Abstract
Background/Objectives: Patients with myasthenia gravis (MG) are at increased risk of osteoporotic fractures due to long-term oral corticosteroid use and disease-related muscle weakness. FRAX(®) estimates 10-year fracture probability but does not incorporate falls or MG-specific functional impairment. To explore heterogeneity of fracture occurrence within MG patients classified as high risk by FRAX major osteoporotic fracture (MOF) probability. Methods: In a single-center retrospective cohort of 68 MG patients assessed in 2012, FRAX MOF with femoral neck BMD was calculable in 54 patients; the 29 patients with FRAX MOF ≥ 9.0% (the median of these 54 patients) comprised the high-FRAX cohort. Patients were stratified by the cohort medians of age (67 years) and MG-ADL (2 points) into four strata (HH, HL, LH, LL). This median-based stratification was exploratory and not intended as a clinically meaningful threshold. The primary outcome was time to first MOF (up to 10 years). We compared fracture occurrence using both proportions and Kaplan-Meier analyses (log-rank test) and performed exploratory univariable Cox models for selected predictors. No multivariable confounder adjustment was performed. Results: Eight of twenty-nine patients (27.6%) experienced an MOF. The proportions with MOF were HH 25.0%, HL 40.0%, LH 57.1%, and LL 0.0% (global p = 0.068). Kaplan-Meier curves differed across strata (log-rank p = 0.03), with separation most evident between LH and LL. For univariable Cox analyses, age was associated with shorter time to MOF (hazard ratio [HR] 1.13 per year, p = 0.041), and baseline difficulty rising from a chair (MG-ADL item) was associated with higher hazard rates (HR 3.45, p = 0.048). Conclusions: In this small, selected high-FRAX MG cohort, fracture events appeared to cluster in patients with impaired ADL and fall-related MG-ADL abnormalities, whereas FRAX values remained strongly age-driven. These findings are exploratory and hypothesis-generating and should not be interpreted as evidence of FRAX miscalibration; confirmation in larger, prospectively followed cohorts is needed.