Pretreatment predictors of weight loss in a 12-month behavioral obesity treatment: a signal detection analysis of DIETFITS

Abstract

Objective

The objective of this study was to identify pretreatment predictors of weight loss in a 12-month behavioral obesity treatment that restricted either fat or carbohydrates.

Methods

Participants were 436 adults with overweight or obesity from the Diet Intervention Examining The Factors Interacting with Treatment Success (DIETFITS) trial. Signal detection analysis was used to identify which combinations of 51 pretreatment demographic, clinical, behavioral, and psychosocial variables, along with diet type (healthy low-fat vs. healthy low-carbohydrate), formed subgroups that varied in proportion of those achieving at least 5% weight loss at 12 months.

Results

Overall, 51% of participants achieved at least 5% weight loss at 12 months, with eight subgroups identified through signal detection. Diet type was not a key factor. Among racial and ethnic minority participants, the best predictors of weight loss were lower levels of emotional eating, less friend discouragement, and presence of metabolic syndrome. Among non-Hispanic White participants, the best predictors were high confidence in participating fully in the intervention, more family encouragement, and lower outcome expectations.

Conclusions

We found that psychosocial and clinical factors, along with race and ethnicity, successfully differentiated subgroups that varied in their 12-month weight loss. Given the heterogeneity in response to behavioral obesity treatment, these results can help generate hypotheses to move intervention science toward a precision medicine approach by matching individuals to their most suitable obesity treatments.