8377 Associations of Serum Testosterone and Sex Hormone-Binding Globulin with Incident Fractures in Middle-Aged to Older Men

• Published in: Journal of the Endocrine Society Vol. 8; no. Supplement_1
• Main Authors: Grahnemo, Louise, Marriott, Ross J, Murray, Kevin, Tyack, Lauren, Nethander, Maria, Matsumoto, Alvin M, Orwoll, Eric S, Vanderschueren, Dirk, Yeap, Bu Beng, Ohlsson, Claes
• Format: Journal Article
• Language: English
• Published: US Oxford University Press 05-10-2024
• Subjects: Abstract
• ISSN: 2472-1972; 2472-1972
• DOI: 10.1210/jendso/bvae163.1565

Disclosure: L. Grahnemo: None. R.J. Marriott: None. K. Murray: None. L. Tyack: None. M. Nethander: None. A.M. Matsumoto: None. E.S. Orwoll: None. D. Vanderschueren: None. B.B. Yeap: None. C. Ohlsson: None.
As men age, their circulating testosterone (T) decreases, circulating sex-hormone binding globulin (SHBG) increases, and risk of bone fracture increases. Unexpectedly, a recent large clinical trial reported that T treatment increased fracture risk in men. It is unclear if circulating T, independently of comorbidities, is associated with male fracture risk. We tested the hypothesis that serum T concentration is an independent risk predictor for incident bone fractures in men. We analyzed associations for T and SHBG with incident fractures in the large and well-characterized UK Biobank cohort (205,973 male participants median age 58 years, excluding men on androgen-related therapies). Baseline serum total T and SHBG were measured by immunoassay, and incident fractures documented over a median follow-up of 13.6 years (11,088 any fracture cases; 1,680 hip fracture cases; 1,366 forearm fracture cases). Associations of hormonal variables with fracture outcomes were modelled using Cox regression, adjusted for multiple comorbidities/covariates, with imputation for missing data, and non-linearity assessed using cubic splines. For total T, in multivariable models not including SHBG, there was a non-linear association with any fracture and hip fractures but not forearm fractures, with the lowest risk in the second quintile (Q2; hazard ratio (HR), 95% confidence intervals (CIs) Q2 vs Q5 any fracture 0.89, 0.84-0.94; hip fracture 0.76, 0.66-0.87; forearm fracture 0.97, 0.83-1.13). However, in multivariable models analyzing total T which were adjusted for SHBG, lower total T was associated with higher risk for fractures at all evaluated bone sites (HR, 95% CIs Q1 vs Q5 any fracture 1.24, 1.16-1.33; hip fracture 1.28, 1.08-1.51; forearm fracture 1.51, 1.25-1.82). Similar results were seen in multivariable models using calculated free T values. Lower SHBG concentration was associated with lower risk of any fracture, hip fractures, and forearm fractures, in multivariable models not including total T (HR, 95% CIs Q1 vs Q5, any fracture 0.71, 0.67-0.75; hip fracture 0.55, 0.47-0.65; forearm fracture 0.62, 0.52-0.74), with similar results following additional adjustment for total T. In conclusion, associations of total T concentrations with fracture risk without adjustment for SHBG are non-linear and inconsistent, whereas lower total T concentration, adjusted for SHBG, was independently associated with higher risks of any fracture, hip fractures, and forearm fractures. Lower circulating SHBG, without or with adjustment for total T, was strongly associated with lower risk of fractures at all three sites, consistent with SHBG being a major independent biomarker of fracture risk in men. When examining the relationship of endogenous T concentrations with health outcomes influenced by androgen-sensitive tissues in middle-aged to older men, both total T and SHBG concentrations should be assessed.
Presentation: 6/2/2024