Predictive value of insulin resistance related indicators for osteoporosis risk in patients with type 2 diabetes mellitus: a real-world study.
Insulin resistance (IR) contributes to osteoporosis (OP) by impairing bone microarchitecture, posing significant fragility-fracture risk in patients with type 2 diabetes mellitus (T2DM). The triglyceride-glucose (TyG) index and its anthropometry-adjusted derivatives (TyG-body mass index [BMI], TyG-waist circumference [WC]) are simple IR surrogates that may aid OP risk stratification. This study evaluated TyG, TyG-BMI and TyG-WC as independent predictors for OP in T2DM.
In a retrospective cohort of 568 patients with T2DM (36 OP cases), baseline clinical, laboratory (HbA1c, haemoglobin [HGB], high-density lipoprotein cholesterol [HDL], serum calcium [Ca]) and bone-metabolism (procollagen type 1 N-terminal propeptide [P1NP], osteocalcin [OC], bone-specific alkaline phosphatase [BAP]) data were analyzed. Candidate covariates were selected by least absolute shrinkage and selection operator (LASSO) regression with variance inflation factors (VIF) screening. Because the events-per-variable ratio was low (≈4), three Firth's penalised-likelihood logistic regression models were fitted for TyG, TyG-BMI or TyG-WC as the primary analysis. Discrimination was assessed by the area under the receiver-operating-characteristic curve (AUC) with 95% confidence intervals (CIs) and pairwise DeLong tests.
All three indices were significantly higher in OP than non-OP patients (all P < 0.001). After Firth-penalised adjustment, only remained independently associated with TyG-WC (OR 1.004; 95% CI 1.001-1.008; P = 0.013); TyG (P = 0.075) and TyG-BMI (P = 0.105) showed positive but non-significant associations. Higher HbA1c, OC and BAP increased the odds of OP, while male sex and higher HGB were protective. Discrimination was high and similar across models (apparent AUCs: 0.896 [TyG], 0.887 [TyG-BMI] and 0.901 [TyG-WC], with no DeLong-significant difference between indices.
Among the three IR surrogates, only TyG-WC was independently associations with OP in T2DM, although overall discrimination was comparable across indices. TyG-WC may serve as low-cost adjuncts for identifying diabetic patients at elevated OP risk, pending prospective external validation.
In a retrospective cohort of 568 patients with T2DM (36 OP cases), baseline clinical, laboratory (HbA1c, haemoglobin [HGB], high-density lipoprotein cholesterol [HDL], serum calcium [Ca]) and bone-metabolism (procollagen type 1 N-terminal propeptide [P1NP], osteocalcin [OC], bone-specific alkaline phosphatase [BAP]) data were analyzed. Candidate covariates were selected by least absolute shrinkage and selection operator (LASSO) regression with variance inflation factors (VIF) screening. Because the events-per-variable ratio was low (≈4), three Firth's penalised-likelihood logistic regression models were fitted for TyG, TyG-BMI or TyG-WC as the primary analysis. Discrimination was assessed by the area under the receiver-operating-characteristic curve (AUC) with 95% confidence intervals (CIs) and pairwise DeLong tests.
All three indices were significantly higher in OP than non-OP patients (all P < 0.001). After Firth-penalised adjustment, only remained independently associated with TyG-WC (OR 1.004; 95% CI 1.001-1.008; P = 0.013); TyG (P = 0.075) and TyG-BMI (P = 0.105) showed positive but non-significant associations. Higher HbA1c, OC and BAP increased the odds of OP, while male sex and higher HGB were protective. Discrimination was high and similar across models (apparent AUCs: 0.896 [TyG], 0.887 [TyG-BMI] and 0.901 [TyG-WC], with no DeLong-significant difference between indices.
Among the three IR surrogates, only TyG-WC was independently associations with OP in T2DM, although overall discrimination was comparable across indices. TyG-WC may serve as low-cost adjuncts for identifying diabetic patients at elevated OP risk, pending prospective external validation.