This study investigated the cardiometabolic effects of hormone therapies initiated within 3 years of onset of menopause after a 14-year follow-up of participants of the Kronos Early Estrogen Prevention Study (KEEPS) clinical trial. There was no evidence of cardiovascular and/or metabolic benefits or adverse effects associated with 4 years use of oral or transdermal forms of hormone therapy by recently menopausal women with good cardiovascular health after 10 years.
Key Words: Cardiovascular risk, Hormone therapy, Menopause, Metabolic risk, Diabetes
Abstract
Objective
This study aimed to determine long-term cardiometabolic effects of hormone therapies initiated within 3 years of onset of menopause after a 14-year follow-up study of participants of the Kronos Early Estrogen Prevention Study (KEEPS).
Methods
KEEPS was a multisite clinical trial that recruited recently menopausal women with good cardiovascular health for randomization to oral conjugated equine estrogens (Premarin, 0.45 mg/d) or transdermal 17β-estradiol (Climara, 50 μg/d) both with micronized progesterone (Prometrium, 200 mg/d) for 12 d/mo, or placebo pills and patch for 4 years. KEEPS continuation recontacted KEEPS participants 14 years after randomization and 10 years after the completion of the 4-year clinical trial to attend in-person clinic visits.
Results
Participants of KEEPS continuation (n = 299 of the 727 KEEPS participants; 41%) had an average age of 67 years (range, 58-73 y). Measurements of systolic and diastolic blood pressures, waist-to-hip ratio, fasting levels of glucose, insulin, lipid profiles, and homeostasis model assessment of insulin resistance were not different among the treatment groups at either KEEPS baseline or at KEEPS continuation visits, or for change between these two visits. The frequency of self-reported diabetes (P = 0.007) and use of diabetes medications was higher in the placebo than the oral conjugated equine estrogens (P = 0.045) or transdermal 17β-estradiol (P = 0.02) groups, but these differences were not supported by the laboratory measurements of glycemia or insulin resistance.
Conclusions
There was no evidence of cardiovascular and/or metabolic benefits or adverse effects associated with 4 years use of oral or transdermal forms of hormone therapy by recently menopausal women with good cardiovascular health after 10 years.
Menopausal hormone therapy (HT) is the most effective therapy for the vasomotor symptoms and genitourinary syndromes of menopause,1 as well as for the maintenance of bone mineral density and prevention of osteoporosis.2,3 Early termination of the Women's Health Initiative (WHI) HT trial because of increased risk of cerebrovascular events, breast cancer, and dementia in the combined estrogen plus progestin treatment group4 led to questions on whether HT given early during the menopausal transition would have similar risks.5,6 The Kronos Early Estrogen Prevention Study (KEEPS) was a randomized double-blind placebo controlled trial of HT, which tested the hypothesis that HT administered to recently postmenopausal women (within 3 y of menopause) with good cardiovascular health would slow the progression of atherosclerosis.7 During the 4 years of HT with either oral conjugated equine estrogens (oCEE) or transdermal 17β-estradiol (tE2), the effect of HT on the progression indices of atherosclerosis (coronary artery calcification and carotid intimal medial thickness) was neutral in comparison to placebo.8 A secondary goal in KEEPS was to investigate the cognitive and mood effects of the two forms of HT (oCEE and tE2), administered early after the onset of menopause. Although a lower frequency of depression and anxiety was observed in the oCEE group compared with the placebo group, the effects of either form of HT were neutral for cognitive outcomes in comparison with placebo during the 4 years of intervention in KEEPS.9,10
KEEPS continuation was initiated approximately 10 years after the KEEPS interventions were completed. The primary goal of the KEEPS continuation was to determine the long-term risks and benefits of HT on Alzheimer's disease pathophysiology, and cognitive and mood health. Analysis of the data from KEEPS continuation on the Alzheimer's disease biomarkers, and cognitive and mood health is ongoing. Here, we report on characteristics of the KEEPS participants who returned for in-person follow-up examinations approximately 14 years since randomization and 10 years since completion of KEEPS, to participate in the KEEPS continuation observational study. We examined the change in clinical and laboratory measurements of cardiovascular and metabolic risk factors from KEEPS baseline to KEEPS continuation in relation to the assigned treatment in KEEPS. The objective was to examine if the short-term exposure to menopausal HT when initiated shortly after menopause onset holds any long-term cardiometabolic implications a decade after discontinuation of exposure.
METHODS
KEEPS continuation participants
All women participating in the KEEPS continuation had participated in the original KEEPS trial, which recruited women (n = 727) from nine sites across the United States from 2005 to 2008. At the time of enrollment in the KEEPS trial, participants were 6 to 36 months past their last menses. KEEPS participants were randomized to one of three treatments: (1) oCEE (Premarin, 0.45 mg/d), (2) tE2 (Climara, 50 μg/d), or (3) placebo pills and patch. Estrogens were administered through oral or transdermal routes because it was hypothesized that tE2 would have a different effect on risk factors for atherosclerosis and thromboembolic disease compared with oCEE because of first-pass metabolism of oral products in the liver. Micronized progesterone (Prometrium, 200 mg/d) was given orally for 12 days at the beginning of each month to both of the active treatment groups to protect the endometrium, and placebo progesterone was given to the group assigned placebo estrogen.7
KEEPS continuation recontacted KEEPS participants from 2017 to 2022, approximately 14 years past their randomization to either oCEE, tE2, or placebo and approximately 10 years past the termination of KEEPS interventions. Of the 727 postmenopausal participants of KEEPS, valid contact information was available for 622 (86%), who were invited to participate in the KEEPS continuation study. Of the 622 KEEPS participants invited to participate in KEEPS continuation, 194 did not respond to the invitation, 10 were deceased, and 119 declined to participate. Of those who agreed to participate, KEEPS continuation enrolled 299 of the 727 KEEPS trial participants (overall, 41%; oCEE, 39%; tE2, 43%; placebo, 41%) at seven sites (Banner Alzheimer's Institute, Brigham Women's Hospital, Columbia University, Mayo Clinic, University of California San Francisco, University of Utah, and Yale University). Participants from the Albert Einstein College of Medicine/Montefiore Medical Center site in New York were enrolled at the Columbia University site in New York (Fig. 1). Baseline data from 428 women who either declined to participate in KEEPS continuation or could not be located were available from the original KEEPS database and were considered as a nonparticipant group.
FIG. 1.
KEEPS continuation participation flow chart. *Coronavirus-related concerns, inconvenient to travel, scheduling conflicts, or moved. KEEPS, Kronos Early Estrogen Prevention Study; oCEE, oral conjugated equine estrogen; PI, principal investigator; tE2, transdermal 17β-estradiol.
Outcomes
Demographic variables including race, ethnicity, and education level were available from the KEEPS baseline database.7 Because of deidentification of data for patient confidentiality, date of birth was not available in the original KEEPS database. Date of birth was self-reported during KEEPS continuation. KEEPS continuation participants self-reported on their interval medical history including medications, smoking, diabetes, stroke, fractures, cancers, heart, kidney and liver disease, and type and duration of HT use after KEEPS. Measurements conducted at KEEPS baseline were repeated at the KEEPS continuation study visits. During the in-person visits, height and weight were measured using clinical scales. For KEEPS baseline, the systolic and diastolic blood pressure measurements were taken 10 minutes apart, and the two readings were averaged.7 For the KEEPS continuation blood pressure measurements, three measurements that were 3 minutes apart were averaged. Waist to hip ratio was calculated as a ratio of their waist circumference (smallest part of the waist right above the belly button) divided by the hip circumference (fullest part of the hips/widest part of the buttocks).
Fasting levels of glucose, insulin, high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C), total cholesterol, and triglycerides were measured in the Kronos Science Laboratories at KEEPS baseline and the Mayo Clinic Laboratories at KEEPS continuation using standard clinical protocols. Homeostasis model assessment of insulin resistance (HOMA-IR) was calculated from fasting glucose and insulin levels (HOMA-IR = fasting glucose × fasting insulin/405).11
Statistical analysis
Participant characteristics at KEEPS baseline and at KEEPS continuation were summarized using means and ranges for continuous variables and counts and percentages for categorical variables. The characteristics were compared among oCEE, tE2, and placebo groups in KEEPS using analysis of variance followed by Tukey pairwise tests if significant for continuous variables or Fisher exact tests for categorical variables. Transformations were applied when necessary to meet distributional assumptions in the tests (eg, log transformation was applied to the HOMA-IR). KEEPS continuation was powered to detect small to moderate effect sizes according to Cohen d12 of 0.20 to 0.40, which may imply clinically meaningful differences for the continuous outcomes we studied. The change in outcome variables was calculated by subtracting KEEPS baseline values from values at year 14 for each participant. The change in outcome variables was compared among oCEE, tE2, and placebo groups using analysis of variance followed by Tukey pairwise tests if significant. Statistical analysis was conducted using SAS version 9.4 and R version 4.2.2. Statistical significance was determined by P < 0.05.
RESULTS
Examining the baseline characteristics of the entire KEEPS cohort revealed no differences between those who participated in the KEEPS continuation (n = 299) versus those who did not (n = 428) except for the systolic and diastolic blood pressures being lower (P = 0.04) and HDL-C being higher (P = 0.03) in participants than nonparticipants (Table 1). These baseline characteristics were also similar between participants and nonparticipants in individual treatment groups; hence, better blood pressure control and HDL-C profile in participants compared with nonparticipants in the entire cohort were not differential by treatment group. The only differences in individual treatment groups at baseline were that frequency of smoking was lower in the oCEE group participants (1%) than oCEE group nonparticipants (9%; P = 0.01), and the total cholesterol (P = 0.04) and LDL-C levels (P = 0.01) were higher in the oCEE group participants than oCEE group nonparticipants (Supplemental Table 1, http://links.lww.com/MENO/B178).
TABLE 1.
Baseline characteristics from the original KEEPS trial in participants and nonparticipants of the KEEPS continuation
| Participants (n = 299) |
Nonparticipants (n = 428) |
P | |
|---|---|---|---|
| Age (y)a | 53 (2) | 53 (3) | 0.18 |
| Education level, n (%) | 0.76 | ||
| Grade school | 0 | 3 (0.7) | |
| Graduate/professional degree | 88 (30) | 114 (27) | |
| High school diploma | 22 (7) | 30 (7) | |
| Some college/vocational | 49 (17) | 83 (20) | |
| Some graduate/professional school | 13 (4) | 21 (5) | |
| Some high school | 1 (0.3) | 2 (0.5) | |
| College graduate | 123 (42) | 168 (40) | |
| BMI (kg/m2) | 25.9 (4.3) | 26.3 (4.3) | 0.20 |
| Waist/hip ratio | 0.8 (0.1) | 0.8 (0.1) | 0.13 |
| Current smoker, n (%) | 12 (4) | 24 (7) | 0.09 |
| Systolic BP (mm Hg) | 117 (14) | 120 (15) | 0.04 |
| Diastolic BP (mm Hg) | 74 (9) | 76 (9) | 0.04 |
| Total cholesterol (mg/dL) | 209 (33) | 207 (34) | 0.48 |
| HDL-C (mg/dL) | 73 (14) | 71 (15) | 0.03 |
| LDL-C (mg/dL) | 113 (29) | 110 (27) | 0.17 |
| Triglycerides (mg/dL) | 85 (55) | 89 (56) | 0.33 |
| Glucose (mg/dL) | 80 (9) | 80 (10) | 0.84 |
| Insulin (mcU/mL) | 6.1 (9.4) | 6.1 (8.1) | 0.93 |
| HOMA-IR | 1.3 (2.1) | 1.3 (2.5) | 0.47b |
Characteristics represent data collected between years 2005 and 2008 before the time when women were recontacted for the KEEPS continuation. Data shown are mean (SD) or n (%). P values are from Fisher exact test or Student t test as appropriate.
The statistically significant findings are written in bold format.
BMI, body mass index; BP, blood pressure; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; KEEPS, Kronos Early Estrogen Prevention Study.
aAge was recorded and shared by participant's report of their age and not date of birth in KEEPS. Age was calculated based on date of birth in KEEPS continuation.
bP value is from log-transformed HOMA-IR.
Baseline characteristics of the women enrolled in KEEPS continuation (n = 299) are listed in Table 2. The average age at enrollment was 54 years with a range of 44 to 59 years and was similar across the KEEPS intervention groups. Ethnicity (Hispanic n = 17 [6%]) of participants of KEEPS continuation differed between the KEEPS intervention groups (P = 0.04); 1% of the Hispanic participants were in the oCEE, 9% were in the tE2, and 6% were in the placebo group. However, pairwise between-group comparisons did not reveal any statistically significant differences (oCEE vs placebo, P = 0.12; tE2 vs placebo, P = 0.44; oCEE vs tE2, P = 0.06). Race (African American, n = 17 [6%]; Asian, n = 9 [3%]; White, n = 253 [85%]; declined to answer, n = 20 [7%]) did not differ across the treatment groups (P = 0.23).
TABLE 2.
Characteristics of KEEPS continuation participants at KEEPS baseline by randomization group
| All (n = 299) | oCEE (n = 90) | tE2 (n = 96) | Placebo (n = 113) | Overall P value | |
|---|---|---|---|---|---|
| Age (y) | 54 (2) | 54 (3) | 54 (2) | 54 (2) | 0.77 |
| BMI (kg/m2) | 25.9 (4.3) | 26.1 (4.4) | 25.4 (4.2) | 26.2 (4.3) | 0.39 |
| Waist/hip ratio | 0.8 (0.1) | 0.8 (0.1) | 0.8 (0.1) | 0.8 (0.1) | 0.84 |
| Diabetes, n (%) | 0 | 0 | 0 | 0 | |
| Smoker, n (%) | 12 (4) | 1 (1) | 5 (5) | 6 (5) | 0.22 |
| Systolic BP (mm Hg) | 117 (14) | 118 (14) | 116 (15) | 118 (15) | 0.53 |
| Diastolic BP (mm Hg) | 74 (9) | 75 (8) | 73 (9) | 75 (9) | 0.17 |
| Total cholesterol (mg/dL) | 209 (33) | 213 (32) | 207 (33) | 208 (36) | 0.43 |
| HDL-C (mg/dL) | 73 (14) | 74 (14) | 75 (15) | 72 (14) | 0.39 |
| LDL-C (mg/dL) | 113 (29) | 117 (30) | 110 (29) | 112 (27) | 0.24 |
| Triglycerides (mg/dL) | 85 (55) | 83 (61) | 81 (44) | 88 (59) | 0.64 |
| Glucose (mg/dL) | 80 (9) | 80 (10) | 79 (8) | 80 (10) | 0.96 |
| Insulin (mcU/mL) | 6.1 (9.4) | 4.8 (3.9) | 6.7 (13.8) | 6.7 (7.8) | 0.28 |
| HOMA-IR | 1.3 (2.1) | 1.0 (0.8) | 1.4 (3.1) | 1.4 (1.7) | 0.46a |
| Heart attack or angina, n (%) | 0 | 0 | 0 | 0 | |
| Congestive heart failure, n (%) | 0 | 0 | 0 | 0 | |
| Stroke, n (%) | 0 | 0 | 0 | 0 |
Characteristics represent data collected between years 2005 and 2008 before the time when women were recontacted for the KEEPS continuation study. Data shown are mean (SD) or n (%). Groupwise P values comparing treatment to placebo are from Fisher exact test or analysis of variance followed by Tukey pairwise tests, as appropriate.
BMI, body mass index; BP, blood pressure; HDL-C, high-density lipoprotein cholesterol; KEEPS, Kronos Early Estrogen Prevention Study; LDL-C, low-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance.
aP value is from log-transformed HOMA-IR.
Participant characteristics at the time of enrollment in KEEPS continuation are listed in Table 3. The average participant age was 67 years with an age range of 58 to 73 years. There were no differences in demographic variables or the time interval from KEEPS baseline to KEEPS continuation among the three treatment groups. Self-reported current use of antihypertensive and lipid-lowering therapy was not different among the treatment groups. However, self-reported use of diabetes medications was more common among women who were in the placebo group (n = 9 [8%]) compared with the tE2 group (n = 1 [1%], P = 0.02) and oCEE group (n = 1 [1%], P = 0.045). Furthermore, we observed a higher frequency of diabetes reported by the participants in the placebo group (n = 6 [6%], P = 0.007) compared with the oCEE (n = 0 [0%]) and tE2 groups (n = 0 [0%]). Other medical history, including smoking (ever smoker), stroke, fractures, cancers, heart, kidney and liver disease, measurements of systolic and diastolic blood pressures, waist to hip ratio, laboratory measurements of fasting levels of glucose, insulin, HDL-C, LDL-C, total cholesterol, and triglycerides as well as the HOMA-IR (log transformed) from blood samples collected during the KEEPS continuation visits, was not different among the treatment groups (all P values >0.05). The findings did not change after adjusting the laboratory measurements of fasting levels of glucose, insulin, HDL-C, LDL-C, total cholesterol, and triglycerides as well as the HOMA-IR (log transformed) for age, waist to hip ratio, and smoking status (all P values >0.05).
TABLE 3.
Characteristics of participants at the time of KEEPS continuation
| All (n = 299) | oCEE (n = 90) | tE2 (n = 96) | Placebo (n = 113) | Overall P value | |
|---|---|---|---|---|---|
| Age (y) | 67 (2) | 67 (3) | 67 (2) | 67 (2) | 0.86 |
| Education (y) | 16 (3) | 16 (2) | 16 (2) | 16 (3) | 0.57 |
| Time from KEEPS baseline to continuation (y) | 14 (1) | 14 (1) | 14 (1) | 14 (1) | 0.44 |
| BMI (kg/m2) | 26.8 (5.0) | 26.8 (5.1) | 26.5 (5.0) | 27.0 (5.0) | 0.78 |
| Waist/hip ratio | 0.9 (0.1) | 0.9 (0.1) | 0.9 (0.1) | 0.8 (0.1) | 0.78 |
| Diabetes medication use, n (%) | 11 (4) | 1 (1) | 1 (1) | 9 (8) | 0.01 a |
| Antihypertensive use, n (%) | 88 (29) | 28 (31) | 23 (24) | 37 (33) | 0.35 |
| Lipid modifying medications, n (%) | 60 (20) | 22 (24) | 12 (12) | 26 (23) | 0.07 |
| Diabetes, n (%) | 6 (2) | 0 | 0 | 6 (6) | 0.007 |
| Smoker, n (%) | 15 (5) | 1 (1) | 5 (5) | 9 (8) | 0.07 |
| Systolic BP (mm Hg) | 127 (17) | 126 (17) | 128 (17) | 128 (18) | 0.74 |
| Diastolic BP (mm Hg) | 76 (9) | 76 (10) | 77 (9) | 76 (9) | 0.97 |
| Total cholesterol (mg/dL) | 207 (40) | 205 (38) | 209 (39) | 207 (42) | 0.83 |
| HDL-C (mg/dL) | 68 (18) | 67 (17) | 68 (18) | 69 (17) | 0.69 |
| LDL-C (mg/dL) | 118 (35) | 117 (34) | 119 (35) | 118 (36) | 0.89 |
| Triglycerides (mg/dL) | 107 (50) | 111 (54) | 111 (46) | 102 (50) | 0.34 |
| Glucose (mg/dL) | 94 (11) | 94 (10) | 93 (10) | 94 (12) | 0.88 |
| Insulin (uIU/mL) | 4.6 (4.1) | 4.9 (5.3) | 4.5 (3.8) | 4.4 (3.1) | 0.71 |
| HOMA-IR | 1.0 (1.2) | 1.2 (1.9) | 0.9 (0.7) | 1.0 (0.8) | 0.31b |
| Fracture after KEEPS, n (%)c | 49 (16) | 11 (12) | 18 (19) | 20 (18) | 0.42 |
| Spine or hip fracture after KEEPS, n (%) | 0 | 0 | 0 | 0 | |
| Heart attack or angina, n (%) | 0 | 0 | 0 | 0 | |
| Congestive heart failure, n (%) | 1 (0.3) | 1 (1) | 0 | 0 | |
| Coronary artery disease, n (%) | 3 (1) | 1 (1) | 1 (1) | 1 (1) | >0.99 |
| Stroke, n (%) | 5 (2) | 0 | 2 (2) | 3 (3) | 0.39 |
| Cancer since start of KEEPS, n (%)d | 38 (13) | 10 (11) | 17 (18) | 11 (10) | 0.20 |
| Breast | 11 (4) | 3 (3) | 5 (5) | 3 (3) | 0.69 |
| Ovary | 2 (0.7) | 0 | 1 (1) | 1 (1) | >0.99 |
| Uterus | 4 (1) | 2 (2) | 0 | 2 (2) | 0.47 |
| Other | 29 (10) | 8 (9) | 13 (14) | 8 (7) | 0.30 |
| Age cancer diagnosed | 57 (10) | 57 (9) | 59 (8) | 54 (13) | 0.30 |
| Liver disease, n (%) | 6 (2) | 2 (2) | 3 (3) | 1 (1) | 0.52 |
| Kidney disease, n (%) | 9 (3) | 2 (2) | 1 (1) | 6 (5) | 0.22 |
| Hormone therapy after KEEPS, n (%)e | 56 (19) | 23 (26) | 14 (15) | 19 (17) | 0.14 |
| Vaginal estradiol | 22 (7) | 7 (8) | 5 (5) | 10 (9) | 0.62 |
| Transdermal estradiol | 13 (4) | 5 (6) | 6 (6) | 2 (2) | 0.22 |
| Oral estradiol | 10 (3) | 5 (6) | 1 (1) | 4 (4) | 0.21 |
| Compounded bio-identical estrogen | 7 (2) | 2 (2) | 2 (2) | 3 (3) | >0.99 |
| oCEE | 4 (1) | 4 (4) | 0 | 0 | 0.008 |
Characteristics represent data collected between years 2017 and 2022 when women were recontacted for the KEEPS continuation study. Data shown are mean (SD) or n (%). Groupwise P values comparing treatment to placebo are from Fisher exact test or analysis of variance followed by Tukey pairwise tests, as appropriate.
BMI, body mass index; BP, blood pressure; HDL-C, high-density lipoprotein cholesterol; KEEPS, Kronos Early Estrogen Prevention Study; LDL-C, low-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; oCEE, oral conjugated equine estrogen.
The statistically significant findings are written in bold format.
aPairwise comparison P values between oCEE and placebo, 0.045; tE2 and placebo, 0.02.
bP value is from log-transformed HOMA-IR.
cEleven participants have fractures in multiple places. P value is from “Yes/No” fracture after KEEPS.
dFive participants have cancer in multiple places. P value is from “Yes/No” cancer since KEEPs.
eP value is from “Yes/No” taking hormone therapy after KEEPS.
Because of the higher self-reported use of diabetes medications among the placebo group participants compared with treatment groups and because diabetes medications may modify the metabolic outcomes of fasting levels of insulin, glucose, and HOMA-IR, we performed a sensitivity analysis by removing the participants on diabetes medications (n = 11) and reanalyzed the group differences for these three laboratory measurements. Removing participants who were taking diabetes medications at the time of enrollment in KEEPS continuation did not alter the results (all P values >0.05).
A notable number of the participants (n = 56 [19%]) continued with HT regimens used in the KEEPS trial or switched to another type of HT after the end of KEEPS; the proportion of participants using HT after KEEPS did not differ among treatment groups (P = 0.14). The most common formulation and route of administration were vaginal estradiol (n = 22 [7%]). Others included tE2 (n = 13 [4%]), oral estradiol (n = 10 [3%]), oral compounded bio-identical estrogen (n = 7 [2%]), and oCEE (n = 4 [1%]). All participants who received oCEE after KEEPS were from the group randomized to oCEE. Vaginal estrogens were used at varying time points after KEEPS within a 0- to 13-year time frame. A majority of the participants taking postintervention oral and transdermal HT started treatment within 5 years after KEEPS ended (n = 24 [70%]), and the remainder started HT more than 5 years after the end of KEEPS (n = 10 [30%]). Most of the participants who started HT after the end of KEEPS were still taking HT at the time of KEEPS continuation (n = 28 [82%]). Because HT after KEEPS may have influenced the outcomes that were investigated, we conducted a sensitivity analysis by removing the participants who used HT after KEEPS. Analysis of KEEPS continuation outcomes in participants who did not use HT after KEEPS (n = 243) was repeated. The findings from this sensitivity analyses did not differ from the main analyses (Supplemental Table 2, http://links.lww.com/MENO/B178). In this supplemental analysis, we further investigated the laboratory measurements of fasting levels of glucose, insulin, HDL-C and LDL-C, total cholesterol, and triglycerides as well as the HOMA-IR (log transformed) after adjusting for age, waist to hip ratio, and smoking status, and the results were similar (all P values >0.05). Similarly, we did not find a difference in fasting levels of insulin, glucose, or HOMA-IR after removing participants using diabetes medications (all P values >0.05) in this sensitivity analysis.
Changes in body mass index, waist to hip ratio, systolic and diastolic blood pressures, and laboratory measurements of fasting levels of glucose, insulin, HDL-C and LDL-C, total cholesterol, and triglycerides as well as the HOMA-IR (log transformed) over 14 years from KEEPS baseline to KEEPS continuation were compared among the treatment groups. There was no evidence that there were differences among the treatment groups on the change in these outcomes (P > 0.05) (Table 4). We repeated this analysis in participants who did not use HT after KEEPS (n = 243), and the results did not change qualitatively (Supplemental Table 3, http://links.lww.com/MENO/B178). In addition, there was no difference in fasting levels of insulin, glucose, or HOMA-IR after removing participants on diabetes medications at the time of KEEPS continuation in either the main analysis or the supplemental analysis (all P values >0.05).
TABLE 4.
Change in cardiovascular and metabolic risk factors from KEEPS baseline (during years 2005 to 2008) to KEEPS continuation (during years 2017 to 2022) over an average period of 14 years
| oCEE (n = 90) | tE2 (n = 96) | Placebo (n = 113) | P | |
|---|---|---|---|---|
| Age (y) | 14 (1) | 14 (1) | 14 (1) | 0.44 |
| BMI (kg/m2) | 0.7 (3.3) | 1.0 (3.0) | 0.7 (2.9) | 0.71 |
| Waist/hip ratio | 0.04 (0.1) | 0.04 (0.1) | 0.04 (0.1) | 0.95 |
| Systolic BP (mm Hg) | 8 (19) | 12 (17) | 9 (20) | 0.36 |
| Diastolic BP (mm Hg) | 2 (10) | 4 (11) | 1 (11) | 0.20 |
| Total cholesterol (mg/dL) | −8 (38) | 2 (42) | −0.5 (45) | 0.25 |
| HDL-C (mg/dL) | −8 (12) | −7 (14) | −4 (13) | 0.06 |
| LDL-C (mg/dL) | −1 (35) | 10 (36) | 7 (40) | 0.16 |
| Triglycerides (mg/dL) | 27 (68) | 30 (53) | 14 (52) | 0.09 |
| Glucose (mg/dL) | 13.9 (10.8) | 14.3 (10.8) | 14.0 (9.9) | 0.97 |
| Insulin (mcU/mL) | 0.1 (5.9) | −2.1 (13.6) | −2.5 (6.8) | 0.11 |
| HOMA-IR | 0.2 (2.0) | −0.4 (3.2) | −0.3 (1.3) | 0.11a |
Data shown are mean (SD). Groupwise P values comparing treatment to placebo are from analysis of variance followed by Tukey pairwise tests, as appropriate.
BMI, body mass index; BP, blood pressure; HDL-C, high-density lipoprotein cholesterol; KEEPS, Kronos Early Estrogen Prevention Study; LDL-C, low-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance.
aP value is from log-transformed HOMA-IR.
DISCUSSION
In KEEPS continuation, a cohort of women was evaluated approximately 14 years after randomization in a 4-year HT trial examining two different HT regimens of oCEE and tE2 with 12 days of progesterone each month versus placebo. Women who participated in KEEPS continuation had slightly lower blood pressures and higher HDL-C at KEEPS baseline than those who did not participate, but these differences were not specific to one of the treatment groups. The remainder of participant characteristics and cardiovascular risk factors were comparable between participants and nonparticipants in the KEEPS continuation study. Overall, KEEPS continuation participants had good cardiometabolic health at the time of randomization, with no prior history of cardiovascular diseases or diabetes. The distribution of KEEPS continuation participants was comparable across three treatment allocation groups of KEEPS.
The main findings of the current study were that, in postmenopausal participants of the continuation study who were a decade remote from participation in KEEPS, the clinical (blood pressure, body mass index, waist to hip ratio) and biochemical (fasting levels of glucose, insulin, HOMA-IR, HDL-C, LDL-C, total cholesterol, or triglyceride) indices of cardiometabolic health were comparable between those who a decade prior had completed a 4-year intervention of oral versus transdermal HT or placebo in KEEPS. The self-reported interval occurrences of stroke, fracture, cancer, heart, kidney, and liver disease between completion of KEEPS to enrollment in KEEPS continuation were also comparable across the randomized treatment groups in KEEPS, except for the higher frequency of diabetes and higher use of diabetes medications observed in the placebo group compared with the oCEE and tE2 treatment groups. However, this observation of higher rates of self-reported diabetes and the use of diabetes medications in the placebo group compared with the treatment groups was not supported by the fasting levels of insulin and glucose, and measurements of insulin resistance such as HOMA-IR even after removing the participants on diabetes medications. Thus, in otherwise healthy postmenopausal women, we did not observe any evidence of cardiometabolic risk mitigation that could be attributed to the 4-year intervention of HT use in KEEPS, nor did we find any evidence of adverse effects on cardiometabolic risk factors.
It is important to recognize that KEEPS enrolled women with good cardiovascular health at the time of randomization; women with cardiovascular disease and diabetes were excluded, and they were screened out for coronary calcium scores >50 or hyperlipidemia.7 After 4 years of KEEPS, the effects on atherosclerosis (assessed by carotid intima-medial thickness and coronary artery calcium measurements) were neutral, and a few clinical cardiovascular disease events occurred during the intervention phase.8 Our data in KEEPS continuation indicate that the effects of HT on cardiovascular health remained neutral across the treatment groups because KEEPS participants aged another decade and reached the mean age of 67 years. The only other HT trial that followed participants for more than a decade postintervention is the WHI. The WHI hormone trials had enrolled significantly older women than KEEPS and found greater risk of breast cancer, stroke, and dementia.6,13
In the Early versus Late Intervention Trial with Estradiol (ELITE),14 oral estradiol treatment given to women who were less than 6 years past menopause at the time of randomization slowed the increase of carotid intima-medial thickness, but the effects were neutral among women who started HT 10 years or more past menopause.15 Unlike KEEPS, the ELITE study implies that early intervention may be beneficial for long-term cardiovascular outcomes in women who were less than 6 years past menopause. Our findings in KEEPS continuation in part agree with WHI and ELITE in that the morbidities and risks associated with HT in an older cohort do not apply to recently postmenopausal women with good cardiovascular health.
During the 4 years of the KEEPS trial, HDL-C and LDL-C cholesterol levels improved with oCEE and insulin resistance measured with HOMA-IR decreased with tE2.8 In the current KEEPS continuation, the beneficial metabolic findings were no longer present. The current data suggest that the effects of HT on HDL-C, LDL-C cholesterol levels, and insulin resistance were limited to the time when HT was in use. Furthermore, the change in these the cardiometabolic outcomes over a 14-year time frame did not differ among the treatment groups, indicating that the effects of 4 years of HT on cardiovascular outcomes in women of good cardiovascular health are neutral in the long-term, consistent with the concept of activational (ie, reversible) effects of hormones on cellular processes.16-19
Our study has several strengths. First, we followed a large cohort of postmenopausal women of good cardiovascular health who were randomized to 4 years of treatment comparing two types of hormone therapies and placebo over a 14-year period in a multicenter setting. Second, we recorded the use of HT after KEEPS ended and performed a sensitivity analysis in participants who were not exposed to HT after KEEPS. Findings from the sensitivity analysis were not different from the findings in the entire cohort; thus, results were not a consequence of continued exposure to HT after KEEPS. Finally, although there is a 14-year difference between the KEEPS baseline and KEEPS-continuation laboratory measurements that were performed in two different laboratories, we were able to investigate the change in measurements in individual participants and test for the differences among the treatment groups.
There are limitations to this study that need to be acknowledged. First, we were able to enroll only 41% of the KEEPS cohort to KEEPS continuation. A good majority of enrollment took place during the coronavirus disease 2019 pandemic (2020-2022), which detrimentally impacted recruitment efforts. Participants were reluctant to travel and participate at the sites, and procedures were delayed or canceled with institutional closures, causing numerous scheduling difficulties. Nonetheless, KEEPS continuation recruited 299 participants, which is a relatively large cohort for the follow-up of a clinical trial that randomized to an intervention 14 years earlier. KEEPS continuation participants had a slightly higher HDL-C and lower blood pressures than nonparticipants at KEEPS baseline, and the coronavirus disease 2019 pandemic–related concerns may have further played a role in the recruitment of a slightly healthier participants of the KEEPS cohort, but these differences were not specific to one of the treatment groups, and they may not be clinically meaningful. In addition, the information on health outcomes including diabetes was collected through self-reports, which may be subject to recall bias. Finally, the current study is underpowered to detect differences among treatment groups for rare events that occurred in this cohort such as incident cancers, liver, kidney, and heart diseases.
CONCLUSIONS
KEEPS continuation was designed to determine the long-term risks and benefits of HT use on Alzheimer disease pathophysiology and cognitive health. The findings from this primary aim of KEEPS continuation will be reported when the data analysis is completed. There was no evidence of long-term cardiovascular or metabolic benefits, nor any evidence to suggest adverse effects associated with remote exposure to HT initiated proximate to onset of menopause and administered over a 4-year period in overall healthy postmenopausal women with good cardiovascular health. The finding of a lower frequency of self-reported diabetes and use of diabetes medications in the HT groups should be interpreted cautiously, because this finding was not supported by laboratory measurements of fasting levels of glucose, insulin, or HOMA-IR. These data have important implications for women considering hormone therapies soon after menopause for vasomotor symptoms or genitourinary syndrome, as well as for maintenance of bone mineral density and prevention of osteoporosis. Applicability of these findings from KEEPS continuation is restricted to women who have good cardiovascular health. Further research is needed to determine the long-term effects of menopausal hormone treatment for women who are deemed to be at elevated risk for cardiometabolic disorders such as heart disease, stroke, and diabetes.
Supplementary Material
Footnotes
Funding/support: This study was funded by the NIH RF1AG57547, Alzheimer Drug Discovery foundation, and Aurora Foundation to the Kronos Longevity Research Institute.
Financial disclosure/conflicts of interest: Dr Kantarci served on the data safety monitoring board for Pfizer Inc. and Takeda Global Research & Development Center, Inc. She received research support from Avid Radiopharmaceuticals, Eli Lilly. She consults for Biogen. Dr Santoro is a member of the Scientific Advisory Boards for Astellas, Que Oncology, Amazon (Ember), and Menogenix, Inc. She is a consultant for Ansh Labs and has received grant support (to her institution) from Menogenix, Inc. Dr Kapoor is funded in part by the National Institute on Aging (NIA grant U54 AG044170). Dr Kapoor has no conflicts of interest directly related to the subject of this manuscript. However, over the past 36 months, she has had the following conflicts of interest: She has been a consultant for Astellas and Mithra Pharmaceuticals, Scynexis, and Womaness. She receives grant support form Mithra Pharmaceuticals. She has received payment for development of educational content from Med Learning Group and Academy of Continued Healthcare Learning. She has received honoraria for CME activity from PriMed and OBG Management. Dr Pal is a member of advisory board for Flo Health, receives ongoing funding from Springer, and has received funding in the past from Ferring. Dr Fields receives institutional grant funding from the National Institute of Health for three additional projects. Dr Malek-Ahmadi receives grant funding from Biomedical Research Alliance of New York, LLC. The other authors have nothing to disclose.
Supplemental digital content is available for this article. Direct URL citations are provided in the HTML and PDF versions of this article on the journal’s Website (www.menopause.org).
Contributor Information
Nirubol Tosakulwong, Email: Tosakulwong.Nirubol@mayo.edu.
Timothy G. Lesnick, Email: lesnick@mayo.edu.
Firat Kara, Email: Kara.Firat@mayo.edu.
June Kendall-Thomas, Email: KendallThomas.June15@mayo.edu.
Ekta Kapoor, Email: kapoor.ekta@mayo.edu.
Julie A. Fields, Email: Fields.Julie@mayo.edu.
Taryn T. James, Email: ttjames@medicine.wisc.edu.
Rogerio A. Lobo, Email: ral35@columbia.edu.
JoAnn E. Manson, Email: jmanson@rics.bwh.harvard.edu.
Lubna Pal, Email: lubna.pal@yale.edu.
Dustin B. Hammers, Email: hammersd@iu.edu.
Michael Malek-Ahmadi, Email: michael.MalekAhmadi@bannerhealth.com.
Marcelle I. Cedars, Email: marcelle.cedars@ucsf.edu.
Frederick N. Naftolin, Email: fnaftolin2@gmail.com.
Nanette Santoro, Email: nanette.santoro@cuanschutz.edu.
Virginia M. Miller, Email: millerv60@gmail.com.
Sherman M. Harman, Email: Sherman.harman@va.gov.
N. Maritza Dowling, Email: nmdowling@gwu.edu.
Carey E. Gleason, Email: ceg@medicine.wisc.edu.
REFERENCES
- 1.The Hormone Therapy Position Statement of The North American Menopause Society Advisory Panel . The 2022 hormone therapy position statement of The North American Menopause Society. Menopause 2022;29:767–794. doi: 10.1097/GME.0000000000002028 [DOI] [PubMed] [Google Scholar]
- 2.Banks E Beral V Reeves G Balkwill A Barnes I, Million Women Study Collaborators . Fracture incidence in relation to the pattern of use of hormone therapy in postmenopausal women. JAMA 2004;291:2212–2220. doi: 10.1001/jama.291.18.2212 [DOI] [PubMed] [Google Scholar]
- 3.Christiansen C. Prevention and treatment of osteoporosis with hormone replacement therapy. Int J Fertil Menopausal Stud 1993;38(Suppl 1):45–54. [PubMed] [Google Scholar]
- 4.Rossouw JE Anderson GL Prentice RL, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women's Health Initiative randomized controlled trial. JAMA 2002;288:321–333. doi: 10.1001/jama.288.3.321 [DOI] [PubMed] [Google Scholar]
- 5.Craig MC, Maki PM, Murphy DG. The Women's Health Initiative Memory Study: findings and implications for treatment. Lancet Neurol 2005;4:190–194. doi: 10.1016/S1474-4422(05)01016-1 [DOI] [PubMed] [Google Scholar]
- 6.Espeland MA Shumaker SA Leng I, et al. Long-term effects on cognitive function of postmenopausal hormone therapy prescribed to women aged 50 to 55 years. JAMA Intern Med 2013;173:1429–1436. doi: 10.1001/jamainternmed.2013.7727 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Harman SM Brinton EA Cedars M, et al. KEEPS: the Kronos Early Estrogen Prevention Study. Climacteric 2005;8:3–12. doi: 10.1080/13697130500042417 [DOI] [PubMed] [Google Scholar]
- 8.Harman SM Black DM Naftolin F, et al. Arterial imaging outcomes and cardiovascular risk factors in recently menopausal women: a randomized trial. Ann Intern Med 2014;161:249–260. doi: 10.7326/M14-0353 [DOI] [PubMed] [Google Scholar]
- 9.Gleason CE Dowling NM Wharton W, et al. Effects of hormone therapy on cognition and mood in recently postmenopausal women: findings from the randomized, controlled KEEPS-Cognitive and Affective Study. PLoS Med 2015;12:e1001833; discussion e1001833. doi: 10.1371/journal.pmed.1001833 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Miller VM Naftolin F Asthana S, et al. The Kronos Early Estrogen Prevention Study (KEEPS): what have we learned? Menopause 2019;26:1071–1084. doi: 10.1097/GME.0000000000001326 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wallace TM, Levy JC, Matthews DR. Use and abuse of HOMA modeling. Diabetes Care 2004;27:1487–1495. doi: 10.2337/diacare.27.6.1487 [DOI] [PubMed] [Google Scholar]
- 12.Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed,. Hillsdale, NJ: Lawrence Erlbaum Associates Publishers; 1988. [Google Scholar]
- 13.Vaughan L Espeland MA Snively B, et al. The rationale, design, and baseline characteristics of the Women's Health Initiative Memory Study of Younger Women (WHIMS-Y). Brain Res 2013;1514:3–11. doi: 10.1016/j.brainres.2013.03.047 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Henderson VW St John JA Hodis HN, et al. Cognitive effects of estradiol after menopause: a randomized trial of the timing hypothesis. Neurology 2016;87:699–708. doi: 10.1212/WNL.0000000000002980 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Hodis HN Mack WJ Henderson VW, et al. Vascular effects of early versus late postmenopausal treatment with estradiol. N Engl J Med 2016;374:1221–1231. doi: 10.1056/NEJMoa1505241 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Coker LH Espeland MA Hogan PE, et al. Change in brain and lesion volumes after CEE therapies: the WHIMS-MRI studies. Neurology 2014;82:427–434. doi: 10.1212/WNL.0000000000000079 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Resnick SM Espeland MA Jaramillo SA, et al. Postmenopausal hormone therapy and regional brain volumes: the WHIMS-MRI Study. Neurology 2009;72:135–142. doi: 10.1212/01.wnl.0000339037.76336.cf [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kantarci K Tosakulwong N Lesnick TG, et al. Effects of hormone therapy on brain structure: a randomized controlled trial. Neurology 2016;87:887–896. doi: 10.1212/WNL.0000000000002970 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kantarci K Tosakulwong N Lesnick TG, et al. Brain structure and cognition 3 years after the end of an early menopausal hormone therapy trial. Neurology 2018;90:e1404–e1412. doi: 10.1212/WNL.0000000000005325 [DOI] [PMC free article] [PubMed] [Google Scholar]