Abstract
Graphical abstract
Gender-affirming hormone therapy and cardiovascular disease risk in transgender individuals. CIMT, carotid intima–media thickness. CVD, cardiovascular disease. FMD, flow-mediated dilatation. PWV, pulse wave velocity. TGM, transgender men. Created using BioRender.com.
Abstract
Background
Gender-affirming hormone therapy (GAHT) is used in individuals with gender identity dysphoria to align their secondary sexual characteristics with their affirmed gender. We conducted a systematic review of the literature to explore the mechanisms regarding the effects of GAHT on the vasculature.
Methods
A literature search using PubMed, Embase, Scopus and LILACS was performed using search terms for GAHT, cardiovascular disease (CVD) risk and transgender. Studies were screened by two independent reviewers. Comparison to a cohort of transgender individuals naive or prior to GAHT or a cisgender population was required. Quality assessment was done using the relevant Critical Appraisal Skills Programme checklists.
Results
Out of 2,564 potentially eligible studies, 69 studies met the inclusion criteria. Studies provided evidence of beneficial changes in CVD risk profile, including reduced haemoglobin and pro-inflammatory markers, and atheroprotective changes in lipids in transgender women. In transgender men, there was evidence of negative changes in CVD risk profile, including atherogenic changes in lipids and increased haemoglobin, arterial stiffness and pro-inflammatory markers.
Conclusions
There is a paucity of research across non-traditional measures of CVD risk, which in combination with heterogeneous study design, loss of follow-up, low sample sizes and lack of diversity in age and ethnicity requires the results to be interpreted with caution. More evidence is required to elucidate the mechanisms behind the increased risk of CVD in the transgender population and determine whether GAHT is a contributing factor.
Introduction
Gender-affirming hormone therapy (GAHT) is used to align the secondary sexual characteristics of an individual with their affirmed gender, where the aim is for virilisation in transgender men (TGM) and feminisation in transgender women (TGW). At least 80% of transgender individuals have received or intend to receive GAHT in their lifetime (1).
GAHT consists of oestrogen, in combination with an anti-androgen, in TGW and testosterone in TGM. Oestrogen has been shown to have a protective effect on the vascular system through the reduced incidence of cardiovascular (CV) disease (CVD) in premenopausal cisgender women (CGW). However, long-term oestrogen exposure in the case of hormone replacement therapy in postmenopausal CGW and oral contraceptives have shown increased risk of CVD (2). The effects of androgens in those who were identified male sex at birth can be positive or negative depending on the physiological environment of the vasculature (3). There is a paucity of research investigating the impact of oestrogen on male sex vasculature and testosterone on female sex vasculature.
Recent studies have reported a higher risk of CVD in the transgender population. In TGW, the higher risk of stroke, venous thromboembolism (VTE) and CVD events is well documented throughout the literature (4, 5). In TGM, the evidence is less conclusive with many studies reporting no increased incidence or risk of CVD in terms of stroke, VTE or CVD events compared to CGW and cisgender men (CGM) (4, 5) and others reporting increased risk (1). A recent systematic review and meta-analysis (6) demonstrated that, even with high heterogeneity and conflicting studies, transgender individuals have a 40% higher risk of CVD compared with cisgender individuals. The incidence of stroke, myocardial infarction and VTE was higher in TGW than in TGM in two recent reviews (6, 7), but both groups had a higher incidence and relative risk compared with cisgender individuals of the same sex at birth (6).
The mechanisms responsible for the increased CVD risk are poorly understood, which poses a challenge for clinicians when assessing and managing CVD in this population. Clearly, there is a need for an improved understanding of these mechanisms in order to institute appropriate preventative strategies to prevent the burden of CVD. Already a complex issue clinically due to patients carrying multiple CVD risk factors, the multifactorial approach required in cisgender individuals (8) becomes more complex when taking into account the impact of gender, health inequalities, differences in behavioural risk factors or utilisation of GAHT (9).
Mechanistic studies investigating vascular function and GAHT are scarce, but a previous review (10) investigated the effect of GAHT on the risk of subclinical atherosclerosis. Twelve studies were identified within the review, which reported on measures of vascular function and blood biomarkers of endothelial function, inflammation or coagulation. Overall, there was limited evidence in both TGM and TGW, often indicating no changes in parameters. Due to the limited studies included within this review (10), which reported on studies dated up to 2020, the present review aims to provide additional evidence on CV risk in the transgender population whilst expanding the range of CV phenotypes investigated. The clinical phenotypes selected for this review were traditional CV and metabolic risk factors, including body composition, blood pressure, lipid profile and insulin sensitivity, in addition to the phenotypes investigated in the previous review by Moreira and coworkers (10). These have been closely linked to major CVD (8), are known to be influenced by sex hormones (3) and are useful intermediate phenotypes that can be easily used to assess CV risk clinically. The primary aim of this systematic review is to establish the proposed mechanisms delineated in the literature to date regarding the effects of GAHT on the vasculature in the transgender population.
Methods
Search strategy
This systematic review builds on the review by Moreira Allgayer and coworkers (10). Potential studies investigating the impact of GAHT on the vascular system were searched for on PubMed, Embase, Scopus and LILACS until 02 July 2024. The search strategy consisted of the following terms for GAHT, CV risk and transgender: ‘hormone therapy’ OR ‘gender hormone therapy’ OR ‘transgender’ OR ‘oestrogen’ OR ‘testosterone’ AND ‘vascular’ OR ‘CV’ OR ‘inflammation’ OR ‘metabolic syndrome’ OR ‘metabolic’ OR ‘cytokine’ AND ‘transgender’ in PubMed and equivalent in the other databases. The reference lists of the studies identified by the search were hand-searched by one author (KAM).
Inclusion/exclusion criteria
Studies had to be in humans, published in the English language and with full text available. Studies could be cohort (retrospective or prospective), cross-sectional, case–control or randomised clinical trials.
Studies had to meet the following criteria for inclusion: i) report on a healthy transgender group (either TGW or TGM); ii) contain a transgender group receiving GAHT; iii) investigate the impact of GAHT on the vascular system; iv) consider at least one assessment of clinical or laboratory measurement of CV risk or vascular phenotype; and v) compare a transgender cohort either before and after GAHT or to a cisgender population or transgender population not treated with GAHT. Studies were excluded if they only provided data on the incidence of CVD events/deaths or just reported CV risk ratios.
Study selection
Potentially eligible studies, identified by the search strategy, were inputted into Covidence (https://www.covidence.org/) for screening. Duplicates were automatically removed by Covidence, and any missed studies were manually removed. The studies were screened by both title and abstract based on the inclusion/exclusion criteria independently by two reviewers (KAM and ALH). Any disagreements between the reviewers were resolved by discussion to reach a consensus. Full texts were screened to exclude any further studies that did not meet the criteria and to assess the eligibility for data extraction.
Data extraction and analysis
The general characteristics of each study were extracted, including vascular phenotype (anthropometry, blood pressure, insulin sensitivity, etc.), study design, sample size, age of participants (range and mean), ethnicity, country and time on and administration route of GAHT. Details of study design included type of study, comparator groups, inclusion/exclusion criteria and measurements taken. Statistically significant results were extracted. When P values were not available to identify statistical significance, confidence intervals were used. When papers carried out multiple models accounting for confounding variables, only the most controlled model was reported.
Quality assessment
To determine the quality of the studies, the appropriate Critical Appraisal Skills Programme (CASP) cohort study checklist was utilised (https://casp-uk.net/casp-tools-checklists/cohort-study-checklist/). Methodological quality was assessed by one author (KAM) with input from another (ALH). Each study was defined as low, medium or high risk of bias based on the answers using the CASP tool.
Results
Search results
The initial search yielded 2,564 results from PubMed (n = 952), Embase (n = 944), Scopus (n = 658) and LILACS (n = 10) (Fig. 1). Twenty-two additional studies were added, which were identified from reference lists (11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32). One-hundred eleven studies were left after duplicate removal (n = 1,623, 37.5%), title and abstract screening (n = 1,494, 58.3%) and lack of access (n = 18, 0.7%). An attempt to email the authors of unretrievable studies was unsuccessful; therefore, these studies were excluded without full text screening. Forty-two (1.6%) further studies were excluded by full-text screening due to no GAHT treatment, outcomes not being relevant, not being available in the English language and abstract available only. This resulted in a total of 69 studies for inclusion in the review.
Study characteristics
Table 1 shows the characteristics of the 69 studies included in this review. Thirty-two (46%) studies were identified as low risk of bias and 37 (54%) as medium risk of bias (Table 1).
Study characteristics.
Author | Country/region | Type | Vascular phenotype theme | Age (mean; range (years))/number of participants (n) | Ethnicity of majority | Time (months) and administration method of GAHT | Comparison | Quality assessment |
---|---|---|---|---|---|---|---|---|
Allen et al. 2021 (9) | USA | Retrospective cohort | Haemostasis and lipid profile | TGW = 31.1 (n = 126) | White | ≤60 months | Baseline, then every 3 months for 1 year and then every 6 months up to 5 years | Medium |
TGM = 27.8 (n = 91) | TGW: oral or IM | |||||||
TGM: IM or transdermal | ||||||||
Anike et al. 2024 (33) | USA | Retrospective cohort | Insulin sensitivity | TGW/TGM /CGW/CGM = 18–55 (n = 2,425/2,127/33,995/38,913) | Non-Hispanic White | ≤120 months | Baseline and up to 120 months | Low |
No information on administration | ||||||||
Balik et al. 2022 (34) | Turkey | Cross-sectional | Vascular and inflammation | CGW = 28.4 ± 4.7 (n = 30) | No information | >12 months | CGW to TGM | Medium |
TGM = 27.4 ± 5.1 (n = 30) | No information on administration | |||||||
Banks et al. 2021 (35) | USA | Retrospective cohort | Blood pressure | TGW = 29.3 ± 10.1 (n = 247) | White | <57 months | Baseline and up to 57 months | Low |
TGM = 26.1 ± 7.1 (n = 223) | TGW: oral or IM | |||||||
TGM: IM or transdermal | ||||||||
Berra et al. 2006 (27) | Europe | Prospective cohort | Anthropometric, lipid profile, inflammation and insulin sensitivity | TGM = 30.4 (5.4) (n = 16) | No information | 6 months | Baseline and after 6 months | Medium |
IM | ||||||||
Boekhout-Berends et al. 2023 (36) | Europe | Retrospective cohort | Haematology | TGW = 30.8 (23.5–43.5) (n = 1178) | White | 12 months | Baseline and after 12 months | Low |
TGM = 23.3 (20.1–30.9) (n = 1023) | TGW: oral or transdermal | |||||||
TGM: IM or transdermal | ||||||||
Bretherton et al. 2021 (37) | Australia | Cross-sectional | Anthropometric and insulin sensitivity | TGM = 28.8; 25.0–33.0 (n = 43) | No information | TGW: oral or transdermal (39 months) | TGM to CGW and CGM | Low |
CGW = 28.1; 24.0–38.7 (n = 48) | TGM: oral or IM (44 months) | TGW to CGW and CGM | ||||||
TGW = 41.1; 26.4–52.7 (n = 41) | ||||||||
CGM = 32.0; 26.3–40.9 (n = 30) | ||||||||
Chandra et al. 2010 (26) | USA | Prospective cohort | Anthropometric, lipid profile, blood, blood pressure and insulin sensitivity | TGM = 29 (9) (n = 12) | Caucasian | 12 months | Baseline and after 12 months | Low |
TGM: IM | ||||||||
Cocchetti et al. 2021 (38) | Europe | Prospective cohort | Anthropometric, blood pressure and lipid profile | TGW = 31.8 ± 11.46 (n = 144) | No information | 24 months | Baseline, 12 and 24 months | Low |
TGM = 26.78 ± 7.48 (n = 165) | TGW: oral or transdermal | |||||||
TGM: IM or transdermal | ||||||||
Colizzi et al. 2015 (25) | Europe | Retrospective cohort | Anthropometric, blood pressure, lipid profile and insulin sensitivity | TGW = 30.24 (9.57) (n = 79) | No information | 24 months | Baseline, 12 and 24 months | Low |
TGM = 28.77 (5.62) (n = 43) | TGW: oral | |||||||
TGM: IM | ||||||||
Cunha et al. 2023 (39) | Brazil | Cross-sectional | Vascular | TGM = 44 ± 10; 26–61 (n = 33) | No information | 168 ± 96 months | TGM to CGW and CGM | Medium |
CGW = 43 ± 10 (n = 55) | TGM: IM | |||||||
CGM = 42 ± 10 (n = 56) | ||||||||
Defreyne et al. 2018 (24) | Europe | Prospective cohort | Blood | TGW = 28.5; 16–69 (n = 239) | No information | 36 months | Baseline, 3, 6, 9, 12, 18, 24 and 36 months | Low |
TGM = 22.5; 17–62 (n = 192) | TGW: oral or transdermal or sublingual | |||||||
TGM: IM or transdermal | ||||||||
Deischinger et al. 2023 (40) | Europe | Cross-sectional | Anthropometric, insulin sensitivity, blood pressure, lipid profile and vascular | TGM = 24; 21–27 (n = 22) | White | TGW: oral or transdermal (16.8; 12–26.4) | TGW to CGM TGM to CGW | Medium |
TGW = 35; 29–46 (n = 16) | TGM: IM or transdermal (13.2; 12–24 months) | |||||||
CGW = 23; 22–27 (n = 16) | ||||||||
CGM = 33; 29–48 (n = 17) | ||||||||
Deutsch et al. 2015 (41) | USA | Prospective cohort | Anthropometric, blood pressure and lipid profile | TGW = 29 (9.4); 19–50 (n = 16) | White | 6 months | Baseline and after 6 months | Medium |
TGM = 27 (6.9); 18–45 (n = 31) | TGW: IM or sublingual or transdermal | |||||||
TGM: Transdermal or subcutaneous injection | ||||||||
Duro et al. 2023 (42) | USA | Cross-sectional | Vascular | TGW = 40; 36–50 (n = 25 (24)) | No information | TGW = 38.4; 24–166.8 months; TGM = 66; 22.8–148.8 months | TGW and TGM to CG | Medium |
TGM = 38; 32–43 (n = 22 (16)) | No information on administration | |||||||
Elbasan et al. 2023 (43) | Turkey | Prospective cohort | Anthropometric and blood | TGM = 24; 21–29 (n = 45) | No information | 12 months | TGM to CGW | Low |
CGW = 27; 23–29 (n = 28) | TGM: IM | Baseline to 6 and 12 months | ||||||
Elbers et al. 1997 (23) | Europe | Prospective cohort | Anthropometric, lipid profile, insulin sensitivity and blood pressure | TGW = 26 (6); 18–36 (n = 20) | No information | 12 months | Baseline and 12 months | Medium |
TGM = 23 (5); 16–34 (n =17) | TGW: oral | |||||||
TGM: IM | ||||||||
Elbers et al. 2003 (22) | Europe | Prospective cohort | Anthropometric, lipid profile, insulin sensitivity and blood pressure | TGW = 26 (6); 18–36 (n = 20) | No information | 12 months | Baseline and 12 months | Medium |
TGM = 23 (5); 16–34 (n = 17) | TGW: oral | |||||||
TGM: IM | ||||||||
Elbers et al. 1999 (21) | Europe | Prospective cohort | Anthropometric and insulin sensitivity | TGW = 26 (6); 18–37 (n = 20) | No information | 12 months | Baseline and 12 months | Medium |
TGM = 23 (5); 16–34 (n = 17) | TGW: oral | |||||||
TGM: IM | ||||||||
Emi et al. 2008 (29) | Asia | Cross-sectional | Anthropometric, blood pressure, blood, haemostasis, lipid profile and vascular | TGM (GAHT) = 27.9 (5.5) (n = 48) | No information | 45 (38.1) months | Compared to no GAHT treatment | Medium |
TGM (no GAHT) = 26.5 (5.5) (n = 63) | TGM: IM | |||||||
Fernandez et al. 2016 (20) | USA | Retrospective cohort | Anthropometric, blood pressure, lipid profile and blood | TGW = 31; 16–56 (n = 33) | White (non-Hispanic) | 18 months | Baseline to 3–6 and 6–18 months | Medium |
TGM = 27; 19–47 (n = 19) | TGM: oral or transdermal or IM | |||||||
TGM: IM | ||||||||
Giltay et al. 1998 (17) | Europe | Prospective cohort | Anthropometric, blood and lipid profile | TGW = 28; 18–40 (n = 17) | No information | 4 months | Baseline to 4 months | Medium |
TGM = 22; 16–34 (n = 17) | TGW: oral | |||||||
TGM: IM | ||||||||
Goh et al. 1995 (16) | Asia | Cross-sectional | Lipid profile | TGM (GAHT) = 31.7 (0.88) (n = 39) | No information | 33 (6–120) months | Compared to no GAHT treatment | Medium |
TGM (no GAHT) = 25.6 (0.78) (n = 29) | TGM: IM | |||||||
Gulanski et al. 2020 (40) | USA | Cross-sectional | Vascular | TGM = 27 ± 4; 20–31 (n = 11) | White | ≥3 months | TGM and CGW | Medium |
CGW = 28 ± 5; 18–34 (n = 20) | TGM: IM | |||||||
Iannantuoni et al. 2020 (44) | Europe | Prospective cohort | Vascular | TGM = 26.2 ± 7.5 (n = 157) | No information | 3 months | Baseline and 3 months | Medium |
TGM: IM | ||||||||
Jacobeit et al. 2007 (45) | Europe | Prospective cohort | Blood and lipid profile | TGM = 33 (6); 26–44 (n = 12) | No information | 12 months | Baseline and 12 months | Medium |
TGM: IM | ||||||||
Jacobeit et al. 2009 (46) | Europe | Prospective cohort | Blood and lipid profile | TGM = 32 (7); 23–47 (n = 17) | No information | 36 months | Baseline, 6, 12, 18, 24, 30 and 36 months | Medium |
TGM: IM | ||||||||
Klaver et al. 2018 (47) | Europe | Prospective cohort | Anthropometric | TGW = 29; 18–66 (n = 179) | No information | 12 months | Baseline and 12 months | Low |
TGM = 24; 18–58 (n = 162) | TGW: oral or transdermal | |||||||
TGM: IM or transdermal | ||||||||
Klaver et al. 2018 (48) | Europe | Retrospective cohort | Anthropometric | TGW = 16.4 (1.1) (n = 71) | Caucasian | TGW: oral (33.6; 19.2–40.8 months) | Before and after GAHT TGM to CGW/CGW | Medium |
TGM = 16.9 (0.9) (n = 121) | TGM: IM (36; 22.8–40.8) | TGW to CGM/CGW | ||||||
Klaver et al. 2020 (49) | Europe | Retrospective cohort | Anthropometric, blood pressure, lipid profile and insulin sensitivity | TGW/TGM = 22; 20.5–23.5 (n = 71/121) | White | TGW: oral (26.4 (IQR: 13.2–37.2 months) | Before and after GAHT TGM to CGW/CGW | Low |
TGM: IM (34.8 (IQR: 20.4–40.8) months) | TGW to CGM/CGW | |||||||
Klaver et al. 2022 (50) | Europe | Prospective cohort | Anthropometric, lipid profile and insulin sensitivity | TGW = 29 (IQR: 23–42) (n = 179) | White | 12 months | Baseline and 12 months | Low |
TGM = 24 (IQR: 21–33) (n = 162) | TGW: oral or transdermal | |||||||
TGM: IM or transdermal | ||||||||
Korpaisarn et al. 2021 (51) | Asia | Retrospective cohort | Blood pressure, haematology and lipid profile | TGM = 27.8 ± 6.0 (n = 39) | Thai | 25.52 ± 12.9 months | Baseline and ∼24 months | Medium |
TGM: IM | ||||||||
Kulprachakarn et al. 2020 (52) | Asia | Cross-sectional | Blood pressure, insulin sensitivity, lipid profile and vascular | TGW (GAHT) = 24.0 ± 5.1 (n = 100) | Thai (inferred) | 79.4 ± 6.2 | Compared to no GAHT treatment | Medium |
TGW (no GAHT) = 24.3 ± 5.7 (n = 100) | No information on administration | |||||||
Lake et al. 2022 (53) | USA | Cross-sectional | Haemostasis, insulin sensitivity, inflammation and vascular | TGW = 40; 32–48 (GAHT = 31/no GAHT = 16) | Latina | No information on time | Compared to no GAHT treatment TGW to CGM | Medium |
CGM = 42; 35–52 (n = 40) | TGW: oral | |||||||
Lake et al. 2023 (54) | USA | Retrospective cohort | Insulin sensitivity, haemostasis, inflammation and vascular | TGW = 27 (n = 138) | Black and Hispanic | 12 months | Baseline compared to 6 and 12 months | Medium |
No information on administration | ||||||||
Leemaqz et al. 2023 (55) | USA | Retrospective cohort | Lipid profile | TGW = 29.9 ± 9.5 (n = 170) | White | <57 months | Baseline up to 57 months | Medium |
TGM = 26.4 ± 7.2 (n = 196) | TGW: oral or IM | |||||||
TGM: IM or transdermal | ||||||||
Lim et al. 2019 (56) | Australia | Cross-sectional | Haemostasis | TGW = 32.8; 26.7–44.7 (n = 26) | No information | 25.5 months | TGW to CGM | Low |
CGM = 28.7; 24.9–57.3 (n = 55) | TGW: oral or transdermal | TGW to CGW | ||||||
CGW = 44.9; 25.5–58.3 (n = 98) | ||||||||
Liu et al. 2021 (57) | Asia | Retrospective cohort | Anthropometry, blood pressure, lipid profile, insulin sensitivity and haematology | TGM = 27.9 ± 0.7 (n = 65) | No information | ≥12 months | Baseline, 3–6 (V1), 6–12 (V2) and 12–24 (V3) months | Medium |
TGW = 26.0 ± 1.1 (n = 45) | TGW: oral | |||||||
TGM: IM | ||||||||
Martinez-Martin et al. 2023 (58) | Europe | Retrospective cohort | Anthropometric, insulin sensitivity, blood pressure and lipid profile | TGW = 19.1 (5.0) (n = 149) | No information | 60 months | Baseline and 60 months | Low |
TGM = 20.2 (5.6) (n = 153) | TGW: transdermal | |||||||
TGM: parenteral | ||||||||
McCredie et al. 1998 (30) | Australia | Cross-sectional | Lipid profile, blood pressure and vascular | TGW = 33 (5); 26–46 (n = 12) | No information | 38 (52); 2–177 months | TGM to CGW | Medium |
CGW = 31 (7); 26–41 (n = 12) | TGM: implant or IM | |||||||
Meyer et al. 2020 (28) | Europe | Retrospective cohort | Haematology and lipid profile | TGW = 25; 15–61 (n = 155) | No information | 48 months | Baseline, 3–4, 10–14 months and 3–4 years | Medium |
TGM = 21; 15–52 (n = 233) | TGW: oral or transdermal | |||||||
TGM: IM or transdermal | ||||||||
Millington & Chan 2021 (59) | USA | Cross-sectional | Lipid profile | TGM = 18.4 (IQR: 17.6–19.5) (n = 17) | White | 14.4 ± 9.6 months | TGM to CGW | Medium |
CGM = 17.8 (IQR: 17–19.4) (n = 33) | TGM: subcutaneous injection | TGM to CGM (models 1 and 2)* | ||||||
CGW = 17.6 (IQR: 17.1–18.5) (n = 32) | ||||||||
Millington et al. 2024 (60) | USA | Prospective cohort | Blood and insulin sensitivity | TGW = 17.3; 16.1–18.6 (n = 93) | No information | 24 months | Baseline, 6, 12 and 24 months | Low |
TGM = 16.2; 15.1–17.6 (n = 200) | TGW: oral or transdermal or IM | |||||||
TGM: subcutaneous or transdermal | ||||||||
Mueller et al. 2007 (12) | Europe | Prospective cohort | Anthropometric, lipid profile, blood, blood pressure, insulin sensitivity and haemostasis | TGM = 29.63 (8.95) (n = 35) | No information | 12 months | Baseline and 12 months | Low |
TGM: IM | ||||||||
Mueller et al. 2010 (11) | Europe | Prospective cohort | Anthropometric, lipid profile, blood pressure and blood | TGW = 30.4 (9.1) (n = 45) | Caucasian | 24 months | Baseline, 12 and 24 months | Low |
TGM: IM | ||||||||
New et al. 1997 (32) | Australia | Cross-sectional | Anthropometric, lipid profile, blood pressure and vascular | TGW = 41 (9) (n = 14) | No information | 61 (70) months | TGW to CGM/CGW | Medium |
CGM = 41 (7) (n = 14) | TGW: oral | |||||||
CGW = 40 (8) (n = 15) | ||||||||
Nokoff et al. 2020 (61) | USA | Cross-sectional | Anthropometric, blood pressure, lipid profile and insulin sensitivity | TGM = 17 ± 1.4 (n = 19) (CGW = 15.2 ± 1.9 (n = 42), CGM = 15.3 ± 1.6 (n = 19)) | White (and non-Hispanic) | TGW: oral or sublingual (12.3 ± 9.9 months) | TGM with matched CGW and CGM | Medium |
TGW = 16.3 ± 1.4 (n = 11) (CGW = 15.9 ± 1.4 (n = 23), CGM = 15.7 ± 1.4 (n = 24)) | TGM: IM or subcutaneous injection (11.2 ± 5.2 months) | TGW with matched CGW and CGM | ||||||
Olson-Kennedy et al. 2018 (62) | USA | Prospective cohort | Anthropometric, blood, insulin sensitivity, lipid profile and blood pressure | TG = 18; 12–23 (TGW = 23/TGM = 35) | Caucasian and Latino(a) | 24 months | Baseline and 24 months | Low |
TGW: oral or IM | ||||||||
TGM: subcutaneous | ||||||||
Ott et al. 2011 (19) | Europe | Retrospective cohort | Anthropometric and lipid profile | TGW = 35.7 (11.4) (n = 89) | No information | 60 months | Baseline and 60 months | Medium |
TGM = 25 (6.3) (n = 80) | TGW: transdermal or oral | |||||||
TGM: IM or oral | ||||||||
Pallotti et al. 2023 (63) | Europe | Prospective cohort | Haematology, insulin sensitivity and lipid profile | TGM = 24.8 ± 8; 18–49 (n = 52) | Caucasian | 12 months | Baseline, 6 and 12 months | Low |
TGM: IM | ||||||||
Pei et al. 2024 (64) | Asia | Retrospective cohort | Anthropometric, lipid profile and insulin sensitivity | TGW (GAHT) = 23 (3) (n = 59) | No information | 12–26 months | Compared to no GAHT treatment | Low |
TGW (no GAHT) = 25 (5) (n = 40) | TGW: oral or transdermal | |||||||
Resmini et al. 2008 (31) | Europe | Cross-sectional | Insulin sensitivity | TGW = 33.21 (2.1) (n = 15) | No information | TGW: oral or transdermal (108 (28) months) | TGW/TGM and CGW/CGM | Medium |
TGM = 30.0 (1.81) (n = 11) | TGM: IM (120 (18) months) | |||||||
CGM/CGW = age-matched (n = 15/14) | ||||||||
Roberts et al. 2014 (65) | USA | Retrospective cohort | Blood and lipid profile | TGW = 46; 27–67 (n = 55) | No information | 48 (>6) months | TGW to CGM/CGW | Medium |
CGM = 58; 21–84 (n = 20) | No information on administration | |||||||
CGW = 56; 23–88 (n = 20) | ||||||||
Sanchez-Toscano et al. 2023 (66) | Europe | Retrospective cohort | Anthropometric, insulin sensitivity, lipid profile and blood pressure | TGW = 18; 16–22 (n = 91) | No information | 12 months | Baseline and 12 months | Low |
TGM = 18; 16–23 (n = 136) | TGW: oral | |||||||
TGM: IM | ||||||||
Scheres et al. 2021 (2) | Europe | Prospective cohort | Haemostasis | TGM = 26.9 ± 9.7 (n = 100) | No information | 12 months | Baseline and 12 months | Low |
TGW = 33.7 ± 12.9 (n = 98) | No information on administration | |||||||
Schutte et al. 2022 (67) | Europe | Prospective cohort | Inflammation and haemostasis | TGM = 23; 20–26 (n = 47) | No information | 12 months | Baseline, 3 and 12 months | Low |
TGW = 30; 24–39 (n = 48) | TGW/TGM: transdermal | |||||||
Shadid et al. 2020 (68) | Europe | Prospective cohort | Anthropometric, lipid profile and insulin sensitivity | TGM = 26.1 ± 1.3; 18–64 (n = 55) | No information | ≥12 months | Baseline and after 12 months | Low |
TGW = 34.4 ± 1.5; 18–64 (n = 35) | TGW: oral or transdermal | |||||||
TGM: IM | ||||||||
SoRelle et al. 2019 (13) | USA | Retrospective cohort | Insulin sensitivity, blood, lipid profile and blood pressure | TGW (no GAHT) = 31 (12) (n = 87) | White | >6 months | Compared to no GAHT treatment | Medium |
TGW (GAHT) = 33 (12) (n = 133) | TGW: oral or transdermal | |||||||
TGM (no GAHT) = 27 (9) (n = 62) | TGM: IM or transdermal | |||||||
TGM (GAHT) = 30 (9) (n = 89) | ||||||||
Stoffers et al. 2019 (69) | Europe | Retrospective cohort | Anthropometric, blood pressure, lipid profile and blood | TGM = 17.2; 14.9–18.4 (n = 62) | No information | 12 (5–33) months | Baseline, 6, 12 and 24 months | Medium |
TGM: IM or transdermal | ||||||||
Suppakitjanusant et al. 2020 (70) | USA | Retrospective cohort | Anthropometric | TGW (GAHT) = 43.9 ± 15.6 (n = 46) | Caucasian | TGW: oral or transdermal or IM (79 ± 112 months) | Over 84-month period | Low |
TGW (no GAHT) = 32.8 ± 11.7 (n = 59) | TGM: IM or transdermal (44 ± 41 months) | |||||||
TGM (GAHT) = 40.4 ± 13.1 (n = 15) | ||||||||
TGM (no GAHT) = 27.4 ± 8.8 (n = 25) | ||||||||
Tebbens et al. 2023 (71) | Europe | Intervention study (cohort study) | Anthropometric, lipid profile and insulin sensitivity | TGW = 26; 20–27 (n = 8) | No information | 12 months | Baseline and 12 months | Low |
TGM = 22; 19–25 (n = 10) | TGW: oral and transdermal | |||||||
TGM: transdermal or IM | ||||||||
Tsatsanis et al. 2023 (72) | Europe | Cross-sectional | Inflammation | TGW = 41.6 (CI: 36.1–47.3) (n = 16) | No information | ≥36 months | TGW to CGW/CGM TGM to CGW/CGM | Medium |
TGM = 33.4 (CI: 30.1–36.6) (n = 27) | TGW: oral or transdermal or IM | |||||||
CGW = 35.9 (CI: 32.5–39.2) (n = 26) | TGM: transdermal or IM | |||||||
CGM = 46.0 (CI: 43.6–48.4) (n = 30) | ||||||||
Valentine et al. 2021 (73) | USA | Retrospective cohort | Anthropometric and lipid profile | TGM = 16.6 ± 1.3 (n = 42) | TGM = White CGW = Black | 10.8 (2.6–25.7) months | Baseline TGM to CGW baseline and 12 months | Low |
CGW = 15.5 ± 1.8 (n = 82) | TGM: IM | |||||||
Van Caenegem et al. 2015 (18) | Europe | Prospective cohort | Anthropometric | TGM = 27 (9) (n = 23) | Caucasian | 12 months | Baseline and 12 months TGM to CGW | Low |
CGM = 27 (9) (n = 23) | TGM: IM | |||||||
van Velzen et al. 2021 (74) | Europe | Prospective cohort | Lipid profile | TGM = 28 ± 13 (n = 15) | No information | <12 months | Baseline and after 12 months | Low |
TGW = 33 ± 12 (n = 15) | TGW: oral or transdermal | |||||||
TGM: transdermal or IM | ||||||||
Victor et al. 2014 (75) | Europe | Prospective cohort | Anthropometric, blood pressure, blood, lipid profile, insulin sensitivity and vascular | TGM = 29.4 (9.7) (n = 57) | No information | 3 months | Baseline and 3 months | Medium |
TGM: IM | ||||||||
Vita et al. 2018 (76) | Europe | Retrospective cohort | Anthropometric, blood pressure, blood, insulin sensitivity, lipid profile | TGW = 25.2 (7) (n = 21) | No information | TGW: oral (31.3 (34.2) months) | Baseline and on GAHT | Low |
TGM = 25.1 (3.7) (n = 11) | TGM: IM (22.2 (11.5) months) | |||||||
Wierckx et al. 2014 (14) | Europe | Prospective cohort | Anthropometric, blood pressure, blood, insulin sensitivity and lipid profile | TGM = 21.7 (5.1)/21.7 (5.1) (n = 53) | No information | 12 months | Baseline and 12 months oral vs transdermal in TGW | Low |
TGW = 31.7 (14.8)/19.3 (2.4) (n = 53) | TGW: oral or transdermal | |||||||
TGM: IM | ||||||||
Yun et al. 2021 (15) | Asia | Prospective cohort | Anthropometric | TGW = 28.5 (8.1) (n = 11) | No information | 6 months | Baseline and 6 months | Low |
TGW: oral or IM |
Vascular phenotype denotes the CV risk phenotypes assessed within the studies. Time on GAHT refers to the time on GAHT in both cross-sectional and cohort studies. In the cohort studies, this also refers to the follow-up time/duration of the study. Quality assessment is defined as low, medium and high risk based on answers from the CASP quality assessment checklist for cohort studies. *Model 1 is the unadjusted P value, and model 2 is based on an adjusted P value based on BMI, race and age (78). For (32), the number of individuals in the brackets is the number of individuals of which the CAC data came from.
Abbreviations: CGM, cisgender men; CGW, cisgender women; GAHT, gender-affirming hormone therapy; IM, intramuscular; IQR, interquartile range; TGM, transgender men; TGW, transgender women.
The 69 studies consisted of 14,363 transgender individuals, where there was an even split of TGW (50%) and TGM (50%). The sample size ranged from as low as 8 TGW and 10 TGM (71) to as high as 2424 TGW and 2127 TGM (33). A high loss of follow-up was identified in some studies, with up to 79% loss in TGW and 82% in TGM in one study (35). Most studies were carried out across Europe (36/69, 52%) or the US (19/69, 28%). Nine (13%) of the studies (2, 14, 18, 24, 38, 48, 67, 68, 74) were part of the European Network for the Investigation of Gender Incongruence (ENIGI), which is a prospective study aiming to provide data-driven information on the safety of GAHT (77). Many studies (43/69, 62%) provided no information on ethnicity, and the rest were mostly of a white (12/69, 17%), Caucasian (9%, 6/69) or white non-Hispanic (3/69, 4%) ethnicity (Table 1).
The mean age range was similar in both TGW and TGM at 16.3–46 years and 16.2–44 years, respectively. The mean time of GAHT was between 3 months and 14 years. In the 62 studies in TGM, most TGM received intramuscular (81%) or transdermal testosterone injections (31%), with two studies providing no information (33, 42). In the 47 studies in TGW, there was more variety in the administration of GAHT, including oral (83%), intramuscular (23%), transdermal (51%) and sublingual (5%), with two studies providing no information (33, 65).
Cohort studies investigating before and after GAHT treatment
Most studies were cohort studies (51/69, 74%), comparing transgender individuals before and after receiving hormone therapy.
Traditional CV and metabolic risk factors
Thirty-eight (38/51, 75%) studies reported on anthropometric measurements of body composition (Table 2). Studies either indicated a significant increase in body mass index or reported no changes in both TGW and TGM. Few studies reported increases in bodyweight in TGW and TGM. Many studies provided more in-depth analysis of body composition (11, 14, 15, 18, 21, 22, 23, 27, 47, 48, 50, 64, 71). In TGW, evidence suggests a reduced lean body mass and visceral adipose tissue with an increase in total body fat and subcutaneous adipose tissue compared to that before GAHT (14, 15, 21, 22, 23, 48, 50, 47, 64, 71). In TGM, the opposite was identified in comparison with the baseline before GAHT (11, 14, 18, 21, 22, 23, 27, 48, 47, 50, 71).
Anthropometric measurements of body composition.
Authors | Study design | Statistically significant results | ||||
---|---|---|---|---|---|---|
Measurements | Participants | Time on GAHT (months) | Comparison | TGW | TGM | |
Cohort study | ||||||
Iannantuoni et al. 2020 (44) | BMI, BW, WC | TGM = 157 | 3 | Baseline and 3 months | – | n.s. |
Victor et al. 2014 (75) | BW, BMI | TGM = 57 | 3 | Baseline and 3 months | – | n.s. |
Giltay et al. 1998 (17) | BMI, LBM | TGW = 17 | 4 | Baseline to 4 months | ↑BMI | ↑BMI, LBM |
TGM = 17 | ||||||
Berra et al. 2006 (27) | BW, BMI, WC, WHR, TFM%, TFM kg, LBM %, LBM kg | TGM = 16 | 6 | Baseline and after 6 months | – | ↑BW, BMI, WC, LBM kg, LBM % ↓TFM % |
Deutsch et al. 2015 (41) | BW, BMI | TGW = 16 | 6 | Baseline and after 6 months | n.s. | ↑BW, BMI |
TGM = 31 | ||||||
Yun et al. 2021 (15) | BW, BMI, LBM (total body, arm, leg, trunk, head, skeletal muscle mass index), adipose tissue (TBF (kg), arm, leg, trunk, head, android, gynoid, TBF (%), android/gynoid ratio, fat mass ratio, estimated visceral fat mass) | 38 | 6 | Baseline and 6 months | ↓Lean body mass (total body, arm, leg, skeletal muscle mass index), adipose tissue (android/gynoid ratio, fat mass ratio) | – |
↑Adipose tissue (TBF kg, leg, gynoid, TBF %) | ||||||
Valentine et al. 2021 (73) | BMI | TGM = 42 | 10.8 (2.6–25.7) | Baseline and 12 months | – | ↑BMI (short and long-term) |
CGW = 82 | ||||||
Stoffers et al. 2019 (69) | BW, BMI, BMI SDS | TGM = 62 | 12; 5–33 | Baseline, 6, 12, 24 months | – | ↑BW (6 and 12 months), BMI (6 months) |
Chandra et al. 2010 (26) | BMI | TGM = 12 | 12 | Baseline and after 12 months | – | n.s. |
Elbasan et al. 2023 (43) | BMI | TGM = 45 | 12 | TGM to CGW | – | n.s. |
CGW = 28 | Baseline to 6 and 12 months | |||||
Elbers et al. 1997 (23) | BW, BMI, body fat (skinfolds kg), body fat (bioimpedance kg), SAT area, sum of skinfolds | TGW = 17 | 12 | Baseline and 12 months | ↑BW, BMI, body fat (skinfolds kg), body fat (bioimpedance kg), subcutaneous fat area, sum of skinfolds | ↑BW, BMI, ↓body fat (skinfolds kg), body fat (bioimpedance kg), SAT area, sum of skinfolds |
TGM = 15 | ||||||
Elbers et al. 1999 (21) | BW, BMI, skinfolds mm (triceps, biceps, subscapular, suprailiac, para-umbilical, sum), % BF-SF, resistance, % BF-BIA, circumferences cm (arm, abdomen, hip, thigh), WHR, SF (abdomen, hip, thigh), visceral fat, V/S, muscle area, total area cm (abdomen, hip, thigh), WHR area | TGW = 20 | 12 | Baseline and 12 months | ↑BW, BMI, skinfolds mm (triceps, biceps, subscapular, suprailiac, para-umbilical, sum), % BF-SF, % BF-SF, resistance, % BF-BIA, circumferences cm (abdomen, hip, thigh), SF (abdomen, hip, thigh), visceral fat, total area (abdomen, hip, thigh), ↓V/S, muscle area | ↑BW, BMI, circumferences cm (arm), WHR, visceral fat, V/S, muscle area, WHR area, ↓skinfolds mm (triceps, biceps, suprailiac, para-umbilical, sum), % BF-SF, resistance, % BF-BIA, circumferences (hip), subcutaneous fat (abdomen, hip, thigh) |
TGM = 17 | ||||||
Elbers et al. 2003 (22) | BW, BMI, MRI (visceral fat, subcutaneous abdominal fat, subcutaneous gluteofemoral fat) | TGW = 20 | 12 | Compared to no GAHT treatment | ↑BW, BMI, MRI (visceral fat, subcutaneous abdominal fat, subcutaneous gluteofemoral fat) | ↑BW, BMI, MRI (visceral fat), ↓MRI (subcutaneous abdominal fat, subcutaneous gluteofemoral fat) |
TGM = 17 | ||||||
Klaver et al. 2018 (47) | BW, BMI, WC, hip circumference, WHR, body fat % change (TBF, arm, leg, trunk, android, gynoid), LBM % change (total, arm, leg, trunk, android, gynoid) | TGW = 179 | 12 | Baseline and 12 months | ↑BW, body fat % change (TBF, arm, trunk, android, gynoid, leg), hip (cm), ↓lean body mass % change (total, arm, trunk, gynoid, leg), WHR | ↑BW, lean body mass % change (total, arm, trunk, android, gynoid, leg), WHR, ↓body fat % change (total, arm, trunk, gynoid, leg), hip (cm) |
TGM = 162 | ||||||
Klaver et al. 2022 (50) | VAT, VAT/TBF, TBF, LBM | TGW = 179 | 12 | Baseline and 12 months | ↓VAT/TBF, LBM and ↑TBF | ↑VAT/TBF, LBM and ↓TBF |
TGM = 162 | ||||||
Mueller et al. 2007 (12) | BMI | TGM = 35 | 12 | Baseline and 12 months | – | ↑BMI |
Sanchez-Toscano et al. 2023 (66) | BW, BMI | TGW = 91 | 12 | Baseline and 12 months | ↑BW, BMI | ↑BW, BMI |
TGM = 136 | ||||||
Tebbens et al. 2023 (71) | BMI, liver fat content, VAT, SAT, VAT/SAT ratio | TGW = 8 | 12 | Baseline and 12 months | ↓Liver fat content (18/58 weeks), VAT/SAT (6/58 weeks) and ↑SAT (8–58 weeks) | ↑Lver fat content (52 weeks), VAT (52 weeks), SAT (at 6/52 weeks), VAT/SAT ratio (12/52 weeks) |
TGM = 10 | ||||||
Van Caenegem et al. 2015 (18) | BW, BMI, waist (cm), hip circumference (cm), WHR, TFM (kg), TFM (%), trunk fat mass, fat CSA forearm, fat CSA calf, LBM (kg), LBM (%), muscle CSA forearm, muscle CSA calf | TGM = 23 | 12 | Baseline and 12 months | – | ↓TFM (kg), TFM (%), trunk fat mass, fat CSA calf, ↑LBM (kg), LBM (%), muscle CSA forearm, muscle CSA calf |
CGW = 23 | ||||||
Wierckx et al. 2014 (14) | BW, BMI, TFM (kg), TFM (kg), WC, hip circumference, WHR | TGW = 53 | 12 | Baseline and 12 months oral vs transdermal in TGW | (Oral): ↑hip circumference, ↓WHR | ↑BW, BMI, LBM, WHR, ↓TFM, hip circumference |
TGM = 53 | (All): ↑TFM (kg), ↓LBM (kg) | |||||
Liu et al. 2021 (57) | BW, BMI | TGM = 65 | ≥12 | Baseline, 3–6, 6–12 and 12–24 months | n.s. | ↑BW and BMI |
TGF = 45 | ||||||
Shadid et al. 2020 (68) | BW, FFM, WHR, BMI | TGW = 55 | ≥12 | Baseline and after 12 months | ↑BMI, ↓WHR | ↑BW, BMI, WHR |
TGM = 35 | TGM: ↑BW, BMI, WHR | |||||
van Velzen et al. 2021 (74) | BMI | TGM = 15 | >12 | Baseline and after 12 months | ↑BMI | n.s. |
TGW = 15 | ||||||
Pei et al. 2024 (64) | BW, BMI, body fat kg (TBF, arm, leg, gynoid, corrected leg, android/gynoid, visceral), body fat % (TBF, arm, leg, gynoid, corrected leg, android/gynoid) | TGW (GAHT) = 59 | 12–36 | Compared to no GAHT treatment | ↑Body fat kg (TBF, arm, leg, gynoid, corrected leg), body fat % (TBF, arm, leg, gynoid, corrected leg) | – |
TGW (no GAHT) = 40 | ↓Body fat kg (android/gynoid, visceral), body fat % (android/gynoid) | |||||
Fernandez et al. 2016 (20) | BMI | TGW = 33 | 18 | Baseline to 3–6 and 6–18 months | n.s. | (3–6 months): ↑BMI |
TGM = 19 | ||||||
Vita et al. 2018 (76) | BMI | TGW = 21 | TGW = 31.3 (34.2) | Baseline and on GAHT | n.s. | n.s. |
TGM = 11 | TGM = 22.2 (11.5) | |||||
Cocchetti et al. 2021 (38) | BMI, BW, WC | TGW = 144 | 24 | Baseline, 12 and 24 months | ↓WC | n.s. |
TGM = 165 | ||||||
Colizzi et al. 2015 (25) | BMI, WC | TGW = 79 | 24 | Baseline, 12 and 24 months | ↑BMI (0–12, 12–24, 0–48), WC (0–12,0–24) | ↑BMI (0–12, 0–24), WC (0–24, 12–24) |
TGM = 43 | ||||||
Mueller et al. 2010 (11) | BMI, TFM, LBM | TGM = 45 | 24 | Baseline, 12 and 24 months | – | ↑LBM (12 and 24 months) |
Olson-Kennedy et al. 2018 (62) | BMI | TGW = 23 | 24 | Baseline and 24 months | n.s. | n.s. |
TGM = 35 | ||||||
Korpaisarn et al. 2021 (51) | BW | TGM = 39 | 25.52 ± 12.9 | Baseline and ∼24 months | – | n.s. |
Klaver et al. 2020 (49) | BMI | TGW = 71 | TGW = 26.4 (IQR: 13.2–37.2) | Before and after GAHT | ↑BMI | ↑BMI |
TGM = 121 | TGM = 34.8 (IQR: 20.4–40.8) | TGM to CGW/CGW | ||||
TGW to CGM/CGW | ||||||
Klaver et al. 2018 (48) | BW, BMI, WC, hip circumference, WHR, body fat % (TBF, android, gynoid), LBM %, SDS (for CG comparison) | TGW = 71 | TGW = 33.6; 19.2–40.8 | Before and after GAHT | ↑Body fat % (TBF, android, gynoid), WC, hip circumference, ↓LBM %, WHR | ↓Body fat % (TBF, gynoid), ↑LBM %, WHR, WC, hip circumference |
CGW: BMI, WC, hip circumference, WHR, % LBM, ↓% body fat (TBF, android, gynoid) SDS | ||||||
TGM = 121 | TGM = 36; 22.8–40.8 | TGM to CGW/CGW | CGM: ↑WC, hip circumference, WHR, % LBM, ↓% TBF | CGM: ↑BMI, hip circumference, % body fat (TBF, android, gynoid), ↓WHR, % LBM SDS | ||
TGW to CGM/CGW | SDS CGW: ↑hip circumference, % body fat (TBF, android, gynoid) ↓WHR, % LBM SDS | |||||
Suppakitjanusant et al. 2020 (70) | BMI | TGW (GAHT) = 46 | TGW = 79 ± 112 | Over 84-month period | ↑BMI | n.s. |
TGW (no GAHT) = 59 | TGM = 44 ± 41 (at baseline) | |||||
TGM (GAHT) = 15 | ||||||
TGM (no GAHT) = 25 | ||||||
Meyer et al. 2020 (28) | BMI | TGW = 155 | 48 | Baseline, 3–4, 10–14 months and 3–4 years | ↑BMI (10–14, 36–48) | ↑BMI (3–4, 10–14) |
TGM = 233 | ||||||
Martinez-Martin et al. 2023 (58) | BW | TGW = 149 | 60 | Baseline and 60 months | n.s. | ↑BW |
TGM = 153 | ||||||
Ott et al. 2011 (19) | BMI | TGW = 89 | 60 | Baseline and 60 months | n.s. | n.s. |
TGM = 80 | ||||||
Allen et al. 2021 (9) | BMI (absolute and % change) | TGW = 126 | ≤60 | Baseline, then every 3 months for 1 year, then every 6 months up to 5 years | n.s. | n.s. |
TGM = 91 | ||||||
Cross-sectional – cisgender cohort comparison | ||||||
Valentine et al. 2021 (73) | BMI | TGM = 42 | 10.8 (2.6–25.7) | Baseline TGM to CGW | – | ↓BMI (short and long term) |
CGW = 82 | ||||||
Nokoff et al. 2020 (61) | BMI, % LBM, % TFM | TGM = 19 (CGW = 42, CGM =19) | TGM = 11.2 ± 5.2 | TGM with matched CGW and CGM. TGW with matched CGW and CGM. | CGW: ↑% LBM and ↓% TFM | CGW: ↑% LBM and ↓% TFM |
TGW = 11 (CGW = 23, CGM = 24) | TGW = 12.3 ± 9.9 | CGM: ↓% LBM and ↑% TFM | CGM: ↓% LBM and ↑% TFM | |||
Van Caenegem et al. 2015 (18) | BW, BMI, waist (cm), hip circumference (cm), WHR, TFM (kg), TFM (%), trunk fat mass, fat CSA forearm, fat CSA calf, LBM (kg), LBM (%), muscle CSA forearm, muscle CSA calf | TGM = 23 | 12 | TGM to CGW | – | ↑WHR |
CGW = 23 | ||||||
Balik et al. 2022 (34) | BMI | CGW = 30 | >12 | CGW to TGM | - | n.s. |
TGM = 30 | ||||||
Deischinger et al. 2023 (40) | BMI, WC, intramyocardial, pancreatic, hepatic fat content, SAT/VAT | TGW = 16 | TGM = 13.2; 12–24 | TGW to CGM | CGM: SAT/VAT | n.s. |
TGM = 22 | TGW = 16.8; 12–26.4 | TGM to CGW | ||||
CGW = 16 | ||||||
CGM = 17 | ||||||
Tsatsanis et al. 2023 (72) | BMI | TGW = 16 | ≥36 | TGW to CGW/CGM | n.s. | CGW: ↑BMI |
TGM = 27 | TGM to CGW/CGM | |||||
CGW = 26 | ||||||
CGM = 30 | ||||||
Bretherton et al. 2021 (37) | LBM, android fat mass, gynoid fat mass, android:gynoid fat ratio, TFM | TGW = 41 | TGM = 44 | TGM to CGW and CGM | CGM: ↑TFM, android fat mass, gynoid fat mass and ↓android: gynoid fat ratio, LBM | CGW: ↑android:gynoid fat ratio, LBM |
TGM = 43 | TGW = 39 | CGW: ↑android fat mass, android:gynoid fat ratio, LBM | CGM: ↓LBM | |||
CGW = 48 | TGW to CGW and CGM | |||||
CGM = 30 | ||||||
New et al. 1997 (32) | BMI, WHR | TGW = 14 | 61 (70) | TGW to CGM/CGW | CGW: ↑WHR | – |
CGM = 14 | ||||||
CGW = 15 | ||||||
Cunha et al. 2023 (39) | BMI | TGM = 33 | 168 ± 96 | TGM to CGW and CGM | – | n.s. |
CGW = 55 | ||||||
CGM = 56 |
The arrows indicate either a significant increase/higher (↑) or significant decrease/lower (↓) parameter in transgender individuals in comparison with either the baseline (before GAHT) values or a cisgender control population. Statistical significance was indicated by a reported P value of <0.05 or a non-overlapping confidence interval (CI). Table is sorted in order of time on GAHT and study type.
Abbreviations: BF-SF, body fat-skinfolds; BIA, bioimpedance method; BMI, body mass index; BW, body weight; CGM, cisgender men; CGW, cisgender women; CSA, cross-sectional area; FFM, fat-free mass; GAHT, gender-affirming hormone therapy; LBM, lean body mass; SAT, subcutaneous adipose tissue; SDS, standard deviation score; SF, skinfolds; TBF, total body fat; TFM, total fat mass; TGM, transgender men; TGW, transgender women; VAT, visceral adipose tissue; WC, waist circumference; WHR, waist–hip ratio.
Twenty-two (22/51, 41%) studies investigated systolic blood pressure (SBP) and/or diastolic blood pressure (DBP) (9, 11, 12, 13, 14, 20, 22, 25, 26, 35, 38, 41, 44, 49, 51, 57, 58, 62, 66, 69, 75, 76). Four studies reported an increase in SBP (22, 25, 58, 66) and eight an increase in DBP (12, 25, 35, 49, 58, 62, 66, 69) in TGW. Three studies reported a conflicting decrease in SBP and/or DBP (35, 38, 41) with the rest showing no significant changes. In TGM, nine studies indicated increased SBP (11, 12, 25, 35, 49, 58, 62, 66, 69) and DBP (12, 49, 62, 66). In TGM, one study (76) reported a conflicting decrease in SBP and the rest showed no significant changes. Only one study reported on mean arterial pressure (26) and pulse pressure (13) with no statistical difference found.
Thirty-five (35/51, 69%) studies investigated lipid profile (Table 3). Some studies reported no difference in parameters of lipid profile in TGW, whilst the rest demonstrated a potential shift towards an anti-atherogenic profile through either a decrease in triglycerides (TG), total cholesterol (TC) and low-density lipoprotein (LDL) or an increase in high-density lipoprotein (HDL) (9, 14, 19, 20, 22, 25, 38, 39, 41, 50, 55, 57, 62, 64, 66, 68, 74). In TGM, most studies indicated a pro-atherogenic change, i.e., increased TG, TC and LDL and decreased HDL, under at least one measurement (9, 11, 12, 13, 14, 19, 22, 25, 26, 27, 38, 39, 44, 49, 50, 55, 57, 62, 63, 66, 68, 69, 73, 74, 75) (Table 3).
Lipid profile in transgender individuals.
Authors | Study design | Statistically significant results | ||||
---|---|---|---|---|---|---|
Measurements | Participants | Time on GAHT (months) | Comparison | TGW | TGM | |
Cohort study | ||||||
Iannantuoni et al. 2020 (44) | TC, LDL, HDL, TG, AIP | TGM = 157 | 3 | Baseline and 3 months | – | ↓HDL, ↑AIP |
Victor et al. 2014 (75) | TC, LDL, HDL, TG, AIP | TGM = 57 | 3 | Baseline and 3 months | – | ↓HDL, ↑AIP |
Giltay et al. 1998 (17) | TC, LDL | TGW = 17 | 4 | Baseline to 4 months | n.s. | n.s. |
TGM = 17 | ||||||
Berra et al. 2006 (27) | TC, LDL, HDL, TG | TGM = 16 | 6 | Baseline and after 6 months | – | ↓HDL |
Deutsch et al. 2015 (41) | TC, LDL, HDL, TG | TGW = 16 | 6 | Baseline and after 6 months | ↑HDL, TG | ↓HDL |
TGM = 31 | ||||||
SoRelle et al. 2019 (13) | TC, LDL, HDL, TG | TGW (no GAHT) = 87 | >6 | TGW no GAHT to GAHT | n.s. | ↑HDL, TG |
TGW (GAHT) = 133 | TGM no GAHT to GAHT | |||||
TGM (no GAHT) = 62 | ||||||
TGM (GAHT) = 89 | ||||||
Valentine et al. 2021 (73) | TC, LDL, HDL, TG | TGM = 42 | 10.8 (2.6–25.7) | Baseline TGM to CGW | – | ↓HDL |
CGW = 82 | Baseline and 12 months | |||||
Stoffers et al. 2019 (69) | TC, LDL, HDL, TG | TGM = 62 | 12; 5–33 | Baseline, 6, 12 and 24 months | – | ↓TC (6 months), HDL (6, 12 and 24 months) |
Chandra et al. 2010 (26) | TC, LDL, HDL, TG | TGM = 12 | 12 | Baseline and after 12 months | – | ↓HDL |
Elbers et al. 2003 (22) | TC, HDL, HDL-2, HDL-3, VLDL, FFA, TG, LDL, LDL size, LDL composition (TC, free, cholesterol ester, phospholipids, TG) | TGW = 20 | 12 | Compared to no GAHT treatment | ↑HDL, HDL-2, HDL-3, TG, LDL composition (TG), ↓LDL, LDL size, LDL composition (TC, free, cholesterol ester) | ↑TG, ↓HDL, HDL-2, HDL-3, LDL size, LDL composition (free, phospholipids) |
TGM = 17 | ||||||
Jacobeit et al. 2007 (46) | TC, LDL, HDL | TGM = 12 | 12 | Baseline and 12 months | – | ↓TC, LDL |
Klaver et al. 2022 (50) | TC, LDL, HDL, TG | TGW = 179 | 12 | Baseline and 12 months | ↓TC, TG, LDL, HDL | ↓HDL and ↑TG, LDL |
TGM = 162 | ||||||
Mueller et al. 2007 (12) | TC, LDL, HDL, TG | TGM = 35 | 12 | Baseline and 12 months | – | ↑TG, ↓HDL |
Pallotti et al. 2023 (63) | TC, LDL, HDL, TG | TGM = 52 | 12 | Baseline, 6 and 12 months | – | 6 months: ↓HDL |
Sanchez-Toscano et al. 2023 (66) | TC, LDL, HDL, TG | TGW = 91 | 12 | Baseline and 12 months | ↓TC | ↑TG, ↓HDL |
TGM = 136 | ||||||
Tebbens et al. 2023 (71) | TC, LDL, HDL, TG | TGW = 8 | 12 | Baseline and 12 months | n.s. | n.s. |
TGM = 10 | ||||||
Wierckx et al. 2014 (14) | TC, LDL, HDL, TG | TGW = 53 | 12 | Baseline and 12 months oral vs transdermal in TGW | (Oral): ↓TC, LDL, HDL (transdermal): ↓TC, LDL, TG | ↑TC, LDL, TG, ↓HDL |
TGM = 53 | ||||||
Liu et al. 2021 (57) | TC, LDL, HDL, TG | TGM = 65 | ≥12 | Baseline, 3–6, 6–12 and 12–24 months | ↓LDL (3–6 months) | ↑LDL (12–24 months), ↓HDL |
TGF = 45 | ||||||
Shadid et al. 2020 (68) | TC, LDL, HDL, TG | TGW = 55 | ≥12 | Baseline and after 12 months | ↓TC, LDL, HDL, TG | ↑LDL, ↓HDL |
TGM = 35 | ||||||
van Velzen et al. 2021 (74) | TC, LDL, HDL, TG, total HDL-CEC, aqueous diffusion HDL-CEC, ABCA1 HDL-CEC, ABCG1 HDL-CEC, HDL-CLC | TGM = 15 | >12 | Baseline and after 12 months | ↓TC, LDL, HDL, total HDL-CEC, ABCA1 HDL-CEC, aqueous diffusion HDL-CEC | ↑TG and ↓HDL, aqueous diffusion HDL-CEC |
TGW = 15 | ||||||
Pei et al. 2024 (64) | TC, LDL, HDL, TG | TGW (GAHT) = 59 | 12–36 | Compared to no GAHT treatment | ↓TC | – |
TGW (no GAHT) = 40 | ||||||
Fernandez et al. 2016 (20) | TC, LDL, HDL, TG | TGW = 33 | 18 | Baseline to 3–6 and 6–18 months | (3–6 months): ↑HDL | n.s. |
TGM = 19 | ||||||
Vita et al. 2018 (76) | TC, LDL, HDL, TG | TGW = 21 | TGW = 31.3 (34.2) | Baseline and on GAHT | n.s. | n.s. |
TGM = 11 | TGM = 22.2 (11.5) | |||||
Cocchetti et al. 2021 (38) | TC, LDL, HDL, TG | TGW = 144 | 24 | Baseline, 12 and 24 months | ↓TC, TG, LDL | ↓HDL and ↑TC, TG, LDL |
TGM = 165 | ||||||
Colizzi et al. 2015 (25) | TC, LDL, HDL, TG, AIP | TGW = 79 | 24 | Baseline, 12 and 24 months | ↑TG, TC, LDL, AIP (0–12, 12–24, 0–24), ↓HDL (0–12, 12–24, 0–24) | ↑TG (0–24, 12–24), ↑TC, LDL (0–12, 0–24, 12–24), AIP (0–12, 0–24) ↓HDL (12–24) |
TGM = 43 | ||||||
Mueller et al. 2010 (11) | TC, LDL, HDL, TG | TGM = 45 | 24 | Baseline, 12 and 24 months | – | ↑TG (12 and 24 months), ↓HDL (12 and 24 months) |
Olson-Kennedy et al. 2018 (62) | TC, HDL, TG | TGW = 23 | 24 | Baseline and 24 months | ↑HDL | ↓HDL, ↑TG |
TGM = 35 | ||||||
Korpaisarn et al. 2021 (51) | TC, LDL, HDL, TG | TGM = 39 | 25.52 ± 12.9 | Baseline and ∼24 months | – | n.s. |
Klaver et al. 2020 (49) | TC, LDL, HDL, TG | TGW = 71 | TGW = 26.4 (IQR: 13.2–37.2) | Before and after GAHT TGM to CGW/CGW | ↑TG | ↑TC, LDL, TG and ↓HDL |
TGM = 121 | TGM = 34.8 (IQR: 20.4–40.8) | TGW to CGM/CGW | ||||
Jacobeit et al. 2009 (46) | TC, LDL, HDL, TG | TGM = 17 | 36 | Baseline, 6, 12, 18, 24, 30, 36 months | – | 18 months; ↓TC, LDL |
Meyer et al. 2020 (28) | LDL, HDL, TG | TGW = 155 | 48 | Baseline, 3–4, 10–14 months and 3–4 years | n.s. | n.s. |
TGM = 233 | ||||||
Leemaqz et al. 2023 (55) | TC, LDL, HDL, TG | TGF = 170 | <57 | Baseline up to 57 months | ↑HDL and ↓LDL (22–57) | ↑TG, TC (22–57), LDL (22–57) and ↓HDL, TC (2–10) |
TGM = 196 | ||||||
Allen et al. 2021 (9) | LDL, HDL, TC, TG (absolute and % change) | TGW = 126 | ≤60 | Baseline, then every 3 months for 1 year, then every 6 months up to 5 years | ↑HDL at 3 and 60 months (relative) | ↑LDL after 36 months, and 48–60 months (% change), LDL at 54 months (absolute) and TC after 60 months, ↓HDL between 3, 9 and 18 months (relative), HDL 6–24 months (absolute) |
TGM = 91 | ||||||
Martinez-Martin et al. 2023 (58) | LDL, TG | TGW = 149 | 60 | Baseline and 60 months | ↑LDL | n.s. |
TGM = 153 | ||||||
Ott et al. 2011 (19) | TC, LDL, HDL, TG, TC/HDL | TGW = 89 | 60 | Baseline and 60 months | ↑TG, TC, HDL | ↑TG, TC, LDL, TC/HDL and ↓HDL |
TGM = 80 | ||||||
Cross-sectional – cisgender cohort comparison | ||||||
Gulanski et al. 2020 (40) | TC, LDL, HDL, TG | TGM = 11 | ≥3 | TGM to CGW | - | n.s. |
CGF = 20 | ||||||
Lim et al. 2020 (56) | TC | TGF = 26 | ≥6 (25.5) | TGW and CGM/CGW | CGM/CGW:↓TC | - |
CGM = 55 | ||||||
CGF = 98 | ||||||
Nokoff et al. 2020 (61) | TC, LDL, HDL, TG | TGM = 19 | TGM = 11.2 ± 5.2 | TGM with matched CGW and CGM | CGM: ↑HDL | CGW: ↓HDL |
(CGW = 42, CGM =19) | ||||||
TGW = 11 | TGW = 12.3 ± 9.9 | TGW with matched CGW and CGM | ||||
(CGW = 23, CGM = 24) | ||||||
Deischinger et al. 2023 (40) | TC, LDL, HDL, TG | TGW = 16 | TGM = 13.2; 12–24 | TGW to CGM | ↓LDL-C | ↓HDL-C |
TGM = 22 | TGW = 16.8; 12–26.4 | TGM to CGW | ||||
CGW = 16 | ||||||
CGM = 17 | ||||||
Millington & Chan 2021 (59) | TRL particle (total, very large, large, medium, small, very small) particles, LDL-C (mean and particle diameter), LDL particle (total, large, medium, small), ApoB, HDL-C (mean and particle diameter), HDL particle (total, large, medium, small), ApoA1 | TGM = 17 | 14.4 ± 9.6 | TGM to CGW | – | CGW: ↑LDL-C mean (model 2), total LDL, small LDL, ApoB (model 1/2) and ↓HDL (model 1/2), HDL particle diameter, large HDL, ApoA1 (model 1/2) |
CGM = 33 | TGM to CGM (model 1 and 2)* | CGM: ↑LDL-C mean (model 1), total LDL (model 1/2), ApoB (model 1) | ||||
CGF = 32 | ||||||
McCredie et al. 1998 (30) | TC, HDL | TGM = 12 | 38 (52); 2–177 | TGM to CGW | – | ↓HDL |
CGW = 12 | ||||||
Roberts et al. 2014 (65) | TC, LDL, HDL, TG | TGW = 55 | 48 (>6) | TGW to CGM/CGW | CGM: ↓LDL, ↑TG | – |
CGM = 20 | CGW: ↑TG | |||||
CGW = 20 | ||||||
New et al. 1997 (32) | TC, HDL, LDL, LDL particle size, TG | TGW = 14 | 61 (70) | TGW to CGM/CGW | CGM: ↑HDL, ↓LDL | – |
CGM = 14 | CGM/CGW: ↑TG, ↓LDL particle size | |||||
CGW = 15 | ||||||
Cunha et al. 2023 (39) | TC, LDL, HDL, TG | TGM = 33 | 168 ± 96 | TGM to CGW and CGM | – | CGM:↑HDL |
CGW = 55 | CGW:↓HDL | |||||
CGM = 56 |
The arrows indicate either a significant increase/higher (↑) or significant decrease/lower (↓) parameter in transgender individuals in comparison with either the baseline (before GAHT) values or a cisgender control population. Significance was indicated by a reported P-value of <0.05 or a non-overlapping confidence interval (CI). Table is sorted in order of time on GAHT and type of study. *Model 1 is the unadjusted P value, and model 2 is based on an adjusted P value based on BMI, race and age (59).
Abbreviations: ABCA1, ATP-binding cassette A1; ABCG1, ATP-binding cassette G1; AIP, atherogenic index of plasma; ApoA, apolipoprotein A; ApoAII, apolipoprotein II; ApoB, apolipoprotein B; CEC, cholesterol efflux capacity; CGM, cisgender men; CGW, cisgender women; CLC, cholesterol loading capacity; FFA, free fatty acids; GAHT, gender-affirming hormone therapy; HDL, high-density lipoprotein; LDL, low-density lipoprotein; TC, total cholesterol; TG, triglycerides; TGM, transgender men; TGW, transgender women; TRL, triglyceride-rich lipoprotein; VLDL, very low-density lipoprotein.
Twenty-seven (27/51, 53%) studies reported on a number of measures of insulin sensitivity, including fasting plasma glucose (FPG) or glucose (9, 13, 14, 22, 25, 26, 27, 33, 44, 49, 51, 54, 57, 58, 62, 63, 64, 66, 68, 71, 75, 76), insulin (14, 21, 22, 25, 27, 44, 49, 54, 57, 68, 75), homeostatic model assessment of insulin resistance (HOMA-IR) (25, 27, 44, 49, 50, 54, 57, 68, 75) and glycated haemoglobin (HbA1c) (33, 57, 60, 63, 66), with some studies carrying out a more focused analysis, such as a hyperinsulinaemic clamp test (22), gastric inhibitory polypeptide (GIP) (68), glucagon-like peptide-1 (GLP-1) (68), resistin (26), adiponectin (27, 54, 67) and leptin (23, 26, 27, 67). In TGW, six studies found no difference in any measures of insulin resistance (9, 33, 49, 62, 71, 76), four found an increased HOMA-IR (25, 50, 54, 68), six reported increased insulin levels (14, 21, 22, 25, 54, 68), and five demonstrated increased glucose (13, 25, 54, 58, 64). Adiponectin (67) and leptin (23, 67) were reported to increase with GAHT in TGW. Conflicting results with HbA1c were identified in TGW, where one study found lower levels (57), one study found higher levels (60), and other studies found no difference (33, 66). Eleven studies (9, 13, 21, 26, 33, 44, 60, 62, 63, 71, 75) demonstrated no changes across any measurements in TGM. Some studies indicated a reduced HOMA-IR (49, 50, 57), insulin (14, 49, 57) and FPG (51, 76) and an increased HbA1c (66) in TGM, with a conflicting increase in insulin (25) and FPG (58) also reported. TGM showed reduced leptin (23, 27, 67) and adiponectin (27, 67). An additional study indicated a reduced fasting GIP and area under curve (AUC) glucose, GIP and GLP-1 in TGW and increased AUC GIP and reduced GLP-1 in TGM (68). Another study showed reduced glucose utilisation in TGW and M value by hyperinsulinaemic clamp in both TGM and TGW (22).
Blood biomarkers of haematology, haemostasis and inflammation
Twenty-one (21/51, 41%) studies investigated blood parameters, all reporting in TGM and only 11 (2, 9, 14, 17, 24, 28, 36, 57, 60, 62, 76) in TGW; measurements included haematocrit (2, 9, 11, 12, 14, 20, 24, 26, 28, 36, 43, 45, 46, 51, 57, 60, 62, 63, 69, 76), haemoglobin (9, 11, 12, 17, 20, 28, 36, 43, 45, 46, 51, 57, 60, 62, 63, 69, 76), red blood cell count (RBC) (9, 20, 63, 76), mean corpuscle volume (MCV) (9, 43, 51, 76), red cell width distribution (9), packed cell volume (17), total homocysteine (tHcy) (17), platelet count (9, 26, 43, 63, 76), mean platelet volume (MPV) (43, 63) and white blood cell count (WBC) (9, 43, 51, 57, 63, 76). Neutrophils, lymphocytes and the neutrophil/lymphocyte ratio (NLR) were reported in one study (43). In TGW, there was consistent reporting of reduced haematocrit and haemoglobin with GAHT (2, 9, 14, 17, 24, 36, 57, 60, 62, 76), with the exception of one study that showed no difference in clinically relevant levels of haematocrit and haemoglobin (28). All studies reporting on packed cell volume (17), tHcy (17) and RBC (9, 76) in TGW indicated a reduction in these parameters, with one report of increased platelets (9). In TGM, 18 studies (2, 11, 12, 14, 20, 26, 36, 43, 45, 46, 60, 63, 69, 76) reported a significant increase in haematocrit and 14 studies (9, 11, 12, 20, 28, 36, 43, 45, 57, 60, 62, 63, 69, 76) also identified an increase in haemoglobin. Increased RBC (9, 20, 63, 76) and tHcy (17) were also reported in TGM.
Four (4/51, 8%) studies investigated blood markers of haemostasis and coagulation (2, 12, 54, 67); parameters included were fibrinogen (2, 12, 67), plasminogen activator-1 (PAI-1) (67), platelet factor 4 (PR-4) (54, 67), β-thromboglobulin (67), P-selectin (67), activated protein C resistance (APCr) ratio (2), coagulation factor II (FII) (2), coagulation factor IX (FIX) (2), coagulation factor XI (FXI) (2), protein S (2), protein C (2), D-dimer (54) and von Willebrand factor (vWF) (54), prothrombin time (PT) (12), activated partial thromboplastin time (APTT) (12) and thrombin time (12). In TGW, studies identified a significant increase in fibrinogen (67), PF-4 (67), β-thromboglobulin (67), FIX (2) and FXI (2) and a significant decrease in protein C (2) and D-dimer (54). In TGM, there was a reported increase in APCr ratio (2), FIX (2) and protein S (2) and a decrease in PF-4 (67), β-thromboglobulin (67), FII (2) and FXI (2).
Five (5/51, 10%) studies reported on blood markers of inflammation (43, 44, 54, 67, 75). Studies in TGW investigated high sensitivity C-reactive protein (hs-CRP) (43, 67), α-1-antitrypsin (67), tumor necrosis factor alpha (TNF-α) (54, 67), interleukin 6 (IL-6) (54, 67), IL-8 (54, 67), IL-10 (67), IL-22 (67), sCD14 (54), sCD163 (54) and fatty acid binding protein 4 (FABP-4) (54). Across these inflammatory markers, there was a reported reduction in hs-CRP, TNF-α, IL-6 and IL-8 (67). Studies in TGM investigated hs-CRP (43, 44, 67, 75), IL-6 (44, 67), TNF-α (44, 67), α-1-antitrypsin (67), IL-8 (67), IL-10 (67) and IL-22 (67). There was a reported elevation in hs-CRP (43, 67), TNF-α (44) and IL-6 (44) across these parameters in TGM.
Vascular function
Blood markers and studies investigating vascular function are shown in Table 4. In TGM, one study reported an increase in E-selectin and vascular cell adhesion protein 1 (VCAM-1) (44). Another study reported no difference in VCAM-1 or other measurements of endothelial function after GAHT treatment in TGM (67). In TGW, a decrease in VCAM-1 was found in a study (67); however, another study reported no difference (54). No changes in vascular function with GAHT in TGW were demonstrated; the only parameter investigated was heart rate, also showing change in TGM (52). One study investigated leucocyte–endothelium cell interactions in TGM, reporting a reduced rolling velocity and increased rolling flux and adhesion (44). In TGM, there was a reported reduction in oxygen consumption, membrane potential, glutathione and glutathione–glutathione disulphide ratio and an increased reactive oxygen species production (75).
Blood markers of vascular function and vascular function studies in transgender individuals.
Authors | Study design | Statistically significant results | ||||
---|---|---|---|---|---|---|
Measurements | Participants | Time on GAHT (months) | Comparison | TGW | TGM | |
Blood markers of vascular function | ||||||
Cohort study | ||||||
Iannantuoni et al. 2020 (44) | VCAM-1, ICAM-1, E-selectin | TGM = 157 | 3 | Baseline and 3 months | – | ↑E-selectin and VCAM-1 |
Lake et al. 2023 (54) | ET-1, VCAM-1 | TGW = 138 | 12 | Baseline, 6 and 12 months | n.s. | |
Schutte et al. 2022 (67) | VCAM-1 , p-selectin | TGF = 48 | 12 | Baseline, 3 and 12 months | (3 months): VCAM-1 | n.s. |
TGM = 47 | (12 months): ↓VCAM-1 | |||||
Cross-sectional – cisgender cohort comparison | ||||||
Lake et al. 2022 (53) | ET-1, VCAM-1, ENRAGE, oxLDL | TGW (GAHT) = 31 | TGW = 40; 32–48 | Compared to no GAHT | ↑ENRAGE, oxLDL and ↓ET-1 | – |
TGW (no-GAHT) = 16; CGM = 40 | CGM = 42; 35–52 | Treatment TGW to CGM | ||||
Vascular function studies | ||||||
Cohort studies | ||||||
Iannantuoni et al. 2020 (44) | Leucocyte–endothelium cell interaction assay: PMN rolling velocity, flux and PMN adhesion | TGM = 157 | 3 | Baseline and 3 months | – | ↑Leucocyte–endothelium interaction: ↓rolling velocity, ↑rolling flux and adhesion |
Victor et al. 2014 (75) | O2 consumption, membrane potential, ROS production, GSH, GSH/GSSG ratio | TGM = 57 | 3 | Baseline and 3 months | – | ↓O2 consumption, membrane potential, GSH, GSH/GSSG ratio, ↑ROS production |
Allen et al. 2021 (9) | HR (absolute and % change) | TGW = 126 | ≤60 | Baseline, then every 3 months for 1 year, then every 6 months up to 5 years | n.s. | n.s. |
TGM = 91 | ||||||
Cross-sectional – cisgender cohort comparison | ||||||
Gulanski et al. 2020 (40) | FMD | TGM = 11 | ≥3 | TGM to CGW | – | ↓FMD |
CGF = 20 | ||||||
Balik et al. 2022 (34) | CIMT | CGW = 30 | >12 | TGM to CGW | – | ↑CIMT |
TGM = 30 | ||||||
Deischinger et al. 2023 (40) | HR, CRP, pro-BNP | TGW = 16 | TGM = 13.2; 12–24 | TGW to CGM | ↑HR, pro-BNP | ↓pro-BNP |
TGM = 22 | TGW = 16.8; 12–26.4 | TGM to CGW | ||||
CGW = 16 | ||||||
CGM = 17 | ||||||
McCredie et al. 1998 (30) | Baseline flow, vessel size, FMD, NTG, hyperaemia (%) | TGM = 12 | 38 (52); 2–177 | TGM to CGW | – | ↑Vessel size, ↓NTG (%) |
CGW = 12 | ||||||
Duro et al. 2023 (42) | CAC % participants score 0, 1–99 and >100 | TGW = 25 (24) | TGW = 38.4; 24–166.8 | TGW and TGM to CG | ↓1–99 ,↑>100 | ↑1–99, ↓>100 |
TGM = 22 (16) | ||||||
TGM = 66; 22.8–148.8 | ||||||
New et al. 1997 (32) | FMD, baseline vessel size, aFMD, baseline flow, hyperaemia (% inc in flow), GTN, aGTN | TGW = 14 | 61 (70) | TGW to CGM/CGW | CGM: ↑FMD, aFMD, GTN, aGTN, ↓baseline vessel size | – |
CGM = 14 | ||||||
CGW = 15 | ||||||
Cunha et al. 2023 (39) | cf-PWV | TGM = 33 | 168 ± 96 | TGM to CGW and CGM | – | CGF/CGM: ↑cf-PWV |
CGW = 55 | ||||||
CGM = 56 |
The arrows indicate either a significant increase/higher (↑) or significant decrease/lower (↓) parameter in transgender individuals in comparison with either the baseline (before GAHT) values or a cisgender control population. Significance was indicated by a reported P value of <0.05 or a non-overlapping confidence interval (CI). Table is sorted in order of time on GAHT and study design.
Abbreviations: ABI, ankle-brachial index; aFMD, flow-mediated dilation adjusted for vessel size (percent flow-mediated vasodilation divided by baseline diameter (mm)); baPWV, brachial-ankle pulse wave velocity; CAC, coronary artery calcium score; CAI, carotid augmentation index; CAVI, cardio-ankle vascular index; cf-PWV, carotid-femoral pulse wave velocity; CGM, cisgender men; CGW, cisgender women; CIMT, carotid intima–media thickness; CRP, c-reactive protein; ET-1, endothelin-1; FMD, flow-mediated dilatation; GAHT, gender-affirming hormone therapy; GSH, glutathione; GSSG, glutathione disulphide; HR, heart rate; ICAM-1, intercellular cell adhesion molecule-1; NTG/GTN, nitroglycerine-induced dilatation; PMN, polymorphonuclear leucocytes; PWV, pulse wave velocity; ROS, reactive oxygen species; TGM, transgender men; TGW, transgender women; VCAM-1, vascular adhesion molecule-1.
Cross-sectional studies comparing treated and untreated transgender individuals
Two studies were cross-sectional, comparing a GAHT-treated and untreated group of TGM (16, 29) and TGW (52, 53) across all investigated CVD phenotypes (Table 5). In TGW, there are few changes in parameters with evidence of reduced waist circumference (52), RBC (13), Hb (13), Hct (13), soluble tumor necrosis factor receptor I and II (TNFRI/II) (53), IL-8 (53), endothelin 1 (ET-1) (53) and carotid intima–media thickness (CIMT) (52) and increased platelets (13), PAI-1 (53), extracellular newly-identified receptor for advanced glycation end-products (ENRAGE) (53), ox-LDL (53), HR (52), cardio-ankle vascular index (CAVI) (52) and cardiac troponin (52) (Table 5). No significant differences were reported in markers of insulin sensitivity, lipid profile or blood pressure. In TGM, there was evidence of increased SBP (29), DBP (29), MAP (29), TG (16, 29), TC (16, 29), LDL (16), ApoB (16), LDL/HDL (16), Hct (29), brachial-ankle pulse wave velocity (baPWV) (29) and reduced HDL (16, 29) and ApoA1/ApoB (16) (Table 5). No studies reported on any markers of inflammation or endothelial function, and no significant differences were reported on anthropometric measures or markers of insulin sensitivity and haemostasis.
Results of studies investigating CV risk factors in a GAHT treatment-naive population compared to GAHT-treated population.
Authors | Study design | Statistically significant results | |||
---|---|---|---|---|---|
Measurements | Participants | Time on GAHT (months) | TGW | TGM | |
Anthropometric | |||||
Emi et al. 2008 (29) | BW, BMI | TGM (GAHT) = 48 | 45 (38.1) | – | n.s. |
TGM (no GAHT) = 63 | |||||
Kulprachakarn et al. 2020 (52) | BMI, WC | TGW (GAHT) = 100 | 79.4 ± 6.2 | ↓WC | – |
TGW (no GAHT) = 100 | |||||
Blood pressure | |||||
Emi et al. 2008 (29) | SBP, DBP, MAP | TGM (GAHT) = 48 | 45 (38.1) | – | ↑SBP, DBP, MAP |
TGM (no GAHT) = 63 | |||||
Kulprachakarn et al. 2020 (52) | SBP, DBP | TGW (GAHT) = 100 | 79.4 ± 6.2 | n.s. | – |
TGW (no GAHT) = 100 | |||||
Lipid profile | |||||
Goh et al. 1995 (16) | TG, TC, LDL, HDL, Apo A1, ApoAII, ApoB, LDL/HDL, ApoA1/ApoB | TGM (GAHT) = 39 | 33; 6–120 | – | ↑TG, TC, LDL, ApoB, LDL/HDL, ↓HDL, ApoA1/ApoB |
TGM (no GAHT) = 29 | |||||
Emi et al. 2008 (29) | TC, HDL, TG | TGM (GAHT) = 48 | 45 (38.1) | – | ↑TC, TG and ↓HDL |
TGM (no GAHT) = 63 | |||||
Kulprachakarn et al. 2020 (52) | TC, LDL, HDL, TG | TGW (GAHT) = 100 | 79.4 ± 6.2 | n.s. | – |
TGW (no GAHT) = 100 | |||||
Insulin sensitivity | |||||
Lake et al. 2022 (53) | High molecular weight adiponectin | TGW (GAHT) = 31 | TGW = 40; 32–48 | n.s. | – |
TGW (no-GAHT) = 16 | CGM = 42; 35–52 | ||||
CGM = 40 | |||||
Emi et al. 2008 (29) | Glucose, HbA1c | TGM (GAHT) = 48 | 45 (38.1) | – | n.s. |
TGM (no GAHT) = 63 | |||||
Kulprachakarn et al. 2020 (52) | FPG | TGW (GAHT) = 100 | 79.4 ± 6.2 | n.s. | – |
TGW (no GAHT) = 100 | |||||
Haematology | |||||
SoRelle et al. 2019 (13) | WBC, RBC, Hb, Hct, MCV, red cell diameter width, platelets | TGW (no GAHT) = 87 | >6 | ↓RBC, Hb, Hct, ↑platelets | ↑RBC, Hb, Hct |
TGW (GAHT) = 133 | |||||
TGM (no GAHT) = 62 | |||||
TGM (GAHT) = 89 | |||||
Emi et al. 2008 (29) | Hct % | TGM (GAHT) = 48 | 45 (38.1) | – | ↑Hct % |
TGM (no GAHT) = 63 | |||||
Haemostasis | |||||
Lake et al. 2022 (53) | PAI-1, vWF, d-dimer | TGW (GAHT) = 31 | TGW = 40; 32–48 | GAHT: ↑PAI-1 | – |
TGW(no-GAHT) = 16 | CGM = 42; 35–52 | ||||
CGM = 40 | |||||
Emi et al. 2008 (29) | PT, PT-INR, APTT, fibrinogen | TGM (GAHT) = 48 | 45 (38.1) | – | n.s. |
TGM (no GAHT) = 63 | |||||
Inflammation | |||||
Lake et al. 2022 (53) | sCD14, sCD163, IL-6, IL-8, sTNFR I/II, P-selectin, LpPLA2 | TGW (GAHT) = 31 | TGW = 40; 32-48 | GAHT: ↓sTNFR II, IL-8 | – |
TGW(no-GAHT) = 16 | CGM = 42; 35–52 | ||||
CGM = 40 | |||||
Endothelial function | |||||
Lake et al. 2022 (53) | ET-1, VCAM-1, ENRAGE, oxLDL | TGW (GAHT) = 31 | TGW = 40; 32–48 | ↑ENRAGE, oxLDL and ↓ET-1 | – |
TGW (no-GAHT) = 16 | CGM = 42; 35–52 | ||||
CGM = 40 | |||||
Vascular function | |||||
Emi et al. 2008 (29) | HR, ABI, baPWV, CAI | TGM (GAHT) = 48 | 45 (38.1) | – | ↑baPWV |
TGM (no GAHT) = 63 | |||||
Kulprachakarn et al. 2020 (52) | HR, PWV, ABI, CAVI, CIMT, CRP, cardiac troponin I, pro-BNP | TGW (GAHT) = 100 | 79.4 ± 6.2 | ↑HR (model 1), CAVI (model 1), cardiac troponin (model) and ↓CIMT (model 1 and 2) | – |
TGW (no GAHT) = 100 |
The arrows indicate either a significant increase/higher (↑) or significant decrease/lower (↓) parameter in transgender individuals in comparison with an untreated transgender control population. Significance was indicated by a reported P-value of <0.05 or a non-overlapping confidence interval (CI). Table is sorted in order of time on GAHT.
Abbreviations: ABI, ankle-brachial index; ApoA, apolipoprotein A; ApoAII, apolipoprotein II; ApoB, apolipoprotein B; APTT, activated partial thromboplastin time; baPWV, brachial-ankle pulse wave velocity; BMI;,body mass index; BW, body weight; CAI, carotid augmentation index; CAVI, cardio-ankle vascular index; CGM, cisgender men; CGW, cisgender women; CIMT, carotid intima–media thickness; CRP, c-reactive protein; DBP, diastolic blood pressure; ET-1, endothelin-1; FPG, fasting plasma glucose; GAHT, gender-affirming hormone therapy; Hb, haemoglobin. HbA1c, glycated haemoglobin; Hct, haematocrit; HDL, high-density lipoprotein; HR, heart rate; LDL, low-density lipoprotein; LpPLA2, lipoprotein-associated phospholipase A2; MAP, mean arterial pressure; MCV, mean corpuscle volume; PAI-1, plasminogen activator-1; PT, prothrombin time; PT-INR, prothrombin time-international normalised ratio; PWV, pulse wave velocity; RBC, red blood cell; SBP, systolic blood pressure; TC, total cholesterol; TG, triglycerides; VCAM-1, vascular adhesion molecule-1; vWF, von Willebrand factor; WBC, white blood cell; WC, waist circumference.
Cross-sectional studies comparing to a cisgender population
Cross-sectional studies also compared transgender individuals on GAHT to cisgender counterparts (18/69, 26%) (Table 2). The same pattern is observed as in cohort studies for anthropometric measures (Table 2) and lipid profile (Table 3) when TGW are compared to CGM and TGM are compared to CGW. When compared to cisgender individuals of the affirmed gender (CGW in the case of TGW, for example), the opposite pattern was observed in both anthropometric measures (37, 61) and lipids ( 39, 65). Most studies provide no evidence for significant differences in MAP (78), PP (39) or SBP and DBP in both TGW (32, 40) and TGM (30, 40, 78). Nokoff and coworkers (61) indicated a decrease in SBP compared to CGM in both TGW and TGM, whereas Cunha and coworkers (39) indicated an increase in both SBP and DBP compared to CGW and CGM in TGM individuals.
Nine studies (31, 37, 39, 40, 53, 56, 61, 72, 78) compared to a cisgender cohort when considering measures of insulin sensitivity with no differences reported in TGW (37, 53, 72) or TGM (31, 37, 39) in three studies each. Reduced HbA1c (40, 56) and 1/(fasting insulin) (61) and increased insulin (40), HOMA-IR (40, 61), leptin (61), adiponectin (40) and betatrophin (40) were found in TGW in comparison with CGM. When comparing TGW to CGW, HbA1c (56) and leptin (31) were lower and fasting insulin (40) and HOMA-IR (40) were higher. In comparison with CGW, studies identified an increase in HbA1c (78) and HOMA-IR (72) and a reduced leptin (61) in TGM. Compared to CGM, only adiponectin (40) and betatrophin (40) were higher in TGM. Additional measures, such as C-peptide (37), IGF-1 (37) and ghrelin (31), and assessment of insulin resistance through the Matsuda index and Stumvoll I and II indices (40) indicated no difference in either TGW or TGM in comparison with cisgender individuals.
Only two studies (56, 65) compared haematological parameters in TGW to CGM and CGW, reporting on Hb, Hct and platelet count: Hb and Hct were higher, and platelets were lower compared to CGM. Four studies reported differences between TGM and CGW (34, 39, 43, 78), and one compared Hct to CGM showing an increase (39). Parameters compared to CGW were Hct (39, 43, 78), Hb (43, 78), WBC (34, 43), neutrophils (43), lymphocytes (43), NLR (34, 43), MCV (43), platelets (43), MPV (34, 43), platelet–lymphocyte ratio (PLR) (34) and lymphocyte–monocyte ratio (34). Hct (39, 78), WBC (34, 43) and neutrophil (43) counts were higher, and PLR (34) was lower in TGM compared to CGW.
No studies investigated markers of haemostasis in TGM compared to cisgender cohorts. In TGW, one study (56) investigated haemostatic function through global coagulation assays, which included analysis of efficiency of coagulation through thromboelastography, thrombin generation through a calibrated automated thrombogram and fibrin generation through an overall haemostatic potential assay in addition to PT and APTT (56). Global coagulation assays indicated hypercoagulability and increased fibrinolytic potential compared to CGM through an elevated maximum amplitude, endogenous thrombin potential, peak thrombin, overall fibrinolytic potential and lower K time. Compared to CGW, there were fewer reported differences, but these included reductions in alpha-angle, overall coagulation and haemostatic potential (56). This study and one other also investigated blood parameters of haemostasis, such as PT (56), APTT (56), FVIII (56), vWF (53, 56), D-dimer (53, 56), fibrinogen (56) and PAI-1 (53) in comparison with cisgender individuals. In comparison with CGM, studies reported an elevated D-dimer (56), fibrinogen (56) and PAI-1 (53) in TGW. In contrast to CGW, TGW had a higher D-dimer and reduced FVII and PT (56).
Two studies (53, 72) reported on blood markers of inflammation in comparison with cisgender populations: one study made comparison to both CGM and CGW, finding lower B-cell activating factor (BAFF) and higher lipopolysaccharide-TNF (LPS–TNF) when comparing TGM to CGW and higher BAFF and lower TNF when comparing TGW and TGM to CGM (72), and the other study (53) investigated a wider range of inflammatory parameters in TGW compared to CGM, including sCD14, sCD163, IL-6, IL-8, sTNFR I/II, P-selectin and LpPLA2, reporting an increase in sTNFRII and IL-8.
No cross-sectional studies compared blood markers of vascular function in TGM. TGW reported a higher oxidative LDL and ENRAGE and a lower ET-1 than CGM (53). Three studies investigated vascular function in TGW, showing an increased coronary artery score (32) HR, brain natriuretic peptide (pro-BNP) (40), flow-mediated dilatation (FMD), and nitroglycerine-induced dilatation and a reduced baseline vessel size (32). Six studies investigated vascular function parameters in TGM, showing an increased CIMT (34), PWV (39), vessel size (30) and a reduced FMD (78), pro-BNP (40), nitroglycerine-induced dilatation (30) and coronary artery calcium (CAC) scoring (42).
Discussion
This systematic review identified 69 studies, which reported on traditional CV risk factors and vascular function in transgender individuals. In TGM, evidence suggested potentially negative changes in CV risk profile and, in TGW, potentially positive changes, as shown in Fig. 2. The direction of evidence in the cohort and cisgender comparison cross-sectional studies within each CV phenotype followed the same pattern. More weight can be attributed to studies comparing to a baseline transgender population due to the presence of confounding variables, such as behavioural risk factors, socio-economic differences and health inequalities, which are not directly comparable to cisgender counterparts. Studies comparing to an untreated transgender cohort were scarce but corroborated these findings. In TGW, some of the research opposed this, indicating a procoagulant shift and an increased insulin sensitivity. The studies were inconsistent concerning blood pressure in both TGW and TGM. Across factors such as body composition, insulin sensitivity, haemostasis, endothelial function and vascular function, the evidence was inconclusive for TGM or TGW or both, with studies reporting on a high number of markers with inconsistencies in results. Currently, no standard practice exists in study design, which is required to gain further knowledge in this field. It would be best to compare to a hormone-naive transgender cohort, but where this is not possible, it is always beneficial to compare to both a CGM and CGW control group in both cross-sectional and cohort studies. Of note is the study by Suppakitjanusant and coworkers (70), where transgender individuals were separated by prior hormone use. In addition, great heterogeneity was identified in terms of reported outcomes, suggesting that development of a core outcome set for standardisation of research into the CV effects of gender identity dysphoria would be useful for future study comparison and synchronicity.
Another important finding in the research to be addressed is the substantial dropout rate reported by some studies (9, 20, 35, 51, 55, 69). These attrition rates and low sample sizes pose a major problem in the ability to assess CV risk in this population. It is important to build cohorts of transgender individuals to follow for longer term whilst on GAHT. Within these cohorts, real-time feedback from transgender individuals can be given to provide insight into their health priorities.
In addition, in relation to ethnicity, there is a lack of diversity within transgender research. All studies have different inclusion and exclusion criteria, making it difficult to make meaningful comparisons. A number of the included studies investigating differences in CV risk controlled for previous hormone exposure (2, 11, 14, 17, 18, 20, 21, 22, 23, 24, 26, 36, 38, 43, 44, 48, 57, 60, 63, 67, 68, 71, 74). Some studies (11, 12, 17, 22, 30, 37, 39, 42, 43, 44, 52, 56, 57, 59, 61, 72, 75, 78) reported on exclusion based on CV risk to varying extents, whereas the rest did not mention excluding for CV disease or CV risk at baseline.
Overall interpretation of findings
The conclusion of the review by Moreira Allgayer and coworkers (10) was a reported neutral or beneficial effect in TGW and an increased risk of subclinical atherosclerosis in TGM. Considering additional CV risk phenotypes provides more evidence to support these conclusions. Some results have been conflicting, such as the evidence of insulin sensitivity being increased in TGM and reduced in TGW and the inconsistencies in published work concerning BP. There is a paucity of research across non-traditional measurements of CV risk in transgender individuals; however, the research on vascular function is severely lacking. In cross-sectional studies comparing to CGW or CGM, there were often more differences between CGM and less between CGW in TGW, indicating a shift towards the affirmed gender. The converse was often observed in TGM. This was particularly evident concerning lipids, insulin and blood cells, where there are known sex differences. More evidence is required to determine whether GAHT is conferring the risk of the affirmed gender or posing additional CVD risk. No clear patterns were observed regarding time to development of CV risk in individuals who were on GAHT for a longer period. The longest study of transgender individuals before and after GAHT spanned over 5 years (9) indicating stabilisation of haemoglobin and HDL as early as 3–6 months. Some values, such as platelets and LDL, increased later at 18 and 36 months, respectively, providing some evidence of longer-term changes with GAHT. However, based on the lack of a clear trend in both that study and the other studies of different lengths included as part of this review, it is likely that most changes in CV risk are established early into treatment. Much more research is required to confirm this hypothesis.
The changes in parameters of CVD risk reported were mostly either neutral or beneficial in the case of TGW and negative in the case of TGM. However, studies reporting a higher incidence of CVD events in TGW are well documented with less conclusive evidence in TGM (1, 4, 5). The incidence of CV events does not coincide with the mechanistic data. It is unclear whether this is because studies reporting on incidence report on a slightly higher age range with as high as 65+ years (4) and that the early changes are not clinically relevant or become more apparent with age. Studies on CVD incidence report on transgender individuals as a whole and not only on individuals on GAHT, which may explain the inconsistency. This hypothesis was supported in a recent systematic review and meta-analysis (6), which compared two of the studies that frequently reported opposing effects on CVD risk. In the study where approximately half of the participants were on GAHT, a smaller risk of CVD events was reported; however, the study in which all individuals were treated with GAHT reported a larger risk of CVD events. Haemostatic and insulin sensitivity markers in TGW supported an elevated incidence of CVD. It is possible that these play an important role in CVD in transgender individuals. Elevated haematocrit has been associated with CV disease, through increased blood viscosity and association with platelet adhesion, but it is not clear what the strength of this association is and its implications clinically (79). Elevated haemoglobin levels have been associated with increased TC and TG, also conferring an additional CV disease risk (79). Reduced insulin sensitivity, otherwise termed insulin resistance, has been shown to be an independent predictor of high CV risk, including hypertension, obesity and dyslipidaemia (80) which is in direct conflict with the potential CV risk profile in TGW in this review.
Strengths and limitations of this review
As far as we are aware, this is the first review to investigate such a wide range of phenotypes. There have been a variety of reviews on a more limited number of phenotypes (80, 81, 82, 83) published as early as 2017, resulting in gaps within the literature. A recent review by Stobbe and coworkers (7) summarised studies investigating blood pressure, lipid profile and glucose from 2001 to July 2019. Studies reported changes in blood pressure but within normal range, inconsistent change in lipids with pro-atherogenic changes in TGM and an absence of data on glucose, corroborating these results. This review includes data on body composition, blood pressure, lipid profiles, insulin sensitivity and blood cell count, thereby expanding on the number of phenotypes. This review was unable to carry out a meta-analysis due to a wide range of measurements and units of measurement across the studies, but this is similar to other reviews within this field. We did not consider the addition of anti-androgens in TGW, which means oestrogen may not be the only influencing factor on the measure of CV risk or vascular function.
Future work and implications
Future work is required to address the gaps within the data in this field. In particular, there is a need for studies to determine mechanisms behind the increased CV risk and assessment of laboratory values to assess what the reference values should be to assess CV risk. Study design must be refined to allow comparable results. All studies should aim to compare to a cisgender population of both men and women for each transgender individual. Studies should be carried out over a longer term. The production of a more ethnically diverse, wider age range and larger sample size of transgender individuals is of high importance. It is important to distinguish between lifestyle, social factors and biological factors by appropriate inclusion/exclusion criteria. Studies should exclude prior CV disease at the baseline and, if possible, recruit individuals before they begin GAHT. Incentivisation and public outreach will be required to recruit transgender individuals and make individuals aware of their CV health risks due to the high loss of follow-up across studies (84).
Understanding the mechanisms behind the increased risk of CVD in transgender individuals will allow screening of those who are at higher risk, provide therapies or lifestyle interventions and provide evidence-based knowledge to individuals undergoing GAHT. Monitoring and interpreting clinical laboratory values is important as these are used to support clinical decisions, and it is important that transgender individuals receive the same high-quality evidence-based care as CGM and CGM. If physicians are aware of the mechanisms and risk factors for CVD in TGM and TGW, preventative measures can be implemented to reduce the burden of CVD in this population. With further understanding of the laboratory and clinical markers of CVD risk, this information can be utilised in risk stratification of transgender individuals when they seek healthcare treatment. The goal would be to produce a review similar to that of Bays and coworkers (8) to provide an overview of the clinical considerations in preventative cardiology in the transgender population in the future. This will also extend further than the transgender population to those who have received sex hormone-based therapies for breast cancer, prostate cancer, HRT for menopause, oral contraceptives and puberty blockers.
Conclusion
Evidence suggests beneficial or neutral changes in the CVD risk profile in TGW and negative changes in TGM with administration of GAHT. The evidence is, however, often conflicting with studies reporting results contradicting this within each CV phenotype. This suggests that additional, larger and more diverse studies with a standardised study design would allow comparison and meta-analyses to be conducted. This would help elucidate why there is a higher rate of CV events in the transgender population and whether GAHT is a contributing factor. In the meantime, there is a need to counsel TGM regarding lifestyle modifications for CVD.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the work reported.
Funding
This work has been supported by a collaborative PhD studentship to KAM funded by the Universities of Glasgow and Sydney. KAM and CD are also supported by a Programme of Research on long-term health outcomes of people accessing gender identity healthcare funded by the Scottish Government. PC, ALH and CD are supported by the British Heart Foundation (RE/18/6/34217).
Author contribution statement
KAM helped with methodology design, data collection, investigation, formal analysis and writing the original draft. ALH contributed to data collection, formal analysis, investigation, writing, reviewing, editing and supervision. PC helped with conceptualisation, writing, reviewing and editing. CD helped with conceptualisation, methodology design, writing, reviewing, editing and supervision.
Data availability
The data underlying this article will be shared on reasonable request to the corresponding author.
References
- 1↑
Alzahrani T, Nguyen T, Ryan A, et al. Cardiovascular disease risk factors and myocardial infarction in the transgender population. Circ Cardiovasc Qual Outcomes 2019 12 e005597. (https://doi.org/10.1161/circoutcomes.119.005597)
- 2↑
Scheres LJJ, Selier NLD, Nota NM, et al. Effect of gender-affirming hormone use on coagulation profiles in transmen and transwomen. J Thromb Haemost 2021 19 1029–1037. (https://doi.org/10.1111/jth.15256)
- 3↑
Vitale C, Mendelsohn ME & Rosano GMC. Gender differences in the cardiovascular effect of sex hormones. Nat Rev Cardiol 2009 6 532–542. (https://doi.org/10.1038/nrcardio.2009.105)
- 4↑
Caceres BA, Jackman KB, Edmondson D, et al. Assessing gender identity differences in cardiovascular disease in US adults: an analysis of data from the 2014-2017 BRFSS. J Behav Med 2020 43 329–338. (https://doi.org/10.1007/s10865-019-00102-8)
- 5↑
Howerton I & Harris JK. Transgender identity and cardiovascular disease. Transgend Health 2022 7 407–415. (https://doi.org/10.1089/trgh.2020.0188)
- 6↑
van Zijverden LM, Wiepjes CM, van Diemen JJK, et al. Cardiovascular disease in transgender people: a systematic review and meta-analysis. Eur J Endocrinol 2024 190 S13–S24. (https://doi.org/10.1093/ejendo/lvad170)
- 7↑
Stobbe PR, Tatit UL, Guedes MH, et al. Cardiometabolic impacts of gender-affirming hormone therapy in transgender persons: a systematic review. Rev Caribeña Cienc Soc 2024 13 e3856. (https://doi.org/10.55905/rcssv13n4-013)
- 8↑
Bays HE, Taub PR, Epstein E, et al. Ten things to know about ten cardiovascular disease risk factors. Am J Prev Cardiol 2021 5 100149. (https://doi.org/10.1016/j.ajpc.2021.100149)
- 9↑
Allen AN, Jiao R, Day P, et al. Dynamic impact of hormone therapy on laboratory values in transgender patients over time. J Appl Lab Med 2021 6 27–40. (https://doi.org/10.1093/jalm/jfaa192)
- 10↑
Moreira Allgayer RMC, Borba GDS, Moraes RS, et al. The effect of gender-affirming hormone therapy on the risk of subclinical atherosclerosis in the transgender population: a systematic review. Endocr Pract 2023 29 498–507. (https://doi.org/10.1016/j.eprac.2022.12.017)
- 11↑
Mueller A, Haeberle L, Zollver H, et al. Effects of intramuscular testosterone undecanoate on body composition and bone mineral density in female-to-male transsexuals. J Sex Med 2010 7 3190–3198. (https://doi.org/10.1111/j.1743-6109.2010.01912.x)
- 12↑
Mueller A, Kiesewetter F, Binder H, et al. Long-term administration of testosterone undecanoate every 3 months for testosterone supplementation in female-to-male transsexuals. J Clin Endocrinol Metab 2007 92 3470–3475. (https://doi.org/10.1210/jc.2007-0746)
- 13↑
SoRelle JA, Jiao R, Gao E, et al. Impact of hormone therapy on laboratory values in transgender patients. Clin Chem 2019 65 170–179. (https://doi.org/10.1373/clinchem.2018.292730)
- 14↑
Wierckx K, Van Caenegem E, Schreiner T, et al. Cross-sex hormone therapy in trans persons is safe and effective at short-time follow-up: results from the European network for the investigation of gender incongruence. J Sex Med 2014 11 1999–2011. (https://doi.org/10.1111/jsm.12571)
- 15↑
Yun Y, Kim D & Lee ES. Effect of cross-sex hormones on body composition, bone mineral density, and muscle strength in trans women. J Bone Metab 2021 28 59–66. (https://doi.org/10.11005/jbm.2021.28.1.59)
- 16↑
Goh HH, Loke DF & Ratnam SS. The impact of long-term testosterone replacement therapy on lipid and lipoprotein profiles in women. Maturitas 1995 21 65–70. (https://doi.org/10.1016/0378-5122(94)00861-z)
- 17↑
Giltay EJ, Hoogeveen EK, Elbers JM, et al. Effects of sex steroids on plasma total homocysteine levels: a study in transsexual males and females. J Clin Endocrinol Metab 1998 83 550–553. (https://doi.org/10.1210/jcem.83.2.4574)
- 18↑
Van Caenegem E, Wierckx K, Taes Y, et al. Body composition, bone turnover, and bone mass in trans men during testosterone treatment: 1-year follow-up data from a prospective case-controlled study (ENIGI). Eur J Endocrinol 2015 172 163–171. (https://doi.org/10.1530/eje-14-0586)
- 19↑
Ott J, Aust S, Promberger R, et al. Cross-sex hormone therapy alters the serum lipid profile: a retrospective cohort study in 169 transsexuals. J Sex Med 2011 8 2361–2369. (https://doi.org/10.1111/j.1743-6109.2011.02311.x)
- 20↑
Fernandez JD & Tannock LR. Metabolic effects of hormone therapy in transgender patients. Endocr Pract 2016 22 383–388. (https://doi.org/10.4158/ep15950.or)
- 21↑
Elbers JM, Asscheman H, Seidell JC, et al. Effects of sex steroid hormones on regional fat depots as assessed by magnetic resonance imaging in transsexuals. Am J Physiol 1999 276 E317–E325. (https://doi.org/10.1152/ajpendo.1999.276.2.e317)
- 22↑
Elbers JM, Giltay EJ, Teerlink T, et al. Effects of sex steroids on components of the insulin resistance syndrome in transsexual subjects. Clin Endocrinol 2003 58 562–571. (https://doi.org/10.1046/j.1365-2265.2003.01753.x)
- 23↑
Elbers JM, Asscheman H, Seidell JC, et al. Reversal of the sex difference in serum leptin levels upon cross-sex hormone administration in transsexuals. J Clin Endocrinol Metab 1997 82 3267–3270. (https://doi.org/10.1210/jcem.82.10.4284)
- 24↑
Defreyne J, Vantomme B, Van Caenegem E, et al. Prospective evaluation of hematocrit in gender-affirming hormone treatment: results from European network for the investigation of gender incongruence. Andrology 2018 6 446–454. (https://doi.org/10.1111/andr.12485)
- 25↑
Colizzi M, Costa R, Scaramuzzi F, et al. Concomitant psychiatric problems and hormonal treatment induced metabolic syndrome in gender dysphoria individuals: a 2 year follow-up study. J Psychosom Res 2015 78 399–406. (https://doi.org/10.1016/j.jpsychores.2015.02.001)
- 26↑
Chandra P, Basra SS, Chen TC, et al. Alterations in lipids and adipocyte hormones in female-to-male transsexuals. Int J Endocrinol 2010 2010 1–4. (https://doi.org/10.1155/2010/945053)
- 27↑
Berra M, Armillotta F, D’Emidio L, et al. Testosterone decreases adiponectin levels in female to male transsexuals. Asian J Androl 2006 8 725–729. (https://doi.org/10.1111/j.1745-7262.2006.00204.x)
- 28↑
Meyer G, Mayer M, Mondorf A, et al. Safety and rapid efficacy of guideline-based gender-affirming hormone therapy: an analysis of 388 individuals diagnosed with gender dysphoria. Eur J Endocrinol 2020 182 149–156. (https://doi.org/10.1530/eje-19-0463)
- 29↑
Emi Y, Adachi M, Sasaki A, et al. Increased arterial stiffness in female-to-male transsexuals treated with androgen. J Obstet Gynaecol Res 2008 34 890–897. (https://doi.org/10.1111/j.1447-0756.2008.00857.x)
- 30↑
McCredie RJ, McCrohon JA, Turner L, et al. Vascular reactivity is impaired in genetic females taking high-dose androgens. J Am Coll Cardiol 1998 32 1331–1335. (https://doi.org/10.1016/s0735-1097(98)00416-1)
- 31↑
Resmini E, Andraghetti G, Rebora A, et al. Leptin, ghrelin, and adiponectin evaluation in transsexual subjects during hormonal treatments. J Androl 2008 29 580–585. (https://doi.org/10.2164/jandrol.108.004952)
- 32↑
New G, Timmins KL, Duffy SJ, et al. Long-term estrogen therapy improves vascular function in male to female transsexuals. J Am Coll Cardiol 1997 29 1437–1444. (https://doi.org/10.1016/s0735-1097(97)00080-6)
- 33↑
Anike O, Zhang Q, Bhasin S, et al. Association between gender-affirming hormone therapy and measures of glucose metabolism: a longitudinal study. J Clin Endocrinol Metab 2024 dgae353. (https://doi.org/10.1210/clinem/dgae353)
- 34↑
Balik AO, Vural F, Alpogan O, et al. The effects of testosterone on transgender males on carotid intima-media thickness and serum inflammatory markers compared within patients with polycystic ovary syndrome. Gynecol Endocrinol 2022 38 771–775. (https://doi.org/10.1080/09513590.2022.2112942)
- 35↑
Banks K, Kyinn M, Leemaqz SY, et al. Blood pressure effects of gender-affirming hormone therapy in transgender and gender-diverse adults. Hypertension 2021 77 2066–2074. (https://doi.org/10.1161/hypertensionaha.120.16839)
- 36↑
Boekhout-Berends ET, Wiepjes CM, Nota NM, et al. Changes in laboratory results in transgender individuals on hormone therapy – a retrospective study and practical approach. Eur J Endocrinol 2023 188 457–466. (https://doi.org/10.1093/ejendo/lvad052)
- 37↑
Bretherton I, Spanos C, Leemaqz SY, et al. Insulin resistance in transgender individuals correlates with android fat mass. Ther Adv Endocrinol Metab 2021 12 2042018820985681. (https://doi.org/10.1177/2042018820985681)
- 38↑
Cocchetti C, Castellini G, Iacuaniello D, et al. Does gender-affirming hormonal treatment affect 30-year cardiovascular risk in transgender persons? A two-year prospective European study (ENIGI). J Sex Med 2021 18 821–829. (https://doi.org/10.1016/j.jsxm.2021.01.185)
- 39↑
Cunha FS, Bachega T, Costa EMF, et al. Arterial stiffness in transgender men receiving long-term testosterone therapy. J Endocr Soc 2023 7 bvad040. (https://doi.org/10.1210/jendso/bvad040)
- 40↑
Deischinger C, Slukova D, Just I, et al. Effects of gender-affirming hormone therapy on cardiovascular risk factors focusing on glucose metabolism in an Austrian transgender cohort. Int J Transgend Health 2023 24 499–509. (https://doi.org/10.1080/26895269.2022.2123425)
- 41↑
Deutsch MB, Bhakri V & Kubicek K. Effects of cross-sex hormone treatment on transgender women and men. Obstet Gynecol 2015 125 605–610. (https://doi.org/10.1097/aog.0000000000000692)
- 42↑
Duro T, McClain M, Aragon KG, et al. Risk assessment and coronary artery calcium scoring in transgender and gender-diverse individuals receiving gender-affirming hormone therapy. Endocr Pract 2023 29 229–234. (https://doi.org/10.1016/j.eprac.2022.12.007)
- 43↑
Elbasan O & Ustay O. Effect of cross-sex hormone therapy on hematological parameters in transmen: a 1-year follow-up study. Endocrinol Res Pract 2023 27 221–226.
- 44↑
Iannantuoni F, Salazar JD, Martinez de Maranon A, et al. Testosterone administration increases leukocyte-endothelium interactions and inflammation in transgender men. Fertil Steril 2021 115 483–489. (https://doi.org/10.1016/j.fertnstert.2020.08.002)
- 45↑
Jacobeit JW, Gooren LJ & Schulte HM. Long-acting intramuscular testosterone undecanoate for treatment of female-to-male transgender individuals. J Sex Med 2007 4 1479–1484. (https://doi.org/10.1111/j.1743-6109.2007.00556.x)
- 46↑
Jacobeit JW, Gooren LJ & Schulte HM. Safety aspects of 36 months of administration of long-acting intramuscular testosterone undecanoate for treatment of female-to-male transgender individuals. Eur J Endocrinol 2009 161 795–798. (https://doi.org/10.1530/eje-09-0412)
- 47↑
Klaver M, de Blok CJM, Wiepjes CM, et al. Changes in regional body fat, lean body mass and body shape in trans persons using cross-sex hormonal therapy: results from a multicenter prospective study. Eur J Endocrinol 2018 178 163–171. (https://doi.org/10.1530/eje-17-0496)
- 48↑
Klaver M, de Mutsert R, Wiepjes CM, et al. Early hormonal treatment affects body composition and body shape in young transgender adolescents. J Sex Med 2018 15 251–260. (https://doi.org/10.1016/j.jsxm.2017.12.009)
- 49↑
Klaver M, de Mutsert R, van der Loos M, et al. Hormonal treatment and cardiovascular risk profile in transgender adolescents. Pediatrics 2020 145. (https://doi.org/10.1542/peds.2019-0741)
- 50↑
Klaver M, Van Velzen D, De Blok C, et al. Change in visceral fat and total body fat and the effect on cardiometabolic risk factors during transgender hormone therapy. J Clin Endocrinol Metab 2022 107 E153–E164. (https://doi.org/10.1210/clinem/dgab616)
- 51↑
Korpaisarn S, Chiewchalermsri D, Arunakul J, et al. Effects of testosterone treatment on transgender males: a single-institution study. SAGE Open Med 2021 9 20503121211051546. (https://doi.org/10.1177/20503121211051546)
- 52↑
Kulprachakarn K, Ounjaijean S, Rerkasem K, et al. Cardiovascular disease risk factors among transgender women in Chiang Mai, Thailand. Am J Cardiovasc Dis 2020 10 124–130.
- 53↑
Lake JE, Wang R, Barrett BW, et al. Trans women have worse cardiovascular biomarker profiles than cisgender men independent of hormone use and HIV serostatus. AIDS 2022 36 1801–1809. (https://doi.org/10.1097/qad.0000000000003346)
- 54↑
Lake JE, Miao H, Bowman ER, et al. Gender-affirming hormone therapy decreases d-dimer but worsens insulin sensitivity in transgender women. HIV Med 2023 24 1144–1149. (https://doi.org/10.1111/hiv.13522)
- 55↑
Leemaqz SY, Kyinn M, Banks K, et al. Lipid profiles and hypertriglyceridemia among transgender and gender diverse adults on gender-affirming hormone therapy. J Clin Lipidol 2023 17 103–111. (https://doi.org/10.1016/j.jacl.2022.11.010)
- 56↑
Lim HY, Leemaqz SY, Torkamani N, et al. Global coagulation assays in transgender women on oral and transdermal estradiol therapy. J Clin Endocrinol Metab 2020 105 e2369–e2377. (https://doi.org/10.1210/clinem/dgaa262)
- 57↑
Liu YH, Wu TH, Chu CH, et al. Metabolic effects of cross-sex hormone therapy in transgender individuals in Taiwan. J Chin Med Assoc 2021 84 267–272. (https://doi.org/10.1097/jcma.0000000000000475)
- 58↑
Martinez-Martin FJ, Kuzior A, Hernandez-Lazaro A, et al. Incidence of hypertension in young transgender people after a 5-year follow-up: association with gender-affirming hormonal therapy. Hypertens Res 2023 46 219–225. (https://doi.org/10.1038/s41440-022-01067-z)
- 59↑
Millington K & Chan YM. Lipoprotein subtypes after testosterone therapy in transmasculine adolescents. J Clin Lipidol 2021 15 840–844. (https://doi.org/10.1016/j.jacl.2021.09.051)
- 60↑
Millington K, Lee JY, Olson-Kennedy J, et al. Laboratory changes during gender-affirming hormone therapy in transgender adolescents. Pediatrics 2024 153 e2023064380. (https://doi.org/10.1542/peds.2023-064380)
- 61↑
Nokoff NJ, Scarbro SL, Moreau KL, et al. Body composition and markers of cardiometabolic health in transgender youth compared with cisgender youth. J Clin Endocrinol Metab 2020 105 e704–e714. (https://doi.org/10.1210/clinem/dgz029)
- 62↑
Olson-Kennedy J, Okonta V, Clark LF, et al. Physiologic response to gender-affirming hormones among transgender youth. J Adolesc Health 2018 62 397–401. (https://doi.org/10.1016/j.jadohealth.2017.08.005)
- 63↑
Pallotti F, Senofonte G, Conflitti AC, et al. Safety of gender affirming treatment in assigned female at birth transgender people and association of androgen and estrogen β receptor polymorphisms with clinical outcomes. Endocrine 2023 81 621–630. (https://doi.org/10.1007/s12020-023-03421-8)
- 64↑
Pei Q, Song Y, Huang Z, et al. Effects of gender-affirming hormone therapy on body fat: a retrospective case-control study in Chinese transwomen. Lipids Health Dis 2024 23 146. (https://doi.org/10.1186/s12944-024-02131-y)
- 65↑
Roberts TK, Kraft CS, French D, et al. Interpreting laboratory results in transgender patients on hormone therapy. Am J Med 2014 127 159–162. (https://doi.org/10.1016/j.amjmed.2013.10.009)
- 66↑
Sanchez-Toscano E, Dominguez-Riscart J, Larran-Escandon L, et al. Cardiovascular risk factors in transgender people after gender-affirming hormone therapy. J Clin Med 2023 12 6141. (https://doi.org/10.3390/jcm12196141)
- 67↑
Schutte MH, Kleemann R, Nota NM, et al. The effect of transdermal gender-affirming hormone therapy on markers of inflammation and hemostasis. PLoS One 2022 17 e0261312. (https://doi.org/10.1371/journal.pone.0261312)
- 68↑
Shadid S, Abosi-Appeadu K, De Maertelaere AS, et al. Effects of gender-affirming hormone therapy on insulin sensitivity and incretin responses in transgender people. Diabetes Care 2020 43 411–417. (https://doi.org/10.2337/dc19-1061)
- 69↑
Stoffers IE, de Vries MC & Hannema SE. Physical changes, laboratory parameters, and bone mineral density during testosterone treatment in adolescents with gender dysphoria. J Sex Med 2019 16 1459–1468. (https://doi.org/10.1016/j.jsxm.2019.06.014)
- 70↑
Suppakitjanusant P, Ji Y, Stevenson MO, et al. Effects of gender affirming hormone therapy on body mass index in transgender individuals: a longitudinal cohort study. J Clin Transl Endocrinol 2020 21 100230. (https://doi.org/10.1016/j.jcte.2020.100230)
- 71↑
Tebbens M, Schutte M, Troelstra MA, et al. Sex steroids regulate liver fat content and body fat distribution in both men and women: a study in transgender persons. J Clin Endocrinol Metab 2023 109 e280–e290. (https://doi.org/10.1210/clinem/dgad409)
- 72↑
Tsatsanis C, Elenkov A, Leijonhufvud I, et al. Sex hormone-dependent and -independent regulation of serum BAFF and TNF in cohorts of transgender and cisgender men and women. Endocr Connect 2023 12 e220456. (https://doi.org/10.1530/ec-22-0456)
- 73↑
Valentine A, Nokoff N, Bonny A, et al. Cardiometabolic parameters among transgender adolescent males on testosterone therapy and body mass index-matched cisgender females. Transgender Health 2021 6 369–373. (https://doi.org/10.1089/trgh.2020.0052)
- 74↑
van Velzen DM, Adorni MP, Zimetti F, et al. The effect of transgender hormonal treatment on high density lipoprotein cholesterol efflux capacity. Atherosclerosis 2021 323 44–53. (https://doi.org/10.1016/j.atherosclerosis.2021.03.008)
- 75↑
Victor VM, Rocha M, Banuls C, et al. Mitochondrial impairment and oxidative stress in leukocytes after testosterone administration to female-to-male transsexuals. J Sex Med 2014 11 454–461. (https://doi.org/10.1111/jsm.12376)
- 76↑
Vita R, Settineri S, Liotta M, et al. Changes in hormonal and metabolic parameters in transgender subjects on cross-sex hormone therapy: a cohort study. Maturitas 2018 107 92–96. (https://doi.org/10.1016/j.maturitas.2017.10.012)
- 77↑
Cocchetti C, Romani A, Collet S, et al. The ENIGI (European network for the investigation of gender incongruence) study: overview of acquired endocrine knowledge and future perspectives. J Clin Med 2022 11 1784. (https://doi.org/10.3390/jcm11071784)
- 78↑
Gulanski BI, Flannery CA, Peter PR, et al. Compromised endothelial function in transgender men taking testosterone. Clin Endocrinol 2020 92 138–144. (https://doi.org/10.1111/cen.14132)
- 79↑
Madjid M & Fatemi O. Components of the complete blood count as risk predictors for coronary heart disease: in-depth review and update. Tex Heart Inst J 2013 40 17–29
- 80↑
Spanos C, Bretherton I, Zajac JD, et al. Effects of gender-affirming hormone therapy on insulin resistance and body composition in transgender individuals: a systematic review. World J Diabetes 2020 11 66–77. (https://doi.org/10.4239/wjd.v11.i3.66)
- 81↑
Defreyne J, Van de Bruaene LDL, Rietzschel E, et al. Effects of gender-affirming hormones on lipid, metabolic, and cardiac surrogate blood markers in transgender persons. Clin Chem 2019 65 119–134. (https://doi.org/10.1373/clinchem.2018.288241)
- 82↑
Connelly PJ, Clark A, Touyz RM, et al. Transgender adults, gender-affirming hormone therapy and blood pressure: a systematic review. J Hypertens 2021 39 223–230. (https://doi.org/10.1097/hjh.0000000000002632)
- 83↑
Velho I, Fighera TM, Ziegelmann PK, et al. Effects of testosterone therapy on BMI, blood pressure, and laboratory profile of transgender men: a systematic review. Andrology 2017 5 881–888. (https://doi.org/10.1111/andr.12382)
- 84↑
Rytz CL, Beach LB, Saad N, et al. Improving the inclusion of transgender and nonbinary individuals in the planning, completion, and mobilization of cardiovascular research. Am J Physiol Heart Circ Physiol 2023 324 H366–H372. (https://doi.org/10.1152/ajpheart.00494.2022)