Trabecular bone score and bone mineral density in systemic sclerosis: Prevalence of low bone quality and associated clinical factors - a case-control study.
BACKGROUND Systemic sclerosis (SSc) is associated with multifactorial bone loss and elevated vertebral fracture risk. Dual-energy X-ray absorptiometry (DXA) measures areal bone mineral density (aBMD), which reflects bone mass rather than bone strength and provides no information on trabecular microarchitecture. The trabecular bone score (TBS) provides complementary, indirect information on trabecular microarchitecture that may be independently impaired in SSc. The clinical determinants of low TBS and its relationship to vertebral fractures in SSc remain incompletely characterized. OBJECTIVES To compare TBS and BMD between SSc patients and age- and sex-matched healthy controls, overall and separately in women and men; to identify clinical factors independently associated with low TBS in SSc; and to describe vertebral fracture prevalence at 5-year radiographic follow-up. METHODS A case-control study enrolled 130 adult SSc patients and 390 age- and sex-matched healthy controls at Srinagarind Hospital, Khon Kaen University, Thailand. All participants underwent lumbar spine DXA with TBS analysis on the same densitometer. Matched comparisons of continuous outcomes used generalized estimating equations with exchangeable working correlation within matched set; matched comparisons of binary outcomes used conditional logistic regression stratified on matched set. Effect modification by sex was assessed with formal SSc-by-sex interaction terms. Factors associated with low TBS (TBS <1.350) were examined by univariate and multivariable logistic regression, with odds ratios, confidence intervals and p-values derived from the same model. Vertebral fractures were graded by Genant's semiquantitative method at 5-year follow-up in 80 patients with radiographs; after pre-analysis exclusion of two patients with tuberculous vertebral destruction, 78 patients contributed to the fracture analysis. RESULTS Low TBS was significantly more prevalent in SSc patients than controls (53.1% vs. 34.1%, p < 0.001). Mean BMD was significantly lower at all skeletal sites in SSc patients (all p < 0.001). The excess of low TBS in SSc was significant in women (63.2%vs. 37.9% of controls; OR 3.77, 95% CI 2.13-6.69) but not in men (25.7%vs. 23.8%; OR 1.12, 95% CI 0.45-2.78), and the SSc-by-sex interaction was statistically significant, albeit marginally (p=0.047 for continuous TBS; interaction OR 3.38, 95% CI 1.15-9.95, p=0.027 for low TBS). Deficits persisted after BMI adjustment. Within SSc patients, age ≥60 years (adjusted OR 3.99, 95% CI 1.63-9.72; p = 0.002) and female sex (adjusted OR 5.29, 95% CI 2.12-13.18; p < 0.001) were each independently associated with low TBS, and both remained so under Firth penalized-likelihood estimation. Among the 78 evaluable patients, vertebral fracture prevalence at 5-year follow-up was 42.3% (95% CI 32.0-53.4%); fractures occurred across all BMD and TBS categories, including in 8 of 20 patients (40%) with both normal BMD and normal TBS. Of 35 patients meeting the T-score criterion for osteoporosis, 11 (31.4%) received no anti-osteoporotic treatment. CONCLUSIONS SSc patients exhibit significantly impaired bone microarchitecture compared with matched controls. Age and female sex were each independently associated with low TBS, and the microarchitectural deficit relative to controls was more marked in women, though the male subgroup was small and the interaction evidence marginal. TBS may provide information complementary to BMD in SSc, and its role in fracture risk stratification warrants prospective evaluation. The high vertebral fracture prevalence and the substantial treatment gap identify priorities for bone health care in SSc.