Concurrent healthy behavior adoption and diabetic retinopathy in the United States.

OBJECTIVE: Emerging work suggests an independent association of physical activity and healthy eating on diabetic retinopathy. No study, however, has examined whether physical activity and healthy eating have an additive and/or additive interaction effect on diabetic retinopathy. METHODS: Data from 2005-2006 NHANES were used (N = 223). Physical activity was assessed via accelerometry; healthy eating was assessed from an interview with the Healthy Eating Index calculated to represent healthy eating; and diabetic retinopathy was assessed from the Canon Non-Mydratic Retinal Camera CR6-45NM. RESULTS: Physical activity (OR = 0.70, p = 0.42) and healthy eating (OR = 0.36, p = 0.16) were not independently associated with moderate-to-severe retinopathy. However, individuals with both health behaviors, compared to none, had a reduced odds of moderate-to-severe retinopathy (OR = 0.03, p = 0.02). Further, the attributable proportion (AR = 0.57, 95% CI 0.02-1.12, p < 0.05) was significant, suggesting that a significant proportion of retinopathy may be attributed to the additive interaction between inactivity and unhealthy eating. CONCLUSION: The concurrent presence of physical activity and healthy eating was associated with reduced odds of diabetic retinopathy, with the additive interaction effects suggesting that this observed association is more than summation.