Men with prostate cancer who had a record of androgen deprivation therapy were more likely to have a vertebral compression fracture within five years than matched men without a recorded therapy exposure, according to a large electronic-health-record study. The fracture rate was 1.95% in the therapy group, compared with 1.29% in the comparison group. That is an absolute difference of 0.66 percentage points, so the gap was small in absolute terms.
The study also found more use of vertebral augmentation among men who developed these fractures. Utilization was 7.4% among those with recorded androgen deprivation therapy and 5.1% among those without it, a 2.3 percentage-point difference and a 45% relative increase.
The gap appeared early and widened over time
The higher fracture rate was present at every time point the researchers reported. At three months, vertebral compression fractures occurred in 0.47% of the therapy group and 0.34% of the comparison group. At one year, the figures were 0.93% and 0.60%; at three years, 1.63% and 1.03%; and at five years, 1.95% and 1.29%. The reported p-value was below 0.01 at each time point.
Among patients who had a fracture, the difference in augmentation was also visible at shorter follow-up. Rates were 5.64% versus 4.24% at three months and 6.53% versus 4.44% at six months, comparing the therapy and non-therapy groups. The reported p-value was 0.0134 at three months and below 0.01 at six months.
How the comparison was made
Researchers examined de-identified electronic-health-record data from more than 170 healthcare organizations, covering 2008 through 2025. The analysis began with 1,222,622 adult men with prostate cancer. After matching, the main comparison included 122,511 men with recorded androgen deprivation therapy and 122,511 without a recorded exposure.
The study was a retrospective cohort analysis, meaning the researchers looked back at existing records rather than assigning treatment. They used propensity-score matching, a statistical technique that pairs people with similar recorded characteristics. Each therapy patient was matched with one comparison patient, using information on demographics, medical conditions and bone-health indicators. All standardized mean differences, a measure of how well the groups were balanced, were below 0.1 after matching.
The researchers included adults aged 18 or older with prostate cancer and separated them according to whether their records showed androgen deprivation therapy. Long-term systemic corticosteroid users were excluded. People with a prior vertebral compression fracture were excluded from the main fracture analysis, and those with prior augmentation were excluded from the post-fracture analysis.
Other outcomes showed a less clear pattern
The higher augmentation rate did not extend to every measure of later care. Posterior spinal instrumentation occurred in 2.1% of the therapy group and 2.0% of the comparison group, while dependency was recorded in 7.5% and 7.7%, respectively. The study reported no statistically significant difference for either outcome.
Among patients with vertebral compression fractures, the average number of fracture events per patient was 3.48 with recorded therapy and 3.17 without it. The difference was not statistically significant. A separate analysis of intravenous bisphosphonate exposure also found no statistically significant difference in fracture incidence. At five years, the reported rates were 2.09% in the exposure group and 2.13% in the non-exposure group; at one year, 1.12% and 1.0%; and at three years, 1.79% and 1.87%.
The findings show an association, not a cause
Because the study used medical records retrospectively, it cannot establish that androgen deprivation therapy caused the fractures or the procedures. Matching can balance measured characteristics, but it cannot remove every difference between the groups. The database may also contain coding errors, and medication records do not show whether patients took the drugs as prescribed, at what dose or for how long.
The records did not include imaging findings, fracture severity, symptoms, bone mineral density or the specific reason an augmentation procedure was chosen. The researchers also noted that occult metastatic disease could not be fully excluded and that more frequent cancer imaging in the therapy group might have led to more fractures being detected or treated. The study population may not represent patients treated in smaller practices.
The authors said the pattern was consistent with bone fragility and a potentially preventable procedural burden, and supported bone-health screening and preventive therapy when androgen deprivation therapy begins. They also identified surveillance and confounding by indication as alternative explanations. Prospective research is still needed to determine whether prevention at treatment initiation reduces fractures and augmentation procedures, and whether better use of bone-active therapy changes outcomes.
Paper data and sources
Original title: Fracture and Vertebral Augmentation Utilization Following Androgen Deprivation Therapy in Men With Prostate Cancer: A Propensity Score–matched Cohort Study
Authors: Taka T, Cummings C, Meng S et al.
Journal/Repository: Global Spine Journal
Status: Peer-reviewed
First online: 2026-10-13
DOI: Not available
Original paper · Full text