Body mass index changes across the trajectory of parkinsonism: results from the PRIME-UK cross-sectional study

Background: Body mass index (BMI) trajectory prior to and after diagnosis of Parkinson’s disease (PD) is uncertain with conflicting findings between studies. BMI may be used as an early biomarker,and change prior to diagnosis suggests a metabolic aetiology or reverse causation.

Methods: Weight and height were collected retrospectively from primary and secondary care electronic records for patients with parkinsonism based in a single centre and enrolled in the PRIME-UK cross-sectional study. BMI was derived for both the period before and after the year of parkinsonism diagnosis either from participants’ self-report or clinical records. Questionnaire assessment included the SCREEN-I4 screening tool for nutritional risk. Change in BMI was explored using multilevel linear regression models. Associations between BMI and SCREEN-14 score were evaluated using linear regression.

Results: 256 patients had one or more recorded weights (3378 BMI readings in total) and 172 had a diagnosis BMI, defined as ± 1 year from diagnosis. In the period following parkinsonism diagnosis, BMI decreased by 0.31 kg/m2 per year (P < 0.0001), with no significant change in BMI found during the period before diagnosis (P = 0.93). When modelled as a non-linear relationship, there was evidence for a non-clinically significant reduction in BMI pre-diagnosis (P = 0.04). BMI was not associated with self-reported nutritional risk score.

Conclusions: BMI was shown to decrease following diagnosis with only a weak pre-diagnosis decline, more suggestive that BMI change may be due to reverse causation. Nutritional screening should not depend solely on BMI. Future cohort studies with repeat BMI measures and detailed confounders should be undertaken.