Long-Term Consumption of Ultraprocessed Foods and Prodromal Features of Parkinson Disease.

BACKGROUND AND OBJECTIVES

Consumption of ultraprocessed foods (UPFs) has been associated with a higher risk of various chronic diseases, but its relation to prodromal Parkinson disease (PD) remains unclear. We aimed to assess the association between long-term UPF consumption and nonmotor features suggestive of prodromal PD.

METHODS

This longitudinal analysis included participants without a history of PD from the Nurses' Health Study and Health Professionals Follow-Up Study. UPF consumption was assessed using repeated food frequency questionnaires (1984-2006) and grouped based on Nova classification. Participants provided data on probable REM sleep behavior disorder (pRBD) and constipation in 2012. Between 2014 and 2015, a subset of participants provided data on 5 additional nonmotor features, including hyposmia, impaired color vision, excessive daytime sleepiness, body pain, and depressive symptoms. The primary outcome was the combination of all 7 prodromal features and further categorized as 0 (reference), 1, 2, and ≥3 features. The secondary outcomes were all features except constipation, a combination of 3 commonly recognized features (constipation, pRBD, and hyposmia), and individual features. Multinomial logistic regression was used to estimate the association of UPF consumption with the combination of prodromal features. The association between UPF consumption and each individual feature was further examined using logistic regression.

RESULTS

The study analyzed 42,853 participants (25,095 women [58.6%]; mean [SD] age, 47.8 [5.2] years). Comparing extreme quintiles of UPF consumption, the multivariable-adjusted odds ratio (OR) for having ≥3 vs 0 prodromal features was 2.47 (95% CI 1.89-3.23, < 0.0001) for cumulative average intake and 1.50 (95% CI 1.18-1.89, = 0.0009) for baseline intake. Similar results were observed for combinations of all features except constipation (OR 2.00, 95% CI 1.29-3.11, < 0.0001) and combinations of 3 features (OR 2.47, 95% CI 1.41-4.34, = 0.008). In addition, higher UPF consumption was associated with increased odds of individual prodromal features, including pRBD, constipation, body pain, and depressive symptoms.

DISCUSSION

Long-term UPF consumption was positively associated with nonmotor prodromal PD features. More studies are warranted to confirm whether lowering UPF consumption may prevent the occurrence of nonmotor symptoms that often precede PD diagnosis.