The positive association between difficult work conditions and mental illness may reflect the fact that difficult work conditions cause mental disorders (the causation hypothesis) or that people with prior mental illness select or are selected into difficult work conditions (the selection hypothesis). Thus, under certain circumstances where individuals with prior mental illnesses may remain ill and that they select or are selected into difficult work conditions, the effect of difficult work conditions on current mental illness would become causally spurious. In this study the classical, competing hypotheses of selection and causation and others are assessed with a rarely studied, albeit important group of individuals, youths who have recently entered the work force. Causation and selection effects will be examined prospectively using a longitudinal community sample of 615 youths from upstate New York. The sample was first interviewed in 1975 when the cohort was between the ages of 1 and 10. In 1984 and 1992 mental health and family environment data were collected and information on work conditions was collected in the latter wave. Mental illness is conceptualized by this dissertation in terms of anxiety, depression and alcohol abuse disorders, disorders that are the most prevalent among workers. Work conditions are measured in terms of excessive job demands, lack of decision latitude and conflictual work relationships. Regression and logistic regression models will be used to analyze the data. In addition, structural equation models will be used to estimate indirect and direct causal effects. Controls for prior mental illness, gender, race, family income, and external social support (family, friends not in the work place) will be included in all models. Finally, the present study will investigate any specificity that might exists between the type of work place condition and type of mental illness and separate models will be run for women, men, and youth from low and high income families.
