Lipid/lipoprotein cholesterol values and sex-hormone-binding globulin levels were determined in 40 transsexual males aged 20-38, 20 castrated and 20 non-castrated, taking conjugated estrogens to induce female characteristics. Variables controlled included dose of estrogen, age, weight, smoking, alcohol intake, exercise and diet history. Transsexual males on estrogens had significantly higher mean (+/- SE) HDL cholesterol levels (69.0 +/- 7.1 mg/dl) respectively, for castrated males and (53.8 +/- 6.2 mg/dl) for non-castrated males, respectively compared to normal control males not on hormonal therapy (41.5 +/- 5.4) (p < 0.001), regardless of dose of estrogen received. The total cholesterol/HDL ratio was 3.31-4.05 in transsexual males on estrogens compared to 5.03 for normal males (p < 0.001). Transsexual males had mean SHBG levels in the female range (63.4 to 71.8 nmol/ml), significantly higher than controls (26.7 nmol/ml) (p < 0.001). SHBG levels were correlated with estrogen use, dose and HDL cholesterol levels. We conclude that exogenous estrogens administered to transsexual males results in a female pattern of lipid/ lipoprotein cholesterol and SHBG concentration. The decreased total cholesterol/HDL ratio may imply a lower atherogenic potential and a lessened cardiovascular risk in males who take estrogens.

(J. Endocrinol. Invest. 12:449-454, 1989)

The well known sex difference in plasma lipoprotein profiles and in cardiovascular disease of atherosclerotic origin has been attributed to the higher mean circulating estrogen levels of premenopausal women. The widespread use of female hormones has prompted several studies of plasma lipid and lipoprotein levels in patients on exogenous estrogen therapy, particularly females on oral contraceptives. Administration of estrogen to men for carcinoma of the prostate resulted in an increase in mortality from cardiovascular disease. Other studies, however, have suggested that administration of estrogens to men may lower the incidence of cardiovascular disease.

The risk of developing coronary heart disease is inversely related to the level of circulating high density lipoprotein (HDL) cholesterol. Lowered HDL levels, such as in obesity and in males, are associated with a higher incidence of clinical coronary artery disease. Endogenous estrogen production in the premenopausal woman is reflected in a higher serum HDL cholesterol value than in age-matched males.

Administration of estrogens to gonadally intact and to castrated men may raise HDL cholesterol and mimic the lipoprotein distribution characteristic of premenopausal women. In addition the suppression of testosterone production after castration may further augment the HDL rise caused by estrogen administration. This study examines the effects of exogenous estrogen administration on HDL cholesterol and the cholesterol/HDL ratio to determine whether estrogen produces a protective alteration in cardiovascular risk in males.

MATERIALS AND METHODS

Forty transsexuals, 20 castrated and 20 non-castrated taking an estrogen preparation for three or more years to induce female secondary sexual characteristics were studied. Controls consisted of 28 age- and weight- matched normal males on no hormonal therapy and were matched for alcohol, smoking status, exercise and dietary intake. A complete history for controlled variables and estrogen dosages was obtained, and physical examination including weight, height and blood pressure was performed. Plasma low density lipoprotein (LDL) cholesterol, HDL cholesterol, triglycerides, total cholesterol and SHBG levels were obtained in fasting subjects and repeated in all subjects one month later at the same time of day. Study subjects were taking Premarin (estrone sulfate 48%, equilin sulfate 26%, 17a-dehydroequilin sulfate 15%) in doses ranging from 1.25 mg/day to 10 mg/day. For analysis of dose related variables, patients were divided into low dose (1.25 to 2.5 mg/day) and high dose (5 to 10 mg/day) groups. One patient taking Premarin in a dose greater than 50 mg/day was eliminated due to the extreme dosage.

Plasma lipids and lipoproteins were measured using the methods of the Lipid Research Clinics Program as modified for use with enzymatic cholesterol and triglyceride analysis. After an overnight fast of at least 12 h, blood was drawn into evacuated tubes containing solid EDTA (final concentration 1.5 mg/ml blood). The samples were mixed gently by inversion, placed in an ice bath and brought promptly to the Lipoprotein Laboratory. Cells were removed within 3h and plasma was stored in sealed vials at 4 C until analyzed. Cholesterol was measured using a commercially available enzymatic method (Cholesterol CHOD-PAP method, Cat. No. 70412, Boehringer-Mannheim Diagnostics, Indianapolis, IN). Triglycerides were also analyzed enzymatically using a commercially available method (A-Gent Triglycerides Reagent Set, Cat. No. 6097, Abbott Laboratories, Irving, TX). Triglyceride blanks were measured using the same reagent but without lipase (Abbott A-Gent, Free Glycerol Reagent Set, Cat. No. 6087-03). For the analysis of HDL-cholesterol, an aliquot of plasma was treated with heparin and MnCl2 (final concentrations 1.3 mg/ml and 0.046 M, respectively) to precipitate apoB-containing lipoproteins (VLDL, LDL, HDL). After standing for 30 min at 4 C, the precipitate was sedimented by centrifugation at 1500 x g for 30 min at 4 C, and the clear supernate was recovered. An aliquot of the supernatant was treated with NaHCO3 (final concentration, 0.1 M) to precipitate excess Mn+2 free supernate which interferes with the enzymatic cholesterol assay. The precipitate was sedimented by centrifugation at 10,000 x g for 10 min in a bench top centrifuge, and an aliquot of the Mn+2 free supernate was taken for the measurement of HDL-cholesterol. The HDL cholesterol values were corrected for dilution that occurs when adding the reagents. LDL-cholesterol was calculated from the cholesterol, triglyceride, and HDL-cholesterol values using the empirical relationship of Friedewald et al: (LDL-cholesterol) = (total cholesterol) - (HDL-cholesterol) (-triglycerides/5) where all concentrations are expressed in mg/dl. During this study, the laboratory was standardized for the measurement of cholesterol, triglycerides and HDL-cholesterol according the criteria of the Centers for Disease Control - National Heart, Lung and Blood Institute Lipid Standardization Program. SHBG was measured using methodology described by Mickelson and Petra.

A multivariate analysis of variance was used to test simultaneously for differences in mean levels of each lipid parameter in control vs patient groups. Classes for the ANOVA were the categorical covariables dose, estrogen use, castration status, alcohol use and smoking. Linear regression analysis for each lipid parameter was also performed, with total estrogen dose, weight, diastolic and systolic blood pressure. Also performed were backward stepwise multiple regression procedures with each lipid parameter using total dose of estrogen in milligrams a day. Further analysis of the relationship between estrogen dose and the lipid parameters was conducted with Duncan's multiple range test statistic for pairwise comparisons. A two sample t test was conducted on all continuous variables for all binary categorical variables (estrogen, castration status, alcohol use, smoking), with a level of significance a p value of < 0.05. For verification of the ANOVA results, a non-parametric analysis was performed with the Kruskall- Wallis statistic and the Chi-square distribution with a level of significance a p value of < 0.05 .

Height, weight, Quetelet index and blood pressure showed no significant differences among the 3 study groups by multivariate analyses of variances and the two sample t-test (p = NS). The castrated group showed slightly higher mean body weight, although not significant. The major confounding variables of alcohol intake, smoking, exercise, diet and family history showed no significant differences between the castrated groups and normal males by the ANOVA (F = 1.06) and two sample t-test (p = NS) (Table 1).

Characteristics of patient group and normal men with respect to confounding variables (MEAN +/- SE)

	Castrated	Non-castrated	Normal Controls	p value
Systolic BP	121 +/- 6	119 +/- 5	120 +/- 6	NS
Diastolic BP	80 +/- 3	78 +/- 2	79 +/- 2	NS
Quetelet Index	2.7+/-0.4	2.6+/-0.4	2.6+/-0.3	NS
Smokers %	66	54	48	NS
Median ETOH	50 +/- 3	52 +/- 3	40 +/- 2	NS
g sat fat/day	13 +/- 2	8 +/- 0.8	10 +/- 1	NS
Exercise (h/wk)	1.5+/-0.5	2.5+/-0.7	3.0+/-0.9	NS

The ANOVA results were verified by the Kruskall-Wallis statistic and the Chi-square distribution (H + 4.90 < X-square = 7.27, p = NS). Males taking estrogens showed higher mean HDL-C concentration than controls by two sample t-test (p < 0.001) (Table 2).

Concentrations (mg/dl) of plasma lipids, lipoproteins, SHBG and cholesterol/HDL (nmol/ml) ratio (mean +/- SE) in Study Groups

Concentrations (mg/dl) of plasma lipids, lipoproteins, and SHBG (mean +/- SE) in Estrogen Dose Groups

The significant difference among the dose classes was between those on estrogen therapy (p < 0.01) (Fig. 1), and not correlated with absolute mg dose of estrogen received, as determined by backward stepwise multiple regression (r + .174, r-square = 0.03) and confirmed by Duncan's multiple range test for pair-wise comparisons. Mean total cholesterol values were significantly higher in the estrogen groups, with castrated individuals having higher mean levels than non-castrated by two-sample t-test (p < 0.01). Triglycerides and LDL cholesterol were similar in all three groups of subjects, with the highest value in control subjects, but not significant after adjustment for body weight. Mean SHBG levels were elevated significantly in all patients taking estrogens by two sample t test (p < 0.01) and were correlated with the dose of estrogen administered according to backward stepwise regression analysis (r = 0.836, r-square = 0.699) (Table 2).

No previous studies have addressed the effects of sex hormone use on lipid/ lipoprotein cholesterol and cardiovascular risk in transsexual men. This study shows that estrogen treatment resulted in a significant elevation of HDL cholesterol levels and that the effect appeared to be augmented by castration. Transsexualism has thus allowed a model of study of hormone specific preparations and metabolic effects in the gonadally intact and open loop hypothalamic-pituitary system in the case of the castrated individual.

Differences in lipid and lipoprotein concentrations can be attributed to body composition, age, exercise or the other population factors. When these variations were controlled in this study, the finding of a female pattern of lipids and lipoproteins in transsexual males appeared to be associated with the estrogenic hormones administered. The levels of HDL cholesterol in males using equine estrogen estrogens were significantly higher than in non-users. This cross-sectional observation is consisted with the serial observation in this study of a 66.3% increase in HDL cholesterol in castrated and a 29.8% increase in HDL cholesterol in non-castrated men treated with estrogen preparations. These results are consistent with the altered HDL cholesterol levels observed with administration of equine estrogens to women.

The observation that an elevation in HDL cholesterol is associated with estrogen administration and not confounding variables is further supported by the parallel rise in SHBG observed in all study patients on estrogen. SHBG levels reflect tissue sensitivity and the impact of exposure to sex hormones for a period of days or weeks. The association of HDL cholesterol with androgen/estrogen ratio changes may actually reflect relationships of HDL with SHBG, since SHBG has been found to be positively correlated with HDL cholesterol (21). In this study the increase in SHBG appear to be correlated with estrogen dose indicating a dose-response effect with this hormone dependent parameter. In the one subject receiving over 50 mg Premarin/day, SHBG levels were markedly elevated (80.0 mg/dl) as was HDL cholesterol (91.6 mg/dl).

The effect of estrogen administration on HDL cholesterol is present regardless of dose received, and implies an alteration in circulating androgen/estrogen ratios rather than a dose and response effect. In males and in females testosterone or more specifically, the androgen/estrogen ratio may regulate HDL production and composition. An androgenic environment may favor a decrease in HDL cholesterol and may be modulated by exogenous estrogen, as shown in this study. Suppression of endogenous testosterone has also been associated with an increase in HDL cholesterol. Estrogen administration has produced a significant inhibition of plasma and testicular levels of testosterone in the in vitro perfused testis as well as in the intact male. The effect of castration on the HDL parameters in these transsexual males may account for the higher levels of HDL cholesterol compared to the gonadally intact individuals who have not had the endogenous source of testosterone removed. Although the castrated individuals as a group were slightly heavier than the non-castrated or control males, castration seemed to override the potential negative effect of weight.

The cardiovascular consequences of the female pattern of altered lipoprotein concentrations of transsexual males is unknown. The total cholesterol to HDL cholesterol ratio may be predictive as to the development of coronary artery disease. The calculated risks for cardiovascular disease in the transsexual groups of males studies was 3.49 and 4.05, below the range of standard risks for men (5.0) and slightly lower than the standard risk for women (4.44).

In conclusion, a female pattern of HDL cholesterol results from exogenous estrogen administration to transsexual males. The clinical impact of these altered lipoprotein parameters and the exact role of altered androgens in gonadally intact males on estrogen requires an evaluation of a large number of subjects. The impact of a lowered cholesterol/HDL ratio as observed in these men remains to be evaluated longitudinally.

Becky's Home Page