Psychosocial correlates of women's health in middle adulthood

Psychosocial variables are not only highly salient to health, but also potentially modifiable if women receive accurate information from care providers. Employing a multivariate multitheoretical approach, psychosocial predictors of health (locus of control, well-being, role quality, stress, social network ties, and optimism) were examined in a secondary analysis of data collected by Baruch and Barnett on 238 midlife women. Mean age was 43.6 years; 60% were college educated; 76% were employed outside the home. The sample included women from four family statuses: never married (n = 50), married without children (n = 54), married with children (n = 88), and divorced with children (n = 46). These family statuses were crossed by employment status: half the married women were employed, as were all of the never married or divorced women. The employed women were equally distributed by occupational prestige into high, medium, and low groupings; unemployed women were classified according to the prestige of their husbands' occupations. Of the 134 women who were mothers, 23% had younger children (< or = 7 years), while 66% had no child under age eight at home and 11% had an "empty nest." Health was regressed on psychosocial and demographic predictors using a backward elimination procedure. Twenty-seven percent of the variance in self-reported health of mothers (whether married or divorced) was accounted for by stress (beta = -.29), optimism (beta = .17), employment outside the home (beta = .14), occupational prestige (beta = .18), and quality of experience in the maternal role (beta = .14). Twenty-two percent of the variance in health of wives was explained by stress (beta = -.36), quality of experience in the wife role (beta = -.12), employment (beta = .15) and occupational prestige (beta = .14). Using t tests, healthiest and least healthy women were compared. Women in better health in middle adulthood had fewer concerns regarding their roles as workers (t = -2.03, p = .04), wives (t = -4.57, p = .0001) and mothers (t = -3.89, p = .0002), as compared to their less healthy counterparts. Women in the least healthy group appeared to be more unsure of themselves as mothers, worrying more and having more arguments with their children. They also reported more conflicts with spouse about housework and children. To summarize, stress was the strongest predictor of health in this sample. Concerns reported by women in connection with their primary roles differentiated healthiest and least healthy women. Implications for research and practice were discussed.