Assessment of triglyceride-glucose index and its modified indices in adults with type 2 diabetes mellitus and their relationship with dinner timing

Abstract

Type 2 diabetes mellitus (T2DM) is associated with various micro- and macrovascular complications that lead to premature morbidity and mortality. Insulin resistance disrupts glucose disposal, elevates hepatic glucose output, and accelerates β-cell dysfunction, thereby increasing cardiometabolic risk. The triglyceride-glucose (TyG) index and its anthropometric derivatives provide simple and cost-effective surrogate measures of insulin resistance. Habitual late dinner timing may impair glucose and lipid metabolism due to circadian variation. However, the relationship between TyG, TyG modified indices, and chrononutrition behaviour such as dinner timing remains limited in adults with T2DM. The cross-sectional study conducted in Vadodara city, India included 441 adults aged 35–60 years with T2DM of less than five years’ duration and not requiring insulin therapy. Participants were categorized as early (< 21:00; n  = 200) or late (≥ 21:00; n  = 241)diners. Anthropometric measurements were obtained using standardized protocols, body composition was assessed by bioelectrical impedance analysis, and dietary intake was evaluated using a 3-day 24-hour dietary recall and food frequency questionnaire. Biochemical parameters were obtained from recent medical records (< 3 days). Data were analyzed using SPSS version 24. Mean age of participants was 49.3 ± 7.1years and 232 (52.6%)were males. The mean values of TyG, TyG-BMI, TyG-WC (waist circumference), TyG-WHr (waist-to-hip ratio), and TyG-WHtr (waist-to-height ratio) were 4.95 ± 0.24, 135.58 ± 23.14, 478.72 ± 62.82, 4.58 ± 0.47, and 2.98 ± 0.39,respectively. TyG, TyG-BMI, TyG-WC, TyG-WHr, and TyG-WHtr increased progressively across quartiles (all p  < 0.001) from 4.64 ± 0.11 to 5.26 ± 0.12, 106.23 ± 9.43 to 165.29 ± 12.49, 400.86 ± 25.57 to 561.67 ± 31.30, 4.04 ± 0.17 to 5.21 ± 0.27, and 2.51 ± 0.15 to 3.51 ± 0.24, respectively, accompanied by progressively adverse lipemic, glycaemic, and adiposity profiles. TyG index showed the strongest differentiation for biochemical markers triglycerides (F = 229.870), TG/HDL ratio (F = 133.410), fasting blood glucose (F = 91.928), and HbA1c (F = 32.240). The modified indices showed strongest differentiation for their respective anthropometric components: TyG-BMI for BMI (F = 638.905), TyG-WC for WC (F = 457.098), TyG-WHr for WHr (F = 276.633), and TyG-WHtr for WHtr (F = 398.272). Beyond these formula-derived relations, TyG-BMI showed the strongest differentiation for visceral fat percentage (F = 117.639), TyG-WHr for HDL-C (F = 34.643), and TyG-WHtr for muscle percentage (F = 17.395). Participants with late dinner timing had significantly higher index values than early diners (all p  < 0.001), with moderate effect sizes (Cohen’s d = 0.55–0.77). The largest between-group differences were observed for TyG (Cohen’s d = 0.77), TyG-WHtr (0.76),and TyG-WC (0.75). Multiple linear regression analyses showed that late dinner timing was an independent positive predictor of TyG (β = 0.365), TyG-BMI (β = 0.265), TyG-WC (β = 0.350), TyG-WHr (β = 0.308), and TyG-WHtr (β = 0.342) after adjusting for sex, age, energy intake, and physical activity (all p  < 0.001), with the models explaining 7–15% of the variability in these indices. Partial ( r  = 0.26–0.37) and part ( r  = 0.25–0.36) correlation showed its independent relation with indices after adjustment. Female sex was significantly negatively related with TyG-WC (β=-0.112) and significantly positively related with TyG-WHtr (β = 0.133). Increasing age was significantly positively related with TyG-WHr (β = 0.090) and TyG-WHtr (β = 0.093). Physical activity showed a weak inverse, although non-significant, relation with anthropometry related TyG indices (TyG-BMI, TyG-WC, and TyG-WHtr). Habitual late dinner timing was independently related with higher TyG and modified TyG indices. Not applicable.