Contextual influences on women's health concerns and attitudes toward menopause.

Contextual influences on women's health concerns and attitudes toward menopause. The purpose of this study was to better understand the ways inwhich the social context of midlife affects women's attitudestoward menopausal aging. This study builds on prior research thatidentifies contextual factors. Adaptation to menopause has many facets,including adjustment to the loss of fertility as well as concerns aboutphysical attractiveness and health (Rossi, 2004). What is not clear iswhether women's adjustment to menopause reflects a singleunderlying dimension that has a common set of antecedents or if it ismore meaningful to differentiate between adjustment to the loss offertility on the one hand and women's concerns regarding health andattractiveness on the other. One contextual factor--having multiple roles--has the potential toenhance women's adaptation to menopause. According torole-enhancement theory (Thoits, 1983), multiple roles can promotewomen's development by increasing self-esteem, providing purposeand meaning in life, increasing one's ability to develop deepconnections with others, and buffering against strain or difficultiesexperienced in one role. The perspective of role accumulation theory hasreceived support from numerous studies of women's well-being(Barnett & Hyde, 2001; Chrouser-Ahrens & Ryff, 2006; Crosby,1991). Although little attention has been given to the potentialcontribution of multiple roles to women's adjustment to menopause,some initial studies suggest that adaptation is better among women whooccupy multiple roles. For example, paid work outside of the home mayhave a positive impact on women's experience of menopausal symptoms(Polit & Larocco, 1980; Prill, 1977). Paid work may be particularlyimportant as a buffer against loss of fertility during menopause, aswork can provide women with a meaningful identity that is not rooted intheir capacity to carry children. Other contextual factors that have been found to be related towomen's adjustment to menopause include age, education, financialsecurity, and symptom levels. Younger women may be more fearful of theimpending experience of menopause than older women (Eisner & Kelly,1980; Kresovich, 1980; Neugarten, Wood, Kraines, & Loomis, 1963)and, therefore, have more negative attitudes toward menopause (Dege& Gretzinger, 1982; Perlmutter & Bart, 1982; Wilbur, Miller,& Montgomery, 1995). Women with lower levels of education and incomehave been found to report more intense menopausal symptoms (Hunter,1990; Hunter, Battersey, & Whitehead, 1986; Jaszmann, Van Lith,& Zatt, 1969; Leiblum & Swartzman, 1986; McKinlay, McKinlay,& Brambilla, 1987; Polit & Laracco, 1980) and tend to showgreater increases in depression during menopause (McKinlay et al.,1987). Women who perceive menopause as a medical problem, rather than asa part of normative aging, tend to show lower levels of psychologicalwell-being (Avis & McKinlay, 1991). Although these studies suggestthat women's adjustment to menopause can be shaped by the socialcontext in which this life transition occurs, it is not clear whethermultiple contextual antecedents have the same adaptational significancefor all facets of adjustment to menopause (for example, loss offertility, concerns about attractiveness and health). The present study examined the influence of contextual factors in asample of women who were born in to the baby boom generation and whoexperienced menopause at the turn of the century. The present study maytherefore provide a more contemporary perspective on the contributionsof potential protective factors on adjustment to menopause. Many of themajor menopause studies cited previously were conducted during the1960s, 1970s, and 1980s. The women in these earlier studies were bornbefore the postwar baby boom generation. The experiences of women bornduring the postwar baby boom may not be represented in these earlierstudies. For the generation born after 1945, opportunities for women toexpand their repertoire of multiple roles have increased greatly due tothe steadily increasing participation of mothers in fulltime careeremployment. These women have also achieved higher levels of formaleducation and have more often attained greater levels of independentfinancial security compared with women born in earlier decades. Hence,multiple roles and other contextual factors may be even more relevant tothe adjustment of women to menopause in a contemporary sample. METHOD Sample The study used data from the National Survey of Midlife Developmentin the United States (MIDUS). This study contained two waves: MIDUS 1(1994 to 1996) and MIDUS 2 (2004 to 2006). The sample for the presentstudy consisted of 1,037 women who completed both waves of the MIDUSstudy and who were born between 1946 and 1964. Technical informationconcerning the sample and data collection procedures are available fromthe MIDMAC Web site (http://midmac.med.harvard. edu/tech.html). Measures Attitudes toward Menopause. The items used for the present studyare based on the work of Rossi (2004), who looked at three basicconcerns women have during the menopausal transition: (1) ability toreproduce, (2) attractiveness, and (3) concern with being morevulnerable to physical health problems. To measure women's level ofconcern about these three issues, the women were asked to respond to thequestions "How much do you worry about the following?":"being too old to have children," "being less attractiveas a woman," and "having more illness as you get older."Women were asked to rate their level of concern on a four-point scale,with 1 = a lot, 2 = some, 3 = a little, and 4 = not at all. Oneadditional question asked women whether they felt more relief or moreregret when their menstrual periods ended, on a six-point scale, whichwas recoded so that higher ratings indicated more relief. Roles. Roles were computed by adding the number of roles occupiedby a woman. The roles of mother, caregiver, employee, and spouse wereincluded in the count of roles. Age. Women were asked to indicate the year in which they were born.Age was calculated by the reported year subtracted from the year of thesurvey. Financial Comfort. Financial comfort was measured by one questionthat asked, "In general, would you say you (and your family livingwith you) have more money than yon need, just enough for your needs, ornot enough to meet your needs?" For the present study, this itemwas recoded so that higher scores indicated greater financial comfort. Education. The education variable was combined into four discretelevels: (1) high school graduate or less, (2) some college or anassociate's degree, (3) college graduate, and (4) postgraduatedegree. Symptoms. A menopausal symptoms scale based on the work of Rossi(2004) was used to obtain a score of symptoms for each respondent. Womenwere asked to rate how often they experienced each of five menopausalsymptoms in the past 30 days (insomnia, heavy sweating, painfulintercourse, hot flashes, and irritability). Women responded on asix-point scale, with 1 = almost every day, 2 = several times a week, 3= once a week, 4 = several times a month, 5 = once a month, and 6 =never. The symptom measure was scored so that higher scores indicatedbetter health (that is, less frequent symptoms). RESULTS The main analyses of this article contrasted taro models ofattitudes toward menopausal aging and their antecedents. The first stageof the analysis examined a one-factor model of attitudes towardmenopausal aging. The one-factor model assumed that concerns aboutfertility, attractiveness, and health and women's affectiveresponse toward menopause are all indicators of a single underlyingdimension of attitudes toward menopausal aging that is shaped by a coreset of contextual conditions in women's lives. The second stage ofthe analysis examined a more differentiated model of attitudes towardmenopausal aging and their contextual antecedents. Descriptivestatistics for the independent and dependent variables are shown inTable 1. One-Factor Model of Attitudes toward Menopause and TheirAntecedents The one-factor model of attitudes toward menopause and theirantecedents assumed that attitudes toward menopausal aging reflect asingle underlying dimension and are related to a common set ofantecedent conditions. Structural equation modeling (SEM) was used totest the fit of this model to the covariance matrix among antecedentsand measures of attitudes. Separate analyses were conducted on the datafrom the wave 1 and wave 2 cohorts. The one-factor model did not exhibitadequate levels of fit to the wave 1 data (goodness-of-fit index [GFI] =.924, normed fit index [NFI] = .510, comparative fit index [CFI] = .509,root mean residual [RMR] = .138, root mean square error of approximation[RMSEA] = .146, p < .001) or to the wave 2 data (GFI = .958, NFI =.764, CFI = .775, RMR = .164, RMSEA =. 100, p < .001). The values ofthese indices do not meet the criteria for adequate fit proposed by Huand Bentler (1999). In particular, the NFI and CFI are too small,whereas the RMR and the RMSEA are too large. Differentiation of Two Factors: Childbearing and Health andWellness The two-factor model of attitudes toward menopause and theirantecedents is shown in Figure 1. This model assumes that differentattitudes toward menopausal aging reflect different factors and arerelated to a common set of antecedent conditions. The first factorreflects attitudes pertaining to loss of childbearing capacity, asreflected in concerns about loss of fertility and affective response tothe last period. The second factor reflects concerns about changes inhealth and physical attractiveness. The antecedents to the childbearingdimension include having multiple roles, age, education, and financialsecurity. The prime antecedent to the health and wellness dimension isthe experience of menopausal symptoms. The two factors of menopausalaging are associated: In this model, attitudes toward loss ofchildbearing capacity influence the health and wellness dimension. SEMwas used to test the fit of this model to the matrix of correlationsamong antecedents and measures of attitudes. Separate analyses wereconducted on the data from the wave 1 and wave 2 cohorts. The two-factormodel exhibited adequate levels of fit to the wave 1 data (GFI = .989,NFI = .938, CFI = .950, RMR = .108, RMSEA = .048, not significant) andto the wave 2 data (GFI = .986, NFI = .922, CFI = .939, RMR = .069,RMSEA = .054, not significant). The levels of these fit indices attainthe standards proposed by Hu and Bender (1999). [FIGURE 1 OMITTED] The regression parameters for the model are shown for wave 1 andwave 2 in Tables 2 and 3, respectively. In both waves, more positiveattitudes toward the loss of childbearing capacity were found amongwomen who occupied more roles and who were older, less educated, andfinancially secure. More positive attitudes concerning health andappearance were found among women who had lower symptom levels. Theeffects found for multiple roles, age, financial security, and symptomswere consistent with the hypotheses offered earlier. However, therelationship of education to attitudes was contrary to the hypothesisthat women with higher levels of education would have more positiveattitudes to the loss of fertility. DISCUSSION The main findings of this study were consistent with a two-factormodel of adjustment to fertility. Different contextual antecedentsshaped adjustment to changes in fertility as distinct from changes inhealth and appearance. Better adjustment to the loss of fertility wasfound among women who occupied more roles and were older, less educated,and more financially secure. Women had fewer concerns about health andappearance when they experienced lower symptom levels. The antecedentsof adaptation to loss of fertility and health/appearance were distinctfrom one another in the sense that symptom levels were not associatedwith adaptation to loss of fertility, and multiple roles, age,education, and financial security were not associated significantly withconcerns about the impact of menopause on health and appearance. The predictions of role enhancement theory were partially supportedby the findings of this study. Consistent with role enhancement theory,women who occupied multiple roles had better adjustment to the loss offertility. Women who have been parents may cope with the loss offertility better than women who have never been parents. The loss offertility may be particularly problematic for some women who have neverbeen parents due to involuntary factors (that is, lifelong difficultiesin conception, spousal infertility). For these women, the loss offertility may represent closure of the chance to conceive any childrenin their lifetime. In addition, roles in paid employment or as acaregiver may provide meaningful social identities for menopausal womenthat are not based on fertility and childbirth. By contrast, having multiple roles did not facilitate adjustment tochanges in health and appearance: Women with multiple roles had the samelevel of concern about the effects of menopause on health and physicalappearance as those who had fewer roles. Further research should examinethe possibility that the contextual antecedents to menopausal symptoms,and adjustment to these symptoms, are different from the antecedents tobroader physical and mental health, where multiple roles appear to servea protective function. The loss of fertility may have been easier for older women to copewith than younger ones because of normative expectations aboutchildbearing. The younger women in this sample were in their 30s duringwave 1 and in their early to mid 40s in wave 2. For many of theseyounger women, conception is possible and may still be desired. Bycontrast, the older women in this sample were in their mid to late 40sin wave 1 and in their 50s in wave 2. For most of these women,conception (other than by in vitro fertilization) is not physicallypossible. The absence of a relationship between age and attitudes to thehealth impact of menopause is noteworthy. Although prior researchsuggests that age brings greater acceptance of and experience withphysical changes in menopause, the findings of this study suggest thatconcerns with the health effects of menopause are not allayed byexperience. Women who were less educated and more financially secure had betteradaptation to loss of fertility. Because education and financialsecurity have opposite effects on adaptation, it is difficult tointerpret this trend as an effect of a woman's overallsocioeconomic status. It is noteworthy that the effects of financialsecurity were significant even when the effects of multiple roles wereconsidered in the structural equation model. The significant coefficientfor financial security indicates that security has significant effectson adaptation after controlling statistically for the other exogenousvariables in the model (that is, multiple roles and age). This findingsuggests that the effects of financial security do not arise merelybecause women who occupy multiple roles (as spouses and as paidemployees) are often more financially secure than women who are singleor do have paid employment. Rather, financial security itself appears tofacilitate adjustment to the loss of fertility. Evidence for the two-factor model suggests that menopause is notonly an issue concerning fertility, but also one in which changes inhealth pose a distinct constellation of challenges. The findings of thisstudy suggest that issues concerning the health effects of menopause areshaped by women's experiences of menopausal symptoms, regardless oftheir educational background, income, or employment. Future researchshould consider the degree to which women's concerns about thehealth effects of menopause are shaped specifically by menopausalsymptoms or by general physical health. To the extent that menopause isdefined as a medical problem, rather than a normative feature of aging,women may attribute changes in their overall level of physical health tomenopause. However, to the extent that menopause is perceived as adistinct set of challenges and changes, women might be less likely toattribute changes in physical health to menopause. Our understanding ofthe different factors affecting adjustment to loss of fertility andhealth concerns may be increased by considering these dual aspects ofadjustment to menopause in the context of concurrent physical andpsychological symptoms. The findings of the present study provide some guidance for socialwork interventions among women in this generation. The primaryimplication of this work is that women facing issues around menopauseare not a homogeneous population. Women may face difficulties concerningthe loss of fertility, or health and attractiveness, or both. For womenwho are facing issues centered on the loss of fertility, exploration ofthe roles that these women occupy and efforts to increase the variety,meaning, and value of these multiple roles may be a fruitful directionto pursue for social work practitioners. However, for women who are primarily concerned with the impact ofmenopause on their health and attractiveness, an emphasis on roleenhancement may not be as pertinent. Adjustment to menopause for thesewomen is more closely related to their unique experience of the changesto their bodies and body images as well as reactions to commonmenopausal symptoms, such as hot flashes and the impact on their health.Hence, social work interventions and social work advocacy might befocused on and include education to better inform younger,perimenopausal women about the emerging stage of menopause to alleviatefears often seen in younger women in regard to menopause. Socialservices that offer women information on safe and available remedies toprevent weight gain or body changes or methods to reduce commonmenopausal symptoms, such as hot flashes, would be more relevant to theneeds of women who are concerned more with attractiveness and health. It should be noted that the assessment of multiple roles in thepresent study is limited by the information that is available in theMIDUS data set. The analysis focused on the number of roles that a womanoccupied, rather than the quality of her role as a paid employee,spouse, parent, caregiver, or community member. Current research on roleenhancement theory suggests that the quality, more than the mere number,of roles is critical to a woman's well-being. Thus, the impact ofmultiple roles might have been underestimated in the present work. Inaddition, the definition of marriage does not include long-termrelationships that fell outside of the prevalent legal definition ofmarriage when the data were collected (for example, same-sexrelationships) that may confer the protective effects posited by roleenhancement theory. In the United States today, public policy and negative attitudestoward aging have not fostered the development of specialized health andsocial services geared to the needs of the population discussed in thisarticle. Public support for such services as education, early diseasedetection, and intervention in emerging health problems during themenopausal transition may prevent or reduce the onset and severity ofhealth problems, such as bone breakages from osteoporosis, or thebeginning of other diseases in later life that become more common duringand after menopause. Addressing the needs of menopausal baby boomers islikely to have radiating effects across the family and benefit membersof the older and younger generation, by enhancing women's roles toprovide better care for children and elderly people. The findings of the present research have potential benefits forsocial workers who work in a variety of settings. Illustratively, socialworkers who practice in the workplace (for example, in employeeassistance programs) are working "with women who are in paidemployment, and would benefit from an understanding of the ways in whichroles inside and outside of the workplace assist women in theiradjustment to the loss of fertility. Social workers located withinprimary care facilities may be working with midlife women who areundergoing the menopausal transition, or they may be working with anolder population whose primary caretakers are daughters who areundergoing menopause while still caring for an elderly parent. In eithercase, understanding the differentiated needs of women during menopausecan guide efforts to increase resources and competencies for coping withboth the medical and the social implications of this major lifetransition. Original manuscript received June 22, 2010 Final revision received October 5, 2010 Accepted November 9, 2010 REFERENCES Avis, N. E., & McKinlay, S. M. (1991). A longitudinal analysisof women's attitudes to menopause: Results from the MassachusettsWomen's Health Study. Maturitas, 13, 65-79. Barnett, R. C., & Hyde, J. S. (2001).Women, men, work, andfamily: An expansionist theory. American Psychologist, 56, 781-796. Chrouser-Aherens, C.J., & Ryff., C. D. (2006). Multiple rolesand well-being: Socio-demographic and psychological moderators. SexRoles, 55, 801-815. Crosby, F. J. (1991). Juggling: The unexpected advantage ofbalancing career and home for women and their families. New York: FreePress. Dege, K., & Gretzinger, J. (1982). Attitudes of familiestowards menopause. In A. M. Voda, M. Dennerstien, & S. R.O'Donnell (Eds.), Changing perspectives on menopause (pp. 60-69).Austin: University of Texas Press. Eisner, H., & Kelly, L. (1980, November 22). Attitudes of womentoward the menopause, Paper presented at the annual meetings of theAmerican Gerontological Society, San Diego. Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fitindices in covariance structure analysis: Conventional criteria versusnew alternatives. Structual Equation Modeling, 6, 1-55. Hunter, M. S. (1990). Psychological and somatic experience of themenopause: A prospective study [corrected]. Psychosomatic Medicine, 52,357-367. Hunter, M. S., Battersey, R., & Whitehead, M. (1986).Relationships between psychological symptoms, somatic complaints andmenopausal status. Maturitas, 8, 217-228. Jaszmann, L., Van Lith, N. D., & Zatt, J.C.A. (1969). Theperimenopausal symptoms: The statistical analysis of a survey. MedicalGynaecology and Sociology, 4, 268-277. Kresovich, E.A.S. (1980). A comparison of attitudes towardmenopause held by women during three phases of the climacterium. Issuesin Mental Health Nursing, 2(3), 59-69. Leiblum, S. R., & Swartzman, L. C. (1986). Women'sattitudes toward the menopause: An update. Maturitas, 8, 47-56. McKinlay, J. B., McKinlay, S. M., & Brambilla, D, (1987). Therelative contributions of endocrine changes and social circumstances todepression in mid-aged women. Journal of Health and Social Behavior, 28,345-363. Neugarten, B.,Wood, V., Kraines, R., & Loomis, B. (1963).Women's attitudes toward the menopause. International Journal ofHuman Development, 6, 140-151. Perlmutter, E., & Bart, P. (1982). Changing views of thechange. In A. Coda, M. Dinnerstein, & S.A. O'Donnell (Eds.),Changing perspectives on menopause (pp. 187-199). Austin: University ofTexas Press. Polit, D. F., & LaRocco, S.A. (1980). Social and psychologicalcorrelates of menopausal symptoms. Psychosomatic Medicine, 42, 335-345. Prill, H.J. (1977). A study of the socio-medical relationship atthe climacteric in 2232 women. Current Medical Research and Opinion, 4,4. Rossi, A. (2004). Menopause and sexuality. In O. G. Brim, C. D.Ryff, & R. Kessler (Eds.), How healthy are we? A national study ofwell-being at midlife. Chicago: University of Chicago. Thoits, P.A. (1983). Multiple identities and psychologicalwell-being: A reformulation and test of the social isolation hypothesis.American Sociological Review, 48, 174-187. Wilbur, J, Miller, A, & Montgomery, A. (1995).The influence ofdemographic characteristics, menopausal status, and symptoms onwomen's attitudes toward menopause. Women and Health, 23(3), 19-39. Judy R. Strauss, PhD, LMSW, is area chair of social sciences,University of Phoenix; e-mail: jrstr@email.phoenix.edu.Table 1: Sample Characteristics Wave 1 Wave 2Characteristic * M SD M SDAge (years) 39.92 5.53 48.87 5.48Number of roles 2.37 1.01 2.16 0.96Symptoms (b) 1.90 0.87 2.58 1.04Physical concerns (c) 2.69 0.90 2.60 0.91Attractiveness (c) 2.91 0.92 2.95 0.95Fertility (c) 3.56 0.84 3.80 0.62Characteristic n % n %Education High school graduate 354 34.2 314 30.3 or less Some college 314 30.3 316 30.5 Four- or five-year 203 19.6 204 19.7 college graduate Postcollege graduate 164 15.8 202 19.5 Total 1,035 100.0 1,036 100.0Characteristic n % n %Financial comfort More money than 144 14.0 241 23.4 needed Just enough money 573 55.5 543 52.6 Not enough money 315 30.5 256 24.0 Total 1,032 100.0 1,040 100.0(a) Sample size: Wave 1 n = 1,010; wave 2 n = 1,009.(b) Scale for symptoms: 1 = never, 2 = once a month, 3 = severaltimes a month, 4 = once a week, 5 = several times a week, 7 =almost every day.(c) Scale for Health, Attractiveness, and Fertility concerns: 1 =a lot, 2 = some, 3 = a little, 4 = not at all.Table 2: Model Parameters forTwo-Factor Model: Wave 1Path BAntecendants to latent endogenous variables Roles [right arrow] Childbearing .203 *** Education [right arrow] Childbearing -.159 ** Age [right arrow] Childbearing .378 *** Financial security [right arrow] Childbearing .085 * Symptoms Health/wellness -.264 ***Latent endogenous variables to indicators Childbearing [right arrow] Fertility .915 *** Childbearing [right arrow] Relief .229 *** Health/wellness [right arrow] Attractiveness .723 *** Health/wellness [right arrow] Physical health .701 ***Latent endogenous variables Childbearing [right arrow] Health/wellness .224 **** p<.05. ** p< .01. *** p < .001.Table 3: Model Parameters forTwo-Factor Model: Wave 2Path B BAntecedents to latent endogenous variable Roles [right arrow] Childbearing .156 *** Education [right arrow] Childbearing -.173 Age [right arrow] Childbearing .339 *** Financial security [right arrow] Childbearing .140 ** Symptoms [right arrow] Health/wellness .334 ***Latent endogenous variables to indicators Childbearing [right arrow] Fertility .679 *** Childbearing [right arrow] Relief .408 *** Health/wellness [right arrow] Attractiveness .845 *** Health/wellness [right arrow] Physical health .590 ***Latent endogenous variables Childbearing [right arrow] Health/wellness .384 ***** p<.01. *** p < .001.