Bleeding patterns during the menopausal transition in the multi-ethnic Study of Women's Health Across the Nation (SWAN): a prospective cohort study

Authors


Abstract

Objective

Previous studies describing menses duration and heaviness of flow during the menopausal transition (MT) have been short in duration and limited to white women. We estimated the frequency of and risk factors for prolonged bleeding, spotting and heavy bleeding during the MT in an ethnically diverse population.

Design

Prospective community-based cohort study.

Setting

USA: southeastern Michigan, northern California and Los Angeles, California.

Population

A total of 1320 midlife women who participated in the Study of Women's Health Across the Nation (SWAN) Menstrual Calendar Substudy. Participants included African-American, white, Chinese, and Japanese women.

Methods

Women completed daily menstrual calendars from 1996 to 2006, and provided information on hormone therapy, smoking and physical activity. Annual measures included height and weight. Kaplan–Meier survival analysis and multivariable regression were used to analyse the data.

Main outcome measures

Menses of 10+ days, spotting of 6+ days, heavy bleeding of 3+ days.

Results

At least three occurrences of menses 10+ days was reported by 77.7% (95% confidence interval [95% CI] 56.7–93.2), of 6+ days of spotting by 66.8% (95% CI 55.2–78.0) and of 3+ days of heavy bleeding by 34.5% (95% CI 30.2–39.2) of women. Menses of 10+ days, 6+ days of spotting, and 3+ days of heavy bleeding were associated with MT stage, uterine fibroids, hormone use and ethnicity. Body mass index was associated with 3+ days of heavy bleeding.

Conclusions

These data provide clinicians and women with important information about the expected frequency of prolonged and heavy bleeding and spotting during the menopausal transition that may facilitate clinical decision making.

Introduction

Gynecological disorders, including abnormal uterine bleeding, are a major cause of inpatient hospital admissions in the USA, accounting for 14% of all hospitalisations among women aged 45–54 years old from 1998 through to 2005.[1] The annual direct cost of treating abnormal uterine bleeding is estimated to be US$ 1 billion.[2] While the burden of abnormal uterine bleeding is substantial, consensus has only recently been reached on proper terminology and definitions, as proposed by the International Federation of Gynecology and Obstetrics (FIGO) Menstrual Disorders Working Group.[3-6]

The FIGO system classifies bleeding into nine categories based on underlying structural and nonstructural features.[5] Suggested definitions for normal menstrual bleeding in the midreproductive years included duration of menstrual flow between 4.5 and 8 days.[3, 6] Previous recommendations defined prolonged menstrual bleeding as 10 or more days[7] or menses exceeding 14 days.[8]

Chronic abnormal uterine bleeding, defined as abnormal volume, regularity or timing for a period of 6 months, associated with ovulatory dysfunction may occur with some frequency in the late reproductive years and the menopausal transition (MT) as a consequence of increased frequency of ovulation and altered patterns of progesterone production.[5, 6, 9-12] However, data on the bleeding experience of women at the end of reproductive life is limited, although the greater variability in menstrual bleeding patterns after age 40 is well recognised.[13-15]

The classic study by Hallberg et al. of menses volume demonstrated that 50-year-old women have a higher median menstrual blood flow than younger women.[16] Two European studies, of 1-year duration, have reported the cumulative incidence of 10+ days of bleeding to be 14.7%[17] and the period prevalence to be 36–49%,[18] while the cumulative incidence of heavy periods was 53%.[18] Previous perimenopausal studies of menses,[17-20] which included mostly white participants, were short in duration and unable to characterise bleeding during the whole MT.

This paper describes the distribution of menses duration and heaviness of flow (number of heavy days, number of spotting days) during the MT based on data from prospectively collected menstrual diaries in a large, multi-ethnic cohort of women. It also assesses the association of race/ethnicity and body mass index (BMI) with these menstrual bleeding characteristics.

Methods

This study used data from The Study of Women's health Across the Nation (SWAN), a prospective cohort study of midlife women living in the USA. It was restricted to three SWAN sites that conducted a menstrual calendar substudy: northern California, Los Angeles and southeastern Michigan. The design of SWAN's cohort study has been previously described.[21] Eligibility included age 42–52 years, self-designation as a member of the specified racial/ethnic groups, having an intact uterus and, in the 3 months prior to screening, having had at least one menstrual period and not currently using exogenous hormones affecting ovarian function, i.e. hormone therapy or hormonal contraceptives. Each site recruited white women and women from one minority group (African-Americans in southeastern Michigan, Japanese in Los Angeles, and Chinese in northern California); a total of 1498 women were enrolled into the cohort at the three study sites. Institutional Review Boards at each study site approved the study protocol and all participants provided written informed consent.

The SWAN cohort study began in 1996; baseline and annual follow-up visits consisted of interviewer-administered and self-administered questionnaires and height and weight measurements.

A self-administered daily menstrual calendar began in 1996 and continued through to 2006. Heaviness of flow was recorded using the following prespecified categories: spotting (bleeding not requiring the use of a sanitary product or not filling a regular-sized sanitary product); light to moderate bleeding (sanitary protection was required but only needed changing a few times a day or every 3–4 hours); or very heavy bleeding (needing to change sanitary product every 1–2 hours for more than 4 hours during the day).[22] Each month women answered questions about hormone use, gynaecological procedures, cigarette use and physical exercise. Women completed monthly calendars for 2 years after their final menstrual period (FMP).

Menstrual cycle length was calculated using bleeding definitions recommended by the World Health Organization[23, 24] as adapted for perimenopausal women by the ReSTAGE Collaboration,[25-27] which include the requirement of a bleed-free interval of at least 3 days. A menstrual cycle consisted of a bleeding episode and subsequent bleed-free interval of at least 3 days. A bleeding episode (menses) was defined as at least 1 day of bleeding or spotting. A bleeding episode could contain two consecutive bleed-free days between bleeding days. The duration of bleeding, total number of spotting days recorded, and total number of heavy days recorded were determined for each bleeding episode. We defined the cut-off for prolonged menses (10+ days), spotting (6+ days) and heavy bleeding (3+ days) by the upper 5th centile of their respective distributions. We also used two additional definitions of prolonged menses: the FIGO definition of menses exceeding 8 days[3, 6] and the definition based on data from the World Health Organization of menses exceeding 2 weeks (15+ days).[8]

Onset of the early MT and late MT were defined from the calendar data using the STRAW + 10 definitions.[9-12] Early MT was defined by the persistent difference of at least 7 days in the length of consecutive menstrual cycles. Persistence was defined as recurrence within ten cycles of the first variable length cycle. Menstrual cycles before the onset of the early MT were defined as premenopausal. Late MT was defined as the first occurrence of a menstrual cycle length of at least 60 days. The FMP was defined as the first day of a menses followed by at least 12 months of amenorrhoea. If neither the onset of the early MT nor the onset of the late MT was observed for a woman, then all of her menstrual cycles were defined as of unknown stage.

Menstrual cycles were coded as treated cycles for all cycles during which women reported hormone use. For months with missing hormone information, menstrual cycles were coded as not treated if the menstrual cycle occurred before the first report of hormone therapy use, if the woman reported no hormone therapy use in the calendar year of the missing information or if a woman never reported hormone therapy use.

Race/ethnicity was self-defined as African-American, Chinese, Japanese or white. Highest education (high school graduate or less than high school versus at least some college) and marital status (single, married/partnered, or separated, widowed, or divorced) were assessed at baseline. Economic strain was assessed at baseline with the question ‘how hard is it to pay for basics?’ and categorised as very hard, somewhat hard, or not hard. For each menstrual cycle, weight was linearly interpolated between the previous and subsequent annual visits and BMI was categorised as normal weight (≤24.9 kg/m2), overweight (25.0–29.9 kg/m2), or obese (≥30.0 kg/m2). The 47 women with BMI < 18.5 were included in the normal weight category. Results do not differ when they are excluded.

A woman was considered diabetic if she self-reported diabetes, was taking medication for diabetes, or had an annual visit (baseline, visit 1 and visits 3–7) fasting serum glucose level of ≥126 mg/dl (≥7 mmol/l). Uterine fibroids were based on self-reported diagnosis. Current cigarette use (yes/no) was defined by whether the woman reported having smoked at least one cigarette a day or a total of 30 cigarettes in the last month. Baseline smoking status was assessed by interview at the start of the study. Average hours of physical activity per week were calculated based on the average number of times per week and the average number of minutes per episode of moderate to vigorous physical activity reported.

Statistical analyses

Of the 1498 women enrolled at the three SWAN sites, 1320 (88.2%) were eligible, defined as having at least one menstrual cycle recorded in the menstrual calendar without hormone use, and were included in this analysis. Menstrual cycles during hormone use and during a 1 month wash-out period were excluded (n = 4266), as were menstrual cycles during pregnancy or where pregnancy was suspected (n = 30), and menstrual cycles with missing covariate information (n = 4266). Post-hysterectomy or post-FMP bleeding events were also excluded. Statistical analyses were performed using sas 9.2 (SAS Institute Inc., Cary, NC, USA). To determine the number of women who had menstrual events, cumulative percentages and associated 95% confidence intervals were calculated using Kaplan–Meier methods for survival probabilities to account for differential duration of observation and right-censoring. Time to bleeding event was calculated from the start of the MT stage until the first, second, or third occurrence of a bleeding event for that MT stage. For women without an event, time was calculated from start of the MT stage until the last menses recorded during that stage. For the all-stages analysis, time to bleeding event was calculated from the start of the study until the first, second or third occurrence of a bleeding event regardless of stage. For women without an event, time was calculated from study start until the last menses recorded in the menstrual calendar study.

Since the distribution of menses duration during the MT is right-skewed, the median is more informative about central tendency than the mean. We used quantile regression[28] to examine factors associated with the median duration of menses. Quantile regression models evaluate factors associated with the percentile (or quantile) of interest. Quantile regression coefficients are interpreted similarly to linear regression coefficients from a linear regression. For example, the coefficients of a categorical predictor represent the difference in menses duration in days for the 50th centile between one category and the reference group, adjusted for all other covariates in the model. Bootstrap sampling with 500 repetitions based on samples of women was conducted to construct 95% confidence intervals. The multivariable model included MT stage, ethnicity and BMI as well as covariates associated in the crude analysis (at < 0.20) with menses duration.

Generalised estimating equation methods with a first-order autoregressive working correlation structure, selected by using the quasi-likelihood information criterion, were used to model the association between menses of 10+ days duration, heavy bleeding of 3+ days, or 6+ days of spotting and covariates. Multivariable models included MT stage, ethnicity and BMI as well as covariates associated in the crude analysis (at < 0.20). In order to examine the association with study site, multivariable models were run among white women with study site as a variable instead of race/ethnicity.

Results

Women who were excluded were more likely to be from southeastern Michigan, African-American, report low educational attainment and economic strain, be overweight/obese, be a current smoker, and be less likely to be married than were included women (Table 1).

Table 1. Baseline demographics of women who were included and excluded from the Study of Women's Health Across the Nation (SWAN) Menstrual Calendar Substudy
 Included (n = 1320) n (%)Excluded (n = 178) n (%)P value
  1. Study was conducted in three SWAN sites: Los Angeles, Northern California, Southeastern Michigan.

  2. a

    Sixteen excluded have missing information.

  3. b

    One included and three excluded have missing information.

  4. c

    Seven excluded have missing information.

  5. d

    Twenty included and three excluded had missing information.

  6. e

    Sixteen included and four excluded had missing information.

  7. f

    Does not include oral contraceptive pills.

  8. g

    Diagnosed during the first ten annual visits.

Study site
Southeastern Michigan430 (32.6)113 (63.5) 
Northern California432 (32.7)27 (15.2)<0.001
Los Angeles458 (34.7)38 (21.3) 
Race/ethnicity
African-American236 (17.9)89 (50.0) 
Chinese232 (17.6)18 (10.1)<0.001
Japanese262 (19.8)19 (10.7) 
White590 (44.7)52 (29.2) 
Education a
Less than High School66 (5.0)14 (8.6) 
High School graduate219 (16.6)35 (21.6) 
Some College/Vocation455 (34.5)60 (37.1)0.01
College graduate296 (22.4)35 (21.6) 
Post College284 (21.5)18 (11.1) 
Marital status b
Single172 (13.0)22 (12.6) 
Married911 (69.1)103 (58.9) 
Separated44 (3.3)10 (5.7)0.02
Widowed24 (1.8)6 (3.4) 
Divorced168 (12.7)34 (19.4) 
How hard is it to pay for basics c
Very hard91 (6.9)15 (8.8) 
Somewhat hard349 (26.4)63 (36.8)0.01
Not hard880 (66.7)93 (54.4) 
BMI, kg/m 2 d
Normal (≤24.9)759 (58.4)67 (38.3) 
Overweight (25.0–29.9)251 (19.3)54 (30.9)<0.001
Obese(≥30.0)290 (22.3)54 (30.9) 
Baseline smoking status e
Never836 (64.1)104 (59.8) 
Past277 (21.2)31 (17.8)0.03
Current191 (14.7)39 (22.4) 
Ever taken hormones before study f 154 (11.7)17 (9.7)0.43
Ever diabetes during study g 211 (16.0)33 (18.5)0.39
Ever fibroids during study g 417 (31.6)47 (26.4)0.17
Age at screener years, mean (SD) 45.7 (2.7)46.3 (3.0)0.02

Of the 1320 eligible women, 963 (73.0%) had onset of the early MT identified and 815 (61.7%) had onset of late MT identified. In all, 431 (32.7%) women had their FMP identified in the menstrual calendars. Additionally, 19 (1.4%) had a hysterectomy, 233 (17.7%) began using hormones and had no further untreated menstrual cycles observed, and 637 (48.3%) withdrew from the study before their FMP was identified.

The 1320 women contributed from one to 166 cycles (median = 29) for a total of 51,606 menstrual cycles. Menses duration ranged from 1 to 132 days with a mean of 5.9 days and a median of 6 days (see Supporting information, Table S1). Menses exceeding 8 days was observed in 5287 (10.2%), menses, of 10+ days was observed in 3401 (6.6%), and menses of 15+ days in 838 (1.6%) menstrual cycles. The range of spotting days was 0–82 days, with a mean of 2.6 days and a median of 2 days. Spotting of 6+ days occurred in 3181 (6.2%) menstrual cycles, and in 1991 (58.5%) cycles with menses of 10+ days. The range of heavy days during menses was 0–19 days, with a mean of 0.7 days and a median of 0 days. Heavy bleeding of 3+ days occurred in 2746 (5.3%) of menstrual cycles but in 561 (16.5%) cycles with menses of 10+ days. Heavy bleeding of 3+ days occurred in 218 (6.9%) of the menstrual cycles with 6+ spotting days, with almost all (n = 215) occurring in menses of 10+ days.

Cumulative percentage of prolonged menses, prolonged spotting and heavy bleeding

In the 10.5 years of observation, the cumulative percentages of women with at least one, two or three occurrences of menses of 10+ days were 91.0%, 87.5%, and 77.7% (see Supporting information, Table S2). During the early MT, with a median duration of 3.0 years (interquartile range [IQR] 1.3–5.3 years), the cumulative percentage of women with at least three occurrences of menses of 10+ days was 51.6%. During the late MT, with a median duration of 2.6 years (IQR 1.8–3.9), the cumulative percentage of women with at least three episodes of menses of 10+ days was 54.6%. The median time between 10+-day bleeding episodes was 125 days (IQR 54–301 days). Among women experiencing such episodes, 27.5% experienced three episodes within a 6-month period. Median time between 8+-day bleeding episodes was 88 days (IQR 35–197 days), with 39.0% of women with such episodes experiencing three 8+ day bleeding episodes within a 6-month period.

During the study, 66.8% of women had at least three occurrences of 6+ days of spotting, with 51.2% of women recording at least three such occurrences during the early MT and 75.2% during the late MT. The median time between 6+-day spotting episodes was 148 days (IQR 60–266 days). Among women experiencing three or more episodes of spotting, 23.6% recorded three episodes within a 6-month time window. The cumulative percentage of women with at least three occurrences of menses with three or more days of heavy bleeding was 34.5%, while 30.7% and 34.7% had at least three such occurrences during the early MT and late MT, respectively. The median time between heavy bleeding episodes was 99 days (IQR 31–231 days), with 39.8% of women who recorded three or more episodes of heavy bleeding experiencing them within a 6-month time window.

Multivariable quantile and logistic regression

Based on the multivariable quantile regression, women who were diagnosed with uterine fibroids had a 1.00 (95% confidence interval 0.17–1.83) day longer median menses duration than women who were not diagnosed with uterine fibroids. Median menses duration was not associated with MT stage, ethnicity, education, economic strain, diabetes, BMI, physical activity, current smoking or reported hormone use during the study.

In multivariable logistic regression models, being in the MT was associated with menses duration of 10+ days, 6+ days of spotting and 3+ heavy bleeding days (Table 2.) Compared with premenopausal cycles, women were more likely to report menses of 10+ days, 6+ days of spotting, or 3+ days of heavy bleeding in both the early and late MT, with the odds slightly higher in the late than early MT. Women with uterine fibroids were more likely to report each of these three types of bleeding episodes than women not diagnosed with uterine fibroids; however, the 95% confidence interval for heavy menstrual bleeding included the null (1.00). Women who reported using hormones during the study were also more likely to report each of the three bleeding types than women who did not.

Table 2. Adjusted logistic regression models for potentially abnormal uterine bleeding,[3-6] The Study of Women's Health Across the Nation Menstrual Calendar Substudy.
 Menses 10 +  daysaSpotting 6 +  daysbHeavy bleeding 3 +  daysc
Adjusted OR (95% CI)Adjusted OR (95% CI)Adjusted OR (95% CI)
  1. a

    Multivariable model contained menopausal transition stage, race/ethnicity, uterine fibroids, BMI, ever hormone use, and adjusted for repeated measures.

  2. b

    Multivariable model contained menopausal transition stage, race/ethnicity, hard to pay for basics, uterine fibroids, BMI, ever hormone use, and adjusted for repeated measures.

  3. c

    Mulitvariable model contains menopausal transition stage, race/ethnicity, hard to pay for basics, uterine fibroids, BMI, current smoker, ever hormone use, and adjusted for repeated measures.

Menopausal transition stage
PretransitionReferenceReferenceReference
Early transition3.15 (2.47–4.02)2.66 (1.98–3.57)1.38 (1.10–1.73)
Late transition4.30 (3.26–5.68)3.95 (2.83–5.52)1.75 (1.31–2.34)
Unknown1.46 (0.82–2.59)0.94 (0.50–1.77)1.41 (0.90–2.22)
Race/ethnicity
African-American0.64 (0.46–0.89)0.52 (0.37–0.73)1.26 (0.87–1.84)
Chinese0.97 (0.72–1.32)1.23 (0.90–1.69)1.04 (0.71–1.51)
Japanese0.80 (0.60–1.07)1.01 (0.76–1.34)0.59 (0.39–0.88)
WhiteReferenceReferenceReference
Very or somewhat hard to pay for basics n/a1.24 (0.96–1.60)1.57 (1.17–2.11)
Uterine fibroids 1.52 (1.22–1.88)1.28 (1.04–1.58)1.34 (0.97–1.85)
BMI, kg/m 2
Normal weight (≤24.9)ReferenceReferenceReference
Overweight (25.0–29.9)0.94 (0.73–1.20)0.88 (0.70–1.11)1.40 (1.02–1.91)
Obese (≥30)0.91 (0.69–1.19)0.80 (0.58–1.10)2.24 (1.63–3.08)
Current smoker n/an/a1.63 (1.16–2.28)
Ever hormone use during study 1.59 (1.27–2.00)1.32 (1.06–1.64)1.62 (1.21–2.16)

African-American women were less likely to report menses of 10+ days or 6+ days of spotting compared with white women. Japanese women were less likely to report 3+ days of heavy bleeding than white women. Women for whom it was very or somewhat hard to pay for basics were more likely to record 3+ days of heavy bleeding than other women. Overweight women and obese women were more likely to experience 3+ days of heavy bleeding than normal weight women. Current smokers were more likely to record 3+ days of heavy bleeding than non-smokers.

Education, economic strain, diabetes, BMI, physical activity or current smoking were not associated with menses of 10+ days or 6+ days of spotting. In the unadjusted analyses, both diabetes and thyroid conditions were associated with 3+ days of heavy bleeding, but the relationships did not persist after adjustment. Education and physical activity were not associated with 3+ days of heavy bleeding. Among white women, study site was not associated with menses of 10+ days, 6+ days of spotting, or 3+ days of heavy bleeding.

Discussion

Main findings

This study is one of the first to examine the frequency of prolonged menses and heavy bleeding during the MT and to evaluate race/ethnicity, BMI and other factors in relation to these menstrual events. Three out of four of these mid-life women experienced at least three occurrences of menses of 10+ days. Slightly more than one-quarter of these women experienced three such bleeding episodes within a 6-month time window. Menses with spotting of 6+ days occurred in over half the episodes of menses of 10+ days. Menses with heavy bleeding of 3+ days were less common, yet one in three women recorded three or more occurrences during their MT, approximately 40% of whom experienced three heavy bleeding episodes with a 6-month time window.

Strengths and limitations

Study limitations included left-censoring, which may have biased our results. The mean age of women eligible for this study was 45.7 years and, for some women, the onset of the MT occurred before enrolment into the study. Women who did not have a menstrual cycle in the last 3 months were ineligible to enrol into the SWAN cohort study, therefore women who were near the end of the MT, who experienced surgical amenorrhoea or who had already experienced their FMP were excluded. Nonetheless, these data provide one of the first evaluations of the frequency of long and heavy menses during the MT. In the population from which our cohort study was enrolled, African-American women were more likely to have had a hysterectomy.[29] Information on uterine fibroids was obtained by self-report and may be underreported.

Given that these data are drawn from a population-based study that did not include medical record linkage, a further limitation of this study is that information on other conditions that may affect uterine bleeding such as polyps or adenomyosis was not available. Hence we could not differentiate pathological from nonpathological bleeding. As we cannot correlate these bleeding patterns with the likelihood of specific pathology or subsequent cancer rates, these data are informative about expected frequency of prolonged and heavy bleeding during the menopausal transition, but definitive inferences about threshold indications for clinical evaluation are not possible.

Interpretation

This study is the only multi-ethnic study of menstrual bleeding patterns during the menopausal transition and it is based on a large sample of prospectively collected menstrual calendar data recorded for up to 10 years. We found that menses of 10+ days were more likely to occur during the early and late MT than in late reproductive life. In contrast, the median duration of menses was not associated with reproductive stage, suggesting that it is not the average menses length but rather the likelihood of prolonged bleeding which changes during the MT. Our results are consistent with previous studies and provide more comprehensive quantitative data. The Melbourne Women's Midlife Health Project did not find a change in mean heaviness of flow from the first to second year of their study.[20] The Massachusetts Women's Health Study found a higher percentage of menses exceeding 8 days in perimenopausal women than premenopausal women.[19]

The FIGO Menstrual Disorders Working Group suggested the normal limits for the duration of menstrual flow in the mid-reproductive years to be between 4.5 and 8 days[3, 6] with prolonged menstrual bleeding defined as bleeding exceeding 8 days. Previous recommendations defined prolonged menstrual bleeding as 10 or more days[7] or menses exceeding 14 days.[8] We found that prolonged menses of 10+ days to be common occurrences during the menopausal transition.

Only one other calendar study, among postmenarcheal girls, has examined ethnic differences in menstrual bleeding in the USA. Mean duration of bleeding was a half-day less among African-American girls compared with white girls, but African-American girls were more likely to report heavy bleeding.[30] In our SWAN Menstrual Calendar Study, menses of 10+ days were less likely in African-American women compared with white women but African-American women were more likely to experience three or more days of heavy bleeding, although the latter association was attenuated in the multivariable analysis.

The BMI was associated with heaviness of flow but not menses duration in this analysis. A similar result was seen in a Danish study, which observed that obese women were more likely to report menstrual flooding but not prolonged menstrual bleeding.[18] In the SWAN Daily Hormone Study, obesity was also associated with three or more days of heavy bleeding.[31] The Michigan Bone Health and Metabolism Study did not find an association between menses duration and BMI in women aged 25–50 years.[32] However, several studies of adolescent and young adult women report an association between low BMI and longer menses[30, 33, 34] or high BMI and short menses.[30, 35-37] The associations we found with obesity may be the result of the differences seen in hormone profiles among obese women during the MT. In the SWAN Daily Hormone Study, obese women were shown to have longer follicular phases, so longer duration of estrogen stimulation, but shorter luteal phases. Obese women in the SWAN Daily Hormone Study had lower luteinising hormone, and progesterone metabolites as well as lower follicle-stimulating hormone levels,[38] as was also found in the longitudinal SWAN cohort.[39]

Ever being on hormones during the study was positively associated with all three bleeding events, consistent with the fact that hormones are used to treat alterations in uterine bleeding.[40] Uterine fibroids were also positively associated with all three bleeding events. In our study, women self-reported a physician diagnosis of uterine fibroids. Typically, women are only diagnosed with uterine fibroids when they present with a specific complaint and are evaluated. However, it is estimated that 50% of women with uterine fibroids are asymptomatic.[41] Hence, this association may only reflect identified uterine fibroids that cause more symptoms.

Conclusion

We found that the large majority of women in SWAN, over 50% of whom were not white, reported menses duration of 10+ days, spotting of 6+ days and/or 3+ days of heavy bleeding during the MT. The likelihood of experiencing these menstrual bleeding events varied by race/ethnicity, BMI and reported uterine fibroids. These data confirm that two types of bleeding, longer menses with more days of spotting and heavier menses, occur in most women during the MT. These data provide clinicians and women with important normative data regarding the expected frequency of these bleeding changes during the menopausal transition, data that may facilitate decision-making in clinical practice. Such information, particularly when coupled with the emerging information about duration of the stages of the menopausal transition,[25, 27, 42] will be of great value to women in this life-stage who wish to be active participants in their own health care. Such normative data can greatly alleviate concerns about midlife changes in ones bleeding patterns and facilitate coping with what, in most cases, will probably be a time-limited experience. Further research on the correlation between such bleeding and documented pathology is warranted to better support recommendations for watchful waiting versus clinical intervention.[43]

Additional studies to better define ‘normal’ bleeding patterns in women aged 40–55 years should be encouraged. As one of the only papers on menstrual bleeding patterns during the menopausal transition, this paper provides important information for women trying to assess changes in their menstrual experience during the menopausal transition.

Disclosure of interests

The authors report no conflict of interest. The content of this paper is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health (NIH), National Institute on Aging (NIA), the National Institute of Nursing Research (NINR) and the NIH Office of Research on Women's Health (ORWH).

Contribution to authorship

Each of the authors contributed to the concept of the project, made important contributions to the design of the project, participated in drafting and reviewing the article and provided expertise on menstrual cycle characteristics and the menopausal transition. Additionally, two of the authors (GG and JR) provided additional clinical expertise, three of the authors provided technical expertise and supervised data collection at the study sites (SH, GG, EG) and three of the authors provided expertise on statistics (SC, ME, LL). PP conducted data programming, all statistical analyses, and took the lead on drafting the manuscript.

Details of ethics approval

Northern California: Institutional review Board Administration, University of California at Davis 28 December 2012 (original study approval date 6 February 1996), IRB no. 260339-8 (Original protocol number 96-466R) and Institutional Review Board for the Protection of Human Subjects, Kaiser Foundation Research Institute 28 May 2013 (Original approval date August 18, 1995) CN-94BSter-01-H.

Los Angeles: Office of Human Research Protection, University of California at Los Angeles, 23 May 2013 (Original study approval date 4 May 1995), IRB#11-002274.

Southeastern Michigan: The Health Sciences and Behavioral Sciences Institutional Review Boards (IRB-HSBS), University of Michigan 20 December 2012 (Original study approval date 1 July 2005) H03-00000531.

Funding

The Study of Women's Health Across the Nation (SWAN) has grant support from the National Institutes of Health (NIH), DHHS, through the National Institute on Aging (NIA), the National Institute of Nursing Research (NINR) and the NIH Office of Research on Women's Health (ORWH) (Grants U01NR004061; U01AG012505, U01AG012535, U01AG012531, U01AG012539, U01AG012546, U01AG012553, U01AG012554, U01AG012495). The first author was also supported by the ReSTAGE collaboration, which has grant support from the NIA (Grant R01AG021543) and the University of Michigan Rackham Graduate School.

Acknowledgements

We thank the study staff at each site and all the women who participated in SWAN. Clinical Centers: University of Michigan, Ann Arbor – Siobán Harlow, PI 2011–present, MaryFran Sowers, PI 1994-2011; Massachusetts General Hospital, Boston, MA – Joel Finkelstein, PI 1999–present; Robert Neer, PI 1994–99; Rush University, Rush University Medical Center, Chicago, IL – Howard Kravitz, PI 2009–present; Lynda Powell, PI 1994–2009; University of California, Davis/Kaiser – Ellen Gold, PI; University of California, Los Angeles – Gail Greendale, PI; Albert Einstein College of Medicine, Bronx, NY – Carol Derby, PI 2011–present, Rachel Wildman, PI 2010–11; Nanette Santoro, PI 2004–10; University of Medicine and Dentistry – New Jersey Medical School, Newark – Gerson Weiss, PI 1994–2004; and the University of Pittsburgh, Pittsburgh, PA – Karen Matthews, PI. NIH Program Office: National Institute on Aging, Bethesda, MD – Winifred Rossi 2012–present; Sherry Sherman 1994–2012; Marcia Ory 1994–2001; National Institute of Nursing Research, Bethesda, MD – Program Officers. Central Laboratory: University of Michigan, Ann Arbor – Daniel McConnell (Central Ligand Assay Satellite Services). Coordinating Center: University of Pittsburgh, Pittsburgh, PA – Maria Mori Brooks PI 2012–present; Kim Sutton-Tyrrell, PI 2001–2012; New England Research Institutes, Watertown, MA – Sonja McKinlay, PI 1995–2001. Steering Committee: Susan Johnson, Current Chair; Chris Gallagher, Former Chair.

Ancillary