The bottom line in one paragraph
TRT and fracture risk became a real conversation in 2024, when the TRAVERSE bone substudy (Snyder et al., New England Journal of Medicine, January 2024, DOI 10.1056/NEJMoa2308836) reported that men randomized to transdermal testosterone had a higher rate of clinical fractures than men on placebo, despite the older Testosterone Trials data showing testosterone improves bone mineral density. Over a mean follow-up of approximately 3.2 years, clinical fractures occurred in 3.50% of the testosterone arm versus 2.46% of the placebo arm, with a hazard ratio of 1.43 (95% CI 1.04–1.97). The signal was unexpected, the absolute difference is small but real, and the mechanism is not settled. This article walks through what the substudy actually measured, why bone density and fracture rate can move in different directions, how guideline bodies have responded, and what bone monitoring on long-term TRT looks like in practice.
For the broader cardiovascular safety picture from the same parent trial, see the companion TRAVERSE trial explained article. For the lab side of long-term monitoring, the TRT blood work schedule covers the full panel and timing.
Educational disclaimer: This article summarizes published peer-reviewed data for educational purposes. It is not medical advice and is not a recommendation for or against TRT. Bone health and fracture risk are individual — decisions about starting, continuing, or modifying testosterone therapy should be made with a licensed clinician who knows your full history, current labs, and fall and fracture risk profile.
What the TRAVERSE bone substudy was
The TRAVERSE bone substudy is a pre-specified secondary analysis of the larger TRAVERSE cardiovascular safety trial. The parent trial — TRAVERSE (Lincoff et al., NEJM, June 2023, PMID 37326322) — randomized 5,246 men aged 45 to 80 with hypogonadism and either established cardiovascular disease or high cardiovascular risk to transdermal testosterone gel 1.62% or matching placebo, titrated to a mid-normal serum testosterone target.
The bone substudy was led by Peter J. Snyder of the University of Pennsylvania — the same investigator who led the original 2017 Testosterone Trials bone substudy — and was published in The New England Journal of Medicine in January 2024 (DOI 10.1056/NEJMoa2308836). It used the full TRAVERSE cohort to count incident clinical fractures, with adjudication by a blinded outcomes committee.
What was counted as a fracture
- Clinical fractures: Fractures that came to medical attention, were confirmed radiographically, and were adjudicated by the blinded clinical events committee
- Skeletal sites included: Hip, vertebrae, wrist, humerus, ribs, pelvis, ankle, foot, and other long bones — essentially the full spectrum of clinically relevant fracture sites
- Mechanism documentation: Investigators recorded fall versus non-fall, low-trauma versus high-trauma, and circumstances where available
- Excluded: Pathologic fractures attributable to malignancy and isolated vertebral fractures detected only on routine imaging without symptoms
That definition matters. The substudy was tracking real-world clinical fractures — events that a man notices, presents to care for, and is treated for — not the radiographic-only vertebral microfractures that some bone trials count. The endpoint is patient-relevant.
The headline finding: more fractures on TRT
Over a mean follow-up of approximately 3.2 years, clinical fractures occurred in 91 of 2,601 men in the testosterone group (3.50%) and 64 of 2,603 men in the placebo group (2.46%). The hazard ratio was 1.43, with a 95% confidence interval of 1.04 to 1.97 — a result that reaches conventional statistical significance. In absolute terms, this is roughly one additional clinical fracture per 100 men treated over three years.
The number that drove the headlines
Most headlines led with the relative risk — "43% higher fracture risk on TRT." That framing is technically accurate but absolute risk is the more useful number for individual decision-making. A 1.04 percentage-point absolute difference is what a man and his prescriber are actually deciding about. The investigators reported both numbers; honest summary requires both.
Clinical fracture rates from the TRAVERSE bone substudy primary analysis (Snyder et al., NEJM 2024, DOI 10.1056/NEJMoa2308836).
What kinds of fractures occurred
The substudy reported fracture distribution by site, which matters because hip and vertebral fragility fractures behave differently from wrist, ankle, and rib fractures clinically and prognostically.
Distribution across sites
Fractures were spread across the full skeletal map rather than concentrated at classical osteoporotic sites. Rib, wrist, ankle, and foot fractures were prominent in both arms, consistent with falls and physical-activity mechanisms. Hip fractures and vertebral fractures, the most concerning fragility sites, did not show an isolated dramatic excess that would have suggested rapid bone loss in the testosterone group.
Mechanism: most fractures were trauma-related
The investigators recorded the circumstances of injury where available. The majority of fractures in both arms were associated with a fall or other identifiable mechanical cause rather than spontaneous fragility events. That is part of why the "testosterone weakens bones" framing does not fit the data well. Bones did not fail under low loads — bodies fell, and bones broke under those falls. Whether testosterone changes the rate of falls is the live mechanistic question.
Why fracture site matters:Vertebral and hip fractures are the canonical osteoporotic fragility sites and predict mortality and disability. Wrist, rib, and ankle fractures from falls are common but generally have better functional outcomes. The TRAVERSE substudy distribution looked more like a fall-driven pattern than an osteoporosis-driven pattern, which shifts the mechanistic conversation away from "bone strength" toward "fall risk and exposure."
The bone mineral density paradox
The single most discussed feature of the TRAVERSE bone substudy is the apparent paradox between bone density imaging and clinical fracture rate. The Testosterone Trials bone substudy (Snyder et al., JAMA Internal Medicine, 2017, PMID 28241248) had measured volumetric bone mineral density and estimated bone strength on quantitative CT in 211 older hypogonadal men randomized to testosterone or placebo for one year. Testosterone increased spine and hip volumetric BMD and estimated bone strength relative to placebo, with the largest gains at the spine.
Those imaging gains were the basis for the prevailing view that TRT helps bone. The 2024 TRAVERSE bone substudy did not contradict that imaging signal — it observed a different endpoint, on a different timescale, in a different (though overlapping) population, and saw more clinical fractures.
How density and fractures can decouple
- BMD predicts fracture risk imperfectly even in osteoporosis trials. The relationship between BMD change and fracture reduction in osteoporosis treatment trials is real but modest. Drugs that improve BMD reduce fractures by less than the BMD gain alone would predict — fall biomechanics, bone microarchitecture, and behavior all contribute.
- The studied populations differ. The 2017 Testosterone Trials bone substudy enrolled 211 men with mean age around 72. TRAVERSE enrolled 5,246 men aged 45 to 80 with cardiovascular risk. The TRAVERSE cohort was younger on average and selected for cardiovascular comorbidity — a population in which mechanical and behavioral factors may dominate.
- Time horizon differs. The 2017 imaging substudy was 12 months. The TRAVERSE clinical fracture follow-up averaged 3.2 years — long enough to accumulate falls and accidents in a way short imaging studies cannot capture.
- Behavioral mediation is plausible. If testosterone restores energy, ambition, and physical activity, fall exposure rises mechanically. A man who returns to weekend hiking, ladder work, or cycling has more chances to break a wrist than the same man sitting on a couch.
What the parallel cardiovascular finding suggests
TRAVERSE also reported small numerical increases in pulmonary embolism, atrial fibrillation, and acute kidney injury in the testosterone arm despite a non-inferior MACE primary endpoint. The pattern across multiple endpoints — small absolute increases in specific adverse events while the headline composite endpoint was reassuring — is consistent with TRT having a real but narrow effect on certain physiologic systems rather than being either uniformly beneficial or uniformly harmful. The bone substudy fits that pattern.
Biological plausibility — why this could be real
The substudy did not test mechanism. The investigators discussed several candidate explanations but explicitly declined to claim a definitive cause. The candidate mechanisms below are hypotheses worth knowing rather than confirmed pathways.
Activity and exposure
If TRT improves energy and motivation in symptomatic hypogonadal men, total physical activity and the variety of activities both rise. Higher activity equals higher fall opportunity. This is the most parsimonious explanation that fits the trauma-driven fracture distribution observed in TRAVERSE — bones did not fail spontaneously, men fell more.
Vestibular, vision, or postural changes
Some androgens have been examined for effects on inner ear function and postural control. Evidence is preliminary, but a small adverse effect on balance or reaction time would be enough to nudge fall rate upward across a large cohort. TRAVERSE was not designed to measure balance, gait speed, or fall frequency, so this hypothesis remains speculative.
Sleep and alertness shifts
TRT can worsen obstructive sleep apnea in some men (covered in our companion piece on TRT and sleep apnea), and sleep fragmentation independently raises fall risk through daytime drowsiness and reaction-time impairment. Whether this contributed to TRAVERSE fractures is unknown but biologically reasonable.
Bone quality vs density
Bone mineral density on DEXA or quantitative CT measures mineralization. It does not capture all aspects of bone quality — cortical porosity, trabecular microarchitecture, collagen cross-linking, and matrix mineralization heterogeneity all affect fracture resistance. A scenario in which testosterone improves measured BMD while subtly shifting microarchitecture in an unfavorable direction is not impossible. Direct evidence is lacking.
What the substudy does and does not establish
Honest interpretation requires being clear about what TRAVERSE bone substudy proves and what it leaves open.
What the substudy establishes
- In men aged 45–80 with hypogonadism and cardiovascular risk, clinical fractures occurred about 1 percentage point more often on transdermal testosterone over ~3.2 years
- The fracture excess reached statistical significance (HR 1.43, 95% CI 1.04–1.97)
- Most fractures were trauma-associated rather than spontaneous fragility events
- The signal exists despite prior imaging evidence that testosterone improves BMD
- Clinicians cannot assume that improved BMD on imaging guarantees fewer fractures on testosterone therapy
What the substudy does not establish
- Whether the same fracture excess applies to younger men, men without cardiovascular risk, or men on intramuscular injection protocols
- Whether the mechanism is fall risk, bone microarchitecture, vestibular effect, or unmeasured confounding
- Whether different testosterone formulations or doses produce the same signal
- Whether the fracture rate continues to diverge, plateaus, or narrows over follow-up beyond 3.2 years
- Whether targeted fall prevention or balance training would attenuate the excess
Interpretation note:The TRAVERSE bone substudy is the most reliable randomized evidence on TRT and clinical fractures we have. It does not justify the framing "testosterone causes osteoporosis." It does justify the framing "in older men with cardiovascular risk on transdermal TRT for ~3 years, clinical fractures occurred slightly more often than on placebo, and we do not yet know the mechanism." That distinction matters for clinical decisions and for honest patient counseling.
How this compares to prior testosterone bone data
The pre-TRAVERSE bone literature on TRT was overwhelmingly centered on bone mineral density rather than fractures. That is partly because BMD is easier to measure in trials of practical size and duration, and partly because no large randomized trial had been powered specifically for fracture endpoints in TRT-treated men.
Key prior findings
| Study | Endpoint | Direction |
|---|---|---|
| Snyder et al., JAMA Internal Medicine 2017 (PMID 28241248) — Testosterone Trials bone substudy | Volumetric BMD and estimated bone strength on QCT, 12 months | Increased on testosterone vs placebo |
| Isidori et al., Clinical Endocrinology 2005 (PMID 15730420) — meta-analysis | Lumbar spine BMD on DEXA | Increased on testosterone in hypogonadal men |
| Tracz et al., JCEM 2006 (PMID 16720668) — meta-analysis | BMD across multiple sites | Modest increases on testosterone, heterogeneous by site |
| Snyder et al., NEJM 2024 (DOI 10.1056/NEJMoa2308836) — TRAVERSE bone substudy | Incident clinical fractures, mean 3.2 years | Higher on testosterone (HR 1.43, 95% CI 1.04–1.97) |
The honest synthesis
Imaging endpoints suggest testosterone is good for bone tissue. The largest randomized clinical-fracture trial suggests testosterone is associated with more clinical fractures over three years in older men with cardiovascular risk. Both findings can coexist if fracture risk is mediated by factors other than mineral density alone — most plausibly fall exposure and behavior — and if the populations and time horizons differ. The pre-TRAVERSE confidence that "TRT protects bone" was based on density data only and now needs the asterisk that clinical fractures may move in a different direction.
How guideline bodies have responded
The major endocrine and urology bodies have acknowledged the TRAVERSE bone substudy without dramatically changing their core TRT recommendations. The integration is ongoing — guidelines move slower than primary publications.
Endocrine Society
The Endocrine Society's 2018 clinical practice guideline (Bhasin et al., JCEM, DOI 10.1210/jc.2018-00229) predates the TRAVERSE bone substudy. The guideline already recommended baseline DEXA in men with low testosterone and a history of low-trauma fracture or osteoporosis, and noted that testosterone increases BMD in hypogonadal men. Post-TRAVERSE bone substudy commentary from Endocrine Society members has called for incorporating the fracture data into shared decision-making and for reconsidering whether the BMD-focused framing has overstated bone benefit. A formal guideline update reflecting the 2024 fracture data is anticipated but not yet final.
American Urological Association
The AUA testosterone deficiency guideline (Mulhall et al., originally 2018, amended 2024) integrates the parent TRAVERSE cardiovascular safety data and acknowledges the bone substudy fracture signal in commentary. The AUA continues to recommend individualized risk-benefit discussion and has not added universal DEXA screening for TRT patients. The guideline language around bone has shifted from "testosterone improves bone health" toward more measured statements about BMD gains coexisting with the new fracture signal.
Bone-specific organizations
The American Society for Bone and Mineral Research and Endocrine Society joint discussions following the 2024 publication have emphasized that the TRAVERSE bone substudy does not justify withdrawing TRT from men with osteoporosis or low BMD, but does justify clearer counseling about fall risk and clinical fracture risk during informed consent. Fall prevention guidance — strength training, balance work, vision and medication review, home hazard assessment — has gained emphasis in the post-substudy commentary.
What practical bone monitoring looks like on TRT
Post-TRAVERSE bone substudy, practical bone monitoring on TRT remains individualized rather than protocol-driven. Most prescribers continue to follow Endocrine Society and AUA frameworks while adding fall-risk attention.
Pre-TRT considerations
- Baseline DEXA: Generally indicated in men with prior low-trauma fracture, known osteoporosis, prolonged hypogonadism (multi-year low T), chronic glucocorticoid use, or other secondary osteoporosis risks. Not universally required for every TRT candidate.
- Fall risk screen: Brief assessment of prior falls, gait, balance, vision, and medications that affect alertness — particularly important in men over 65
- Vitamin D and calcium status: Standard pre-TRT panels often include 25-hydroxyvitamin D; deficiency is correctable and matters for bone independent of testosterone
- Documentation: Baseline fracture history, family history of osteoporosis, lifestyle factors (alcohol, smoking, activity level)
On-therapy monitoring
- Routine bone-specific monitoring: No major society currently recommends universal serial DEXA for every TRT patient. For men with baseline osteoporosis or fragility fracture history, follow-up DEXA at 1–2 year intervals is reasonable.
- Standard TRT labs: The full TRT blood work schedule covers what is monitored on therapy — testosterone, hematocrit, PSA, lipids, A1C, sensitive estradiol — none of which directly assess bone but which inform overall risk stratification
- Fall conversation at follow-up: Many prescribers now ask explicitly about falls, near-falls, and any new dizziness, balance concerns, or vision changes since the last visit. This is a low-cost addition that aligns with the TRAVERSE fracture pattern.
- Activity counseling: Resistance training, balance work, and adequate protein intake support both bone density and fall resistance. These are useful adjuncts regardless of TRT status.
When to consider DEXA on therapy
- New low-trauma fracture occurs while on TRT
- Men over 70 with no prior baseline DEXA, after several years on therapy
- Development of secondary osteoporosis risk factors (chronic steroid course, malabsorption, hypogonadism recurrence on inadequate dose)
- Patient or clinician preference based on personal risk profile
Pro tip on counseling: The most useful single addition to TRT informed consent post-TRAVERSE bone substudy is a brief, honest mention that clinical fractures occurred about 1 percentage point more often on testosterone than placebo over ~3 years in the largest randomized trial, that the mechanism is not settled, and that fall prevention and resistance training help. That framing respects the data, supports patient autonomy, and avoids both alarmism and false reassurance.
Who should pay closest attention
The TRAVERSE bone substudy applies most directly to men resembling the trial population. Some men have additional reasons to weigh the fracture signal more heavily.
Men aged 65 and older
Age is the dominant fracture risk factor regardless of testosterone status. Men in their late 60s, 70s, and 80s start with higher baseline fall and fragility risk. A 1 percentage point absolute increase on top of an already elevated baseline rate translates into more meaningful absolute risk than the same relative increase in a 50-year-old.
Men with prior fragility fractures
A previous low-trauma fracture is one of the strongest predictors of a subsequent fragility fracture. Men with prior hip, vertebral, or wrist fragility fractures should have a documented bone health discussion before starting TRT and benefit from structured fall prevention regardless of TRT decision.
Men with established osteoporosis
Men already diagnosed with osteoporosis on DEXA are not excluded from TRT candidacy, but the conversation should integrate bone-active therapy considerations (bisphosphonates, denosumab, anabolic agents like teriparatide) where indicated. TRT is not a substitute for bone-targeted therapy in men with osteoporosis.
Men on supraphysiological doses
TRAVERSE used transdermal gel titrated to mid-normal serum testosterone. The substudy does not directly characterize what happens at supraphysiological doses or with sharp peak-to-trough swings of high-dose intramuscular injection regimens. For broader context on dosing, our 200mg testosterone per week article discusses where physiological replacement ends and supraphysiological territory begins.
Men with sleep apnea
Untreated obstructive sleep apnea contributes to fall risk through daytime drowsiness, and TRT can worsen sleep apnea in some men. Coupling those two effects raises fall plausibility. Our TRT and sleep apnea article covers screening and CPAP synergy in detail.
Plain-English takeaway for men on long-term therapy
If you are a man on TRT or considering it and you came here asking whether testosterone causes fractures, here is the honest synthesis the data support.
- The largest randomized trial found about 1 extra clinical fracture per 100 men treated over three years. That is the absolute risk number to anchor on. The relative 43% framing is real but absolute risk drives decisions.
- Bone density on imaging tends to improve on TRT. The 2017 Testosterone Trials bone substudy is not wrong. The 2024 TRAVERSE bone substudy is not wrong. The two are measuring different things, and clinical fractures matter more for daily life than imaging numbers.
- Most fractures in TRAVERSE were trauma-related, not spontaneous fragility events. That points more toward fall exposure than toward bones losing strength. Mechanism is not proven, but the pattern matters for what to do about it.
- Fall prevention is the practical lever. Resistance training, balance work, vision and medication review, home hazard assessment, and treating any contributing sleep apnea are useful regardless of TRT status and may be especially worth doing if you stay on therapy long-term.
- Universal DEXA on every TRT patient is not standard. Baseline DEXA is reasonable for men with prior fragility fracture, prolonged hypogonadism, age over 70, or other secondary osteoporosis risks. Talk with your prescriber about whether your profile warrants it.
- The fracture signal does not invalidate TRT. It adds nuance. Men who are correctly diagnosed with hypogonadism and who derive symptomatic benefit from carefully monitored TRT do not need to stop based on TRAVERSE bone data alone. They do benefit from honest counseling about the fracture rate, fall prevention, and individualized bone monitoring.
- Watch the literature for the next several years. Subgroup analyses, longer follow-up, mechanistic studies, and replication in other cohorts will refine what the substudy means. The field is actively integrating it.
A brief scenario
Consider a 67-year-old man with hypogonadism, controlled hypertension, no prior fractures, and a sedentary baseline who starts TRT and finds himself energized enough to return to weekly tennis and weekend home repair projects. His bone density on DEXA improves over two years. In the third year, he falls off a step stool while clearing gutters and breaks his wrist. Pre-TRAVERSE bone substudy, this would have been classified as a freak accident. Post-substudy, the more honest interpretation is that his exposure to fall scenarios increased meaningfully on therapy, his bones held up at the imaging level, and the wrist fracture is part of the small absolute excess the trial documented. The clinical recommendation that follows is not "stop TRT" but "keep going, add a balance and grip-strength program, reconsider ladders and step stools, get the cataract evaluated, and document the fracture."
Where this fits in the broader TRT conversation
The TRAVERSE bone substudy is one piece of a larger TRAVERSE evidence base that has reshaped TRT counseling since 2023. The cardiovascular safety primary endpoint (covered in our TRAVERSE trial explained piece) was reassuring on MACE but flagged pulmonary embolism, atrial fibrillation, and acute kidney injury signals. The bone substudy added the fracture finding. The PSA and prostate cancer secondary endpoints were reassuring. The honest aggregate message is that TRT in middle-aged and older men with cardiovascular risk is not the cardiovascular catastrophe the 2015 FDA labeling implied, but it carries small, real, specific signals that reward careful monitoring rather than complacency.
Next steps if you want to go deeper: The primary publication is Snyder PJ, Bhasin S, Cunningham GR, et al. "Lessons from the Testosterone Trials." NEJM 2024 — full text is freely accessible via the journal. For the cardiovascular companion data from the same parent trial, see our TRAVERSE trial explained. For the lab side of long-term monitoring, the TRT blood work schedule covers what gets retested and when. For the broader safety framework — hematocrit, sleep apnea, cardiovascular monitoring, and estrogen — the side effects pillar guide integrates these into a coherent monitoring approach.
Sources referenced in this article
- Snyder PJ, Bhasin S, Cunningham GR, et al. "Effects of Testosterone Treatment on Clinical Fractures in Older Men with Hypogonadism" (TRAVERSE bone substudy). New England Journal of Medicine. 2024. DOI 10.1056/NEJMoa2308836.
- Lincoff AM, Bhasin S, Flevaris P, et al. "Cardiovascular Safety of Testosterone-Replacement Therapy" (TRAVERSE primary results). New England Journal of Medicine. 2023;389(2):107-117. PMID 37326322.
- Snyder PJ, Kopperdahl DL, Stephens-Shields AJ, et al. "Effect of Testosterone Treatment on Volumetric Bone Density and Strength in Older Men with Low Testosterone: A Controlled Clinical Trial" (Testosterone Trials bone substudy). JAMA Internal Medicine. 2017;177(4):471-479. PMID 28241248.
- Bhasin S, Brito JP, Cunningham GR, et al. "Testosterone Therapy in Men With Hypogonadism: An Endocrine Society Clinical Practice Guideline." Journal of Clinical Endocrinology & Metabolism. 2018;103(5):1715-1744. DOI 10.1210/jc.2018-00229.
- Mulhall JP, Trost LW, Brannigan RE, et al. "Evaluation and Management of Testosterone Deficiency: AUA Guideline." American Urological Association. Originally 2018, amended 2024.
- Isidori AM, Giannetta E, Greco EA, et al. "Effects of Testosterone on Body Composition, Bone Metabolism and Serum Lipid Profile in Middle-Aged Men: A Meta-Analysis." Clinical Endocrinology. 2005;63(3):280-293. PMID 15730420.
- Tracz MJ, Sideras K, Boloña ER, et al. "Testosterone Use in Men and Its Effects on Bone Health: A Systematic Review and Meta-Analysis of Randomized Placebo-Controlled Trials." Journal of Clinical Endocrinology & Metabolism. 2006;91(6):2011-2016. PMID 16720668.
This article was written by the TRT FAQ Editorial Team and reviewed against the published NEJM TRAVERSE bone substudy primary paper, the parent TRAVERSE cardiovascular publication, the Endocrine Society 2018 clinical practice guideline, and the AUA 2018/2024 testosterone deficiency guideline. It is educational content, not medical advice. Last content review: May 2026.
Frequently Asked Questions
Does TRT cause fractures?
The TRAVERSE bone substudy (Snyder et al., New England Journal of Medicine, 2024, DOI 10.1056/NEJMoa2308836) reported a higher rate of clinical fractures in the testosterone group compared with placebo over a mean follow-up of about 3.2 years. Clinical fractures occurred in 3.50% of men on testosterone versus 2.46% on placebo (hazard ratio 1.43, 95% CI 1.04–1.97). The finding was unexpected because testosterone improved bone mineral density on imaging in the parent Testosterone Trials. The substudy did not establish that testosterone directly weakens bone — fall risk and behavioral factors are plausible mediators — but the absolute increase is real and is now part of the TRT risk-benefit conversation.
Does testosterone weaken bones?
Testosterone does not appear to weaken bone tissue itself. In the original Testosterone Trials bone substudy (Snyder et al., JAMA Internal Medicine, 2017, PMID 28241248), one year of testosterone increased volumetric bone mineral density and estimated bone strength on quantitative CT in older hypogonadal men. The TRAVERSE bone substudy (NEJM 2024) reproduced no signal of bone loss on imaging — yet still observed more clinical fractures. The current working hypothesis is that fracture risk on TRT is mediated by factors other than bone density alone, including possible increases in physical activity, fall risk, or unmeasured confounders. Bone density is not the same outcome as fracture, and TRAVERSE is the trial that made that gap concrete.
Is TRT safe for bone density?
Imaging-based bone outcomes are reassuring on TRT in hypogonadal men. Multiple controlled trials, including the original Testosterone Trials bone substudy (Snyder JAMA IM 2017, PMID 28241248) and the meta-analysis by Isidori et al. in Clinical Endocrinology (2005, PMID 15730420), have shown that testosterone replacement increases lumbar spine bone mineral density in men with low testosterone. The TRAVERSE bone substudy did not contradict the BMD literature. What it added was the unexpected fracture signal — a clinical outcome that bone density alone does not fully predict. For most men on appropriately monitored TRT, density is not the worry; fall risk, balance, and overall musculoskeletal health are the levers that probably matter more.
Why did TRAVERSE show more fractures on testosterone?
The TRAVERSE investigators (Snyder NEJM 2024) explicitly flagged the finding as unexpected and discussed several candidate explanations. Testosterone may increase physical activity, exposure to fall situations, or risk-taking behavior in older men, raising mechanical fracture risk independent of bone strength. Vestibular or vision changes are speculative. The trial was not designed to test mechanism, and the substudy authors did not claim a definitive cause. What the trial did establish is the absolute fracture rate difference (3.50% vs 2.46% over ~3.2 years), the hazard ratio (1.43), and the fact that the imaging endpoints did not explain the clinical signal. Mechanism is open; the rate difference is the data.
How big is the absolute fracture risk difference on TRT?
Over a mean follow-up of about 3.2 years in TRAVERSE, the absolute difference in clinical fractures between testosterone and placebo was approximately 1 percentage point — 3.50% versus 2.46%. That translates to roughly 1 additional clinical fracture per 100 men treated over three years in the studied population. The men enrolled were aged 45 to 80 with hypogonadism and pre-existing cardiovascular risk, so the absolute number does not necessarily extrapolate to younger men or men without those baseline risks. For an individual decision, the relevant question is whether that absolute difference is meaningful given the symptomatic benefit the man expects from TRT and his specific fall and fracture risk profile.
Should men on long-term TRT get DEXA scans?
Endocrine Society 2018 clinical practice guidance (Bhasin et al., JCEM, DOI 10.1210/jc.2018-00229) recommends DEXA bone density assessment in men with low testosterone and a history of low-trauma fracture or osteoporosis, but does not mandate routine DEXA in every man on TRT. Post-TRAVERSE bone substudy, no major society has changed that baseline recommendation to universal DEXA screening for TRT patients. In practice, prescribers may consider baseline DEXA in older men, men with a personal or family history of fragility fracture, men with prolonged hypogonadism prior to treatment, or men with additional secondary osteoporosis risk factors (chronic steroid use, alcohol excess, malabsorption). Individualized assessment, not universal screening, is the current standard.
Did the TRAVERSE bone substudy contradict the older Testosterone Trials bone results?
Not directly. The 2017 Testosterone Trials bone substudy (Snyder JAMA IM, PMID 28241248) measured volumetric bone mineral density and estimated bone strength on quantitative CT after 12 months of testosterone in 211 men, and reported significant improvements over placebo. The 2024 TRAVERSE bone substudy (Snyder NEJM, DOI 10.1056/NEJMoa2308836) measured a different endpoint — incident clinical fractures — in a much larger cohort followed for ~3.2 years. The two studies asked different questions. TRAVERSE did not show that testosterone harms bone tissue; it showed that the clinical fracture rate was higher despite no documented harm to density. The honest synthesis is that bone strength imaging and fracture incidence are partially decoupled, and clinicians should consider both.