Initial BMI and Weight Loss over Time Predict Mortality in Parkinson Disease

Abstract

Objectives: Although weight loss is a frequent symptom in Parkinson disease (PD), there have been few studies on the association between body mass index (BMI) and mortality. The objective of this study was to investigate the association between BMI and change in BMI at diagnosis in patients with PD and all-cause mortality.

Design: Cohort study using Korean National Health Insurance Service-Elderly Cohort data.

Setting and participants: Patients with new-onset PD were selected using the International Classification of Diseases 10th edition code (G20). Then, patients who were diagnosed more than 3 times with PD and had been prescribed anti-parkinsonian medication for ≥30 days were included. Those with a combined diagnosis of atypical parkinsonism and secondary parkinsonism were excluded.

Methods: The primary outcome was all-cause mortality. Anthropometric data, including height and weight, were obtained from the health screening data to calculate BMI. The Cox proportional hazards model was used to assess mortality risk by BMI.

Results: Among the 2703 patients with PD, 492 (18.20%) died during the 11-year follow-up period. There was a significant inverse dose-response relationship between baseline BMI and mortality [<18.5 kg/m2: hazard ratio (HR), 1.872, 95% CI, 1.338-2.494; 23-25 kg/m2: HR, 0.695, 95% CI, 0.546-0.886; 25-30 kg/m2: HR, 0.644, 95% CI, 0.476-0.869; ≥30 kg/m2: HR, 0.396, 95% CI, 0.165-0.950]. Change in BMI of 10% revealed a significant association with mortality. Subgroup analyses by sex showed a significant inverse dose-response relationship between BMI and all-cause mortality only in women.

Conclusions and implications: We demonstrated an inverse dose-response association between BMI at diagnosis and mortality in patients with PD, especially in women. Early detection of PD before weight loss progression and proper management might improve mortality. The small number of obese PD participants in our study should be considered when interpreting and generalizing results.