Sommer, Barbara PhD; Avis, Nancy PhD; Meyer, Peter PhD; Ory, Marcia PhD; Madden, Tom MD; Kagawa-Singer, Marjorie PhD; Mouton, Charles MD; Rasor, Niki O’Neill MA; Adler, Shelley PhD
HRT = hormone replacement therapy, HSD = Tukey’shonestly significant difference, SWAN = Study of Women’s HealthAcross the Nation.,
Menopause, the cessation of menstruation, is a psychosocial as well as a biological event. Attitudes, perceptions, and expectations are part of the psychosocial phenomena surrounding menopause (1). Both women in midlife and health professionals believe that attitudes play a role in the experience of menopause (2), with results from three longitudinal studies supporting this belief (3–5). However, the majority of the studies have been conducted among western and mostly white women.
Some ethnic differences in attitude have been documented, particularly between Asian and western women (6, 7). The literature on Asian samples, conducted across nations and religions, repeatedly invokes the theme of postmenopausal women being released from menstrual taboos and gaining increased power and respectability, thereby producing positive attitudes toward aging (6, 8–11). Among the Japanese, for whom there is not a specific word for menopause, the life phase of “kônenki,” a period ranging from the early forties to the mid fifties, is one in which women are perceived to be more vulnerable to emotional and physical difficulties than are men (6).
Within the United States, the few studies that have included African American women have suggested that African American women have a more positive attitude that that expressed by white women (12–14). Studying Hispanic women, Bell (15) found generally positive attitudes similar to those reported by non-Hispanic women, but menopause was perceived as potentially disturbing with accompanying irritability and depression.
Much of the data on attitudes from women of various ethnic backgrounds come from investigations using different design methodologies, thereby hampering cross-study comparisons. Studies differ in terms of specific questions asked, study design, and representativeness (ie, some studies use convenience samples or volunteers; others are population based). Furthermore, one cannot make cross-ethnic comparisons without controlling for important variables that affect attitude (eg, education).
Menopause status seems to have an impact on attitude. Postmenopausal and older women consistently express more positive feelings about menopause than do younger women, both somewhat younger (eg, in their forties) (3, 16–19) and much younger (teens and twenties) (20–22). Avis and McKinlay (3) found that surgically menopausal women held more negative attitudes toward menopause than those going through the normal transition. Wilbur et al. (19) did not confirm this finding in their sample, which included more African American women.
SWAN, a multiethnic, community-based, national study initiated in the fall of 1995, provided the opportunity to address variations in attitude across ethnic groups and menopausal statuses within a single study. The large sample size also made it possible to take into account other variables likely to influence attitudes, such as acculturation, education, and other demographic variables.
The specific questions addressed in this study are as follows: 1) Do attitudes toward aging and menopause within this US community-based study differ across ethnic groups, and, if so, does acculturation play a role? 2) Does the positive relationship between menopausal status and attitude hold across ethnic groups?
The SWAN cross-sectional sample consists of 16,065 women who were interviewed by telephone or in person from seven sites across the United States during 1995 and 1996. The women were selected using a variety of random, list-based, and snowball sampling techniques (23). In addition to sampling the white population, a minority population was sampled at each site (African Americans in Pittsburgh, PA, Boston, MA, Ann Arbor, MI, and Chicago, IL; Japanese Americans in Los Angeles, CA; Chinese Americans in the San Francisco, CA, East Bay region; and Hispanics in New Jersey). Ethnic group was self-defined. To ensure representativeness, only those women belonging to the principal primary ethnic groups at each site were included in the final sample.
Women who were pregnant, breast-feeding, or had used hormones within the past 3 months were excluded, reducing the sample to 12,275. Attitude scores were missing from 49 women, resulting in a final sample of 12,226 women ranging in age from 40 to 55 years.
Professional translators constructed Cantonese, Japanese, and Spanish versions of the interview items, which were originally created in English. All three versions were checked word by word with the English version by a second bilingual speaker and were back-translated. Revisions went back to the original translator, and discrepancies were resolved.
Ten agree/disagree statements pertaining to attitudes toward menopause and aging were drawn from existing instruments and were modified to provide a balance of negative and positive wording (18, 22, 24–36). The interview protocol also contained questions pertaining to menstrual history, symptom occurrence within the preceding 2 weeks, and demographic information. A complete description of the interview has been published elsewhere (23).
The participants were asked whether they 1) agreed, 2) felt neutral, or 3) disagreed with each of 10 statements concerning menopause and aging. An independent Thurstone procedure revealed that 7 of the 10 statements show a clear positive or negative direction. (Each of the 10 statements was rated by 154 university students using an 11-point scale ranging from “very negative” through “neutral” to “very positive.” Three items were clearly positive, having mean and median scores of 9. Four items had an average score of 4 or less (negative). The remaining three were judged as neutral with regard to menopause and aging.)
The statements were as follows: 1) “The older a woman is, the more valued she is.” 2) “A woman is less attractive after menopause.” 3) “Women who no longer have menstrual periods feel free and independent.” 4) “Overall, going through the menopause or change of life will be, or was, a positive experience for me.” 5) “As I age, I feel worse about myself.” 6) “During the menopause or the change of life, I became, or expect to become, irritable or depressed.” 7) “I will feel, or felt, regret when my periods stopped for the last time.” The remaining three statements were 8) “Menopause is a midlife change that generally does not need medical attention.” 9) “Women with little free time hardly notice the menopause.” 10) “I don’t, or didn’t, know what to expect with the menopause.” Because there was not a consensus with regard to either a positive or negative direction for these last three items, they were not included in the analysis.
Scoring was modified to provide a consistent direction. Agreement with positive items and disagreement with negative items was scored as 3. The converse was scored as 1. All seven items were positively correlated (Cronbach’s α = 0.51). A composite attitude toward menopause and aging score (attitude) was created by reversing the scores on the negative items and calculating the seven-item mean for each respondent. The resulting attitude score varied from 1 (negative) to 3 (positive). Respondents who failed to answer at least four of the items were excluded from the analysis. In the other cases, the mean of the answered items was used.
Ethnic group was determined by self-identification and categorized as African American, white, Chinese American, Japanese American, or Hispanic (the Hispanic category included women who identified themselves as Puerto Rican, Mexican or Mexican American, Dominican, Central American, Cuban or Cuban American, South American, Spanish, and other Hispanic).
Menopause status was defined as follows: postmenopausal, no menstrual period in past 12 months; late perimenopausal, bleeding in past 12 months but not in the past 3 months; early perimenopausal, menstrual period in past 3 months, experiencing irregularity; premenopausal, menstrual period in past 3 months, not experiencing irregularity; and surgical menopause, hysterectomy or bilateral ovariectomy. (Technically, a woman with intact ovaries is not surgically menopausal; however, pilot testing revealed considerable confusion among respondents with regard to the distinctions between ovariectomy and hysterectomy. Thus, a clear distinction could not be made.)
The early and late perimenopausal distinctions were made because the SWAN study also involves hormonal markers, and each stage is assumed to reflect a somewhat different hormonal substrate (an empirical issue to be addressed in other SWAN reports).
Known predictors and possible covariates of attitudes toward menopause and aging needed to be taken into account to examine the effects of ethnic group and menopause status. One set consisted of demographic variables: age, education, marital status, and parity (number of children), and site (geographical locale). Age was measured in years. Education was divided into five levels: less than high school, high school graduate, post high school, college graduate, postgraduate. Marital status was coded in five categories: single, never married; currently married or living as if married; separated; widowed; and divorced. Three levels of parity were used: no children, 1 child, and 2 or more children. Site covers the seven locations from which samples were drawn.
The experience of symptoms might affect reported attitude. Therefore, they formed a second set of potential covariates. Hot flashes and night sweats were entered into the analysis by creating a dichotomous “vasomotor” variable of either hot flashes and/or night sweats or neither during the preceding 2 weeks. If the respondent answered “no” to one of the two symptoms and left the other blank, the data were coded as missing.
Three symptoms related to mood were used to construct a “dysphoric mood” variable: feeling tense or nervous, feeling blue or depressed, and irritability or grouchiness. A factor analysis showed that the three mood symptoms loaded >0.60 on one factor (NE Avis, unpublished data, 1999). To distinguish between a possibly typical variation in mood and a more severe dysphoria, log linear models were applied to the data. There was a break point between two and three symptoms. If a woman reported one symptom, she was likely to report a second but not a third. Thus, the variable was scored dichotomously, contrasting women reporting three symptoms vs. those reporting fewer than three. The criterion was also adopted to be consistent with that of another SWAN report (J Bromberger, unpublished data, 1999).
Two final variables included in the model measured acculturation. “Language” refers to the language used in the telephone interview and was coded as English, Spanish, Chinese, or Japanese. “Country of education” reflects subjects’ responses to the question, “In what country did you complete your education, excluding vocational school or college?” (available options were the United States (including US territories), China, Japan, and other). The variable was dichotomized into more acculturated (interview language as English, schooling in United States) vs. less acculturated (interview in non-English language or schooling outside the United States).
A three-stage model using a backward stepwise regression procedure was used to test the effects of ethnic group and menopause status on attitude. We first entered the demographic, symptom, and acculturation variables. The second stage involved ethnic group. To control for the possible effects of geographic location, we included a site-by-ethnic group interaction term in the model. The third step was to test the effect of menopause status, having controlled for demographic, symptom, ethnic, and site variables. A general description of the sample for each variable is presented in Table 1. We also examined responses to the individual items.
Table 2 shows the univariate results for attitude by ethnic group. Although small, all of the differences between means were statistically significant at the p= .001 level (HSD) (37), with the exception of the Hispanic vs. Japanese American difference, which had a probability of 0.03 (overall F (4/12,221) = 234.20, p< .001). These results show that African American women had the most positive attitudes and that the Chinese American and Japanese American women had the least positive attitudes. It is important to note that all the means were on the positive side of neutral.
The univariate results for menopause status are shown in Table 3. The differences, although quite small, were significant (overall F (4/12,221) = 44.05, p< .001). The premenopausal and early perimenopausal groups differed from the late perimenopausal and postmenopausal groups, with the surgical menopause group showing the most positive attitude (HSD, p< .001). Differences in attitude across menopause statuses were smaller than differences across ethnic groups.
The modeling procedure, described in Analysis Plan, was applied to take into account the effects of potential confounds. Table 4 shows the results of the final model, which included age, education, parity, dysphoric mood, acculturation, ethnic group, site, ethnic group by site, and menopause status. Because they were not associated with attitude (using p< .05), vasomotor symptoms and marital status were dropped from the model. The adjusted R2 value was 0.158. The ethnic and menopause status effects remain, with the former being the stronger of the two. The African American women had a more positive attitude toward menopause and aging than any of the other four ethnic groups, and the effect held up across menopause status (Figure 1).
Acculturation seems to play a variable role in attitude. Women who were interviewed in Chinese or Japanese and who had received their basic education outside the Unites States scored significantly lower in attitude (Table 5). There was no consistent acculturation effect for Hispanic women.
We tested whether there was a significant interaction between ethnic group and menopause status in the final model. The p value associated with that interaction was .001 (F (16,11956) = 2.4), indicating that there were significant differences between ethnic groups. To more clearly understand the role of menopause status within particular ethnic groups, a model was constructed for each ethnic group, first including confounders that were significant for the particular ethnic group and then testing menopause status. These analyses showed menopause status to be a significant predictor of attitude for African American, Hispanic, Japanese, and white women but not for Chinese American women. Figure 1 shows the estimated means based on these models for each ethnic group. As shown in Figure 1, there was a tendency for the latter stages of menopause to be associated with a more positive attitude.
To gain a more specific sense of ethnic differences in attitudes, we also examined responses to specific items.
Figure 2 shows the mean response by ethnic group to each of the three items related to aging. The mean for African American women was significantly more positive than those for the other four ethnic groups on each of the three items. The other four groups did not differ substantially in their level of agreement regarding the value of older women. The Chinese Americans scored significantly lower than all other groups on the other two aging items and tended to have the most neutral attitudes. In general, women’s attitudes about aging were positive (Figure 2).
Affective quality of the menopausal experience.
Although African Americans were less likely to agree with the expectation of irritability or depression, they were not as sanguine as the Hispanics with regard to menopause as a positive experience (Figure 3). More of the Chinese American and Japanese American respondents were, or expected to be, irritable or depressed.
Sense of freedom.
The white and African American women were more likely to endorse the view that absence of menstruation confers freedom and independence (Figure 4). There were no significant differences among the other three groups.
Hispanic women were most likely to express regret, and African American women, the least likely (Figure 4).
Attitudes toward menopause and aging differed across ethnic groups, with African American women being the most positive and the less acculturated Chinese American and Japanese American women the least positive. The general finding that attitudes toward menopause ranged from neutral to positive is consistent with findings of previous research and can be interpreted in light of the general life context of women experiencing the transition. Many have lived through major life events and transitions and lead busy lives with other things going on, such as changes in family structure, work and career shifts, readjustments in interpersonal relationships, economic hardship or gain, loss of loved ones, and personal triumphs and defeats. For less resilient individuals and those lacking economic and social support, the menopausal transition may be difficult to accommodate (J Bromberger, unpublished data, 1999). For others, menopause may simply be one more transition in life’s course. Community discussions suggested that one reason the African American women were more positive was that compared with the consequences of racism they had experienced throughout their lives, menopause seemed a minor stressor. Family patterns might also contribute; for example, perhaps African American women spend more time with their older female relatives and witness the menopausal transition more closely and thus become less susceptible to false stereotypes (14).
The finding that more African American and white women agreed with the perception of menopause signaling freedom and independence may reflect a higher valuation of these qualities than is the case in the Asian and Hispanic cultures, which are generally thought of as more communal (38). One can only speculate as to why a greater proportion of Hispanic women agreed that they would feel regret when their periods stopped. Perhaps there are group differences in the salience of menstruation as a marker of womanhood or of the centrality of the motherhood role.
The result that less acculturated women of Japanese and Chinese extraction were substantially less positive runs counter to the traditional concept of respect for the elderly in the Asian culture, which might be expected to translate into a more positive attitude toward menopause as a manifestation of aging (8, 39). The finding from SWAN differs from those described in the Introduction and raise questions about the assumption that Asian women are more positive about aging and menopause than western women because of a more positive cultural attitude toward the aged. The picture appears to be more complex. Part of the difference in findings may reflect different contexts and methods. The cited studies were conducted abroad, and several used qualitative methods, perhaps reaching deeper levels of meaning. The SWAN data have the advantage of comparing the 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 toward menopause.
Menopause status was not a strong predictor of attitude. Contrary to the finding of Avis and McKinlay (3), surgically menopausal women were not more negative than the other women, a finding in accord with that of Wilbur et al. (19). The latter study had a higher proportion of African American women in the surgical menopause group than did the former.
For the sample as a whole, the premenopausal and early perimenopausal women tended to be less positive in attitude, a finding in accord with previous research. However, the difference by menopause status was neither substantial nor consistent across ethnic groups, suggesting that factors other than direct experience with menopause or its imminence are playing a role in attitude.
There have been studies of women’s attitudes toward HRT (35, 40–42) but not of the relationship between menopausal attitudes and whether one takes HRT. It is conceivable that the exclusion of women using HRT would inflate the attitude scores; that is, that women with more negative perceptions and expectations toward menopause and aging might be more likely to use hormonal treatment.
The issue of comparability of questions across cultures remains. When translated instruments are used, there is always the possibility that concepts change when described in a different language. Even when the language is clear, items may be read within a cultural context that influences how they are interpreted (43, 44). It may be no accident that the variables included in this study were least effective in accounting for attitude in the Japanese and Chinese American samples and more effective for the African American and white women. A substantial amount of the variance remains to be explained. Because the SWAN project is designed to obtain a broad overview of the menopausal phenomenon, there were limitations on inclusion of specific variables relevant to attitude, such as the influence of the menopausal experience of one’s mother, relatives, or friends; life events; history of menstrual symptoms and/or premenstrual dysphoria; attitudes toward femininity; coping styles; and personality. An extensive evaluation of some of these variables and of the relationship between psychological symptoms and attitudes will be provided by longitudinal data forthcoming in the SWAN study.
1. Avis NE. Women’s perceptions of the menopause. Eur Menopause J 1996; 3: 80–3.
2. Cowan G, Warren LW, Young JL. Medical perceptions of menopausal symptoms. Psychol Women Q 1985; 9: 3–14.
3. Avis NE, McKinlay SM. A longitudinal analysis of women’s attitudes toward the menopause: results from the Massachusetts Women’s Health Study. Maturitas 1991; 13: 65–79.
4. Hunter MS. Psychological and somatic experience of the menopause: a prospective study. Psychosom Med 1990; 52: 357–67.
5. Matthews KA. Myths and realities of the menopause. Psychosom Med 1992; 54: 1–9.
6. Lock M. Menopause in cultural context. Exp Gerontol 1994; 29: 307–17.
7. Lock M. Menopause: lessons from anthropology. Psychosom Med 1998; 60: 410–9.
8. Chen Y-LD, Voda AM, Mansfield PK. Chinese midlife women’s perceptions and attitudes about menopause. Menopause 1998; 5: 28–34.
9. Flint M. Male and female menopause: a cultural put-on. In: Voda AM, Dinnerstein M, O’Donnell SR, editors. Changing perspectives on menopause. Austin (TX): University of Texas Press; 1982. p. 363–78.
10. Punyahotra S, Dennerstein L. Menopausal experiences of Thai women: part 2. The cultural context. Maturitas 1997; 26: 9–14.
11. Tang GWK. The climacteric of Chinese factory workers. Maturitas 1994; 19: 177–82.
12. Holmes-Rovner M, Padonu G, Kroll J, Rovner DR, Talarczyk G, Rothert M. African-American women’s attitudes and expectations of menopause. Am J Prev Med 1996; 12: 420–3.
13. Padonu G, Holmes-Rovner M, Rothert M, Schmitt N, Kroll J, Rovner D, Talarczyk G, Breer L, Ransom S, Gladney E. African-American women’s perception of menopause. Am J Health Behav 1996; 20: 242–51.
14. Standing TS, Glazer G. Attitudes of low-income clinic patients toward menopause. Health Care Women Int 1992; 13: 271–80.
15. Bell ML. Attitudes toward menopause among Mexican American women. Health Care Women Int 1995; 16: 425–35.
16. Abraham S, Llewellyn-Jones D, Perz J. Changes in Australian women’s perception of the menopause and menopausal symptoms before and after the climacteric. Maturitas 1994; 20: 121–8.
17. Frey K. Middle-aged women’s experience and perceptions of menopause. Women Health 1981; 6: 24–36.
18. Neugarten BL, Wood V, Kraines RJ, Loomis B. Women’s attitudes toward the menopause. Vita Hum 1963; 6: 140–51.
19. Wilbur J, Miller A, Montgomery A. The influence of demographic characteristics, menopausal status, and symptoms on women’s attitudes toward menopause. Women Health 1995; 23: 19–39.
20. Bowles C. Measure of attitude toward menopause using the semantic differential model. Nurs Res 1986; 35: 81–5.
21. Gannon L, Ekstrom B. Attitudes toward menopause: the influence of sociocultural paradigms. Psychol Women Q 1993; 17: 275–88.
22. Leiblum SR, Swartzman LC. Women’s attitudes toward the menopause: an update. Maturitas 1986; 8: 47–56.
23. Sowers M, Crawford S, Sternfeld B, Morganstein D, Gold E, Greendale G, Evans D, Neer R, Matthews K, Sherman S, Lo A, Weiss G, Kelsey J. Design, survey, sampling and recruitment methods of SWAN: a multi-center, multi-ethnic, community-based cohort study of women and the menopausal transition. In: R Logo, R Marcus, and Kelsey J, editors. Menopause. New York: Academic Press. In Press.
24. Groeneveld FPMJ, Bareman FP, Barentsen R, Dokter HJ, Drogendijk AC, Hoes AW. Relationships between attitude toward menopause, well-being and medical attention among women aged 45–60 years. Maturitas 1993; 17: 77–88.
25. Kafer RA, Rakowski W, Lachman M, Hickey T. Aging opinion survey: a report on instrument development. Int J Aging Hum Dev 1980; 11: 319–33.
26. Kogan N. Attitudes toward old people: the development of a scale and an examination of correlates. J Abnorm Soc Psychol 1961; 62: 44–54.
27. Lasher KP, Faulkender PJ. Measurement of aging anxiety: development of the Anxiety About Aging Scale. Int J Aging Hum Dev 1993; 37: 247–59.
28. Lawton MP, Kleban MH, Rajagopal D, Dean J. Dimensions of affective experience in three age groups. Psychol Aging 1992; 7: 171–84.
29. Maoz B, Dowty N, Antonovsky A, Wijsenbeek H. Female attitudes to menopause. Soc Psychiatry 1970; 5: 35–40.
30. Matthews KA, Bromberger J, Egeland G. Behavioral antecedents and consequences of the menopause. In: Koreman SG, editor. The menopause. Nowell (MA): Serano Symposia; 1989. p. 1–15.
31. Matthews KA, Wing RR, Kuller LH, Meilahn EN, Kelsey SF, Costello EJ, Caggiula AW. Influences of natural menopause on psychological characteristics and symptoms of middle-aged healthy women. J Consult Clin Psychol 1990; 58: 345–51.
32. Neugarten BL, Wood V, Kraines RJ, Loomis B. Women’s attitudes toward the menopause. In: Neugarten BL, editor. Middle age and aging: a reader in social psychology. Chicago: University of Chicago Press; 1968. p. 195–200.
33. Neugarten BL, Kraines RJ. “Menopausal symptoms” in women of various ages. Psychosom Med 1965; 27: 266–73.
34. Rosencranz HA, McNevin TE. A factor analysis of attitudes toward the aged. Gerontologist 1969; 9: 55–9.
35. Schmitt N, Gogate J, Rothert M, Rovner D, Holmes M, Talarcyzk G, Given B, Kroll J. Capturing and clustering women’s judgment policies: the case of hormonal therapy for menopause. J Gerontol 1991; 46: P 92–101.
36. Speas K, Obenshain B. Images of aging in America. Washington DC: American Association of Retired Persons; 1995.
37. Hays WL. Statistics. New York: Holt, Rinehart, and Winston; 1988.
38. Rogler LH. International migrations: a framework for directing research. Am Psychol 1994; 49: 701–8.
39. Hong GK, Lee BS, Lorenzo MK. Somatization in Chinese American clients: implications for psychotherapeutic services. J Contemp Psychother 1995; 25: 105–18.
40. Pham K-TC, Freeman EW, Grisso JA. Menopause and hormone replacement therapy: focus groups of African-American and Caucasian women. Menopause 1997; 4: 71–9.
41. Quine L, Rubin R. Attitude, subjective norm and perceived behavioural control as predictors of women’s intentions to take hormone replacement therapy. Br J Health Psychol 1997; 2: 199–216.
42. Ferguson KJ, Hoegh C, Johnson S. Estrogen replacement therapy: a survey of women’s knowledge and attitudes. Arch Intern Med 1989; 149: 133–6.
43. Munet-Vilaro F, Egan M. Reliability issues of the Family Environment Scale for cross-cultural research. Nurs Res 1990; 39: 244–7.
44. Reid PT, Kelly E. Research on women of color: from ignorance to awareness. Psychol Women Q 1994; 18: 477–86.