ABSTRACT
Background
It is unknown whether behavioral interventions to improve diet are effective in people with a genetic predisposition to obesity.
Objectives
To examine associations between BMI genetic risk and changes in weight and workplace purchases by employees participating in a randomized controlled trial of an automated behavioral workplace intervention to promote healthy food choices.
Methods
Participants were hospital employees enrolled in a 12-mo intervention followed by a 12-mo follow-up. Hospital cafeterias utilized a traffic-light labeling system (e.g., green = healthy, red = unhealthy) that was used to calculate a validated Healthy Purchasing Score (HPS; higher = healthier). A weighted genome-wide BMI genetic score was generated by summing BMI-increasing alleles.
Results
The study included 397 adults of European ancestry (mean age, 44.9 y; 80.9% female). Participants in the highest genetic quartile (Q4) had a lower HPS and higher purchases of red-labeled items relative to participants in the lowest quartile (Q1) at baseline [Q4–Q1 Beta HPS, –4.66 (95% CI, –8.01 to –1.32); red-labeled items, 4.26% (95% CI, 1.45%–7.07%)] and at the 12-mo [HPS, –3.96 (95% CI, –7.5 to –0.41); red-labeled items, 3.20% (95% CI, 0.31%–6.09%)] and 24-mo [HPS, –3.70 (95% CI, –7.40 to 0.00); red-labeled items, 3.48% (95% CI, 0.54%–6.41%)] follow-up periods. In the intervention group, increases in HPS were similar in Q4 and Q1 at 12 mo (Q4–Q1 Beta, 1.04; 95% CI, –2.42 to 4.50). At the 24-mo follow-up, the change in BMI from baseline was similar between Q4 and Q1 (0.17 kg/m2; 95% CI, –0.55 to 0.89 kg/m2) in the intervention group, but higher in Q4 than Q1 (1.20 kg/m2; 95% CI, 0.26–2.13 kg/m2) in the control group. No interaction was evident between the treatment arm and genetic score for BMI or HPS.
Conclusions
Having a high BMI genetic risk was associated with greater increases in BMI and lower quality purchases over 2 y. The 12-mo behavioral intervention improved employees’ food choices, regardless of the genetic burden, and may have attenuated weight gain conferred by having the genetic risk.