Frequency and type of premature or early menopause in a geographically defined American population

The study sample was derived from an existing cohort study, the Mayo Clinic Cohort Study of Oophorectomy and Aging – 2 (MOA-2) [ 6 , 7 ]. As part of that study, we identified 1,653 women of age 18–50 years who resided in Olmsted County, MN between 1988 and 2007. These women were originally identified randomly from a complete enumeration of the population to serve as potential age-matched (±1 year) referent persons for 1,653 women who had undergone bilateral oophorectomy in the same time window. Women were excluded and randomly replaced for the MOA-2 study if they did not provide authorization to use their medical records for research or for other administrative reasons. In addition, randomly selected women who had undergone bilateral oophorectomy before the time of recruitment were excluded and replaced for the MOA-2 study. By contrast, women who had undergone bilateral oophorectomy were relevant for the current study and were retained in the current sample. Women were retained for this study regardless of any gynecologic disease or any gynecologic surgery. Therefore, our final sample can be considered a random sample of the complete Olmsted County female population in the defined age range over a 20-year time window (1988–2007). All research activities were approved by the Mayo Clinic and Olmsted Medical Center Institutional Review Boards. Even though menopause is defined by cessation of menses, cessation of menses is not always the result of ovarian insufficiency. To distinguish the cessation of menses from menopause (the cessation of ovarian follicular activity), we defined age at menopause as: 1) the age of reaching ≥12 months of continuous spontaneous amenorrhea, 2) the age of bilateral oophorectomy, 3) or the age of reaching ≥12 months of continuous amenorrhea induced by chemotherapy or radiation therapy during the treatment or within one year of the end of treatment (in women with an intact uterus). For these three groups of women, age at cessation of menses and age at menopause coincided. Spontaneous menopause was diagnosed in women who experienced ≥12 months of continuous amenorrhea in the absence of a surgical or medical cause. For women who underwent hysterectomy with conservation of at least one ovary or endometrial ablation before reaching spontaneous menopause, the age at cessation of menses was known but the age at ovarian insufficiency was unknown. In agreement with the World Health Organization and the North American Menopause Society, we used the term premature when menopause occurred before age 40 years, early for ages 40–44, normal for ages 45–54, and late after age 54 years [ 8 , 9 ]. The complete medical records included in the Rochester Epidemiology Project medical-records-linkage system were reviewed by a physician (LGR) to confirm the menopausal status through December 31, 2021, and to collect demographic and clinical characteristics. Details about the Rochester Epidemiology Project were reported elsewhere [ 10 – 13 ]. For most women, the onset of menopause was documented by a physician or a nurse in the medical records, but hormone tests were not performed as part of routine medical care (absence of a specific indication for hormonal studies). Women who had not yet reached cessation of menses or had unknown menopausal status at the end of follow-up were excluded from the analyses. Given the time frame of the sampling and of data collection, the measure of occurrence is best described as a cumulative frequency (percent). For some women, the onset of menopause occurred before the time of recruitment, but for most women the onset occurred during follow-up. Women were divided into three groups by type of cessation of menses: 1) women with known age at onset of spontaneous ovarian insufficiency, 2) women with known age at onset of induced ovarian insufficiency (induced by bilateral oophorectomy, chemotherapy, or radiation therapy performed before the onset of spontaneous menopause), and 3) women with unknown age at onset of ovarian insufficiency after undergoing hysterectomy or endometrial ablation. Women who underwent spontaneous or induced menopause were also divided into four groups by age at menopause: 1) premature (54 years). Characteristics for all women were compared by type and by age at menopause using chi-square or Kruskal-Wallis tests. Age at cessation of menses or age at menopause with ovarian insufficiency were reported as median and interquartile range (IQR). Age at cessation of menses or menopause were also displayed graphically using cumulative percentage and kernel density. The frequencies of premature or early menopause were reported as percentages and 95% Wilson score confidence intervals (95% CI). All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC).

Figure 1 shows a flowchart of inclusions and exclusions for the study. The medical records of the 1,653 women included in the study were reviewed to assess menopause status through December 31, 2021. A total of 196 women had not yet reached cessation of menses at the end of follow-up and the menopause status or the age at onset of menopause could not be determined for 33 women. A total of 1,424 women experienced cessation of menses within the study time frame. Cessation of menses was premature for 174 women (12.2%), early for 138 women (9.7%), normal for 1,030 women (72.3%), and late for 85 (5.8%). Table 1 summarizes the demographic and clinical characteristics of all women who experienced cessation of menses, stratified by type of cessation of menses. The three groups by type of cessation of menses differed significantly by years of education, smoking, age at menarche, and oral contraceptive use. Spontaneous menopause was observed for 1,015 women (71.3%). An additional 138 women (9.7%) underwent bilateral oophorectomy (with or without concurrent hysterectomy) and in 17 women (1.2%) the onset of menopause was induced by chemotherapy or radiation therapy prior to reaching spontaneous menopause. Age at ovarian insufficiency was unknown for 231 women (16.2%) who underwent hysterectomy with at least one ovary conserved and 23 women (1.6%) who underwent endometrial ablation. Among the 1,015 women who underwent spontaneous menopause (71.3% of the total sample), 4 (0.4%) were premature, 53 (5.2%) were early, and 78 (7.7%) were late. Of 155 women who underwent menopause induced by oophorectomy, chemotherapy, or radiation therapy (10.9% of the total sample), 32 (20.6%) were premature, 20 (12.9%) were early, and 4 (2.6%) were late (but prior to reaching spontaneous menopause). Of 254 women who underwent hysterectomy with at least one ovary conserved or endometrial ablation (17.8% of the total sample), 138 (54.3%) were premature, 65 (25.6%) were early, and none (0.0%) were late at cessation of menses. Among women with premature cessation of menses, hysterectomy or endometrial ablation were the most common causes (138 of 174 women; 79.3%), and among women with early cessation of menses, more than half underwent hysterectomy, endometrial ablation, or induced menopause (85 of 138 women; 61.6%). Among women in the combined strata of premature or early cessation of menses, the most common causes were hysterectomy, endometrial ablation, bilateral oophorectomy, chemotherapy, or radiation therapy combined (255 of 312 women; 81.7%). Considering spontaneous and induced menopause, the frequency was 3.1% (95% CI, 2.2–4.2) for premature menopause and 6.2% (95% CI, 5.0–7.8) for early menopause. Including also the women who underwent hysterectomy with at least one ovary conserved or endometrial ablation, the overall frequency substantially increased to 12.2% (95% CI, 10.6–14.0) for premature cessation of menses and 9.7% (95% CI, 8.3–11.3) for early cessation of menses. Considering only spontaneous menopause, the frequency reduced to 0.4% (95% CI, 0.2–1.0) for premature and 5.2% (95% CI, 4.0–6.8) for early menopause. Table 2 shows the demographic and clinical characteristics of women stratified by age at onset of menopause. Women with hysterectomy or endometrial ablation were excluded from Table 2 . The three groups by age at menopause differed significantly by years of education, smoking, oral contraceptive use, and number of pregnancies. Among women with premature menopause, the majority experienced induced menopause (32 of 36 women; 88.9%), and bilateral oophorectomy was the single most common cause (28 of 36 women; 77.8%). Figure 2 , panel A shows the cumulative percent of women reaching cessation of menses or menopause separately by type. The median age at menopause was 51.0 (IQR, 49.0–52.0) years for spontaneous menopause and 46.0 (IQR, 41.0–49.0) years for induced menopause. The median age at cessation of menses for women who underwent hysterectomy or endometrial ablation was 38.0 (IQR, 33.0–44.0) years. Figure 2 , panel B shows the same distribution as in panel A using the kernel density. The distribution by age at cessation of menses was shifted earlier for women who underwent bilateral oophorectomy, chemotherapy, or radiation therapy and even further for women who underwent hysterectomy or endometrial ablation.

We reported a complete picture of the occurrence of cessation of menses and of menopause in a geographically defined American population. When we considered only the spontaneous occurrence of menopause, our frequency of premature or early menopause was lower than in previous studies (0.4% for premature and 5.2% for early) [ 1 ]. However, if we included women who underwent menopause induced by bilateral oophorectomy, chemotherapy, or radiation therapy, the frequency increased to 3.1% for premature and 6.2% for early menopause. Interestingly, the most common causes of premature or early cessation of menses were hysterectomy with at least one ovary conserved, endometrial ablation, or bilateral oophorectomy. Therefore, when we included women who underwent hysterectomy or endometrial ablation (and had an unknown age at ovarian insufficiency), our overall frequency further increased to 12.2% for premature and 9.7% for early cessation of menses. Historically, the definition of menopause based on amenorrhea has incorrectly implied that the uterus is the key organ [ 2 , 14 , 15 ]. Historically, the term menopause was introduced when the only way to infer the functioning of the ovaries was through their effect on the uterus (endometrium). However, if the uterus is removed but at least one ovary is conserved, the woman will undergo permanent cessation of menses, but she may not experience ovarian insufficiency. The onset of ovarian insufficiency may be difficult to determine in the absence of the uterus and of menses [ 15 ]. Therefore, a clear separation of the cessation of menses from the cessation of ovarian function is a useful clarification for both clinicians and the lay public [ 2 ]. As with most biological events, the transition from premature to early, from early to normal, and from normal to late menopause is continuous rather than discrete. Therefore, all cut-off ages are arbitrary and based on the conventional use of 5-year age groups in the medical literature. We used the cut-off ages suggested by the World Health Organization and the North American Menopause Society [ 8 , 9 ]. The authors of a 2019 meta-analysis emphasized that the frequency of spontaneous premature or early menopause may be higher than previously believed [ 1 ]. They reported 3.7% for premature and 12.2% for early spontaneous menopause. Our frequencies for spontaneous menopause were a bit lower (0.4% for premature and 5.2% for early). However, when we included women who underwent induced menopause (10.9% of the total sample), the frequency was 3.1% for premature and 6.2% for early menopause. In addition, when we also included the women who underwent hysterectomy or endometrial ablation (17.8% of the total sample), the overall frequency was even higher: 12.2% for premature and 9.7% for early cessation of menses. Therefore, the direct comparison of the frequency of premature or early menopause across populations or in the same population over time may be problematic. The frequency of hysterectomy and of bilateral oophorectomy have changed over time and across populations [ 3 – 5 ]. It is possible that some of the women who underwent hysterectomy or bilateral oophorectomy were destined to experience premature or early spontaneous menopause. Therefore, the gynecologic surgery may have concealed the early manifestation of spontaneous menopause in these women. Similarly, changes over time in the use of chemotherapy or of agents to preserve ovarian function during chemotherapy may influence trends over time. There are several strengths to our study. First, to the best of our knowledge, this is one of the few studies in which age at menopause was derived from medical record abstraction rather than from self-report by women. Women may not be able to recall the precise timing of menopause (onset and duration of amenorrhea or onset of menopausal symptoms) and may not be aware of the possible iatrogenic or surgical causes of menopause. Second, we clearly separated cessation of menses from menopause with ovarian insufficiency. Although in most women the two events co-occur, in 17.8% of the women in our population they did not co-occur (254 of 1,424 women). Third, by using a sample representative of the general population, we were able to provide frequencies by type of premature menopause quite different from the frequencies observed in studies based on specialty clinics [ 16 ]. There are also several limitations to our study. First, our study involved a single geographically defined population in the United States, and the observed findings may differ in other populations. Second, for some women who experienced spontaneous premature menopause, we may have reported a delayed age at onset of menopause because we focused on cessations of menses rather than on menstrual irregularities and blood tests were not routinely performed. Third, for some of the women, the menopausal status or the age at menopause could not be extracted from the medical records. Fourth, the ability to observe the occurrence of spontaneous menopause in each population is related to the frequency and type of gynecologic surgeries performed. The surgeries may conceal the spontaneous occurrence of ovarian insufficiency. Therefore, comparison of age at menopause across populations within the United States or worldwide may be problematic. We provided a full description of the causes and types of cessations of menses in a geographically defined population. Our study may serve as a model for future studies in other populations in the United States and worldwide. Approximately 3% of women in our population underwent spontaneous or induced premature menopause. Because of the variability in clinical presentation and the inconsistency of symptoms, the diagnosis of premature spontaneous menopause is often delayed, and women do not receive timely hormone replacement therapy. Even though clinical trials data are not available, observational data suggest that hormone replacement therapy may improve the quality of life and may reduce the risk of future osteoporosis, cardiovascular disease, dementia, and other chronic diseases caused by premature estrogen deprivation. Primary care providers and specialists should be aware of premature menopause in young women presenting with menstrual irregularities [ 17 ]. There remains some confusion about the need of hormone replacement therapy for women who experience premature ovarian insufficiency, particularly when they do not manifest major menopausal symptoms [ 18 ]. Consistent with several clinical guidelines, our recommendation is that these women should be treated at least until the median age of spontaneous menopause (age 50–51 years) for the prevention of chronic diseases, even in the absence of menopause symptoms [ 18 – 20 ]. In our study, the most common cause of premature ovarian insufficiency was bilateral oophorectomy (28 of 36 women; 77.8%). Because bilateral oophorectomy is performed most commonly for the prevention of ovarian cancer, hormone replacement therapy may be delayed, discontinued too early, or completely avoided by some women for fear of cancer [ 20 ]. For women already affected by cancer, particularly hormone-dependent cancers like breast or endometrial, systemic hormone replacement therapy may not be an option and the treatment of premature menopause may be particularly complex. It has been hypothesized that women who undergo hysterectomy with at least one ovary conserved before the onset of spontaneous menopause may experience premature or early ovarian insufficiency [ 21 – 23 ]. However, in the absence of sequential testing of hormone levels or of the use of some peripheral biomarker of ovarian reserve, it is difficult to confirm the hypothesis. The appearance of clinical manifestations of menopause, such as vasomotor symptoms, is not a valid marker of the time of onset of ovarian insufficiency [ 24 ]. These women may be at higher risk of subsequent cardiovascular and metabolic conditions, depression, and anxiety [ 22 , 25 ]. Unfortunately, 17.8% of women in our population underwent hysterectomy or endometrial ablation before reaching spontaneous menopause and may have suffered from delayed recognition and treatment of menopause symptoms and may have missed the opportunity for prevention of several chronic diseases.

Because of the growing body of literature demonstrating that women who undergo premature or early menopause are at increased risk for several specific diseases, for accelerated aging and multi-morbidity, and for increased mortality, our findings should prompt preventive interventions [ 2 , 4 , 26 ]. The first and most urgent intervention is changing the clinical practice of removing the ovaries for the prevention of ovarian cancer in women at average risk. The harm-to-benefit balance is clearly in favor of ovarian conservation [ 2 ]. A second urgent intervention is a complete reassessment of the practice of hysterectomy for benign conditions. We emphasize that, in our study, only approximately 4% of the hysterectomies with conservation of at least one ovary were performed for a uterine malignancy. It has been suggested by us and by others that the benefit-to-harm balance is again in favor of uterine conservation for a majority of women [ 4 ]. We hope that this study will contribute to a shift toward a more conservative use of gynecologic surgeries for benign conditions in the population. A third urgent intervention is to increase awareness and clinical expertise for the management of premature and early menopause, regardless of its etiology [ 17 ]. Prompt recognition and treatment of the symptoms may improve the outcomes.

In 2016, the authors of an international meta-analysis concluded that there was a scarcity of data on the prevalence of premature and early menopause. The meta-analysis focused on spontaneous menopause and the authors used the term natural menopause [ 1 ]. However, menopause can follow the spontaneous development of ovarian insufficiency, can be surgically induced by bilateral oophorectomy, or can be medically induced by chemotherapy or radiation therapy [ 2 ]. In addition, women who undergo hysterectomy with conservation of at least one ovary or endometrial ablation also experience cessation of menses (≥12 months of continuous amenorrhea) but not ovarian insufficiency, and age at onset of ovarian insufficiency remains difficult to determine in these women [ 2 ]. In this study, we investigated age at cessation of menses as well as age at menopause with ovarian insufficiency in a geographically defined American population. We defined menopause as the cessation of menses due to spontaneous or induced ovarian insufficiency. However, for completeness, we also studied women who underwent hysterectomy with at least one ovary conserved or endometrial ablation. The study of women who underwent induced menopause, hysterectomy, or endometrial ablation is essential because these medical or surgical practices conceal the time of spontaneous menopause (censoring), and the frequency of these practices changes across populations and over time [ 3 – 5 ]. We also studied the frequency of premature and early menopause.