Predictors of Stress-Related Illness in City Bus Drivers

This study examines the relation between stress and illness among bus drivers in a large American city. Several factors are identified that predict stress-related ill health for this occupational group. Canonical correlation techniques are used to combine daily work stress and recent stressful life events into a single life/work stress variate. Likewise, somatic symptoms and serious illness reports are combined into a single canonical illness variate. This procedure simplifies the analysis of multiple stress and illness indicators and also permits the statistical control of potential contaminating influences on stress and illness measures (eg, neuroticism). Discriminant function analysis identified four variables that differentiate bus drivers who get ill under high stress (N= 137) from those who remain healthy under stress (N = 137). Highly stressed and ill bus drivers use more avoidance coping behaviors, report more illness in their family medical histories, are low in the disposition of "personality hardiness" and are also low in social assets. The derived stepwise discriminant function correctly classified 71% of cases in an independent "hold-out" sample. These results suggest fruitful areas of attention for health promotion and stress management programs in the public transit industry.