Abstract
STUDY QUESTION
Is physical activity associated with incident early menopause?
SUMMARY ANSWER
Physical activity is not associated with incident early menopause.
WHAT IS KNOWN ALREADY
Lifestyle factors such as physical activity may influence menopause timing, but results from prior research are inconsistent.
STUDY DESIGN, SIZE, DURATION
We evaluated the association between physical activity and the occurrence of early natural menopause in a prospective cohort study, the Nurses’ Health Study II. Women were followed prospectively from 1989 to 2011.
PARTICIPANTS/MATERIALS, SETTING, METHODS
Our analysis included 107 275 women who were premenopausal at baseline. Menopause status was self-reported biennially. Time per week participating in specific activities was reported approximately every 4 years and used to calculate metabolic task hours per week (MET h/week). We used Cox proportional hazards model to evaluate the association between physical activity and incidence of natural menopause before age 45 years while controlling for potential confounding factors.
MAIN RESULTS AND THE ROLE OF CHANCE
There were 2 786 study members who experienced menopause before the age of 45. After adjustment for age, smoking and other factors, we observed no association between adulthood physical activity and early menopause. For example, compared to women reporting <3 MET h/week, the hazard ratio for women in the highest category (≥42 MET h/week) of cumulatively-averaged total physical activity was 0.89 (95% confidence interval: 0.76–1.04; P-trend: 0.26). Neither moderate nor strenuous activity in adolescence and young adulthood were related to risk. The relation of physical activity and early menopause did not vary across strata of body mass index or smoking status.
LIMITATIONS, REASONS FOR CAUTION
Physical activity and menopausal status were self-reported, but repeated assessment of physical activity and prospective report of menopause status likely reduce the potential for non-differential misclassification. While the majority of our study participants were white, it is unlikely that the physiological relation of activity and early menopause varies by ethnicity.
WIDER IMPLICATIONS OF THE FINDINGS
Findings from our large prospective study do not support an important association between physical activity and early menopause.
STUDY FUNDING/COMPETING INTEREST(S)
This project was supported by UM1CA176726 and R01HD078517 from the National Institutes of Health, Department of Health and Human Services. No competing interests are declared.
TRIAL REGISTRATION NUMBER
N/A.
Keywords: exercise, menopause, ovarian reserve, metabolic equivalent, prospective studies, early menopause
Introduction
Early menopause, defined as the cessation of menstruation before age 45, is experienced by approximately 10% of women (Jacobsen et al., 2003; Broekmans et al., 2009; Shuster et al., 2010). Early menopause is associated with an increased risk of premature mortality, cardiovascular disease and other chronic conditions (van Der Voort et al., 2003; Miller, 2012; Wellons et al., 2012; Bleil et al., 2013; Torrealday and Pal, 2015). Genetic factors alone do not explain the development of early menopause, and recent studies suggests that modifiable lifestyle factors may also influence risk, including physical activity (Aydin et al., 2005; Ortega-Ceballos et al., 2006; Dratva et al., 2007; Dorjgochoo et al., 2008; Nagata et al., 2012; Emaus et al., 2013; Gold et al., 2013).
In premenopausal women, physical activity may influence ovarian function by altering production of estrogens and other steroid hormones via the hypothalamic–pituitary–ovarian axis (Tworoger et al., 2007). Intense physical activity can disrupt regular ovarian cycles, leading to luteal phase insufficiency, anovulation and amenorrhea; these may be associated with a lower risk of early menopause as they are associated with a lower frequency of ovulation and a slower rate of exhaustion of the follicle pool (Gudmundsdottir et al., 2009). Most prior research of physical activity and menopause has primarily addressed a potential relation with menopausal timing more generally. Prospective studies have yielded inconsistent findings. Only two studies of physical activity and early menopause specifically have been conducted; both were based on small sample sizes and observed point estimates consistent with a strong inverse association, but the results were not statistically significant. It thus remains unclear whether physical activity may be importantly related to risk of early menopause.
We evaluated the relation of physical activity at multiple ages with risk of early menopause in the prospective Nurses’ Health Study II (NHSII).
Materials and Methods
Participants
The NHSII cohort includes of 116 429 female US registered nurses aged 25–42 years in 1989. Participants completed questionnaires assessing lifestyle factors and medical conditions every two years thereafter. Follow-up rates for each questionnaire cycle were least 89%. The study protocol was approved by the Institutional Review Board at Brigham, Women Hospital in Boston, MA and the Harvard TH Chan School of Public Health.
Assessment of physical activity
In study questionnaires in 1989, 1991, 1997, 2001, 2005 and 2009, women reported the average time per week during the past year spent on recreational activities. In 1989, activities included walking, jogging, running, bicycling (including on a stationary machine), racquet sports, lap swimming, calisthenics and other aerobic activities. In later years, activities were expanded to include weight training, low intensity exercises like yoga and high intensity activities like lawn mowing. Response options ranged from 0 minutes to ≥11 h/week. In 1997, women also reported the average amount of time per week spent on moderate and strenuous recreational activities at three ages: grades 7–8 (ages 12–13), grades 9–12 (ages 14–17) and ages 18–22, 23–30 and 30–34. Response options ranged from none to ≥11 h/week. To incorporate the frequency, duration and intensity of activity, we multiplied hours per week of each activity by its metabolic equivalent (MET) score, with one MET equal to 1 kcal/kg/h or the energy expended by sitting quietly (Ainsworth et al., 2000), and then summed values for all activities to create total MET hours/week.
Assessment of early menopause
In 1989 and on each subsequent biennial questionnaire, women were asked if their menstrual periods had ceased permanently, and if so, at what age, and whether cessation was related to surgery, radiation or chemotherapy or occurred naturally. Information regarding use of hormone therapy (HT) was collected as well. Age at menopause was defined as age after 12 consecutive months of amenorrhea. A small number of women reported being postmenopausal on one questionnaire and then subsequently reported being premenopausal. For these women, we defined age at menopause as age after which periods were absent for 12 months or more, and then confirmed that this status persisted for at least three consecutive questionnaires.
Eligibility for the current analysis was limited to women who were premenopausal and reported physical activity at the baseline 1989 questionnaire (n = 107 275). Women were followed for incidence of menopause until June 2011. We defined cases as those women who reported natural menopause before age 45.
Covariate assessment
Information on race, ethnicity, age, education, height and age at menarche was collected at baseline. Information on parity, oral contraceptive (OC) use, HT use and smoking was collected biennially since baseline. Current weight and body mass index (BMI; weight (kg)/height (m)2) were also assessed biennially. Information on diet was collected using semi-quantitative food frequency questionnaires (FFQ) in 1991 and every 4 years thereafter (Willet et al., 1985). Additional information on dietary supplement use, including vitamin D and calcium, was collected on main NHSII questionnaires. Nutrient intake was adjusted for total energy using the residual method (Willett, 2012).
Statistical analysis
We divided women into categories of total MET hours/week and then compared the distribution of covariates at baseline with physical activity in age-adjusted general linear models. We evaluated the relation of category of physical activity with incidence of early menopause using Cox proportional hazard models to calculate hazard ratios (HR) and 95% confidence intervals (CI). Women contributed person-time in months from the date of return of the baseline questionnaire (1989) until menopause, hysterectomy, oophorectomy, diagnosis of cancer (other than non-melanoma skin cancer), death, loss to follow-up, age 45 or June 2011, whichever came first.
We modeled total physical activity participation in adulthood in three ways: (1) MET h/week at baseline (1989) only; (2) updating MET h/week data every 4 years; (3) cumulatively averaging MET h/week, updated every 4 years; this method uses the mean MET h/week from all prior physical activity assessments as a measure of long-term physical activity (Eliassen et al., 2010). In addition, we evaluated associations of early menopause with the cumulative average of time spent walking, as walking was the most common activity reported by participants. Risk associated with moderate and strenuous activity in adolescence, early and middle adulthood was also assessed.
In addition to models adjusting for age, we built a multivariate model controlling for time-varying factors including age, pack-years of smoking, parity, breastfeeding duration, alcohol intake, percentage of total calories from vegetable protein, and vitamin D from dairy intake, as well as age at menarche (time invariant); these covariates were selected a priori based on previous studies of early menopause in our population (Purdue-Smithe et al., 2017, Boutot et al., 2018, Whitcomb et al., 2018). We created a second multivariate model additionally controlling for BMI (time-varying), given that adiposity could be a confounder or intermediate in the pathway; BMI and physical activity are closely related, and prior research in the same population found that underweight women are at increased risk for early menopause (Szegda et al., 2017). To test for linear trend, we assigned the median activity level of each category to all women in that category and used median values in regression models.
We conducted several sensitivity analyses. We stratified analyses to evaluate whether hazard ratio estimates were comparable across categories of BMI, smoking status and current OC use, using likelihood ratio tests to compare models with and without multiplicative interaction terms. Finally, because HT use may contribute to misclassification of menopause status, we ran additional models censoring at first use of HT. Statistical analyses were conducted with SAS v9.3 software (SAS Institute Inc, Cary, NC).
Results
During 20 years of follow-up, 2786 members of the analytic cohort experienced incident early menopause. Overall, women reporting higher physical activity were somewhat younger, had lower BMI at baseline and recalled BMI at age 18, and lower parity compared to those reporting lower activity overall (Table I). Greater physical activity was associated with slightly longer duration of OCs use, and higher intake of alcohol, vegetable protein and vitamin D.
Table I.
Age-Standardized characteristics of premenopausal women according to category of total physical activity at baseline (1989): Nurses’ Health Study II, 1989
| Total physical activity (MET h/week) | ||||||
|---|---|---|---|---|---|---|
| Characteristic | <3 (n = 16,086) | 3 – <9 (n = 24,298) | 9 – <18 (n = 22,413) | 18 – <27 (n = 14,133) | 27 – <42 (n = 12,867) | ≥42 (n = 17,478) |
| Mean (SE) | Mean (SE) | Mean (SE) | Mean (SE) | Mean (SE) | Mean (SE) | |
| Age, ya | 34.7 ± 4.5 | 34.4 ± 4.6 | 34.2 ± 4.6 | 34.1 ± 4.6 | 33.8 ± 4.6 | 33.2 ± 4.7 |
| Age at menarche, y | 12.4 (0.01) | 12.4 (0.01) | 12.4 (0.01) | 12.4 (0.01) | 12.4 (0.01) | 12.5 (0.01) |
| BMI at age 18, kg/mb | 21.4 (0.03) | 21.3 (0.02) | 21.3 (0.02) | 21.2 (0.03) | 21.2 (0.03) | 21.1 (0.03) |
| BMI, kg/mb | 25.2 (0.04) | 24.5 (0.03) | 24.0 (0.03) | 23.6 (0.04) | 23.3 (0.04) | 23.1 (0.04) |
| Pack-years of smokingb | 12.2 (0.10) | 11.6 (0.08) | 10.8 (0.09) | 10.7 (0.11) | 10.6 (0.11) | 10.5 (0.10) |
| Duration of oral contraceptive use (m)c | 52.5 (0.4) | 52.6 (0.3) | 53.1 (0.3) | 53.6 (0.4) | 53.7 (0.4) | 54.6 (0.4) |
| Alcohol intake (g/d) | 2.4 (0.05) | 2.7 (0.04) | 2.9 (0.04) | 3.2 (0.05) | 3.4 (0.05) | 3.7 (0.04) |
| Parity (pregnancies > 6 m)d | 2.1 (0.01) | 2.1 (0.01) | 2.1 (0.01) | 2.0 (0.01) | 2.0 (0.01) | 2.0 (0.01) |
| Breastfeeding duration (m)d | 12.0 (0.1) | 12.8 (0.1) | 13.2 (0.1) | 13.0 (0.1) | 12.9 (0.2) | 12.5 (0.1) |
| Dietary vitamin D intake (IU/d) | 235 (1.1) | 246 (0.9) | 255 (0.9) | 260 (1.1) | 263 (1.2) | 266 (1.1) |
| Vegetable protein intake (% of kcal) | 4.8 (0.01) | 4.9 (0.01) | 5.0 (0.01) | 5.1 (0.01) | 5.2 (0.01) | 5.2 (0.01) |
| % | % | % | % | % | % | |
| Oral contraceptive use | ||||||
| Past | 69.5 | 70.1 | 69.5 | 69.4 | 68.4 | 67.2 |
| Current | 13.0 | 12.6 | 13.6 | 13.8 | 14.2 | 15.5 |
| Nulliparous | 24.2 | 25.3 | 27.3 | 31.4 | 33.4 | 38.5 |
| Smoking status | ||||||
| Past | 21.3 | 21.6 | 23.2 | 24.0 | 24.5 | 25.0 |
| Current | 13.9 | 12.3 | 10.3 | 9.8 | 9.3 | 10.5 |
aFor age, values presented are unadjusted mean and standard deviation. All characteristics were calculated with the use of generalized linear models adjusted for the age of participants in 1989. All variables are significant at the P < 0.05 level.
bAmong ever-smokers.
cAmong ever users of oral contraceptives.
dAmong parous women.
In results from age-adjusted Cox proportional hazard models, physical activity at baseline was not associated with risk of early menopause (Table II). For example, compared to women who reported <3 MET h/week, women reporting ≥42 MET h/week of activity had a HR of early menopause of 0.92 (95% CI = 0.81–1.06; P-trend = 0.34). Results adjusting for covariates (Model 2) were similar, and additional adjustment for BMI had minimal impact. Results from analyses modeling updated and cumulatively-averaged physical activity were consistent with models of baseline activity, suggesting no association with early menopause risk. No association was observed for time spent walking and early menopause risk. In results from fully adjusted models, women reporting a cumulative average of 4 or more h/week of walking had a HR of early menopause of 0.89 (95% CI = 0.76–1.04) compared to women reporting no walking (P-trend = 0.26). Time spent per week in moderate and strenuous activity at ages 23–39 and 30–34 were also unrelated to risk of early menopause (results not shown).
Table II.
Hazard ratios (95% CIs) for early menopause by level of physical activity: Nurses’ Health Study II (1989–2011)a
| n Cases | Age-adjusted | MV1a | MV2b | |
|---|---|---|---|---|
| HR (95% CI) | HR (95% CI) | HR (95% CI) | ||
| Physical activity, 1989 (MET h/week) | ||||
| <3 | 437 | 1 | 1 | 1 |
| 3 – <9 | 648 | 0.98 (0.87, 1.10) | 1.00 (0.88, 1.13) | 1.00 (0.88, 1.13) |
| 9 – <18 | 574 | 0.94 (0.83, 1.06) | 0.97 (0.85, 1.10) | 0.96 (0.85, 1.09) |
| 18 – <27 | 363 | 0.94 (0.81, 1.08) | 0.97 (0.84, 1.11) | 0.96 (0.83, 1.10) |
| 27 – <42 | 336 | 0.97 (0.84, 1.12) | 1.01 (0.87, 1.16) | 0.99 (0.86, 1.14) |
| ≥42 | 418 | 0.92 (0.81, 1.06) | 0.94 (0.82, 1.08) | 0.92 (0.81, 1.06) |
| P-trend | 0.34 | 0.42 | 0.24 | |
| Physical activity, updated (MET h/week) | ||||
| <3 | 515 | 1 | 1 | 1 |
| 3 – <9 | 795 | 1.01 (0.90, 1.12) | 1.03 (0.92, 1.16) | 1.03 (0.92, 1.15) |
| 9 – <18 | 438 | 0.94 (0.82, 1.07) | 0.98 (0.86, 1.12) | 0.97 (0.84, 1.10) |
| 18 – <27 | 284 | 1.02 (0.88, 1.19) | 1.07 (0.92, 1.24) | 1.05 (0.90, 1.22) |
| 27 – <42 | 241 | 0.98 (0.83, 1.14) | 1.01 (0.86, 1.19) | 0.99 (0.84, 1.16) |
| ≥42 | 234 | 0.91 (0.78, 1.07) | 0.94 (0.80, 1.11) | 0.91 (0.77, 1.07) |
| P-trend | 0.26 | 0.41 | 0.21 | |
| Physical activity, cumulative average (MET h/week) | ||||
| <3 | 308 | 1 | 1 | 1 |
| 3 – <9 | 634 | 0.93 (0.81, 1.06) | 0.94 (0.82, 1.08) | 0.94 (0.82, 1.08) |
| 9 – <18 | 667 | 0.90 (0.79, 1.04) | 0.94 (0.82, 1.07) | 0.93 (0.81, 1.07) |
| 18 – <27 | 466 | 1.03 (0.89, 1.19) | 1.06 (0.91, 1.22) | 1.04 (0.90, 1.21) |
| 27 – <42 | 358 | 0.92 (0.79, 1.08) | 0.95 (0.81, 1.11) | 0.93 (0.80, 1.09) |
| ≥42 | 348 | 0.89 (0.76, 1.04) | 0.91 (0.78, 1.07) | 0.89 (0.76, 1.04) |
| P-trend | 0.42 | 0.53 | 0.26 | |
aMultivariable (MV) model adjusted for age (continuous, months), pack-years of smoking (0–10, 11–20 or ≥21), age at menarche (continuous), parity (nulliparous, 1–2, or ≥3), breastfeeding duration (in months; continuous), oral contraceptive use (never/past/current), % of calories from vegetable protein (quintiles), alcohol intake (<10, ≥10 g/day), and intake of vitamin D from dairy source (quintiles).
bMV model additionally adjusted for BMI (<18.5, 18.5–<25, 25–<30 or ≥30 kg/m2).
In results stratified by BMI and OC use, estimates were very similar between strata and we observed no evidence of effect modification (Table III). In results stratified by smoking status, we observed no association among ever-smokers. Among never smokers, we observed some evidence of an inverse relation between activity and risk, with women averaging at least 42 MET h/week having a 23% lower risk of early menopause compared to women reporting <3 MET h/week (HR = 0.77; 95% CI = 0.63–0.95; P-trend = 0.07), but the test for interaction was not significant (P-interaction = 0.34). Results from models censoring at first HT use were highly similar to those of the main analysis (results not shown).
Table III.
Hazard Ratios (95% CIs) for early menopause by level of physical activity (cumulatively averaged), stratified by current body mass index, smoking status and oral contraceptive use: Nurses’ Health Study II (1989–2011)
| Physical activity, (MET h/week) | Current body mass index | |||
|---|---|---|---|---|
| BMI < 25 kg/m2 | BMI ≥ 25 kg/m2 | |||
| n cases | HR (95% CI)a | n cases | HR (95% CI)a | |
| <3 | 182 | 1 | 126 | 1 |
| 3 – <9 | 394 | 0.94 (0.79, 1.12) | 240 | 0.95 (0.76, 1.18) |
| 9 – <18 | 453 | 0.94 (0.79, 1.12) | 214 | 0.91 (0.73, 1.14) |
| 18 – <27 | 327 | 1.04 (0.86, 1.25) | 139 | 1.07 (0.84, 1.38) |
| 27 – <42 | 261 | 0.93 (0.77, 1.13) | 97 | 0.95 (0.72, 1.25) |
| ≥42 | 265 | 0.87 (0.72, 1.06) | 83 | 0.94 (0.71, 1.25) |
| P-trend | 0.21 | 0.98 | ||
| P-interaction | 0.98 | |||
| Current smoking status | ||||
| Never smokers | Ever-smokers | |||
| Physical activity, (MET h/week) | n Cases | HR (95% CI)b | n Cases | HR (95% CI)b |
| <3 | 187 | 1 | 118 | 1 |
| 3 – <9 | 373 | 0.87 (0.73, 1.04) | 261 | 1.07 (0.85, 1.33) |
| 9 – <18 | 403 | 0.86 (0.72, 1.03) | 265 | 1.02 (0.82, 1.28) |
| 18 – <27 | 256 | 0.89 (0.73, 1.08) | 210 | 1.28 (1.01, 1.61) |
| 27 – <42 | 212 | 0.87 (0.71, 1.06) | 144 | 1.01 (0.79, 1.30) |
| ≥42 | 191 | 0.77 (0.63, 0.95) | 155 | 1.00 (0.78, 1.28) |
| P-trend | 0.07 | 0.71 | ||
| P-interaction | 0.34 | |||
| Current oral contraceptive use | ||||
| Current | Not current | |||
| Physical activity, (MET h/week) | n Cases | HR (95% CI)c | n Cases | HR (95% CI)c |
| <3 | 12 | 1 | 274 | 1 |
| 3 – <9 | 39 | 1.41 (0.70, 2.83) | 564 | 0.95 (0.82, 1.10) |
| 9 – <18 | 49 | 1.33 (0.68, 2.63) | 582 | 0.93 (0.80, 1.07) |
| 18 – <27 | 25 | 1.02 (0.48, 2.13) | 426 | 1.10 (0.94, 1.28) |
| 27 – <42 | 29 | 1.39 (0.67, 2.88) | 309 | 0.93 (0.80, 1.10) |
| ≥42 | 23 | 1.08 (0.51, 2.29) | 305 | 0.92 (0.77, 1.08) |
| P-trend | 0.56 | 0.49 | ||
| P-interaction | 0.08 | |||
aAll models adjusted for age, pack-years of smoking (0–10, 11–20 or ≥21), BMI (continuous in kg/m2), age at menarche (continuous), parity (nulliparous, 1–2 or ≥3), breastfeeding duration (in months; continuous), oral contraceptive use (never/past/current), % of total calories from vegetable protein (quintiles), alcohol intake (<10, ≥10 g/day), and intake of vitamin D from dairy source (quintiles).
bAdjusted for age, pack-years of smoking (continuous), BMI [in kg/m2 (<18.5, 18.5–<25, 25–<30 or ≥30)], age at menarche (continuous), parity (nulliparous, 1–2 or ≥3), breastfeeding duration (in months; continuous), oral contraceptive use (never/past/current), % of total calories from vegetable protein (quintiles), alcohol intake (<10, ≥10 g/day), and intake of vitamin D from dairy source (quintiles).
cAdjusted for age, pack-years of smoking (0–10, 11–20 or ≥21), BMI [in kg/m2 (<18.5, 18.5–<25, 25–<30 or ≥30)], age at menarche (continuous), parity (nulliparous, 1–2 or ≥3), breastfeeding duration (in months; continuous), % of total calories from vegetable protein (quintiles), alcohol intake (<10, ≥10 g/day), and intake of vitamin D from dairy source (quintiles).
We did not find moderate or strenuous activity at any age (12–13 years, 14–17 years or 18–22 years) to be associated with risk of early menopause (Table IV).
Table IV.
Hazard ratios (95% CIs) for early menopause by level of moderate and strenuous activity in adolescence and young adulthood; Nurses’ Health Study II (1989–2011)
| Physical activity (h/week) | n Cases | Moderate activity | n Cases | Strenuous activity |
|---|---|---|---|---|
| HR (95% CI)a | HR (95% CI)a | |||
| Age 12–13 (grades 7–8) | ||||
| None | 118 | 1 | 243 | 1 |
| >0–<1 | 177 | 1.12 (0.89, 1.41) | 179 | 1.08 (0.89, 1.31) |
| 1–2 | 645 | 1.01 (0.83, 1.23) | 500 | 0.96 (0.82, 1.12) |
| 3–4 | 585 | 1.08 (0.89, 1.32) | 479 | 1.03 (0.88, 1.20) |
| 5–6 | 284 | 0.98 (0.79, 1.22) | 349 | 1.03 (0.87, 1.21) |
| 7–10 | 140 | 0.91 (0.71, 1.17) | 206 | 0.97 (0.80, 1.17) |
| ≥11 | 110 | 0.99 (0.76, 1.28) | 138 | 0.98 (0.79, 1.20) |
| P-trend | 0.29 | 0.80 | ||
| Age 14–17 (grades 9–12) | ||||
| None | 106 | 1 | 222 | 1 |
| >0–<1 | 174 | 1.14 (0.89, 1.45) | 183 | 1.12 (0.92, 1.36) |
| 1–2 | 639 | 1.04 (0.84, 1.27) | 463 | 0.91 (0.77, 1.06) |
| 3–4 | 588 | 1.11 (0.90, 1.36) | 448 | 0.93 (0.79, 1.09) |
| 5–6 | 290 | 1.01 (0.80, 1.26) | 360 | 0.97 (0.82, 1.14) |
| 7–10 | 152 | 1.00 (0.78, 1.28) | 238 | 0.94 (0.78, 1.13) |
| ≥11 | 109 | 1.10 (0.84, 1.43) | 187 | 0.97 (0.79, 1.17) |
| P-trend | 0.85 | 0.76 | ||
| Age 18–22 | ||||
| None | 127 | 1 | 299 | 1 |
| >0–<1 | 207 | 0.99 (0.79, 1.24) | 290 | 1.03 (0.88, 1.21) |
| 1–2 | 665 | 1.04 (0.86, 1.25) | 575 | 1.01 (0.87, 1.16) |
| 3–4 | 577 | 1.10 (0.91, 1.33) | 449 | 1.05 (0.90, 1.21) |
| 5–6 | 283 | 1.03 (0.83, 1.27) | 249 | 0.97 (0.82, 1.15) |
| 7–10 | 118 | 0.94 (0.73, 1.21) | 137 | 1.06 (0.86, 1.30) |
| ≥11 | 77 | 1.06 (0.79, 1.40) | 88 | 1.17 (0.92, 1.49) |
| P-trend | 0.97 | 0.31 | ||
aAdjusted for age (continuous, months), BMI (<18.5, 18.5-<25, 25-<30 or ≥30 kg/m2), pack-years of smoking (0–10, 11–20 or ≥21), age at menarche (continuous), parity (nulliparous, 1–2, or ≥3), breastfeeding duration (in months; continuous), oral contraceptive use (never/past/current), % of calories from vegetable protein (quintiles), alcohol intake (<10, ≥10 g/day) and intake of vitamin D from dairy source (quintiles).
Discussion
In this prospective study, we did not find any substantial evidence that participation in physical activity was associated with risk of early menopause. In our main analyses, the results were null across all levels of physical activity, even for women reporting the equivalent to 4 or more hours of running or 8 or more hours of brisk walking per week compared to inactive women. Analyses limited to never smokers suggested the possibility of a modestly lower risk among women reporting consistently high levels of activity. High levels of moderate and intense physical activity in adolescence and early adulthood were unrelated to risk of early menopause.
The association of physical activity and early menopause was assessed previously in a prospective study among 3 435 women followed for approximately 12 years, during which time 124 cases of early menopause were observed (Gudmundsdottir et al., 2013). Compared to women reporting no activity, HRs of early menopause (95% CI) among women reporting high and very high activity were 0.68 (0.28–1.68) and 0.21 (0.02–1.78), respectively. In a cross-sectional analysis of 1 829 women in the Oslo Health Study that observed early menopause in 9.6% of the cohort, the odds ratio for early menopause among women reporting ≥3 h/week of activity versus none was 0.70 (95% CI = 0.38–1.30) (Mikkelsen et al., 2007). Despite low point estimates in these two studies, small numbers of cases and low statistical power resulted in wide confidence intervals; our large prospective study including more than 2 700 incident early menopause cases suggests no inverse association.
Associations from previous cohort studies of activity and menopause timing have generally been small in magnitude, if not entirely null (Bromberger et al., 1997; Nilsson et al., 1997; Nagata et al., 2000; Palmer et al., 2003; Nagel et al., 2005; Dorjgochoo et al., 2008; Nagata et al., 2012; Emaus et al., 2013; Gold et al., 2013; Gudmundsdottir et al., 2013). For example, in a cross-sectional analysis in the Breakthrough Generations Study (n = 50 678) women, women reporting any strenuous exercise at ages 30–49 had a HR for menopause of 0.96 (95% CI = 0.93–0.98; P = 0.003) compared to women with no strenuous activity (Morris et al., 2012). Though statistically significant, the magnitude of effect was very small, and the median age at menopause was 52 years in both strenuously active and inactive women.
Results were similar in magnitude in the California Teachers Study (n > 97 000 women), the largest prospective study conducted to date of activity and menopause timing (Emaus et al., 2013). Women in the highest quartile of long-term physical activity (equivalent to ≥6 h/week each year from high school through age 45–54) had a HR for menopause of 0.95 (95% CI: 0.92–0.98; P-trend = 0.005) compared to the least active women. Results differed by smoking status, but in the opposite direction from what we observed; associations were null for never smokers but were significant in heavy smokers (HR = 0.89; 95% CI = 0.80–0.99 for current heavy smokers). Cumulatively, physical activity appears either unassociated with early menopause or only modestly linked, with magnitudes of association of likely limited clinical relevance.
A potential limitation of our study is our use of self-reported physical activity, rather than objective measurement, which likely contributes to misclassification. However, our physical activity questionnaire has demonstrated good reproducibility and validity as compared with 7-day diaries (Wolf et al., 1994). Also, these physical activity measures have been significantly associated with other health outcomes such as depression, breast cancer and colon cancer in the NHS2, suggesting that misclassification is likely not substantial enough to obscure important associations (Maruti et al., 2008; Lucas et al., 2011; Morikawa et al., 2011; Zhang et al., 2011; Du et al., 2012). While menopause timing was self-reported as well, the prospective report of onset of menopause has been demonstrated to have high reproducibility over multiple questionnaire cycles (Colditz et al., 1987). We note that residual confounding by psychosocial stress, employment status and other factors potentially related to menopause timing may persist if these were also related to physical activity levels in our population. Additionally, while we know of no data suggesting that the physiology underlying a relation between physical activity and menopausal timing differs by race/ethnicity, we were unable to evaluate potential effect modification by race/ethnicity, as the majority of NHS2 participants were white.
In summary, although physical activity has been shown to have wide ranging positive health effects overall, results of this large, prospective study do not support an important role of physical activity with regard to risk of early menopause.
Authors’ roles
Conception and design of study: M.Z. and E.R.B.J.; analysis and interpretation of data: M.Z., B.W.W., A.C.P.S., J.E.M., S.E.H., B.A.R. and E.R.B.J.; drafting of manuscript or revising critically for important intellectual content: M.Z., B.W.W., A.C.P.S., J.E.M., S.E.H., B.A.R. and E.R.B.J.; and final approval of version to be published: M.Z., B.W.W., A.C.P.S., J.E.M., S.E.H., B.A.R. and E.R.B.J.
Funding
This project was supported by UM1CA176726 and R01HD078517 from the US Department of Health and Human Services, National Institutes of Health.
Conflict of interest
No competing interests declared.
References
- Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, Strath SJ, Brien O, Bassett WL, Schmitz DR, Emplaincourt KH. PO, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc 2000;32:498–504. [DOI] [PubMed] [Google Scholar]
- Aydin ZD, Erbas B, Karakus N, Aydin O, Sule K. Sun exposure and age at natural menopause: a cross-sectional study in Turkish women. Maturitas 2005;52:235–248. [DOI] [PubMed] [Google Scholar]
- Bleil ME, Gregorich SE, Mcconnell D, Rosen MP, Cedars MI. Does accelerated reproductive aging underlie premenopausal risk for cardiovascular disease? Menopause 2013;20:1139–1146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Boutot ME, Purdue-Smithe A, Whitcomb BW, Szegda KL, Manson JE, Hankinson SE, Rosner BA, Bertone-Johnson ER. Dietary Protein intake and early menopause in the Nurses’ Health Study II. Am J Epidemiol 2018;187:1341–1342. (in press). [DOI] [PMC free article] [PubMed] [Google Scholar]
- Broekmans FJ, Soules MR, Fauser BC. Ovarian aging: mechanisms and clinical consequences. Endocr Rev 2009;30:465–493. [DOI] [PubMed] [Google Scholar]
- Bromberger JT, Matthews KA, Kuller LH, Wing RR, Meilahn EN, Plantinga P. Prospective study of the determinants of age at menopause. Am J Epidemiol 1997;145:124–133. [DOI] [PubMed] [Google Scholar]
- Colditz GA, Stampfer MJ, Willett WC, Stason WB, Rosner B, Hennekens CH, Speizer FE. Reproducibility and validity of self-reported menopausal status in a prospective cohort study. Am J Epidemiol 1987;126:319–325. [DOI] [PubMed] [Google Scholar]
- Dorjgochoo T, Kallianpur A, Gao YT, Cai H, Yang G, Li H, Zheng W, Shu XO. Dietary and lifestyle predictors of age at natural menopause and reproductive span in the Shanghai Women’s Health Study. Menopause 2008;15:924–933. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dratva J, Zemp E, Staedele P, Schindler C, Constanza M, Gerbase M, Probst-Hensch N, Rochat T, Ackermann-Liebrich U. The Sapaldia-Team. Variability of reproductive history across the Swiss SAPALDIA cohort—Patterns and main determinants. Ann Hum Biol 2007;34:437–453. [DOI] [PubMed] [Google Scholar]
- Du M, Prescott J, Kraft P, Han J, Giovannucci E, Hankinson SE, De Vivo I. Physical activity, sedentary behavior, and leukocyte telomere length in women. Am J Epidemiol 2012;175:414–422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Eliassen AH, Hankinson SE, Rosner B, Holmes MD, Willett WC. Physical activity and risk of breast cancer among postmenopausal women. Arch Intern Med 2010;170:1758–1764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Emaus A, Dieli-Conwright C, Xu X, Lacey JV Jr, Ingles SA, Reynolds P, Bernstein L, Henderson KD. Increased long-term recreational physical activity is associated with older age at natural menopause among heavy smokers: the California Teachers Study. Menopause 2013;20:282–290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gold EB, Crawford SL, Avis NE, Crandall CJ, Matthews KA, Waetjen LE, Lee JS, Thurston R, Vuga M, Harlow SD. Factors related to age at natural menopause: longitudinal analyses from SWAN. Am J Epidemiol 2013;178:70–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gudmundsdottir S, Flanders W, Augestad L. Physical activity and fertility in women: the North-Trendelag Health Study. Hum Reprod 2009;24:3196–3204. [DOI] [PubMed] [Google Scholar]
- Gudmundsdottir SL, Flanders WD, Augestad LB. Physical activity and age at menopause: the North-Trendelag population-based health study. Climacteric 2013;16:78–87. [DOI] [PubMed] [Google Scholar]
- Jacobsen BK, Heuch I, Kvale G. Age at natural menopause and all-cause mortality: a 37-year follow-up of 19,731 Norwegian Women. Am J Epidemiol 2003;157:923–929. [DOI] [PubMed] [Google Scholar]
- Lucas M, Mekary R, Pan A, Mirzaei F, O’Reilly ÉJ, Willett WC, Koenen K, Okereke OI, Ascherio A. Relation between clinical depression risk and physical activity and time spent watching television in older women: a 10-year prospective follow-up study. Am J Epidemiol 2011;174:1017–1027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Maruti SS, Willett WC, Feskanich D, Rosner B, Colditz GA. A prospective study of age-specific physical activity and premenopausal breast cancer. J Natl Cancer Inst 2008;100:728–737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mikkelsen TF, Graff-Iversen S, Sundby J, Bjertness E. Early menopause, association with tobacco smoking, coffee consumption and other lifestyle factors: a cross-sectional study. BMC Public Health 2007;7:149. [DOI] [PMC free article] [PubMed]
- Miller V. Faculty of 1000 evaluation for Early menopause predicts future coronary heart disease and stroke: the Multi-Ethnic Study of Atherosclerosis. Menopause 2012;19:1081–1087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Morikawa T, Kuchiba A, Yamauchi M, Meyerhardt JA, Shima K, Nosho K, Chan AT, Giovannucci E, Fuchs CS, Ogino S. Association of CTNNB1 (β-catenin) alterations, body mass index, and physical activity with survival in patients with colorectal cancer. JAMA 2011;305:1685–1694. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Morris DH, Jones ME, Schoemaker MJ, Mcfadden E, Ashworth A, Swerdlow AJ. Body mass index, exercise, and other lifestyle factors in relation to age at natural menopause: analyses from the breakthrough generations study. Am J Epidemiol 2012;175:998–1005. [DOI] [PubMed] [Google Scholar]
- Nagata C, Takatsuka N, Kawakami N, Shimizu H. Association of diet with the onset of menopause in Japanese women. Am J Epidemiol 2000;152:863–867. [DOI] [PubMed] [Google Scholar]
- Nagata C, Wada K, Nakamura K, Tamai Y, Tsuji M, Shimizu H. Associations of physical activity and diet with the onset of menopause in Japanese women. Menopause 2012;19:75–81. [DOI] [PubMed] [Google Scholar]
- Nagel G, Altenburg HP, Nieters A, Boffetta P, Linseisen J. Reproductive and dietary determinants of the age at menopause in EPIC-Heidelberg. Maturitas 2005;52:337–347. [DOI] [PubMed] [Google Scholar]
- Nilsson P, Möller L, Köster A, Hollnagel H. Social and biological predictors of early menopause: a model for premature aging. J Intern Med 1997;242:299–305. [DOI] [PubMed] [Google Scholar]
- Ortega-Ceballos PA, Morán C, Blanco-Muñoz J, Yunes-Díaz E, Castañeda-Iñiguez MS, Salmerón J. Reproductive and lifestyle factors associated with early menopause in Mexican women. Salud Publica Mex 2006;48:300–307. [DOI] [PubMed] [Google Scholar]
- Palmer JR, Rosenberg L, Wise LA, Horton NJ, Adams-Campbell LL. Onset of natural menopause in African American women. Am J Public Health 2003;93:299–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Purdue-Smithe AC, Whitcomb BW, Szegda KL, Boutot ME, Manson JE, Hankinson SE, Rosner BA, Troy LM, Michels KB, Bertone-Johnson ER. Vitamin D and calcium intake and risk of early menopause. Am J Clin Nutr 2017;105:1493–1501. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shuster LT, Rhodes DJ, Gostout BS, Grossardt BR, Rocca WA. Premature menopause or early menopause: long-term health consequences. Maturitas 2010;65:161–166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Szegda KL, Whitcomb BW, Purdue-Smithe AC, Boutot ME, Manson JE, Hankinson SE, Rosner BA, Bertone-Johnson ER. Adult adiposity and risk of early menopause. Hum Reprod 2017;32:2522–2531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Torrealday S, Pal L. Premature menopause. Endocrinol Metab Clin North Am 2015;44:543–557. [DOI] [PubMed] [Google Scholar]
- Tworoger SS, Missmer SA, Eliassen AH, Barbieri RL, Dowsett M, Hankinson SE. Physical activity and inactivity in relation to sex hormone, prolactin, and insulin-like growth factor concentrations in premenopausal women. Cancer Causes Control 2007;18:743–752. [DOI] [PubMed] [Google Scholar]
- van Der Voort DJ, van Der Weijer PH, Barentsen R. Early menopause: increased fracture risk at older age. Osteoporos Int 2003;14:525–530. [DOI] [PubMed] [Google Scholar]
- Wellons M, Ouyang P, Schreiner PJ, Herrington DM, Vaidya D. Early menopause predicts future coronary heart disease and stroke: the multi-ethnic study of atherosclerosis. Menopause 2012;19:1081–1087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Whitcomb BW, Purdue-Smithe AC, Szegda KL, Boutot ME, Hankinson SE, Manson JE, Rosner B, Willett WC, Eliassen AH, Bertone-Johnson ER. Cigarette smoking and risk of early natural menopause. Am J Epidemiol 2018;187:696–704. (in press). [DOI] [PMC free article] [PubMed] [Google Scholar]
- Willet WC, Sampson L, Stampfer MJ, Rosner B, Bain C, Witschi J, Hennekens CH, Speizer FE. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol 1985;122:51–65. [DOI] [PubMed] [Google Scholar]
- Willett WC. Nutritional Epidemiology, 3rd ed Oxford, New York: Oxford University Press, 2012. [Google Scholar]
- Wolf AM, Hunter DJ, Colditz GA, Manson JE, Stampfer MJ, Corsano KA, Willett WC. Reproducibility and validity of a self-administered physical activity questionnaire. Int J Epidemiol 1994;23:991–999. [DOI] [PubMed] [Google Scholar]
- Zhang X, Smith-Warner SA, Chan AT, Wu K, Spiegelman D, Fuchs CS, Willett WC, Giovannucci EL. Aspirin use, body mass index, physical activity, plasma C-peptide, and colon cancer risk in US health professionals. Am J Epidemiol 2011;174:459–467. [DOI] [PMC free article] [PubMed] [Google Scholar]