Systemic glucocorticoid use increases fracture risk similarly in individuals with and without type 2 diabetes: a Danish nationwide cohort study.

Abstract

Type 2 diabetes (T2D) and glucocorticoid treatment both increase risk of bone fractures. We found that glucocorticoid use increases fracture risk similarly regardless of T2D status. Findings were consistent over time and across subgroups. Care should be taken when prescribing glucocorticoids to individuals with T2D already at increased fracture risk. Type 2 diabetes (T2D) is associated with increased risk of bone fractures, which may be caused by suppressed bone formation. Similarly, glucocorticoid (GC) treatment leads to suppressed bone formation, in addition to increased resorption. It is unknown whether GC treatment alters fracture risk equally in subjects with and subjects without T2D. We performed a nationwide registry-based cohort study with a within-subject pre-treatment reference period to investigate whether T2D status modified the effects of systemic GCs on risk of first major osteoporotic fractures (MOFs)-defined as fractures of the hips, vertebrae, humerus, or forearm-within each observation period. Individuals with or without T2D receiving systemic GC treatment (≥ 1 redeemed prescription) between 2000 and 2018 were identified and followed during treatment. The reference period for GC treatment consisted of the 2-year period immediately prior to treatment in any given individual. For confounding control, we used standardized morbidity ratio (SMR) weighting. Each group (with/without T2D and during/before GC) was reweighted to resemble the T2D group in the GC treatment period. Incidence rate ratios (IRR) were estimated using a zero-inflated negative binomial model with predictors being T2D, GC treatment, and a product term between the two. A product term estimate > 1 represents a larger effect of GC among the T2D group than in the non-T2D group. We included 477,605 individuals (27,942 with T2D) with a total follow-up time of 1,358,739 person-years. A total of 21,581 MOFs were observed. Median follow-up time was 35 days (with a mean of 212 days) during GC treatment and 731 days in the pre-treatment period. After weighting, GC treatment was associated with an IRR of 1.70 [1.55-1.85], with no evidence of effect modification by T2D (product term IRR 1.01 [0.871.17]). These results were similar across subgroup and sensitivity analyses; when stratifying by sex (males vs. females) or age (< 65 years, 6574 years, and ≥ 75 years), when counting recurring fractures during treatment, when allowing the effects of GC treatment to vary smoothly as a function of time-since-initiation, when accounting for differences in daily dosage with different durations of treatment, and under alternative modeling specifications. We found that systemic GC use increases fracture risk to a similar degree in individuals with and individuals without T2D. While no potentiating effect was observed, individuals with T2D receiving GC treatment experience increased risk through multiple paths, and care should be taken when prescribing GCs in this population.