Testing integrated behavioural and biomedical models of activity and activity limitations in a population-based sample.

Purpose: The predictive utility of an integrated model of disability is tested. The integrated model incorporates an impairment based model (International Classification of Functioning, Disability and Health (ICF)) and the behavioural models. Methods: Community dwelling adults (n = 628) completed a postal questionnaire measuring the integrated model. The ability of the model to predict disability in the form of activity limitations (ALs) and walking, in the full community sample and in respondents reporting chronic pain was tested. Results: In both the community and chronic pain samples each version of the integrated model explained a majority (55%-67%) of the variance in ALs but only 11%-29% of the variance in walking behaviour (WB). Impairment directly predicted ALs but did not directly predict WB. Control related cognitions were direct predictors, and mediators, of the relationship between bodily impairment and both ALs and WB. In addition, intentions and outcome expectancies predicted WB. Self-efficacy (SE) was the most consistent predictor of both ALs and WB. Conclusions: An integrated model which combines psychological constructs and impairment is required for an adequate understanding of ALs. By contrast, behavioural models, but not degree of impairment, are necessary to explain activity levels. [ABSTRACT FROM AUTHOR]