Association Between Self-Reported Diabetic Diet and LDL Cholesterol Levels in U.S. Overweight and Obese Adults: A Cross-Sectional Analysis of NHANES 2017–2018

Résumé fourni par la source

Background: High levels of low-density lipoprotein cholesterol (LDL-C) and diabetes are major cardiovascular risk factors. Dietary modifications are commonly recommended as part of diabetes management. However, in real-world populations, the association between self-reported adherence to a diabetes-focused diet and LDL-C levels remains unclear. Objective: To examine the cross-sectional association between a self-reported diabetic diet (SRDD) and LDL-C levels in U.S. overweight and obese adults, and to explore potential mediators and effect modifiers of this relationship using nationally representative data. Methods: We performed a cross-sectional analysis of 2,150 adults aged ≥18 years, overweight and obese, from the 2017–2018 National Health and Nutrition Examination Survey (NHANES). Exposure was SRDD and the outcome was laboratory-measured LDL-C. Linear regression models were used to assess associations. Multivariable models included age, statin use, diabetes diagnosis, and HbA1c to evaluate whether the observed association persisted after accounting for metabolic status and treatment-related factors. NHANES survey weights were applied. Results: Of the 2,150 adults, 52 participants reported following a diabetic diet. In unadjusted analyses, SRDD was associated with significantly lower LDL-C levels (mean difference = –18.6 mg/dL; 95% CI –28.4 to –8.8; p < 0.001). After adjustment for metabolic and treatment-related variables, including diabetes status, statin use, and HbA1c, the association was attenuated and no longer statistically significant (β = –8.4; p = 0.3). Statin use showed the strongest association with lower LDL-C levels (β = –26.6 mg/dL; 95% CI –30.7 to –22.4; p < 0.001). Effect modification analysis indicated that diabetes status significantly modified the association between diet and LDL-C (p for interaction = 0.024), with an inverse association observed primarily among participants with diabetes. Mediation analyses indicated that body mass index (BMI) significantly mediated the association between diet and LDL-C (p = 0.007), whereas HbA1c did not demonstrate evidence of mediation. Conclusions: While adherence to a self-reported diabetic diet was associated with lower LDL-C levels in unadjusted analyses, this association was attenuated after accounting for metabolic and treatment-related factors. Mediation analyses suggest that body mass index may partially explain this relationship, indicating that the association between diet and LDL-C could operate through broader metabolic mechanisms. In contrast, HbA1c did not appear to mediate the association. Overall, these findings highlight the complex interplay between dietary behaviors, metabolic status, and clinical management in cross-sectional populations and underscore the need for longitudinal studies to clarify causal pathways.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.