Background & aims: Plant-based diets benefit human health, while the deficient in some nutrients limits its application. We aimed to examine whether balanced diets could be better in reducing diabetes risk than plant-based diets.
Methods: In cross-sectional analysis of Environment-Inflammation-Metabolic-Diseases Study (EIMDS), we used a questionnaire to investigate the habit of balanced and plant-based diets. In the prospective analysis of UK Biobank, we used the plant-based diet index of health (hPDI) and unhealth (uPDI) to evaluate the plant-based diets, and defined the balanced diet as a daily intake of 5 categories and at least 12 types of food. After analyzing proteomic data in UK Biobank, we explored the causal relationship between signature proteins of balanced diets and incident diabetes based on summary-data-based Mendelian randomization (SMR).
Results: Compared to participants who had plant-based diets, those who had balanced diets showed a lower risk of diabetes in EIMDS (Odd Ratio 0.65, 95%CI 0.44-0.95). In UK Biobank, after excluding participants with unhealthy plant-based diets, participants with balanced diets still showed a lower diabetes risk than participants with plant-based diets (Hazard Ratio 0.86, 95%CI 0.77-0.95). Proteomic analysis identified 107 downregulated and 2 upregulated proteins that were associated with higher and lower risk of diabetes, respectively. In SMR analyses, the downregulated signature proteins of balanced diets (AGR2, DBI, IL17RA and SERPINH1) were causally associated with diabetes incidence.
Conclusion: Adhered to a balanced diet is associated with a lower risk of diabetes compared to plant-based diet, which might be attributed to signature proteins such as AGR2, DBI, IL17RA and SERPINH1.
Keywords: Balanced diets; Causal Relationship; Diabetes; Mendelian Randomization; Plant-based diets.
Copyright © 2024 Elsevier B.V. All rights reserved.