Cardiometabolic diseases in transgender populations

Abstract

The number of people seeking gender-affirming medical care and gender-affirming hormone therapy (GAHT) is increasing. As sex steroids shape cardiometabolic physiology, GAHT is expected to meaningfully alter cardiometabolic risk. The aim of this review was to systematically determine and narratively synthesize evidence on the cardiometabolic effects of GAHT and identify clinically relevant knowledge gaps. We searched PubMed for English-language primary human studies published through November 26, 2025. Of 346 screened records, 96 unique original investigations were included. Most studies had less than 3 years of follow-up, the longest being 20 years in one electronic health record-based study. In transgender women, BMI was often reported as stable with higher fat and lower lean mass compared to baseline; insulin resistance tended to increase while fasting glucose and HbA1c were usually longitudinally stable. Lipid changes often showed reductions in total and low-density lipoprotein cholesterol as well as triglyceride levels. Blood pressure generally remained stable or decreased. In transgender women, VTE incidence was higher than in both cisgender women and men, whereas stroke and MI incidence was higher primarily in comparison with cisgender women. In transgender men, BMI was increased or stable and lean mass increased with variable fat mass changes and some long-term increases in visceral and liver fat. Glucose usually remained stable while insulin resistance findings were mixed. Lipid profiles commonly shifted toward lower high-density lipoprotein cholesterol and higher low-density lipoprotein cholesterol and triglycerides. Blood pressure findings were mixed with some cohorts showing an increase. In transgender men, MI incidence was higher than in cisgender women but generally not higher than in cisgender men; evidence for other cardiovascular outcomes was limited and inconsistent. Overall, evidence on systemic inflammation, renal and hepatic outcomes is sparse. In conclusion, GAHT is associated with quantifiable cardiometabolic changes, but magnitude and clinical relevance vary by regimen, population, baseline risk and study follow-up duration. Heterogeneous regimens, study methods, and confounding contribute to mixed findings. Major gaps persist with limited long-term prospective data, sparse mechanistic insights, underrepresentation of adolescents and non-European ancestry populations, and insufficient information on cardiometabolic outcomes with ageing.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s40842-026-00336-x.

Highlights

Sex- and gender identity-stratified evidence synthesis: Distinguishes feminizing and masculinizing GAHT while presenting parallel outcome analyses for transgender women and transgender men, enabling more clinically precise interpretation of cardiometabolic effects.

Systematic characterization of heterogeneity: Integrates regimen-, population-, and context-specific findings, including discordant results, into structured domain-specific evidence tables, improving transparency and interpretability across a heterogeneous literature base.

Novel evidence mapping framework: Introduces a domain-level framework quantifying both evidence volume (number of primary studies) and directional concordance of findings, visualized within a sex-stratified cardiometabolic risk continuum to identify areas of consistency, uncertainty, and evidence gaps.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s40842-026-00336-x.

Introduction

Globally, the number of individuals identifying as transgender and seeking gender-affirming care has increased in recent years. While estimates vary widely by methodology, recent multi-country survey data suggest that ~ 3% of adults report a transgender or gender-diverse identity (including transgender, non-binary/gender-fluid, or another non-binary identity), with substantial variation by country and age/generation (www.ipsos.com). Between 1972 and 2015, according to “The Amsterdam Cohort of Gender Dysphoria Study” the number of people presenting at a gender identity clinic has increased substantially over time [1], a finding mirrored by other studies [2, 3]. These observations likely reflect improved awareness and greater access to care rather than a sudden change in underlying prevalence, which may still be underestimated in research due to stigma and structural barriers [4]. Transgender individuals therefore represent a distinct patient population that is often discriminated against and, additionally, experiences health disparities. This highlights a societal and medical responsibility to ensure the availability of evidence-based care.

Gender affirming hormone therapy (GAHT), the cornerstone of gender-affirming treatment, is routinely applied with the purpose of aligning physical appearance and gender identity [5]. For individuals assigned male sex at birth GAHT consists of feminizing hormone therapy, which most often comprises a combination of estradiol and antiandrogens [5, 6]. In transgender men the treatment mainly consists of testosterone administered either via injections or transdermal preparations [5, 6].

Cardiometabolic diseases, amongst which obesity, type 2 diabetes mellitus, hypertension, atherosclerotic cardiovascular disease, chronic kidney disease, and steatotic liver disease constitute a major and growing global health and socio-economic burden with increasing prevalence and incidence which is largely tracking the substantial worldwide increase in obesity [7]. These conditions frequently present as clusters within individuals, resulting in multimorbidity and complex risk profiles that are difficult to manage from a clinical perspective and substantially increase morbidity and mortality [7]. Sexual dimorphism in cardiovascular and metabolic physiology is well established and described extensively in the literature [8, 9]. This dimorphism is partly attributed to sex hormone differences between cisgender males and females [9, 10], as well as to chromosomal, epigenetic, and body composition-related factors [11, 12]. This is backed by the mechanistic insight that sex steroid hormones, amongst which testosterone and estrogen, act via nuclear receptors, modulating pathways that are impactful on lipid metabolism, glucose homeostasis, vascular function, coagulation, and inflammatory signaling, thereby influencing cardiovascular and metabolic homeostasis [13]. This complexity has practical implications for clinical risk prediction, as widely used CVD risk estimators operationalize sex as a binary input (male/female) and have not been formally validated in transgender populations. In baseline data from the LITE Plus cohort of Black and Latina transgender women with HIV, estimated 10-year ASCVD risk and resulting statin eligibility differed depending on whether the Pooled Cohort Equation was calculated using sex assigned at birth, current gender identity, or current hormone use, highlighting potential misclassification in routine care [14].

Given the link between sex hormones and sexual dimorphism in cardiometabolic physiology, an impact of GAHT on cardiometabolic outcomes is conceivable. However, despite the increase in demand for GAHT [15], its effect on cardiometabolic outcomes and the actual shifts in risk profiles are not fully understood. This is due to an overall limited number of studies, insufficiently powered cohorts with limited participant numbers, short follow-up periods, and GAHT regimen heterogeneity [10]. The limited literature is reflected by limited clinical guidelines for cardiometabolic monitoring of individuals undergoing GAHT. A good example of that is the literature regarding chronic kidney disease (CKD) in transgender cohorts, where, importantly, no consensus has been reached regarding the most appropriate method for estimating glomerular filtration rate (eGFR) in transgender individuals. This issue is clinically relevant considering that CKD is a well-established risk factor for cardiovascular morbidity and mortality, with even mild reductions in eGFR associated with graded increases in cardiovascular events [16]. Calculations based on sex assigned at birth versus affirmed gender can yield substantially different results, leading to potentially significant misclassification of kidney function [17]. The described misclassification may therefore not only affect CKD risk assessment but also CVD risk estimates, therapeutic decision making, and eligibility for treatment or preventive interventions [16]. This complicates the assessment of chronic kidney disease and poses challenges for accurately monitoring renal outcomes in the context of GAHT.

To address these knowledge gaps, we performed a systematic literature search with narrative synthesis to summarize and critically appraise the current evidence on the impact of GAHT on cardiometabolic outcomes in transgender individuals. This review will analyze the impact of feminizing and masculinizing hormone therapy, administered via different regimens, on specific cardiometabolic outcomes. These include anthropometry, glucose and insulin metabolism, lipid profiles, blood pressure and vascular outcomes, inflammation and coagulation markers, renal and hepatic outcomes, and clinical cardiovascular disease events.

Methods

This narrative review was informed by a systematic PubMed literature search. Given substantial heterogeneity in study design, GAHT regimens, comparator populations, follow-up duration, and outcome definitions, findings were synthesized narratively rather than pooled quantitatively. Studies published up to Nov 26th, 2025, were considered for inclusion. The search strategy combined MeSH terms and keywords related to transgender populations, hormone therapy, and cardiometabolic outcomes as detailed in the online supplement. According to the scope of this review, only human studies were included. Eligible outcomes included anthropometry and body composition, glucose and insulin metabolism, lipid profiles, blood pressure and vascular measures, inflammation and hemostasis, renal and hepatic outcomes, and clinical cardiovascular events. The population of interest included both transgender adults and adolescents undergoing feminizing or masculinizing hormone therapy encompassing regimens based on estradiol, testosterone and combinations with antiandrogens. Eligible study designs included randomized trials, prospective and retrospective cohort studies, and cross-sectional studies.

Ex vivo and animal studies were excluded. Studies focused on reproductive, skeletal, or mental health outcomes were not included unless they also reported eligible cardiometabolic outcomes. Secondary sources, case reports, publications in languages other than English, and publications for which full-text access could not be obtained were excluded from this review. Full texts of potentially eligible articles were reviewed for inclusion. Studies yielding mechanistic insights such as vascular imaging and metabolomics were included as secondary outcomes to support and strengthen clinical interpretations. The initial search strategy yielded 346 records out of which 96 were included in this review. To improve interpretability, findings were synthesized separately for feminizing and masculinizing GAHT and reported in parallel for transgender women and transgender men. Study-level regimen details and comparator populations were summarized in domain-specific tables. In addition, we used an evidence-mapping framework based on study quantity and directional agreement across cohorts, as described in the Supplemental Methods and Supplementary Table 1. This GAHT-focused synthesis is therefore intended to complement broader cardiometabolic reviews across sexual and gender minority populations [18] and guideline-style statements on cardiovascular health in transgender and gender-diverse people [19], and should be read synergistically with these works. For each of the included studies, a systematic extraction of outcomes was carried out summarizing study design and setting, information regarding the study population, details regarding the type of GAHT, metabolic and anthropometric parameters, organ function and parameters, and thrombotic markers and clinical events.

Results

Overview of included studies

This review included 96 studies, predominantly retrospective, prospective, or cross-sectional cohort analyses. Most cohorts comprised young adults, with mean or median ages generally between 23 and 35 years, and most follow-up periods were shorter than three years; one electronic health record-based study included up to 20 years of retrospective observation [20]. It is worth mentioning that studies with extensive retrospective follow-up periods may include participants exposed to earlier estrogen formulations and dosages such as ethinyl estradiol, which when used as a contraceptive was associated with significantly higher risk of myocardial infarction and thrombotic stroke [21] therefore posing potential interpretation bias in long-term cohorts. Multiple studies used data extracted from electronic health records.

Feminizing GAHT in the included studies aimed at maintaining estradiol within typical female reference ranges while suppressing testosterone levels [22]. This treatment typically comprised estradiol, administered orally, transdermally, or intramuscularly, and combined with anti-androgens primarily CPA or spironolactone. Some studies additionally reported occasional use of Gonadotropin Releasing Hormone (GnRH) agonists or progesterone.

Masculinizing GAHT aimed at suppressing female secondary sex characteristics by most commonly administering testosterone esters or undecanoate via intramuscular or subcutaneous injection preparations or alternatively by using transdermal testosterone gel preparations [23].

Some studies also explored the effects of puberty suppression by means of GnRH agonists on cardiometabolic outcomes [24, 25]. Regardless of the therapy, the dosages were commonly adjusted individually to achieve target hormone levels.

Anthropometry

Anthropometric outcomes reflect changes in classical measures of obesity such as body weight, body mass index (BMI), and waist circumference as well as body composition, fat distribution, and lean body mass. In this subsection, reported changes in the body mass index (BMI), total and regional fat mass, lean and muscle mass, and body composition in transgender people are summarized and contextualized by the type of GAHT administered. More detailed study-level information is summarized in Tables 1A and B.

Table 1.

A Anthropometric changes in transgender women receiving feminizing gender affirming hormone therapy: direction of effect, regimen modifiers, and key supporting studies. B Anthropometric changes in transgender men receiving masculinizing GAHT: direction of effect, regimen modifiers, and key supporting studies

Anthropometric Parameter	Direction of Change	Regimen / Clinical Context	Supporting References
A
BMI / Body weight	No significant change	Estradiol (oral, transdermal, injected) + anti-androgens (spironolactone or CPA)	[9, 26–34]
BMI / Body weight	No significant change	Comparison between orchiectomy vs. testes retention	[26]
BMI / Body weight	No significant change	CPA or leuprolide acetate in combination with estradiol	[93]
BMI / Body weight	Increase	Initiation of GAHT; Age 17–29 after 21 months of treatment	[94]
BMI / Body weight	Increase	Estradiol along with spironolactone or CPA	[35, 36]
Body weight	No significant change	First 21 months of treatment (Age 17–29); all durations for Age > 30	[94]
Waist circumference	Decrease	Feminizing GAHT (24 months)	[27]
Total fat mass	Increase	Consistent gains reported (range + 4.0 kg to + 9.8 kg)	[26, 37, 38]
Total fat mass	Lower	Compared to cisgender women	[41]
Lean mass	Decrease	Consistent losses reported (range − 1.7 kg to -6.9 kg)	[26, 37]
Muscle fat infiltration	Increase	4% increase at 5-year follow-up	[38]
Visceral Adipose Tissue (VAT)	Mixed / Decrease	Reported decrease in volume or VAT/total fat ratio	[37, 38]
Subcutaneous/Visceral Fat Ratio	Increase (More favorable)	Significant difference compared to cisgender males	[39]
Appendicular lean mass	Lower	Compared to cisgender men	[39]
Body fat %	Lower fat %	Lower total and subcutaneous fat % in orchiectomy compared to GAHT with testes retention	[40]
B
BMI / Body weight	Increase	Testosterone (injectable esters or transdermal gels)	[28, 29, 31, 33, 36, 41, 42]
BMI / Body weight	No significant change	Thai cohort (testosterone enanthate); multiple other cohorts	[27, 30, 34, 38, 43–45]
Waist circumference	No significant change	Masculinizing GAHT	[27, 43]
Lean mass	Increase	Significant gains (range + 4.7 kg to + 7.8 kg); total lean tissue volume	[26, 37, 38]
Muscle mass	Increase	Thigh fat-free muscle volume at 6 years	[38]
Total body fat	Decrease / Stable	Variable results; some reports of stable total adipose volumes	[37, 38, 43]
Visceral fat (VAT)	Increase	Higher than cisgender men reference values; +14% VAT/total fat ratio	[37, 38]
Liver fat	Increase	Reported increase at 6-year follow-up (NMR)	[38]

Findings are presented within specific clinical contexts, such as age at initiation, duration of treatment, and hormone regimen (e.g., use of cyproterone acetate [CPA] or spironolactone). Additionally, the tables highlight shifts in fat distribution, such as visceral adipose tissue (VAT) and subcutaneous-to-visceral fat ratios, relative to cisgender comparators or surgical status (e.g., orchiectomy). Data are synthesized from primary studies to characterize the systematic physical remodeling associated with GAHT

Transgender women

With respect to body weight and BMI, data show heterogeneous results tending towards stable maintenance. Multiple studies detected no longitudinal statistically significant changes in BMI or body weight during feminizing GAHT with estradiol administered orally, transdermal or injected combined with anti-androgens such as spironolactone or CPA [9, 26–34]. Meanwhile, in one cohort, significant decreases in waist circumference were observed [27]. One single-center longitudinal retrospective cohort study with a follow-up of 7 years reported that transgender women entering the study already on GAHT experienced no significant change in BMI while individuals who initiated GAHT after inclusion into the cohort experienced a significant increase in BMI [35]. One study where patients were treated with estradiol along with spironolactone or CPA noted statistically significant increases in weight and BMI [36]. On the other hand, the results on the impact on body composition appear to be more homogenous. Significant increases in total fat mass were consistently reported, with gains of + 9.8 kg [26] and + 4.0 kg [37]. On the other hand, significant decreases in lean mass were observed, ranging from − 1.7 kg [37] to -6.9 kg [26]. One study reported an increase in muscle fat infiltration at 5 years by 4% [38].

With respect to specific fat depots, findings regarding visceral adipose tissue (VAT) were mixed with one study reporting a decrease in the VAT/total body fat ratio [37]. Another study indicated a statistically significant difference in transgender women as compared to cisgender males with respect to the subcutaneous/visceral fat ratio which was more favorable in the transgender women. Further, transgender women had lower appendicular lean mass than cisgender men and lower total fat mass than cisgender women [39]. Another study reported that transgender women developed increased total adiposity but decreased visceral adipose tissue volumes upon GAHT [38]. Interestingly, participants who underwent orchiectomy had significantly lower total fat and subcutaneous fat percentages compared to those with testes retention [40].

Transgender men

Within transgender men cohorts, results were more homogenous with respect to BMI compared to the transgender women cohorts. Multiple studies using testosterone, either as injectable esters or as transdermal gels, reported significant increases in body weight and BMI [28, 29, 31, 36, 41, 42]. One study noted a 7.4% increase in BMI at 6 months followed by a plateau after one year [33]. When determined, waist circumference generally showed no significant change [27, 43]. Conversely, another study conducted on a Thai cohort treated with testosterone enanthate 250 mg every 4 weeks described no significant change in body weight or BMI with a mean follow-up of 25.2 months [44]. Other work also observed no significant changes in BMI [27, 30, 34, 38, 43, 45].

With respect to lean and fat mass, results varied. Significant increases in lean mass, presumably attributable to increases in skeletal muscle mass, were reported in multiple studies, with gains of + 7.8 kg [26], + 4.7 kg [37], and an increase in total lean tissue volume at 6 years measured by NMR, a method with a particularly small coefficient of variation of 0.8% [38]. Further, an increase of liver and visceral fat accompanied by stable total adipose tissue volumes was observed in transgender men [38]. Various studies reported a decrease in total body fat in transgender men [37, 43]. Lundberg et al. also highlight a significant increase in visceral adipose tissue (VAT) in transgender men at 6 years follow-up. In this paper, it is suggested that the VAT observed in transgender men is higher than the reference value for healthy cisgender men [38]. Another study observed a 14% increase in VAT/total body fat ratio in transgender men [37].

In summary, while GAHT effectively facilitates a phenotypic shift in body composition toward the affirmed gender, the frequently observed stability of BMI suggests that traditional weight-based metrics mask significant underlying metabolic remodeling and visceral fat accumulation and highlights the importance of NMR-based imaging studies.

Glucose metabolism and insulin resistance

Although sex steroid hormones influence glucose homeostasis by having (i) direct effects on key target tissues such as skeletal muscle, liver, and adipose, and (ii) indirect effects on body composition in its entirety, the effect of GAHT on glucose metabolism and insulin resistance remains unclear [46]. The following section summarizes evidence from the current literature examining changes in fasting glucose, insulin, homeostatic model assessment of insulin resistance (HOMA-IR), and the risk of developing insulin resistance or diabetes during feminizing and masculinizing GAHT. Despite HOMA-derived indices not having been formally validated in transgender populations, the calculation on which they are based is derived from fasting glucose and insulin concentrations, not incorporating sex-specific coefficients and thus making the assessment presumably reliable also in individuals subject to GAHT [47].

Transgender women

Changes in fasting glucose were described as non-significant in multiple studies [29, 30, 32, 34] with one study also finding no significant change in the triglyceride-glucose index, a measure of insulin resistance alternative to HOMA-IR [38]. Similarly, changes in glycated hemoglobin (HbA1c) were also mostly described as non-significant [32, 36] with one study observing a small age-adjusted increase of HbA1c upon GAHT initiation that was deemed not clinically relevant when compared to corresponding plots for cisgender referents [48]. Additionally, one study found that the Diabetes Risk Index, an NMR-based score calculated from a number of relevant metabolic parameters that change when insulin resistance develops, decreased significantly in transgender women [30].

Concerning insulin resistance and sensitivity, most studies suggest a trend towards a worsening impact of GAHT in transgender women regardless of the used medication. One study observed an average 47% increase in insulin resistance at one year of GAHT, estimated by HOMA-IR, in 179 participants with oral estradiol valerate or transdermal patches combined with CPA [37]. This study further highlighted that this change occurred independent of relative VAT mass [37]. This result is supported by another study finding HOMA2-IR to be 1.5 times and significantly higher than in age-matched cisgender males [26]. This study also underlined that the HOMA2-IR was significantly higher in transgender women as compared to cisgender women [26]. Similarly, a regimen comprising estradiol valerate and spironolactone was associated with worsening insulin sensitivity measured by HOMA-IR [49]. This result was confirmed in another study to also be independent of HIV serostatus in a cohort comparing 32 HIV-positive with 138 HIV-negative participants [50]. Another study conducted on transgender women treated with estrogen and CPA found a decrease in insulin sensitivity measured by HOMA2-%S and a significant increase in β-cell function, determined by means of homeostasis model assessment for β-cell function (HOMA2-%β), determined at 6 months after GAHT initiation [32]. Other studies reported no significant change in insulin sensitivity (HOMA-IR) [31, 38, 51, 52] though one of these expressed that power analyses revealed the necessity of larger sample sizes in order to reliably detect changes in these parameters [51]. One work that specifically investigated the impact of orchiectomy in comparison to GAHT alone observed that orchiectomy was associated with lower HOMA-IR and fasting insulin levels than GAHT alone therefore describing a potential protective metabolic effect of orchiectomy in transgender women that should be mechanistically further explored [40].

Transgender men

Concerning fasting glucose, the results are rather heterogenous. Some studies reported no significant changes [31, 48, 53] while other studies observed a decrease [34, 38, 54]. One study also found that fasting glucose levels in a transgender men cohort were consistently lower than the levels observed in cisgender men controls [48]. One study observed a statistically significant increase of glucose levels at 12 months of GAHT and an increase of the NMR metabolomics-based diabetes risk index compared to baseline [30].

With respect to insulin resistance some studies described that the comparison between transgender men and cisgender men yielded no significant difference [38, 49, 53], with one study reporting a stable HOMA-IR over a period of 5 years in groups (mean age ~ 30) treated with testosterone undecanoate or enanthate [53]. A comparison with cisgender women yielded also non-significance [26]. However, there are also studies describing a significant decrease in HOMA-IR in transgender men at 1 year after initiating testosterone treatment [31, 37].

With respect to HbA1C levels, effects of GAHT also seem to differ amongst studies. Some observed no significant changes [31, 48] while one study conducted in a cohort with a median treatment duration of 5 years observed a statistically significantly higher HbA1C as compared to cisgender women controls [41]. A slight but significant increase from 32 to 33 (mmol/mol)² (p = 0.04) was also observed at one year of GAHT in one study conducted on a cohort of 136 young transgender men treated with intramuscular testosterone [36]. However, despite this change, no significant increase in prevalent or incident diabetes or prediabetes was reported in the same study [36].

Combined, available evidence indicates that while GAHT particularly in feminizing regimens is associated with a trend toward increased insulin resistance, overall fasting glucose and HbA1c levels remain largely stable in predominantly young, metabolically healthy cohorts, suggesting a degree of glycemic compensation. With longer follow-ups and the impact of ageing, however, diabetic phenotypes could precipitate.

Lipid profile

The effects of GAHT on lipid profiles have been rather extensively studied, as clinical chemistry lipid panels are widely available and alterations in lipid metabolism represent a more direct and established pathway to link sex hormones to cardiometabolic risk [55]. This section summarizes observed changes in routinely determined plasma lipids reporting total cholesterol, low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), triglycerides, and, where available, apolipoproteins across the included studies. More detailed study-level information and regimen-specific modifiers are summarized in Tables 2A and B.

Table 2.

A Lipid profile changes in transgender women receiving feminizing GAHT: direction of effect, regimen modifiers, and key supporting studies. B Lipid profile changes in transgender men receiving masculinizing GAHT: direction of effect, regimen modifiers, and key supporting studies

Lipid Parameter	Direction of Change	Regimen / Clinical Context	Supporting  References
A
Total Cholesterol	Decrease	Estradiol sometimes combined with CPA or spironolactone	[9, 17, 27, 32, 37, 51, 56]
Total Cholesterol	Non-significant decrease	Estradiol combined with CPA	[36]
Total Cholesterol	Increase	Observed in study comparing leuprolide acetate and CPA (greater increase than in CPA group)	[93]
Total Cholesterol	Negative correlation	Higher estradiol doses associated with lower levels	[9]
LDL-C	Decrease	Estradiol sometimes combined with CPA or spironolactone	[9, 17, 27, 32, 37, 51, 56]
LDL-C	Greater decrease	CPA group compared to spironolactone group	[9]
Triglycerides	Decrease	Most longitudinal follow-up studies	[17, 27, 37, 52, 56]
Triglycerides	Non-significant increase	Specific cohort observation	[28]
HDL-C	Increase	Observed specifically in users of spironolactone	[9, 28]
HDL-C	No significant change	Treated with CPA and estrogen combinations	[27, 32, 52]
HDL-C	Decrease	Reported in cohorts treated with CPA	[17, 37, 56, 93]
HDL-C	Negative correlation	Higher estradiol doses associated with lower levels	[9]
HDL Function	Decrease	Reduction in cholesterol efflux capacity (pro-atherogenic)	[30, 56]
ApoB: ApoA1 Ratio	Decrease	Observed in CPA and spironolactone groups (anti-atherogenic)	[9]
Particle Size	No change	Estrogen did not influence LDL or HDL particle size	[30]
Dyslipidemia	Higher prevalence	Transdermal users (47%) vs. oral users (3%)	[57]
Dyslipidemia	Significant increase	Feminizing GAHT per ESC/EAS Guidelines (p = 0.035)	[36]
B
Total Cholesterol	Increase	Frequent shift toward pro-atherogenic profile	[27, 31, 38, 41, 56]
Total Cholesterol	No significant change	Reported in some cohorts	[17, 32, 37]
LDL-C	Increase	Development of atherosclerosis-promoting profile	[17, 27], 36– [38, 41, 56]
LDL-C	No significant change	Reported in some cohorts	[17, 32, 37]
Triglycerides	Increase	Homogenous results across most studies	[17, 27], 36– [38, 41, 56]
Triglycerides	No significant change	Reported in 6-month follow-up	[28, 32]
HDL-C	Decrease	Consistent longitudinal reduction	[17, 27, 28, 31], 36– [38, 41, 56]
HDL-C	No significant change	Single study observation	[32, 54]
HDL-C	Comparable / Lower	Comparable to cis-males; lower than cis-females	[95]
HDL Function	No change	Stable cholesterol efflux capacity up to 11 months	[30, 59]
HDL Subtypes	Decrease	Lower concentrations of large HDL particles	[30, 58]
LDL Subtypes	Increase	Higher concentrations of small LDL	[30, 58]
Dyslipidemia	Significant increase	Monthly testosterone cypionate (p = 0.001)	[36]
Regimen Effect	No difference	Similar effects between transdermal, IM esters, and IM undecanoate, testosterone undecanoate and enanthate	[53, 56]

The synthesized data highlight directional changes in total cholesterol, low-density lipoprotein cholesterol (LDL-C), triglycerides, and high-density lipoprotein cholesterol (HDL-C) across diverse cohorts. Findings are contextualized by specific treatment regimens, including the influence of hormone dosage, route of administration (e.g., oral, transdermal, or injectable), and the use of adjunctive therapies such as anti-androgens (cyproterone acetate or spironolactone). Additionally, the tables capture shifts in advanced cardiovascular risk markers, including HDL function, particle size, and apolipoprotein ratios, and the comparative prevalence of clinical dyslipidemia as defined by international guidelines

Transgender women

The impact of feminizing GAHT on lipid metabolism appears heterogeneous with most studies showing favorable reductions in atherogenic lipids.

Several studies observed significant decreases in total cholesterol and LDL-C in groups using estradiol in combination with CPA or spironolactone [9, 17, 27, 32, 37, 51, 56]. With respect to triglycerides most studies reported a decrease in longitudinal follow-up [17, 27, 37, 52, 56]. The changes in HDL-C seemed to be highly variable and dependent on the GAHT regimen. While increases were observed in users of spironolactone [9, 28], no significant change was observed in transgender women treated with CPA and estrogen combinations [27, 32, 52]. Decreases were also reported in some cohorts treated with CPA [17, 37, 56]. Not only did the anti-androgenic component of feminizing GAHT modulate lipid profiles but also the estrogen regimen. One study found a significantly higher prevalence of dyslipidemia in transdermal estradiol users (47%) compared to oral users (3%) [57]. In another study higher estradiol doses were negatively correlated with total cholesterol and HDL-C [34]. Only limited data exist with respect to lipid and lipoprotein analyses using nuclear magnetic resonance (NMR) spectroscopy. One study reported a decrease in the ApoB: ApoA1 ratio, interpreted as anti-atherogenic change, in CPA and spironolactone using groups [9]. With respect to the emerging concept of HDL function, two studies observed that in transgender women the HDL cholesterol efflux capacity decreased, i.e. the properties of HDL to mobilize cholesterol from macrophages indicating a pro-atherogenic change in HDL function in these individuals [30, 56]. Lei et al. also describe that estrogen did not seem to influence LDL and HDL particle size in transgender women [30].

Transgender men

Testosterone therapy is frequently associated with a shift toward a more atherogenic lipid profile, characterized primarily by decreased HDL-C and increased LDL-C as well as triglyceride levels, the latter mainly representing a marker of the pro-atherogenic remnant lipoproteins in the fasting state [27]. Increases in LDL-C and/or total cholesterol were frequently seen pointing toward the development of an atherosclerosis-promoting lipid profile [27, 31, 37, 38, 41, 56]. On the other hand some studies reported no significant change in LDL-C or total cholesterol [17, 32, 37]. With respect to triglycerides, results were rather homogenous with increases observed in most studies [17, 27, 36–38, 41, 56] while only two studies with a 6 months follow-up reported no significant change [28, 32]. Notably, no study saw a decrease in triglycerides. In line with the well appreciated inverse relationship between triglycerides and HDL-C, a consistent decrease in HDL-C was observed in multiple studies [17, 27, 28, 31, 36–38, 41, 56] with only one reporting no significant change [32]. GAHT regimen-specific effects in transgender men were also explored in the literature with one study stating that no differences were observed in lipid related alterations between different testosterone formulations comprising transdermal testosterone, intramuscular testosterone esters and intramuscular testosterone undecanoate [56]. A similar conclusion was reached in another study comparing testosterone undecanoate and enanthate [53]. NMR spectroscopy-based analyses revealed additional pro-atherogenic shifts in the lipoprotein profile. Two studies observed lower concentrations of large HDL particles in transgender men [30] compared to cisgender females [58]. The same studies also found transgender male adolescents to have higher concentrations of small LDL in longitudinal analyses [30] compared to cisgender females [58]. Regarding the HDL cholesterol efflux function no change was discernible in transgender men up to 11 months after initiating GAHT as compared to baseline [30]. Using a different protocol to determine HDL-mediated cholesterol efflux, another group confirmed these results [59].

Taken together, feminizing GAHT generally induces a favorable reduction in atherogenic lipids, whereas masculinizing therapy consistently results in a pro-atherogenic shift characterized by decreased HDL-C and increased LDL-C and triglycerides, mimicking a cisgender male cardiovascular risk profile.

Blood pressure and vascular function parameters

Next to circulating lipids, blood pressure and vascular function are key determinants of cardiovascular health [60]. In the following section findings about changes in factors such as systolic blood pressure and arterial stiffness are explored.

Transgender women

The literature on blood pressure changes in response to feminizing GAHT shows mixed results with some studies reporting no longitudinal significant changes while others observe decreases in blood pressure. However, it is worth noting that within the included studies, none reported an increase in blood pressure. The majority of available work described either no significant longitudinal or no clinically meaningful changes above the hypertensive threshold in blood pressure in transgender women using estradiol with anti-androgens such as CPA or spironolactone [29–31, 33, 56]. Further, a study looking at systolic and diastolic blood pressure as well as mean arterial pressure in a cohort of HIV-positive transgender women reported no significant differences upon follow-up conducted at variable time points with a minimum of 6 months exposure to GAHT [51]. Other work observed a decrease in systolic blood pressure in transgender women treated by means of oral or transdermal formulations for 24 months [27]. Similarly, another study reported a decrease in systolic blood pressure but no decrease in diastolic blood pressure [34]. A small decrease in diastolic and systolic blood pressure was also observed in a cohort treated with CPA and estrogen valerate or transdermal patches [56]. Another study in which 88% of the cohort was treated with sublingual micronized 17-beta estradiol 2 mg twice daily found a statistically significant decrease in both diastolic and systolic blood pressure [28].

One study conducted in HIV positive transgender women reported that estradiol long-term, as well as short-term use was negatively associated with carotid intima media thickness (CIMT) [61]. Conversely, another study found no significant difference in CIMT at 11 months of GAHT as compared with baseline values [30]. Additionally, one study exploring the impact on aortic pulse wave velocity reported no significant changes [38]. However, longer follow-up times are likely required to come to clinically meaningful conclusions.

An elegant, although not sufficiently powered, study that performed coronary computed tomography (CT) angiography on 29 transgender women found that this cohort had a lower prevalence of coronary plaques of any type as compared to cisgender males regardless of HIV serostatus [62]. This work further observed that estradiol but not testosterone levels were moderately and negatively correlated with the prevalence of coronary plaques of any type and with the prevalence of coronary stenosis [62].

Transgender men

In transgender men, the findings with respect to changes in blood pressure are more heterogeneous than in transgender women. No statistically significant blood pressure alterations were observed in several studies [27, 28, 31, 33, 43] with one also mentioning no increase in arterial stiffness [38]. In one study that compared testosterone undecanoate and testosterone enanthate with a follow-up period of 5 years no significant change in blood pressure was observed [53]. On the other hand, systolic blood pressure increased significantly in a number of cohorts using different testosterone formulations [29, 30, 36, 56, 63].

One study exploring the impact on aortic pulse wave velocity reported no significant changes in response to GAHT [38]. Another study assessing arterial stiffness by measuring carotid-femoral pulse wave velocity observed that long-term testosterone therapy was associated with significantly higher values compared to cisgender males and females, thereby showing that transgender men exposed to GAHT for a cohort average exposure time of 14 years had higher aging-related aortic stiffening compared to control groups [64]. As described by two studies, internal and common CIMT were significantly higher in transgender men longitudinally in a cohort with 11 months of GAHT exposure [30] and compared to controls in a cohort with a median GAHT exposure of 5 years [41]. Likely owing to the young age of transgender individuals included in most of these studies and the relatively short follow-up times, more research is required to assess the impact of GAHT on functional and structural aspects of the arterial branch of the cardiovascular system.

In conclusion, vascular findings remain heterogeneous especially in transgender women, although a worsening of vascular function appears unlikely in this group. Transgender men on the other hand may experience increases in systolic pressure and arterial stiffness, particularly with longer exposure durations.

Inflammation and thrombotic markers

Gender-affirming hormone therapy can influence the coagulation cascade as well as the basic inflammatory state, both key contributors to cardiometabolic risk [65]. However, most data come from unchallenged, young and otherwise healthy cohorts. Below we summarize how GAHT alters markers of inflammation and hemostasis, highlighting potential shifts toward pro- or anti-thrombotic states and their potential relevance for cardiovascular outcomes in transgender populations.

Transgender women

Reported effects of GAHT on chronic inflammation markers are overall limited. One study in HIV seropositive individuals observed higher interleukin-6 levels in transgender as compared to cisgender women [66]. Similarly, other work in transgender women observed significantly higher levels of interleukin-6 and endothelin compared to cisgender men and significantly lower levels of plasminogen activator inhibitor compared to cisgender men [62]. However, in the same cohort, other inflammatory markers, such as EN-RAGE, FABP4, sTNFRII, CD163, CD14, total adiponectin, VCAM-1, and IL-8 did not differ significantly to cisgender males [62]. Another study conducted on 48 transgender women using estradiol patches in combination with CPA found that the classical inflammation marker high sensitivity (hs) CRP decreased with increasing time of GAHT administration [67]. The same study observed that interleukin-6 and interleukin-8 decreased at 8 months while other inflammatory cytokines showed no significant changes. Further, a decrease in TNF-α was observed at 3 months, but an increase of the same marker was described at 12 months of GAHT [67]. The endothelial inflammatory marker soluble VCAM-1 decreased as well over the 12 month period [67], which was confirmed in another study conducted on 40 transgender women and in total 50 age- and BMI matched cisgender controls [68]. One study also observed significantly higher leukocyte numbers in transgender women when compared to cisgender females and males [68]. Increases were reported in leptin and adiponectin levels. These changes were considered to be consistent with the reference values of the affirmed gender [67], although, in other work such changes were described to be independent of body fat and therefore attributed to changes in the hormone milieu [69].

Literature on hemostasis and coagulation biomarkers is likewise limited. One study described an increase in platelet factor 4 (PF-4) and β-thromboglobulin at 12 months while P-selectin and PAI-1 showed no significant changes [67]. Further, fibrinogen transiently increased at 3 months of GAHT but returned to baseline at 12 months [67]. Another study found significantly higher plasminogen activator inhibitor 1 in transgender women compared to cisgender women with no significant difference in antithrombin, protein C, prothrombin activity, and thrombin time, and significantly lower free protein S in transgender women compared to cisgender men [68].

Transgender men

Similar to transgender women, the literature on biomarkers of chronic inflammation and coagulation in transgender men appears to be limited at present. One study described a significant increase in hs-CRP in transgender men after 12 months of GAHT [67]. This observation is supported by the results of another study conducted on 18 transgender men with a follow-up of 52 weeks. In this study, a relative increase of 9 out of 12 inflammatory markers was observed with highest increases in hs-CRP (+ 68% at 52 weeks), interleukin-5 (+ 37% at 52 weeks) and interferon-gamma (+ 109% at 52 weeks) [10]. Soluble VCAM-1 was shown to remain stable over 12 months of GAHT in transgender males [67]. Decreases were reported in leptin and adiponectin in transgender men which, just as mentioned above for transgender women, changed towards reference values of the affirmed gender [67].

With respect to pro-thrombotic and coagulation markers, a slight increase was observed in platelet activation markers and platelet leukocyte complexes at 52 weeks [10]. However, another study with a similar follow-up period observed no significant change in platelet factor 4, β-thromboglobulin, P-selectin, fibrinogen, and PAI-1 [67]. Another study conducted in transgender men treated with testosterone undecanoate and testosterone enanthate and with a follow-up period of 5 years reported no changes in the functional coagulation parameters aPTT and INR [53].

Renal outcomes

The relationship between decreased kidney function and an increased CVD risk is well-established [16, 70]. Chronic kidney disease (CKD), defined by reduced estimated glomerular filtration rate (eGFR) and/or increased albuminuria [71], is strongly associated with a higher prevalence and incidence of hypertension, diabetes, dyslipidemia, and cardiovascular risk [72]. As described above, GAHT seems to impact cardiometabolic factors on multiple levels and, for that reason, it is also interesting and important to consider changes in kidney function in transgender populations. However, the current evidence on the impact of GAHT on kidney function is very limited as no studies describe the long-term impact of GAHT on kidney function and CKD incidence or progression. Current literature is limited to short-term studies using creatinine levels and cystatin C [17, 31, 56, 73, 74] with some studies providing estimates on eGFR, but calculations are complicated by whether sex assigned at birth or affirmed gender is used for the formula, which is clinically relevant because commonly used eGFR equations include a sex coefficient, so the same biomarker value can yield substantially different eGFR estimates depending on the sex input. A creatinine-based eGFR is particularly vulnerable in this setting because creatinine generation is tightly linked to lean muscle mass, which shifts in opposite directions in transgender women and transgender men during GAHT [26, 37, 38]. Therefore, changes in creatinine and creatinine-based eGFR may reflect altered creatinine production rather than true changes in glomerular filtration [17, 73]. Although cystatin C is less dependent on muscle mass, it is not a pure filtration marker and can be influenced by non-GFR determinants (e.g., inflammation and adiposity), which may also change during GAHT [75]. Accordingly, short-term changes in eGFR around GAHT initiation should be interpreted cautiously, particularly when based on creatinine alone. An additional limitation is that available equations were not formally validated in transgender cohorts across GAHT regimens and durations, and there is no consensus on which sex input should be used for routine clinical reporting [17]. Although in clinical practice these considerations likely do not have a large impact on medical decision making, this is expected to change rapidly with increasing ageing of transgender populations. This is exemplified by a recent longitudinal cohort study, where cystatin C decreased in transgender women and increased in transgender men, corresponding to an approximate 7 ml/min/1.73 m² increase and 6 ml/min/1.73 m² decrease in eGFR, respectively, while creatinine changed in parallel with expected GAHT-induced shifts in muscle mass. Importantly, the authors emphasized that these alterations likely reflect body composition changes rather than true changes in glomerular filtration, highlighting the limitations of equation-based eGFR estimates in this population [17]. A potential impact of GAHT on urinary albumin excretion was not found in the current literature. Additionally, studies longitudinally comparing measured GFR by means of e.g., iohexol clearance are thus far lacking.

Hepatic outcomes

Metabolic dysfunction-associated steatotic liver disease (MASLD) and metabolic dysfunction-associated steatohepatitis (MASH) are closely related to obesity, insulin resistance, dyslipidemia, metabolic syndrome, and cardiovascular disease and therefore relevant to consider in the general context of this review [76]. MASLD is defined by hepatic fat accumulation and can be detected non-invasively using imaging modalities such as ultrasound or MRI, whereas the diagnosis of MASH mostly requires histological confirmation by liver biopsy and is therefore invasive and rarely assessed in GAHT cohorts.

In transgender women, the enzyme profile tends to remain stable or improve. Although transaminases are not very sensitive biomarkers of steatotic liver disease most literature reports unchanged or decreasing AST and ALT levels [38, 52, 77]. One study also reported a significant increase in the FIB-4 fibrosis index and, even more relevant than indirect biomarkers, an increase in liver fat determined by NMR at 6 years of follow-up which was significant in transgender men and not significant (p = 0.129) in transgender women [38]. Further, one cross-sectional study comparing transgender women (median age 40.5) with cisgender males (median age 49) and females (median age 54), thus reflecting a statistically significant age difference among groups, that used FibroScan technology to assess liver steatosis and fibrosis found that transgender women subject to GAHT had lower levels of liver fat as compared to transgender women that were not being treated with GAHT (265 vs. 277 dB/m) [66]. The same study observed similar liver stiffness across all groups [66]. Additionally, one study comparing transgender women who had undergone orchiectomy with transgender women who were subject to GAHT found orchiectomy and GAHT to be significantly protective (p = 0.02) against hepatic steatosis assessed by ¹H magnetic resonance spectroscopy [40].

In contrast, in transgender men, the impact of testosterone therapy on liver enzymes (AST/ALT), relatively poor-performing markers to detect MASLD/MASH, appears to be heterogeneous. Although one study found no significant or clinically relevant alterations [54], other investigations have reported statistically significant increases over time [38, 77]. One study also reported a significant increase in the FIB-4 fibrosis index and, even more relevant than indirect biomarkers, an increase in liver fat determined by NMR at 6 years of follow-up which was significant in transgender men and not significant (p = 0.129) in transgender women [38]. Other work reported that the prevalence of coded liver abnormalities defined by elevated AST/ALT values and diagnoses of nonalcoholic fatty liver disease or steatohepatitis was higher in transgender males receiving testosterone therapy with or without GnRHa than in controls [78].

Clinical events and composite outcomes

The analysis of clinical events is the most relevant endpoint for assessing the long-term effects of GAHT. Data in the literature are heterogeneous and limited with respect to study designs, exposure standardization and ascertainment, outcome definitions, comparator selection and follow-up duration in transgender women and men. The data are summarized below and presented in greater detail in Tables 3A and B.

Table 3.

A: Clinical cardiovascular events in transgender women receiving feminizing GAHT: summary of study design, exposure definition, follow-up, and effect estimates (versus cisgender comparators where available). B: Clinical cardiovascular events in transgender men receiving masculinizing GAHT: summary of study design, exposure definition, follow-up, and effect estimates (versus cisgender comparators where available)

Study (Author, year)	Design / setting	Outcome	Population	GAHT exposure definition	Regimen details	Comparator	Follow-up		Key result
A
Getahun et al., 2018; [79]	Retrospective EMR-based cohort (integrated US health systems; Kaiser Permanente regions) with matched cisgender referents	VTE, ischemic stroke, MI	n = 2,118 Matched reference cohorts: 48,686 cisgender men and 48,775 cisgender women	GAHT exposure defined from EMR: estradiol ± antiandrogen in AMAB (TW) and testosterone in AFAB (TM); initiator cohort indexed at first filled estrogen/testosterone prescription.	Not reported	Matched reference cohorts: 48,686 cisgender men and 48,775 cisgender women	Mean follow-up (SD): 3.6 (2.7)		VTE 3.1/1,000 PY (95% CI 2.0–4.6); adj HR 1.6 (0.9–2.9) vs. cis men (NS), 1.1 (0.6–2.1) vs. cis women (NS) Ischemic stroke 2.1/1,000 PY (1.3–3.5); adj HR 1.1 (0.6–2.0) vs. cis men (NS), 1.3 (0.7–2.5) vs. cis women (NS) MI 1.2/1,000 PY (0.6–2.3); adj HR 0.7 (0.3–1.8) vs. cis men (NS), 1.3 (0.5–3.9) vs. cis women (NS)
Poteat et al., 2021 [83]	Cross-sectional, population-based survey analysis (pooled from two probability-based U.S. studies; weighted logistic regression)	Any CVD condition (composite)VTE	n = 25 (analytic sample age ≥ 40) n = 964 cisgender	GAHT exposure (self-report): “ever hormone use” = ever took hormones for gender identity/transition; lifetime hormone use 28.8% (95% CI 11.3–56.2)	Not reported	Cisgender women and cisgender men (age ≥ 40)	Cross-sectional, no follow-up		Any CVD condition: aOR 0.61 (0.16–2.26) vs. cis women (NS); aOR 0.49 (0.13–1.81) vs. cis men (NS) VTE: aOR 1.60 (95% CI 0.32–7.95) vs. cis women (NS); aOR 0.77 (0.14–4.25) vs. cis men (NS)
Lopez et al., 2023; [87]	Retrospective claims-based cohort (Optum deidentified Clinformatics Data Mart; 2007–2021)	Composite CVD, CAD, stroke, CHF, PAD, MI	Transgender people on TTh n = 1,580 Cisgender women on TTh n = 25,796	TTh identified via NDC + CPT codes; binary exposed vs. unexposed Index date = first TTh prescription for exposed; unexposed assigned an index date post 1:1 matching on birth year Continuous enrollment required (6 months pre + 6 months post index)	Not reported	Within-cohort non-TTh users after 1:1 matching on birth year	Overall follow-up distribution: ≤1y 314/1,580 (19.9%); 2–3y 809/1,580 (51.2%); ≥4y 457/1,580 (28.9%)		Composite CVD: HR 0.99 (95% CI 0.69–1.42) NS CAD: HR 1.19 (95% CI 0.59–2.7) NS Stroke: HR 0.97 (95% CI 0.56–1.70) NS CHF: HR 0.41 (95% CI 0.13–1.22) NS PAD / MI: not estimable (counts suppressed; models not converged)
Mullins et al., 2021; [89]	Retrospective chart review at a single pediatric hospital–associated transgender health clinic	Venous/arterial thrombosis during GAHT	n = 429 (70.2% of GAHT initiators); median age at GAHT start 17y (IQR 16–19)	Clinic EMR–confirmed initiation of GAHT Hormones titrated toward physiologic ranges with testosterone monitoring	Testosterone routes: subcutaneous 312 (72.7%), IM 105 (24.4%), gel 11 (2.8%), transdermal 1 (0.7%). Median testosterone dose 70.0 mg (IQR 60.0–80.0). Most recent measured levels (overall testosterone users): testosterone 413.0 ng/dL (IQR 225.0–581.2)	Single cohort, no comparator	Median 577 days (IQR 283–923)		No thrombosis events during GAHT (0 events; no effect estimate). Anticoagulation during GAHT in 5 individuals overall (2 prior thrombosis, 3 prophylaxis) not stratified to TM
Streed Jr. et al., 2025; [88]	Retrospective cohort (VHA EHR 2009–2019); gender identity via validated NLP algorithm applied to clinical text; time-to-event analyses	Incident ASCVD	n = 7,127; cis men n = 918,843; cis women n = 144,090	Gender identity determined via validated algorithm using gender identity/sex-related ICD codes + GAHT use; baseline ASCVD-free	Masculinizing GAHT category (regimen details not reported)	Cis men and cis women; secondary within-trans analysis uses Absent GAHT as reference	Median follow-up 9.39 years		Incident ASCVD: Transmasculine vs. cis women: adj HR 1.11 (1.00–1.22), p = 0.045; vs. cis men: adj HR 0.86 (0.79–0.94), p = 0.001. Within trans: HR 0.86 (0.77–0.95), p = 0.004 vs. Absent GAHT.
van Zijverden et al., 2025; [80]	Retrospective cohort (Amsterdam Cohort of Gender Dysphoria linked to Statistics Netherlands registries; outcomes captured 2012–2022)	MI, ischemic CVA, VTE (hospital-diagnosed acute events)	n = 1,617 Netherlands reference population (residents 2012–2022; ~15.7 million, born 1917–2004) Median age at GAHT start 23.8 (IQR 20.3–31.7) Median age at observation start 28.5 (21.4–43.3)	Inclusion required starting GAHT before 31 Dec 2018; registry-based event follow-up available from 1 Jan 2012. Masculinizing GAHT defined as testosterone therapy	Testosterone gel 20–100 mg/day IM testosterone esters 125–250 mg every 2–3 w IM testosterone undecanoate 1000 mg every 10–14 w Oral testosterone undecanoate 40–160 mg/day (oral largely phased out from 2006)	General population women and men in the Netherlands. Age/sex/calendar-year standardized incidence ratios adjusted for socioeconomic status (SIRa)	Median follow-up 8.4 years (IQR 5.9–11.0); total time at risk 13,457 PY		Event counts + incidence rates (per 1000 PY): * MI: 25 events; 1.9/1000 PY (1.2–2.7) * CVA: 28 events; 2.1/1000 PY (1.4–3.0) * VTE: 17 events; 1.3/1000 PY (0.7–2.0) MI * vs. cis-women: SIRa 4.25 (2.75–6.06), P < 0.001 * vs. cis-men: SIRa 1.50 (0.97–2.15), P = 0.08 Ischaemic cerebrovascular accident (CVA) * vs. cis-women: SIRa 1.61 (1.07–2.26), P = 0.04 * vs. cis-men: SIRa 1.32 (0.87–1.85), P=0.21 Venous thromboembolism (VTE) * vs. cis-women: SIRa 1.20 (0.70–1.83), P = 0.56 * vs. cis-men: SIRa 1.40 (0.82–2.14), P = 0.27
Karalexi et al., 2022; [81]	Population-based nationwide cohort, Swedish healthcare registers; matched general-population reference cohorts (1:10, matched on age + county); Cox proportional hazards time-to-event analyses	Any CVD, MACE, conduction disorder, cerebrovascular event, thrombotic event, CVD death	AMAB with GD + GAHT: n = 847 Reference cohorts: two matched groups without GD/GAHT, matched separately to AMAB and AFAB comparators	AFAB GAHT = testosterone prescriptions (ATC-class level; no route/dose/anti-androgen detail)	Estrogens via ATC codes only; no dosing/route/ information	Matched general-population cohorts without GD/GAHT: No-GD AMAB reference and No-GD AFAB reference	Median follow-up (IQR): 0.73 (0.97)		Any CVD event (Cox HR): * GD with GAHT: HR vs. cis men 1.7 (0.95–2.9), p = 0.073; vs. cis women 2.4 (1.3–4.2), p = 0.003. * GD without GAHT: HR vs. cis men 1.8 (1.0–3.3), p = 0.048; vs. cis women 2.8 (1.5–5.3), p = 0.001. Conduction disorder (Cox HR): * GD with GAHT: HR vs. cis men 4.2 (1.4–12), p = 0.009; vs. cis women 3.4 (1.2–9.6), p = 0.024. * GD without GAHT: HR vs. cis men 1.7 (0.75–4.1), p = 0.196; vs. cis women 2.3 (0.96–5.3), p = 0.062 Other outcomes presented as incidence rates
de Blok CJM et al., 2021; PMID: [85]	Retrospective single-centre clinical cohort (Amsterdam UMC) with linkage to Statistics Netherlands (CBS) for vital status/cause of death; GAHT start 1972–2018	All-cause mortality; cause-specific mortality (incl CVD)	Transgender men (TM): n = 1,641	Receipt of masculinizing GAHT; person-time from start to death, 31 Dec 2018, or last clinic visit	Testosterone (gel 20–100 mg/d; IM esters 125–250 mg q2–3w; TU oral 40–160 mg/d or IM 1000 mg q10–14w); ± progestogens (lynestrenol 5–10 mg/d)	General Dutch population mortality rates (CBS), stratified by age + calendar period; SMRs vs. CM and vs. CW	Median 5y (IQR 2–17); 17,285 PY		All-cause mortality: SMR 1.2 (0.9–1.6) vs. CM; 1.8 (1.3–2.4) vs. CW. Cause-specific mortality (incl CVD): not significantly different vs. population references.
Bonnet et al., 2025; [86]	Retrospective, EMR-based cohort study using the TriNetX Research Collaborative network (66 healthcare organizations). Study period: Jan 1, 2017 to Sep 1, 2024	All-cause death; CVA; AMI; atrial fibrillation (AF); acute pulmonary embolism/deep vein thrombosis (APE/DVT)	1. Trans men treated with testosterone vs. trans men not treated: n = 6,251 vs. n = 6,251 (after 1:1 PSM) 2. Trans men treated with testosterone vs. cis men not treated with testosterone: n = 6,986 vs. n = 6,986 (after 1:1 PSM) 3. Trans men treated with testosterone vs. cis women under contraceptive pill and not treated with testosterone: n = 9,714 vs. n = 9,714	Treated index date = initiation of testosterone Untreated trans men index date = first medical consultation with a transgender diagnosis	No formulation, route, or dose details Testosterone levels were compared 12 months pre-index vs. after 12–24 months of treatment, showing an increase from 4.68 ± 7.90 to 14.98 ± 9.57 nmol/L (P < 0.001)	Untreated trans men Untreated cis men Cis women under contraceptive pill	Trans men treated vs. trans men untreated: Follow-up median 2.1; IQR 3.3 Trans men treated vs. cis men untreated: Follow-up median 2.6; IQR 2.9		TM + GAHT vs. trans men not treated (PSM 1:1; n = 6,251 per group): * Death: NS * Ischemic stroke or thromboembolism: NS * Acute MI: NS * Atrial fibrillation: NS * APE/DVT: NS TM + GAHT vs. cis men not treated (PSM 1:1; n = 6,986 per group): * Death: HR 0.30 (0.17–0.52), P<0.0001 * Ischemic stroke or thromboembolism: NS * Acute myocardial infarction: NS * Atrial fibrillation: HR 0.30 (0.15–0.63), P=0.001 * APE/DVT: NS TM + GAHT vs. cis women under contraceptive pill (matched; n = 9,714 per group): * Myocardial infarction: HR 2.82 (1.12–7.03), P=0.02 (higher in trans men) * Ischemic stroke: NS * Atrial fibrillation: NS * APE/DVT: HR 0.46 (0.22–0.93), P=0.03 (lower in trans men)
Gava et al., 2018; [53]	Single-centre retrospective observational pilot cohort (University of Bologna, Italy); 5-year follow-up; two injectable testosterone regimens	Death, MI, stroke, VTE, pulmonary oil microembolism, other adverse events	Trans men n = 50 (Caucasian, Age 21–42) TU N = 25, TE n=25 Gonadectomy during follow-up 72% overall (reported 72% in each group)	Testosterone-therapy naïve at baseline; regimen defined by TU vs. TE injections	TU 1000 mg IM week 0, week 6, then every 12–16 weeks; TE 250 mg IM every 3–4 weeks; injection interval extended by 1–2 weeks if pre-injection testosterone > 320 ng/dL; doses not modified after gonadectomy	Active comparator TU vs. TE; within-subject baseline comparator	Follow-up duration 5 years testosterone administration		No deaths, MI, stroke, VTE, pulmonary microembolism, or other adverse events in either group over 5 years
Wierckx et al., 2013; [84]	Single specialist center (Ghent University Hospital, Belgium); cross-sectional questionnaire study; morbidity verified in medical files; pre-HRT morbidity from files	VT/PE; MI; TIA/cerebrovascular disease	TM n = 138 respondents; controls for TM: 414 cis men + 414 cis women	Inclusion required ≥ 3 months masculinizing GAHT; treated since 1986–Jun 2012; surveyed Aug–Dec 2012; events classified as occurring “during hormone therapy”	TM regimens: IM mixed testosterone esters (46.4%); testosterone undecanoate (44.9%); transdermal (6.5%); oral TU (1.4%); hysterectomy/oophorectomy 118/138 (86%)	Age-matched cis men and cis women from Flemish population-based survey		Mean hormone therapy duration 7.4 years (overall cohort)	VT/PE 14.5/1000 during HRT. MI 0/1000 (controls: cis men 7.3/1000; cis women 0/1000; both NS). TIA/CVD 0/1000 (controls: cis men 7.3/1000; cis women 7.3/1000; both NS).
B
Getahun et al., 2018; [79]	Retrospective EMR-based cohort (integrated US health systems; Kaiser Permanente regions) with matched cisgender referents	VTE, ischemic stroke, MI	n = 2,842 Matched reference cohorts: 48,686 cisgender men and 48,775 cisgender women	GAHT exposure defined from EMR: estradiol ± antiandrogen in AMAB; initiator cohort indexed at first filled estrogen/testosterone prescription.	Average maximum daily oral estradiol dose during follow-up: * 4.1 mg (range 1–10) among those with VTE or ischemic stroke (n = 11) * 4.2 mg (range 0.3–10) among those without VTE/stroke (n = 391)	Matched reference cohorts: 48,686 cisgender men and 48,775 cisgender women		Mean follow-up (SD): 4.0 years (3.0)	VTE 5.5/1,000 PY (95% CI 4.3–7.0); adj HR 1.9 (1.4–2.7) vs. cis men, 2.0 (1.4–2.8) vs. cis women (both sig) Ischemic stroke 4.8/1,000 PY (3.7–6.3); HR 1.2 (0.9–1.7) vs. cis men (NS), 1.9 (1.3–2.6) vs. cis women (sig) MI 2.9/1,000 PY (2.1–4.1); HR 0.9 (0.6–1.5) vs. cis men (NS), 1.8 (1.1–2.9) vs. cis women (sig)
Poteat et al., 2021; [83]	Cross-sectional, population-based survey analysis (pooled from two probability-based U.S. studies; weighted logistic regression)	Any CVD condition (composite), VTE	n = 70 (analytic sample age ≥ 40) 964 cisgender	GAHT exposure (self-report): “ever hormone use” = ever took hormones for gender identity/transition; TW lifetime hormone use 49.8% (95% CI 34.1–65.5)	Not reported	Cisgender women and cisgender men (age ≥ 40)		Cross-sectional, no follow-up	Any CVD condition: aOR 0.94 (0.46–1.93) vs. cis women (NS); aOR 0.76 (0.37–1.55) vs. cis men (NS) VTE: aOR 3.94 (95% CI 1.24–12.51) vs. cis women (sig); aOR 1.90 (0.53–6.81) vs. cis men (NS)
Mullins et al., 2021; [89]	Retrospective chart review at a single pediatric hospital–associated transgender health clinic	Venous/arterial thrombosis during GAHT	n = 182 (29.8% of GAHT initiators); median age at GAHT start 17y (IQR 16–19)	Clinic EMR–confirmed initiation of GAHT Hormones titrated toward physiologic ranges with estradiol monitoring	Majority on oral estradiol 165 (90.7%) Most recent measured levels (overall estradiol users): estradiol 47.2 pg/mL (IQR 27.2–99.5); testosterone 189.5 ng/dL (IQR 15.2–367.0).	Single cohort, no comparator		Median follow-up 554 days (IQR 283–1037)	No thrombosis events during GAHT (0 events; no effect estimate) Anticoagulation during GAHT in 5 individuals overall (2 prior thrombosis, 3 prophylaxis) not stratified to TW
Poteat et al., 2025; [96]	Observational cohort (LITE Plus); cross-sectional baseline analysis (gender-affirming community health centers: Boston, New York, Washington DC; enrollment Dec 2020–Jun 2022)	Estimated 10-year ASCVD risk 2018 AHA revised Pooled Cohort Equation calculated using sex assigned at birth (male at birth)	Black and Latina transgender women with HIV: n = 95 (median age 41.5y, range 21–71)	Baseline interviewer-administered self-report: lifetime GAHT use (yes/no), current GAHT use (yes/no) Median GAHT duration: 5 years (IQR 0.4, 13.5)	Not reported	Single cohort, no comparator		Visits every 6 months for a total of 5 visits over 24 months	GAHT duration positively associated with PCE risk score in bivariate analysis (p < 0.05), but not significant after multivariable adjustment (age, income, education, ± stress term depending on model) PCE not validated in transgender populations; regimen details not available; restricted population
Korpaisarn et al., 2021; [44]	Single university-based transgender clinic, Thailand; retrospective EMR chart review (Jan 2015–Oct 2019)	Thromboembolism, stroke, CAD, cancer, death	Transmasculine adults initiating testosterone: n = 39; mean age 27.8 ± 6.0y; Thai 97.4%	Testosterone exposure = initiation of clinic-prescribed testosterone; dose titrated to serum levels; targets referenced Endocrine Society 2017	IM testosterone enanthate regimens: 50–100 mg weekly OR 100–200 mg q2weeks OR 250 mg every 3–4 weeks; target “midway between injections” 400–700 ng/dL	No comparator (single-cohort safety description)		Follow-up visits q3–6 months; outcomes assessed up to 2 years after initiation	No clinical events observed: no thromboembolism, stroke, coronary artery disease, cancer, or deaths reported during follow-up (no effect estimates).
Balcerek et al., 2021; [57]	Retrospective single-center clinical cohort (Australia); cross-sectional analysis of feminizing GAHT prescribing patterns and cardiometabolic risk factors, with focus on age ≥ 45y	Clinical history/prevalence (IHD/CVA/VTE)	n = 296 Older subgroup n = 55 aged ≥ 45y; excluded ethinyl estradiol users (n = 12) and anti-androgen monotherapy (n = 3)	EMR estradiol treatment ≥ 6 months Median therapy duration: 25.2 months IQR 10.7–32.9	Majority on oral estradiol (85.8%). Median achieved estradiol concentration 325 pmol/L (IQR 233–428)	No cisgender referent cohort; comparisons are within cohort strata by age group and estradiol route/formulation		Cross-sectional set up, duration of feminizing therapy reported Younger group: 25.2 months IQR 10.7–32.9 ≥ 45y subgroup: median 69.0 months (IQR 45.9–134.9)	Clinical history/prevalence ≥ 45y subgroup IHD 5% (3/55); CVA 4% (2/55); VTE 2% (1/55); no incidence rates or HRs (prevalence only)
van Zijverden et al., 2025; [80]	Retrospective cohort (Amsterdam Cohort of Gender Dysphoria linked to Statistics Netherlands registries; outcomes captured 2012–2022)	MI, ischaemic CVA, VTE (hospital-diagnosed acute events)	n = 2,714 Netherlands reference population (residents 2012–2022; ~15.7 million, born 1917–2004) Median age at GAHT start 30.5 (IQR 23.6–41.3) Median age at observation start 42.9 (29.0–52.9)	Inclusion required starting GAHT before 31 Dec 2018; registry-based event follow-up available from 1 Jan 2012. Feminizing GAHT defined as oestradiol ± anti-androgens	Ethinyl oestradiol 25–100 µg/day Conjugated oestrogens 0.625–1.25 mg/day Oral oestradiol valerate/hemihydrate 2–6 mg/day Injections 10–100 mg every 2–4 w Implants 20 mg every 3–6 mo Patches 50–150 µg/24 h Gel 0.75–3 mg/day Spray 3–6 mg/day From 2001 onward mostly valerate/hemihydrate + patches/gel/spray Anti-androgens: CPA 10–100 mg/day, spironolactone 100–200 mg/day, triptorelin 3.75 mg q4w or 11.25 mg q12w; from 2018 onward mostly triptorelin	General population women and men in the Netherlands. Age/sex/calendar-year standardized incidence ratios adjusted for socioeconomic status (SIRa)		Median follow-up 10.9 years (IQR 6.5–11.0); total time at risk 23,907 PY.	Event counts + incidence rates (per 1000 PY): * MI: 31 events; 1.3/1000 PY (95% CI 0.9–1.8) * CVA: 74 events; 3.1/1000 PY (2.4–3.9) * VTE: 57 events; 2.4/1000 PY (1.8–3.1) MI * vs. cis-women: SIRa 1.57 (1.06–2.16), P=0.03 * vs. cis-men: SIRa 0.60 (0.40–0.82), P < 0.001 CVA * vs. cis-women: SIRa 1.31 (1.03–1.63) P = 0.04 * vs. cis-men: SIRa 1.02 (0.80–1.27), P = 0.88 VTE * vs. cis-women: SIRa 1.80 (1.37–2.30), P<0.001 * vs. cis- men: SIRa 1.75 (1.33–2.23), P<0.001
Karalexi et al., 2022; [81]	Population-based nationwide cohort, Swedish healthcare registers; matched general-population reference cohorts (1:10, matched on age + county); Cox proportional hazards time-to-event analyses	Any CVD, MACE, conduction disorder, cerebrovascular event, thrombotic event, CVD death	AMAB with GD + GAHT: N = 847 Reference cohorts: two matched groups without GD/GAHT, matched separately to AMAB and AFAB comparators	AMAB GAHT = oestrogen prescriptions (ATC-class level; no route/dose/anti-androgen detail)	Estrogens via ATC codes only; no dosing/route/anti-androgen information	Matched general-population cohorts without GD/GAHT: No-GD AMAB reference and No-GD AFAB reference		Median follow-up (IQR): 1.8 (2.1)	Any CVD event (Cox HR): * GD with GAHT: HR vs. cis men 2.4 (0.87–6.9), p = 0.091; vs. cis women 2.3 (0.81–6.4), p = 0.118 * GD without GAHT: HR vs. cis men 1.5 (0.71–3.4), p = 0.272; vs. cis women 1.8 (0.81–3.8), p = 0.157 Conduction disorder (Cox HR): * GD with GAHT: HR vs. cis men 2.2 (1.1–4.4), p = 0.035; vs. cis women 2.7 (1.3–5.5), p = 0.008 * GD without GAHT: HR vs. cis men 1.0 (0.32–3.4), p = 0.941; vs. cis women 1.3 (0.41–4.3), p = 0.635
Wierckx et al., 2013; [84]	Single specialist center (Ghent, Belgium); cross-sectional questionnaire study; morbidity verified via medical files.	VT/PE; MI; TIA/cerebrovascular disease	TW n = 214 respondents; controls: 640 cis men + 619 cis women	≥ 3 months GAHT; treated 1986–Jun 2012; events classified as occurring “during hormone therapy”	Regimens: Transdermal gel (35.5%), patch (13.6%), oral EV (42.5%), ethinyl estradiol (3.2%); 65% orchiectomy	Age-matched cis men and cis women from Flemish population-based survey		Mean 7.4 years	VT/PE: 5.1%; MI: 18.7/1000 (Higher vs. CW, p = 0.001; NS vs. CM); TIA/CVD: 23.4/1000 (Higher vs. CM, p = 0.03; NS vs. CW)
Streed Jr. et al., 2025; [88]	Retrospective cohort (VHA EHR 2009–2019); validated NLP algorithm; time-to-event analyses.	Incident ASCVD	n = 8,013; cis men n = 918,843; cis women n = 144,090	Baseline ASCVD-free; Gender identity via algorithm (ICD codes + GAHT use); GAHT-exposure: Absent vs. Feminizing	Feminizing GAHT category (regimen details not reported)	Cis men and cis women; “Absent GAHT” within-trans reference		Median 9.39 years	Incident ASCVD: vs. CW: adj HR 1.28 (1.16–1.42), p < 0.001; vs. CM: adj HR 0.91 (0.83–1.00), p = 0.052. Within trans: GAHT vs. Absent HR 1.02 (p = 0.77).
Nota et al., 2019; [82]	Single-center Amsterdam gender clinic cohort (1972–2015); retrospective record review; age-adjusted SIRs.	Stroke, MI, VTE	TW = 2,517 (median age 30)	Receipt of prescribed hormones with known start date + ≥ 1 follow-up; follow-up accrued during THT.	Estrogens (± antiandrogens); adolescents: puberty suppressors with THT from age 16.	General-population women and men (age/sex-specific incidence rates)		Mean 9.07y; median 5.95y (Total: 22,830 PY)	Stroke: SIR 2.42 (vs. CW, sig), 1.80 (vs. CM, sig). MI: SIR 2.64 (vs. CW, sig), 0.79 (vs. CM, NS). VTE: SIR 5.52 (vs. CW, sig), 4.55 (vs. CM, sig).
de Blok CJM et al., 2021; [85]	Retrospective single-centre clinical cohort (Amsterdam UMC) with linkage to Statistics Netherlands (CBS) for vital status/cause of death; GAHT start 1972–2018	All-cause mortality; CVD mortality; MI mortality (cause-specific)	Transgender women (TW): n = 2,927	Receipt of feminizing GAHT; person-time from start to death, 31 Dec 2018, or last clinic visit	Estrogens (historical ethinyl estradiol/conjugated estrogens; current valerate/patch/gel) ± antiandrogens (CPA 10–100 mg/day or spironolactone 100–200 mg/day)	General Dutch population mortality rates (CBS), stratified by age + calendar period; SMRs vs. CM and vs. CW		Median 11y (IQR 4–22); 40,232 PY	CVD mortality: SMR 1.4(1·4–1·9) (vs. CM, sig); 2.6 (2.1–2.7) (vs. CW, sig). MI mortality: SMR 1.1 (vs. CM, NS); 3.0 (1.7–4.5) (vs. CW, sig)
Getahun et al., 2018; [79]	Retrospective EMR-based cohort (integrated US health systems; Kaiser Permanente regions) with matched cisgender referents	VTE, ischemic stroke, MI	n = 2,842 Matched reference cohorts: 48,686 cisgender men and 48,775 cisgender women	GAHT exposure defined from EMR: estradiol ± antiandrogen in AMAB; initiator cohort indexed at first filled estrogen/testosterone prescription.	Average maximum daily oral estradiol dose during follow-up: * 4.1 mg (range 1–10) among those with VTE or ischemic stroke (n = 11) * 4.2 mg (range 0.3–10) among those without VTE/stroke (n = 391)	Matched reference cohorts: 48,686 cisgender men and 48,775 cisgender women		Mean follow-up (SD): 4.0 years (3.0)	VTE 5.5/1,000 PY (95% CI 4.3–7.0); adj HR 1.9 (1.4–2.7) vs. cis men, 2.0 (1.4–2.8) vs. cis women (both sig) Ischemic stroke 4.8/1,000 PY (3.7–6.3); HR 1.2 (0.9–1.7) vs. cis men (NS), 1.9 (1.3–2.6) vs. cis women (sig) MI 2.9/1,000 PY (2.1–4.1); HR 0.9 (0.6–1.5) vs. cis men (NS), 1.8 (1.1–2.9) vs. cis women (sig)
Poteat et al., 2021; [83]	Cross-sectional, population-based survey analysis (pooled from two probability-based U.S. studies; weighted logistic regression)	Any CVD condition (composite), VTE	n = 70 (analytic sample age ≥ 40) 964 cisgender	GAHT exposure (self-report): “ever hormone use” = ever took hormones for gender identity/transition; TW lifetime hormone use 49.8% (95% CI 34.1–65.5)	Not reported	Cisgender women and cisgender men (age ≥ 40)		Cross-sectional, no follow-up	Any CVD condition: aOR 0.94 (0.46–1.93) vs. cis women (NS); aOR 0.76 (0.37–1.55) vs. cis men (NS) VTE: aOR 3.94 (95% CI 1.24–12.51) vs. cis women (sig); aOR 1.90 (0.53–6.81) vs. cis men (NS)

Tables summarize primary human studies reporting clinical cardiovascular outcomes during GAHT. “Population” indicates cohort sizes (transgender cohort and comparator cohorts where applicable). “GAHT exposure definition” reflects how GAHT (or GAHT category) was operationalized (e.g., prescriptions, pharmacy fills, self-report, clinic-confirmed initiation). “Regimen details” report formulation/route/dose when provided in the source; “Not reported” indicates unavailable regimen information. “Follow-up” is reported as described in the original study (e.g., median [IQR], mean [SD], person-years, or cross-sectional where applicable). “Key result” reports the main outcome-specific effect estimates and significance as provided by the original study (e.g., incidence rates, hazard ratios, standardized incidence ratios, odds ratios) and specifies the reference comparator

Note: de Blok et al. [85] and van Zijverden et al. [80] both use Amsterdam Cohort of Gender Dyspohoria-derived participants but with different inclusion criteria and follow-up windows (GAHT initiation era 1972–2018 vs. registry event capture 2012–2022), resulting in partial overlap with differing regimen eras and event ascertainment; findings should be interpreted as complementary

Cardiovascular event estimates in transgender women differed according to whether cisgender women or cisgender men constituted the comparator population. Evidence indicates an increased risk of incident CVD and conduction disorders [81], with higher cardiovascular mortality ratios than both cisgender men (SMR 1.4) and women (SMR 2.6) [85]. MI incidence is significantly higher in transgender women than in cisgender women [81–84], but typically comparable to or lower than cisgender men [82, 83]. Specifically, one study shows a standardized MI incidence ratio of 1.57 relative to cisgender women [82]. VTE risks are consistently and significantly elevated compared to both cisgender men and women across multiple cohorts [81, 82, 84, 85]. Similarly, risks for ischemic stroke and cerebrovascular accidents (CVA) are higher than in cisgender women [81, 82, 84, 85]. However, stroke risk often aligns with cisgender men [82], particularly after adjusting for socioeconomic factors [84].

In transgender men, testosterone therapy does not appear to significantly increase composite CVD risk compared to untreated individuals [86, 87]. While overall mortality is comparable to cisgender men, it remains elevated relative to cisgender women [85]. MI and CVA risks are significantly higher than in cisgender women exemplified by a four times higher MI risk [80]. GAHT in transgender men was also associated with a significantly higher incidence of atherosclerotic CVD in comparison to cisgender women, but lower when compared to cisgender men [88]. However, these outcomes often align with cisgender men [82]. Transgender men also exhibit an increased risk of conduction disorders [81]. Conversely, findings for VTE are often similar to cisgender women or men [80] or lack statistical significance across smaller cohorts due to low event frequency [44, 89].

Discussion and future directions

The principal finding of this review is that feminizing and masculinizing GAHT are associated with distinct, domain-specific cardiometabolic patterns rather than a uniform shift toward the cardiometabolic profile typically observed in cisgender people of the affirmed gender. While one might intuitively expect that feminizing therapy in transgender women would result in a cardiometabolic risk profile resembling that of cisgender women, and masculinizing therapy in transgender men in a profile comparable to that of cisgender men, the currently available evidence does not support such a simplistic and complete shift in risk patterns. The interpretation of these findings is complicated by substantial heterogeneity across study methods, amongst which differences in GAHT regimens, baseline cardiometabolic risk, age distribution, duration of exposure, and clinical context, which limit direct comparability between cohorts. In the translation of these findings to the clinical perspective, one encounters complications such as the use of sex-specific reference values and equation-based estimates (e.g., creatinine-based eGFR). These introduce methodological uncertainty that may influence both biomarker interpretation and cardiometabolic risk assessment in individuals undergoing GAHT.

In transgender women, anthropometric changes are rather substantial although they are not often reflected in BMI, showing the limitations of BMI as a measure of adiposity and highlighting the need for more granular assessments of body composition and fat distribution. Evidence shows increases in fat mass and muscle fat infiltration and decreases in lean mass [26, 37, 38]. Further, favorable subcutaneous/visceral fat ratios were observed when compared to cisgender men with decreases in waist circumference [27, 37]. This distinction is clinically relevant, as different adipose depots exhibit different metabolic functions: visceral and ectopic fat are more strongly linked to insulin resistance and cardiometabolic risk than subcutaneous fat [90], which may partly explain the discordance between stable BMI and measurable metabolic changes. Results about changes in weight or BMI were heterogeneous with most studies observing no statistically significant change. In transgender men BMI tends to increase but not universally [28, 31, 33, 36, 38, 41–43, 77]; while lean muscle mass increases, fat mass often decreases or remains stable and long-term data suggest increased visceral and liver fat [26, 37, 38].

With respect to glucose and insulin metabolism in transgender women a trend towards higher insulin resistance, predominantly assessed using indirect measures such as HOMA-IR, was observed but was not uniform across studies. Notably, no hyperinsulinemic euglycemic clamp studies, ideally combined with stable isotope tracers to differentiate between hepatic and skeletal muscle insulin resistance, have been performed in transgender individuals receiving GAHT, limiting mechanistic interpretation of these findings. Meanwhile fasting glucose and HbA1c mostly remained stable in the available cohorts, which were predominantly young and metabolically healthy at baseline, limiting conclusions about long-term trajectories as individuals age [26, 30–32, 37, 40, 48, 51, 52, 62, 78]. Ageing-related metabolic deterioration may modify these patterns over time, but longitudinal data extending into older age are currently lacking. On the other hand, in transgender men findings with respect to glucose metabolism and insulin resistance were mixed with often no significant changes reported [31, 38, 48, 53, 54]. It is further important to mention that, overall, in transgender populations the effect of GAHT on dietary intake and the response to glucose-lowering pharmacotherapy have not been systematically assessed and may represent significant contributors to glycemic outcomes with particular relevance in a long-term context and combined with ageing.

Dyslipidemia, characterized by elevated LDL-C, reduced HDL-C, and increased triglycerides, represents a major modifiable cardiovascular risk factor and is a central component of commonly used risk prediction tools such as the Framingham Risk Score and ESC SCORE [91, 92]. These algorithms, however, have not been formally validated in transgender populations. Yet, regarding lipid profile changes in transgender women, the literature rather homogenously reports a favorable shift with decreases recorded in total cholesterol, LDL-C, and triglycerides [9, 17, 27, 32, 37, 51, 56] while changes in HDL-C are seemingly regimen-dependent [9, 27, 28, 32]. In transgender men GAHT was often associated with a shift toward atherogenic lipid profiles characterized primarily by decreased HDL-C and increased LDL-C as well as triglycerides [27, 31, 37, 38, 41, 56]. Additionally, it is worth mentioning that the impact of GAHT on response to lipid-lowering therapies such as statins remains largely unexplored, and it is unclear whether GAHT modifies the clinical course of e.g. familial hypercholesterolemia or other genetically determined dyslipidemias. Furthermore, potential interactions with genetic modifiers of lipid metabolism, including apoE genotype, have not been systematically studied in this population.

Results on blood pressure changes are also mixed. In transgender women blood pressure often remained unchanged or decreased with no studies reporting an increase [29–31, 33, 38, 56]. One study found a significantly lower prevalence of coronary plaques compared to cisgender males [62]. On the other hand, in transgender men the results are also heterogenous, although most studies recorded no significant change [27, 28, 31, 33, 38, 43] or an increase in blood pressure [29, 30, 36, 56, 63].

While evidence on renal and hepatic outcomes is insufficient and calls for further study, more data are available regarding the impact of GAHT on cardiovascular events, a clinically highly relevant endpoint to assess the long-term effects of GAHT. However, these are mostly based on limited and heterogeneous data collected from cohorts insufficiently powered to detect differences in relatively infrequent events. Larger cohorts with longer follow-up durations and more comprehensive assessment of contemporary cardiovascular risk factors are required to draw more definitive conclusions. While in transgender women consistently an elevated incidence of VTE and CVA is observed in large cohorts [10, 79, 82–84], data on coronary heart disease-related events such as myocardial infarction are more heterogeneous. In transgender men MI and CVA incidence were higher than in cisgender women but generally not higher than in cisgender men [81, 82, 84–87].

The anthropometric changes observed reflect the ability of GAHT to remodel the body composition phenotype towards the sex of the affirmed gender. According to a proposed cardiometabolic risk continuum [7], it can be hypothesized that these body composition changes represent underlying drivers of the observed alterations in insulin sensitivity and metabolic risk. Changes in total and regional adiposity, particularly shifts in visceral and ectopic fat depots, are likely to directly influence insulin resistance and cardiometabolic risk, whereas concomitant alterations in skeletal muscle mass may exert counterbalancing effects. This suggests a complex mechanism of action in which GAHT affects cardiometabolic risk through two interacting pathways. Of note, such mechanistic conclusions cannot currently be firmly drawn, and more work is needed to support this proposed concept. In addition, testosterone- or estrogen-induced skeletal muscle hypertrophy, potential alterations in muscle fiber composition, and changes in intramyocellular lipid content may further modulate peripheral insulin sensitivity, although these mechanisms also await experimental confirmation in transgender cohorts. On the one hand, indirect effects occur via GAHT-driven changes in body composition and fat distribution, which are closely linked to insulin sensitivity and lipid metabolism. On the other hand, direct effects occur through sex steroid signaling via nuclear receptors, influencing vascular tone, hepatic lipid handling, coagulation pathways, and inflammatory signaling [13]. These pathways conceivably interact, and their relative contribution may depend on regimen details (e.g., route of estradiol, anti-androgen type, testosterone formulation), baseline risk, and duration of exposure. Lastly, the potential impact of sex chromosomes in this pathway interplay is only beginning to be addressed and understood. Therefore, this review aimed to translate a rather heterogeneous literature into a clinically interpretable structure by separating feminizing and masculinizing GAHT and presenting evidence for transgender women and men in parallel. This synthesis is intended to provide a structured summary of the currently available evidence base, clarifying what is consistently evidence-supported versus what remains uncertain or context dependent.

Combined, the present review concludes that, despite the mounting body of available studies, key uncertainties and gaps remain to be resolved. Firstly, high-quality evidence assessing the long-term effects of feminizing and masculinizing GAHT on cardiometabolic endpoints and events using sufficiently powered, homogenous prospective cohorts is extremely limited. Secondly, effect analysis in the current literature is hindered by inadequate regimen standardization. This lack of regimen standardization and uniform outcome reporting across cohorts makes it currently difficult to derive evidence-based, broadly generalizable monitoring and screening protocols for cardiometabolic disease in individuals receiving GAHT. Consensus-style guidance is available (e.g., the American Heart Association scientific statement), which provides practice- and research-oriented recommendations for cardiovascular risk assessment and care in transgender and gender-diverse populations despite limited longitudinal evidence [19]. Thirdly, integrated longitudinal and mechanistic studies linking body composition shifts to changes in lipids, insulin sensitivity, vascular function, inflammation, and thrombosis, rather than isolated biomarker reporting are necessary. Lastly, data on individuals with long-term exposure, as well as studies exploring the impact on individuals of non-European ancestry, are also very limited. Therefore, large, well-characterized prospective cohort studies and (inter)national registries with standardized phenotyping and structured long-term follow-up are needed to optimize evidence-based health care for transgender populations.

Fig. 1.

a, b: The cardiometabolic risk continuum in gender-affirming hormone therapy. Conceptual schematic illustrating a hypothesized progression from obesity and upstream metabolic disturbances to major clinical cardiometabolic outcomes in transgender women (a) and transgender men (b) receiving gender affirming hormone therapy. Gender-affirming hormone therapy is associated with changes across multiple cardiometabolic domains, including anthropometric changes that conceivably fuel dyslipidemia, insulin resistance/type 2 diabetes, hypertension, hepatic steatosis (MASLD/MASH), and chronic kidney disease (CKD). All these cardiometabolic sequelae share an increase in systemic inflammatory load and converge to increase cardiovascular disease (CVD) risk and respective downstream clinical events (myocardial infarction, stroke, and heart failure). In this schematic, node size reflects the quantity of available evidence, with larger circles indicating a higher number of unique primary research sources identified for that domain (see Supplementary Table 1). Node color intensity represents the unanimity index, where darker shades correspond to a higher percentage of unambiguous evidence supporting the dominant clinical direction. More intense colors signify high directional unanimity, while lighter shades represent lower unanimity due to higher heterogeneity or a greater proportion of mixed findings within the primary human literature (see Supplemental Methods for calculation). Arrows depict proposed pathophysiological links and clinical trajectories and are not intended to imply proven causality; effects are subject to variation based on hormone regimen, route of administration, use of adjunctive therapies (e.g., anti-androgens), baseline risk profiles, and duration of exposure

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

GAHT

Gender-affirming hormone therapy

THT

Transgender hormone therapy

TW

Transgender women

TM

Transgender men

AMAB

Assigned male at birth

AFAB

Assigned female at birth

GD

Gender dysphoria

CVD

Cardiovascular disease

ASCVD

Atherosclerotic cardiovascular disease

CAD

Coronary artery disease

MI

Myocardial infarction

AMI

Acute myocardial infarction

HF

Heart failure

CHF

Congestive heart failure

CVA

Cerebrovascular accident

VTE

Venous thromboembolism

VT

Venous thrombosis

PE

Pulmonary embolism

DVT

Deep vein thrombosis

APE

Acute pulmonary embolism

TIA

Transient ischemic attack

IHD

Ischemic heart disease

MACE

Major adverse cardiovascular event

AF

Atrial fibrillation

IR

Incidence rate

PY

Person-years

HR

Hazard ratio

aHR

Adjusted hazard ratio

SIR

Standardized incidence ratio

SMR

Standardized mortality ratio

OR

Odds ratio

aOR

Adjusted odds ratio

CI

Confidence interval

IQR

Interquartile range

SD

Standard deviation

PSM

Propensity score matching

EMR/EHR

Electronic medical/health record

NDC

National Drug Code

CPT

Current Procedural Terminology

NLP

Natural language processing

CBS

Statistics Netherlands

NS

Not statistically significant

Author contributions

MF, conception and design, acquisition, analysis and interpretation of studies, drafting of the manuscript; UJFT, conception and design, acquisition, analysis and interpretation of studies, drafting of the manuscript. Both authors read and approved the final manuscript.

Funding

UJFT is supported by grants from CIMED (FoUI-963234), the Swedish Heart-Lung Foundation (grant Nos. 20220271, 20241339) and ALF Medicin by Region Stockholm (project grant FOUI-962738).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.