U-Shaped Association Between Testosterone and Fractures

Nelson Vergel · Mar 12, 2020

Interesting. I wonder if the higher T levels cause either more fractures due to more exposure to activities that can cause it (due to higher stamina) or if higher hematocrit can decrease blood flow to bones.

Jinzang · Mar 13, 2020

Whenever I see I u-shaped result, I'm a little suspicious. I'd like to see this study replicated.

Vince · 2025-07-12T10:03:42-0500

We now know that low levels of testosterone does cause issues. High levels we don't know, high levels are definitely much better than low levels.
When I have my next wellness exam. I'm going to get another bone density test.

Nelson Vergel · 2025-07-12T10:34:10-0500

Thread 'Testosterone Treatment and Fractures in Men with Hypogonadism'

Jan 18, 2024

https://www.nejm.org/doi/pdf/10.1056/NEJMoa2308836

Background

Testosterone treatment in men with hypogonadism improves bone density and quality, but trials with a sufficiently large sample and a sufficiently long duration to determine the effect of testosterone on the incidence of fractures are needed.

Methods

In a subtrial of a double-blind, randomized, placebo-controlled trial that assessed the cardiovascular safety of testosterone treatment in middle-aged and older men with hypogonadism, we examined the risk of clinical fracture in a...

Nelson Vergel · 2025-07-12T10:35:18-0500

I find the findings about high T very strange.

Briefing Document: Association of Plasma Testosterone and SHBG with Fracture Incidence in Older Men

Source: Yeap, B. B., et al. "U-Shaped Association of Plasma Testosterone, and no Association of Plasma Estradiol, with Incidence of Fractures in Men." J Clin Endocrinol Metab, vol. 105, no. 5, 2020, pp. 1489–1500.

Date: May 2020

Purpose: To assess the associations of various sex hormones with the incidence of any bone fracture or hip fracture in older men, aiming to understand the role of androgens versus estrogen in maintaining bone health during male aging.

Study Design: This was an observational study utilizing data from the Health In Men Study (HIMS) cohort.

Participants: 3307 community-dwelling men from Perth, Western Australia, with a mean age of 76.8 ± 3.5 years at baseline (2001-2004). Men with known osteoporosis, previous bone fracture, Paget's disease, or receiving specific bone-affecting medications were excluded.
Follow-up: Median follow-up period of 10.6 years, extending until December 31, 2016.
Hormone Assays: Plasma testosterone (T), dihydrotestosterone (DHT), and estradiol (E2) were quantified by liquid chromatography-tandem mass spectrometry (LC-MS). Sex hormone-binding globulin (SHBG) and luteinizing hormone (LH) were measured by immunoassay.
Fracture Ascertainment: Incident fractures were determined via the Western Australian Data Linkage System (WADLS), linking to records from death, hospital, and emergency department registries. Fractures associated with falls were included, excluding pathological fractures.
Outcomes: The primary outcomes were the incidence of any fracture and the incidence of hip fracture.
Statistical Analysis: Cox proportional hazards regression was used, adjusted for age, medical comorbidities (e.g., smoking, alcohol, hypertension, diabetes, CVD, depression), frailty, BMI/waist, creatinine, and vitamin D. Nonlinear relationships were explored using restricted cubic splines.

Key Findings:

Testosterone (T) and Fracture Risk (U-Shaped Association):

The study found a "U-shaped association of plasma T with incidence of any fracture... and hip fracture." This means that men with midrange plasma testosterone concentrations had the lowest risk of fractures.
Specifically, men with T concentrations in the second (Q2) and third (Q3) quartiles had a significantly lower risk of any fracture compared to those in the lowest quartile (Q1):
Q2 vs. Q1: Fully adjusted Hazard Ratio (HR) = 0.69 (95% CI 0.51–0.94, P = .020)
Q3 vs. Q1: Fully adjusted HR = 0.59 (95% CI 0.42–0.83, P = .002)
Similarly, for hip fracture:
Q2 vs. Q1: Fully adjusted HR = 0.60 (95% CI 0.37–0.93, P = .043)
Q3 vs. Q1: Fully adjusted HR = 0.52 (95% CI 0.31–0.88, P = .015)
Men with T concentrations in the highest quartile (Q4) had a similar risk of both any fracture and hip fracture as those in the lowest quartile (Q1), reinforcing the U-shaped relationship. The risk was "not different in men with T concentrations in Q4 versus Q1."
Calculated free testosterone (cFT) results mirrored total T for any fracture but not for hip fracture, suggesting "total T appears to be the more informative measure of circulating androgen."

Sex Hormone-Binding Globulin (SHBG) and Hip Fracture Risk:

Higher SHBG concentrations were associated with an increased risk of hip fracture.
Men with SHBG in the highest quartile (Q4) had a significantly increased risk of hip fracture compared to those in the lowest quartile (Q1):
Q4 vs. Q1: Fully adjusted HR = 1.76 (95% CI 1.05–2.96, P = .033)
This association "was largely unchanged when T was included in the fully adjusted model." This finding is consistent with other studies that have linked higher SHBG with increased fracture risk.

Estradiol (E2) and Other Hormones:

"DHT, E2, and LH were not associated with fracture."
This finding regarding E2 is notable, as it "differs from previous studies" and runs "contrary to previous studies and the reasons for this are unclear." The authors postulate that "in older men, circulating E2 may not be fully indicative of E2 exposure in specific tissues."
LH was not associated with fracture risk, indicating that "circulating T not LH appears to influence fracture risk in older men."

Main Themes/Most Important Ideas:

Midrange Testosterone is Optimal for Bone Health in Older Men: The study provides strong evidence for a U-shaped association between plasma testosterone and fracture risk. This implies that both very low and very high testosterone levels may increase fracture risk in older men, with a "sweet spot" in the middle of the normal range offering the best protection against fractures. This is a significant departure from studies that primarily focused on the risks associated with low T.
Androgens, Not Estrogen, as Primary Biomarker for Fracture Risk in Older Men: The study concludes that "Circulating androgen rather than estrogen represents a biomarker for hormone effects on bone driving fracture risk." The lack of association between E2 and fracture risk, despite estrogen's known role in bone metabolism (including T's aromatization to E2), suggests that circulating E2 levels in older men might not accurately reflect tissue-level E2 exposure or its impact on bone.
SHBG as an Independent Risk Factor for Hip Fracture: Elevated SHBG is independently associated with an increased risk of hip fracture, reinforcing its role as a significant hormonal factor in male bone health, possibly by influencing androgen availability at the tissue level.
Implications for Clinical Practice and Future Research: The findings suggest that future randomized controlled trials for testosterone treatment in older men should "recruit men with baseline T in the lowest quartile of values, and explore whether the degree to which exogenous T is aromatized at the tissue level predicts its beneficial effect on bone." This highlights the complexity of testosterone's role and the need to consider its metabolism and binding in future interventions.

Limitations:

Observational Study: Causality cannot be established.
"Healthy Survivor" Effect: Participants were community-dwelling men who had already participated in an earlier study, potentially leading to a healthier cohort and possibly lower fracture rates than the general population.
Single Baseline Hormone Measurement: Hormonal status was assessed at one point, not longitudinally.
Underestimation of Fractures: Fractures not requiring emergency department visits or hospital admissions (e.g., many vertebral fractures) were likely missed, potentially undercounting the overall fracture incidence.
Lack of Bone Mineral Densitometry Data: The study did not include bone mineral density measurements, which could have provided further insight into the mechanisms.
Homogeneous Cohort: Participants were predominantly Caucasian, limiting generalizability to other ethnicities or women.

Strengths:

Large Sample Size and Long Follow-up: Allowed for a substantial number of fracture events, enhancing the precision of the analysis.
Accurate Hormone Assays: Used mass spectrometry for T, DHT, and E2, which is superior for measuring low hormone concentrations.
Comprehensive Covariate Adjustment: Adjusted for numerous potential confounding factors, including age, medical comorbidities, and frailty.
Robust Outcome Ascertainment: Utilized a comprehensive data linkage system for fracture identification.

Conclusion:

This study provides compelling evidence for a U-shaped association between plasma testosterone levels and fracture risk in older men, indicating that midrange T levels are associated with the lowest incidence of both any fracture and hip fracture. Furthermore, high SHBG levels are associated with an increased risk of hip fracture. Contrary to some previous research, circulating estradiol was not found to be associated with fracture risk in this cohort. These findings underscore the importance of circulating androgen levels, particularly total testosterone and SHBG, as crucial biomarkers for bone health and fracture risk in aging men, warranting careful consideration in clinical assessment and guiding future therapeutic research.

Nelson Vergel · 2025-07-12T10:36:57-0500

What was the purpose of this study?

The study aimed to investigate the relationship between different sex hormones (testosterone, dihydrotestosterone, estradiol), sex hormone-binding globulin (SHBG), and luteinizing hormone (LH) with the incidence of bone fractures, specifically any fracture and hip fracture, in older men. The researchers sought to understand whether androgens, distinct from estrogen, play a crucial role in maintaining bone health during male aging, which has implications for understanding osteoporosis.

Who participated in this study, and how were the data collected?

The study analyzed data from 3307 community-dwelling men from Perth, Western Australia, with a mean age of 76.8 ± 3.5 years. These men were part of the Health In Men Study (HIMS) cohort. Blood samples collected between 2001-2004 served as the baseline for hormone measurements (testosterone, DHT, estradiol, SHBG, LH) using mass spectrometry and immunoassay. Incident fractures were identified over a median follow-up period of 10.6 years using the Western Australian Data Linkage System, which links records from various health registries.

What were the key findings regarding testosterone and fracture risk?

The study found a "U-shaped" association between plasma testosterone (T) concentrations and the incidence of both any fracture and hip fracture. This means that men with midrange plasma T concentrations had the lowest risk of fractures. Both very low and very high testosterone levels were associated with an increased risk of fractures. Specifically, men with T in the second and third quartiles had significantly lower hazard ratios for fractures compared to those in the lowest quartile.

Was estradiol (E2) associated with fracture risk in this study?

Unlike some previous studies, this research found no significant association between circulating estradiol (E2) levels and the incidence of any fracture or hip fracture in older men, even after adjusting for confounding factors. This suggests that in this cohort, circulating androgen, rather than estrogen, was a more direct biomarker for hormone effects on bone driving fracture risk.

What was the role of Sex Hormone-Binding Globulin (SHBG)?

Higher levels of Sex Hormone-Binding Globulin (SHBG) were significantly associated with an increased risk of hip fracture. Men in the highest quartile of SHBG had a higher hazard ratio for hip fracture compared to those in the lowest quartile. This finding is consistent with other studies that have also linked higher SHBG to increased fracture risk in men.

How do these findings compare to previous research on sex hormones and fractures in men?

The study's findings differ from some previous research, particularly concerning estradiol. While some past studies suggested an association between lower E2 and increased fracture risk, this study did not find such a link. However, the association of higher SHBG with increased fracture risk aligns with other major studies like MrOS US and MrOS Sweden. The U-shaped association of testosterone with fracture risk, where midrange levels are optimal, is a distinct finding compared to studies that only reported an increased risk with low testosterone. The larger cohort size and extended follow-up in this study may contribute to these more precise findings.

What are the clinical implications of these results for older men?

The results suggest that maintaining midrange plasma testosterone levels might be beneficial for reducing fracture risk in older men. For older men with low-normal circulating testosterone without clear hypogonadism, the study implies that testosterone treatment could improve bone mineral density, though a randomized controlled trial specifically powered for fracture outcomes is still needed. Furthermore, future research should explore how the aromatization of exogenous testosterone into estradiol at the tissue level influences its beneficial effects on bone. The study also highlights SHBG as an important hormonal factor associated with fracture risk.

What were the limitations of this study?

Several limitations were acknowledged. As an observational study, it cannot establish causality. The cohort might have a "healthy survivor" effect because participants had engaged in previous study waves. Hormone levels were measured from a single baseline blood sample, not serially over time. The overall incidence of fractures was relatively low, and vertebral fractures might be undercounted due to their potentially asymptomatic nature or non-hospital presentation. The study also lacked bone mineral densitometry data and was primarily conducted on Caucasian men, limiting the generalizability to other ethnicities or women.

Nelson Vergel · 2025-07-12T10:39:53-0500

From another study:

We did not expect these results, because most previous studies showed that testosterone improved many measures of bone structure and quality. In studies involving men with severe hypogonadism, testosterone treatment increased areal and volumetric bone mineral density1-4 and improved many structural and mechanical measures of trabecular bone on magnetic resonance microimaging.5-7 In the Testosterone Trials, which involved older men with hypogonadism, testosterone treatment for 1 year increased volumetric bone mineral density and estimated bone strength on quantitative CT.10

Because we did not expect these results, we did not design the trial to assess possible mechanisms by which testosterone would increase the incidence of fractures, so we can only speculate about possible mechanisms. Although previous studies showed that testosterone treatment in men with hypogonadism improved many measures of bone structure, especially of trabecular bone, one study showed that testosterone treatment in men with severe hypogonadism decreased cortical bone volume fraction and cortical bone axial thickness, a measure of bone strength.7

The fact that testosterone was associated with increased fracture risk among middle-aged and older men with hypogonadism should be considered in the context of potential benefits and other risks of testosterone treatment in these men. The Testosterone Trials showed that testosterone treatment improved sexual function26 and mood26 and increased hemoglobin levels27 in older men. In the present trial, testosterone was not associated with an increased risk of major adverse cardiovascular events but was associated with increased risks of atrial fibrillation, pulmonary embolism, and acute kidney injury.15
The Fracture Trial had many strengths, including enrolling more than 5000 men who had two low morning testosterone values, as well as a randomized, placebo-controlled design and a median duration of observation for more than 3 years — a large and long trial of testosterone treatment. Other strengths were the prespecified design to inquire about fractures at every visit, collection of information about reported fractures, and adjudication of the reported fractures centrally by an experienced adjudicator.16,17

This trial also had limitations. Participants were not evaluated for organic causes of hypogonadism, so it is not known whether some men with such causes were included. Adherence to administration of testosterone or placebo was suboptimal, although it was similar in the two trial groups. The increase in serum testosterone concentrations during treatment was less than in some other studies, but the lesser increase could not explain an increase in fractures. Information about falls was not assessed, except in participants who reported fractures. Physical activity and risk taking were also not assessed. Bone density and structure were not evaluated, so the effect of testosterone on these measures cannot be compared with the results in previous studies.

We found that among middle-aged and older men with hypogonadism, testosterone treatment did not result in a lower incidence of clinical fracture than placebo. The fracture incidence was numerically higher among men who received testosterone than among those who received placebo.

TRAVERSE: https://www.nejm.org/doi/full/10.1056/NEJMoa2308836

Vince · 2025-07-12T10:47:59-0500

It would be interesting to see how many of the men were TRT. And what they're bone strength was before starting TRT.

Nelson Vergel · 2025-07-12T11:37:56-0500

Vince said:
It would be interesting to see how many of the men were TRT. And what they're bone strength was before starting TRT.

The TRAVERSE study had guys on Androgel although most were only over 400 ng/dL and not like most of us with T levels higher than that. @madman would know more.

Vince · 2025-07-12T11:50:35-0500

Nelson Vergel said:
The TRAVERSE study had guys on Androgel although most were only over 400 ng/dL and not like most of us with T levels higher than that. @madman would know more.

@madman when using Androgel does your T level usually peek at over 1000 ng/dL ?

madman · 2025-07-12T16:22:44-0500

Vince said:
It would be interesting to see how many of the men were TRT. And what they're bone strength was before starting TRT.

* The subset fracture trial within TRAVERSE included 5,204 men: 2,601 men received the testosterone gel, and 2,603 men received placebo. These numbers reflect the full-analysis population used for the fracture subtrial, which was not a separate cohort but an analysis of all randomized participants in the main trial.

* Bone density and structure were not evaluated, so the effect of testosterone on these measures cannot be compared with the results in previous studies.

Thread 'Testosterone Treatment and Fractures in Men with Hypogonadism'

Jan 18, 2024

https://www.nejm.org/doi/pdf/10.1056/NEJMoa2308836

Background

Testosterone treatment in men with hypogonadism improves bone density and quality, but trials with a sufficiently large sample and a sufficiently long duration to determine the effect of testosterone on the incidence of fractures are needed.

Methods

In a subtrial of a double-blind, randomized, placebo-controlled trial that assessed the cardiovascular safety of testosterone treatment in middle-aged and older men with hypogonadism, we examined the risk of clinical fracture in a...

PARTICIPANTS

Osteoporosis was not a criterion for entry. Among the exclusion criteria were a serum testosterone concentration of less than 100 ng per deciliter (3.5 nmol per liter) and conditions that might be worsened by testosterone treatment, such as prostate cancer, severe lower urinary tract symptoms, a hematocrit of more than 50%, and severe untreated sleep apnea.

Participants were randomly assigned in a 1:1 ratio to receive either a transdermal 1.62% testosterone gel or matching placebo gel. Randomization was stratified according to the presence or absence of preexisting cardiovascular disease. The testosterone gel was supplied in a pump bottle; each depression yielded 20.25 mg of testosterone. Participants applied the gel once per day, initially one depression of the pump bottle to each shoulder. The dose was adjusted, with the use of a prespecified algorithm,14,15 to attempt to maintain a serum testosterone concentration of 350 to 750 ng per deciliter (12.1 to 26.0 nmol perliter) and a hematocrit of less than 54%. The serum testosterone concentration was measured at weeks 2, 4, 12, 26, 52, 78, and 104 and then yearly. The dose was adjusted in participants in the placebo group to maintain blinding. Testosterone or placebo was discontinued if the serum testosterone concentration remained more than 750 ng per deciliter or the hematocrit remained more than 54% at the lowest daily dose of testosterone (20.25 mg) or if prostate cancer developed.

Assessments

Participants were asked at each in-person or telephone visit if they had had a fracture since their previous visit. If they had, they were asked about the nature of the injury and the location of the fracture or fractures; they were also asked for permission to obtain source documents, including radiology reports.

Records of reported fractures were reviewed by an adjudicator at the San Francisco Coordinating Center who was unaware of the trial-group assignments; the adjudicator was trained by the second author, who also reviewed the submitted medical records to ensure agreement with the adjudication. In a manner similar to that used in several large fracture studies, the adjudicator classified the reported fracture as follows: confirmed fracture, confirmed not to be a fracture, fracture uncertain, or insufficient documentation to determine.16,17 When documentation was insufficient, an attempt was made to obtain additional documentation, including radiographs.

Fracture End Points

The main fracture end point, which was assessed in a time-to-event analysis, was the first clinical fracture, defined as a clinical spine or non-spine fracture that was documented by imaging or surgery and confirmed by adjudication. Fractures of the sternum, fingers, toes, facial bones, and skull were excluded. Other prespecified end points were time to first non–high-impact clinical fracture; time to first clinical fracture in participants not taking a medication to treat osteoporosis; time to first non–high-impact clinical fracture in participants not taking a medication to treat osteoporosis; fracture-free survival, for which death as well as clinical fracture counted as an event; time to first clinical fracture not excluding fractures of the sternum, fingers, toes, facial bones, and skull; time to first clinical fracture not excluding those classified as uncertain; time to any major osteoporotic fracture (hip, humerus, wrist, and clinical spine); time to hip fracture; and time to clinical vertebral fracture.

Statistical Analysis

The parent trial was designed to continue until at least 256 major adverse cardiovascular events had occurred, which was estimated to require enrollment of up to 6000 men for a mean of 3 years.14 Before enrollment, we estimated the power of the trial to detect a clinically significant decrease in fracture risk. Assuming an enrollment of at least 5400 men over a period of 3.5 years, an additional 1.0 to 1.5 years of follow-up, and a fracture rate of 3 to 4% per year in the placebo group, we estimated that the trial would have at least 80% power to detect a 30% lower risk of fracture in the testosterone group than in the placebo group.

All event analyses of this subtrial were conducted in the full-analysis population that included all the participants who underwent randomization; the analyses were conducted on an intention-to-treat basis, irrespective of adherence to the trial regimen. Baseline characteristics were also assessed in the full-analysis population. In accordance with the prespecified analysis plan for the main trial, analyses of serum testosterone, dihydrotestosterone, and estradiol concentrations were conducted in the safety population of participants who had undergone randomization and received at least one dose of testosterone or placebo; measurements only within 30 days after the last dose of testosterone or placebo were analyzed.

All analyses in the Fracture Trial used a cause specific Cox proportional-hazards model with terms for trial group and status with respect to previous cardiovascular disease. Data from participants without an event were censored at the date of last contact. Prespecified sensitivity analyses are described in the Supplementary Appendix.18

Descriptive summaries of the adjudication process included the number and percentages of the total events reported. For fracture location and trauma, counts and percentages of participants having at least one event of the indicated type were calculated. Aalen–Johansen estimates of the cumulative incidence of fracture events, with death as a competing risk, were also computed.

In accordance with the statistical analysis plan, no adjustment was made for multiple comparisons. All confidence intervals are unadjusted and are not a substitute for hypothesis tests. Analyses were performed with the use of SAS software, version 9.4 (SAS Institute), and R software, version 4.2.1 (R Foundation for Statistical Computing).

Results

Participants Enrollment was conducted from May 23, 2018 to February 1, 2022. The last participant completed trial assessments on January 19, 2023. Of 5246 patient identification numbers, 42 were attributed to 20 participants with duplicate or triplicate enrollment. After excluding these, the full-analysis population included 5204 participants: 2601 in the testosterone group and 2603 in the placebo group. The safety population included 5198 participants who had received at east one dose: 2596 in the testosterone group and 2602 in the placebo group. The baseline characteristics of the participants have been reported.15 The two trial groups were similar with respect to age, race, serum testosterone and estradiol concentrations, and the use of medications to treat osteoporosis, which was documented in 13 participants (0.50%) in the testosterone group and 11 participants (0.42%) in the placebo group (Table S1 in the Supplementary Appendix). The trial participants appear to be representative of men with hypogonadism in this age range in the United States, except for an intentionally increased prevalence of cardiovascular disease (Table S2).

Interventions and Adherence

Of the participants who were enrolled (safety population), 4804 (92.4%) were followed for at least 1 year, 3842 (73.9%) for at least 2 years, and 2974 (57.2%) for at least 3 years. The median duration of participation was 3.19 years (inter-quartile range, 1.96 to 3.53). Adherence, determined by comparison of the weights of the pump bottles when dispensed and when returned, was approximately 90% in both trial groups. The incidence of early discontinuation of testosterone or placebo while continuing trial assessments (61.6%) and early withdrawal from the trial and having no further assessments (39.0%) was relatively high but was similar in the two trial groups (Fig. S1 and Table S3).

The median serum testosterone concentration in the testosterone group increased from 227 ng per deciliter (inter-quartile range, 189 to 258) (7.8 nmol per liter; interquartile range, 6.6 to 9.0) at baseline to 368 ng per deciliter (inter-quartile range, 266 to 519) (12.8 nmol per liter; interquartile range, 9.2 to 18.0) at month 6 and remained higher than baseline through year 3 (Table S4). The median serum testosterone concentration did not change substantially among the participants assigned to receive placebo. The median serum concentrations of dihydrotestosterone and estradiol (Tables S5 and S6) also increased among the participants assigned to receive testosterone but not among those assigned to receive placebo.

Adjudication of Fractures

During the trial, 309 fractures in 224 participants were reported, including 186 fractures in the testosterone group and 123 in the placebo group (Table 1). Of these, 154 in the testosterone group and 97 in the placebo group were confirmed to be fractures, and 8 in the testosterone group and 6 in the placebo group were confirmed not to be fractures. The remaining 44 reported fractures could not be confirmed to be fractures or not because of insufficient documentation or uncertainty after review of available medical records.

Fracture End Points

A total of 91 of 2601 participants (3.50%) in the testosterone group and 64 of 2603 participants (2.46%) in the placebo group had one or more clinical fractures, excluding fractures of the sternum, fingers, toes, facial bones, and skull (hazard ratio, 1.43; 95% confidence interval [CI], 1.04 to 1.97) (Fig. 1). Results of prespecified sensitivity analyses were consistent with those of the primary analysis (Fig. S3). No departures from the proportional-hazards assumption were observed for any fracture end point. The cumulative incidence of clinical fracture at year 3 was 3.8% (95% CI, 3.0 to 4.6) in the testosterone group and 2.8% (95% CI, 2.1 to 3.5%) in the placebo group (Fig. 2).

Testosterone was also associated with a higher fracture incidence than placebo for other fracture end points. The forest plot in Figure 1 shows the consistency of the association of testosterone treatment with a higher incidence of fractures of all types. The cumulative incidence in the two trial groups of non–high-impact fractures, all clinical fractures (including those that had been excluded from the primary analysis), and clinical fractures in participants not taking medication for osteoporosis is shown in Figure 3.

Trauma, Fracture Location, and Adverse Events

Most fractures in both trial groups were associated with trauma, more commonly with falls (Table 2). The anatomical locations of the fractures, including locations excluded from the primary analysis, are shown in Table S7. The most common sites of fractures were ribs, wrist, and ankle.

Traumatic events and falls were not prespecified end points, but clinically significant trauma was captured by the reporting of serious adverse events. Serious adverse events involving the musculoskeletal system were reported in 66 participants (2.5%) in the testosterone group and 65 participants (2.5%) in the placebo group. Major adverse cardiovascular events, the primary end point of the parent trial, and all serious adverse events have been reported15 and are summarized in Table S8.

Discussion

In this subtrial involving middle-aged and older men with hypogonadism, the 3-year cumulative incidence of all clinical fractures was 3.8% in the testosterone group and 2.8% in the placebo
group. The fracture incidence was also numerically higher in the testosterone group for all other fracture end points.

The end point of all clinical fractures is the same as that used in several trials of treatments for osteoporosis.19-21 The most common anatomical sites of fractures were ribs, wrist, and ankle, findings similar to those in previous studies involving men.22,23 These sites are of clinical significance because fractures at these sites are associated with low bone mineral density22-24 and with previous fractures22,23 and are therefore considered osteoporotic fractures. More important, they are associated with an increased risk of future fractures22 and increased mortality.25

We did not expect these results, because most previous studies showed that testosterone improved many measures of bone structure and quality. In studies involving men with severe hypogonadism, testosterone treatment increased areal and volumetric bone mineral density1-4 and improved many structural and mechanical measures of trabecular bone on magnetic resonance microimaging.5-7 In the Testosterone Trials, which involved older men with hypogonadism, testosterone treatment for 1 year increased volumetric bone mineral density and estimated bone strength on quantitative CT.10

Because we did not expect these results, we did not design the trial to assess possible mechanisms by which testosterone would increase the incidence of fractures, so we can only speculate about possible mechanisms. Although previou studies showed that testosterone treatment in men with hypogonadism improved many measures of bone structure, especially of trabecular bone, one study showed that testosterone treatment in men with severe hypogonadism decreased cortical bone volume fraction and cortical bone axial thickness, a measure of bone strength.7

The fact that testosterone was associated with increased fracture risk among middle-aged and
older men with hypogonadism should be considered in the context of potential benefits and other risks of testosterone treatment in these men. The Testosterone Trials showed that testosterone treatment improved sexual function26 and mood26 and increased hemoglobin levels27 in older men. In the present trial, testosterone was not associated with an increased risk of major adverse cardiovascular events but was associated with increased risks of atrial fibrillation, pulmonary embolism, and acute kidney injury.15

The Fracture Trial had many strengths, including enrolling more than 5000 men who had two low morning testosterone values, as well as a randomized, placebo-controlled design and a median duration of observation for more than 3 years — a large and long trial of testosterone treatment. Other strengths were the prespecified design to inquire about fractures at every visit, collection of information about reported fractures, and adjudication of the reported fractures centrally by an experienced adjudicator.16,17

This trial also had limitations. Participants were not evaluated for organic causes of hypogonadism, so it is not known whether some men with such causes were included. Adherence to administration of testosterone or placebo was suboptimal, although it was similar in the two trial groups. The increase in serum testosterone concentrations during treatment was less than in some other studies, but the lesser increase could not explain an increase in fractures. Information about falls was not assessed, except in participants who reported fractures. Physical activity and risk taking were also not assessed. Bone density and structure were not evaluated, so the effect of testosterone on these measures cannot be compared with the results in previous studies.
We found that among middle-aged and older men with hypogonadism, testosterone treatment did not result in a lower incidence of clinical fracture than placebo. The fracture incidence was numerically higher among men who received testosterone than among those who received placebo.

Thread 'T and Male Bone Health: A Puzzle of Interactions'

Apr 4, 2025

Figure 1 339x190 mm ( x DPI)

https://academic.oup.com/jcem/advance-article-abstract/doi/10.1210/clinem/dgaf191/8090450?redirectedFrom=fulltext&login=false

Bone Structure: The Key Players Under the Magnifying Glass

Testosterone in men is mainly secreted by the gonads, with minor contributions from the adrenal cortex. Testosterone and DHT exert their effects through the androgen receptor (AR). Testosterone can be aromatized to bind estrogen receptors (ER and ER) (20, 21). AR and ERs are expressed on osteoblasts and osteocytes, though their presence...

Conclusion: getting down to the bone

Testosterone influences bone metabolism through complex molecular interactions, acting directly on bone cells while modulating key factors like IGF-1 and vitamin D. Its role evolves throughout life, driving bone density and microarchitecture during puberty and maintaining bone mass in adulthood. Testosterone acts directly, through its more active form, DHT, or via estradiol, produced after aromatization, which is crucial for bone health. This is evident in models of ER insensitivity, aromatase deficiency, cAIS, and therapies involving AIs or ADT plus estrogens. Binding protein levels, peripheral metabolism, and comorbidities affecting both gonadal and bone health further complicate the testosterone-bone axis.

Male hypogonadism, characterized by low testosterone levels, is a well-established risk factor for osteoporosis in men. However, many studies combine diverse severity and causes of hypogonadism, and occasionally include eugonadal men, resulting in significant heterogeneity. While TRT consistently improves BMD in truly hypogonadal men, its effect on fracture risk—a key indicator of bone fragility—remains unclear. Fracture risk assessment, particularly for asymptomatic fractures, is often overlooked, and robust evidence linking TRT to reduced fracture incidence is lacking. In conclusion, a deeper understanding of testosterone's role in bone health—spanning from molecular mechanisms to clinical outcomes—is essential for developing effective therapies to prevent and treat osteoporosis. Such advancements could reduce fracture incidence and improve skeletal health in men across the lifespan.

Thread 'Testosterone Replacement and Bone Health: Making Sense of the Data'

Oct 16, 2024

Thread 'Testosterone, estradiol and bone health in men'

May 13, 2024

The EAU Guidelines Office will provide a state-of-the-art update on male hypogonadism and testosterone replacement therapy. This recording will be of great educational value for urologists as internationally-known experts will focus on the controversial and challenging areas of this rapidly evolving field, and provide you with vital insights.

Testosterone replacement therapy is indicated in a variety of clinical conditions and has shown to improve the quality of life in patients with hypogonadism. In addition, the therapy may also have...

madman · 2025-07-12T16:30:02-0500

Vince said:
@madman when using Androgel does your T level usually peek at over 1000 ng/dL ?

Yes one can achieve a high-end/high peak T level when using the higher-end doses!

Post in thread 'New TRT user and frustrated on 1.62 Gel'

Dec 4, 2022

ADK Hunter said:
I’m 47 and low T, low energy and no gym gains. I’ve had issue with ED and my libido at the moment is nonexistent right now. 2 months ago my TEST was 160 and my free T was 11 pg/ml! My PA started me on 1 pump per day alternating shoulders and then bumped me to 2 pumps per day (one pump each shoulder). I have seen not one bit improvement in any areas. He tells me to be patient and it takes time. I’m frustrated and annoyed.

First off you need labs!

Blood work should have been done 2 weeks in.

We have no clue where such a dose has your TT, FT, estradiol, and DHT levels...

From my reply:

A common starting dose for Androgel 1% was 50mg T/day (5 g of gel/1 packet) which would be roughly 5 mg T/day and in most cases, men would only hit a mid-normal T level at best.

Most men would need the higher-end dose of 100 mg T/day (10 g of gel) which would be roughly 10 mg T/day to achieve a high-end or in some cases very high T level.

Vince · 2025-07-12T16:44:34-0500

madman said:
Yes one can achieve a high-end/high peak T level when using the higher-end doses!

Post in thread 'New TRT user and frustrated on 1.62 Gel'

Dec 4, 2022

ADK Hunter said:

I’m 47 and low T, low energy and no gym gains. I’ve had issue with ED and my libido at the moment is nonexistent right now. 2 months ago my TEST was 160 and my free T was 11 pg/ml! My PA started me on 1 pump per day alternating shoulders and then bumped me to 2 pumps per day (one pump each shoulder). I have seen not one bit improvement in any areas. He tells me to be patient and it takes time. I’m frustrated and annoyed.

Click to expand...

First off you need labs!

Blood work should have been done 2 weeks in.

We have no clue where such a dose has your TT, FT, estradiol, and DHT levels...

madman

From my reply:

A common starting dose for Androgel 1% was 50mg T/day (5 g of gel/1 packet) which would be roughly 5 mg T/day and in most cases, men would only hit a mid-normal T level at best.

Most men would need the higher-end dose of 100 mg T/day (10 g of gel) which would be roughly 10 mg T/day to achieve a high-end or in some cases very high T level.

It's sad that they only use gel.

madman · 2025-07-12T16:52:31-0500

Vince said:
We now know that low levels of testosterone does cause issues. High levels we don't know, high levels are definitely much better than low levels.
When I have my next wellness exam. I'm going to get another bone density test.

What is critical here when it comes to overall bone health is not only maintaining a healthy level of FT and estradiol as we age but also making sure that you are consuming adequate protein, calcium, VitD, VitK, potassium, magnesium, phosphorus, zinc and EFAs let alone implementing some form of weight bearing exercise!

Achieving a healthy FT is what truly matters.

When speaking in terms of high T and more importantly high FT there is a big difference between a natty running around with a short-lived daily peak vs one using exogenous T.

Far cry from the high let alone absurdly high trough FT levels many men are running around with on TTh and we are talking levels beyond/well-beyond their natty genetic set-point.

Too many are amped up on T 24/7 steady-state!

madman · 2025-07-12T16:53:09-0500

Peter Snyder well known and highly respected in the field!

Join host Chase Hendrickson, MD, from Vanderbilt University Medical Center, in a discussion about a recent investigation in The Journal of Clinical Endocrinology & Metabolism that tries to illuminate the relationship between serum testosterone and fractures in men. Ever since the release of the TRAVERSE fracture findings last year, we as endocrinologists have been uncertain in how to think about testosterone and bone health, as that trial made us question what we thought we already knew. So we thought that this article would be an important one to review and are eager to discuss it. Dr. Hendrickson talks with Anna Goldman, MD, from Harvard Medical School and guest expert Peter J. Snyder, MD, from the University of Pennsylvania. The article featured this month, by Grahnemo et al, was first published in JCEM in October 2024: “Associations of Serum Testosterone and SHBG With Incident Fractures in Middle-Aged to Older Men.”

Thread 'Testosterone Levels and Fracture Risk in Men'

Mar 22, 2025

Join host Chase Hendrickson, MD, from Vanderbilt University Medical Center, in a discussion about a recent investigation in The Journal of Clinical Endocrinology & Metabolism that tries to illuminate the relationship between serum testosterone and fractures in men. Ever since the release of the TRAVERSE fracture findings last year, we as endocrinologists have been uncertain in how to think about testosterone and bone health, as that trial made us question what we thought we already knew. So we thought that this article would be an important one to review and are...