Lipoprotein subtypes after testosterone therapy in transmasculine adolescents

Abstract

Differences in lipoprotein-particle subclasses between men and women start in puberty and narrow after menopause, suggesting a role for sex steroids. In this cross-sectional cohort study, we examined lipoprotein subtype profiles in transmasculine adolescents treated with testosterone. Transmasculine adolescents (n = 17) had lipoprotein profiles that were similar to those of cisgender males (n = 33) and more atherogenic than those of cisgender females (n = 32), with higher concentrations of small low-density lipoprotein (LDL) particles (435 ± 222 nmol/L vs. 244 ± 163 nmol/L, p = 0.008) and lower concentrations of large high-density lipoprotein (HDL) particles (1.5 ± 1.3 μmol/L vs 2.7 ± 1.2 μmol/L, p = 0.003) when compared to cisgender females. Thus, testosterone appears to be a major contributor to differences in lipoprotein profiles, a surrogate for cardiovascular disease risk, between cisgender women and both transgender and cisgender men.

INTRODUCTION

Morbidity and mortality from cardiovascular disease are higher in men than in women, a difference that is attributable in part to differences in cholesterol profiles.¹ Compared to women, men have more atherogenic cholesterol profiles characterized both by differences in overall cholesterol concentrations (i.e., higher low-density lipoprotein cholesterol, LDL-C, and lower high-density lipoprotein cholesterol, HDL-C) and by differences in lipoprotein-particle size (i.e., smaller LDL and HDL particles).² These differences in cholesterol profiles have in turn been attributed to the effects of sex steroids such as testosterone, but it is unclear if other factors such as the presence of a Y chromosome play a role independently of sex steroids.³

Transmasculine adolescents have XX chromosomes, have a male gender identity, and often receive testosterone treatment to induce secondary sex characteristics concordant with their gender identity.^4,5 In this study, we examined lipoprotein profiles in individuals with XX chromosomes treated with testosterone.

METHODS

Lipoprotein-particle analysis of previously collected nonfasting plasma samples from the Boston Children’s Hospital BioBank was conducted at Labcorp (Morrisville, NC) by nuclear magnetic resonance (NMR) spectroscopy on the Vantera^® 400 MHz analyzer platform using the NMR LipoProfile-4 deconvolution algorithm.^6,7 Samples were collected between January 2017 and August 2020 from patients age 12 to 23 years. Individuals with medical conditions associated with cholesterol abnormalities (e.g., familial hypercholesterolemia, diabetes mellitus, thyroid disease, eating disorders, mitochondrial disease, liver disease) or who were taking medications known to effect cholesterol metabolism (e.g., statins, carbamazepine, valproic acid, isotretinoin) were excluded. This study was approved by the Boston Children’s Hospital Institutional Review Board.

Continuous variables were compared via ANOVA followed by Dunnett’s test for normally distributed variables and the Mann-Whitney U test for those that were not normally distributed. Categorical variables were compared via Fisher’s exact test. Comparisons of lipoprotein subtypes were made using a multivariable regression model including age, race, and body mass index (BMI). A p value of < 0.01 was considered significant without additional correction for multiple testing, as is standard in the field.⁸ Stata Statistical Software: Release 16 (College Station, TX) was used for analyses.

RESULTS

The lipoprotein profiles of 17 transmasculine adolescents who were treated with testosterone for mean ± standard deviation 1.2 ± 0.8 years (range 0.33 to 3.3 years) were compared to those of 33 cisgender male adolescents and 32 cisgender female adolescents. Of the transmasculine adolescents, 12 received 50 mg of testosterone cypionate subcutaneously weekly, 4 received 60 mg weekly, and 1 individual received 80 mg weekly. The mean serum testosterone in this group was in the adult male range (468 ± 202 ng/dL). None of the transmasculine adolescents had been previously treated with GnRH analogs. The ages of the transmasculine group (median 18.4 years, interquartile range 17.6 to 19.5 years) were similar to those of the cisgender male group (17.8 years, 17.0 to 19.4 years) and cisgender female group (17.6 years, 17.1 to 18.5 years). Most participants identified as white (82% in the transmasculine group, 76% in the cisgender male group, and 72% in the cisgender female group). There was no significant difference between the groups in age, race, or terms of blood pressure. There was, however, a difference in body-mass index (BMI), with transmasculine adolescents having higher BMI (mean ± SD 27 ± 4 kg/m²) than either cisgender males (23 ± 4 kg/m², p = 0.004) or cisgender females (22 ± 4 kg/m², p = 0.001). Because of this difference, all analyses were adjusted for BMI.

Transmasculine adolescents had higher concentrations of small LDL particles, which are considered atherogenic, than cisgender females (435 ± 222 nmol/L vs. 244 ± 163 nmol/L, p = 0.008), but not cisgender males (359 ± 135 nmol/L, p = 0.2) (Table 1, Figure 1). There were no differences in medium or large LDL-particle concentrations between transmasculine adolescents and cisgender females. As a net consequence of higher concentrations of small LDL particles, transmasculine adolescents had higher concentrations of total LDL particles and apolipoprotein B (ApoB, the major lipoprotein on LDL particles), than cisgender females, but not cisgender males (Table 1). LDL-C concentrations were higher in transmasculine adolescents than either cisgender females (81 ± 27 mg/dL vs. 67 ± 19 mg/dL) or cisgender males (64 ± 18 mg/dL), but this difference did not reach statistical significance when controlled for BMI.

Table 1.

Lipoprotein subtypes in transmasculine adolescents treated with testosterone and in cisgender adolescents

	Transmasculine Adolescents (n = 17)	Cisgender Adolescents (n = 65)		Transmasculine Adolescents vs. Cisgender Female Adolescents		Transmasculine Adolescents vs. Cisgender Male Adolescents
		Male (n = 33)	Female (n = 32)	*p-value (unadjusted)	*p-value in model including BMI, race, and age	*p-value (unadjusted)	*p-value in model including BMI, race, and age
Total TRL particles, mean (SD), nmol/L	122 (40)	112 (32)	116 (47)	0.9	-	0.3	-
Very large TRL particles, >90 nm	0.3 (0.4)	0.4 (0.7)	0.2 (0.3)	0.9	-	0.8	-
Large TRL particles, 50-90 nm	1.9 (1.6)	2.5 (3.2)	1.3 (2.0)	0.6	-	0.7	-
Medium TRL particles, 37-49 nm	19 (10)	21 (10)	18 (12)	0.8	-	0.5	-
Small TRL particles, 30-36 nm	56 (28)	45 (25)	51 (39)	0.8	-	0.08	-
Very small TRL particles, 24-29 nm	44 (31)	41 (30)	47 (29)	1.0	-	0.8	-
LDL Cholesterol, mean (SD), mg/dL	81 (27)	64 (18)	67 (19)	0.054	0.012	0.011	-
LDL Particle Diameter, mean (SD), nm	21 (0.3)	21 (0.3)	21 (0.4)	0.01	0.12	0.6	-
Total LDL particles, mean (SD), nmol/L	1156 (361)	925 (244)	908 (228)	0.005	0.001	0.007	0.011
Large LDL particles, 21.5-23 nm	337 (178)	240 (146)	338 (179)	1.0	-	0.07	-
Medium LDL particles, 20.5-21.4 nm	385 (277)	326 (236)	326 (202)	0.6	-	0.5	-
Small LDL particles, 19-20.4 nm	435 (222)	359 (135)	244 (163)	0.001	0.008	0.1	-
Apolipoprotein B, mean (SD), mg/dL	71 (22)	58 (15)	59 (16)	0.04	0.005	0.02	-
HDL Cholesterol, mean (SD), mg/dL	45 (12)	47 (12)	57 (11)	0.002	0.007	0.7	-
HDL Particle Diameter, mean (SD), nm	9 (0.5)	9.0 (0.3)	9.5 (0.3)	<0.001	0.001	0.7	-
Total HDL particles, mean (SD), umol/L	17 (3)	18 (3)	18 (3)	0.4	-	0.5	-
Large HDL particles, 10-13 nm	1.5 (1.3)	1.6 (0.9)	2.7 (1.2)	0.001	0.003	0.2	-
Medium HDL particles, 8.5-10 nm	4.4 (2.2)	4.6 (1.6)	4.6 (1.6)	0.8	-	0.4	-
Small HDL particles, 7.4-8.5 nm	12 (4)	12 (2)	11 (2.8)	0.8	-	0.8	-
Apolipoprotein A1, mean (SD), mg/dL	110 (19)	113 (21)	128 (22)	0.011	0.02	0.7	-

^*Unadjusted p-values were determined by Dunnett’s test; adjusted p-values were determined by multivariate regression in a model including BMI, race, and age; p-values < 0.01 are considered significant and are indicated in bold.

BMI, body-mass index; TRL, triglyceride-rich lipoprotein; LDL, low-density lipoprotein; HDL, high-density lipoprotein

Figure 1.

Lipoprotein profiles in transmasculine, cisgender male, and cisgender female adolescents. A. Concentrations of subclasses of low-density lipoprotein (LDL) particles. B. Concentrations of subclasses of high-density lipoprotein (HDL) particle concentrations. Transmasculine adolescents receiving testosterone therapy are represented by black circles, cisgender male adolescents are represented by black squares, and cisgender female adolescents are represented by open squares. Transmasculine adolescents had significantly higher concentrations of small LDL particles and lower concentrations of large HDL particles compared to cisgender female adolescents (asterisks p < 0.01).

Transmasculine adolescents also had lower concentrations of large HDL particles compared to cisgender females (1.5 ± 1.3 µmol/L vs. 2.7 ± 1.2 µmol/L, p = 0.003) (Table 1, Figure 1). There were no differences in the concentrations of medium or small HDL particles between the groups. As a consequence of the lower concentration of large HDL particles, transmasculine adolescents had lower concentrations of high-density lipoprotein cholesterol (HDL-C) than cisgender females (45 ± 12 mg/dL vs. 57 ± 11 mg/dL, p = 0.007) and a mean HDL particle diameter that was more similar to that of cisgender males (9.0 ± 0.5 nm vs. 9.0 ± 0.3 nm, p = 0.7) than that of cisgender females (9.5 ± 0.3 mg/dL, p = 0.001) (Figure 2).

Figure 2.

Distributions of mean particle sizes among subjects in each group. A. Low density lipoprotein (LDL)-particle size and B. High density lipoprotein (HDL)-particle size. Transmasculine adolescents are represented by the thick solid line, cisgender female adolescents are represented by the thin solid line, and cisgender male adolescents are represented by the dotted line. The distribution of both mean LDL and mean HDL particle size across transmasculine adolescents treated with testosterone is more similar to that of cisgender male adolescents than that of cisgender female adolescents.

There were no significant differences in triglyceride-rich particles between transmasculine adolescents and either cisgender group (Table 1).

DISCUSSION

We have found that transmasculine adolescents treated with testosterone have lipoprotein profiles more similar to male than female cisgender adolescents. Specifically, compared to cisgender females, transmasculine adolescents have higher concentrations of small LDL particles and lower concentrations of large HDL particles. Thus, sex-steroid exposure appears to be the main factor leading to the more atherogenic lipoprotein subtype profiles found in males.

Cardiovascular mortality is higher in men than in women, a difference that has been attributed at least in part to the more atherogenic lipoprotein profile found in men. The difference in lipoprotein profiles starts at puberty and persists until menopause, suggesting a role for sex steroids.^2,9 Indeed, excess cardiovascular risk and abnormal lipid profiles have been associated with both low (i.e., post-menopausal women and hypogonadal men) and elevated sex-steroid states (i.e., women with polycystic ovary syndrome).^10,11 In cisgender men, low serum testosterone has been associated with more atherogenic lipid profiles, with higher total cholesterol and triglycerides, and lower HDL-C as well as lower concentrations of large HDL and large LDL particles^11,12 Lower serum testosterone in cisgender men is also associated with excess cardiovascular risk. Mendelian randomization studies provide evidence that this is an effect of testosterone, but it remains possible that a portion of the observed associations may be due to testosterone serving as a marker of other factors, such as overall health and/or aging, rather than acting as a causal agent.¹³

Testosterone treatment in transmasculine adolescents has been shown to lower HDL-C concentration into or below the typical male range.^14-16 Here we refine these prior findings to show that this difference is due specifically to a decrease in the number of large HDL particles.

In contrast to observations about HDL-C, changes in LDL-C have not been consistently seen in transmasculine individuals treated with testosterone.^15,17,18 We found a higher concentration of LDL-C in transmasculine adolsecents than in cisgender adolescents. Although the LDL-C concentration in transmasculine adolescents remained in the low range for cardiovascular risk, if this elevation persists into adulthood it may result in a greater number of transmasculine individuals with LDL-C above the target range. Additionally, our finding of a higher concentration of small LDL particles, a ‘male-type’ pattern, in transmasculine adolescents treated with testosterone suggests that testosterone may affect a clinically relevant subset of LDL particles without necessarily affecting total LDL-C concentration.¹⁷ Thus, detailed lipoprotein subtype profiling in transmasculine individuals may reveal changes beyond what can be determined through measurement of LDL-C alone and may help clinicians caring for transgender individuals to better assess cardiovascular risk.

Testosterone therapy in transgender male adults has been associated with a more atherogenic lipid profile, with increases in LDL-C and triglycerides and a decrease in HDL-C.^17,19 In transgender male adolescents, studies examining lipid concentrations with testosterone therapy have found mixed results; with some studies finding a increase in LDL-C and decrease in HDL-C, and others reporting no change.^15,18,20,21 We have found in transgender male adolescents a more atherogenic lipid profile associated with testosterone therapy. Despite these effects of testosterone on lipid profiles, an increase in cardiovascular events among transgender men has yet to be reported, although this could be secondary to the young age of available transgender male cohorts.^22,23 Our finding that testosterone treatment in the adolescent period leads to a ‘male-type’ pattern of lipoproteins suggests that the overall cardiovascular risk of transgender men may be similar to that of cisgender men. Future work is needed to determine the extent to which these changes in lipoprotein profiles translate into changes in cardiovascular risk.

Because we studied samples that had been collected previously and samples collected before testosterone treatment were not available, we were unable to compare lipoprotein profiles before and after treatment. We also did not have the opportunity to collect data on other important factors that affect cardiovascular health such as tobacco use, dietary habits, and physical activity. Additionally, because samples were not consistently obtained while the patient was fasting, we were not able to meaningfully analyze triglyceride-rich lipoprotein particles. Because testosterone treatment in transmasculine adolescents results in both an increase in serum testosterone and suppression serum estradiol, this cross-sectional cohort study was not able to determine if the changes in lipoproteins are secondary to the changes in testosterone, estradiol, or both.

This study is an important first look at the effects of testosterone treatment on lipoprotein subtypes in adolescents with XX sex chromosomes and has implications for cardiovascular health in both transgender and cisgender individuals. Future studies are needed to further elucidate factors that contribute to the effects of sex hormones on lipoprotein subtypes and the risk of cardiovascular disease.

Abbreviations:

NMR

Nuclear Magnetic Resonance

BMI

Body Mass Index

LDL

Low-density lipoprotein

LDL-C

Low-density lipoprotein cholesterol

HDL

High-density lipoprotein

HDL-C

High density lipoprotein cholesterol