Archives of Gynecology and Obstetrics

, Volume 274, Issue 5, pp 289–296

The effect of hormone replacement therapy on the levels of serum lipids, apolipoprotein AI, apolipoprotein B and lipoprotein (a) in Turkish postmenopausal women

Authors

    • Biochemistry Dept., Faculty of MedicineHacettepe University
  • Derya (Akaydın) Aldemir
    • Biochemistry Dept., Faculty of MedicineBaşkent University
  • Tülin Bayrak
    • Biochemistry Dept., Faculty of MedicineHacettepe University
  • Aydın Çorakçı
    • Obstetrics and Gynecology Dept., Faculty of MedicineKocaeli University
    • Obstetrics and Gynecology Dept., Faculty of MedicineHacettepe University
Original Article

DOI: 10.1007/s00404-006-0187-2

Cite this article as:
Bayrak, A., Aldemir, D.(., Bayrak, T. et al. Arch Gynecol Obstet (2006) 274: 289. doi:10.1007/s00404-006-0187-2

Abstract

Objectives

Estrogen replacement therapy alters the lipid profiles favorably for delaying atherosclerosis in postmenopausal women. The effects of estrogen plus progesterone combination therapy on lipids are controversial. This study was designed to evaluate the effect of female sex hormones on lipids and lipoproteins and to clarify the influence of progesterone on the effect of estrogen in postmenopausal women.

Methods

Of the 60 postmenopausal women admitted to our menopause clinic, 40 had intact uterus and received continuous 0.625 mg conjugated equine estrogen (CEE) plus 2.5 mg medroxyprogesterone acetate (MPA), whereas the remaining 20 were hysterectomized and received 0.625 mg CEE daily. To assess the alterations in lipids and lipoproteins during menopause, 45 healthy premenopausal women were investigated. Lipid and lipoprotein levels were assessed in each subject at baseline and at the 6th and 18th months of therapy.

Results

In menopause, a shift towards more atherogenic lipid and lipoprotein profiles than those of the premenopausal state was found. Following 18 months of treatment, both regimens reduced total cholesterol (TC) levels as compared with the baseline (6.4 vs. 6.9% in the CEE/MPA and CEE groups, respectively). The CEE group had a more pronounced increase in high-density lipoprotein (HDL) cholesterol than the CEE/MPA group (10.3 vs. 8.8%, respectively). Both groups displayed reduced TC, low-density lipoprotein (LDL) cholesterol and apolipoprotein-B (ApoB) concentrations, whereas triglycerides increased, with a greater tendency to increase in the CEE/MPA group at the end of the trial. Also, the lipoprotein (a) [Lp(a)] levels decreased significantly (27.6 vs. 24.5% in the CEE/MPA and CEE groups, respectively). This decrease was more pronounced in subjects with a relatively higher basal Lp(a) concentration.

Conclusion

Both treatment regimens caused positive alterations in the lipid and lipoprotein profiles. This association might play a pivotal role in the postmenopausal increases in atherosclerotic diseases and cardioprotective effect of estrogen in postmenopausal women.

Keywords

ERTHRTLipidsLipoproteinsAtherosclerotic diseasesCoronary heart diseaseCardioprotective effect of estrogen

Introduction

The lower cardiovascular disease (CVD) rates in women as compared to men may well be due to genetic differences, presence of estrogen, absence of testosterone, or a combination of all of these factors. While women appear to be protected against CVD until menopause, the cessation of hormone synthesis in menopause is accompanied by increased atherosclerotic lesion formation rates and vascular dysfunction incidences. These changes are largely responsible for increased CVD-related morbidity and mortality [1].

The data obtained in the previous studies have suggested that postmenopausal women receiving estrogen supplements may be protected against CVD. However, the mechanism of this effect is not fully understood [2]. Although non-lipid factors such as the action on the vascular wall and endothelium and the synthesis of vasoconstrictor and vasodilator agents have been held responsible, the significance of lipids and lipoproteins cannot be ignored in the pathogenesis of CVD [3]. Favorable alterations in these markers may account for a considerable part of the cardioprotection; however, physiological data such as improved lipid profile or endothelial function assessment values following estrogen treatment have suggested potentially beneficial mechanisms [4]. Several debates have been initiated particularly in the recent years regarding the administration of sex hormone supplements to reduce the risk of CVD in postmenopausal women. According to the results of the large-scale randomized study known as Women’s Health Initiative (WHI) conducted in order to define the risks and benefits of hormone treatment, daily combined estrogen plus progestin treatment regimen not only slightly increases the breast cancer risk, but also causes a more striking increase in the coronary heart diseases (CHD) [5, 6]. However, the information accumulated in the literature throughout several years displays that estrogen influences numerous factors responsible for the initiation and progression of CVDs through multiple logical mechanisms [1, 7]; the fact that hormone replacement therapy (HRT) causes positive alterations in the lipid profiles contradicts with the results of WHI, which denotes that HRT does not lead to a reduction in coronary events. However, considering the influences of estrogen on blood vessels and other contexts, it can be stated that the results of WHI shall be subjected to further debate. In WHI, the mean age of cases included in the study is 63, which is basically much higher than that of the target population receiving HRT. Besides, the selection of an elderly population for a study evaluating primarily the relationship between HRT and CHD in patients who apply menopause clinics has been found controversial [8, 9].

The beneficial effects of estrogen on lipid profiles are thought to be reduced by progestins, which are added in order to reduce endometrial hyperplasia and the risk of cancer [10]. In earlier studies of lipid metabolism, it has been shown that both the dosage and type of progestogen are of importance for the lipoprotein, cholesterol and TG fractions [11]. The progestins used in HRT [particularly medroxyprogesteron acetate (MPA), administered either cyclically (10 mg/day for 12 days per month) or continuously (2.5 mg/day)] have been proposed to counter the increase in HDL cholesterol [12]. Some other investigators have reported that the positive influences of estrogen on lipid profiles may be maintained in HRT with appropriate type, dose and way of administration of progestins [13, 14].

The CEE/MPA regimen is the most commonly prescribed HRT. In one study, it has been shown that MPA almost had no counteracting effects on serum lipid changes induced by conjugated estrogen CEE alone [15]. In a study by Lobo et al. [15], the CEE/MPA regimen increased HDL and TG, and decreased total cholesterol (TC), LDL and serum Lp(a). Lp(a) is a modified LDL particle and is associated with increased cardiac event risk. With its homology and interactions with plasminogen, Lp(a) can increase the risk of thrombotic complications, and high concentrations of Lp(a) are thought to be independent risk factors for CVD [16]. Unlike other lipids and lipoproteins, Lp(a) is not affected by blood lipid-lowering agents other than niacin [17]. In recent epidemiological investigations, it has been reported that the Lp(a) concentrations in postmenopausal women are higher than those in the premenopausal state, and the postmenopausal women receiving sex hormone supplements have less Lp(a) levels than those without the treatment [18, 19].

The effect of the combination therapy of estrogen and progesterone on lipids has been evaluated by many authors; however, the results were inconsistent [20, 21]. Thus, the purposes of this study were to evaluate the effects of estrogen and progesterone on the lipid and lipoprotein concentrations and to clarify the influence of progesterone on lipid levels when administered with estrogen to postmenopausal women.

Materials and methods

Participants and study design

This study was a prospective case-controlled trial. The participants were 60 postmenopausal women aged 47–55 years who initially enrolled in the study from our Menopause Clinic. Forty-five premenopausal women aged 27–35 years were enrolled in the control group in order to define the menopause-related alterations in the lipid and lipoprotein profiles. The definition of menopause is amonerrhea for more than 12 months, or for more than 6 months plus an elevated serum follicle stimulation hormone (FSH) level greater than 40 IU/l and an estradiol level less than 74 pmol/l (20 pg/ml). Postmenopausal women with a prior history of major cardiovascular, cerebrovascular, thrombotic, renal, hepatic, cholecystic or active mammary diseases, known endometrial hyperplasia or endometrial cancer were excluded from the study. None of the participants had received estrogens, progestins, androgens, aspirin, warfarin, or other anticoagulant treatment prior to the trial according to patients’ statements and their medical records. Each participant underwent bimanual examination, PAP smear, transvaginal sonography, breast examination, mammography, thyroid, renal and liver function tests and blood coagulation tests prior to the initiation of study. Records of participants’ daily activities and eating habits were not recorded; however, according to their own statements they followed an ordinary diet and normal activity. In each participant, baseline biochemical assessments were conducted after an informed consent was signed. These patients were divided into two groups; those with intact uteruses received continuous 0.625 mg CEE plus 2.5 mg MPA (Premelle, Wyeth İlaç, Istanbul, Turkey) (n = 40) and those that had undergone hysterectomy received 0.625 mg CEE (Premarin, Wyeth İlaç, Istanbul, Turkey) (n = 20). The study period was 18 months. Follow-up evaluation was performed at the 6th and 18th months (Fig. 1). Blood pressure, weight and height were recorded in the morning in light clothing, and the body mass index (BMI) was computed at each visit, including the initial visit of the study. Fasting blood samples were drawn at 0900 hours for lipid and lipoprotein determination. Samples were immediately centrifuged and serum was stored at −30°C until assayed.
https://static-content.springer.com/image/art%3A10.1007%2Fs00404-006-0187-2/MediaObjects/404_2006_187_Fig1_HTML.gif
Fig. 1

Mean ± SD for changes from baseline of serum lipid and lipoprotein fractions 6 and 18 months following the initiation of treatment in groups receiving (a) 0.625 mg/dl CEE only (n = 20) and (b) 0.625 mg CEE/2.5 mg MPA (n = 20). Cholesterol total cholesterol (mg/dl); LDL low-density lipoprotein (mg/dl); HDL high-density lipoprotein (mg/dl) and Lp(a) lipoprotein(a)

Blood biochemical measurements

Fasting serum Lp(a), ApoA-I, ApoB, TC, LDL-C, HDL cholesterol and triglyceride (TG) levels were assessed at baseline, and at the 6th and 18th months of treatment. Serum TC and TG were assessed by enzymatic methods. HDL cholesterol was measured enzymatically after the precipitation of LDL cholesterol, and VLDL cholesterol with Mg24-dextra. LDL cholesterol was estimated as described by Friedewald (LDL cholesterol = TC − TG/5 − HDL cholesterol). The ApoAI and ApoB levels in plasma were determined by immunoturbidimetric assay in a nephelometer (Behring-Nephelometer-100 analyzer, Frankfurt-Germany). Lp(a) concentrations were measured with a micro-ELISA system using Biopool tint ELISA kits.

Statistical analyses

The statistical analyses were performed using SPSS Version 11.5 (Statistical Package for the Social Sciences, Chicago, IL, USA). For the parameters with normal distribution, the t test was performed, whereas ANOVA for repeated measures and the Friedman test were performed for the parameters without normal distribution. The baseline characteristics were compared between therapy groups by the independent samples t test. Baseline and follow-up lipid and lipoprotein mean levels were compared across the same therapy group using ANOVA for repeated measures and the Friedman test. Following calculation of the percent changes according to the initial values, the independent samples t test and Mann–Whitney U test were used for the comparison of two groups. Statistical significance was set at 0.05.

Results

Mean levels of the baseline demographic parameters are presented in Table 1. No differences regarding age, duration of menopause, BMI, estradiol levels, current smoking, alcohol consumption, level of physical exercise and family history of CVD were found between the groups.
Table 1

Baseline demographic characteristics of the patients

 

Premenopause (n = 45)

Postmenopausal (n = 60)

CEE (n = 20)

CEE/MPA (n = 40)

Continuous variables (mean, SD)

Age (years)

32.3 ± 4.3

50.9 ± 2.2

51.2 ± 2.5

Duration of menopause (years)

3.5 ± 0.6

3.8 ± 1.2

BMI (kg/m2)

21.72 ± 2.1

25.75 ± 3.6

25.5 ± 3.1

Nominal variables (%)

Current smoking

20

20

25

Alcohol consumption

5

0

2.5

Physical exercise (<3 h/week)

20

15

17

Family history of cardiovascular disease

20

20

25

CEE Conjugated equine estrogen, MPA medroxyprogesterone acetate

In menopause, a shift towards more atherogenic lipid and lipoprotein profiles than those of the premenopausal state was found. Comparison of the pre- and postmenopausal women revealed statistically significant increases in the TC, LDL cholesterol, TG, ApoB and Lp(a) and decreases in the HDL cholesterol and ApoAI levels. The lipid and lipoprotein profile of the premenopausal women is displayed in Table 2. Baseline and follow-up unadjusted mean lipid and lipoprotein levels are presented in Table 3. Following both treatment regimens, the TC levels were reduced (6.9 vs. 6.4% in CEE and CEE/MPA groups, respectively). Likewise, the TG levels increased in both groups (4.4 vs. 3.4% in CEE and CEE/MPA groups, respectively). The LDL cholesterol levels decreased in both groups (15.4 vs. 15.6% in CEE and CEE/MPA groups, respectively). The HDL cholesterol and ApoAI levels increased in the CEE and CEE/MPA groups (HDL cholesterol; 10.3 vs. 8.8%, respectively, and ApoAI 9.9 vs. 7.9%, respectively). The levels of Lp(a) and ApoB decreased in the CEE and CEE/MPA groups [Lp(a); 27.6 vs. 24.5%, respectively, and ApoB 4.3 vs. 7.28%, respectively].
Table 2

The lipid profiles and apolipoprotein levels (mean, SD) of women in the premenopausal group

Total cholesterol (mg/dl)

189.47 ± 23.64

Triglycerides (mg/dl)

115.67 ± 46.96

Low-density lipoprotein cholesterol (mg/dl)

122.02 ± 25.35

High-density lipoprotein cholesterol (mg/dl)

44.31 ± 5.68

Apolipoprotein AI (mg/dl)

154.53 ± 14.58

Apolipoprotein B (mg/dl)

142.53 ± 24.46

Lipoprotein (a) (mg/dl)

20.54 ± 8.14

Table 3

Baseline and follow-up lipid profiles and apolipoprotein levels (mean, SD) of the women receiving therapy regimens

 

Month

CEE

CEE/MPA

P value*

Total cholesterol (mg/dl)

0

233.60

218.050

<0.001

6

219.55

205.08

 

18

217.15

203.78

 

P value

P<0.001

P<0.001

 

Triglycerides (mg/dl)

0

142.950

146.550

0.538

6

148.200

152.600

 

12

148.950

151.075

 

P value

P=0.006

P<0.001

 

Low-density lipoprotein cholesterol (mg/dl)

0

167.05

152.30

<0.001

6

149.35

135.38

 

18

145.65

132.40

 

P value

P<0.001

P<0.001

 

High-density lipoprotein cholesterol (mg/dl)

0

37.80

36.28

0.063

6

40.65

39.30

 

18

41.75

41.05

 

P value**

P<0.001

P<0.001

 

Apolipoprotein AI (mg/dl)

0

136.55

137.975

0.731

6

144.10

144.700

 

18

150.30

148.775

 

P value**

P<0.001

P<0.001

 

Apolipoprotein B (mg/dl)

0

144.450

147.425

0.047

6

139.650

140.975

 

18

137.750

138.075

 

P value**

P<0.001

P<0.001

 

Lipoprotein (a) (mg/dl)

0

39.950

28.620

<0.001

6

25.659

22.585

 

18

20.023

18.017

 

P value

P<0.001

P<0.001

 

CEE Conjugated equine estrogen, MPA medroxyprogesterone acetate

*Univariate analysis of variance (independent-samples t test) for baseline differences between groups

**ANOVA for repeated measures and Friedman

Discussion

Throughout their premenopausal periods, women are protected from CVD largely by ovarian hormones. The risk of CVD increases rapidly by menopause due to the alterations in sex steroid metabolism [22]. With menopause, a decrease in HDL cholesterol levels while increases in total and LDL cholesterol levels are observed, and these negative alterations in lipid profiles lead to an increase in the CVD risk [23]. Our findings were compatible with these previous observations. In the postmenopausal group, while the HDL cholesterol and apo AI levels were lower than those of the premenopausals, the total and LDL cholesterol, apo B and Lp(a) levels increased significantly (P<0.001 for each of the parameters). These indicate a shift to more atherogenic lipid profiles after menopause. This situation has been confirmed by longitudinal follow-ups that had demonstrated increases in LDL cholesterol and decreases in HDL cholesterol during menopause that cannot be solely attributed to aging. The increase in the incidence of CVD in the postmenopausal years also reflects this situation [1, 22, 24].

It appears that premenopausal women are protected against atherogenesis as compared with men [25], and it is this gender difference that initially led investigators to propose that estrogens may well be protective against CVD. Though the efficacy of ERT/HRT in reducing the incidence of CVD has been questioned in controlled trials (HERS-I, HERS-II, WHI), extensive data from observational and experimental studies suggest that replacement therapy administered to healthy postmenopausal women may be cardioprotective [2628]. A variety of mechanisms may be involved in this suggested cardioprotection, including effects on the vascular wall and endothelium [29] as well as favorable alterations in the lipid–lipoprotein profile [3]. Several investigators have demonstrated that women receiving ERT have significantly lower rates of developing CVD than those who do not [30].

The results of this study have shown that both ERT and HRT decrease TC. Similar results have been reported in previous studies [22, 31]. Our findings on the effects of CEE on TC and LDL cholesterol are consistent with those of the previous reports about the same regimen [3]. The addition of MPA did not significantly modify the CEE-associated decrease in TC and LDL cholesterol in the CEE/MPA group. For the same regimen, decreased or unchanged TC concentrations and variable decreases in LDL cholesterol have been reported in previous studies [32, 33]. Our results revealed that both HRT regimens and ERT exert their efficacy by reducing TC and LDL cholesterol levels mainly in women with high baseline cholesterol levels. This observation is significant in that most of the postmenopausal women included in our study had high cholesterol levels.

The HRT-related increase in TG found in this study is consistent with several [34, 35] but not all [36] studies. The inconsistency may be attributed to an increase in the hepatic secretion of TG-rich lipoproteins [37]. The implication for cardiovascular risk of the HRT-related increase in TG levels is difficult to interpret because the independent relationship of TG with heart disease is not confirmed in the epidemiologic literature [38]. The levels of TG are influenced by the type, the dose and the route of administration of estrogen [39]. CEE has been shown to increase TG significantly [28, 39], the effect being greater with higher doses [28]. In our study, CEE increased TG levels by 4.4%. The inclusion of a progestin opposes the action of estrogen on TG and the effect varies with the type and dose of progestin: as the androgenicity and dose increase, the opposing effect increases [26]. In our study, MPA did not alter the increasing effect of CEE on TG levels. Our data indicated a lesser degree of decrease than previous studies [3, 35, 40]. The discrepancy can be explained by differences in regimen, dose and duration of follow-up. In addition, the postmenopausal women in our study had serum TG levels increasingly exceeding the upper limit of normal.

Serum HDL cholesterol is inversely related to CVD risk; high levels have been associated with cardioprotection [41]. ERT significantly increases HDL and the magnitude of increase appears to be related to the type and dose of estrogen and the route of administration [28]. Higher doses induce a higher increase [28] and so does oral compared to transdermal administration [42], although not without exception [33]. In our study, HDL cholesterol levels increased in both groups as compared to the baseline; however, the increase was more pronounced in the ERT group. The increase in HDL cholesterol in both treatment groups had been previously reported, though with different magnitudes [3, 43]. The addition of progestin has been reported to have a negative effect on HDL cholesterol [44]. However, in our study MPA did not antagonize the effect of CEE on HDL cholesterol. It has been suggested that the type and dose of progestin is more important than the dose of estrogen in modulating HDL cholesterol levels, and that the lesser the androgenicity of the progestin, the lesser the attenuation of the estrogen-related increase on HDL cholesterol [26, 28].

ApoAI is an important antiatherogenic factor that upregulates HDL production. Apo AI is the major apoprotein of HDL and Apo AI synthesis rate has been shown to be higher in female monkeys than males [45]. This is most likely estrogen mediated, as there is an increase in Apo AI following oral therapy with a synthetic estrogen. Orally administered estrogens increase the hepatic synthesis of ApoAI [46]. In our study, both CEE and CEE/MPA increased ApoAI, the increase being greater with CEE monotherapy. Similar increases have been reported in previous studies [28, 47].

Replacement therapy decreases ApoB levels. The extent of decrease is analogous to that of LDL cholesterol. In our study, both treatment regimens decreased ApoB, the decrease being greater in the CEE/MPA group (P=0.009). ApoB concentrations were reported to be higher when blood was sampled during the combined phase of treatment, suggesting an undesirable effect on the part of progestin [48].

In recent epidemiological studies, it has been reported that Lp(a) concentrations increase similar to TC and LDL cholesterol following menopause and postmenopausal women taking estrogen have lower Lp(a) concentrations than those who do not receive estrogen supplements [16, 49]. Large-sized community-based studies have revealed that Lp(a) lipoprotein levels increase with age, especially in women, and that females have higher levels than males [50, 51]. The Lp(a) lipoprotein levels are associated with female sex hormones. Postmenopausal women have higher levels than premenopausal women [50]. In our study, we have found higher Lp(a) levels in postmenopausal women than the premenopausals. However, the authors of Framingham Offspring Study insist that Lp(a) levels are correlated with age and the influence of menopausal state is not statistically significant after controlling for age [52]. We do not agree with the Framingham Offspring Study. Most studies, including our study, report a significant effect of female sex hormone on Lp(a); moreover, the association of Lp(a) with age may be a consequence of postmenopausal change. Although several studies have suggested that increase in Lp(a) levels predisposes to atherosclerosis and CVD [53, 54], the full clinical implication of this marker on vascular damage has not been completely elucidated [28, 53]. Among ERT regimens, oral CEE induces the greatest reduction, while transdermal estrogen has a lesser effect than orally administered preparations [28]. In our study, CEE decreased Lp(a) to a lesser extent than that reported in the study of Kim et al. [55], but our results were similar to that of Christodoulakos et al. [3]. There is no consensus on whether progestin alters the effect of estrogen on Lp(a) [56]. It has been suggested that the impact of the progestin on Lp(a) may not be as important as on the other lipids [27] that it may augment the decrease in Lp(a) [28] or that it may attenuate the effect of estrogen [55]. Both treatment regimes caused similar decreases in Lp(a). This reduction was compatible with those reported in recent multicentric studies [49, 57]. The basal concentration of Lp(a) was positively related to the difference and percent change of Lp(a) concentration. This finding implies that the Lp(a)-lowering effect is more prominent in subjects with high basal levels. This is clinically very important, because the individuals with higher risk due to higher Lp(a) levels will benefit more from HRT.

In the light of all these data, we conclude that positive lipid profiles can be achieved with postmenopausal estrogen replacement therapy, and that the addition of MPA as progesterone supplement to estrogen therapy does not alter the positive lipid profile achieved with estrogens in postmenopausal women with intact uterus.

Copyright information

© Springer-Verlag 2006