Background Idiopathic Parkinson's disease (PD) is a common, disabling, neurodegenerative disorder. The overall co-morbidity burden associated with PD is unclear, but may be important to adjust for when predicting prognosis or comparing cases and controls.

Aims ▸ To determine how best to assess overall co-morbidity in PD

▸ To compare PD co-morbidity burden to that of age- and sex-matched controls

Methods Data from an incident, community-based cohort of 205 patients with PD and 148 age-, sex- and GP-matched controls (the PINE study) were used. The intra- and inter-rater reliability and mortality predictive ability of three co-morbidity scales (the Charlson Co-Morbidity Index, the Cumulative Illness Rating Scale and a disease count) were evaluated. The co-morbidity burden of cases and controls was compared at baseline and over 5 years of follow-up.

Results and conclusions The Charlson Co-Morbidity Index was more reliable for use in PD and was the only scale that was independently predictive of mortality (hazard ratio=1.20, [95% CI 1.07–1.34]). There was no significant difference between cases and controls at baseline (p=0.20). Charlson Co-Morbidity Index scores increased over time. This increase was greater in patients with PD than controls and greater in patients and controls who died earlier.