Attributing Causes to Disability

Diseases play a major role in the disablement process, especially at older ages. This chapter focuses on the attribution method, which uses cross-sectional data to partition the disability prevalence into additive contribution of causes, taking into account multimorbidity and that disability can occur in the absence of diseases. We present a detailed description of the attribution method, including the definition of the additive hazard models for binary and multinomial disability outcomes. The method is applied to cross-sectional data from Brazil to illustrate the interpretation of the cumulative hazard rates of disability and how the method accounts for multimorbidity and independence. A summary of previous studies that have applied the method are also provided. Finally, the limitations and strengths of this approach compared to alternative methods using cross-sectional data are outlined.