Bilateral Oophorectomy and Accelerated Aging: Cause or Effect?

Background The cause–effect relationship between bilateral oophorectomy and accelerated aging remains controversial. We conducted new analyses to further address this controversy.

The Rochester Epidemiology Project records-linkage system was used to identify all premenopausal women who underwent bilateral oophorectomy for a noncancerous condition before age 50 years between 1988 and 2007 in Olmsted County, MN. Each woman was randomly matched to a referent woman born in the same year (±1 year) who had not undergone bilateral oophorectomy. We studied the rate of accumulation of 18 common chronic conditions over a median of approximately 14 years of follow-up (historical cohort study). Analyses were restricted to women free of any of the 18 chronic conditions at the time of oophorectomy (or index date).

After adjustments for race/ethnicity, education, body mass index, smoking, and age and calendar year at the index date, women who underwent oophorectomy before age 46 years experienced an accelerated rate of accumulation of the 18 chronic conditions considered together (hazard ratio = 1.24; 95% confidence interval: 1.12, 1.37; p < .001). The single-year incidence rate of new conditions was most different in the first 6 years after oophorectomy but the difference attenuated thereafter. Findings did not vary by surgical indication for the oophorectomy.

Bilateral oophorectomy is associated with a higher risk of multimorbidity among women who did not have any of the 18 selected conditions at baseline. The association did not vary by surgical indication for oophorectomy. Our findings suggest that bilateral oophorectomy is causally linked to accelerated aging.

Bilateral oophorectomy, Estrogen therapy, Multimorbidity, Accelerated aging, Cohort study Levine and colleagues recently reported an association between bilateral oophorectomy and accelerated aging as measured by an epigenetic biomarker, the “epigenetic clock.” They concluded that the premature loss of ovarian function may lead to an increase in the extent of DNA methylation, a biological marker of accelerated aging (1,2). Rocca and colleagues recently showed an association between bilateral oophorectomy performed before menopause and the rate of accumulation of multimorbidity using 18 defined chronic conditions (3). Multimorbidity is a complementary clinical and epidemiological marker of accelerated aging (3–6). The two publications by Levine and colleagues and Rocca and colleagues provide important new evidence for a possible causal role of ovarian function in regulating the aging processes in women (1,3). However, it remains difficult in a nonexperimental setting to definitively test whether accelerated aging leads to symptoms that prompt bilateral oophorectomy or whether bilateral oophorectomy prompts accelerated aging (cause–effect uncertainty). In our previous publication, the 1,653 women who underwent bilateral oophorectomy for a noncancer indication had more chronic conditions at the time of oophorectomy than the 1,653 referent women, suggesting that some degree of accelerated aging may have preceded the oophorectomy (3). Therefore, we used inverse probability weights to minimize these differences (3,7,8). Because this statistical method requires a number of assumptions, we are reporting here a set of alternative analyses restricted to women who did not have any of the 18 chronic conditions at the time of oophorectomy and we further stratified analyses by surgical indication.

The study sample and the methods of the Mayo Clinic Cohort Study of Oophorectomy and Aging-2 (MOA-2) have been reported extensively elsewhere (3). In brief, in this historical cohort study, the Rochester Epidemiology Project (REP) records-linkage system was used to identify all premenopausal women who underwent bilateral oophorectomy before age 50 years for a noncancerous condition between 1988 and 2007 in Olmsted County, MN (9). Each woman was randomly matched to a referent woman from the same Olmsted County population who was born in the same year (±1 year) and had not undergone bilateral oophorectomy. We studied the rate of accumulation of 18 common chronic conditions over a median follow-up of approximately 14 years (through December 31, 2014) (3). Details about the ascertainment of chronic conditions through the REP were reported elsewhere (3). In brief, the complete medical records were manually abstracted for demographic, reproductive history, family history of cancer, and life habits information and were searched electronically for all ICD-9 codes that each woman in the study received before and after the index date. In addition, we considered all causes of death listed on the death certificate. These ICD-9 codes were used to define the 20 chronic conditions recommended by the U.S. Department of Health and Human Services (DHHS) to study multimorbidity (5,10–12). However, we excluded from the DHHS list human immunodeficiency virus infections, autism spectrum disorders, and hepatitis and added to the DHHS list anxiety as detailed elsewhere (18 selected conditions) (3,13). Two statistical analysis approaches were used to address the cause–effect uncertainty: (i) we first restricted the analyses to women who did not have any of the 18 chronic conditions at the index date to study de novo development of morbidity and (ii) we then stratified the analyses by surgical indication for oophorectomy (benign ovarian condition vs no ovarian condition). The benign ovarian condition (eg, cysts, endometriosis) was listed by the gynecologist in the medical record as the indication for oophorectomy. Women without a benign ovarian condition were historically considered to have “prophylactic,” “elective,” or “incidental” oophorectomy. Women were censored at death, at the last contact with the system, or on December 31, 2014. The accumulation of multimorbidity was represented graphically using Aalen–Johansen curves (a multistate generalization of cumulative incidence curves; unadjusted curves considering all 18 chronic conditions equally). We also plotted single-year incidence rates for new conditions (per 100 person-years) as a measure of instantaneous density incidence rather than cumulative incidence. We computed hazard ratios (HRs) using Andersen–Gill regression models with age as the time scale (14–16). None of the models violated the proportional hazards assumption. In each analysis, we calculated unadjusted HRs and HRs adjusted using inverse probability weights derived from a logistic regression model including years of education (0–12, 13–16, >16 years), race/ethnicity (white vs other), body mass index (BMI < 30 vs ≥30 kg/m2), cigarette smoking (current or former vs never), and age and calendar year at index date (continuous). We used inverse probability weights to balance the oophorectomy and the referent cohort at baseline on potential confounders and we assessed the balance obtained using the standardized difference of means (absolute value) (7,8). In addition to the analyses stratified by surgical indication, we explored possible interactions using analyses stratified by age at oophorectomy (cutoff ≤ 45 years consistent with the definition of premature or early menopause), estrogen therapy (ET; within age strata), calendar year, and smoking. Referent women were stratified based on the original matching with the oophorectomy women for surgical indication, age at oophorectomy, calendar year, and ET. However, they were stratified for smoking based on their own smoking status.

We restricted the initial full cohorts of 1,653 women who underwent oophorectomy and 1,653 referent women to 659 women who underwent oophorectomy and 888 referent women who did not have any of the 18 chronic conditions at baseline. To increase statistical power, all women free of any condition at the index date were included, regardless of the original matching. In the oophorectomy cohort, the median age at index date was 44.0 (interquartile range [IQR]: 40.0–47.0), 97.3% of the women were white, 66.9% had 13 or more years of education, 41.0% were current or former smokers, and 25.3% had BMI ≥ 30. The median follow-up time was 16.4 years (IQR: 11.0–21.5) and 26 (4.0%) women died during follow-up. In the referent cohort, the median age at index date was 44.0 (IQR: 39.0–47.0), 93.9% of the women were white, 72.9% had 13 or more years of education, 38.1% were current or former smokers, and 22.0% had BMI ≥ 30. The median follow-up time was 15.1 years (IQR: 10.4–20.0) and 23 (2.6%) women died during follow-up. Tables 1 and 2 and Figure 1 show the results of our cohort analyses. The counts for the 18 chronic conditions in the overall analysis were reported in Table 1. Table 1. | Bilateral Oophorectomy (N = 659) | Referent Women (N = 888) | Total (N = 1,547) | |||| |---|---|---|---|---|---|---| | Chronic Condition | N | % | N | % | N | % | | Five mental health conditions | |||||| | Depression | 190 | 28.8 | 194 | 21.8 | 384 | 24.8 | | Anxiety | 133 | 20.2 | 136 | 15.3 | 269 | 17.4 | | Substance abuse disorders | 22 | 3.3 | 21 | 2.4 | 43 | 2.8 | | Dementia | 9 | 1.4 | 9 | 1.0 | 18 | 1.2 | | Schizophrenia | 2 | 0.3 | 5 | 0.6 | 7 | 0.5 | | Seven cardiovascular or metabolic conditions | |||||| | Hyperlipidemia | 315 | 47.8 | 334 | 37.6 | 649 | 42.0 | | Hypertension | 213 | 32.3 | 251 | 28.3 | 464 | 30.0 | | Diabetes | 154 | 23.4 | 161 | 18.1 | 315 | 20.4 | | Cardiac arrhythmias | 121 | 18.4 | 144 | 16.2 | 265 | 17.1 | | Coronary artery disease | 30 | 4.6 | 31 | 3.5 | 61 | 3.9 | | Stroke | 19 | 2.9 | 24 | 2.7 | 43 | 2.8 | | Congestive heart failure | 4 | 0.6 | 10 | 1.1 | 14 | 0.9 | | Six other somatic conditions | |||||| | Arthritis | 276 | 41.9 | 257 | 28.9 | 533 | 34.5 | | Cancer (all types) | 91 | 13.8 | 122 | 13.7 | 213 | 13.8 | | Asthma | 77 | 11.7 | 59 | 6.6 | 136 | 8.8 | | Chronic obstructive pulmonary disease | 121 | 18.4 | 117 | 13.2 | 238 | 15.4 | | Osteoporosis | 60 | 9.1 | 67 | 7.5 | 127 | 8.2 | | Chronic kidney disease | 33 | 5.0 | 26 | 2.9 | 59 | 3.8 | | Total number of eventsa | 1,870 | — | 1,968 | — | 3,838 | — | Note: aThe total numbers of chronic conditions are also reported in Table 2, first row. Table 2. | Strata | Bilateral Oophorectomy | Referent Womena | Unadjusted Models | Adjusted Modelsb | |||||| |---|---|---|---|---|---|---|---|---|---|---| | N at Risk | Person-Years | N of Events | N at Risk | Person-Years | N of Events | Hazard Ratio (95% CI) | p Value | Hazard Ratio (95% CI) | p Value | | | Overall | 659 | 10,191 | 1,870 | 888 | 13,097 | 1,968 | 1.20 (1.11, 1.30) | <.001 | 1.18 (1.09, 1.28) | <.001 | | Age ≤45 y | 420 | 6,569 | 1,154 | 592 | 8,624 | 1,184 | 1.25 (1.13, 1.39) | <.001 | 1.24 (1.12, 1.37) | 45c | 295 | 3,982 | 703 | 337 | 4,307 | 701 | 1.07 (0.94, 1.21) | .33 | 1.07 (0.95, 1.22) | .28 | | No ET or ≤45 | 58 | 552 | 101 | 92 | 988 | 129 | 1.36 (1.02, 1.82) | .04 | 1.18 (0.86, 1.62) | .32 | | Benign conditiond | 233 | 3,720 | 645 | 307 | 4,625 | 615 | 1.27 (1.10, 1.47) | <.001 | 1.24 (1.08, 1.43) | .003 | | No indicatione | 187 | 2,849 | 509 | 285 | 3,998 | 569 | 1.23 (1.06, 1.43) | .006 | 1.24 (1.07, 1.45) | .006 | | Age 46–49 y | 239 | 3,621 | 716 | 296 | 4,473 | 784 | 1.12 (0.99, 1.26) | .06 | 1.11 (0.98, 1.25) | .10 | | ET > 49f | 179 | 2,522 | 476 | 208 | 3,046 | 572 | 0.99 (0.86, 1.14) | .90 | 0.99 (0.85, 1.14) | .83 | | No ET or ≤49 | 53 | 619 | 122 | 69 | 758 | 129 | 1.14 (0.87, 1.49) | .34 | 1.06 (0.80, 1.41) | .68 | | Benign conditiond | 59 | 842 | 157 | 78 | 1,082 | 197 | 1.03 (0.78, 1.36) | .82 | 0.94 (0.70, 1.26) | .66 | | No indicatione | 180 | 2,780 | 559 | 218 | 3,392 | 587 | 1.15 (1.01, 1.31) | .04 | 1.13 (1.00, 1.29) | .06 | | Benign conditiond | 292 | 4,562 | 802 | 385 | 5,707 | 812 | 1.21 (1.07, 1.38) | .003 | 1.18 (1.04, 1.34) | .01 | | No indicatione | 367 | 5,629 | 1,068 | 503 | 7,390 | 1,156 | 1.19 (1.08, 1.31) | <.001 | 1.18 (1.07, 1.30) | .001 | | Years 1988–1997 | 389 | 7,316 | 1,345 | 460 | 8,558 | 1,321 | 1.18 (1.07, 1.29) | <.001 | 1.18 (1.07, 1.29) | <.001 | | Years 1998–2007 | 270 | 2,875 | 525 | 428 | 4,539 | 647 | 1.25 (1.08, 1.45) | .002 | 1.22 (1.05, 1.41) | .008 | | Smokers | 270 | 4,215 | 852 | 338 | 4,988 | 895 | 1.13 (1.01, 1.28) | .04 | 1.15 (1.02, 1.29) | .02 | | Nonsmokers | 389 | 5,976 | 1,018 | 550 | 8,109 | 1,073 | 1.24 (1.12, 1.37) | <.001 | 1.22 (1.10, 1.35) | <.001 | Note: ET = estrogen therapy. aReferent women were stratified based on the original matching with the oophorectomy women for age at oophorectomy, estrogen therapy, surgical indication, and calendar year. However, they were stratified for smoking based on their own smoking status. bHazard ratios calculated using Andersen–Gill regression models with age as the time scale and adjusted using inverse probability weights derived from a logistic regression model including years of education (≤12, 13–16, >16), race/ethnicity (white vs non-white), body mass index (<30 vs ≥30 kg/m2), cigarette smoking (current or former vs never), age at index date (continuous), and calendar year at index date (continuous). These adjustments were done separately in each stratum to maximize the balance at index date. In the overall analysis, the range of the stabilized inverse probability weights was 0.7–2.1 for the oophorectomy women and 0.7–1.4 for the referent women. None of the interactions by age was significant. cWomen who were taking systemic ET (only oral or transdermal) on their 46th birth date, after bilateral oophorectomy. Women who died or were lost to follow-up prior to their 46th birth date, or had not reached age 46 years as of December 31, 2014 were not included in this analysis. Follow-up for these analyses was started at age 46 years. None of the interactions by ET was significant in the ≤45-year age stratum. dThe benign condition (eg, cysts, endometriosis) was listed by the gynecologist in the medical record at the time of oophorectomy but may not have been the sole indication for the surgery. None of the interactions by surgical indication was significant. eWomen without a benign ovarian condition. Historically, the terms “prophylactic,” “elective,” or “incidental” oophorectomy were used; however, we prefer to avoid these terms. f Women who were taking systemic ET (only oral or transdermal) on their 50th birth date, after bilateral oophorectomy. Women who died or were lost to follow-up prior to their 50th birth date, or had not reached age 50 years as of December 31, 2014 were not included in this analysis. Follow-up for these analyses was started at age 50 years. None of the interactions by ET was significant in the 46–49-year age stratum. The two cohorts were not highly imbalanced before the adjustments using inverse probability weights (standardized difference of means < 25% of the SD for all variables) and the adjustments improved the balance successfully (standardized difference of means < 5% of the SD; data not shown). The range of the weights used in the overall analysis was reported in Table 2 (footnote b). The HR for the accumulation of multimorbidity after bilateral oophorectomy was 1.18 overall (95% CI: 1.09, 1.28; p < .001) and 1.24 (95% CI: 1.12, 1.37; p < .001) in women who underwent oophorectomy at age 45 years and younger. The single-year incidence rate of new conditions was most different between oophorectomy and referent women in the first 6 years following the oophorectomy (Figure 2, shaded area) but the difference attenuated thereafter. To address the cause–effect uncertainty further, we also stratified women free of chronic conditions at baseline into those who did and those who did not have a benign ovarian condition listed as the surgical indication. In women who did not have a benign ovarian indication, the HR for accumulation of multimorbidity was 1.18 overall (95% CI: 1.07, 1.30; p = .001) and 1.24 in women aged 45 years and younger at index date (95% CI: 1.07, 1.45; p = .006). The HRs were similar for women who did have a benign ovarian indication. Women who underwent bilateral oophorectomy at age 45 years and younger and received ET through age 45 years had a lesser accumulation of multimorbidity compared to women who did not receive ET. However, the difference between the two strata did not reach statistical significance (Table 2).

Although a subset of women who underwent bilateral oophorectomy in our population had one or more chronic conditions present at the index date, approximately 41% of women who underwent oophorectomy at age 45 years and younger and 57% of the age-matched referent women had none of the 18 study conditions (3). In our analyses restricted to women free of any of the 18 conditions at baseline, women who underwent bilateral oophorectomy experienced subsequent accelerated aging regardless of whether there was an ovarian surgical indication. The 367 women who did not have any of the 18 conditions at index date and did not have an ovarian condition prompting the oophorectomy, can be considered an optimal sample to resolve the cause–effect uncertainty about the association of bilateral oophorectomy and accelerated aging (17). Our findings provide additional observational evidence that bilateral oophorectomy adversely impacts the aging process. It remains unclear which ovarian functions or specific hormones might be involved in accelerating aging after bilateral oophorectomy (17–20). Our analyses stratified by ET did not show a significant effect possibly because of limited statistical power, because of the variable effects of ET across the 18 chronic conditions, and/or because of the variable doses and formulations of ET. In summary, our study suggests that bilateral oophorectomy is causally linked to accelerated aging. Funding This work was supported by the resources of the Rochester Epidemiology Project, which is supported by the National Institute on Aging of the National Institutes of Health under (award number R01 AG034676). This work was also supported by funds from the Mayo Clinic Research Committee (to W.A.R.). W.A.R. was partly supported by the National Institutes of Health (grants P50 AG044170, U01 AG006786, and P01 AG004875). V.M.M. was partly supported by the National Institutes of Health (grant P50 AG044170). J.L.K. was partly supported by the National Institutes of Health (AG013925). Conflict of Interest W.A.R. reports grants from the National Institutes of Health. S.S.F. reports consulting fees from Mithra Pharmaceuticals. E.A.S. reports grants from the National Institutes of Health and consulting fees from AbbVie, Bayer, GlaxoSmithKline, Gynesonics, Astellas Pharmaceuticals, Welltwigs, Viteava Pharmaceuticals, and Allergan, outside the submitted work. In addition, E.A.S. has a patent: Methods and Compounds for Treatment of Abnormal Uterine Bleeding (US 6440445 issued to None). V.M.M. and J.L.K. report grants from the National Institutes of Health.

- 1. Levine ME, Lu AT, Chen BH, et al. Menopause accelerates biological aging. Proc Natl Acad Sci USA. 2016;113:9327–9332. doi:10.1073/pnas.1604558113 [DOI] [PMC free article] [PubMed] [Google Scholar] - 2. Horvath S. DNA methylation age of human tissues and cell types. Genome Biol. 2013;14:R115. doi:10.1186/gb-2013-14-10-r115 [DOI] [PMC free article] [PubMed] [Google Scholar] - 3. Rocca WA, Gazzuola-Rocca L, Smith CY, et al. Accelerated accumulation of multimorbidity after bilateral oophorectomy: a population-based cohort study. Mayo Clin Proc. 2016;91:1577–1589. doi:10.1016/j.mayocp.2016.08.002 [DOI] [PMC free article] [PubMed] [Google Scholar] - 4. Fabbri E, Zoli M, Gonzalez-Freire M, Salive ME, Studenski SA, Ferrucci L. Aging and multimorbidity: new tasks, priorities, and frontiers for integrated gerontological and clinical research. J Am Med Dir Assoc. 2015;16:640–647. doi:10.1016/j.jamda.2015.03.013 [DOI] [PMC free article] [PubMed] [Google Scholar] - 5. St Sauver JL, Boyd CM, Grossardt BR, et al. Risk of developing multimorbidity across all ages in an historical cohort study: differences by sex and ethnicity. BMJ Open. 2015;5:e006413. doi:10.1136/bmjopen-2014-006413 [DOI] [PMC free article] [PubMed] [Google Scholar] - 6. Fabbri E, An Y, Zoli M, et al. Association between accelerated multimorbidity and age-related cognitive decline in older baltimore longitudinal study of aging participants without dementia. J Am Geriatr Soc. 2016;64:965–972. doi:10.1111/jgs.14092 [DOI] [PMC free article] [PubMed] [Google Scholar] - 7. Harder VS, Stuart EA, Anthony JC. Propensity score techniques and the assessment of measured covariate balance to test causal associations in psychological research. Psychol Methods. 2010;15:234–249. doi:10.1037/a0019623 [DOI] [PMC free article] [PubMed] [Google Scholar] - 8. Rubin DB. The design versus the analysis of observational studies for causal effects: parallels with the design of randomized trials. Stat Med. 2007;26:20–36. doi:10.1002/sim.2739 [DOI] [PubMed] [Google Scholar] - 9. St Sauver JL, Grossardt BR, Yawn BP, et al. Data resource profile: the Rochester Epidemiology Project (REP) medical records-linkage system. Int J Epidemiol. 2012;41:1614–1624. doi:10.1093/ije/dys195 [DOI] [PMC free article] [PubMed] [Google Scholar] - 10. Goodman RA, Posner SF, Huang ES, Parekh AK, Koh HK. Defining and measuring chronic conditions: imperatives for research, policy, program, and practice. Prev Chronic Dis. 2013;10:E66. doi:10.5888/pcd10.120239 [DOI] [PMC free article] [PubMed] [Google Scholar] - 11. Cohen JW, Cohen SB, Banthin JS. The medical expenditure panel survey: a national information resource to support healthcare cost research and inform policy and practice. Med Care. 2009;47(7 suppl 1):S44–S50. doi:10.1097/MLR.0b013e3181a23e3a [DOI] [PubMed] [Google Scholar] - 12. Rocca WA, Boyd CM, Grossardt BR, et al. Prevalence of multimorbidity in a geographically defined American population: patterns by age, sex, and race/ethnicity. Mayo Clin Proc. 2014;89:1336–1349. doi:10.1016/j.mayocp.2014.07.010 [DOI] [PMC free article] [PubMed] [Google Scholar] - 13. Bobo WV, Yawn BP, St Sauver JL, Grossardt BR, Boyd CM, Rocca WA. Prevalence of combined somatic and mental health multimorbidity: patterns by age, sex, and race/ethnicity. J Gerontol A Biol Sci Med Sci. 2016;71:1483–1491. doi:10.1093/gerona/glw032 [DOI] [PMC free article] [PubMed] [Google Scholar] - 14. Therneau TM, Grambsch PM. Modeling Survival Data: Extending the Cox Model. New York, NY: Springer; 2000. [Google Scholar] - 15. Jackson C. Multi-state models for panel data: the MSM package for R. J Stat Softw. 2011;38:1–29. doi:10.18637/jss.v038.i08 [Google Scholar] - 16. Al-Khalidi HR, Hong Y, Fleming TR, Therneau TM. Insights on the robust variance estimator under recurrent-events model. Biometrics. 2011;67:1564–1572. doi:10.1111/j.1541-0420.2011.01589.x [DOI] [PMC free article] [PubMed] [Google Scholar] - 17. Rocca WA, Shuster LT, Grossardt BR, et al. Long-term effects of bilateral oophorectomy on brain aging: unanswered questions from the Mayo Clinic Cohort Study of Oophorectomy and Aging. Womens Health (Lond). 2009;5:39–48. doi:10.2217/17455057.5.1.39 [DOI] [PMC free article] [PubMed] [Google Scholar] - 18. 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:10.1016/j.maturitas.2009.08.003 [DOI] [PMC free article] [PubMed] [Google Scholar] - 19. Rocca WA, Grossardt BR, Shuster LT. Oophorectomy, estrogen, and dementia: a 2014 update. Mol Cell Endocrinol. 2014;389:7–12. doi:10.1016/j.mce.2014.01.020 [DOI] [PMC free article] [PubMed] [Google Scholar] - 20. Faubion SS, Kuhle CL, Shuster LT, Rocca WA. Long-term health consequences of premature or early menopause and considerations for management. Climacteric. 2015;18:483–491. doi:10.3109/13697137.2015.1020484 [DOI] [PMC free article] [PubMed] [Google Scholar]