Abstract / Summary
Patients with type 2 diabetes mellitus (T2DM) have increased skeletal fragility despite bone mineral density that is often normal or relatively high. The hemoglobin glycation index (HGI) captures the difference between measured HbA1c and the value expected from glucose levels. Recent studies have reported associations between HGI and bone mineral density, but evidence remains limited across different populations and skeletal outcomes. We examined the association between HGI and lumbar osteoporosis in hospitalized patients with T2DM. We retrospectively analyzed hospitalized patients with T2DM who underwent dual-energy X-ray absorptiometry between January and December 2023. HGI was calculated as the difference between measured HbA1c and HbA1c predicted from fasting plasma glucose (FPG). Lumbar osteoporosis was defined as an L1–L4 T-score ≤ -2.5. Logistic regression was used for the primary analysis, with additional sensitivity analyses adjusting for HbA1c or FPG. Restricted cubic spline models were used to explore nonlinearity. Lumbar spine L1–L4 T-score and bone turnover markers were examined as secondary and exploratory outcomes. A total of 757 participants were included in the baseline analysis, and 687 had complete data for the fully adjusted primary logistic model. Lumbar osteoporosis was present in 208 participants (27.5%) overall and in 190 of the 687 participants included in the fully adjusted model. Its prevalence increased across HGI tertiles from 15.8% to 28.6% and 38.1%. In the fully adjusted primary model, each 1-unit increase in HGI was associated with higher odds of lumbar osteoporosis (OR = 1.342, 95% CI: 1.168–1.546; P < 0.001). This association remained significant after additional adjustment for HbA1c (OR = 1.251, 95% CI: 1.039–1.507; P = 0.018) and after additional adjustment for FPG (OR = 1.337, 95% CI: 1.164–1.540; P < 0.001). Higher HGI was also associated with a lower L1–L4 T-score (β = -0.134, 95% CI: -0.236 to -0.031; P = 0.011). In exploratory analyses, HGI was inversely associated with osteocalcin (β = -0.057, 95% CI: -0.103 to -0.011; P = 0.016), whereas associations with PINP and β-CTX were not significant. Higher HGI was associated with a greater prevalence of lumbar osteoporosis and lower lumbar spine L1–L4 T-scores in hospitalized patients with T2DM. The osteocalcin finding was modest and exploratory. Because of the cross-sectional design and the absence of formal discrimination or reclassification analyses, these results should be interpreted as associations and do not demonstrate causality or incremental predictive value beyond HbA1c. Not applicable.