Femoral Neck and Total Hip BMD Outcomes in Transgender Women on Feminizing Hormone Therapy

What the evidence says

Across the sources on this page, feminizing gender-affirming hormone therapy (GAHT) shows no measured effect on femoral neck (FN) or total hip (TH) bone mineral density (BMD) when studies track change from baseline. Singh Ospina (2017), a meta-analysis of longitudinal studies, found femoral neck BMD change was not statistically significant at 12 and 24 months, and reported no statistically significant change in total hip BMD either, based on evidence from small observational studies at moderate risk of bias. Iwamoto (2024, dgad536) cites a separate 2019 meta-analysis of longitudinal studies confirming this pattern for the first year of treatment: femoral neck and total hip areal BMD were unchanged, in contrast with spine areal BMD, which showed a small increase over the same window. Cross-sectional data are consistent with this picture of stability rather than decline. In Iwamoto's 2024 cross-sectional cohort (Journal of Clinical & Translational Endocrinology), mean femoral neck BMD was 0.931 g/cm² (±0.111), and Z-scores at all three measured sites fell within expected male and female reference ranges. Only 1 of 15 participants had low BMD for age (Z-score ≤ −2.0) at the total hip.

Ciancia (2026) extends this with a larger cross-sectional cohort (n=23 for the transgender-women subgroup) comparing reference standards: femoral neck and total hip areal BMD Z-scores stayed within the normal range but aligned more closely with affirmed-gender reference values than with sex-registered-at-birth values at all sites (femoral neck Z-score −0.26 ± 0.76 against affirmed-gender references versus −0.81 ± 0.63 against sex-registered-at-birth references; total hip Z-score 0.00 ± 1.25 versus −0.60 ± 1.01; both p < 0.001). Clinically low Z-scores were uncommon: at the femoral neck, 2 of 23 participants (8.7%) had Z-scores between −2 and −2.5 under either reference; at the total hip, the same proportion (8.7%) fell in that range under the sex-registered-at-birth reference only. No pathological fractures were reported in the study cohort during the study period; recorded finger, toe, and long-bone fractures were all tied to traumatic events.

Two factors examined for association with FN/TH BMD in transgender women on feminizing GAHT came back null: BMI was not associated with higher femoral neck or total hip BMD in this group (in contrast to transgender men on masculinizing GAHT, where higher BMI was associated with higher BMD at multiple sites), and GAHT duration was not associated with femoral neck or total hip BMD or Z-scores. The one factor consistently linked to BMD in this group was estradiol level: lower serum estradiol was associated with significant decreases in femoral neck and total hip BMD compared to higher estradiol levels, though Iwamoto (2024, dgad536) notes causality is not established and the association could reflect other confounders.

One subgroup comparison stands out for its magnitude and its caveats. Ciancia (2026) found that among transgender women who underwent gonadal hormone suppression, those who received GnRH agonists had lower total hip aBMD Z-scores than those who received cyproterone acetate (Z-score against sex-registered-at-birth reference: −1.37 ± 0.56 versus −0.38 ± 1.01, p=0.015; against affirmed-gender reference: −0.95 ± 0.70 versus 0.26 ± 1.25, p=0.016), with no equivalent difference at the femoral neck. The source itself flags that "definitive conclusions are limited due to the small number of participants who had received GnRHa in both groups."

Where the evidence splits

No contradictions in the current sources.

Thresholds and dose-response

Iwamoto (2024, dgad536) reports that lower estradiol levels — specifically below 182 pmol/L (49.6 pg/mL) — have been associated with significant decreases in femoral neck and total hip BMD compared to higher estradiol levels in transgender women on feminizing GAHT.

Open empirical questions

• No source on this page reports fracture incidence rates or long-term fracture risk in transgender women on feminizing GAHT — only the absence of pathological fractures during the study period is measured.
• The mechanism behind the divergence between spine BMD (small increase in year one) and femoral neck/total hip BMD (unchanged) is not addressed by any source here.
• Whether the total hip Z-score difference between GnRHa and cyproterone acetate use reflects the suppression agent itself, treatment duration, or another unmeasured factor is not resolved — the source's own sample was too small to distinguish these.
• No source here tracks femoral neck or total hip BMD beyond 24 months, so trajectories over multi-year or lifetime feminizing GAHT use are not measured.