Elsevier

Social Science & Medicine

Volume 52, Issue 3, February 2001, Pages 345-356
Social Science & Medicine

Is there a menopausal syndrome? Menopausal status and symptoms across racial/ethnic groups

https://doi.org/10.1016/S0277-9536(00)00147-7Get rights and content

Abstract

In recent years, research on menopausal symptomalogy has focused on identifying symptom groupings experienced by women as they progress from premenopausal to postmenopausal status. However, most of these studies have been conducted among Caucasian women from western cultures. This leaves open the question of whether the findings from these studies can be extended to women of other racial/ethnic groups or cultures. Furthermore, many of the previous studies have been conducted on relatively small samples. This paper addresses the diversity of the menopause experience by comparing symptom reporting in a large cross-sectional survey of women aged 40–55 years among racial/ethnic groups of women in the United States (Caucasian, African-American, Chinese, Japanese, and Hispanic). Evaluation of the extent to which symptoms group together and consistently relate to menopausal status across these five samples provides evidence for or against a universal menopausal syndrome. The specific research questions addressed in this paper are: (1) How does the factor structure of symptoms among mid-aged women compare across racial/ethnic groups? (2) Is symptom reporting related to race/ethnicity or menopausal status? and (3) Does the relation between menopausal status and symptoms vary across racial/ethnic groups?

Analyses are based on 14,906 women who participated in the multi-ethnic, multi-race, multi-site study of mid-aged women called the Study of Women’s Health Across the Nation (SWAN). Study participants completed a 15-min telephone or in-person interview that contained questions on a variety of health-related topics. Items of interest for these analyses include symptoms, menstrual history (to assess menopausal status), health status, and sociodemographics.

Factor analysis results showed that across all five racial/ethnic groups, two consistent factors emerged; one consisting of clearly menopausal symptomshot flashes and night sweats — and the other consisting of psychological and psychosomatic symptoms. Results of regression analyses showed racial/ethnic differences in symptom reporting, as well as differences by menopausal status. Controlling for age, education, health, and economic strain, Caucasian women reported significantly more psychosomatic symptoms than other racial/ethnic groups. African-American women reported significantly more vasomotor symptoms. Perimenopausal women, hormone users, and women who had a surgical menopause reported significantly more vasomotor symptoms. All of these groups, plus postmenopausal women, reported significantly more vasomotor symptoms than premenopausal women. The pattern of results argues against a universal menopausal syndrome consisting of a variety of vasomotor and psychological symptoms.

Introduction

A variety of symptoms are frequently reported as being part of a menopausal syndrome. These include hot flushes or flashes, night sweats, menstrual irregularities, vaginal dryness, as well as other symptoms such as depression, nervous tension, palpitations, headaches, insomnia, lack of energy, difficulty concentrating, and dizzy spells (WHO Scientific Group, 1996). The question of whether a universal “menopausal syndrome” exists has been debated for some time (Donovan, 1951, Neugarten and Kraines, 1965, McKinlay and Jefferys, 1974, Holte and Mikkelsen, 1991). Clinicians have suggested a constellation of symptoms experienced by most women due to declining levels of estrogen as they pass through menopause. An opposing view is that no universal syndrome exists. Rather, women experience a variety of symptoms and the ones experienced and reported by women are due to individual and cultural influences.

The question of whether a single menopausal “syndrome” exists and whether this is a universal phenomenon can be addressed in several ways: (1) use of data reduction techniques such as studies of clusters or groupings of symptoms, (2) the consistency of associations between specific symptoms and menopausal status across studies, and (3) comparisons of symptom reporting according to menopausal status across cultures.

This paper directly addresses the diversity of the menopause experience by comparing symptom reporting among five racial/ethnic groups of women in the US who are all part of a single study and all aged 40–55 at the time of survey: (non-Hispanic) Caucasian, African-American, Chinese, Japanese, and Hispanic. While all the women reside in the United States, the cultural experiences of the various groups differ widely. Evaluation of the extent to which symptoms group together and consistently relate to menopausal status across these five populations provides evidence for or against a menopausal syndrome and the universality of this experience. The specific research questions addressed in this paper are: (1) How does the factor structure of symptoms among mid-aged women compare across racial/ethnic groups? (2) Is symptom reporting related to race/ethnicity or menopausal status? and (3) Does the relation between menopausal status and symptoms vary across racial/ethnic groups?

A number of researchers have used factor analysis and related approaches to determine how menopausal symptoms group together. One of the earliest such studies was conducted by McKinlay and Jefferys (1974) in which they assessed relationships between symptoms, as well as sociodemographic variables among a sample of 638 women aged 45–54. Pairwise correlation coefficients were transformed to represent distances between two-dimensional points such that these distances indicated strength of pairwise associations. Results of this analysis showed that, except for hot flashes, none of the symptoms was related to menopause status. A similar analysis was conducted by Boulet, Oddens, Lehert, Vemer and Visser (1994) among 2992 women from seven south-east Asian countries. This analysis yielded two factors. One factor, consisting of vasomotor symptoms, dizziness, palpitations, and incontinence, was associated with perimenopausal status. The other factor, which included depression, insomnia, headache, anxiety, and irritability, was more closely associated with postmenopausal status.

The first reported factor analysis of symptoms was conducted by Greene (1976) who factor analyzed a list of 30 symptoms from a study of 50 British women aged 40–55 who were attending a hormone replacement clinic in London. Greene found three factors, which he labeled vasomotor, somatic, and psychological. Since Greene’s small, clinic-based study, numerous large, population and clinic-based studies have conducted factor analyses of symptom lists.

Holte and Mikkelsen (1991) reported results of a factor analysis of 21 symptoms from 1566 Norwegian women aged 45–55 who were not taking HRT that yielded five factors which they labeled vasomotor complaints, mood lability, nervousness, vague somatic complaints, and urogenital complaints. Dennerstein et al. (1993) conducted a factor analysis of 22 symptoms among a population-based sample of 1897 women aged 45–55 in Melbourne Australia and found seven first order factors: dysphoria, vasomotor, cardiopulmonary, skeletal, digestive, respiratory, and general somatic. A second order analyses, yielded three factors: vasomotor, general respiratory, and psychosomatic. Collins and Landgren (1995) report results of a study of 1324 Swedish women all aged 48 years. From a 20-item symptom list, they found the following four factors: negative mood, vasomotor symptoms, decreased sexual desire, and positive well-being. This was the only study that included aspects of positive mood. Other factor studies have also been conducted in the US (Mitchell & Woods, 1996) and Britain (Hunter et al., 1986, Kuh et al., 1997).

One study reported by Avis, Kaufert, Lock, McKinlay and Vass (1993) compared the factor structure of symptoms in three distinct populations (Canada, US, and Japan). All three studies were based on similar sample characteristics (i.e., community-based; aged 45–55; and pre-, peri-, and naturally postmenopausal women). Avis et al. found somewhat different factor structures across the three samples, which could have been due to the differences in number of symptoms assessed among the three studies or to true differences in the types of symptoms experienced. For each population, however, the vasomotor symptoms loaded on a factor separate from any of the psychological or somatic symptoms.

These studies differ in terms of the specific symptoms studied, number of symptoms included in the list (ranging from 20 to 36), time frame for symptom reporting (from the past two weeks up to one year), as well as cut-point for factor loadings. Studies also differ in sample characteristics such as age of sample (two studies included only women aged 47 or 48, while others included wider age ranges), composition of sample (some excluded women taking estrogen), and whether the sample was clinic- or community-based. Studies were conducted in a variety of countries including the United States (Avis et al., 1993, Mitchell and Woods, 1996), Canada (Avis et al., 1993, Kaufert et al., 1988), Australia (Dennerstein et al., 1993), Great Britain (Hunter et al., 1986, Kuh et al., 1997), Sweden (Collins & Landgren, 1995), Norway (Holte & Mikkelson, 1991), Japan (Avis et al., 1993), and south-east Asia (Boulet et al., 1994). Despite these differences, the results are overwhelmingly consistent in one respect; in every study, vasomotor symptoms load on a separate factor from psychological symptoms.

Except for the Avis et al. and Boulet et al. studies, these analyses are all based on samples of Caucasian women. We have little data on how symptoms group together for non-Caucasian women.

The majority of studies examining the relation between menopausal status and symptoms find vasomotor symptoms occurring more frequently among postmenopausal women (Collins and Landgren, 1995, Hunter et al., 1986, Kuh et al., 1997, Dennerstein et al., 1993, Dennerstein et al., 1994, Kaufert et al., 1988). However, findings from these and other studies are less consistent for psychological symptoms. The majority of cross-sectional studies do not find a relation between menopausal status and various measures of mood disturbance (Dennerstein et al., 1993, Dennerstein et al., 1994, Holte and Mikkelsen, 1991, Hunter et al., 1986, Kuh et al., 1997, Porter et al., 1996, Ballinger, 1976, Cawood and Bancroft, 1996), although several studies have found greater psychological or mood symptoms among perimenopausal women (Ballinger, 1975, Avis et al., 1993, Hunter et al., 1986). These studies, however, have generally not controlled for vasomotor symptoms. When vasomotor symptoms are included in the model, this relationship has attenuated (Avis et al., 1994, Collins and Landgren, 1995). These studies are based on Western, largely Caucasian, women.

Several studies of non-Western women suggest cultural differences in menopausal symptoms. For example, in a study of 483 Indian women of the Rajput caste in India, Flint (1975) found that few women had any problems with menopause other than cycle changes; they reported no depression, dizziness or incapacitation. Lock has extensively studied Japanese women and finds that only a small proportion of Japanese women aged 45–55 experience depressive symptoms or irritability and these symptoms varied little with menopausal status. Rates of hot flushes and night sweats were low in comparison with those reported in Western culture (Lock, 1986). In a cross-cultural comparison of the rates of somatic and psychological symptoms, Avis and colleagues (1993) reported that rates of almost every symptom were lower in the Japanese than in groups of US and Canadian women of similar ages. Furthermore, the percentage reporting five or more symptoms was substantially lower in the Japanese compared to the US and Canadian women.

Two recent studies found considerable variation in symptom reporting among women in different south-east Asian countries (Boulet et al., 1994, Punyahotra et al., 1997). These samples, however, were not representative and varied considerably with regard to many sociodemographic variables that might affect symptoms. The authors cite these differences, but did not conduct analyses controlling for these variables.

Other cross-cultural studies have significant methodological limitations, such as small sample sizes, use of patient samples, lack of a comparison group, use of non-standardized measures, and/or unclear definitions of menopausal status (Davis, 1982, Indiria and Murthy, 1980; Flint & Garcia, 1979).

Cross-cultural/ethnic studies of menopause are very limited in the United States. One study in the Philadelphia area found similar levels of psychological, vasomotor, genitourinary, and general physical symptoms in a convenience sample of 33 African-American and 35 Caucasian premenopausal women, aged 44–49 (Pham, Grisso & Freeman, 1997). In contrast, the Pittsburgh Healthy Women Study (unpublished data) found that African-American women, compared to Caucasian women, had significantly higher levels of depressive symptoms and tension. The size of the African-American sample, however, was very small.

In summary, epidemiologic studies of menopause have been carried out largely in Western cultures on Caucasian women. Those conducted in other cultures have used differing measures, varying sampling methodology and composition, and yielded inconsistent findings. Thus, we do not know to what extent these findings are generalizable to non-Western and for non-white women.

Section snippets

Methods

Data in these analyses are from the Study of Women’s Health Across the Nation (SWAN), a multi-race, multi-ethnic, multi-site study of middle-aged women from across the United States. SWAN is comprised of two stages: a cross-sectional telephone or in-home survey conducted between November 1995 and October 1997 and a longitudinal investigation to track changes in women’s physical and mental health as they age and traverse the menopause transition. SWAN’s study design has been described in detail

Sample

Table 1 shows the characteristics of the analytic sample for the full sample and by race/ethnicity. All variables, including menopausal status, were related to ethnicity. Compared with all other groups surveyed, African-American women were much more likely to have had a surgical menopause, while Chinese and Japanese women were less likely to have had a surgical menopause than Caucasian or Hispanic women. The Chinese and Japanese women were more likely to be premenopausal. Hormone use was

Discussion

This paper sought to determine the evidence for a menopausal syndrome across five racial/ethnic groups. The results of the five factor analyses reported in this study do not support a single syndrome consisting of both menopausal and psychological symptoms. Across all five racial/ethnic groups, two factors emerged; one consisting of clearly vasomotor symptoms — hot flashes and night sweats — and the other consisting of psychological and psychosomatic symptoms. These results, along with other

Acknowledgements

The Study of Women’s Health Across the Nation (SWAN) was funded by grants from the National Institute on Aging and the National Institute of Nursing Research of the National Institutes of Health to the following participating centers: University of Michigan, Ann Arbor (U01 NR04061; U01 AG12495); Massachusetts General Hospital, Boston, MA (U01 AG12531); Rush University, Rush–Presbyterian–St. Luke’s Medical Center, Chicago, IL (U01 AG12505); University of California, Davis (U01 AG12554);

References (44)

  • E.S. Mitchell et al.

    Symptom experiences of midlife women: observations from the Seattle Midlife Women’s Health Study

    Maturitas

    (1996)
  • S. Punyahotra et al.

    Menopausal experiences of Thai women. Part 1: symptoms and their correlates

    Maturitas

    (1997)
  • G.W.K. Tang

    The climacteric of Chinese factory workers

    Maturitas

    (1994)
  • A. Agresti

    Analysis of ordinal categorical data

    (1984)
  • H. Aldercruetz et al.

    Dietary phytoestrogens and the menopause in Japan

    Lancet

    (1992)
  • Avis, N. E., Kaufert, P. A., Lock, M., McKinlay, S. M., Vass, K. (1993). The evolution of menopausal symptoms....
  • C.B. Ballinger

    Psychiatric morbidity and the menopause: screening of general population sample

    British Medical Journal

    (1975)
  • C.B. Ballinger

    Psychiatric morbidity and the menopause: clinical features

    British Medical Journal

    (1976)
  • E.H.H. Cawood et al.

    Steroid hormones, the menopause, sexuality and well-being of women

    Psychological Medicine

    (1996)
  • A. Collins et al.

    Reproductive health, use of estrogen and experience of symptoms in perimenopausal women: a population-based study

    Maturitas

    (1995)
  • L. Dennerstein et al.

    Menopausal symptoms in Australian women

    The Medical Journal of Australia

    (1993)
  • M.P. Flint et al.

    Culture and the climacteric

    Journal of Biosocial Sciences

    (1979)
  • Cited by (522)

    View all citing articles on Scopus
    View full text