Advertisement

Effects of high-intensity training on cardiovascular risk factors in premenopausal and postmenopausal women

Published:December 23, 2016DOI:https://doi.org/10.1016/j.ajog.2016.12.017

      Background

      Menopause is associated with increased risk of cardiovascular disease and the causal factors have been proposed to be the loss of estrogen and the subsequent alterations of the hormonal milieu. However, which factors contribute to the deterioration of cardiometabolic health in postmenopausal women is debated as the menopausal transition is also associated with increased age and fat mass. Furthermore, indications of reduced cardiometabolic adaptations to exercise in postmenopausal women add to the adverse health profile.

      Objective

      We sought to evaluate risk factors for type 2 diabetes and cardiovascular disease in late premenopausal and early postmenopausal women, matched by age and body composition, and investigate the effect of high-intensity training.

      Study Design

      A 3-month high-intensity aerobic training intervention, involving healthy, nonobese, late premenopausal (n = 40) and early postmenopausal (n = 39) women was conducted and anthropometrics, body composition, blood pressure, lipid profile, glucose tolerance, and maximal oxygen consumption were determined at baseline and after the intervention.

      Results

      At baseline, the groups matched in anthropometrics and body composition, and only differed by 4.2 years in age (mean [95% confidence limits] 49.2 [48.5-49.9] vs 53.4 [52.4-54.4] years). Time since last menstrual period for the postmenopausal women was (mean [95% confidence limits] 3.1 [2.6-3.7] years). Hormonal levels (estrogen, follicle stimulation hormone, luteinizing hormone) confirmed menopausal status. At baseline the postmenopausal women had higher total cholesterol (P < .001), low-density lipoprotein-cholesterol (P < .05), and high-density lipoprotein-cholesterol (P < .001) than the premenopausal women. The training intervention reduced body weight (P < .01), waist circumference (P < .01), and improved body composition by increasing lean body mass (P < .001) and decreasing fat mass (P < .001) similarly in both groups. Moreover, training resulted in lower diastolic blood pressure (P < .05), resting heart rate (P < .001), total cholesterol (P < .01), low-density lipoprotein-cholesterol (P < .01), total cholesterol/high-density lipoprotein-cholesterol index (P < .01), and improved plasma insulin concentration during the oral glucose tolerance test (P < .05) in both groups.

      Conclusion

      Cardiovascular risk factors are similar in late premenopausal and early postmenopausal women, matched by age and body composition, with the exception that postmenopausal women have higher high- and low-density lipoprotein-cholesterol levels. A 3-month intervention of high-intensity aerobic training reduces risk factors for type 2 diabetes and cardiovascular disease to a similar extent in late premenopausal and early postmenopausal women.

      Key words

      Introduction

      The menopausal transition is accompanied by metabolic changes and increasing prevalence of metabolic syndrome (MetS), which is defined as copresence of abdominal obesity, hypertension, dyslipidemia, and insulin resistance.
      • Carr M.C.
      The emergence of the metabolic syndrome with menopause.
      • Ross L.A.
      • Polotsky A.J.
      Metabolic correlates of menopause: an update.
      MetS is associated with development of type 2 diabetes (T2D) (relative risk [RR] 5.0), cardiovascular disease (CVD) (RR 2.35), and all-cause mortality (RR 1.86),
      • O’Neill S.
      • O’Driscoll L.
      Metabolic syndrome: a closer look at the growing epidemic and its associated pathologies.
      and globally it is estimated that 30-55% of postmenopausal women fulfill the diagnostic criteria for MetS.
      • Stefanska A.
      • Bergmann K.
      • Sypniewska G.
      Metabolic syndrome and menopause: pathophysiology, clinical and diagnostic significance.
      It is debated whether the increased prevalence of MetS after menopause is due to hormonal changes, a normal phenomenon of aging, or a consequence of gain in weight and fat-mass during and after the menopausal transition.
      • Matthews K.A.
      • Crawford S.L.
      • Chae C.U.
      • et al.
      Are changes in cardiovascular disease risk factors in midlife women due to chronological aging or to the menopausal transition?.
      It is therefore highly relevant to evaluate factors associated with MetS in nonobese premenopausal and postmenopausal women matched by age and body composition.
      Physical exercise increases cardiorespiratory fitness and reduces the risk of MetS
      • Lin X.
      • Zhang X.
      • Guo J.
      • et al.
      Effects of exercise training on cardiorespiratory fitness and biomarkers of cardiometabolic health: a systematic review and meta-analysis of randomized controlled trials.
      • Pattyn N.
      • Cornelissen V.A.
      • Eshghi S.R.T.
      • Vanhees L.
      The effect of exercise on the cardiovascular risk factors constituting the metabolic syndrome: a meta-analysis of controlled trials.
      but the ability of postmenopausal women to respond to exercise training has been debated. Some cross-sectional studies suggest that postmenopausal women have a reduced fitness level compared to premenopausal women
      • Lynch N.A.
      • Ryan A.S.
      • Berman D.M.
      • Sorkin J.D.
      • Nicklas B.J.
      Comparison of VO2max and disease risk factors between perimenopausal and postmenopausal women.
      • Mercuro G.
      • Saiu F.
      • Deidda M.
      • Mercuro S.
      • Vitale C.
      • Rosano G.M.C.
      Impairment of physical exercise capacity in healthy postmenopausal women.
      and a blunted response to exercise-induced central and peripheral cardiovascular modulations.
      • Parker B.A.
      • Kalasky M.J.
      • Proctor D.N.
      Evidence for sex differences in cardiovascular aging and adaptive responses to physical activity.
      Another study finds similar beneficial effects of brisk walking on body composition and glucose metabolism, irrespective of menopausal status, in overweight to obese women
      • Riesco E.
      • Tessier S.
      • Lacaille M.
      • et al.
      Impact of a moderate-intensity walking program on cardiometabolic risk markers in overweight to obese women: is there any influence of menopause?.
      and a review, investigating the effects of exercise training in early postmenopausal women, found increases in V˙o2max (ranging from 4-32%), diverse effects of exercise on blood pressure (BP) in normotensive women but a lowering effect in hypertensive women, as well as reductions in plasma lipids in dyslipidemic but not in normolipidemic women.
      • Asikainen T.-M.
      • Kukkonen-Harjula K.
      • Miilunpalo S.
      Exercise for health for early postmenopausal women: a systematic review of randomized controlled trials.
      However, very few studies have compared the effect of physical activity on the risk of MetS in premenopausal vs postmenopausal women. The aim of this study was therefore to investigate the effect of a well-controlled high-intensity aerobic training program on risk factors predisposing to T2D and CVD in nonobese, early postmenopausal women and to compare the effect of the training intervention to the effect in late premenopausal women differing, on average, only by 4 years of age.

      Materials and Methods

      Overall study design

      The work was carried out as part of the research program Copenhagen Women Study (cws.ku.dk) funded by the University of Copenhagen Excellence Program for Interdisciplinary Research. This article presents data from work package II where late premenopausal and early postmenopausal women were assigned to 3 months of high-intensity aerobic training, performed as spinning. All participants underwent a health examination before inclusion as well as examinations at baseline and after 3 months, comprising physiological, psychological, and sociological tests (Figure 1). All women participated in the general tests on day 1 and 2 at baseline and test day 4 and 5 after 3 months. At test day 3 (baseline) and 6 (3 months) specific investigations of either cardiovascular function
      • Nyberg M.
      • Egelund J.
      • Mandrup C.M.
      • et al.
      Early postmenopausal phase is associated with reduced prostacyclin-induced vasodilation that is reversed by exercise training.
      (n = 42 women) or adipose tissue and skeletal muscle metabolic function (n = 41 women) were conducted. This article covers the results of the general physiological tests, and all presented outcomes were a priori defined as secondary outcomes in clinicaltrials.gov, registration number: NCT02135575.
      Figure thumbnail gr1
      Figure 1Overall study design of Copenhagen Women Study, menopause
      Horizontal timeline and vertical overview of different examinations. This article only includes physiological data.
      Mandrup et al. High-intensity training for postmenopausal health. Am J Obstet Gynecol 2017.

      Recruitment

      Late premenopausal (n = 43) and early postmenopausal (n = 40) women were recruited from the Copenhagen area through newspaper advertisements. Eligibility was assessed upon first contact by telephone or mail, secondly by evaluation of an online questionnaire, and finally at a health examination (Figure 2). Recruitment was conducted through 4 rounds from August 2013 through August 2015, and an almost equal number of premenopausal and postmenopausal women were recruited in each round to prevent seasonal and investigator-dependent variations. All participants received written and oral information about the study, including risks and discomforts associated with participation, before they gave their written consent to participate. The study was conducted according to the Helsinki Declaration and approved by the ethical committee in the capital region of Denmark, protocol no. H-1-2012-150.
      Figure thumbnail gr2
      Figure 2Flow diagram of inclusion and exclusion
      Participants before and after start of high-intensity aerobic training intervention.
      Mandrup et al. High-intensity training for postmenopausal health. Am J Obstet Gynecol 2017.

      Participants

      Inclusion criteria were healthy, sedentary, normal-weight to overweight (body mass index [BMI] 18.5-30 kg/m2) women, 45-57 years of age, who were either late premenopausal (regular bleeding and plasma estradiol [E2] in the normal fertile range; follicular phase 0.05-0.51 nmol/L, mid cycle 0.32-1.83 nmol/L, luteal phase 0.16-0.78 nmol/L, and plasma follicle-stimulating hormone [FSH] <20 IU/L) or early postmenopausal (no bleeding for at least 1 year, E2 <0.20 nmol/L and FSH 22-138 IU/L). Being sedentary was defined as performing <2 hours of physical activity per week during the last 2 years, and the definition was also supported by a V˙o2max <40 mL O2/min/kg. Exclusion criteria were smoking, use of hormonal contraception, excessive alcohol intake, diagnosis of hypertension or any other chronic disease, daily intake of medication, or blood samples (screening for liver, kidney, and bone-marrow function) outside of normal range. Characteristics of the participants are presented in Table 1.
      Table 1Participant characteristics before and after 3-month high-intensity aerobic training intervention
      VariablesPremenopausalPostmenopausal
      Baselinen3 monthsnBaselinen3 mon
      Anthropometrics
      Age, y
      Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group. Effort of intervention was assessed by 2-way analysis of variance
      49.2 (48.5–49.9)3953.4 (52.4–54.4)38
      Height, m1.68 (1.66–1.70)381.67 (1.65–1.69)37
      Weight, kg
      Statistically significant (P ≤ .05) difference from baseline to 3 mo
      67.7 (65.5–70.0)3867.1 (64.9–69.3)3866.4 (63.6–69.1)3865.8 (62.9–68.7)37
      BMI, kg/m2
      Statistically significant (P ≤ .05) difference from baseline to 3 mo
      23.9 (23.2–24.7)3823.7 (23.0–24.4)3823.7 (22.9–24.4)3723.5 (22.6–24.3)37
      Waist circumference, cm
      Statistically significant (P ≤ .05) difference from baseline to 3 mo
      80 (78–82)3279 (77–81)3279 (76–81)3278 (76–81)31
      Body composition
      Lean body mass, kg
      Statistically significant (P ≤ .05) difference from baseline to 3 mo
      43.5 (42.0–44.9)3844.1 (42.7–45.5)3842.7 (41.3–44.1)3843.2 (41.7–44.7)37
      Fat mass, kg
      Statistically significant (P ≤ .05) difference from baseline to 3 mo
      24.3 (22.8–25.7)3823.0 (21.5–24.5)3823.6 (21.8–25.5)3822.6 (20.6–24.6)37
      Fat, %
      Statistically significant (P ≤ .05) difference from baseline to 3 mo
      35.7 (34.2–37.1)3834.1 (32.7–35.6)3835.3 (33.7–36.9)3833.9 (32.1–35.7)37
      Android fat, %
      Statistically significant (P ≤ .05) difference from baseline to 3 mo
      38.4 (35.8–41.0)3836.4 (33.7–39.0)3835.9 (32.8–39.1)3834.3 (30.9–37.6)37
      Gynoid fat, %
      Statistically significant (P ≤ .05) difference from baseline to 3 mo
      41.5 (40.0–43.0)3839.4 (38.0–40.9)3841.9 (40.5–43.2)3839.7 (38.1–41.4)37
      Android-gynoid fat ratio0.92 (0.87–0.98)380.92 (0.86–0.98)380.85 (0.79–0.91)380.85 (0.79–0.92)37
      Hormones
      P-estradiol, nmol/L0.61 (0.28–1.13)390.54 (0.37–0.98)320.04 (0.04–0.32)380.04 (0.04–0.09)32
      P-follitropin [FSH], IU/L
      Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group. Effort of intervention was assessed by 2-way analysis of variance
      ,
      Statistically significant (P ≤ .05) difference from baseline to 3 mo
      ,
      Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group
      8.5 (5.3–16.3)398.3 (5.4–13.8)3290.0 (74.2–112.0)3879.8 (65.5–103.0)32
      P-lutropin [LH], IU/L
      Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group. Effort of intervention was assessed by 2-way analysis of variance
      ,
      Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group
      11.4 (8.6–15.2)3911.0 (8.0–15.1)3238.3 (35.0–41.9)3833.8 (30.1–38.1)32
      Blood pressure
      Systolic, mm Hg
      Statistically significant (P ≤ .05) interaction between time and menopausal status, meaning that 2 groups responded differently to intervention.
      107 (99–117)39108 (103–118)38111 (103–120)38107 (99–120)37
      Diastolic, mm Hg
      Statistically significant (P ≤ .05) difference from baseline to 3 mo
      70 (66–75)3970 (66–76)3871 (66–79)3870 (65–75)37
      Mean arterial pressure, mm Hg
      Statistically significant (P ≤ .05) interaction between time and menopausal status, meaning that 2 groups responded differently to intervention.
      82 (76–88)3982 (78–90)3884 (79–92)3882 (77–88)37
      Resting heart rate, beats/min–1
      Statistically significant (P ≤ .05) difference from baseline to 3 mo
      66 (61–72)3865 (58–71)3864 (59–68)3861 (56–66)37
      Maximal heart rate, beats/min–1
      Statistically significant (P ≤ .05) difference from baseline to 3 mo
      176 (172–180)35175 (172–177)38175 (171–180)36173 (169–176)34
      Data are mean (95% confidence limits). No statistical analysis has been performed for estrogen as most measurements for postmenopausal women were lower than detection limit. In first inclusion round, waist circumference and postintervention analysis of hormonal levels were not performed, resulting in lower n for those parameters.
      BMI, body mass index; FSH, follicle-stimulating hormone; LH, luteinizing hormone.
      Mandrup et al. High-intensity training for postmenopausal health. Am J Obstet Gynecol 2017.
      Baseline differences assessed by independent t test:
      a Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group. Effort of intervention was assessed by 2-way analysis of variance
      b Statistically significant (P ≤ .05) difference from baseline to 3 mo
      c Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group
      d Statistically significant (P ≤ .05) interaction between time and menopausal status, meaning that 2 groups responded differently to intervention.

      Exercise training intervention

      The intervention consisted of 3 months of high-intensity aerobic training (spinning), conducted 3 times/wk for approximately 1 hour. Two weekly sessions were conducted in our exercise training facility by instructors from our research group and in the beginning of every round, a medical doctor attended the spinning classes. The sessions comprised a warm-up, 3 blocks of varying intervals, with multiple periods of maximum performance, followed by a cool-down period (Figure 3). The intensity of the training sessions increased gradually during the 3-month period. During the sessions, the instructor and the participants were able to monitor their own and their peers’ heart rate (HR) on a big screen, given as a percentage of their individual maximal HR (HRmax). One weekly session took place in a local fitness center. During all training sessions the participants wore HR monitors (FT2; Polar, Kempele, Finland).
      Figure thumbnail gr3
      Figure 3Graphic illustration of training sessions
      Duration and intensity of the high-intensity training intervention.
      HRmax, maximal heart rate.
      Mandrup et al. High-intensity training for postmenopausal health. Am J Obstet Gynecol 2017.

      Measurements and analyses

      Anthropometrics were assessed using a tape measure for hip and waist circumference and a stadiometer for height. Body composition was assessed by dual x-ray absorptiometry scanning (Lunar iDXA; GE Healthcare, Little Chalfont, United Kingdom) at Department of Clinical Physiology, Nuclear Medicine, and Positron Emission Tomography, Rigshospitalet, Glostrup, Denmark, by an investigator blinded for menopausal status. All scans were performed by the same investigator at baseline and after 3 months. Blood samples were obtained from the antecubital vein by a BD Vacutainer system (Becton-Dickinson, Plymouth, United Kingdom) and analyzed at Department of Clinical Biochemistry, Rigshospitalet, Copenhagen, Denmark, by investigators blinded for menopausal status. Plasma E2 was analyzed by competitive electrochemiluminescence immunoassay (ECLIA), and FSH and luteinizing hormone were analyzed by sandwich ECLIA (Cobas 8000, e602 module; F. Hoffmann-La Roche Ltd, Rotkreuz, Switzerland). Plasma high-density lipoprotein (HDL)-cholesterol (HDL-C), low-density lipoprotein-cholesterol (LDL-C), total cholesterol, and triglyceride were measured by enzymatic absorption photometry (Cobas 8000, c702 module; F. Hoffmann-La Roche Ltd). Screening samples, used for evaluation of liver, kidney, and bone-marrow function, were analyzed by standard methods at Department of Clinical Biochemistry, Rigshospitalet, Copenhagen, Denmark.

      Oral glucose tolerance test

      The participants arrived at the laboratory after an overnight fast and an intravenous catheter (Vasofix Safety, 20G; B. Braun Melsungen AG, Melsungen, Germany) was placed in an antecubital vein. Blood was obtained from the catheter, put into EDTA-precoated tubes, and immediately centrifuged (Ole Dich Instrumentmakers ApS, Hvidovre, Denmark) at 4000 rpm for 2 minutes, and plasma was stored at –80°C until analysis. After oral intake of 82.5 g monohydrate glucose dissolved in 250 mL of water, blood samples were collected at 15, 30, 45, 60, 90, and 120 minutes and handled as described above. Glucose was assessed by photometric measurement (Cobas 8000, c702 module), and insulin and C-peptide were determined by sandwich ECLIA (Cobas 8000, e602 module).

      Calculations

      The Matsuda index was calculated as a surrogate measure of insulin sensitivity during the oral glucose tolerance test (OGTT) using the formula
      • Matsuda M.
      • DeFronzo R.A.
      Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp.
      :
      Matsuda index=10,000(fasting glucose(mgdL)fasting insulin(μUmL)mean glucose(mgdL)mean insulin(μUmL))0.5


      BP was measured using an upper arm, automatic BP monitor (M2 HEM-7121-E; Omron, Hoofddorp, The Netherlands) with the participant in a supine position, after at least 15 minutes of rest. The reported values are means of 7 consecutive measurements.

      VO2-max

      Maximal oxygen consumption was assessed during an incremental bicycle ergometer protocol (839E; Monark Exercise AB, Vansbro, Sweden) using an automated online system, measuring breath-by-breath pulmonary oxygen uptake and carbon-dioxide production (Oxycon Pro, Intramedic, Gentofte, Denmark). Before the test, the bicycle ergometer and the Oxycon Pro were calibrated and the saddle and handlebars were adjusted to fit the participant. HR monitor (Team2 Transmitter, Polar) was placed, and the mask was fitted. The participant was informed about the test, told to remain seated during the entire test, and to keep a pace of 60-70 rpm. Before the test was started, there was an 8-minute warm-up period with a workload of 50 W. The test workload started at 50 W and was increased by 25 W every minute, until exhaustion. The test was approved if 2 of 3 criteria were fulfilled: plateau in V˙o2 despite an increasing workload, respiratory exchange ratio >1.1, and HR >90% of predicted value, calculated as 220 minus age in years. The maximal watt load was registered at the end of the test.

      Statistical methods

      Statistical calculations were performed using software (SAS Enterprise Guide 7.1; SAS Institute Inc, Cary, NC). Descriptive statistics for parametric data are given as means (95% confidence limits) and for nonparametric data as median (25th-75th percentile). Baseline comparisons were made using independent t tests. The effects of menopausal status and exercise training were assessed using a 2-way repeated measures analysis of variance. Alpha was set to 0.05.

      Results

      A total of 43 premenopausal and 40 postmenopausal women were included, but 4 (3 premenopausal and 1 postmenopausal) women were excluded in the run-in period, as it was not possible to schedule the test days. Time since last menstrual period for the postmenopausal women was (mean [95% confidence limits] 3.1 [2.6-3.7] years). There were a few dropouts in each group, but only 1 due to low adherence to the training intervention (Figure 2). Data from a participant who was later excluded due to pregnancy and 1 who was excluded due to low training adherence were included in the baseline analyses. At baseline, 1 V˙o2max test was not approved due to subjective exhaustion before plateau of V˙o2. At 3 months, 2 tests were not included: 1 due to anemia and 1 test was not performed due to logistic difficulties. No adverse events occurred in response to the high-intensity training.

      High-intensity aerobic training intervention

      The 2 groups of premenopausal and postmenopausal women adhered equally well to the training intervention, as assessed by total training sessions, duration of the sessions, and exercise intensity during the sessions (Table 2).
      Table 2Training adherence for 3-month high-intensity aerobic training intervention
      GroupTraining sessionsPercentage of time spent in each interval of HRmax, %
      TotalLength, min<61%61–70%71–80%81–85%86–90%91–95%96–100%
      Premenopausal (n = 38)37 (35–39)54 (53–54)2 (1–2)7 (6–8)26 (24–29)23 (21–24)26 (23–28)14 (11–16)1 (1–2)
      Postmenopausal (n = 37)38 (36–39)53 (51–54)1 (1–2)7 (6–9)26 (22–29)23 (21–25)27 (24–31)13 (10–16)1 (1–2)
      Data are mean (95% confidence limits). Equal number, length, and intensity of training sessions.
      HRmax, maximal heart rate.
      Mandrup et al. High-intensity training for postmenopausal health. Am J Obstet Gynecol 2017.

      V˙o2max

      At baseline, there was no difference in V˙o2max between premenopausal and postmenopausal women (Figure 4, A). After the training intervention, the premenopausal and postmenopausal women had increased their V˙o2max by 8.8% and 9.4% (P < .001), respectively. Nonresponders were observed in both groups (Figure 5).
      Figure thumbnail gr4
      Figure 4Maximal oxygen consumption (V˙o2max) and maximal watt load (Wattmax)
      A, V˙o2max and B, Wattmax for premenopausal and postmenopausal women at baseline (light green/blue) and after high-intensity aerobic training intervention (green/blue). ‡Difference from baseline to 3 months, P ≤ .05.
      Mandrup et al. High-intensity training for postmenopausal health. Am J Obstet Gynecol 2017.
      Figure thumbnail gr5
      Figure 5Individual training response
      Change in maximal oxygen (O2) consumption.
      Mandrup et al. High-intensity training for postmenopausal health. Am J Obstet Gynecol 2017.

      Maximal watt load

      We observed no group difference in maximal watt load at baseline and both groups increased wattmax by 13.2% (P < .001) (Figure 4, B).

      Blood pressure

      At baseline, no difference between the groups in systolic or diastolic BP, mean arterial pressure, or resting HR was observed (Table 1). After training, both groups had lowered their diastolic BP (0.3% and 4.6% for premenopausal and postmenopausal group, respectively) (P < .05) and their resting HR (mean [95% confidence limits] 2.0 [3.7-0.4] beats/min for both groups) (P < .001). For systolic BP and mean arterial pressure a significant interaction between the 2 groups was found as the postmenopausal women had a lower systolic BP after the intervention, while the premenopausal women had a slight elevation (nonsignificant changes).

      Lipids

      At baseline, the concentrations of total cholesterol (P < .001), LDL-C (P = .01), and HDL-C (P < .001) differed between the premenopausal and postmenopausal women (Table 3). The training intervention reduced total cholesterol (P < .001), LDL-C (P < .001), and the total cholesterol/HDL-C index (P < .001) in both premenopausal and postmenopausal women. No significant change in triglyceride level was observed after training.
      Table 3Glucose metabolism and lipids
      VariablesPremenopausalPostmenopausal
      BaselineN3 monBaselinen3 mon
      Fasting glucose, mg/dL
      Nonparametric data are presented as median (25th–75th percentile).
      95.1 (87.6–98.4)3893.5 (91.0–97.1)3893.0 (89.9–97.5)3792.8 (86.1–97.1)37
      Fasting insulin, pmol/L
      Statistically significant (P ≤ .05) difference from baseline to 3 months
      ,
      Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group
      ,
      Statistically significant (P ≤ .05) interaction between time and menopausal status, meaning that 2 groups responded differently to intervention
      ,
      Nonparametric data are presented as median (25th–75th percentile).
      46.6 (33.7–59.0)3645.2 (33.3–57.9)3740.7 (32.2–64.4)3731.9 (24.5–45.0)37
      Fasting C-peptide, pmol/L
      Nonparametric data are presented as median (25th–75th percentile).
      617 (506–705)36610 (487–719)37536 (477–689)37515 (455–636)37
      Oral glucose tolerance test
      Glucose, AUC, mg/dL * 103
      Nonparametric data are presented as median (25th–75th percentile).
      14.5 (12.1–16.7)3414.7 (12.6–16.6)3513.9 (12.7–15.5)3513.9 (12.6–16.5)35
      Insulin, AUC, pmol/L * 103
      Statistically significant (P ≤ .05) difference from baseline to 3 months
      ,
      Nonparametric data are presented as median (25th–75th percentile).
      34.5 (29.7–41.9)3632.6 (28.6–41.9)3732.3 (25.9–44.6)3731.3 (24.7–40.0)37
      C-peptide, AUC, pmol/L *103
      Nonparametric data are presented as median (25th–75th percentile).
      249 (208–280)34238 (207–288)35252 (200–300)35251 (209–286)35
      Matsuda index
      Statistically significant (P ≤ .05) difference from baseline to 3 months
      ,
      Nonparametric data are presented as median (25th–75th percentile).
      6.0 (4.5–7.2)346.3 (4.7–7.8)356.7 (4.5–9.1)357.9 (5.4–9.7)35
      Lipids
      Total cholesterol, mg/dL
      Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group. Effect of intervention was assessed by 2-way analysis of variance
      ,
      Statistically significant (P ≤ .05) difference from baseline to 3 months
      ,
      Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group
      ,
      Parametric data are presented as mean (95% confidence limits)
      188 (179–197)37184 (175–192)37218 (209–227)37212 (204–220)36
      LDL-cholesterol, mg/dL
      Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group. Effect of intervention was assessed by 2-way analysis of variance
      ,
      Statistically significant (P ≤ .05) difference from baseline to 3 months
      ,
      Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group
      ,
      Parametric data are presented as mean (95% confidence limits)
      111 (102–120)37105 (97–112)37126 (118–134)37123 (114–131)36
      HDL-cholesterol, mg/dL
      Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group. Effect of intervention was assessed by 2-way analysis of variance
      ,
      Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group
      ,
      Parametric data are presented as mean (95% confidence limits)
      64 (59–68)3767 (61–72)3777 (72–83)3779 (73–84)36
      Triglycerides, mg/dL
      Nonparametric data are presented as median (25th–75th percentile).
      74 (57–96)3771 (61–97)3775 (59–97)3769 (59–93)36
      Total/HDL-cholesterol ratio
      Statistically significant (P ≤ .05) difference from baseline to 3 months
      ,
      Nonparametric data are presented as median (25th–75th percentile).
      3.1 (2.4–3.6)372.7 (2.3–3.3)372.8 (2.5–3.2)372.7 (2.4–3.1)36
      AUC, area under the curve; HDL, high-density lipoprotein; LDL, low-density lipoprotein.
      Mandrup et al. High-intensity training for postmenopausal health. Am J Obstet Gynecol 2017.
      Baseline difference assessed by independent t test:
      a Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group. Effect of intervention was assessed by 2-way analysis of variance
      b Statistically significant (P ≤ .05) difference from baseline to 3 months
      c Statistically significant (P ≤ .05) difference between premenopausal and postmenopausal group
      d Statistically significant (P ≤ .05) interaction between time and menopausal status, meaning that 2 groups responded differently to intervention
      e Parametric data are presented as mean (95% confidence limits)
      f Nonparametric data are presented as median (25th–75th percentile).

      Body composition

      At baseline, we did not find any differences in body composition between the premenopausal and postmenopausal women (Table 1). After the training period, both groups had experienced a decrease in weight, BMI, and waist circumference (P < .01). Additionally, both groups had similar reductions in fat mass (P < .001), fat percentage (P < .001), android fat percentage (P < .001), and gynoid fat percentage (P < .001), with no change in android-gynoid fat ratio. Lean body mass increased in both groups (P < .001).

      OGTT

      Fasting glucose was similar for the premenopausal and postmenopausal women at baseline and no changes were seen after the intervention (Table 3). Likewise, the total glucose load during an OGTT expressed as area under the curve was similar between groups and unchanged after the intervention (Figure 6). No difference was seen in fasting insulin at baseline, but the training intervention entailed a decrease in the fasting insulin level across groups (P < .05). However, the postmenopausal women had a better response to the training intervention as they had lower fasting insulin: 27% compared to 3% in the premenopausal women (P < .05 for interaction). The total insulin concentration (area under the curve) during the OGTT was lowered after the intervention in both groups (P < .05) (Figure 6), which was also reflected by the increase in Matsuda index after the intervention (P < .05). No difference was seen between groups in the C-peptide concentration either in the fasting state or during glucose stimulation at baseline or after the intervention.
      Figure thumbnail gr6
      Figure 6Oral glucose tolerance test (OGTT)
      Area under curve for glucose, insulin, and C-peptide during OGTT. Data are given as median and 25th-75th percentile. ‡Difference from baseline to 3 months, P ≤ .05.
      Mandrup et al. High-intensity training for postmenopausal health. Am J Obstet Gynecol 2017.

      Comment

      This study evaluated the effects of menopause and high-intensity aerobic training in late premenopausal and early postmenopausal women on cardiovascular and metabolic risk factors in a controlled and unique study design with narrow inclusion criteria for age (45-57 years) and BMI (18.5-30 kg/m2) to avoid confounding from those parameters. The main finding of the study was that 3 months of high-intensity aerobic training induces substantial beneficial effects on aerobic fitness and a number of risk factors in middle-aged women and that the postmenopausal women experienced the same positive adaptations to training as premenopausal women.

      Study strengths and limitations

      A strength of our study is that the exercise training intervention was highly controlled with continuous monitoring and supervision, resulting in optimal compliance. To avoid confounding factors such as obesity, smoking, and chronic disease when observing the effect of menopausal status on training response, we had narrow inclusion criteria. This led to a study population more healthy than the general population, and thereby reducing the generalizability of our study. However, since moderate-intensity exercise has been shown to reduce BP in hypertensive early postmenopausal women, and the cholesterol level in dyslipidemic early postmenopausal women,
      • Asikainen T.-M.
      • Kukkonen-Harjula K.
      • Miilunpalo S.
      Exercise for health for early postmenopausal women: a systematic review of randomized controlled trials.
      it seems likely that also less healthy postmenopausal women would benefit from high-intensity training to reduce cardiovascular risk factors. A selection bias may have been introduced, as the women who volunteered to participate were motivated to be physically active. Due to ethical considerations a nontraining control group was not included in the study design, as our primary objective was to evaluate differences in training response between premenopausal and postmenopausal women.
      The high intensity of our training intervention may be pivotal to the beneficial effects of exercise as moderate-intensity exercise training did not improve the plasma lipid profile in nondyslipidemic, early postmenopausal women.
      • Asikainen T.-M.
      • Kukkonen-Harjula K.
      • Miilunpalo S.
      Exercise for health for early postmenopausal women: a systematic review of randomized controlled trials.
      Other studies, concluding that estrogen is essential for exercise-induced improvements in risk factors for CVD, consisted of moderate-intensity walking interventions with HR between 50-65% and 65-80% of V˙o2max.
      • Kretzschmar J.
      • Babbitt D.M.
      • Diaz K.M.
      • et al.
      A standardized exercise intervention differentially affects premenopausal and postmenopausal African-American women.
      • Moreau K.L.
      • Stauffer B.L.
      • Kohrt W.M.
      • Seals D.R.
      Essential role of estrogen for improvements in vascular endothelial function with endurance exercise in postmenopausal women.
      Earlier, we showed that floorball training, also classified as high-intensity exercise, caused an increase in V˙o2max among postmenopausal women.
      • Nyberg M.
      • Seidelin K.
      • Andersen T.R.
      • Overby N.N.
      • Hellsten Y.
      • Bangsbo J.
      Biomarkers of vascular function in premenopausal and recent postmenopausal women of similar age: effect of exercise training.
      Which also implies that intense exercise training provides an effective stimulus for central cardiovascular adaptations.
      Biomarkers of vascular function change rapidly after menopause suggesting a strong influence by the hormonal milieu
      • Nyberg M.
      • Seidelin K.
      • Andersen T.R.
      • Overby N.N.
      • Hellsten Y.
      • Bangsbo J.
      Biomarkers of vascular function in premenopausal and recent postmenopausal women of similar age: effect of exercise training.
      but the impact of menopause on the development of essential hypertension has been debated. Longitudinal observational studies such as the Melbourne Women’s Midlife Health Project
      • Do K.-A.
      • Green A.
      • Guthrie J.R.
      • Dudley E.C.
      • Burger H.G.
      • Dennerstein L.
      Longitudinal study of risk factors for coronary heart disease across the menopausal transition.
      and the Study of Women Across the Nation
      • Matthews K.A.
      • Crawford S.L.
      • Chae C.U.
      • et al.
      Are changes in cardiovascular disease risk factors in midlife women due to chronological aging or to the menopausal transition?.
      indicate that increasing age and not postmenopausal status causes hypertension in contrast to cross-sectional findings.
      • Zanchetti A.
      • Facchetti R.
      • Cesana G.C.
      • Modena M.G.
      • Pirrelli A.
      • Sega R.
      Menopause-related blood pressure increase and its relationship to age and body mass index: the SIMONA epidemiological study.
      The current result of similar BP levels in the 2 groups of women at baseline does not support an influence of estrogen; however, in a previous publication we showed that vascular function is impaired and intravascular BP is higher in postmenopausal compared to premenopausal women in a subgroup of the same women.
      • Nyberg M.
      • Egelund J.
      • Mandrup C.M.
      • et al.
      Early postmenopausal phase is associated with reduced prostacyclin-induced vasodilation that is reversed by exercise training.
      However, importantly, 3 months of exercise training was found to normalize vascular function in the postmenopausal women.
      The differences in the lipid profile at baseline is somewhat consistent with the literature, as we observed that the LDL-C concentration was highest in the postmenopausal women.
      • Carr M.C.
      The emergence of the metabolic syndrome with menopause.
      • Singla A.
      • Bliden K.P.
      • Jeong Y.-H.
      • et al.
      Platelet reactivity and thrombogenicity in postmenopausal women.
      However, HDL-C was also higher in the postmenopausal compared with the premenopausal women in accordance with our earlier observations
      • Nyberg M.
      • Seidelin K.
      • Andersen T.R.
      • Overby N.N.
      • Hellsten Y.
      • Bangsbo J.
      Biomarkers of vascular function in premenopausal and recent postmenopausal women of similar age: effect of exercise training.
      but in disagreement with most other observations.
      • Carr M.C.
      The emergence of the metabolic syndrome with menopause.
      It may be speculated that the balance between the more antiatherogenic HDL2 levels and the less antiatherogenic HDL3 levels differed in the 2 groups, as menopause has been associated with decreases in HDL2 and increases in HDL3.
      • Carr M.C.
      The emergence of the metabolic syndrome with menopause.
      Physical exercise is generally associated with an increase in HDL-C,
      • Cornelissen V.A.
      • Fagard R.H.
      Effects of endurance training on blood pressure, blood pressure-regulating mechanisms, and cardiovascular risk factors.
      but like previous findings in postmenopausal women
      • Binder E.F.
      • Birge S.J.
      • Kohrt W.M.
      Effects of endurance exercise and hormone replacement therapy on serum lipids in older women.
      no change was observed in our study after the intervention. The observed training-induced reductions in LDL-C, total cholesterol, and the total cholesterol/HDL-C index, which is a valid predictor for CVD in women,
      • Millán J.
      • Pintó X.
      • Muñoz A.
      • et al.
      Lipoprotein ratios: physiological significance and clinical usefulness in cardiovascular prevention.
      contrasts conclusions from STRRIDE
      • Slentz C.A.
      • Houmard J.A.
      • Johnson J.L.
      • et al.
      Inactivity, exercise training and detraining, and plasma lipoproteins. STRRIDE: a randomized, controlled study of exercise intensity and amount.
      and a meta-analysis from the American Heart Association (2015) stating that there is no effect of moderate- or vigorous-intensity exercise training on LDL-C concentrations.
      • Lin X.
      • Zhang X.
      • Guo J.
      • et al.
      Effects of exercise training on cardiorespiratory fitness and biomarkers of cardiometabolic health: a systematic review and meta-analysis of randomized controlled trials.
      Apparently, high-intensity aerobic training can entail lowering of LDL-C and presumably the amount, regularity, and intensity of the exercise are important for the health-related effects.
      The matching body compositions between premenopausal and postmenopausal women at baseline are not consistent with observations in the literature, showing that menopause is associated with increased abdominal and total adipose tissue.
      • Davis S.R.
      • Castelo-Branco C.
      • Chedraui P.
      • et al.
      Understanding weight gain at menopause.
      • Polotsky H.N.
      • Polotsky A.J.
      Metabolic implications of menopause.
      Nevertheless, we observed similar reductions in fat mass and fat percentage in premenopausal and postmenopausal women after the intervention despite previous indications that menopause decreases the ability to oxidize fat during exercise.
      • Abildgaard J.
      • Pedersen A.T.
      • Green C.J.
      • et al.
      Menopause is associated with decreased whole body fat oxidation during exercise.
      In contrast to previous observations, where exercise did not improve glucose homeostasis or insulin secretion in postmenopausal women,
      • Chapman J.
      • Garvin A.W.
      • Ward A.
      • Cartee G.D.
      Unaltered insulin sensitivity after resistance exercise bout by postmenopausal women.
      we find an intact ability to reduce the plasma insulin response to an OGTT in early postmenopausal women. Interestingly, plasma C-peptide did not decline, suggesting that the decrease in insulin concentration was due to higher peripheral clearance of insulin rather than a decrease in secretion.
      • Björntorp P.
      The effects of exercise on plasma insulin.
      • Meshkani R.
      • Adeli K.
      Hepatic insulin resistance, metabolic syndrome and cardiovascular disease.

      Clinical implication

      Adverse metabolic changes appear in more than one third of postmenopausal women. In this study we show that even in postmenopausal women without cardiovascular risk factors, substantial health improvements can be achieved by only 3 months of exercise training, thereby lowering the risk of T2D and CVD beyond weight loss. This knowledge is of importance for recommendations on physical activity for prevention and treatment of lifestyle-related diseases in middle-aged women. The clinical challenge is to motivate postmenopausal as well as premenopausal women to perform high-intensity aerobic training and to adhere to an active lifestyle.

      References

        • Carr M.C.
        The emergence of the metabolic syndrome with menopause.
        J Clin Endocrinol Metab. 2003; 88: 2404-2411
        • Ross L.A.
        • Polotsky A.J.
        Metabolic correlates of menopause: an update.
        Curr Opin Obstet Gynecol. 2012; 24: 402-407
        • O’Neill S.
        • O’Driscoll L.
        Metabolic syndrome: a closer look at the growing epidemic and its associated pathologies.
        Obes Rev. 2015; 16: 1-12
        • Stefanska A.
        • Bergmann K.
        • Sypniewska G.
        Metabolic syndrome and menopause: pathophysiology, clinical and diagnostic significance.
        Adv Clin Chem. 2015; 72: 1-75
        • Matthews K.A.
        • Crawford S.L.
        • Chae C.U.
        • et al.
        Are changes in cardiovascular disease risk factors in midlife women due to chronological aging or to the menopausal transition?.
        J Cardiol. 2010; 54: 2366-2373
        • Lin X.
        • Zhang X.
        • Guo J.
        • et al.
        Effects of exercise training on cardiorespiratory fitness and biomarkers of cardiometabolic health: a systematic review and meta-analysis of randomized controlled trials.
        J Am Heart Assoc. 2015; 4: 1-29
        • Pattyn N.
        • Cornelissen V.A.
        • Eshghi S.R.T.
        • Vanhees L.
        The effect of exercise on the cardiovascular risk factors constituting the metabolic syndrome: a meta-analysis of controlled trials.
        Sports Med. 2013; 43: 121-133
        • Lynch N.A.
        • Ryan A.S.
        • Berman D.M.
        • Sorkin J.D.
        • Nicklas B.J.
        Comparison of VO2max and disease risk factors between perimenopausal and postmenopausal women.
        Menopause. 2002; 9: 456-462
        • Mercuro G.
        • Saiu F.
        • Deidda M.
        • Mercuro S.
        • Vitale C.
        • Rosano G.M.C.
        Impairment of physical exercise capacity in healthy postmenopausal women.
        Am Heart J. 2006; 151: 923-927
        • Parker B.A.
        • Kalasky M.J.
        • Proctor D.N.
        Evidence for sex differences in cardiovascular aging and adaptive responses to physical activity.
        Eur J Appl Physiol. 2010; 110: 235-246
        • Riesco E.
        • Tessier S.
        • Lacaille M.
        • et al.
        Impact of a moderate-intensity walking program on cardiometabolic risk markers in overweight to obese women: is there any influence of menopause?.
        Menopause. 2013; 20: 185-193
        • Asikainen T.-M.
        • Kukkonen-Harjula K.
        • Miilunpalo S.
        Exercise for health for early postmenopausal women: a systematic review of randomized controlled trials.
        Sports Med. 2004; 34: 753-778
        • Nyberg M.
        • Egelund J.
        • Mandrup C.M.
        • et al.
        Early postmenopausal phase is associated with reduced prostacyclin-induced vasodilation that is reversed by exercise training.
        Hypertension. 2016; 68: 1011-1020
        • Matsuda M.
        • DeFronzo R.A.
        Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp.
        Diabetes Care. 1999; 22: 1462-1470
        • Kretzschmar J.
        • Babbitt D.M.
        • Diaz K.M.
        • et al.
        A standardized exercise intervention differentially affects premenopausal and postmenopausal African-American women.
        Menopause. 2014; 21: 579-584
        • Moreau K.L.
        • Stauffer B.L.
        • Kohrt W.M.
        • Seals D.R.
        Essential role of estrogen for improvements in vascular endothelial function with endurance exercise in postmenopausal women.
        J Clin Endocrinol Metab. 2013; 98: 4507-4515
        • Nyberg M.
        • Seidelin K.
        • Andersen T.R.
        • Overby N.N.
        • Hellsten Y.
        • Bangsbo J.
        Biomarkers of vascular function in premenopausal and recent postmenopausal women of similar age: effect of exercise training.
        Am J Physiol Regul Integr Comp Physiol. 2014; 306: R510-R517
        • Do K.-A.
        • Green A.
        • Guthrie J.R.
        • Dudley E.C.
        • Burger H.G.
        • Dennerstein L.
        Longitudinal study of risk factors for coronary heart disease across the menopausal transition.
        Am J Epidemiol. 2000; 151: 584-593
        • Zanchetti A.
        • Facchetti R.
        • Cesana G.C.
        • Modena M.G.
        • Pirrelli A.
        • Sega R.
        Menopause-related blood pressure increase and its relationship to age and body mass index: the SIMONA epidemiological study.
        J Hypertens. 2005; 23: 2269-2276
        • Singla A.
        • Bliden K.P.
        • Jeong Y.-H.
        • et al.
        Platelet reactivity and thrombogenicity in postmenopausal women.
        Menopause. 2013; 20: 57-63
        • Cornelissen V.A.
        • Fagard R.H.
        Effects of endurance training on blood pressure, blood pressure-regulating mechanisms, and cardiovascular risk factors.
        Hypertension. 2005; 46: 667-675
        • Binder E.F.
        • Birge S.J.
        • Kohrt W.M.
        Effects of endurance exercise and hormone replacement therapy on serum lipids in older women.
        J Am Geriatr Soc. 1996; 44: 231-236
        • Millán J.
        • Pintó X.
        • Muñoz A.
        • et al.
        Lipoprotein ratios: physiological significance and clinical usefulness in cardiovascular prevention.
        Vasc Health Risk Manag. 2009; 5: 757-765
        • Slentz C.A.
        • Houmard J.A.
        • Johnson J.L.
        • et al.
        Inactivity, exercise training and detraining, and plasma lipoproteins. STRRIDE: a randomized, controlled study of exercise intensity and amount.
        J Appl Physiol. 2007; 103: 432-442
        • Davis S.R.
        • Castelo-Branco C.
        • Chedraui P.
        • et al.
        Understanding weight gain at menopause.
        Climacteric. 2012; 15: 419-429
        • Polotsky H.N.
        • Polotsky A.J.
        Metabolic implications of menopause.
        Semin Reprod Med. 2010; 28: 426-434
        • Abildgaard J.
        • Pedersen A.T.
        • Green C.J.
        • et al.
        Menopause is associated with decreased whole body fat oxidation during exercise.
        Am J Physiol Endocrinol Metab. 2013; 304: E1227-E1236
        • Chapman J.
        • Garvin A.W.
        • Ward A.
        • Cartee G.D.
        Unaltered insulin sensitivity after resistance exercise bout by postmenopausal women.
        Med Sci Sports Exerc. 2002; 34: 936-941
        • Björntorp P.
        The effects of exercise on plasma insulin.
        Int J Sports Med. 1981; 2: 125-129
        • Meshkani R.
        • Adeli K.
        Hepatic insulin resistance, metabolic syndrome and cardiovascular disease.
        Clin Biochem. 2009; 42: 1331-1346