- Email: [email protected]
Associations of postmenopausal hormone therapy with metabolic syndrome among diabetic and non-diabetic women
Maturitas 121 (2019) 76–82
Contents lists available at ScienceDirect
Maturitas journal homepage: www.elsevier.com/locate/maturitas
Associations of postmenopausal hormone therapy with metabolic syndrome among diabetic and non-diabetic women
T
Ji-Eun Kima, Jaesung Choia, JooYong Parka, Jong-koo Leeb,c, Aesun Shind,e, Sang Min Parka,c, ⁎ Daehee Kanga,d,e,f, Ji-Yeob Choia,d,e, a
Department of Biomedical Sciences, Seoul National University Graduate School, Seoul, Republic of Korea JW Lee Center for Global Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea c Department of Family Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea d Department of Preventive Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea e Cancer Research Institute, Seoul National University, Seoul, Republic of Korea f Institute of Environmental Medicine, Seoul National University Medical Research Center, Seoul, Republic of Korea b
ARTICLE INFO
ABSTRACT
Keywords: Hormone therapy Menopause Metabolic syndrome Diabetes
A lack of estrogen due to menopause changes metabolic homeostasis, which increases the risk of metabolic syndrome (MetS) in postmenopausal women. Hormone therapy (HT) has beneficial effects on chronic diseases as well as on menopause symptoms. The aim of this study was to investigate the association of HT use with MetS and its components by diabetes status in middle-aged postmenopausal Korean women. A cross-sectional analysis was undertaken of a total of 39,295 non-diabetic and 3,359 diabetic postmenopausal women aged 40 to 69 years from the Health Examinees-Gem (HEXA-G) study (2004–2013). The mean differences in the MetS components by HT use were assessed using a general linear model and Tukey’s multiple comparisons tests. The prevalence odds ratio (POR) and 95% confidence intervals (CIs) were estimated using the logistic regression model. HT use was associated with lower fasting glucose level, total cholesterol, systolic blood pressure, body mass index, waist circumference, and waist-to-hip ratio among both diabetic and non-diabetic women. In non-diabetic women, HT ‘ever’ use was negatively associated with the prevalence of MetS (POR = 0.80, 95% CI: 0.75-0.85), and current users had a lower prevalence of MetS (POR = 0.68, 95% CI: 0.60-0.76). A longer duration of HT use was associated with a decreasing prevalence of MetS. We did not find heterogeneity by age regarding MetS prevalence. Our results suggest that HT use is negatively associated with the prevalence of MetS among postmenopausal women. However, further longitudinal studies are required to investigate the effect of HT on MetS in Korean women.
1. Introduction
(HT) on chronic diseases and related factors [7–12], the use of HT for the prevention of chronic diseases has not been recommended [13]. Because previous RCTs included specific women who were older, Caucasian or had coronary heart diseases, they reported uncertainties in the generalizability to all postmenopausal women [7–10]. In addition, the necessity of studies targeting younger healthy menopausal women with varying ethnicities was emphasized [14]. Diabetic women have a higher risk of dyslipidemia, hypertension, and CVD than non-diabetic women [15–17]. It was suggested that diabetes could modify the association of HT with clinical outcomes. There are several reports on the differential effects of HT use on each MetS component according to diabetes status [15,16,18]. A review of HT use and diabetes suggested that most of the evidence was
With the rising prevalence of metabolic syndrome (MetS), diabetes and mortality from cardiovascular disease (CVD) have also increased [1–3]. As people age, sex differences in the prevalence of MetS and each of its component are observed [4]. Furthermore, women after menopause have a higher prevalence of MetS compared with premenopausal women regardless of age [5]. This phenomenon is related to estrogen deficiency caused by menopause. Thus, the alteration of metabolic homeostasis via estrogen deficiency predisposes many women to MetS [6]. Although both observational studies and randomized controlled trials (RCTs) have suggested the beneficial effects of hormone therapy
⁎ Corresponding author at: Department of Biomedical Sciences, Seoul National University Graduate School, 103 Daehak-ro, Jongno-gu, Seoul, 03080, Republic of Korea. E-mail address: [email protected] (J.-Y. Choi).
https://doi.org/10.1016/j.maturitas.2018.12.012 Received 7 September 2018; Received in revised form 10 December 2018; Accepted 19 December 2018 0378-5122/ © 2018 Elsevier B.V. All rights reserved.
Maturitas 121 (2019) 76–82
J.-E. Kim et al.
insufficient to recommend HT because the evidence resulted from studies with limited power [17]. Considering both the lack of research on younger healthy Asian women and the differences according to diabetes status, the aim of this study is to investigate the association between the use of HT and MetS and its components among postmenopausal Korean women.
2.3. Diabetes and MetS as outcomes Diabetes status was assessed on the basis of the measured fasting glucose (FG) level (mg/dL) and the use of medication for diabetes diagnosed by a doctor before recruiting at baseline. Diabetic women were defined as having an FG level ≥126 mg/dL or those who reported the use of hypoglycemic medication or insulin after diabetes diagnosis. Non-diabetic women were defined as those with an FG level < 126 mg/ dL and without diagnosed diabetes. As the components of MetS, FG, total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), triglyceride (TG), systolic blood pressure (SBP), diastolic blood pressure (DBP), body mass index, waist circumference (WC), waist-to-hip ratio (WHR), and visceral fat were considered. All of these were measured at recruitment. MetS was defined according to the National Cholesterol Education Program Adult Treatment Panel III [22], except for the criterion of WC which followed the Asian guidelines. Among non-diabetic women, those who met three or more of the following criteria were defined as having MetS: (1) WC ≥ 80 cm, (2) TG ≥ 150 mg/dL or drugs and treatment for elevated TG, (3) HDL-C ≤50 md/dL, (4) FG ≥ 100 mg/dL, and (5) SBP ≥ 130, DBP ≥ 85 mmHg, or drug treatment for elevated BP.
2. Methods 2.1. Study population The Health Examinees (HEXA) study in Korea is a large-scale genomic epidemiologic cohort study that recruited subjects aged 40–69 years between 2004 and 2013. The information on HEXA’s rationale and study design has been described in previous reports [19,20]. This current study used updated HEXA-Gem (HEXA-G) data that applied additional eligibility criteria to the HEXA participants. Details of the HEXA-G study can be found in the published report [21]. Among the 139,345 HEXA-G subjects, 92,368 women were identified. We excluded those who lacked menopause information or premenopausal women (N = 38,382), had unknown HT use status (N = 761) or age at menopause (N = 1,600), had unknown MetS components or diabetes information (N = 6,753), and had premature ovarian failure (age at menopause less than 40 years, N = 2,218). Finally, 42,654 postmenopausal women were included in the current analysis. Among them, 3,359 (7.9%) diabetic women and 39,295 (92.1%) non-diabetic women were identified (Fig. 1). This study was approved by the institutional review board of Seoul National Hospital, Seoul, Korea (IRB No. E-1503-103-657). All participants provided written informed consent before participating in the study.
2.4. Covariates Demographic factors such as age (40–49, 50–59, and 60–69 years), education (less or equal to middle school, high school graduate, and greater than or equal to college), income (< 2.0, 2.0–3.9, and ≥4.0 million Korean won), occupation (manual labor, office, and unemployed/ house-wives), and marital status (living with spouse and living alone) were included. Behavioral factors such as smoking status (never and ever users), drinking status (never and ever users), and physical activity (none, < 150 min/week, and ≥150 min/week) were assessed. Reproductive factors such as menopause cause, age at menopause, hysterectomy, and oophorectomy were included. Menopause causes were categorized as natural, surgical, and others by chemotherapy or radiotherapy. Hysterectomy and oophorectomy were categorized as yes and no. Among the women who underwent oophorectomy, detailed information was investigated as follows: one side, partial, and complete resection. The information on diagnosis of hypertension, dyslipidemia, cancer, stroke and myocardial infarction (MI) or family history of hypertension and dyslipidemia were categorized as yes and no. Hypertension was defined as having BP ≥ 140/ 80 mmHg or taking antihypertensive agents. Dyslipidemia was defined as having TC ≥ 200 mg/dL or HDL ≤ 50 mg/dL or TG ≥ 150 mg/dL or in cases reporting the use of lipid-lowering agents. History of diagnosed cancer was considered as history of any cancer, including breast and cervical cancer.
2.2. Menopause and HT as exposure Menopausal status was assessed by whether the women were still having their periods. Postmenopausal women were defined as women who reported no periods for more than 12 months except for the women recruited in 2004. They were asked at the baseline survey whether they had had periods during the prior 6 months. Postmenopausal women were asked about the use of HT (pills or injections) after menopause. HT use was categorized as follows: never, past use, current use, and ever use (past and current use). Those who reported past or current use of HT were asked about the duration of HT use (months), respectively.
Fig. 1. HEXA-G study population included in the current analysis Abbreviations: HT, hormone therapy; MetS, metabolic syndrome. 77
Maturitas 121 (2019) 76–82
J.-E. Kim et al.
2.5. Statistical analysis
value by F test = 0.018). Table 3 shows the POR of MetS and HT use among the non-diabetic women. HT use was significantly associated with a decreasing prevalence of MetS among the ever (POR = 0.80, 95% CI: 0.75-0.85), past (POR = 0.84, 95% CI: 0.78-0.89), and current users (POR = 0.68, 95% CI: 0.60-0.76) compared with the never users. A longer duration of HT use was significantly associated with a decreasing prevalence of MetS compared with the HT never users (< 5 years vs. never use: POR = 0.83, 95% CI: 0.78-0.88; ≥5 years vs. never use: POR = 0.63, 95% CI: 0.56-0.72). Table 4 shows the associations between MetS prevalence and HT use by the strata of age. There were few differences in MetS by HT use among the women aged 50–59 years and 60–69 years, while MetS was not significantly associated with HT use among those aged 40–49 years. Although there was no differential association with MetS prevalence by age group among the HT ever users (P-value by Q-test > 0.05 and I2 < 30%), the HT current users were slightly heterogeneous by age group (P-value by Q-test > 0.05 and I2 > 50%). The results of a sensitivity analysis after excluding the women who had no periods for more than 6 months and those who reported stroke and MI were similar to the main results (data not shown).
The associations between HT use and each covariate and the associations between HT use and MetS components were separately investigated among the diabetic and non-diabetic women. Prevalence odds ratios (POR) and 95% confidence intervals (CI) were used to evaluate the association between each covariate and HT use using the multivariable logistic regression model among HT ever and never users and the multinomial regression model among HT past, current, and never users. If covariates affected the marked differential association between HT use and outcomes or if covariates showed an association with both of HT use and outcomes from previous studies, they were considered as the final confounding factors for the multivariable model. The comparison of the mean differences of the MetS components was investigated using the general linear model (GLM) adjusted for age, education, income, smoking, drinking, physical activity, menopause cause, hysterectomy, oophorectomy, hypertension, dyslipidemia, and cancer. The significant level of mean differences (P < 0.05) was evaluated using Student’s t-test and Tukey’s test. The false discovery rate (FDR) corrected P-values by considering the number of MetS components. Furthermore, whether diabetes status influenced the associations between HT use and MetS components was evaluated using the F test. To evaluate the MetS prevalence by HT use among non-diabetic women, the POR and 95% CI were estimated using the multivariable logistic regression model adjusted for the same covariates that were applied in the GLM model. Stratification analyses by age were conducted to investigate whether there were differential associations between age and MetS prevalence, and a heterogeneity test by age group was evaluated using the Q-test and I2 statistic. Whether a longer duration of HT use had a stronger association with MetS was evaluated by comparing the POR between < 5 and ≥5 years. The sensitivity analyses were conducted after excluding the postmenopausal women who reported no periods for more than 6 months, and the subjects who reported stroke or MI. The statistical analysis was conducted using SAS statistical software version 9.4 (SAS Institute, Cary, NC, USA).
4. Discussion In the present study, a significant association between HT use and the MetS components was identified among diabetic and non-diabetic postmenopausal Korean women. HT use was associated with a decreasing prevalence of MetS among the non-diabetic women. A longer duration of HT use was associated with a decreasing prevalence of MetS. There were few differences between HT use and MetS risk by age group. According to the guidelines for HT prescription, HT should be avoided in women with a history of CVD-related diseases such as stroke, MI, or transient ischemic attack, whereas the evaluation of CVD risk factors was recommended to determine the use of HT among those without CVD-related diseases [23]. After comparing the distribution between HT use and CVD risk factors, we found that the characteristics of the HT current users met the guidelines; for example, women who were aged 60 years older or who had hypertension, stroke, or MI were likely to have decreased HT current use. Although the non-diabetic women with MI were likely to have increased HT current use, we found that the concordance rate of self-reported MI compared with the medical records was lower than the estimates of stroke through internal identification within the HEXA-G (in preparation). Previous studies showed consistent effects of HT on glucose metabolism but inconsistent effects of HT on TC, HDL-C, and TG by diabetes status [15,16,18,24,25]. These associations of HT use with MetS components were also consistent in the study designs (RCT vs. observational studies) [13]. In addition, studies including healthy postmenopausal Korean women reported better associations with the MetS components [11,12]. However, these studies had comparatively small sample sizes of less than 2,500 or did not evaluate the various confounding factors related to HT use such as age, duration of HT use, or diabetes status. It has been reported that estrogen plays a role in regulating energy homeostasis and body fat distribution and improves insulin resistance and -cell dysfunction [26]. Diabetic women are known to be hyperandrogenic, which is caused by their decrease in estrogen compared with non-diabetic women, and HT use has been shown to have beneficial effects on glucose and lipid metabolism in diabetic women [27]. This might reflect not only the effects of HT but also healthy lifestyles and medication use among HT users with diabetes [15] because we included prevalent diabetic cases. Previous studies showed that TG and HDL-C had differential associations depending on the regimen or route of administration of HT [12,16,24,28]: oral estrogen increased TG, whereas transdermal estrogen decreased TG [24]; tibolone decreased TG and HDL-C [12]; 17 -estradiol (E2) decreased TG compared with
3. Results Table 1 shows the characteristics of the study population by HT use according to diabetes status. There were 20.9% HT ever users among the 3,359 diabetic women and 26.5% HT ever users among the 39,295 non-diabetic women. Regardless of diabetes status, the proportion of HT ever use increased among the older women, those who regularly exercised, and those with a family history of hypertension compared with the HT never users. The diabetic women with HT ever use had an increased association with living alone and having a family history of diabetes. The non-diabetic women with HT ever use had an increased association with higher level of education and income, being housewives, ever smoking and drinking, menopause caused by surgery or other reasons (chemotherapy or radiotherapy), and having undergone hysterectomy or oophorectomy, whereas decreased associations with the women who had hypertension, dyslipidemia, and cancer history were identified. Similar patterns were found among past and current HT use versus never use (Supplementary Tables 1 and 2). Table 2 shows the mean differences in the MetS components by HT use among the diabetic and non-diabetic women. The diabetic women were associated with statistically significant lower FG, TC, SBP, WC, and WHR among HT ever use as compared with among HT never use (Pvalue by FDR < 0.05). The non-diabetic women had stronger associations in all of the MetS components among HT ever use compared with never use (P-value by FDR < 0.05). HT ever use had differential associations with FG by diabetes status (P-value by F test < 0.001). Supplementary Tables 3 and 4 show the mean differences in the MetS components by HT use status comparing HT past and current users with HT never users. HT current use had a differential association with FG and WC by diabetes status (FG: P-value by F test < 0.001 and WC: P78
Maturitas 121 (2019) 76–82
J.-E. Kim et al.
Table 1 Characteristics of diabetic and non-diabetic postmenopausal women according to hormone therapy use. Diabetic women (N = 3,359, 7.9%)
HT never (N = 2,658, 79.1%)
N
(%)
Age (years, mean (SD)) 59.4 ± 5.7 40–49 105 (4.0) 50–59 1,200 (45.2) 60–69 1,353 (50.9) Education ≤Middle school 1,692 (63.7) High school 771 (29.0) ≥College 163 (6.1) Income (10,000 won) < 200 1,251 (47.1) 200–400 788 (29.7) ≥400 287 (10.8) Unknown 332 (12.5) Current occupation Manual labor 599 (22.5) Office 110 (4.1) House-wives 1,889 (71.1) Marital status Living spouse 2,180 (82.0) Living alone 467 (17.6) Smoking Never 2,557 (96.2) Ever 93 (3.5) Drinking Never 2,182 (82.1) Ever 471 (17.7) Physical activity (min/week) No 1,250 (47.0) < 150 281 (10.6) ≥150 1,048 (39.4) Menopause cause Natural 2,303 (86.6) Surgery 327 (12.3) Others 15 (0.6) Hysterectomy No 2,307 (86.8) Yes 348 (13.1) Oophorectomy No 2,248 (84.6) Yes 184 (6.9) Unknown 226 (8.5) Cancer No 2,547 (95.8) Yes 106 (4.0) Stroke No 2,587 (97.3) Yes 69 (2.6) Myocardial infarction No 2,491 (93.7) Yes 162 (6.1) Hypertension No 1,242 (46.7) Yes 1,416 (53.3) Dyslipidemia No 549 (20.7) Yes 2,109 (79.3) Family history of diabetes No 1,676 (63.1) Yes 967 (36.4) Family history of hypertension No 1,877 (70.6) Yes 771 (29.0) Family history of dyslipidemia No 1,573 (59.2) Yes 33 (1.2) Unknown 1,052 (39.6)
HT ever (N = 701, 20.9%) N 59.8 12 312 377
Non-diabetic women (N = 39,295. 92.1%)
Adjusted POR (95% CI)1
(%)
HT never (N = 28,898, 73.5%) N
(%)
HT ever (N = 10,397, 26.5%) N
Adjusted POR (95% CI)1
(%)
± 5.0 (1.7) (44.5) (53.8)
1.00 2.94 3.42
Reference (1.57–5.50) (1.82–6.45)
56.8 ± 5.7 2,506 (8.7) 17,101 (59.2) 9,291 (32.2)
57.4 ± 5.4 673 (6.5) 6,033 (58.0) 3,691 (35.5)
1.00 1.65 2.04
Reference (1.50–1.81) (1.84–2.26)
430 204 55
(61.3) (29.1) (7.9)
1.00 1.08 1.33
Reference (0.88–1.32) (0.93–1.91)
14,614 10,509 3,388
(50.6) (36.4) (11.7)
5,084 3,900 1,272
(48.9) (37.5) (12.2)
1.00 1.07 1.05
Reference (1.02–1.13) (0.96–1.14)
309 236 59 97
(44.1) (33.7) (8.4) (13.8)
1.00 1.19 0.81
Reference (0.97–1.46) (0.58–1.13) –
10,573 9,602 4,429 4,294
(36.6) (33.2) (15.3) (14.9)
3,612 3,381 1,653 1,751
(34.7) (32.5) (15.9) (16.8)
1.00 1.06 1.11
Reference (1.00–1.12) (1.03–1.20) –
131 29 527
(18.7) (4.1) (75.2)
1.00 1.19 1.22
Reference (0.73–1.93) (0.97–1.53)
7,451 2,242 18,523
(25.8) (7.8) (64.1)
2,255 797 7,021
(21.7) (7.7) (67.5)
1.00 1.10 1.12
Reference (0.99–1.22) (1.06–1.19)
541 156
(77.2) (22.3)
1.00 1.42
Reference (1.14–1.76)
24,369 4,416
(84.3) (15.3)
8,699 1,660
(83.7) (16.0)
1.00 1.04
Reference (0.97–1.11)
678 22
(96.7) (3.1)
1.00 0.90
Reference (0.55–1.47)
27,979 794
(96.8) (2.8)
10,019 320
(96.4) (3.1)
1.00 1.15
Reference (1.01–1.32)
559 139
(79.7) (19.8)
1.00 1.23
Reference (0.99–1.54)
21,436 7,327
(74.2) (25.4)
7,393 2,945
(71.1) (28.3)
1.00 1.22
Reference (1.16–1.28)
303 64 313
(43.2) (9.1) (44.7)
1.00 0.91 1.20
Reference (0.67–1.24) (1.00–1.45)
14,119 3,275 10,569
(48.9) (11.3) (36.6)
4,339 1,341 4,287
(41.7) (12.9) (41.2)
1.00 1.27 1.26
Reference (1.18–1.37) (1.20–1.33)
530 163 5
(75.6) (23.3) (0.7)
1.00 1.49 1.47
Reference (0.92–2.41) (0.52–4.16)
24,595 3,915 194
(85.1) (13.6) (0.7)
8,103 2,108 79
(77.9) (20.3) (0.8)
1.00 1.22 1.45
Reference (1.06–1.40) (1.10–1.90)
526 173
(75.0) (24.7)
1.00 1.44
Reference (0.90–2.32)
24,923 3,933
(86.2) (13.6)
8,224 2,141
(79.1) (20.6)
1.00 1.17
Reference (1.02–1.35)
547 94 60
(78.0) (13.4) (8.6)
1.00 1.32
Reference (0.96–1.80) –
23,397 1,625 3,876
(81.0) (5.6) (13.4)
7,383 1,268 1,746
(71.0) (12.2) (16.8)
1.00 2.07
Reference (1.89–2.26) –
658 43
(93.9) (6.1)
1.00 1.31
Reference (0.90–1.91)
27,616 1,246
(95.6) (4.3)
9,903 473
(95.3) (4.6)
1.00 0.88
Reference (0.79–0.99)
685 16
(97.7) (2.3)
1.00 0.87
Reference (0.49-1.52)
28,600 295
(99.0) (1.02)
10,306 91
(99.1) (0.9)
1.00 0.82
Reference (0.64–1.04)
663 38
(94.6) (5.4)
1.00 0.85
Reference (0.58–1.23)
28,228 666
(97.7) (2.3)
10,105 291
(97.2) (2.8)
1.00 1.15
Reference (1.00–1.33)
341 360
(48.6) (51.4)
1.00 0.86
Reference (0.73–1.03)
20,146 8,752
(69.7) (30.3)
7,357 3,040
(70.8) (29.2)
1.00 0.89
Reference (0.85–0.94)
131 570
(18.7) (81.3)
1.00 1.12
Reference (0.90–1.39)
6,144 22,754
(21.3) (78.7)
2,424 7,973
(23.3) (76.7)
1.00 0.88
Reference (0.83–0.93)
409 290
(58.4) (41.4)
1.00 1.21
Reference (1.02–1.45)
24,171 4,580
(83.6) (15.9)
8,606 1,723
(82.8) (16.6)
1.00 1.05
Reference (0.99–1.12)
463 236
(66.1) (33.7)
1.00 1.25
Reference (1.04–1.51)
20,071 8,744
(69.5) (30.3)
7,055 3,309
(67.9) (31.8)
1.00 1.08
Reference (1.02–1.13)
398 12 291
(56.8) (1.7) (41.5)
1.00 1.41
Reference (0.71–2.80) –
16,134 426 12,338
(55.8) (1.5) (42.7)
5,160 173 5,064
(49.6) (1.7) (48.7)
1.00 1.23
Reference (1.02–1.48) –
Abbreviations: HT, hormone therapy; POR, prevalence odds ratio; CI, confidence interval. 1 Adjusted for age (continuous), education, job, income, smoking, drinking, physical activity, menopause cause, hysterectomy, oophorectomy, hypertension, dyslipidemia and cancer. 79
Maturitas 121 (2019) 76–82
J.-E. Kim et al.
Table 2 Mean differences in metabolic syndrome components by hormone therapy. Diabetic women
Fasting glucose (mg/dl)* Total cholesterol (mg/dl) High density lipoprotein (mg/dl) Triglyceride (mg/dl) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Body mass index (kg/m2) Waist circumference (cm) Waist-to-hip ratio Visceral fat (kg)
Non-diabetic women
HT never Adjusted mean (95% CI)1
HT ever Adjusted mean (95% CI)1
HT never Adjusted mean (95% CI)1
HT ever Adjusted mean (95% CI)1
142.06 (121.06, 163.07) 200.54 (180.94, 220.13) 57.40 (52.27, 62.54) 150.34 (101.55, 199.12) 123.41 (116.50, 130.32) 78.61 (74.31, 82.90) 24.94 (23.41, 26.47) 84.74 (80.85, 88.62) 0.89 (0.86, 0.92) 2.66 (2.17, 3.15)
137.05 (115.93, 158.17)2,3 196.71 (177.01, 216.41)2,3 57.47 (52.31, 62.63) 141.89 (92.83, 190.95)2 121.57 (114.62, 128.52) 2,3 78.04 (73.72, 82.36) 24.66 (23.12, 26.19) 2 83.68 (79.78, 87.58) 2,3 0.88 (0.86, 0.91) 2,3 2.58 (2.09, 3.08)
89.29 (88.0, 90.58) 196.27 (192.02, 200.51) 57.16 (55.57, 58.75) 113.79 (104.36, 123.21) 122.82 (121.17, 124.47) 78.08 (76.99, 79.17) 23.91 (23.55, 24.28) 79.23 (78.24, 80.22) 0.84 (0.83, 0.85) 2.27 (2.15, 2.39)
88.23 (86.94, 89.52)2,3 193.23 (188.98, 197.48)2,3 57.45 (55.85, 59.04)2,3 108.31 (98.87, 177.75)2,3 121.17 (119.52, 122.83) 2,3 77.20 (76.11, 78.28) 2,3 23.58 (23.21, 23.94) 2,3 78.27 (77.28, 79.26) 2,3 0.83 (0.82, 0.84) 2,3 2.18 (2.06, 2.30) 2,3
Abbreviations: HT, hormone therapy; CI, confidence interval. 1 Adjusted for age (continuous), education, job, income, smoking, drinking, physical activity, menopause cause, hysterectomy, oophorectomy, hypertension, dyslipidemia and cancer. 2 Significant differences between never and ever users (P-value < 0.05). 3 Significant differences between never and ever users by multiple comparison test (P-value by FDR < 0.05). * Significant interaction between diabetes status and HT ever use (FG: P-value using the F-test < 0.001).
Table 3 Associations between the prevalence of metabolic syndrome and hormone therapy use among non-diabetic women. MetS
No MetS
(N = 10,896, 27.7%)
(N = 28,399, 72.3%)
N Status of HT use Never 8,338 Ever 2,558 Past 2,065 Current 493 Duration of HT use (years) Never 8,338 <5 1,979 ≥5 447
Age adjusted POR (95% CI)
Multivariable adjusted POR (95% CI)1
(%)
N
(%)
(76.5) (23.5) (19.0) (4.5)
20,560 7,839 5,702 2,137
(72.4) (27.6) (20.1) (7.5)
1.00 0.77 0.82 0.62
Reference (0.73–0.81) (0.77–0.87) (0.56–0.69)
1.00 0.80 0.84 0.68
Reference (0.75–0.85) (0.78–0.89) (0.60–0.76)
(76.5) (18.2) (4.1)
20,560 6,009 1,506
(72.4) (21.2) (5.3)
1.00 0.81 0.61
Reference (0.76–0.85) (0.54–0.68)
1.00 0.83 0.63
Reference (0.78–0.88) (0.56–0.72)
Abbreviations: MetS, metabolic syndrome; HT, hormone therapy; POR, prevalence odds ratio; CI, confidence interval. 1 Adjusted for age (continuous), education, job, income, smoking, drinking, physical activity, menopause cause, hysterectomy, oophorectomy, hypertension, dyslipidemia, and cancer.
conjugated equine estrogen (CEE), whereas CEE increased HDL-C compared with E2 [29]. These beneficial effects of HT use on MetS components results in a decreasing prevalence of MetS with a cluster of its components. Although the Women’s Health Initiative suggested that age and time since menopause are important factors for CVD risk [30], we did not identify significant results in women aged 40 years from the stratification analysis by age. This may be due to insufficient statistical power as a result of the small sample size of participants with diabetes or differential characteristics, such as menopause cause, history of hysterectomy or oophorectomy among women aged 40 years. There are limitations to the current study. First, it used a crosssectional design. It is difficult to evaluate whether the prescription of HT was weighted to the healthy women or whether HT use benefitted MetS. Second, the current guidelines on HT use have suggested considering the patient’s health prior to prescription. Our study population was shown to meet the current guideline, and metabolically healthier women may have more prescriptions of HT. Therefore, we could not totally exclude the healthy-user bias. Although, in the present study, the diabetes status did not influence the associations between HT use and MetS components, except for FG level, we still need to consider a healthy-user bias. Third, we could not rule out potential recall bias because this study used self-reported information regarding menopause and HT use. Finally, information about the regimen types or obvious routes of HT administration was not collected at baseline. Thus, we
could not consider the associations by regimen types or routes of HT use. Nevertheless, to the best of our knowledge, this is the first study to assess the association of between HT use and MetS and its components stratified by diabetes status or age group in Korean women. In addition, we included healthy postmenopausal women who are likely to consider using HT due to menopausal symptoms. In conclusion, our results suggest that HT use is associated with lower prevalence of MetS among postmenopausal Korean women. Although there are controversies regarding the benefits and risks of HT, it has been recommended that the prescription of HT take into consideration the history of disease and the characteristics of the patient. Therefore, if scrupulous evaluation is conducted before initiating HT, it is expected to have beneficial effects for postmenopausal women. However, additional longitudinal studies to investigating the association between HT use and MetS are required. Contributors Ji-Eun Kim participated in analyzing and interpreting the data and writing and editing the manuscript and discussion. Jaesung Choi and JooYong Park contributed to reviewing the manuscript and discussion. Jong-koo Lee contributed to conducting, designing and supervising the study and reviewed the manuscript. 80
Maturitas 121 (2019) 76–82 Reference 0.82 (0.74–0.91) 0.59 (0.50–0.70)
Reference 0.76 (0.69–0.83) 0.80 (0.72–0.88) 0.56 (0.46–0.69)
5,703 (69.4) 1,673 (20.4) 739 (9.0)
The authors declare that they have no conflict of interest. Funding This work was supported by the Research Program funded by the Korea Centers for Disease Control and Prevention (2004-E71004-00, 2005-E71011-00, 2005-E71009-00, 2006-E71001-00, 2006-E71004-00, 2006-E71010-00, 2006-E71003-00, 2007-E71004-00, 2007-E71006-00, 2008-E71006-00, 2008-E71008-00, 2009-E71009-00, 2010-E71006-00, 2011-E71006-00, 2012-E71001-00, and 2013-E71009-00), Education and Research Encouragement Fund of Seoul National University Hospital (2018) and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No.2018R1A2A3075397).
Abbreviations: MetS, metabolic syndrome; HT, hormone therapy; POR, prevalence odds ratio; CI, confidence interval. 1 Adjusted for education, job, income, smoking, drinking, physical activity, menopause cause, hysterectomy, oophorectomy, hypertension, dyslipidemia, and cancer.
3,588 (75.2) 850 (17.8) 271 (5.7) 4,358 (77.2) 1,050 (18.6) 172 (3.1) Reference 0.83 (0.62–1.11) 0.63 (0.20–2.01) 2,114 (78.3) 514 (19.0) 35 (1.3)
12,743 (72.9) 3,822 (21.9) 732 (4.2)
Reference 0.85 (0.78–0.92) 0.72 (0.59–0.87)
5,703 (69.4) 2,511 (30.6) 2,063 (25.1) 448 (5.5) 3,588 (75.2) 1,180 (24.8) 1,023 (21.5) 157 (3.3) Reference 0.84 (0.77–0.91) 0.88 (0.80–0.96) 0.73 (0.63–0.84) 12,743 (72.9) 4,744 (27.1) 3,333 (19.1) 1,411 (8.1) 4,358 (77.2) 1,289 (22.8) 989 (17.5) 300 (5.3) Reference 0.85 (0.64–1.12) 0.93 (0.66–1.32) 0.75 (0.51–1.12) 2,114 (78.3) 584 (21.7) 306 (11.3) 278 (10.3)
No MetS (N = 2,698, 84.9%) MetS (N = 481, 15.1%)
Aesun Shin and Sang Min Park contributed to reviewing the manuscript and discussion. Daehee Kang contributed to conducting, designing and supervising the study and reviewed the manuscript. Ji-Yeob Choi participated in the study design, reviewed and edited the manuscript, contributed to the discussion, and supervised the study. Conflict of interest
Status of HT use Never 392 (81.5) Ever 89 (18.5) Past 53 (11.0) Current 36 (7.5) Duration of HT use (years) Never 392 (81.5) <5 79 (16.4) ≥5 4 (0.8)
No MetS (N = 8,214, 63.3%) MetS (N = 4,768, 36.7%) No MetS (N = 17,487, 75.6%) MetS (N = 5,647, 24.4%)
50–59 years
Adjusted POR (95% CI)1 40–49 years
Table 4 Associations between the prevalence of metabolic syndrome and hormone therapy use by age group among non-diabetic women.
Adjusted POR (95% CI)1
60–69 years
Adjusted POR (95% CI)1
J.-E. Kim et al.
Ethical approval This study was approved by the institutional Review Board of the Seoul National Hospital, Seoul, Korea (IRB No. E-1503-103-657). Provenance and peer review This article has undergone peer review. Research data (data sharing and collaboration) There are no linked research data sets for this paper. The authors do not have permission to share data. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.maturitas.2018.12. 012. References [1] S.E. Lee, K. Han, Y.M. Kang, S.O. Kim, Y.K. Cho, K.S. Ko, J.Y. Park, K.U. Lee, E.H. Koh, A. Taskforce Team of Diabetes Fact Sheet of the Korean Diabetes, Trends in the prevalence of metabolic syndrome and its components in South Korea: findings from the Korean National Health Insurance Service Database (2009–2013), PLoS One 13 (3) (2018) e0194490. [2] K.C. Sung, E.J. Rhee, S. Ryu, B.J. Kim, B.S. Kim, W.Y. Lee, K.W. Oh, Y.B. Kim, P.W. Chung, H. Kim, C.D. Byrne, K.B. Lee, S.W. Park, Increased cardiovascular mortality in subjects with metabolic syndrome is largely attributable to diabetes and hypertension in 159,971 Korean adults, J. Clin. Endocrinol. Metab. 100 (7) (2015) 2606–2612. [3] J. Noh, The diabetes epidemic in Korea, Endocrinol. Metab. (Seoul) 31 (3) (2016) 349–353. [4] F. Mauvais-Jarvis, Sex differences in metabolic homeostasis, diabetes, and obesity, Biol. Sex Differ. 6 (2015) 14. [5] H.M. Kim, J. Park, S.Y. Ryu, J. Kim, The effect of menopause on the metabolic syndrome among Korean women: the Korean National Health and Nutrition Examination Survey, 2001, Diabetes Care 30 (3) (2007) 701–706. [6] M.C. Carr, The emergence of the metabolic syndrome with menopause, J. Clin. Endocrinol. Metab. 88 (6) (2003) 2404–2411. [7] A.M. Kanaya, D. Herrington, E. Vittinghoff, F. Lin, D. Grady, V. Bittner, J.A. Cauley, E. Barrett-Connor, S. Heart, Estrogen/progestin Replacement, Glycemic effects of postmenopausal hormone therapy: the Heart and Estrogen/progestin Replacement Study. A randomized, double-blind, placebo-controlled trial, Ann. Intern. Med. 138 (1) (2003) 1–9. [8] K.L. Margolis, D.E. Bonds, R.J. Rodabough, L. Tinker, L.S. Phillips, C. Allen,
81
Maturitas 121 (2019) 76–82
J.-E. Kim et al.
[9]
[10] [11]
[12] [13] [14] [15]
[16] [17] [18]
T. Bassford, G. Burke, J. Torrens, B.V. Howard, I. Women’s Health Initiative, Effect of oestrogen plus progestin on the incidence of diabetes in postmenopausal women: results from the Women’s Health Initiative Hormone Trial, Diabetologia 47 (7) (2004) 1175–1187. J.E. Manson, R.T. Chlebowski, M.L. Stefanick, A.K. Aragaki, J.E. Rossouw, R.L. Prentice, G. Anderson, B.V. Howard, C.A. Thomson, A.Z. LaCroix, J. Wactawski-Wende, R.D. Jackson, M. Limacher, K.L. Margolis, S. WassertheilSmoller, S.A. Beresford, J.A. Cauley, C.B. Eaton, M. Gass, J. Hsia, K.C. Johnson, C. Kooperberg, L.H. Kuller, C.E. Lewis, S. Liu, L.W. Martin, J.K. Ockene, M.J. O’Sullivan, L.H. Powell, M.S. Simon, L. Van Horn, M.Z. Vitolins, R.B. Wallace, Menopausal hormone therapy and health outcomes during the intervention and extended poststopping phases of the Women’s Health Initiative randomized trials, JAMA 310 (13) (2013) 1353–1368. Effects of estrogen or estrogen/progestin regimens on heart disease risk factors in postmenopausal women. The postmenopausal estrogen/progestin interventions (PEPI) trial. The writing Group for the PEPI Trial, JAMA 273 (3) (1995) 199–208. E.Y. Ki, S.Y. Hur, J.S. Park, K. Do Han, Y.G. Park, Differences in the lipid profile and hormone replacement therapy use in Korean postmenopausal women: the Korea National Health and Nutrition Examination Survey (KNHANES) 2010-2012, Arch. Gynecol. Obstet. 294 (1) (2016) 165–173. S.M. Kim, S.H. Kim, J.R. Lee, B.C. Jee, S.Y. Ku, C.S. Suh, Y.M. Choi, J.G. Kim, S.Y. Moon, The effects of hormone therapy on metabolic risk factors in postmenopausal Korean women, Climacteric 14 (1) (2011) 66–74. D. Lovre, S.H. Lindsey, F. Mauvais-Jarvis, Effect of menopausal hormone therapy on components of the metabolic syndrome, Ther. Adv. Cardiovasc. Dis. (2016). J. Marjoribanks, C. Farquhar, H. Roberts, A. Lethaby, J. Lee, Long-term hormone therapy for perimenopausal and postmenopausal women, Cochrane Database Syst. Rev. 1 (2017) CD004143. C.J. Crespo, E. Smit, A. Snelling, C.T. Sempos, R.E. Andersen, Nhanes III, Hormone replacement therapy and its relationship to lipid and glucose metabolism in diabetic and nondiabetic postmenopausal women: results from the Third National Health and Nutrition Examination Survey (NHANES III), Diabetes Care 25 (10) (2002) 1675–1680. P. Manwaring, L. Morfis, T. Diamond, L.G. Howes, The effects of hormone replacement therapy on plasma lipids in type II diabetes, Maturitas 34 (3) (2000) 239–247. C.A. Stuenkel, Menopause, hormone therapy and diabetes, Climacteric 20 (1) (2017) 11–21. Y. Zhang, B.V. Howard, L.D. Cowan, T.K. Welty, C.F. Schaefer, R.A. Wild, J. Yeh, E.T. Lee, Associations of postmenopausal hormone therapy with markers of
[19] [20] [21]
[22]
[23] [24]
[25] [26] [27] [28] [29] [30]
82
hemostasis and inflammation and lipid profiles in diabetic and nondiabetic American Indian women: the strong heart study, J. Womens Health (Larchmt) 13 (2) (2004) 155–163. G. Health Examinees Study, The Health Examinees (HEXA) study: rationale, study design and baseline characteristics, Asian Pac. J. Cancer Prev. 16 (4) (2015) 1591–1597. Y. Kim, B.G. Han, G.E.Sg. Ko, Cohort profile: the Korean genome and epidemiology study (KoGES) consortium, Int. J. Epidemiol. 46 (2) (2017) e20. J. Choi, J.Y. Choi, A. Shin, S.A. Lee, K.M. Lee, J. Oh, J.Y. Park, J.K. Lee, D. Kang, Trends and correlates of high-risk alcohol consumption and types of alcoholic beverages in middle-aged Korean adults: results from the HEXA-G study, J. Epidemiol. (2018). E. National Cholesterol Education Program Expert Panel on Detection, A, Treatment of high blood cholesterol in, third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III) final report, Circulation 106 (25) (2002) 3143–3421. C.A. Stuenkel, S.R. Davis, A. Gompel, M.A. Lumsden, M.H. Murad, J.V. Pinkerton, R.J. Santen, Treatment of symptoms of the menopause: an endocrine society clinical practice guideline, J. Clin. Endocrinol. Metab. 100 (11) (2015) 3975–4011. S.R. Salpeter, J.M. Walsh, T.M. Ormiston, E. Greyber, N.S. Buckley, E.E. Salpeter, Meta-analysis: effect of hormone-replacement therapy on components of the metabolic syndrome in postmenopausal women, Diabetes Obes. Metab. 8 (5) (2006) 538–554. Y. Xu, J. Lin, S. Wang, J. Xiong, Q. Zhu, Combined estrogen replacement therapy on metabolic control in postmenopausal women with diabetes mellitus, Kaohsiung J. Med. Sci. 30 (7) (2014) 350–361. F. Mauvais-Jarvis, J.E. Manson, J.C. Stevenson, V.A. Fonseca, Menopausal hormone therapy and type 2 diabetes prevention: evidence, mechanisms, and clinical implications, Endocr. Rev. 38 (3) (2017) 173–188. B. Andersson, P. Marin, L. Lissner, A. Vermeulen, P. Bjorntorp, Testosterone concentrations in women and men with NIDDM, Diabetes Care 17 (5) (1994) 405–411. P. Anagnostis, S.A. Paschou, N. Katsiki, D. Krikidis, I. Lambrinoudaki, D.G. Goulis, Menopausal hormone therapy and cardiovascular risk. Where are we now? Curr. Vasc. Pharmacol. (2018). D. Pu, R. Tan, Q. Yu, J. Wu, Metabolic syndrome in menopause and associated factors: a meta-analysis, Climacteric 20 (6) (2017) 583–591. G. Rosano, C. Vitale, I. Spoletini, M. Fini, Cardiovascular health in the menopausal woman: impact of the timing of hormone replacement therapy, Climacteric 15 (4) (2012) 299–305.
Contents lists available at ScienceDirect
Maturitas journal homepage: www.elsevier.com/locate/maturitas
Associations of postmenopausal hormone therapy with metabolic syndrome among diabetic and non-diabetic women
T
Ji-Eun Kima, Jaesung Choia, JooYong Parka, Jong-koo Leeb,c, Aesun Shind,e, Sang Min Parka,c, ⁎ Daehee Kanga,d,e,f, Ji-Yeob Choia,d,e, a
Department of Biomedical Sciences, Seoul National University Graduate School, Seoul, Republic of Korea JW Lee Center for Global Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea c Department of Family Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea d Department of Preventive Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea e Cancer Research Institute, Seoul National University, Seoul, Republic of Korea f Institute of Environmental Medicine, Seoul National University Medical Research Center, Seoul, Republic of Korea b
ARTICLE INFO
ABSTRACT
Keywords: Hormone therapy Menopause Metabolic syndrome Diabetes
A lack of estrogen due to menopause changes metabolic homeostasis, which increases the risk of metabolic syndrome (MetS) in postmenopausal women. Hormone therapy (HT) has beneficial effects on chronic diseases as well as on menopause symptoms. The aim of this study was to investigate the association of HT use with MetS and its components by diabetes status in middle-aged postmenopausal Korean women. A cross-sectional analysis was undertaken of a total of 39,295 non-diabetic and 3,359 diabetic postmenopausal women aged 40 to 69 years from the Health Examinees-Gem (HEXA-G) study (2004–2013). The mean differences in the MetS components by HT use were assessed using a general linear model and Tukey’s multiple comparisons tests. The prevalence odds ratio (POR) and 95% confidence intervals (CIs) were estimated using the logistic regression model. HT use was associated with lower fasting glucose level, total cholesterol, systolic blood pressure, body mass index, waist circumference, and waist-to-hip ratio among both diabetic and non-diabetic women. In non-diabetic women, HT ‘ever’ use was negatively associated with the prevalence of MetS (POR = 0.80, 95% CI: 0.75-0.85), and current users had a lower prevalence of MetS (POR = 0.68, 95% CI: 0.60-0.76). A longer duration of HT use was associated with a decreasing prevalence of MetS. We did not find heterogeneity by age regarding MetS prevalence. Our results suggest that HT use is negatively associated with the prevalence of MetS among postmenopausal women. However, further longitudinal studies are required to investigate the effect of HT on MetS in Korean women.
1. Introduction
(HT) on chronic diseases and related factors [7–12], the use of HT for the prevention of chronic diseases has not been recommended [13]. Because previous RCTs included specific women who were older, Caucasian or had coronary heart diseases, they reported uncertainties in the generalizability to all postmenopausal women [7–10]. In addition, the necessity of studies targeting younger healthy menopausal women with varying ethnicities was emphasized [14]. Diabetic women have a higher risk of dyslipidemia, hypertension, and CVD than non-diabetic women [15–17]. It was suggested that diabetes could modify the association of HT with clinical outcomes. There are several reports on the differential effects of HT use on each MetS component according to diabetes status [15,16,18]. A review of HT use and diabetes suggested that most of the evidence was
With the rising prevalence of metabolic syndrome (MetS), diabetes and mortality from cardiovascular disease (CVD) have also increased [1–3]. As people age, sex differences in the prevalence of MetS and each of its component are observed [4]. Furthermore, women after menopause have a higher prevalence of MetS compared with premenopausal women regardless of age [5]. This phenomenon is related to estrogen deficiency caused by menopause. Thus, the alteration of metabolic homeostasis via estrogen deficiency predisposes many women to MetS [6]. Although both observational studies and randomized controlled trials (RCTs) have suggested the beneficial effects of hormone therapy
⁎ Corresponding author at: Department of Biomedical Sciences, Seoul National University Graduate School, 103 Daehak-ro, Jongno-gu, Seoul, 03080, Republic of Korea. E-mail address: [email protected] (J.-Y. Choi).
https://doi.org/10.1016/j.maturitas.2018.12.012 Received 7 September 2018; Received in revised form 10 December 2018; Accepted 19 December 2018 0378-5122/ © 2018 Elsevier B.V. All rights reserved.
Maturitas 121 (2019) 76–82
J.-E. Kim et al.
insufficient to recommend HT because the evidence resulted from studies with limited power [17]. Considering both the lack of research on younger healthy Asian women and the differences according to diabetes status, the aim of this study is to investigate the association between the use of HT and MetS and its components among postmenopausal Korean women.
2.3. Diabetes and MetS as outcomes Diabetes status was assessed on the basis of the measured fasting glucose (FG) level (mg/dL) and the use of medication for diabetes diagnosed by a doctor before recruiting at baseline. Diabetic women were defined as having an FG level ≥126 mg/dL or those who reported the use of hypoglycemic medication or insulin after diabetes diagnosis. Non-diabetic women were defined as those with an FG level < 126 mg/ dL and without diagnosed diabetes. As the components of MetS, FG, total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), triglyceride (TG), systolic blood pressure (SBP), diastolic blood pressure (DBP), body mass index, waist circumference (WC), waist-to-hip ratio (WHR), and visceral fat were considered. All of these were measured at recruitment. MetS was defined according to the National Cholesterol Education Program Adult Treatment Panel III [22], except for the criterion of WC which followed the Asian guidelines. Among non-diabetic women, those who met three or more of the following criteria were defined as having MetS: (1) WC ≥ 80 cm, (2) TG ≥ 150 mg/dL or drugs and treatment for elevated TG, (3) HDL-C ≤50 md/dL, (4) FG ≥ 100 mg/dL, and (5) SBP ≥ 130, DBP ≥ 85 mmHg, or drug treatment for elevated BP.
2. Methods 2.1. Study population The Health Examinees (HEXA) study in Korea is a large-scale genomic epidemiologic cohort study that recruited subjects aged 40–69 years between 2004 and 2013. The information on HEXA’s rationale and study design has been described in previous reports [19,20]. This current study used updated HEXA-Gem (HEXA-G) data that applied additional eligibility criteria to the HEXA participants. Details of the HEXA-G study can be found in the published report [21]. Among the 139,345 HEXA-G subjects, 92,368 women were identified. We excluded those who lacked menopause information or premenopausal women (N = 38,382), had unknown HT use status (N = 761) or age at menopause (N = 1,600), had unknown MetS components or diabetes information (N = 6,753), and had premature ovarian failure (age at menopause less than 40 years, N = 2,218). Finally, 42,654 postmenopausal women were included in the current analysis. Among them, 3,359 (7.9%) diabetic women and 39,295 (92.1%) non-diabetic women were identified (Fig. 1). This study was approved by the institutional review board of Seoul National Hospital, Seoul, Korea (IRB No. E-1503-103-657). All participants provided written informed consent before participating in the study.
2.4. Covariates Demographic factors such as age (40–49, 50–59, and 60–69 years), education (less or equal to middle school, high school graduate, and greater than or equal to college), income (< 2.0, 2.0–3.9, and ≥4.0 million Korean won), occupation (manual labor, office, and unemployed/ house-wives), and marital status (living with spouse and living alone) were included. Behavioral factors such as smoking status (never and ever users), drinking status (never and ever users), and physical activity (none, < 150 min/week, and ≥150 min/week) were assessed. Reproductive factors such as menopause cause, age at menopause, hysterectomy, and oophorectomy were included. Menopause causes were categorized as natural, surgical, and others by chemotherapy or radiotherapy. Hysterectomy and oophorectomy were categorized as yes and no. Among the women who underwent oophorectomy, detailed information was investigated as follows: one side, partial, and complete resection. The information on diagnosis of hypertension, dyslipidemia, cancer, stroke and myocardial infarction (MI) or family history of hypertension and dyslipidemia were categorized as yes and no. Hypertension was defined as having BP ≥ 140/ 80 mmHg or taking antihypertensive agents. Dyslipidemia was defined as having TC ≥ 200 mg/dL or HDL ≤ 50 mg/dL or TG ≥ 150 mg/dL or in cases reporting the use of lipid-lowering agents. History of diagnosed cancer was considered as history of any cancer, including breast and cervical cancer.
2.2. Menopause and HT as exposure Menopausal status was assessed by whether the women were still having their periods. Postmenopausal women were defined as women who reported no periods for more than 12 months except for the women recruited in 2004. They were asked at the baseline survey whether they had had periods during the prior 6 months. Postmenopausal women were asked about the use of HT (pills or injections) after menopause. HT use was categorized as follows: never, past use, current use, and ever use (past and current use). Those who reported past or current use of HT were asked about the duration of HT use (months), respectively.
Fig. 1. HEXA-G study population included in the current analysis Abbreviations: HT, hormone therapy; MetS, metabolic syndrome. 77
Maturitas 121 (2019) 76–82
J.-E. Kim et al.
2.5. Statistical analysis
value by F test = 0.018). Table 3 shows the POR of MetS and HT use among the non-diabetic women. HT use was significantly associated with a decreasing prevalence of MetS among the ever (POR = 0.80, 95% CI: 0.75-0.85), past (POR = 0.84, 95% CI: 0.78-0.89), and current users (POR = 0.68, 95% CI: 0.60-0.76) compared with the never users. A longer duration of HT use was significantly associated with a decreasing prevalence of MetS compared with the HT never users (< 5 years vs. never use: POR = 0.83, 95% CI: 0.78-0.88; ≥5 years vs. never use: POR = 0.63, 95% CI: 0.56-0.72). Table 4 shows the associations between MetS prevalence and HT use by the strata of age. There were few differences in MetS by HT use among the women aged 50–59 years and 60–69 years, while MetS was not significantly associated with HT use among those aged 40–49 years. Although there was no differential association with MetS prevalence by age group among the HT ever users (P-value by Q-test > 0.05 and I2 < 30%), the HT current users were slightly heterogeneous by age group (P-value by Q-test > 0.05 and I2 > 50%). The results of a sensitivity analysis after excluding the women who had no periods for more than 6 months and those who reported stroke and MI were similar to the main results (data not shown).
The associations between HT use and each covariate and the associations between HT use and MetS components were separately investigated among the diabetic and non-diabetic women. Prevalence odds ratios (POR) and 95% confidence intervals (CI) were used to evaluate the association between each covariate and HT use using the multivariable logistic regression model among HT ever and never users and the multinomial regression model among HT past, current, and never users. If covariates affected the marked differential association between HT use and outcomes or if covariates showed an association with both of HT use and outcomes from previous studies, they were considered as the final confounding factors for the multivariable model. The comparison of the mean differences of the MetS components was investigated using the general linear model (GLM) adjusted for age, education, income, smoking, drinking, physical activity, menopause cause, hysterectomy, oophorectomy, hypertension, dyslipidemia, and cancer. The significant level of mean differences (P < 0.05) was evaluated using Student’s t-test and Tukey’s test. The false discovery rate (FDR) corrected P-values by considering the number of MetS components. Furthermore, whether diabetes status influenced the associations between HT use and MetS components was evaluated using the F test. To evaluate the MetS prevalence by HT use among non-diabetic women, the POR and 95% CI were estimated using the multivariable logistic regression model adjusted for the same covariates that were applied in the GLM model. Stratification analyses by age were conducted to investigate whether there were differential associations between age and MetS prevalence, and a heterogeneity test by age group was evaluated using the Q-test and I2 statistic. Whether a longer duration of HT use had a stronger association with MetS was evaluated by comparing the POR between < 5 and ≥5 years. The sensitivity analyses were conducted after excluding the postmenopausal women who reported no periods for more than 6 months, and the subjects who reported stroke or MI. The statistical analysis was conducted using SAS statistical software version 9.4 (SAS Institute, Cary, NC, USA).
4. Discussion In the present study, a significant association between HT use and the MetS components was identified among diabetic and non-diabetic postmenopausal Korean women. HT use was associated with a decreasing prevalence of MetS among the non-diabetic women. A longer duration of HT use was associated with a decreasing prevalence of MetS. There were few differences between HT use and MetS risk by age group. According to the guidelines for HT prescription, HT should be avoided in women with a history of CVD-related diseases such as stroke, MI, or transient ischemic attack, whereas the evaluation of CVD risk factors was recommended to determine the use of HT among those without CVD-related diseases [23]. After comparing the distribution between HT use and CVD risk factors, we found that the characteristics of the HT current users met the guidelines; for example, women who were aged 60 years older or who had hypertension, stroke, or MI were likely to have decreased HT current use. Although the non-diabetic women with MI were likely to have increased HT current use, we found that the concordance rate of self-reported MI compared with the medical records was lower than the estimates of stroke through internal identification within the HEXA-G (in preparation). Previous studies showed consistent effects of HT on glucose metabolism but inconsistent effects of HT on TC, HDL-C, and TG by diabetes status [15,16,18,24,25]. These associations of HT use with MetS components were also consistent in the study designs (RCT vs. observational studies) [13]. In addition, studies including healthy postmenopausal Korean women reported better associations with the MetS components [11,12]. However, these studies had comparatively small sample sizes of less than 2,500 or did not evaluate the various confounding factors related to HT use such as age, duration of HT use, or diabetes status. It has been reported that estrogen plays a role in regulating energy homeostasis and body fat distribution and improves insulin resistance and -cell dysfunction [26]. Diabetic women are known to be hyperandrogenic, which is caused by their decrease in estrogen compared with non-diabetic women, and HT use has been shown to have beneficial effects on glucose and lipid metabolism in diabetic women [27]. This might reflect not only the effects of HT but also healthy lifestyles and medication use among HT users with diabetes [15] because we included prevalent diabetic cases. Previous studies showed that TG and HDL-C had differential associations depending on the regimen or route of administration of HT [12,16,24,28]: oral estrogen increased TG, whereas transdermal estrogen decreased TG [24]; tibolone decreased TG and HDL-C [12]; 17 -estradiol (E2) decreased TG compared with
3. Results Table 1 shows the characteristics of the study population by HT use according to diabetes status. There were 20.9% HT ever users among the 3,359 diabetic women and 26.5% HT ever users among the 39,295 non-diabetic women. Regardless of diabetes status, the proportion of HT ever use increased among the older women, those who regularly exercised, and those with a family history of hypertension compared with the HT never users. The diabetic women with HT ever use had an increased association with living alone and having a family history of diabetes. The non-diabetic women with HT ever use had an increased association with higher level of education and income, being housewives, ever smoking and drinking, menopause caused by surgery or other reasons (chemotherapy or radiotherapy), and having undergone hysterectomy or oophorectomy, whereas decreased associations with the women who had hypertension, dyslipidemia, and cancer history were identified. Similar patterns were found among past and current HT use versus never use (Supplementary Tables 1 and 2). Table 2 shows the mean differences in the MetS components by HT use among the diabetic and non-diabetic women. The diabetic women were associated with statistically significant lower FG, TC, SBP, WC, and WHR among HT ever use as compared with among HT never use (Pvalue by FDR < 0.05). The non-diabetic women had stronger associations in all of the MetS components among HT ever use compared with never use (P-value by FDR < 0.05). HT ever use had differential associations with FG by diabetes status (P-value by F test < 0.001). Supplementary Tables 3 and 4 show the mean differences in the MetS components by HT use status comparing HT past and current users with HT never users. HT current use had a differential association with FG and WC by diabetes status (FG: P-value by F test < 0.001 and WC: P78
Maturitas 121 (2019) 76–82
J.-E. Kim et al.
Table 1 Characteristics of diabetic and non-diabetic postmenopausal women according to hormone therapy use. Diabetic women (N = 3,359, 7.9%)
HT never (N = 2,658, 79.1%)
N
(%)
Age (years, mean (SD)) 59.4 ± 5.7 40–49 105 (4.0) 50–59 1,200 (45.2) 60–69 1,353 (50.9) Education ≤Middle school 1,692 (63.7) High school 771 (29.0) ≥College 163 (6.1) Income (10,000 won) < 200 1,251 (47.1) 200–400 788 (29.7) ≥400 287 (10.8) Unknown 332 (12.5) Current occupation Manual labor 599 (22.5) Office 110 (4.1) House-wives 1,889 (71.1) Marital status Living spouse 2,180 (82.0) Living alone 467 (17.6) Smoking Never 2,557 (96.2) Ever 93 (3.5) Drinking Never 2,182 (82.1) Ever 471 (17.7) Physical activity (min/week) No 1,250 (47.0) < 150 281 (10.6) ≥150 1,048 (39.4) Menopause cause Natural 2,303 (86.6) Surgery 327 (12.3) Others 15 (0.6) Hysterectomy No 2,307 (86.8) Yes 348 (13.1) Oophorectomy No 2,248 (84.6) Yes 184 (6.9) Unknown 226 (8.5) Cancer No 2,547 (95.8) Yes 106 (4.0) Stroke No 2,587 (97.3) Yes 69 (2.6) Myocardial infarction No 2,491 (93.7) Yes 162 (6.1) Hypertension No 1,242 (46.7) Yes 1,416 (53.3) Dyslipidemia No 549 (20.7) Yes 2,109 (79.3) Family history of diabetes No 1,676 (63.1) Yes 967 (36.4) Family history of hypertension No 1,877 (70.6) Yes 771 (29.0) Family history of dyslipidemia No 1,573 (59.2) Yes 33 (1.2) Unknown 1,052 (39.6)
HT ever (N = 701, 20.9%) N 59.8 12 312 377
Non-diabetic women (N = 39,295. 92.1%)
Adjusted POR (95% CI)1
(%)
HT never (N = 28,898, 73.5%) N
(%)
HT ever (N = 10,397, 26.5%) N
Adjusted POR (95% CI)1
(%)
± 5.0 (1.7) (44.5) (53.8)
1.00 2.94 3.42
Reference (1.57–5.50) (1.82–6.45)
56.8 ± 5.7 2,506 (8.7) 17,101 (59.2) 9,291 (32.2)
57.4 ± 5.4 673 (6.5) 6,033 (58.0) 3,691 (35.5)
1.00 1.65 2.04
Reference (1.50–1.81) (1.84–2.26)
430 204 55
(61.3) (29.1) (7.9)
1.00 1.08 1.33
Reference (0.88–1.32) (0.93–1.91)
14,614 10,509 3,388
(50.6) (36.4) (11.7)
5,084 3,900 1,272
(48.9) (37.5) (12.2)
1.00 1.07 1.05
Reference (1.02–1.13) (0.96–1.14)
309 236 59 97
(44.1) (33.7) (8.4) (13.8)
1.00 1.19 0.81
Reference (0.97–1.46) (0.58–1.13) –
10,573 9,602 4,429 4,294
(36.6) (33.2) (15.3) (14.9)
3,612 3,381 1,653 1,751
(34.7) (32.5) (15.9) (16.8)
1.00 1.06 1.11
Reference (1.00–1.12) (1.03–1.20) –
131 29 527
(18.7) (4.1) (75.2)
1.00 1.19 1.22
Reference (0.73–1.93) (0.97–1.53)
7,451 2,242 18,523
(25.8) (7.8) (64.1)
2,255 797 7,021
(21.7) (7.7) (67.5)
1.00 1.10 1.12
Reference (0.99–1.22) (1.06–1.19)
541 156
(77.2) (22.3)
1.00 1.42
Reference (1.14–1.76)
24,369 4,416
(84.3) (15.3)
8,699 1,660
(83.7) (16.0)
1.00 1.04
Reference (0.97–1.11)
678 22
(96.7) (3.1)
1.00 0.90
Reference (0.55–1.47)
27,979 794
(96.8) (2.8)
10,019 320
(96.4) (3.1)
1.00 1.15
Reference (1.01–1.32)
559 139
(79.7) (19.8)
1.00 1.23
Reference (0.99–1.54)
21,436 7,327
(74.2) (25.4)
7,393 2,945
(71.1) (28.3)
1.00 1.22
Reference (1.16–1.28)
303 64 313
(43.2) (9.1) (44.7)
1.00 0.91 1.20
Reference (0.67–1.24) (1.00–1.45)
14,119 3,275 10,569
(48.9) (11.3) (36.6)
4,339 1,341 4,287
(41.7) (12.9) (41.2)
1.00 1.27 1.26
Reference (1.18–1.37) (1.20–1.33)
530 163 5
(75.6) (23.3) (0.7)
1.00 1.49 1.47
Reference (0.92–2.41) (0.52–4.16)
24,595 3,915 194
(85.1) (13.6) (0.7)
8,103 2,108 79
(77.9) (20.3) (0.8)
1.00 1.22 1.45
Reference (1.06–1.40) (1.10–1.90)
526 173
(75.0) (24.7)
1.00 1.44
Reference (0.90–2.32)
24,923 3,933
(86.2) (13.6)
8,224 2,141
(79.1) (20.6)
1.00 1.17
Reference (1.02–1.35)
547 94 60
(78.0) (13.4) (8.6)
1.00 1.32
Reference (0.96–1.80) –
23,397 1,625 3,876
(81.0) (5.6) (13.4)
7,383 1,268 1,746
(71.0) (12.2) (16.8)
1.00 2.07
Reference (1.89–2.26) –
658 43
(93.9) (6.1)
1.00 1.31
Reference (0.90–1.91)
27,616 1,246
(95.6) (4.3)
9,903 473
(95.3) (4.6)
1.00 0.88
Reference (0.79–0.99)
685 16
(97.7) (2.3)
1.00 0.87
Reference (0.49-1.52)
28,600 295
(99.0) (1.02)
10,306 91
(99.1) (0.9)
1.00 0.82
Reference (0.64–1.04)
663 38
(94.6) (5.4)
1.00 0.85
Reference (0.58–1.23)
28,228 666
(97.7) (2.3)
10,105 291
(97.2) (2.8)
1.00 1.15
Reference (1.00–1.33)
341 360
(48.6) (51.4)
1.00 0.86
Reference (0.73–1.03)
20,146 8,752
(69.7) (30.3)
7,357 3,040
(70.8) (29.2)
1.00 0.89
Reference (0.85–0.94)
131 570
(18.7) (81.3)
1.00 1.12
Reference (0.90–1.39)
6,144 22,754
(21.3) (78.7)
2,424 7,973
(23.3) (76.7)
1.00 0.88
Reference (0.83–0.93)
409 290
(58.4) (41.4)
1.00 1.21
Reference (1.02–1.45)
24,171 4,580
(83.6) (15.9)
8,606 1,723
(82.8) (16.6)
1.00 1.05
Reference (0.99–1.12)
463 236
(66.1) (33.7)
1.00 1.25
Reference (1.04–1.51)
20,071 8,744
(69.5) (30.3)
7,055 3,309
(67.9) (31.8)
1.00 1.08
Reference (1.02–1.13)
398 12 291
(56.8) (1.7) (41.5)
1.00 1.41
Reference (0.71–2.80) –
16,134 426 12,338
(55.8) (1.5) (42.7)
5,160 173 5,064
(49.6) (1.7) (48.7)
1.00 1.23
Reference (1.02–1.48) –
Abbreviations: HT, hormone therapy; POR, prevalence odds ratio; CI, confidence interval. 1 Adjusted for age (continuous), education, job, income, smoking, drinking, physical activity, menopause cause, hysterectomy, oophorectomy, hypertension, dyslipidemia and cancer. 79
Maturitas 121 (2019) 76–82
J.-E. Kim et al.
Table 2 Mean differences in metabolic syndrome components by hormone therapy. Diabetic women
Fasting glucose (mg/dl)* Total cholesterol (mg/dl) High density lipoprotein (mg/dl) Triglyceride (mg/dl) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Body mass index (kg/m2) Waist circumference (cm) Waist-to-hip ratio Visceral fat (kg)
Non-diabetic women
HT never Adjusted mean (95% CI)1
HT ever Adjusted mean (95% CI)1
HT never Adjusted mean (95% CI)1
HT ever Adjusted mean (95% CI)1
142.06 (121.06, 163.07) 200.54 (180.94, 220.13) 57.40 (52.27, 62.54) 150.34 (101.55, 199.12) 123.41 (116.50, 130.32) 78.61 (74.31, 82.90) 24.94 (23.41, 26.47) 84.74 (80.85, 88.62) 0.89 (0.86, 0.92) 2.66 (2.17, 3.15)
137.05 (115.93, 158.17)2,3 196.71 (177.01, 216.41)2,3 57.47 (52.31, 62.63) 141.89 (92.83, 190.95)2 121.57 (114.62, 128.52) 2,3 78.04 (73.72, 82.36) 24.66 (23.12, 26.19) 2 83.68 (79.78, 87.58) 2,3 0.88 (0.86, 0.91) 2,3 2.58 (2.09, 3.08)
89.29 (88.0, 90.58) 196.27 (192.02, 200.51) 57.16 (55.57, 58.75) 113.79 (104.36, 123.21) 122.82 (121.17, 124.47) 78.08 (76.99, 79.17) 23.91 (23.55, 24.28) 79.23 (78.24, 80.22) 0.84 (0.83, 0.85) 2.27 (2.15, 2.39)
88.23 (86.94, 89.52)2,3 193.23 (188.98, 197.48)2,3 57.45 (55.85, 59.04)2,3 108.31 (98.87, 177.75)2,3 121.17 (119.52, 122.83) 2,3 77.20 (76.11, 78.28) 2,3 23.58 (23.21, 23.94) 2,3 78.27 (77.28, 79.26) 2,3 0.83 (0.82, 0.84) 2,3 2.18 (2.06, 2.30) 2,3
Abbreviations: HT, hormone therapy; CI, confidence interval. 1 Adjusted for age (continuous), education, job, income, smoking, drinking, physical activity, menopause cause, hysterectomy, oophorectomy, hypertension, dyslipidemia and cancer. 2 Significant differences between never and ever users (P-value < 0.05). 3 Significant differences between never and ever users by multiple comparison test (P-value by FDR < 0.05). * Significant interaction between diabetes status and HT ever use (FG: P-value using the F-test < 0.001).
Table 3 Associations between the prevalence of metabolic syndrome and hormone therapy use among non-diabetic women. MetS
No MetS
(N = 10,896, 27.7%)
(N = 28,399, 72.3%)
N Status of HT use Never 8,338 Ever 2,558 Past 2,065 Current 493 Duration of HT use (years) Never 8,338 <5 1,979 ≥5 447
Age adjusted POR (95% CI)
Multivariable adjusted POR (95% CI)1
(%)
N
(%)
(76.5) (23.5) (19.0) (4.5)
20,560 7,839 5,702 2,137
(72.4) (27.6) (20.1) (7.5)
1.00 0.77 0.82 0.62
Reference (0.73–0.81) (0.77–0.87) (0.56–0.69)
1.00 0.80 0.84 0.68
Reference (0.75–0.85) (0.78–0.89) (0.60–0.76)
(76.5) (18.2) (4.1)
20,560 6,009 1,506
(72.4) (21.2) (5.3)
1.00 0.81 0.61
Reference (0.76–0.85) (0.54–0.68)
1.00 0.83 0.63
Reference (0.78–0.88) (0.56–0.72)
Abbreviations: MetS, metabolic syndrome; HT, hormone therapy; POR, prevalence odds ratio; CI, confidence interval. 1 Adjusted for age (continuous), education, job, income, smoking, drinking, physical activity, menopause cause, hysterectomy, oophorectomy, hypertension, dyslipidemia, and cancer.
conjugated equine estrogen (CEE), whereas CEE increased HDL-C compared with E2 [29]. These beneficial effects of HT use on MetS components results in a decreasing prevalence of MetS with a cluster of its components. Although the Women’s Health Initiative suggested that age and time since menopause are important factors for CVD risk [30], we did not identify significant results in women aged 40 years from the stratification analysis by age. This may be due to insufficient statistical power as a result of the small sample size of participants with diabetes or differential characteristics, such as menopause cause, history of hysterectomy or oophorectomy among women aged 40 years. There are limitations to the current study. First, it used a crosssectional design. It is difficult to evaluate whether the prescription of HT was weighted to the healthy women or whether HT use benefitted MetS. Second, the current guidelines on HT use have suggested considering the patient’s health prior to prescription. Our study population was shown to meet the current guideline, and metabolically healthier women may have more prescriptions of HT. Therefore, we could not totally exclude the healthy-user bias. Although, in the present study, the diabetes status did not influence the associations between HT use and MetS components, except for FG level, we still need to consider a healthy-user bias. Third, we could not rule out potential recall bias because this study used self-reported information regarding menopause and HT use. Finally, information about the regimen types or obvious routes of HT administration was not collected at baseline. Thus, we
could not consider the associations by regimen types or routes of HT use. Nevertheless, to the best of our knowledge, this is the first study to assess the association of between HT use and MetS and its components stratified by diabetes status or age group in Korean women. In addition, we included healthy postmenopausal women who are likely to consider using HT due to menopausal symptoms. In conclusion, our results suggest that HT use is associated with lower prevalence of MetS among postmenopausal Korean women. Although there are controversies regarding the benefits and risks of HT, it has been recommended that the prescription of HT take into consideration the history of disease and the characteristics of the patient. Therefore, if scrupulous evaluation is conducted before initiating HT, it is expected to have beneficial effects for postmenopausal women. However, additional longitudinal studies to investigating the association between HT use and MetS are required. Contributors Ji-Eun Kim participated in analyzing and interpreting the data and writing and editing the manuscript and discussion. Jaesung Choi and JooYong Park contributed to reviewing the manuscript and discussion. Jong-koo Lee contributed to conducting, designing and supervising the study and reviewed the manuscript. 80
Maturitas 121 (2019) 76–82 Reference 0.82 (0.74–0.91) 0.59 (0.50–0.70)
Reference 0.76 (0.69–0.83) 0.80 (0.72–0.88) 0.56 (0.46–0.69)
5,703 (69.4) 1,673 (20.4) 739 (9.0)
The authors declare that they have no conflict of interest. Funding This work was supported by the Research Program funded by the Korea Centers for Disease Control and Prevention (2004-E71004-00, 2005-E71011-00, 2005-E71009-00, 2006-E71001-00, 2006-E71004-00, 2006-E71010-00, 2006-E71003-00, 2007-E71004-00, 2007-E71006-00, 2008-E71006-00, 2008-E71008-00, 2009-E71009-00, 2010-E71006-00, 2011-E71006-00, 2012-E71001-00, and 2013-E71009-00), Education and Research Encouragement Fund of Seoul National University Hospital (2018) and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No.2018R1A2A3075397).
Abbreviations: MetS, metabolic syndrome; HT, hormone therapy; POR, prevalence odds ratio; CI, confidence interval. 1 Adjusted for education, job, income, smoking, drinking, physical activity, menopause cause, hysterectomy, oophorectomy, hypertension, dyslipidemia, and cancer.
3,588 (75.2) 850 (17.8) 271 (5.7) 4,358 (77.2) 1,050 (18.6) 172 (3.1) Reference 0.83 (0.62–1.11) 0.63 (0.20–2.01) 2,114 (78.3) 514 (19.0) 35 (1.3)
12,743 (72.9) 3,822 (21.9) 732 (4.2)
Reference 0.85 (0.78–0.92) 0.72 (0.59–0.87)
5,703 (69.4) 2,511 (30.6) 2,063 (25.1) 448 (5.5) 3,588 (75.2) 1,180 (24.8) 1,023 (21.5) 157 (3.3) Reference 0.84 (0.77–0.91) 0.88 (0.80–0.96) 0.73 (0.63–0.84) 12,743 (72.9) 4,744 (27.1) 3,333 (19.1) 1,411 (8.1) 4,358 (77.2) 1,289 (22.8) 989 (17.5) 300 (5.3) Reference 0.85 (0.64–1.12) 0.93 (0.66–1.32) 0.75 (0.51–1.12) 2,114 (78.3) 584 (21.7) 306 (11.3) 278 (10.3)
No MetS (N = 2,698, 84.9%) MetS (N = 481, 15.1%)
Aesun Shin and Sang Min Park contributed to reviewing the manuscript and discussion. Daehee Kang contributed to conducting, designing and supervising the study and reviewed the manuscript. Ji-Yeob Choi participated in the study design, reviewed and edited the manuscript, contributed to the discussion, and supervised the study. Conflict of interest
Status of HT use Never 392 (81.5) Ever 89 (18.5) Past 53 (11.0) Current 36 (7.5) Duration of HT use (years) Never 392 (81.5) <5 79 (16.4) ≥5 4 (0.8)
No MetS (N = 8,214, 63.3%) MetS (N = 4,768, 36.7%) No MetS (N = 17,487, 75.6%) MetS (N = 5,647, 24.4%)
50–59 years
Adjusted POR (95% CI)1 40–49 years
Table 4 Associations between the prevalence of metabolic syndrome and hormone therapy use by age group among non-diabetic women.
Adjusted POR (95% CI)1
60–69 years
Adjusted POR (95% CI)1
J.-E. Kim et al.
Ethical approval This study was approved by the institutional Review Board of the Seoul National Hospital, Seoul, Korea (IRB No. E-1503-103-657). Provenance and peer review This article has undergone peer review. Research data (data sharing and collaboration) There are no linked research data sets for this paper. The authors do not have permission to share data. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.maturitas.2018.12. 012. References [1] S.E. Lee, K. Han, Y.M. Kang, S.O. Kim, Y.K. Cho, K.S. Ko, J.Y. Park, K.U. Lee, E.H. Koh, A. Taskforce Team of Diabetes Fact Sheet of the Korean Diabetes, Trends in the prevalence of metabolic syndrome and its components in South Korea: findings from the Korean National Health Insurance Service Database (2009–2013), PLoS One 13 (3) (2018) e0194490. [2] K.C. Sung, E.J. Rhee, S. Ryu, B.J. Kim, B.S. Kim, W.Y. Lee, K.W. Oh, Y.B. Kim, P.W. Chung, H. Kim, C.D. Byrne, K.B. Lee, S.W. Park, Increased cardiovascular mortality in subjects with metabolic syndrome is largely attributable to diabetes and hypertension in 159,971 Korean adults, J. Clin. Endocrinol. Metab. 100 (7) (2015) 2606–2612. [3] J. Noh, The diabetes epidemic in Korea, Endocrinol. Metab. (Seoul) 31 (3) (2016) 349–353. [4] F. Mauvais-Jarvis, Sex differences in metabolic homeostasis, diabetes, and obesity, Biol. Sex Differ. 6 (2015) 14. [5] H.M. Kim, J. Park, S.Y. Ryu, J. Kim, The effect of menopause on the metabolic syndrome among Korean women: the Korean National Health and Nutrition Examination Survey, 2001, Diabetes Care 30 (3) (2007) 701–706. [6] M.C. Carr, The emergence of the metabolic syndrome with menopause, J. Clin. Endocrinol. Metab. 88 (6) (2003) 2404–2411. [7] A.M. Kanaya, D. Herrington, E. Vittinghoff, F. Lin, D. Grady, V. Bittner, J.A. Cauley, E. Barrett-Connor, S. Heart, Estrogen/progestin Replacement, Glycemic effects of postmenopausal hormone therapy: the Heart and Estrogen/progestin Replacement Study. A randomized, double-blind, placebo-controlled trial, Ann. Intern. Med. 138 (1) (2003) 1–9. [8] K.L. Margolis, D.E. Bonds, R.J. Rodabough, L. Tinker, L.S. Phillips, C. Allen,
81
Maturitas 121 (2019) 76–82
J.-E. Kim et al.
[9]
[10] [11]
[12] [13] [14] [15]
[16] [17] [18]
T. Bassford, G. Burke, J. Torrens, B.V. Howard, I. Women’s Health Initiative, Effect of oestrogen plus progestin on the incidence of diabetes in postmenopausal women: results from the Women’s Health Initiative Hormone Trial, Diabetologia 47 (7) (2004) 1175–1187. J.E. Manson, R.T. Chlebowski, M.L. Stefanick, A.K. Aragaki, J.E. Rossouw, R.L. Prentice, G. Anderson, B.V. Howard, C.A. Thomson, A.Z. LaCroix, J. Wactawski-Wende, R.D. Jackson, M. Limacher, K.L. Margolis, S. WassertheilSmoller, S.A. Beresford, J.A. Cauley, C.B. Eaton, M. Gass, J. Hsia, K.C. Johnson, C. Kooperberg, L.H. Kuller, C.E. Lewis, S. Liu, L.W. Martin, J.K. Ockene, M.J. O’Sullivan, L.H. Powell, M.S. Simon, L. Van Horn, M.Z. Vitolins, R.B. Wallace, Menopausal hormone therapy and health outcomes during the intervention and extended poststopping phases of the Women’s Health Initiative randomized trials, JAMA 310 (13) (2013) 1353–1368. Effects of estrogen or estrogen/progestin regimens on heart disease risk factors in postmenopausal women. The postmenopausal estrogen/progestin interventions (PEPI) trial. The writing Group for the PEPI Trial, JAMA 273 (3) (1995) 199–208. E.Y. Ki, S.Y. Hur, J.S. Park, K. Do Han, Y.G. Park, Differences in the lipid profile and hormone replacement therapy use in Korean postmenopausal women: the Korea National Health and Nutrition Examination Survey (KNHANES) 2010-2012, Arch. Gynecol. Obstet. 294 (1) (2016) 165–173. S.M. Kim, S.H. Kim, J.R. Lee, B.C. Jee, S.Y. Ku, C.S. Suh, Y.M. Choi, J.G. Kim, S.Y. Moon, The effects of hormone therapy on metabolic risk factors in postmenopausal Korean women, Climacteric 14 (1) (2011) 66–74. D. Lovre, S.H. Lindsey, F. Mauvais-Jarvis, Effect of menopausal hormone therapy on components of the metabolic syndrome, Ther. Adv. Cardiovasc. Dis. (2016). J. Marjoribanks, C. Farquhar, H. Roberts, A. Lethaby, J. Lee, Long-term hormone therapy for perimenopausal and postmenopausal women, Cochrane Database Syst. Rev. 1 (2017) CD004143. C.J. Crespo, E. Smit, A. Snelling, C.T. Sempos, R.E. Andersen, Nhanes III, Hormone replacement therapy and its relationship to lipid and glucose metabolism in diabetic and nondiabetic postmenopausal women: results from the Third National Health and Nutrition Examination Survey (NHANES III), Diabetes Care 25 (10) (2002) 1675–1680. P. Manwaring, L. Morfis, T. Diamond, L.G. Howes, The effects of hormone replacement therapy on plasma lipids in type II diabetes, Maturitas 34 (3) (2000) 239–247. C.A. Stuenkel, Menopause, hormone therapy and diabetes, Climacteric 20 (1) (2017) 11–21. Y. Zhang, B.V. Howard, L.D. Cowan, T.K. Welty, C.F. Schaefer, R.A. Wild, J. Yeh, E.T. Lee, Associations of postmenopausal hormone therapy with markers of
[19] [20] [21]
[22]
[23] [24]
[25] [26] [27] [28] [29] [30]
82
hemostasis and inflammation and lipid profiles in diabetic and nondiabetic American Indian women: the strong heart study, J. Womens Health (Larchmt) 13 (2) (2004) 155–163. G. Health Examinees Study, The Health Examinees (HEXA) study: rationale, study design and baseline characteristics, Asian Pac. J. Cancer Prev. 16 (4) (2015) 1591–1597. Y. Kim, B.G. Han, G.E.Sg. Ko, Cohort profile: the Korean genome and epidemiology study (KoGES) consortium, Int. J. Epidemiol. 46 (2) (2017) e20. J. Choi, J.Y. Choi, A. Shin, S.A. Lee, K.M. Lee, J. Oh, J.Y. Park, J.K. Lee, D. Kang, Trends and correlates of high-risk alcohol consumption and types of alcoholic beverages in middle-aged Korean adults: results from the HEXA-G study, J. Epidemiol. (2018). E. National Cholesterol Education Program Expert Panel on Detection, A, Treatment of high blood cholesterol in, third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III) final report, Circulation 106 (25) (2002) 3143–3421. C.A. Stuenkel, S.R. Davis, A. Gompel, M.A. Lumsden, M.H. Murad, J.V. Pinkerton, R.J. Santen, Treatment of symptoms of the menopause: an endocrine society clinical practice guideline, J. Clin. Endocrinol. Metab. 100 (11) (2015) 3975–4011. S.R. Salpeter, J.M. Walsh, T.M. Ormiston, E. Greyber, N.S. Buckley, E.E. Salpeter, Meta-analysis: effect of hormone-replacement therapy on components of the metabolic syndrome in postmenopausal women, Diabetes Obes. Metab. 8 (5) (2006) 538–554. Y. Xu, J. Lin, S. Wang, J. Xiong, Q. Zhu, Combined estrogen replacement therapy on metabolic control in postmenopausal women with diabetes mellitus, Kaohsiung J. Med. Sci. 30 (7) (2014) 350–361. F. Mauvais-Jarvis, J.E. Manson, J.C. Stevenson, V.A. Fonseca, Menopausal hormone therapy and type 2 diabetes prevention: evidence, mechanisms, and clinical implications, Endocr. Rev. 38 (3) (2017) 173–188. B. Andersson, P. Marin, L. Lissner, A. Vermeulen, P. Bjorntorp, Testosterone concentrations in women and men with NIDDM, Diabetes Care 17 (5) (1994) 405–411. P. Anagnostis, S.A. Paschou, N. Katsiki, D. Krikidis, I. Lambrinoudaki, D.G. Goulis, Menopausal hormone therapy and cardiovascular risk. Where are we now? Curr. Vasc. Pharmacol. (2018). D. Pu, R. Tan, Q. Yu, J. Wu, Metabolic syndrome in menopause and associated factors: a meta-analysis, Climacteric 20 (6) (2017) 583–591. G. Rosano, C. Vitale, I. Spoletini, M. Fini, Cardiovascular health in the menopausal woman: impact of the timing of hormone replacement therapy, Climacteric 15 (4) (2012) 299–305.