Older adults with osteoarthritis have been found to be impaired in physical functioning and report higher levels of depression. This study examined the relationships between pain, functional disability, and depression to test the activity restriction model in a cohort of 176 older adults in China. This model states that disability is a mediator for the relationship between pain and depression. Other investigators have found that pain and disability were two independent correlates of depression. In both cross-sectional and longitudinal analyses, the authors found that disability is a mediator, using commonly accepted methods (indirect effect 44%, Sobel Z = 4.07, P < .001; 41%, Sobel Z = 3.45, P < .001, respectively). However, this mediation effect was not seen when the outcome was residualized with the baseline value. When the baseline level of depression is residualized, the effect size of the relationship is reduced, requiring larger sample size to test its effect.