Affectivity, quality of life and health resources utilization in arthritis

The main objective of this dissertation was to analyze the role of psychosocial variables, especially positive affect, on health outcomes and health care services utilization (HCSU) in arthritis. The identification of additional psychological variables as predictive or mediating factors is critical for gaining a better understanding of the dynamics of treatment response, especially in the setting of interventions that focus on behavioral changes, and health care seeking behavior. The analyses performed were based on a theoretical model that assumed affective status as a direct effector on the use of specific coping strategies and patients' perception of control over their health, which would, in turn, influence health outcomes and health resources utilization. Using random effects modeling, we found positive affect (PA) to have a negative (β = -4.27; p < 0.001) and negative affect (NA) a positive (β = 5.33; p < 0.001) longitudinal association with levels of perceived pain, adjusted for other covariates. The final, most parsimonious, model explained 22% of the overall variance and 28% of the between-subjects variance of the level of pain perceived by subjects with osteoarthritis. Subjects with high PA levels were more likely to achieve minimal clinically important differences in pain (OR = 7.1; 95% CI, 1.4 to 36.2), global assessment of disease activity (OR = 6.1; 95% CI, 1.7 to 21.8), and physical function (OR = 13.27; 95% CI, 1.7 to 102.1) than subjects with low PA. Finally, in a multivariate stepwise linear regression model, higher levels of positive affect significantly predicted less HCSU (B = -0.29; p = 0.04). The final model also included social support and anxiety and explained slightly over 17% of the variance of long term HCSU (R = 0.51; Adjusted R-square = 0.217)