Pubertal Suppression, Bone Mass, and Body Composition in Youth With Gender Dysphoria

Medical records of the 198 youth <18 years of age seen at the endocrine diversity clinic at CHEO from January 2006 to April 2017 were retrospectively reviewed, and the 172 youth (86.9%) with at least 1 DXA measurement were included in this study. To allow a minimum follow-up time of 18 months, the study population’s medical records were last reviewed in January 2019. GD diagnosis was based on adolescent medicine specialist assessments, which typically preceded the referral to the endocrine diversity clinic.

Our center uses the GnRHa formulation of leuprolide acetate, historically starting with 3 doses of 7.5 mg intramuscularly every 4 weeks, followed by 11.25 mg intramuscularly every 12 weeks after confirmation of puberty suppression clinically and biochemically. GD management includes areal bone mineral density (aBMD) measurements by DXA with Lunar Prodigy system, and aBMD z scores are determined on the basis of birth-assigned sex, age, and ethnicity. Lumbar spine (LS) (L2–L4) and left total hip (LTH) aBMD z scores are usually assessed at baseline and every 12 months; however, youth with an aBMD z score below 2 SDs or significant drop (≥ 1 SD) in LS aBMD z score undergo BMD every 6 months and a lateral spine radiograph for vertebral fracture assessment. Youth with poor calcium intake are advised to take calcium carbonate 500 mg twice daily to meet the Institute of Medicine’s¹⁹  recommended dietary allowance. All youth are advised to take vitamin D 1000 to 2000 IU daily. Serum 25-hydroxyvitamin D (25OHD) levels are assessed at baseline and monitored every 6 to 12 months. Vitamin D status is classified on the basis of 25OHD level, with <30 nmol/L indicating deficiency, 30 to 50 nmol/L indicating insufficiency, and >50 nmol/L indicating sufficiency.²⁰ 

Given the retrospective nature of the study and variable timing of baseline and follow-up DXA results relative to GnRHa initiation, pre-GnRHa DXA was defined as DXA scan within “−180 to +30” days of GnRHa initiation. Post-GnRHa DXA was defined as DXA scan at “≥ +180” days of GnRHa initiation, as long as this assessment was completed within 90 days of gender-affirming hormone initiation or GnRHa discontinuation.

Birth-assigned sex, affirmed gender, Tanner stage at initial assessment, and anthropometrics (height and weight) at baseline and follow-up visits were retrieved from medical records. BMI z score was calculated on the basis of 2014 revised World Health Organization growth charts for Canada.²¹ 

Pre- and post-GnRHa DXA results were used to retrieve aBMD (grams per square centimeters) of LS, LTH, and total body less head (TBLH). Volumetric BMD was calculated as bone mineral apparent density (BMAD) (grams per cubic centimeters) at LS by using the method of Kroger et al²²  as “LS aBMD × [4/(π × width)].” LS BMAD z score was calculated on the basis of available reference for age-matched birth-assigned gender BMAD mean and SD.²³ 

DXA is the most readily available and valid method of 3 compartmental body analyses.²⁴  In this study, we used pre- and post-GnRHa DXA results to extract the following variables: lean body mass (LBM) (kilograms), total body fat (TBF) (percentage), TBF (kilograms), regional fat mass ratios (ie, trunk/total, legs/total, and extremities/total), fat distribution in the form of android (fat percentage) and gynoid (fat percentage), as well as bone mineral content (BMC) (kilograms) and BMC z score. TBF (percentage) z score was calculated on the basis of age-matched birth-assigned sex TBF (percentage), with mean and SD as “logarithm natural (Ln) TBF (%) – (Ln TBF (%) mean/Ln TBF (%) SD)” for transgender females and Ln “(TBF (%) +10) – (Ln TBF (%) mean/Ln TBF (%) SD)” for transgender males.²³  LBM z score was calculated on the basis of age-matched birth-assigned sex LBM mean and SD.²³ 

Normally distributed data were expressed as mean ± SD and compared with the paired sample t test. Data that were not normally distributed were expressed as median and interquartile range and the Wilcoxon rank test used for comparison. P < .05 was considered statistically significant. However, bond percolation threshold was also used to determine statistical significance, corrected for multiple testing.

The study included 172 youth with GD; 119 (69.2%) youth self-identified as transgender males (age 15.2 ± 1.8 [SD] years; 90.7% Tanner 4–5), 51 (29.7%) youth as transgender females (age 15.4 ± 2.0 years; 80.3% Tanner 4–5), and 2 (1.1%) youth as nonbinary. Baseline body composition and BMD profiles are summarized in Table 1; transgender females had lower z scores at LS aBMD, LS BMAD, LTH aBMD, and BMC than transgender males.

At baseline, the majority (55.2%) of transgender youth were found to have vitamin D deficiency or insufficiency (Table 2). Baseline vitamin D status was associated with baseline LS aBMD, LS BMAD, and LTH z scores (Fig 1). Supplementation with 1000 to 2000 IU of vitamin D daily improved the vitamin D status with no case of vitamin D toxicity (Table 2).

A subgroup of the study population, namely, 36 (30.5%) transgender females and 80 (67.8%) transgender males with pre- and post-GnRHa DXA, was analyzed for GnRHa-associated changes. Pre-GnRHa DXA was done −51.4 ± 41.3 days (range −158 to +28 days) relative to GnRHa initiation; post-GnRHa DXA was done 355.2 ± 96.7 days or median of 352.5 (294.5, 385.8) days (range 188–676 days) after GnRHa initiation. The mean time interval between pre- and post-DXA scans was 406.7 ± 98.3 days (range 210–720 days).

BMD profiles and body composition after GnRHa as well as corresponding changes relative to baseline characteristics are summarized in Table 3. LS, LTH, and TBLH aBMD z scores dropped significantly among both transgender males and transgender females, whereas LS BMAD z scores decreased significantly among transgender males. Considering the limited number of transgender youth at early stages of puberty,^2,3  no statistical comparison of subgroups based on puberty status was done.

Four transgender youth (one birth-assigned girl and 3 birth-assigned males) had a baseline LS aBMD z score below 2 SDs, with 20 transgender youth (all transgender males) having >1-SD drop (1.37 ± 0.26) in their follow-up LS aBMD z score. Each of these youth had a subsequent lateral spine radiograph, none of which revealed a vertebral fracture.

Transgender males’ post-GnRHa body composition analysis revealed significant increases in BMI, LBM (kilograms), TBF (percentage), android (fat percentage), and gynoid (fat percentage); there was no significant change in BMI z score, LBM z score, or TBF (%) z score. Transgender females showed a significant increase in TBF (percentage), TBF (percentage) z score, and gynoid (fat percentage), as well as a drop in LBM z score. GnRHa-associated changes in body compartment z scores and BMI z score were not different between transgender youth with baseline BMI percentiles above and below the obesity risk (85%) cutoff percentile (Table 4).

With this study, we reviewed BMD and body composition changes in youth with GD after 355.2 ± 96.7 days of GnRHa monotherapy. Youth identifying as transgender males constituted 69.2% of study population in keeping with the current trend of overrepresentation (50%–90%) of birth-assigned female patients in GD clinics.^3,25,26 

Baseline BMD characteristics were remarkable for transgender females’ lower LS aBMD and BMAD z scores compared with those of transgender males. Similar baseline differences were previously observed in studies among youth^11,14,15,17  and adult²⁷  transgender females. Klink et al¹⁵  suggested that lower engagement in physical activities, as is typical for transgender females, is a factor, as well as pubertal stage. Despite lack of consistent records of physical activity among our study population, literature supports reduced involvement of sexual minorities in moderate to vigorous physical activities and team sports.²⁸  We believe it is prudent to consider evaluating physical activity level as part of baseline and follow-up assessments of youth with GD, although lack of objective scales or specific recommendations from the Endocrine Society or World Professional Association for Transgender Health^5,6  are challenges in clinical practice.

Baseline body composition comparison of transgender males and transgender females was remarkable for lower BMC z score among transgender females and no difference in LBM and TBF (percentage) z scores. Of note, both transgender males and transgender females had higher baseline TBF and lower baseline LBM relative to the birth-assigned sex population; however, only transgender females had a lower baseline BMC z score. The latter is likely secondary to factors like physical activity, LBM, and BMI rather than delayed bone accrual, and the noted higher TBF among transgender females may have contributed to it, as some evidence suggests that higher body fat attenuates bone density in males while supporting bone density in females.²⁹ 

At baseline, the majority (55.2%) of transgender youth had vitamin D deficiency or insufficiency with vitamin D status associated with BMD at LS and LTH. Some evidence suggests potential bone protective effects of physical activity in the presence of very low serum 25OHD.³⁰  However, the conceivably lower physical activity in transgender youth may heighten vitamin D deficiency effects on BMD. In addition, the combination of low vitamin D and low calcium intake (<600 mg/day) may impair LS mineralization in postmenarchal females.³¹  These all suggest that vitamin D and calcium supplementation should be considered for all youth with GD regardless of whether they are considering puberty suppression.

GnRHa monotherapy negatively affected transgender youth aBMD z scores at LS, LTH, and TBLH levels and transgender males’ LS BMAD z scores. The most pronounced decrease in z score (>0.5) was noted in transgender males’ LS aBMD and BMAD. The majority of the youth were late pubertal; this observed decrease is in line with other studies^11,14  and explained by the significant decrease of estrogen on the trabecular rich bone at the spine. Transgender males at early stages of puberty seemed to have less pronounced LS aBMD z score changes. Schagen et al¹¹  found that transgender youth in both early and late stages of puberty had a drop in aBMD z scores after 24 months of GnRHa regardless of puberty status, but this was more pronounced in transgender males in late puberty, assessing BMD z scores on the basis of sex assigned at birth. Similarly, because we were assessing the effect of GnRHa monotherapy, BMD z scores were based on sex assigned at birth. In future studies, researchers should consider whether to use sex assigned at birth or identified gender when assessing the effect of gender-affirming hormones on bone health in transgender youth. Given the limited evidence on transgender youth in early puberty, the degree of BMD changes after puberty suppression in early puberty is not known. This is an important clinical concern for transgender youth in early stages of puberty who are treated with a longer course of GnRHa therapy before gender-affirming hormones are considered.^5,6 

Current evidence on GnRHa effects on body composition of youth with GD suggests a stable or slightly increased BMI z score, as well as an increase in fat percentage and drop in LBM.^8,15,16,32  In our study, GnRHa was not associated with any significant changes in BMI z score. Among transgender females, GnRHa was associated with a significant increase in TBF (percentage) z score and drop in LBM z score. If transgender females were on GnRHa longer, a significant change in BMI and TBF (percentage) z scores could occur because of loss of testosterone effects on LBM and TBF^33,34  at advanced stages of puberty. The observed fat redistribution (android versus gynoid) is in keeping with youth affirmed gender’s typical fat distribution.^33–35 

Evidence on GnRHa-associated changes in body composition and BMD will help health care professionals involved in the care of youth with GD to counsel appropriately and optimize their bone health. Given the absence of vertebral fractures detected in those with significant decreases in their LS z scores, the significance of BMD effects of GnRHa in transgender youth needs further study, as well as whether future spine radiographs are needed on the basis of BMD trajectory. This study can advance clinical practice guidelines and research in the field.

The study strengths include the large study population with baseline assessments and pre-post GnRHa subgroup analysis. This is the first study used to examine vitamin D status and provide a comprehensive assessment of body composition after GnRHa monotherapy. Lack of consistent records of physical activity at baseline and follow-up visits limited analysis of physical activity’s role as a potential contributing factor to bone health and body composition. The small sample size of youth in early puberty prevented comparison of GnRHa effects in early versus late puberty.

BMI z score does not appear to increase significantly in transgender youth on GnRHa monotherapy; however, it is associated with an increase in TBF (percentage) z score and reduction in LBM z score of transgender females. Transgender youth body fat redistribution (android versus gynoid) after GnRHa and transgender females’ body composition changes after GnRHa were in keeping with youth affirmed gender. GnRHa monotherapy in transgender youth is associated with a decrease in LS, LTH, and TBLH aBMD z scores. Vitamin D sufficiency was uncommon among transgender youth. Vitamin D and calcium supplementation should be considered for all youth with GD, particularly those who seek pubertal suppression. Given the negative effects of GnRHa monotherapy on BMD, bone density and 25OHD should be considered at baseline and follow-up care of adolescents with GD on GnRHa monotherapy.

aBMD

areal bone mineral density

BMAD

bone mineral apparent density

BMC

bone mineral content

BMD

bone mineral density

CHEO

Children’s Hospital of Eastern Ontario

DXA

dual-energy radiograph absorptiometry

GD

gender dysphoria

GnRHa

gonadotropin-releasing hormone agonist

LBM

lean body mass

Ln

logarithm natural

LS

lumbar spine

LTH

left total hip

TBF

total body fat

TBLH

total body less head

25OHD

25-hydroxyvitamin D