research paper

Women, Work and Stress: A Review and Agenda for the Future Author(s): Mary Ann Haw Source: Journal of Health and Social Behavior, Vol. 23, No. 2 (Jun., 1982), pp. 132-144 Published by: American Sociological Association Stable URL: https://www.jstor.org/stable/2136510 Accessed: 25-09-2018 00:28 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide

range of content in a trusted digital archive. We use information technology and tools to increase productivity and

facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

https://about.jstor.org/terms

American Sociological Association is collaborating with JSTOR to digitize, preserve and extend access to Journal of Health and Social Behavior

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

Women, Work and Stress: A Review and Agenda for the Future

MARY ANN HAW

California Nurses Association

Journal of Health and Social Behavior 1982, Vol. 23 (June):132-144

A review olfthe literature on women and work-related stress was conducted to identify conclu- sions regarding the link betw een job conditions conducive to stress and disease outcomes, and to suggest directions fr jfiture research. Defined as an imbalance between perceived demand and pe)-ceived capability, stress is viewed as an intervening variable between conditions c(onducil'e to stress, and responses and the more enduring disease outcomes. Research findings tire inconclusilve but suggest that work may have a beneficial effect on mental health, and that certain tvpes of-jobs in combination with family responsibilities may lead to increased risk or actual development of 'cardiovascular disease. However, studies on women lacked specificity on Work environment and onfamily responsibilitieslattitudes. Future research on women should involve (I) longitudinal studies before, during and after cessation of employment, (2) specificity about job environment and family responsibilities, (3) length and continuity of exposure to potentially stressful conditions, and (4) individual perceptions and coping responses.

The past 30 years have witnessed a dramatic change in the participation of women in the workforce. During this period the number of women workers in the United States has more than doubled (U.S. Department of Labor, 1975). Women's work participation rates have risen from 33% in 1950 to 53% in 1975 (U.S. Department of Labor, 1977). Since 1965, changes in the rate of participation have been most accentuated among women in their twen- ties and early thirties, and especially among mothers of young children. Today, over 37% of women in the workforce have children under the age of six, as compared to 29% in 1969 (U.S. Department of Labor, 1977).

Coupled with a quantitative change in the workforce participation of women has been a qualitative change. Women have moved up the occupational hierarchy, assuming jobs with higher status and greater responsibility. In ad- dition, increasing numbers of women are en- tering nontraditional jobs, widening the scope of occupations in which women are employed. In 1950, women accounted for 13.8% of per- sons occupying managerial/administrative po- sitions. By 1976, women occupied 20.8% of these positions. In the professions, the number of women lawyers and physicians roughly

Address communications to: Mary Ann Haw, R.N., Ph.D., Consultant, California Nurses Asso- ciation Region XI, Emeryville, CA 94608.

doubled between 1950 and 1976 (U.S. Depart- ment of Labor, 1977).

Much research over the course of the last 20 years suggests that work may be a significant source of stress, and that stress may be tied to serious consequences in regard to mental and physical ill health (Cooper and Marshall, 1976; House, 1974; Jenkins, 1971a, b; Kahn et al., 1964; Kasl, 1978; Margolis et al., 1974). Fur- thermore, occupational mobility has been as- sociated with cardiovascular disease (Jenkins, 1971a, b; Syme and Reeder, 1967), although the reported findings have not always been consistent (Hinkle et al., 1968), and contro- versy exists over the adequacy of the data analysis in these studies (Horan and Gray, 1974). However, a point that must be empha- sized is that, until recently, studies on work- related stress either excluded women as sub- jects or did not analyze sex differences.

In addition to the inattention to women and work-related stress, there are a number of other compelling reasons for reviewing the av- ailable evidence and for pursuing this area of research:

(1) It may provide bases on which to gener- alize morel broadly the findings of previous re- search on work-related stress. Evidence that women respond to this source of stress simi- larly to men would extend the external validity of previous studies.

(2) Conversely, evidence that a particular

132

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

WOMEN, WORK AND STRESS 133

finding does not generalize across sex may suggest that there are important individual or social environmental variables at play; for in- stance, there is some evidence that women are less likely to respond to emotionally arousing stimuli by the release of epinephrine (Gray, 1971).

(3) As increasing numbers of women enter the workforce, the more favorable mortality rates for women may show a change. Women may have enjoyed a more favorable mortality rate, in part, because of their having a limited exposure to a noxious environment at work (Waldron, 1976).

(4) As women gain increasing occupational mobility, they not only may be exposed to the same physical and emotional hazards of the work environment as men, but also may be exposed to the pressures created by multiple role demands and conflicting expectations. For example, the burdens of housework and child care continue to fall more heavily on women than on men, regardless of employment status (Hedges and Barnett, 1972; Vanek, 1974).

(5) Women, because of job segregation, may be exposed to different work hazards than are men. Despite evidence of an increased occu- pational mobility for women, there has been a continued concentration of women in a rela- tively small number of areas that have traditionally been considered women's fields

(Stellman, 1978). In 1973, more than 40% of all women workers were concentrated in 10 occu- pations, such as secretary, waitress, nurse, and school teacher (U.S. Department of Labor, 1975).

A MODEL FOR REVIEWING THE LITERATURE

In this model for reviewing the literature, stress is viewed as a complex interrelationship among a number of variables, rather than as a unitary concept. The model encompasses both the Cox and Mackay transactional model of stress

(Cox, 1978) and the House paradigm of stress research (1974). House identified five classes of variables necessary for a comprehensive paradigm of stress research which are included in the present model (Figure 1): (1) objective social conditions conducive to stress; (2) indi- vidual perceptions of stress; (3) individual re- sponses to perceived stress; (4) more enduring outcomes of perceived stress, such as mental ill health and cardiovascular disease; and (5) individual and situational conditioning vari- ables that specify the relationships among the four sets of factors.

Embodying the Cox and Mackay trans- actional model, the present model views stress as an intervening variable between con- ditions conducive to stress or a potentially

FIGURE 1. Model of stress for reviewing literature. The solid arrows between the boxes indicate hypothesized causal relationships; the dotted arrows indicate that social or individual variables condition or specify the nature of the relationships

5. Individual or situational conditioning variables, e.g., commitment to work,

childcare and housekeeping responsibilities outside of work.

2. Perceived stress-- 3. Stress Responses--

imbalance between per- physiologic, cognitive, ceived demand and per- ? affective, behavioral,

ceived capability, e.g. , e.g., BP, cholesterol perceived work load in (physiologic); denial relation to perceived (cognitive); depression, job skill to accomplish anger (affective); smok-

4,, it. efenses ing (behavioral).

1. Conditions conducive to stress--actual work demands, actual capabili- cO9 4. Outcomes--physio- ties of the individual, logic, affective, i.e., objective work load behavioral, e.g., men- objective job skill. tal illness, cardio-

vascular disease.

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

134 JOURNAL OF HEALTH AND SOCIAL BEHAVIOR

disturbing environment (box 1) and an individ-

ual's response to the environment (box 3).

Stress is defined in terms of an imbalance be- tween the perceived demand and the person's

perception of his or her ability to meet that

demand (box 2). An imbalance between per-

ceived demand versus perceived capability,

when coping is important, may generate the

experience of stress and stress responses

(box 3). Stress may be generated when the perceived demand exceeds perceived capabil- ity, as well as when the demand is perceived to

fall short of that capability. Responses to stress are both physiological and psychological, the

latter involving cognitive, affective, or behav-

ioral responses, or a combination of these.

These responses are attempts at directly alter-

ing the source of stress (coping), altering the

perception of it (defenses), or both. If coping techniques or defense mechanisms are ineffec-

tive, stress is prolonged, which may lead to abnormal responses such as elevated blood

pressure, smoking, or psychological depres-

sion. The occurrence of these abnormal re-

sponses in conjunction with prolonged expo-

sure to stress may lead to permanent functional and structural damage (box 4), such as chronic

mental illness or cardiovascular disease. The

development of a consequent disease outcome

may, in turn, influence the perception of stress

on the job (box 2), emphasizing the feedback

element in the model.

However, not all individuals perceive a par- ticular work situation as stressful, nor do those who perceive it as stressful react with similar responses or with the same type and degree of outcomes, such as physical disease or mental illness. Important individual or situational variables (box 5) may condition or specify the nature of the relationships in the model. For instance, individuals with a high commitment to work may find underutilization of their skills on the job frustrating and stressful, whereas individuals with a low commitment to work

may find it tolerable. Situationally, women with major family responsibilities and heavy role demands at home may find a particular workload on the job overwhelming, whereas their unmarried counterparts may find it chal- lenging. With regard to responses, some work- ers may react to an overwhelming workload by delegating work and making other direct at- tempts to reduce it, whereas others may handle

the same situation by working at a more rapid pace and putting in overtime hours. Finally, some individuals subjected to stress over long periods of time may develop rheumatoid ar- thritis and others, cardiovascular disease. Out- come may be influenced not only by psycho- logical predispositions and coping techniques, but also by genetic and possibly other physio- logical predispositions to disease.

REVIEW OF RELEVANT STUDIES

A computer search of all the relevant medi- cal and social science journals was conducted for studies that concerned women, employ- ment, and work-related stress. In addition, studies were included that concerned women, employment, and variables that have been as- sociated with work-related stress in previous research primarily involving male respondents. This broad-based search strategy was adopted because of the different classes of variables that must be considered in a model of stress research. Evidence was sought regarding all links in the model. However, some links es- sentially were not addressed in the studies, most notably the first link between conditions conducive to stress and perceived stress. Most of the studies reviewed concerned employment or some facet of the work environment among women in relation to responses and outcomes that have been associated with work-related stress, primarily among male respondents in previous research, such as psychological symptoms, cardiovascular risk factors, mental illness, and cardiovascular disease. Although work-related stress was rarely a major focus in the studies surveyed, it was often presumed or implied. In addition, there was a sizable number of studies concerning situational vari- ables that primarily were presented under the rubric of family responsibilities and prolifera. tion of role demands for the working woman. Because of the gaps in the literature on women and work-related stress, this review is orga- nized in the following fashion:

* Conditions conducive to stress and responses/outcomes

* Perceived stress and responses/outcomes * The link between responses and outcomes * Situational variables (primarily family re-

sponsibilities) * Male-female comparisons

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

WOMEN, WORK AND STRESS 135

The purpose of this review is to identify con- clusions that may be drawn about work-related stress and women, and to provide a basis on which to broaden existing literature on work- related stress. Furthermore, the purpose is to suggest the direction of future research on women and work-related stress.

In this review, the word "work" used in re- lation to women refers to work or employment outside the home. This is not meant to imply, however, that women who remain at home do not work.

GENERAL OBSERVATIONS

A number of general observations can be made about studies of work-related stress and women in comparison to the literature on job stress in general. First, there are many fewer studies on work-related stress concerning women: for every study concerning women and work-related stress in the last decade, there have been roughly six concerning men.

Second, the studies on work-related stress and women tend to concern less specific fac- tors of the job environment than do the studies concerning men. Many of the studies on women concern employment/occupational status or rough measures of workload in rela- tion to a number of dependent variables, the implication being that employment or oc- cupation per se is stressful for women. These studies on men, however, tend to concern more specific variables on the job, such as work overload, underutilization of skills, and role conflict.

Third, studies on women often span both work and home-related roles. The literature on job stress and men rarely concerns the overlap between work and family. Although this inat- tention may highlight a fertile area for investi- gation among men, it suggests differential role demands for the working woman vis-h-vis the working man.

Finally, over one-third of the studies re- viewed concern, in some way, the proliferation of role demands for the working woman. In addition, a number of the studies point to the more negative attitudes toward work and lower job satisfaction among women in comparison to men. Yet few of these studies link multiple role demands or work attitudes to the more

enduring outcome variables, such as car- diovascular disease or mental illness.

Similar to the weaknesses in studies of work-related stress on men, two general limi- tations may be found in the literature on women. With one notable exception, the studies are primarily cross-sectional in design and correlational in analysis. As with most studies of this type, the sequence of events frequently is indeterminant. Which variable precedes another in time cannot be inferred with any degree of confidence. In addition, these studies generally were conducted on small or non-representative groups, limiting the ability to generalize the findings to larger groups of women.

CONDITIONS CONDUCIVE TO STRESS AND RESPON SES/O UTCOM ES

Employment and Physiological Response

The only physiological response to work- related stress addressed in the literature on women was blood pressure. Two studies-the Framingham study (Haynes and Feinleib, 1980), and one among black and white women living in Detroit (Hauenstein et al., 1977)- addressed the link between employment and blood pressure levels. In both studies, working women and housewives showed no differences in blood pressure levels. Although Hauenstein et al. found that currently unemployed women (those looking for work) had significantly lower blood pressure than either working women or housewives (p < .025 and .05, respectively), this finding was observed only among women living in high-stress neighborhoods (neighbor- hood stress was defined in terms of such vari- ables as poverty, crime rate, and family insta- bility).

Findings from both studies suggest that job instability may be related to blood pressure for some groups of women. Among black women in the study of Hauenstein et al., job instability (job and line of work changes) was positively correlated with blood pressure (p < .05) and negatively correlated among white women, although the correlation did not reach statisti- cal significance. Among white-collar women in the Framingham study (racial composition not specified), job instability (line of work changes)

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

136 JOURNAL OF HEALTH AND SOCIAL BEHAVIOR

was negatively correlated with blood pressure (p <.05).

Although employment per se was unrelated to blood pressure levels, additional findings in the study of Hauenstein et al. lead to the speculation that individual or situational vari- ables may moderate the relationship between employment and blood pressure levels. Work- ing women over 40 years of age with a strong commitment to work (as measured by working by choice versus working reluctantly) evi- denced higher blood pressure than did reluc- tant workers (p < .01). In contrast, both job dissatisfaction and "not having done well on the job" were associated with higher blood pressure among reluctant workers but not among the women working by choice.

Overall there is meager evidence linking conditions conducive to stress and physiologi- cal response among women. In addition to the paucity of studies in this area, limitations in the two studies available preclude any conclu- sions. The Framingham study did not discrimi- nate between full-time or part-time workers or those currently employed or temporarily un- employed at the time the physiologic indica- tions were taken. Although Hauenstein et al. found no relationship between workload and blood pressure (workload as measured by full-time or part-time work in conjunction with number of children and number of hours spent on housework), workload on the job was not investigated.

Employment and Behavioral Response

There are several studies linking employ- ment or occupational status with Type A be- havior (Haynes and Feinleib, 1980; Shekelle et al., 1976; Waldron, 1978). In previous pros- pective studies, Type A behavior has shown an independent association with coronary heart disease among both men and women, even when other standard coronary risk factors were taken into account (Haynes et al., 1978b; Rosenman et al., 1964). Because it is difficult to ascertain whether Type A behavior is an effect of work, a personality variable, or a re- flection of a combination of these two factors, the following findings should be interpreted with caution in regard to cause and effect. High occupational status has been associated with

Type A behavior among men and women (Shekelle et al., 1976; Waldron, 1978). Fur- thermore, higher Type A scores were found among working women, as compared with housewives (Haynes and Feinleib, 1980), and among full-time working women as compared to part-time workers (Waldron, 1978). How- ever, both Type A working wives and Type A housewives had similar rates of coronary heart disease (CHD) (Haynes and Feinleib, 1980).

Employment and Affective Response

Generally, the evidence suggests that em- ployed married women fare better emotionally than do housewives on a number of affective indicators. Employed married women have greater life satisfaction (Rose, 1955), show greater self-acceptance (Feld, 1963), and have fewer psychiatric symptoms (Gove and Geer- ken, 1977) than do housewives. Radloff (1975), controlling for happiness with job and mar- riage, found that housewives were significantly more depressed than were working wives. With regard to symptoms of stress, however, working women reported more daily stress than did housewives (Haynes and Feinleib, 1980).

Although employed wives generally fared better emotionally than did housewives, the comparisons with working men along affective indicators were not as favorable. One study found that working women experienced greater physical and emotional distress than did the men (Cohen, 1976), and another showed that

they were nearly twice as likely as men to express negative attitudes toward their work (Work in America, 1973). However, these two studies reported general differences in re- sponse to work between men and women who

probably differed on a number of variables, such as occupational status, salary, and mobil- ity, without examining the effect of these vari- ables. Gordon and Strober (1978), comparing men and women at similar occupational levels, found that women reported more symptoms of stress than men, such as feeling depressed, having nightmares, feeling overwhelmed, and experiencing stomach distress.

Few conclusions can be drawn from these studies on employment among women in rela- tion to affective responses. First, the generally

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

WOMEN, WORK AND STRESS 137

more favorable affective response among em- ployed wives as compared to housewives may reflect the healthy worker effect. The criterion variables may have been those that contributed to seeking and maintaining employment, rather than a result of employment. Second, the more negative response among working women as compared to working men may reflect the gen- eral tendency for women to show a more nega- tive affective response than do men. Previous studies have shown that women generally are more depressed than men and report more psychiatric, as well as physical, symptoms (Nathanson, 1975). However, the fact that working women report more daily stress than nonworking women (Haynes and Feinleib, 1980) suggests that work itself, above and be- yond gender tendencies, contributes to symp- toms of stress.

Employment and Disease Outcomes

Although the Framingham study (Haynes and Feinleib, 1980) was the only, one in the literature addressing the link between em- ployment and disease outcomes, its prospec- tive design adds strength to the findings. A psy- chological questionnaire was administered to a subsample of the Framingham cohort (350 housewives, 387 working women, and 580 men) at their eighth or ninth biennial medical examination. The respondents were followed for the development of coronary heart disease over the next 8 years.

Working women did not have a significantly higher incidence of CHD than did the house-

wives (7.8% and 5.4%, respectively). In the analysis of occupational categories in relation to CHD, white-collar working women were di- vided into two categories: (1) clerical workers

(e.g., secretaries, stenographers, bookkeepers, bank clerks), and (2) white-collar professionals (e.g., teachers, nurses, librarians). Clerical workers were found to be twice as likely to have CHD, as compared to housewives (10.6% and 5.4%, respectively). No such excess risk was found among other categories of working women (white-collar professionals and blue- collar workers). Of interest was that the in-

creased risk among clerical workers occurred only among those with significant family re- sponsibilities (Haynes and Feinleib, 1980).

This finding will be discussed in more detail in a later section (see Situational Variables).

Examining the specific facets of the job

among clerical workers, decreased job mobility (fewer changes of job and line of work) was associated with higher rates of CHD (p < .001), as was having a nonsupportive boss (p < .001).

None of the standard coronary risk factors (age, blood pressure, serum cholesterol, or cigarette smoking) was associated with CHD among these clerical workers (Haynes and Feinleib, 1980).

Perceived Stress and

ResponseslOutcomes

The links among perceived stress, re- sponses, and outcomes were sparsely ad- dressed in the literature. Only two studies have been reported, and serious limitations in each preclude conclusive statements.

In a cross-sectional study of 799 Australian

workers (including male and female high

school teachers, factory and clerical workers, and men in managerial and other high level positions in industry), lack of need-value at- tainment (the discrepancy between a person's perceived needs, cultural values, and per- ceived attainment at work and in life in general) was associated with self-reports of symptom awareness and visits to the doctor (p < .01 and

.001, respectively) (Otto, 1979).

It is noteworthy that semi-skilled women in Otto's study (1979) scored the lowest in need-value attainment, as compared with other men and women, including women cleri- cal workers. Although the Framingham study (Haynes and Feinleib, 1980) did not compare need-value attainment on perceived stress among the various occupational categories of employed women, clerical workers, not semi- skilled workers, were the most disadvantaged with regard to subsequent CHD.

A retrospective study in Sweden comparing women with ischemic heart disease (IHD) to those free of IHD provides some evidence re- garding the link between perceived stress and outcomes (Bengtsson, 1973). Nearly two-thirds of the 1,642 women in the study were em- ployed or had been employed during most of their adult years. IHD was defined as having one of the following: (1) myocardial infarction

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

138 JOURNAL OF HEALTH AND SOCIAL BEHAVIOR

(MI), (2) angina pectoris (AP), or (3) coronary

electrocardiogram (EKG). Respondents were interviewed regarding psychosocial stress

factors (similar to the Holmes and Rahe (1967) "life-styles" inventory) occurring in the year preceding the study or MI attack. The number

of stressors tended to be larger in the MI group than in the reference group (p < .05), although this finding did not reach statistical significance in the AP group. The respondents' subjective feelings of stress also were obtained in inter-

view by asking them if, during the previous

year, they had a "feeling of stress for a month or longer, including tension, fear, anxiety or

sleep disturbances in connection with conflicts in the family or at work." Severe stress was

defined as "a continuous feeling of stress dur- ing the year preceding the study or the MI."

Subjective stress was more often reported among women with MI and AP than in the reference group (p < .001 and .01, re-

spectively). "Severe stress" was also signifi- cantly more common in the MI and AP group. However, women with a "coronary EKG"

(without symptoms, but having coronary dis-

ease based on an EKG) did not differ signifi- cantly from women in the reference group. The latter finding illustrates the weakness of re- trospective studies, and could suggest that the presence of symptoms (AP) or actual disease

(MI) may influence the perception of stress, rather than vice-versa.

The Link between Responses

and Outcomes

Over a period of time, abnormal physio- logical, affective, and behavioral responses to work-related stress may lead to outcomes in- volving permanent structural and functional damage. Previous research on cardiovascular

disease indicates that responses such as ele- vated blood pressure and cholesterol levels,

smoking, and Type A behavior are indepen- dent predictors of CHD (Haynes et al., 1980; Rosenman et al., 1964).

Among working women of ages 45 to 64 who

participated in the Framingham study (Haynes et al., 1980), Type A behavior and reactions-( I) suppressing anger, (2) taking anger out on others, (3) discussing anger, and (4) physiologic reactions to anger such as

headaches, tension-were significant predic-

tors of CHD. Systolic blood pressure signifi-

cantly predicted CHD among all women 45 to

64 years of age. However, only Type A be-

havior and suppressed hostility remained inde-

pendent predictors among working women in

multivariate analysis (Haynes et al., 1980). The

exact mechanism by which suppressed hostil-

ity leads to CHD is unknown. However, anger

symptoms and discussing anger were corre-

lated with diastolic blood pressure in white-

collar women younger than 65 years old (r =

.12 and .14, respectively), and taking out anger

on others was negatively correlated with

cholesterol levels (r = .14) (Haynes et al., 1978a).

Although the Framingham study provides

the only evidence available regarding the link

between responses and outcomes, its pro-

spective nature strengthens the inferences that

may be drawn about antecedent-consequent

relationships.

Situational Variables: Family Responsibilities

Although there are a number of situational

variables that might moderate the relationship

between job conditions conducive to stress and

both responses and outcomes (such as social

support on the job and economic adversity in the family situation), family responsibilities were the only variables given significant atten- tion in the literature reviewed.

A number of studies document the increased

overall workload for the employed married woman vis-h-vis the employed married man.

Employed married women have 17% less free time (Szalai, 1972) and more often report in- sufficient time for rest and recreation than em-

ployed married men (Otto, 1979). These women spend on the average 5 more hours during the week on paid work, commuting, housework, and family tasks than men (Hedges

and Barnett, 1972). Among men and women with an MBA degree who were employed full- time, the women with families were more likely to assume responsibility for household man- agement and child care than their male coun- terparts, and more often than men reported worrying about household responsibilities while at work (Gordon and Strober, 1978).

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

WOMEN, WORK AND STRESS 139

Johnson and Johnson (1977) interviewed 28

dual-career families with young children and

found that wives but not husbands experienced

strain between their work and home roles. In a

study of physicians, one-third of the doctors

who were women and none of those who were

men reported that marriage and family respon-

sibilities provided the impetus to change career

directions (Nadelson et al., 1979). Relatedly, a

study of women in the professions (law,

medicine, and college teaching) found that over

half of them coped with conflict between pa-

rental and work roles by temporarily lowering

their career ambitions and perceived that their

professional involvement had less priority than

their husbands' involvement (Poloma, 1972).

Turning to studies that link multiple role de-

mands to mental health outcomes, we find that

the results are inconsistent. Gove and Geerken (1977) found that psychiatric symptoms in-

creased monotonically with an increase in number of children among employed women

but not among employed men or -unemployed

women. Radloff (1975) found working wives

more depressed than working husbands, but

found no significant relationship between

amount of housework and depression. In a

study of 144 married women in North Carolina,

Woods (1978) found that women's involvement

in multiple roles did not have a deleterious

effect on their mental health, except wfien they

received little support from significant others

and when' they evaluated their role perfor- mance negatively. However, the cross-

sectional nature of this study makes it impossi-

ble to determine the antecedent-consequent relationship.

With regard to role demands in relationship

to pathophysiological responses and outcomes among women, Hauenstein et al. (1977) found

no relationship between number of children or

number of hours spent on housework and

blood pressure levels among employed

women. Although housewives who reported tension about housework and being critical of

their own performance had higher blood pres- sure, no such relationship was found among working women, for whom the outside job "presumably reduces the centrality of house- work."

In the Framingham study (Haynes and

Feinleib, 1980), women who worked outside

the home with three or more children were

more likely to develop CHD than working women who had no children (11% versus 6.5%),

and were two and one-half times as likely to

develop CHD as were housewives with the same number of children (4.4%). Similarly, in the Bengtsson study (1973), in which nearly two-thirds of the sample population was em-

ployed, significantly more women in the MI

group than in the reference group had four or

more children.

Returning to the Framingham study, the ex-

cess risk of CHD among working women was

confined to clerical workers and only those clerical workers with children. Clerical work- ers who were single or married without chil- dren were at no greater risk than were other workers. Moreover, clerical workers with chil- dren who had blue-collar husbands were over three times as likely to develop CHD than non-clerical working mothers (21.3% and 6%, respectively, p = .004). No such excess risk was observed among clerical workers married to white-collar workers, suggesting that "cer-

tain life style behaviors and attitudes" (not measured in the study) may contribute to the excess risk among those clerical workers mar-

ried to blue-collar husbands. Perhaps blue-

collar husbands were less likely to help with child care and household tasks than other hus- bands. Economic necessity may have also

played a part; i.e., the element of having to work (the reluctant worker). In the study of Hauenstein et al. (1977), the reluctant worker

who was dissatisfied with being unable to use her best skills on the job had significantly higher blood pressure levels than did workers by choice. In addition, women married to blue-collar workers may have been more likely to have worked full-time, with fewer interrup- tions in employment, out of a need for income than had other working women, and therefore may have had greater exposure to the pres- sures created by work and home roles. How- ever, the Framingham study did not measure the amount of exposure to work-related stress

among working women in terms of part-time versus full-time work, and uninterrupted ver-

sus segmented employment. In summary, the results are mixed regarding

family responsibilities as situational variables moderating relationships among job conditions conducive to stress and responses/illness out- comes. However, the Framingham study

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

140 JOURNAL OF HEALTH AND SOCIAL BEHAVIOR

(Haynes and Feinleib, 1980) provides the most

convincing evidence that the combination of

certain job conditions and family respon-

sibilities may lead to structural and functional

damage.

Several problems in conceptualization and

methodology may account for the inconsis-

tency in findings among the studies regarding

family responsibilities. The measure of the amount of housework in the Radloff study

(1975) may have been too imprecise. Respon-

dents were asked to estimate how often they

had worked around the house and yard in the

past week (the four choices ranged from none

to more than once a day). In the study of

Hauenstein et al. (1977), respondents indicated

the number of hours each day they spent on

housework, but were not questioned on per-

ceived overload. Similarly, Woods's study (1978) examined the proliferation of roles

among women but did not include a measure of

perceived overload. In previous studies on work-related stress with male respondents,

perceived overload showed stronger relation- ships with dependent physiological and psy-

chological measures than did objective over-

load (French and Caplan, 1973; Modigliani,

1966). Finally, none of the studies examined

the "executive" aspect of housework and child

care responsibilities. It is possible that having primary (executive) responsibility for house- work and child care may be an even more

important variable in the genesis of stress than

is the amount of time spent doing the related

tasks.

MALE-FEMALE COMPARISONS

The review of the relevant studies on women

yielded little to either increase the gener-

alizations that can be drawn from previous re- search on work-related stress (primarily in- volving male respondents) or illuminate major

differences between men and women. First of

all, the relative lack of studies involving female

respondents limited the possible comparisons. Second, studies on women, for the most part, concerned different variables from those in the studies on men. In many instances, any com- parative statements to be made about men and women in regard to work-related stress must

be based at this point on one or two studies for

each class of variables. Therefore, the follow-

ing discussion of similarities and differences

between men and women should be regarded as speculative.

Similarities

Similarities among men and women primar-

ily involved several relationships between re-

sponses to stress and the development of car-

diovascular disease. For both men and women

in the Framingham study (Haynes et al., 1980),

suppressed hostility and Type A behavior were

independent predictors of CHD. This latter

finding is consistent with the results of previ-

ous studies (Kenigsberg et al., 1974; Rosenman

and Friedman, 1961). Similarly to women in

the Bengtsson study (1973), men with MI's

demonstrated higher levels of life changes in

the year before the MI than did the comparison group (Theorell, 1973), although the re-

trospective nature of both of these studies

weakens this finding.

Differences

Differences in regard to work-related stress between men and women primarily concern responses and outcomes in relation to condi- tions conducive to stress. With regard to

blood-pressure response, married working- class men anticipating job loss had elevated blood pressure levels that remained high during the ensuing unemployment period (Kasl and Cobb, 1970), in contrast to unemployed women (those looking for work), who had significantly lower blood pressure than either working women or housewives (Hauenstein, et al., 1977). As previously noted, however, this finding held only for women living in high- stress neighborhoods. Furthermore, job insta- bility (line of work and job changes) was unre- lated to blood pressure levels among men,

contrasting with findings among women in the same study (Haynes et al., 1978a) and in the study of Hauenstein et al. (1977).

Turning to cardiovascular disease outcomes, the review showed that the relationship be- tween occupational work categories and CHD varied between men and women in the Framing- ham study (Haynes and Feinleib, 1980). Un-

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

WOMEN, WORK AND STRESS 141

like women, among whom the highest rates of

CHD were found among clerical workers, the

highest rates of CHD among men were ob-

served among white-collar professionals, and

the lowest among clerical workers. Other

studies regarding occupational work categories in relation to CHD among men show con- tradictory findings; a number support the pres-

ent findings (Breslow and Buell, 1960; Syme

et al., 1964; Wardwell et al., 1964), while

others show no relationships (Bainton and Peterson, 1963; Paul, 1963; Stamler et al.,

1960). Although nonsupport from boss was as-

sociated with CHD among women, there was no relationship between these variables among

men in the same study (Haynes et al., 1980).

With regard to coping patterns in response to work-related stress, men more often reported

problem-solving responses than women (p < .05), although the investigators acknowledged that work settings may not have been compa- rable (Folkman and Lazarus, 1980). Coping

patterns in relation to health/illness outcomes were not investigated in this study.

Finally, in the area of overlap between work

and family roles, the two studies on men dem- onstrated a consistent relationship between

perceived stress (conflict between work and home roles) and self-reports of health (Beck

and Cassel, 1972; Coburn, 1978) compared to

inconsistent findings in similar studies con-

cerning women. However, the reliance of self- reports for all indices in the two work-family overlap studies on men limits the conclusions that may be drawn.

CONCLUSIONS AND FURTHER RESEARCH

The gaps in the literature on women and work-related stress are many, and few conclu-

sions are possible. Beginning with the first class of variables (conditions conducive to stress), the research on women lacks speci- ficity about the job environment. Specific vari- ables such as underutilization of skills, lack of recognition for accomplishment, lack of au-

tonomy, presence of deadlines and excessive hours in relation to responses, and the more enduring outcomes of these stresses warrant more thorough investigation.

There was no evidence in the literature about the link between conditions conducive to stress

and perceived stress in women. Among the

studies on work-related stress in general

(primarily male respondents), there was gener-

ally a weak relationship between conditions

conducive to stress and perceived stress (Kasl, 1978). However, these relationships appeared

to be strengthened when individual personality

variables were taken into account. For in-

stance, Kahn et al. (1964) found that individu-

als who tended to have a high level of anxiety

experienced much more perceived role conflict under objective conditions of role conflict than

did individuals who tended to have low levels

of anxiety. In future research on women and

work-related stress, it will be important to

measure individual and situational variables that might specify or condition the relationship

between objective environment and subjective

perception, and between perceptions and responses/outcomes, such as Type A behavior,

flexibility, supportive relationships with others in the work and home environments. In addi-

tion, both the objective environment and per-

ception of it should be investigated to deter- mine to what extent the actual environment is

implicated in the genesis of stress and stress-

related disease.

Few studies of women investigated per-

ceived stress (the imbalance between per-

ceived demand and perceived capability) as either a dependent or independent variable. Perceived stress was often implied or pre-

sumed, but it was rarely directly measured.

Because the individual's cognitive appraisal of the situation may be one of the crucial links to harmful physiological, affective, and behav-

ioral response-and, over time, the more en-

during illness outcomes-it is essential that

perceived stress be included as a focus of fu- ture research. Moreover, studies dealing with

perceived stress should be prospective in na- ture, as the knowledge of deleterious responses

or illness outcomes may influence the percep-

tion of stress. In addition, future studies should attempt to avoid the "triviality trap" discussed by Kasl (1978), in which the measurement of

independent and dependent variables are so

close operationally that they appear to be tap- ping a singular concept.

Evidence regarding the link between condi- tions conducive to stress and physiological,

affective, and behavioral responses among women is suggestive, but inconclusive because

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

142 JOURNAL OF HEALTH AND SOCIAL BEHAVIOR

it is sparse and it derives from cross-sectional studies. However, it suggests that employment has a beneficial effect on mental health for women. To resolve the question of antecedent-consequent relationships, how- ever, longitudinal cohort studies on house- wives and working women before, during, and after cessation of employment are needed. The present social environment involving the gamut of career involvement among women from housewife status to segmented, intermit- tent careers and uninterrupted lifetime adult employment provides an excellent opportunity for such studies.

Among the studies on women and work- related stress, the Framingham study (Haynes and Feinleib, 1980) provides the strongest evi- dence that employment may lead to functional and structural physiological damage. The sig- nificantly higher rates of CHD among clerical workers suggests that jobs that are charac- terized by underutilization of the individual's skills, lack of autonomy and control over working environment, and lack of recognition for accomplishments may contribute to the genesis of the disease, although these facets of the job environment were not measured. Moreover, the fact that the excess risk of CHD among clerical workers occurred only among women with children and among women mar- ried to blue-collar workers suggests a complex interrelationship among the following variables in the development of disease: (1) the necessity to work; (2) family responsibilities, attitudes, and lifestyles; and (3) job conditions.

Future research should address attitudes toward employment (having to work versus working by choice) and the amount of expo- sure (full-time versus part-time work and un- interrupted employment versus segmented employment). Furthermore, the whole area of work-family overlap needs to be more carefully conceptualized and precisely measured. Role conflict (having to meet conflicting demands) needs to be separated from role overload (having too much to do). Sex-role attitudes and the distribution of responsibility as well as tasks among family members regarding house- work, child care, and related family mainte- nance functions, need to be measured.

The Framingham study (Haynes et al., 1980) suggests that suppression of anger at work may lead to the development of CHD. No other

studies were found in the literature that ad- dressed women's coping styles and work- related stress. The whole area of coping styles at work in relation to health-illness outcomes is fertile for investigation. Future research should address which coping responses to perceived stress (such as problem-solving attempts and defense formation) reduce the effects of stress, and for which individuals and under what cir- cumstances they do so (illustrating the com- plexity of relationships in the genesis of stress-related disease).

Finally, the actual physiological processes through which specific diseases develop need to be conceptualized and measured. Relatedly, outcomes other than CHD and mental illness, such as allergy, rheumatoid arthritis, gastroin- testinal diseases, and a general decline in health-objectively measured by physical ex- amination and laboratory estimation, as op- posed to less reliable self-reports-need to be included.

The trends in the participation of women in the workforce indicate that work outside the home will be an increasingly important part of the lives of women. One cannot conclude at this point that the increasing workforce par- ticipation of women will not lead to increased cardiovascular disease and other illness out- comes among women. More research is needed specifying the complex interrelationships among job conditions, individual responses, and role responsibilities outside of work in the genesis of stress and stress-related disease. With such directions for study, the results of future research can provide a solid foundation for guiding corporate decision-makers in de- signing job conditions, and individuals in structuring a family life conducive to health.

REFERENCES

Bainton, C. R., and D. R. Peterson 1963 "Deaths from coronary heart disease in

persons fifty years of age and younger: A community-wide study." New England Journal of Medicine 268:569-74.

Beck, J. D., and J. C. Cassel 1972 "Role inconsistency and health status." So-

cial Science and Medicine 6:737-51. Bengtsson, C., T. Hallstrom, and G. Tibblin

1973 "Social factors, stress experience, and per- sonality traits in women with ischaemic

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

WOMEN, WORK AND STRESS 143

heart disease." Acta Medica Scandinavica (Suppl.) 549:82-92.

Breslow, L., and P. Buell 1960 "Mortality from coronary heart disease and

physical activity 'of work in California." Journal of Chronic Disease 11:615-26.

Coburn, D. 1978 "Work and general psychological and

physical well-being." International Journal of Health Services 8(2):415-35.

Cohen, J. 1976 "German and American workers: A com-

parative view of worker distress." Interna- tional Journal of Mental Health 5(2): 138-47.

Cooper, C. L., and J. Marshall 1976 "Occupational sources of stress: A review

of the literature relating to coronary heart disease and mental ill health." Journal of Occupational Psychology 49:11-28.

Cox, T. 1978 Stress. Baltimore, Maryland: University

Park Press. Feld, S.

1963 "Feelings of adjustment." In F. I. Nye and L. W. Hoffman (eds.), The Employed Mother in America. Chicago: Rand Mc- Nally.

Folkman, S., and R. Lazarus 1980 "An analysis of coping in a middle-aged

community sample." Journal of Health and Social Behavior 21:219-39.

French, J. R. P., Jr., and R. D. Caplan 1973 "Organizational stress and individual

strain." In A. J. Morrow (ed.), The Failure of Success. New York: AMACOM.

Gordon, F. E., and M. H. Strober 1978 "Initial observations on a pioneer cohort:

1974 women MBAs." Sloan Management Review 19(2):15-25.

Gove, W. R., and M. R. Geerken 1977 "The effect of children and employment on

the mental health of married men and women." Social Forces 56(1):66-76.

Gray, J. A. 1971 "Sex differences in emotional and cognitive

behavior in mammals including man: En- docrine bases." Acta Psychologica 35:29-46.

Hauenstein, L. S., S. V. Kasl, and E. Harburg 1977 "Work status, work satisfaction and blood

pressure among married black and white women." Psychology of Women Quarterly 1(4):334-39.

Haynes, S. G., and M. Feinleib 1980 "Women, work and coronary heart disease:

prospective findings from the Framingham heart study." American Journal of Public Health 70(2): 133-41.

Haynes, S. G., M. Feinleib, and W. B. Kannel 1980 "The relationship of psychological factors

to coronary heart disease in the Framing- ham study. Part III." American Journal of Epidemiology 111(1):37-58..

Haynes, S. G., M. Feinleib, S. Levine, N. Scotch, and W. B. Kannel

1978a "The relationship of psychological factors to coronary heart disease in the Framing- ham study. Part I." American Journal of Epidemiology 107:362-83.

1978b "The relationship of psychological factors to coronary heart disease in the Framing- ham study. Part II." American Journal of Epidemiology 107:384-402.

Hedges, J. N., and J. K. Barnett 1972 "Working women and the division of

household tasks." Monthly Labor Review (April) 9-14.

Hinkle, L. E., Jr., L. H. Whitney, E. W. Lehman, J. Dunn, B. Benjamin, R. King, A. Platum, and B. Flehinger

1968 "Occupation, education and coronary heart disease." Science 161:238-46.

Holmes, T. H., and R. H. Rahe 1967 "The social readjustment rating scale."

Journal of Psychosomatic Research 11:213-18.

Horan, P. M., and B. H. Gray 1974 "Status inconsistency, mobility and coro-

nary heart disease." Journal of Health and Social Behavior 15:300-10.

House, J. 1974 "Occupational stress and coronary heart

disease: A review and theoretical integra- tion." Journal of Health and Social Behav- ior 15:12-27.

Jenkins, C. D. 1971a "Psychologic and social precursors of

coronary disease. Part I." New England Journal of Medicine 284:244-55.

197 lb "Psychologic and social precursors of coronary disease. Part II." New England Journal of Medicine 284:307-17.

Johnson, F. A., and G. L. Johnson 1977, "Attitudes toward parenting in dual-career

families." American Journal of Psychiatry, 134(4):391-95.

Kahn, R. L., D. M. Wolfe, R. P. Quinn, J. D. Snoek, and R. A. Rosenthal

1964 Organizational Stress: Studies in Role Con- flict and Role Ambiguity. New York: Wiley.

Kasl, S. V. 1978 "Epidemiological contributions to the study

of work stress." In C. L. Cooper and R. Payne (eds.), Stress at Work. New York: Wiley.

Kasl, S. V., and S. Cobb 1970 "Blood pressure changes in men undergo-

ing job loss: A preliminary report." Psychosomatic Medicine 32(1):19-38.

Kenigsberg, D., S. J. Zyzanski, C. D. Jenkins, et al. 1974 "The coronary-prone behavior pattern in

hospitalized patients with and without coronary heart disease." Psychosomatic Medicine 36:344-51.

Margolis, B. L., W. H. Kroes, and R. P. Quinn 1974 "Job stress: An unlisted occupational

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms

144 JOURNAL OF HEALTH AND SOCIAL BEHAVIOR

hazard." Journal of Occupational Medicine 16(10):654-61.

Modigliani, A. 1966 Embarrassment and Social Influence. Un-

published doctoral thesis. Ann Arbor: Uni- versity of Michigan.

Nadelson, C. C., M. T. Notman, and P. Lowenstein 1979 "The practice patterns, life styles and stres-

ses of women and men entering medicine: A follow-up study of Harvard Medical School graduates from 1967 to 1977." Journal of the American Medical Women's Association 34(11):400-06.

Nathanson, C. A. 1975 "Illness and the feminine role: A theoretical

review." Social Science and Medicine 9:57-65.

Otto, R. 1979 "Negative and positive life experiences

among men and women in selected occu- pations, symptoms awareness and visits to the doctor." Social Science and Medicine 13A: 151-64.

Paul, 0. 1963 "A longitudinal study of coronary heart dis-

ease." Circulation 28:20-31. Poloma, M. M.

1972 "Role conflict and the married professional woman." In C. Safilios-Rothschild (ed.), Toward a Sociology of Women. Lexington, Mass.: Xerox College Publishing.

Radloff, L. 1975 "Sex differences in depression." Sex Roles

1(3):249-65. Rose, A. M.

1955 "Factors associated with the life satisfac- tion of middle-aged persons." Marriage and Family Living 17:15-19.

Rosenman, R. H., and M. Friedman 1961 "Association of specific behavior pattern in

women with blood and cardiovascular findings." Circulation 24:1173-84.

Rosenman, R. H., M. Friedman, and R. Straus 1964 "A predictive study of CHD." Journal of

the American Medical Association 189:15-22.

Shekelle, R. D., J. A. Schoenberger, and J. Stamler 1976 "Correlates of the JAS Type A behavioral

pattern score." Journal of Chronic Disease 29:381-94.

Stamler, J., M. Kjelsberg, and Y. Hall 1960 "Epidemiologic studies of cardiovascular-

renal disease: I. Analysis of morality by

age-race-sex-occupation." Journal of Chronic Disease 12:440-55.

Stellman, J. M. 1978 "Occupational health hazards of women:

An overview." Preventive Medicine 7:281-93.

Syme, S. L., M. M. Hyman, and P. E. Enterline 1964 "Some social and cultural factors associ-

ated with the occurrence of coronary heart disease." Journal of Chronic Disease 17:277-89.

Syme, S. L., and L. G. Reeder (eds.) 1967 Social Stress and Cardiovascular Disease.

Millbrook Memorial Fund Quarterly 45 (April): Part 2.

Szalai, A. (ed.) 1972 The Use of Time. The Hague: Mouton.

Theorell, T. 1973 "Psychosocial factors and myocardial

infarction-why and how?" Advances in Cardiology 9:117-31.

U.S. Department of Labor, Bureau of Labor Statis- tics

1977 U.S. Working Women: A Databook. Washington, D. C.: U.S. Government Printing Office.

U.S. Department of Labor, Women's Bureau 1975 1975 Handbook on Women Workers.

Washington, D.C.: U.S. Government Printing Office.

Vanek, J. 1974 "Time spent on housework." Scientific

American 231(5): 116-20. Waldron, I.

1976 "Why do women live longer than men? Part II." Journal of Human Stress 2(2):19-30.

Waldron, I. 1978 "The coronary-prone behavior pattern,

blood pressure, employment and socio- economic status in women." Journal of Psychosomatic Research 22:79-87.

Wardwell, W. I., M. M. Hyman, and C. B. Bahnson 1964 "Stress and coronary disease in 3 studies."

Journal of Chronic Disease 17:73-84. Woods, N. F.

1978 Women's Roles, Mental Ill Health and Ill- ness Behavior. Unpublished doctoral the- sis. Chapel Hill, North Carolina: University of North Carolina.

1973 Work in America. Report of a Special Task Force to the Secretary of Health, Education and Welfare. Cambridge, Mass.: MIT Press.

This content downloaded from 63.145.155.130 on Tue, 25 Sep 2018 00:28:16 UTC All use subject to https://about.jstor.org/terms