of hedonic hunger, with weight outcomes and dieting in U.S. young adults.

METHODS

PFS (PFS-aggregate and subscales: PFS-available, PFS-present, PFS-tasted) was assessed in waves (W, years) 5 and 6 of a nationally representative cohort of 10th graders assessed annually (baseline n = 2785, 83% retention at W7). Internal consistency (Cronbach's α), 1-year stability (intraclass correlation coefficient, ICC) and confirmatory factor analysis (CFA) were examined. Analyses accounting for the complex survey design examined cross-sectional associations of W6 PFS with W6 BMI and dieting, and prospective relationships of PFS in each wave with BMI, 1-year BMI change (BMIΔ, W6-W5 and W7-W6), overweight/obesity onset (OWOB, moving to a higher risk weight category) and dieting in the following wave. Multiple imputations addressed missing data.

RESULTS

Baseline participant mean ± SE age was 20.3 ± 0.02 years. Mean BMI increased by approximately 0.6 kg/m from W5 through W7; OWOB occurred in 7.4% of participants between W5-W6; 9.0% between W6-W7. Approximately half the sample reported dieting in each wave. W6 weight outcomes were not associated with W6 PFS, but W6 dieting frequency was positively associated with W6 PFS-available, PFS-present, and PFS-aggregate (but not PFS-tasted) in multivariable models. PFS was not prospectively associated with weight outcomes. Positive prospective associations of PFS with dieting frequency were inconsistent across waves and with respect to inclusion of covariates.

CONCLUSIONS

Greater PFS is associated with dieting cross-sectionally but was not a reliable indicator of susceptibility to future weight outcomes or dieting in young adults.