Abstract
A number of studies have suggested that ethnic background influences a woman's perception of her symptoms. The Study of Women's Health Across the Nation (SWAN) is a multiethnic, longitudinal, cohort study of US women that includes non-Hispanic Caucasian, African–American, Chinese, Japanese and Hispanic women. The initial strategy for this seven-site study involved community-based recruitment of non-Hispanic Caucasians at each site, plus one minority ethnic group. Since ethnicity varies with many other factors, measures of education, acculturation, social status, psychological wellbeing and financial strain were all taken into account in interpreting symptom onset, frequency and severity of the common menopausal symptoms. Biological and physical measures were also assessed and related to symptoms. Most symptoms varied by ethnicity. Vasomotor symptoms were more prevalent in African–American and Hispanic women and were also more common in women with greater BMI, challenging the widely held belief that obesity is protective against vasomotor symptoms. Vaginal dryness was present in 30–40% of SWAN participants at baseline, and was most prevalent in Hispanic women. Among Hispanic women, symptoms varied by country of origin. Acculturation appears to play a complex role in menopausal symptomatology. We conclude that ethnicity should be taken into account when interpreting menopausal symptom presentation in women.
Menopausal symptoms are reported by a majority of women [1]. Their prevalence and severity vary considerably based upon a number of different factors, including race/ethnicity [2] and other demographic and lifestyle factors [3,4]. The variability of these other influences on menopausal symptoms gives rise to many questions:
How do symptoms differ by ethnicity?
What characteristics are related to ethnicity itself and how can we examine ethnicity independently of its common covariates, such as socioeconomic status, acculturation and language?
To what extent does genetic background influence susceptibility to symptoms?
How does perceived discrimination contribute to symptoms?
Do acculturation, discrimination and ‘othering’ contribute to symptoms?
The Study of Women's Health Across the Nation (SWAN) is a longitudinal, multiethnic, community-based, observational study of women traversing the menopause. It was the first NIH observational study of its kind that sought to include women from a variety of ethnic backgrounds in order to encompass the menopausal experience of as many women as possible and, thus, to fully characterize the menopause of American women. This article will review the strategies utilized by SWAN in order to accomplish timely recruitment of minority and nonminority women, highlight some of the challenges that were encountered and describe our initial findings and their interpretations.
Strategy of recruitment: SWAN
The Study of Women's Health Across the Nation recruited its community-based sample by beginning with a population-based, cross-sectional screening survey administered to 16,065 women upon which eligibility for the longitudinal study was based. A total of 3302 women completed the baseline assessment for the longitudinal study. Annual assessments in SWAN are fairly comprehensive and include physical measurements, blood draws, urine sampling and bone density measurements at some but not all sites, and a detailed battery of survey measures, details of which have been previously published [5]. Since the intent was to track the menopausal process as fully as possible, women were not eligible to be in the study if they had less than one menstrual period within the preceding 3 months. This is an accepted epidemiological definition of the ‘early transition’ [6]. SWAN is being conducted at the seven sites, each of which recruits non-Hispanic Caucasian women and women from one other race–ethnic background: (African–American, Hispanic, Japanese or Chinese) Boston (MA, USA), Chicago (IL, USA), Detroit (MI, USA), Los Angeles (CA, USA), Newark (NJ, USA), Pittsburgh (PA, USA) and Oakland (CA, USA). At enrolment, women were required to have an intact uterus and at least one ovary (to allow bleeding patterns to be tracked), were not pregnant or breastfeeding and were between 42 and 52 years of age. All participants were still menstruating, and women using oral contraceptives or HRT within the previous 3 months were excluded. A total of 420 women (278 Hispanic, and 142 non-Hispanic Caucasian) were enrolled from Hudson County, NJ, USA. The 278 Hispanic women were from five subethnic groups; Central American, South American, Cuban, Puerto Rican and Dominican, all drawn from the same geographic area, as were non-Hispanic Caucasians.
The initial sampling strategy of SWAN was to include a non-Hispanic Caucasian group at each site, assuming that these non-Hispanic Caucasian women would be sufficiently similar to each other in order to allow comparisons both across and within sites. Since sites differ geographically as well as by the overall socioeconomic status of their particular sample, this strategy has assisted in disentangling some of the covariates mentioned previously, and allows us to begin to answer questions about different symptoms across ethnic groupings.
Basic menopausal physiology & ethnic variation
Ethnic differences were observed in the onset of premature and early menopause, defined as the cessation of menstruation before the ages of 40 and 45 years, respectively. The cross-sectional survey used by SWAN to determine eligibility for the longitudinal cohort study included questions about the untimely cessation of menses. Highly statistically significant differences were observed in the prevalence of premature and early menopause across the ethnic groups in SWAN. Although previous reports estimated the prevalence of premature menopause at 1% [7], there was variation around this number in SWAN. In Hispanic women, 1.4% reported menopause prior to the age of 40 years; for African–Americans the prevalence was 1.4%, for Caucasian (non-Hispanic) 1%, for Chinese 0.5% and for Japanese 0.1% [8]. Early menopause displayed even more striking differences, with 4.1% of Hispanic, 3.7% of African–American, 2.9% non-Hispanic Caucasian, 2.2% Chinese and 0.8% of Japanese women reporting cessation of menses prior to 45 years of age[8].
Hysterectomy confounds the assessment of menopause because it interrupts bleeding patterns and, thus, the cessation of ovarian function can no longer be observed. Ethnic differences in the indications for hysterectomy have long been reported in the literature [9]. The SWAN cross-sectional screening survey (n = 15,160) sampled 49.9% Caucasian, 28.1% African–Americans, 12.3% Hispanic and 9.8% Asian–American women. Ethnicity was found to be associated with past hysterectomy [10]. The odds ratios (ORs) were: Caucasian = 1.0, African–American = 1.66 (95% CI: 1.46–1.88), Hispanic = 1.64 (95% CI: 1.29–2.07) and Asian–American = 0.44 (95% CI: 0.34–0.56). These results indicate that for benign conditions, such as uterine fibroids, abnormal bleeding, endometriosis and pelvic organ prolapse, the highest rates of hysterectomy occurred in African–American and Hispanic subgroups. These ORs were adjusted for age, education, a diagnosis of fibroids, BMI, marital status, smoking, geographic site and country of education, all of which are believed to affect the likelihood of hysterectomy. The authors concluded that disparity in the form of overuse in disadvantaged minority groups may exist [10]. Previous studies that have attempted to elucidate the reasons for higher hysterectomy rates in African–American women have found earlier onset of large, symptomatic fibroids in this population group [11], which may account for their tendency to have hysterectomies at an earlier age compared with other ethnic groups.
It is important to recognize that there are ethnic differences in the prevalence of premature menopause and hysterectomy because both of these conditions are associated with worse menopausal symptoms [1].
Other possible modifiers may differ by ethnicity and be owing to a different genetic background, such as sex steroid-synthesizing enzymes and metabolism-regulating genes [12]. Although not observed in SWAN, in the Seattle Midlife Women's Health Study, the CYP19 11r polymorphism for the aromatase gene was associated with worse vasomotor symptoms (VMS) [13]. Genetic variations are tempting explanations for symptoms that might be based upon hormone synthesis pathways.
Ethnicity as a covariate influencing symptom vulnerability
Diet, methods of coping and sociopolitical factors may play a role in symptom sensitivity as well as in symptom reporting. It is important to keep in mind that there may be ethnic differences in how some of the SWAN survey measures are processed by women of different backgrounds and literacy.
Vasomotor symptoms
The hot flash is viewed as the cardinal symptom of menopause. Hot flashes affect up to 75% of women at some point in the menopausal transition [1]. However, a minority of women report being severely affected. Their prevalence peaks less than 2 years to 1 year before or after the final menstrual period and returns to near premenopausal levels 8 years later, based on a recent meta-analysis [14]. In the SWAN baseline, 46.5% African–American, 49.4% Hispanic, 36.6% Chinese, 34.3% Japanese and 28.9% Caucasian women reported hot flashes and/or night sweats [2]. Hispanic women report more negative associations (embarrassment) with VMS than do other ethnic groups in SWAN [15]. Hispanic women also reported more baseline anxiety and depressive symptoms related to VMS. However, Chinese women had the fewest VMS overall.
These early findings from SWAN challenged the current thinking about the genesis of VMS and have led to new hypotheses. It has long been believed that increased adiposity is associated with decreased menopausal VMS because of the known conversion of androgens to estrogens in body fat [16]. However, increased adipose tissue may be associated with a greater likelihood of VMS because of its insulating properties. This thermoregulatory model of the pathophysiology of VMS is supported by SWAN. In a sample of 1776 women, a higher percentage of body fat was associated with increased odds of reporting VMS (per standard deviation increase in percent body fat; OR: 1.27; 95% CI: 1.14, 1.42) in both age and site-adjusted models [17]. It is important to recognize the role of adiposity in predicting VMS, because BMI varies considerably in SWAN with respect to ethnicity [18]. African–American and Hispanic women have the highest BMIs in SWAN. Combined with their higher risk for hysterectomy, African–American women are at greatest risk for VMS than the ORs suggest, because the various studies adjust for these additive factors.
Vaginal symptoms
Approximately 30–40% of women experience dyspareunia or vaginal dryness apart from during intercourse, and symptoms persist long after menopause in at least 26% of women [19,20].
In SWAN, Hispanic women complain of more vaginal symptoms than the other ethnic groups [15]. These findings are similar to those from the Women's Health Study, in which postmenopausal Hispanic women reported the highest rates of vaginal irritation [19].
Menopause & mood
Hispanic and African–American women are more likely to report depressive symptoms and Chinese and Japanese women are the least likely to report depressive symptoms in SWAN [21]. Irritability and nervousness, associated with the early transition, appears to be linked to fluctuating hormones. The late menopausal transition is associated with increased depressive symptoms (OR: 1.3–1.71) in SWAN [22] and other studies as compared with premenopause [23,24]. Prior history of depression is predictive of increased body pain, poor social functioning, negative mood (not clinical depression) and treatment for back pain [21,22,25].
Sleep disturbances
Sleep complaints begin to appear in women above 35 years of age. A total of 23.6% of women and 14.4% of men report sleep difficulties between 45–49 years of age and 39.7% women and 15.3% of men report not sleeping well by their early 50s [26]. Whether the large midlife increase in sleep problems in women is due to aging, stress, psychosocial and physical problems or the effects of hormonal changes associated with menopause is still unclear. Among ethnic groups, poor sleep increases with menopausal stage [26]. In SWAN, longitudinal data revealed that African–American and Hispanic women reported the highest percentage of difficulty in sleeping, manifested as staying asleep, and early morning awakenings. Progression through the menopausal transition as indicated by three menopausal characteristics – symptoms, bleeding-defined stages and endogenous hormone levels – is associated with self-reported sleep disturbances [26,27]. These findings are the first to describe ethnic-specific changes in perceived sleep efficiency in middle-aged women.
Relationship of common menopausal symptoms to ethnicity in the Study of Women's Health Across the Nation.
Data are shown as the percentage of women experiencing the symptom or condition, or as an OR, with Caucasian women = 1.0 and appropriate 95% CI.
Hispanic women reported more embarrassment or discomfort associated with hot flashes.
Other ethnic groups not compared.
CI: Confidence interval; OR: Odds ratio.
Discrimination, ‘othering’ & acculturation
As new immigrants become accustomed to the dominant culture of their newly adopted county, there is an expectation of assimilation and acceptance. In an ideal scenario, the USA is viewed as a benign ‘melting pot’ into which immigrants are welcomed and prosper. According to this logic, the more assimilation that takes place, the better the lot of the immigrant. In this model, assimilation is viewed as entirely beneficial. This phenomenon has been observed among the Hispanic women in SWAN [28] inasmuch as increased self-reported language acculturation is related to improved overall quality of life (QoL).
However, increased acculturation has the potential to erode the QoL for immigrants. As individuals assimilate, they have more interaction with the dominant culture and more chances for discrimination [29]. Acculturation may erode immigrant family values. Increased acculturation has been hypothesized to lead to worsening health outcomes in Hispanic populations, another so-called Hispanic paradox attributed to acculturation stress [30]. In this scenario, increased time since immigration results in a loss of healthy behaviors, resulting in an adverse influence on a variety of health outcomes. For example, Central American immigrants who are less acculturated are more likely to be married, and less likely to engage in unhealthy habits, such as cigarette smoking [31]. Adverse effects of migration include a loss of family and social structure, and exposure to street drugs and unmarried pregnancy. These factors result in a deterioration of the nativity advantage in perinatal outcomes [32].
Thus, acculturation may not bring about health benefits if there is a significant increase in exposures that are unhealthy or that lead to the perception of ‘othering’ or frank discrimination. There is evidence that relatively highly acculturated, Hispanic women from the Bronx, NY, USA, who are exposed to high rates of poverty, drug use and HIV infection are especially vulnerable to menopausal symptoms of all kinds [3]. The data suggest that symptom vulnerability cannot be reliably predicted by acculturation alone. Specific studies examining these issues in the context of menopause are sparse.
SWAN: the trouble with generalizing
SWAN Hispanic women & acculturation
Despite the similarity of their geographic sites and low overall socioeconomic status, there are large differences between the Hispanic women in SWAN with respect to acculturation and a number of other factors that are typically related to acculturation, such as education, financial strain (difficulty paying for the basic necessities of life) and health habits (e.g., cigarette smoking). Puerto Rican women are the most acculturated, based upon their response to a language variable ascertained from four questions regarding the language women usually think, read and speak in, talk with their friends in and listen to the radio or watch television in. The acculturation scale allows researchers to quickly identify Hispanics who are low and high in mostly language acculturation. The scale has been referenced with a variety of Hispanic subgroups, including Mexican–Americans, Cuban–Americans, Puerto Ricans, Dominicans and Central and South Americans [33].
Despite their greater acculturation, Puerto Rican women are not better educated than other Hispanic women in SWAN. They also report relatively more menopausal symptoms and more anxiety symptoms among the Hispanic women of SWAN. The Cuban women have the highest socioeconomic status and fewer symptoms than the other Hispanic subgroups, despite being less acculturated. Central American women tend to have the worst overall menopausal symptomatology (VMS and vaginal dryness) [15] and less acculturation.
Thus, for the Central American women, who were more likely to be the more recent immigrants, less acculturation appeared to help explain some of their increased symptoms. However, the Puerto Rican women, who were most acculturated, appear to be suffering from the phenomenon of acculturation stress and did not experience better health as a result of being more assimilated into the US mainstream culture.
Quality of life & race/ethnicity
The question of whether menopausal status is associated with global QoL among multiethnic (Caucasian, African–American, Chinese, Japanese and Hispanic) women aged 40–55 years and whether this association varies by race/ethnicity in SWAN analyses suggests that in this population-based sample of women, early perimenopausal women reported lower QoL compared with premenopausal in unadjusted analyses, but menopausal status was no longer associated with QoL when analyses were adjusted for other variables. This difference was largely explained by perceived stress, which was highest among the early perimenopausal women. Whether this is directly related to menopausal status or explained by a third variable (e.g., life changes) needs further exploration. In multivariate models, being married and having low levels of perceived stress were associated with better QoL across all racial/ethnic groups [28].
Attitudes towards menopause & aging
Attitudes vary across ethnic groups and menopausal status (i.e., premenopausal through to postmenopausal). African–American women were significantly more positive in their attitude towards menopause. The least positive groups were the less acculturated Chinese–American and Japanese–American women. The general finding that attitudes toward menopause ranged from neutral to positive is consistent with findings of previous research. The finding that more African–American women and Caucasian women agreed with the perception of menopause signaling freedom and independence may reflect a higher value of these qualities than is the case in the Asian and Hispanic cultures, which are generally thought of as more communal. The SWAN data have the advantage of comparing responses of presumably more and less acculturated women to the same questions and do not support the conclusion that exposure to Western ideas is the source of negative attitudes towards menopause [34].
Conclusion & future perspective
In summary, examination of epidemiological data on menopausal symptoms indicates clear-cut differences in symptom reporting by ethnicity. Much caution needs to be exercised in interpreting these data since the reasons are unlikely to be innate ethnic-related issues. Genetic hypotheses are under exploration but to date the data have not been explicatory of these relationships. The role of discrimination and ‘othering’ in increasing vulnerability to symptoms needs to be taken into account; clearly further research is needed. As immigrant populations on the whole are targeted to exceed the ‘majority’ of Caucasians over the next several decades, much more attention needs to be focused on the differences in reporting of menopausal symptoms, their treatment and their long-term consequences in women of varying ethnic backgrounds.
Executive summary
The Study of Women's Health Across the Nation (SWAN) is a multiethnic, longitudinal, cohort study of mid-life women – the first NIH observational study of its kind that sought to include women from a variety of ethnic backgrounds in order to encompass the menopausal experience of as many women as possible and, thus, to fully characterize the menopause of American women.
Hispanic and African–American women were most likely to have premature or early menopause.
For benign conditions, such as uterine fibroids, abnormal bleeding, endometriosis and pelvic organ prolapse, the highest rates of hysterectomy were observed in African–American and Hispanic subgroups.
Interactions of adiposity, higher rates of hysterectomy and earlier menopause favor increased vasomotor symptoms in African-American women and may help to explain their almost twofold increased risk of this common menopausal symptom.
Vaginal dryness is a prevalent menopausal symptom and was reported by up to 40% of women in SWAN, most commonly by Hispanics.
Depressive symptoms and major depression are increased over the menopausal transition. Hispanic and African–American women appear to be the most vulnerable.
Sleep disturbances increase during the perimenopause and are worst in African–American and Hispanic women.
Assimilation into the mainstream culture is often associated with improved symptoms, but in the SWAN study, midlife Hispanic women who are more acculturated may not enjoy a reduction in symptoms.
Symptoms differ within and among Hispanic women based upon their country of origin, indicating an ability to resolve the relationships of culture to symptoms at a very fine level.
Appreciation of the role of ethnicity in the expectations, onset and duration of menopausal symptoms is expanding.
Cultural, biological and genetic factors that influence vulnerability to symptoms can be identified by detailed study.
The role of othering or discrimination in predicting symptoms will be best resolved by using a correlative biological marker for acculturation stress.
The definition of Hispanic ethnicity as a single subgroup is a concept that warrants reconsideration, since differences among women who self-identify as Hispanic have been shown to exist based upon their country of origin.
Footnotes
Acknowledgements
The Study of Women's Health Across the Nation (SWAN) has grant support from the National Institutes of Health (NIH), Department of Health and Human Services, 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 NR004061; AG012505, AG012535, AG012531, AG012539, AG012546, AG012553, AG012554 and AG012495. The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the NIA, NINR, ORWH or the NIH.
The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.