The decade-long effect of work insecurity on husbands' and wives' midlife health mediated by anxiety: A dyadic analysis

Abstract

Although the detrimental physical health effects of work insecurity have been noted in previous research, less is known about the mediating processes, such as anxiety symptoms, that link work insecurity to physical health. Even less research has explored these effects at specific life stages and how these effects may impact significant others, even though the impact of this stress may vary across the life course and the mutual influences between married partners may cause dyadic effects stemming from partners' work insecurity. To fill these gaps, the current study incorporates theories that emphasize the stress-work connection, such as stress appraisal theory (Lazarus, 1999) and resource conservation theory (Hobfoll, 1989), into a neurobiological stress-health perspective. This study uses a sample of 330 consistently married, dual-earner husbands and wives who provided data at multiple time points over a 10-year period from 1991 to 2001. Results from a model including growth curves of work insecurity and anxiety symptoms when respondents were in their early middle years and reports of physical illness in their later middle years generally supported the hypothesized model. Both the level and rate of change in work insecurity were related to the change in anxiety symptoms over time. Similarly, the level and rate of change in anxiety symptoms from 1991 to 1994 were linked to subsequent illness years later in 2001. There was only partial support for the existence of partner effects. Findings are discussed as they relate to previous research as well as policy and clinical implications. (PsycINFO Database Record

Figure 1

A Conceptual Dyadic Model Linking Work Insecurity to Husbands' and Wives' Subsequent Illness in Middle Adulthood.

Figure 2

Results from a Dyadic Model of Work Insecurity Explaining Husbands' and Wives' Illness. Notes. WI = Work Insecurity. AS = Anxiety Symptoms. Unstandardized coefficients are shown. There was a statistically significant association between the rates of change in husbands' and wives' work insecurity (r = -.02*). The paths between husbands' WI and wives' AS (level and slope), and path from the slope of wives' WI to husbands' AS slope were not statistically significant and are not shown in the figure. * p < .05. ** p < .01. *** p < .001.