Co-morbidity burden in Parkinson’s disease: Comparison with controls and its influence on prognosis

BACKGROUND Many aspects of co-morbidity burden in Parkinson's disease (PD) are unclear, but it may be an important predictor of prognosis or confounder of associations in epidemiological studies. OBJECTIVES To determine how best to assess co-morbidity burden in PD, to compare with matched controls, and investigate its association with prognostic outcomes. METHODS Data from an incident, community-based cohort with prospective follow-up (the PINE study) were used (198 patients with PD and 151 controls). The reliability of three co-morbidity scales (the Charlson co-morbidity index (CCI), the Cumulative Illness Rating scale and a simple disease count) were evaluated. The association with mortality and development of dependency was assessed with Cox regression. The co-morbidity burden in PD and controls was compared at baseline and over 5 years of follow-up using linear mixed modelling. RESULTS The CCI was more reliable and was an independent predictor of mortality with a time-dependent effect (hazard ratio = 1.27 [1.08-1.49] in first four years of follow-up; no significant association after four years). Associations between the other scales and mortality and between each scale and development of dependency were non-significant once adjusted for confounders. Co-morbidity burden was similar between cases and controls at baseline and there was no evidence of differential accrual of co-morbidity between patients and controls (p = 0.94). CONCLUSIONS The CCI is probably the better scale for measuring co-morbidity burden in PD. There were no differences between PD and controls. Co-morbidity burden at diagnosis was associated with mortality in the early part of the disease course, but not later.