Systemic Delivery of Estradiol, but not Testosterone or Progesterone, Alters Very Low Density Lipoprotein-Triglyceride Kinetics in Postmenopausal Women

Abstract

Context:

Sexual dimorphism in plasma triglyceride (TG) metabolism is well established but it is unclear to what extent it is driven by differences in the sex hormone milieu. Results from previous studies evaluating the effects of sex steroids on plasma TG homeostasis are inconclusive because they relied on orally administered synthetic hormone preparations or evaluated only plasma lipid concentrations but not kinetics.

Objective:

The purpose of this study was to evaluate the effects of systemically delivered 17β-estradiol, progesterone, and T on very low density lipoprotein-triglyceride (VLDL-TG) concentration and kinetics in postmenopausal women.

Setting and Design:

VLDL-TG concentration and kinetics were evaluated by using stable isotope–labeled tracer techniques in four groups of postmenopausal women (n = 27 total) who were studied before and after treatment with either 17β-estradiol (0.1 mg/d via continuous delivery skin patch), progesterone (100 mg/d via vaginal insert) and T (12.5 mg/d via skin gel), or no intervention (control group).

Results:

VLDL-TG concentration and kinetics were unchanged in the control group and not altered by T and progesterone administration. Estradiol treatment, in contrast, reduced VLDL-TG concentration by approximately 30% due to accelerated VLDL-TG plasma clearance (25.1 ± 2.5 vs. 17.4 ± 2.7 mL/min; P < .01).

Conclusions:

Estradiol, but not progesterone or T, is a major regulator of VLDL-TG metabolism.

Sexual dimorphism in the plasma lipid profile is well established (1). Most characteristically, men have higher serum total and very low density lipoprotein-triglyceride (VLDL-TG) and lower high-density lipoprotein cholesterol concentrations than premenopausal women. The mechanisms responsible for these differences are not entirely clear but differences in the sex hormone milieu are considered prime suspects. However, studies evaluating the effect of sex steroids on lipid metabolism have provided conflicting results (1). The discrepancy between sex-hormone intervention studies and naturally occurring sexual dimorphism is probably due to the use of synthetic hormone derivatives, nonspecific drugs used to stimulate endogenous production and/or the oral route of administration of sex-hormone preparations in most previous studies. Several investigators have shown that synthetic estrogen derivatives given orally (2, 3) or ovarian hyperstimulation by administering recombinant human follicle stimulation hormone (4) raise plasma TG concentration while systemically delivered estradiol (via skin patches) has no effect on plasma TG concentration (2) or reduces (3) it. Oral coadministration of progestins attenuates or even reverses the estrogen preparation–induced increases in plasma TG concentration (5), and progestins administered (orally) alone decreases plasma TG concentrations (6, 7). The mechanism(s) responsible for the potential hypotriglyceridemic effect of systemically delivered estradiol have not been evaluated and the effect of systemically delivered progesterone on plasma TG concentration and kinetics is unknown.

The effect of T on plasma TG metabolism is also unclear. In men, T delivered orally or systemically has no effect on TG concentration (1, 8–13), possibly because of the high baseline availability of T. In women, the results are equivocal. We have shown that transdermally delivered T has no effect on VLDL-TG kinetics and plasma TG concentration in premenopausal women (14) whereas different T preparations given orally have been found to decrease, increase, or have no effect on plasma TG concentration in postmenopausal women (15, 16).

The purpose of the present study was to evaluate the effects of systemically delivered 17β-estradiol, progesterone and T on VLDL-TG concentration and kinetics. In addition, we measured free-fatty acid (FFA) rate of appearance (Ra) in plasma (a major source of fatty acids for VLDL-TG synthesis) (14, 17) and glucose Ra, an index of hepatic glucose production, to provide a measure of hepatic insulin sensitivity, an important regulator of hepatic lipoprotein metabolism (18).

Materials and Methods

Subjects and prestudy testing

Twenty seven postmenopausal, sedentary (<1.5 h of exercise/wk), and weight-stable (<2 kg change for ≥6 mo) women participated in this study. None of the women in the study were currently receiving or had received hormone replacement therapy within 1 y of their first study visit. One subject in the T group, one in the estradiol group, and one in the progesterone group had received hormone replacement therapy before that; all others had never received hormone replacement therapy. Written informed consent was obtained from all subjects before participation in the study, which was approved by the Institutional Review Board of Washington University School of Medicine in St. Louis, MO. All subjects were considered to be in good health after completing a comprehensive medical examination, including a detailed history and physical examination, a resting electrocardiogram, standard blood tests, and an oral glucose tolerance test. None of the subjects were taking medications (including hormone replacement therapy) that could interfere with the outcome measures, and none reported excessive alcohol intake or consumed tobacco products.

The lipid kinetic data presented here were obtained simultaneously with muscle protein kinetic data, which we recently reported in the Journal of Clinical Endocrinology and Metabolism (19); three control subjects completed the lipid kinetic study only and were added to this cohort. Subject characteristics are presented in Supplemental Table 1 and the treatment-induced changes in plasma hormone concentrations are reproduced from ref (19) in Supplemental Table 2 to this brief report.

Experimental design

All subjects completed two identical stable isotope–labeled tracer infusion studies each: one before and one after either T (12.5 mg/d of T transdermally for 21 d; i.e. 1.25 g of Androgel 1% per d, Abbvie Inc.), 17β-estradiol (0.1 mg/d by continuous delivery patches in three 14-d on/14-d off cycles; Mylan Pharmaceuticals Inc.), or micronized progesterone (100 mg/d in three 14-d on/14-d off cycles by using a vaginal insert; Endometrin, Ferring Pharmaceuticals Inc.) treatment or a no-intervention control period (see (19) and Supplemental Methods and Materials for details). In the late afternoon on the day before the tracer infusion study, subjects were admitted to the Clinical Research Unit where they consumed a standard dinner between 1800–1900 h, and then fasted, except for water, until the next morning.

At 0600 h, a catheter was inserted into an arm vein for the infusion of stable isotope–labeled tracers, and a second catheter was inserted into a vein of the contralateral hand, which was warmed to 55 C by using a thermostatically controlled box to obtain arterialized blood samples. The sampling catheter was kept open with a slow, controlled infusion of 0.9% NaCl solution (30 mL/h). At 0700 h (time = 0), a bolus of [1,1,2,3,3-²H₅]glycerol (75 μmol/kg body weight), dissolved in 0.9% NaCl solution, was administered and continuous infusions of [6,6-²H₂]glucose (0.22 μmol/kg body weight/min; priming dose, 18 μmol/kg body weight) dissolved in 0.9% NaCl solution and [2,2-²H₂]palmitate (0.03 μmol/kg body weight/min) dissolved in 25% albumin solution, were started and maintained for 5 and 12 h, respectively.

Blood samples were obtained immediately before administering the tracers, at 5, 15, 30, 60, 90, 120, 150, 180, 210, 240, 270 and 300 min after starting the tracer infusions, and then hourly for another 7 h to determine A) plasma insulin, glucose, FFA, and VLDL-TG concentrations; B) glycerol, palmitate, and glucose tracer-to-tracee ratios (TTRs) in plasma; and C) glycerol and palmitate TTR in VLDL-TG.

Sample collection, processing, and analyses

Blood samples to determine plasma glucose, insulin, FFA, and VLDL-TG concentrations were collected, prepared, and analyzed as previously described (14, 17, 19). The TTR of plasma free glycerol, glucose and palmitate, and the TTRs of glycerol and palmitate in VLDL-TG were determined by using electron-impact gas-chromatography mass-spectrometry (MSD 5973 system, Agilent) (14, 17).

Calculations

Palmitate and glucose Ra in plasma were calculated by dividing the tracer infusion rate by the respective average plasma TTR between 180 and 300 min (steady state); FFA Ra was calculated based on the proportional contribution of palmitate to total plasma FFA concentration (20). The hepatic insulin sensitivity index (HISI) was calculated as the inverse of the product of glucose Ra (in μmol/min) and plasma insulin concentration (in mU/L) (21). VLDL-TG kinetics (VLDL-TG secretion rate, plasma VLDL-TG clearance rate, and the relative contribution of systemic plasma FFA to the VLDL-TG fatty acid pool) was determined by fitting the TTR time courses of glycerol and palmitate in plasma and glycerol and palmitate in VLDL-TG to a compartmental model (14, 22).

Statistical analysis

Data are presented as means ± SEM or median (quartiles) for normally distributed and skewed data sets, respectively. Repeated measures ANOVA and Tukey's post hoc procedure were used to evaluate the significance of hormone treatment–induced changes in plasma substrate and hormone concentrations and VLDL-TG kinetics. Skewed data sets were log transformed prior to performing ANOVA. Significant within group differences were confirmed by analysis of covariance (ANCOVA) using the post-treatment value as the dependent variable and the pretreatment value as the covariate to determine whether the change in the treatment group was significantly different from the change in the control group. A P value ≤.05 was considered statistically significant.

Results

Plasma sex hormone concentrations were unchanged in the control group and plasma estradiol, progesterone and T concentration were increased after treatment in the estradiol, progesterone, and T groups, respectively (Supplemental Table 2). None of the treatments affected the sex hormone–binding globulin concentration or resulted in nonspecific sex hormone concentration differences (Supplemental Table 2).

Plasma insulin, glucose, FFA, and cholesterol concentrations, glucose and FFA Ra, and the HISI were all unchanged in the control group and were not affected by any of the treatments (Table 1 and Supplemental Table 2). VLDL-TG concentration and kinetics were unchanged in the control group and not altered by T and progesterone administration. Estradiol treatment, in contrast, reduced VLDL-TG concentration by approximately 30% due to accelerated VLDL-TG plasma clearance from the circulation (Table 1) and this response was significantly different from that in the control group (P < .05). The contribution of systemic FFA sources to total VLDL-TG production was not different between groups at the beginning of the study and remained unchanged in the control and all treatment groups (grand mean: 64 ± 3% in all Before and 62 ± 3% in all After studies; ANOVA group, time, and interaction P ≥ .24).

Table 1.

Effect of Sex Steroids on Plasma Insulin and Substrate Concentrations and Kinetics

	Control (n = 9)		Testosterone (n = 6)		Estradiol (n = 6)		Progesterone (n = 6)
	Before	After	Before	After	Before	After	Before	After
Insuline
Concentration,a mU/L	4.5 (2.7, 4.7)	4.2 (3.0, 6.5)	2.8 (1.6, 4.2)	3.2 (2.1, 3.9)	4.0 (3.0, 5.6)	3.3 (2.7, 4.5)	3.2 (2.7, 3.6)	4.0 (2.8, 4.9)
Concentration,a pM	31 (18, 33)	29 (21, 45)	19 (11, 29)	22 (15, 27)	28 (21, 39)	23 (19, 31)	22 (19, 25)	27 (19, 34)
Glucosee
Concentration,b mg/dL	96 ± 3	94 ± 2	92 ± 2	91 ± 1	96 ± 4	95 ± 3	93 ± 2	95 ± 2
Concentration,b mM	5.36 ± 0.14	5.23 ± 0.13	5.11 ± 0.11	5.07 ± 0.08	5.33 ± 0.24	5.28 ± 0.15	5.17 ± 0.12	5.27 ± 0.11
Ra,b μmol/min	719 ± 30	695 ± 28	692 ± 49	679 ± 40	739 ± 48	692 ± 41	696 ± 15	669 ± 17
HISIb	0.36 ± 0.06	0.35 ± 0.06	0.68 ± 0.18	0.61 ± 0.16	0.36 ± 0.08	0.42 ± 0.07	0.47 ± 0.09	0.44 ± 0.09
FFA
Concentration,b mg/dL	13.5 ± 1.0	13.4 ± 1.3	17.4 ± 1.9	18.3 ± 1.4	15.8 ± 1.1	15.3 ± 1.3	14.9 ± 0.8	15.8 ± 0.9
Concentration,b mM	0.48 ± 0.04	0.48 ± 0.05	0.62 ± 0.07	0.65 ± 0.05	0.56 ± 0.04	0.54 ± 0.04	0.53 ± 0.03	0.56 ± 0.03
Ra,b μmol/min	341 ± 42	347 ± 47	316 ± 13	339 ± 18	369 ± 34	337 ± 26	319 ± 11	320 ± 26
VLDL-TG
Concentration,b mg/dL	35 ± 7	34 ± 6	16 ± 3	16 ± 4	44 ± 10	31 ± 8cd	38 ± 6	39 ± 13
Concentration,b μM	394 ± 78	390 ± 71	175 ± 39	185 ± 40	502 ± 112	352 ± 90cd	424 ± 72	436 ± 147
Secretion rate,a μmol/L/min	3.24 (2.36, 4.08)	3.89 (2.37, 4.25)	1.91 (1.84, 2.24)	1.97 (1.67, 2.77)	2.79 (2.35, 3.39)	2.85 (2.16, 3.07)	2.84 (2.40, 3.70)	3.09 (2.25, 3.84)
Clearance rate,b mL/min	24.9 ± 4.3	23.8 ± 1.7	35.1 ± 6.7	34.5 ± 6.4	17.4 ± 2.7	25.1 ± 2.5cd	21.2 ± 3.0	24.9 ± 5.5

^aData expressed as median (quartiles) for skewed data sets.

^bData are expressed as mean ± SEM for normally distributed data sets.

^cValue significantly different from corresponding Before value (P < .01).

^dValue significantly different from control group value after accounting for differences in Before values by using analysis of covariance (P < .05).

^ePlasma glucose and insulin concentrations in the three treatment groups are reproduced from (19).

Discussion

The purpose of the present study was to evaluate the effects of sex hormones on VLDL-TG concentration and kinetics to evaluate their potential role in mediating the difference in plasma TG concentration between men and women (women < men). We found that 17β-estradiol lowers VLDL-TG concentration in plasma and demonstrates that this is due to increased plasma VLDL-TG clearance whereas VLDL-TG secretion rate is unaffected by estradiol. We also found that T and progesterone have no effect on VLDL-TG concentration or kinetics and therefore do not seem to contribute to the sexual dimorphism in plasma TG metabolism.

The hypotriglyceridemic effect of systemically delivered estradiol in our study is consistent with observations made by other investigators (1, 3) and our kinetic data now provides a mechanistic explanation for this phenomenon. The lack of effect of T on VLDL-TG concentration and kinetics in the present study is also consistent with the results from studies by other investigators who report no effect of T on plasma TG concentration or kinetics in men (8, 9, 11–13) and confirms our previous findings in premenopausal women (14). Thus it is unlikely that failure of T treatment to affect plasma TG concentration was due to a “ceiling effect” in men or that female sex hormones interfere with or oppose any putative T-induced effect on plasma TG metabolism. As far as we know, ours is the first study to evaluate the effect of systemically delivered progesterone on plasma TG homeostasis and our results suggest that it is not involved in regulating VLDL-TG metabolism. These findings are consistent with the results from previous work in our laboratory demonstrating that a 40-fold difference in plasma progesterone concentration between the late follicular and luteal phases of the menstrual cycle in premenopausal women had no effect on VLDL-TG concentration or kinetics (23). It is possible but unlikely that the current study was underpowered to detect an effect of progesterone and/or T on VLDL-TG metabolism. The change in VLDL-TG concentration in the progesterone and T arms in our study were so small (∼3–5% increase) that it would have required >7,000 and >700 subjects respectively to detect it with a sufficiently small probability of Type I error (α < 0.05) and sufficient power (≥0.80) to reject the null hypothesis.

The hypertriglyceridemic effect of estrogen preparations (2) and the hypotriglyceridemic effect of orally administered progesterone preparations (6, 7) in postmenopausal women are therefore most likely nonspecific consequences of either the oral route of delivery and a hepatic “first pass” effect or the modifications necessary to make the hormones fit for oral consumption rather than hormone effects themselves. Nevertheless, it is possible that the differences in outcomes are simply artifacts due to differences in study design and/or subject characteristics.

The stimulatory effect of estradiol on VLDL-TG plasma clearance helps explain the lower plasma TG concentration in women compared with men (1). However, it is unlikely that estradiol availability is the sole factor involved in the sexually dimorphic plasma TG metabolism phenotype because it only partially explains the differences in VLDL-TG kinetics between men and women (24), most likely due to a complex interplay of multiple regulatory factors (including insulin sensitivity, fatty acid availability from visceral and hepatic fat stores, energy expenditure, and fatty acid channeling toward oxidative and nonoxidative pathways (25–30)) that ultimately affect VLDL-TG homeostasis. The differential response of VLDL-TG metabolism in men and women to metabolic stimuli such as hyperglycemia and insulinemia or fructose ingestion and obesity (17, 31, 32) support this concept.

In summary, T and progesterone have no effect on VLDL-TG concentration and therefore do not seem to contribute to the sexual dimorphism in plasma TG metabolism whereas estradiol lowers VLDL-TG concentration by increasing VLDL-TG plasma clearance.