Parkinson’s disease and food expenditure in Italy: stochastic and non-stochastic analyses

Abstract

Background The correlation between food and Parkinson's disease (PD) indicates that there are positive benefits from the Mediterranean diet (MeD). Objective To find the correlation between PD and food expenditure/consumption in the various regions of Italy in 2016. Methods The food expenditure/quantities in kilograms of 56 distinct food categories and PD, in terms of Standardised Mortality Ratio (SMR), were computed for the year 2016 for 19,500 families in 540 Italian municipalities. Stochastic analysis based on Partial Least Square Regression (PLSR) and non-stochastic analyses (neural network mapping) were used to compute the associations with PD. Results The following results were obtained by focusing on food categories deemed significant in both stochastic and non-stochastic analyses: Yogurt, fresh vegetables, wine, processed meat, citrus fruit, and butter were identified as causative factors, while cheese, ice cream, beef, grapes & strawberries , bananas, canned fruit, and fruit juice were deemed partially causative. Added sugar, flour, lamb meat, and dried legumes emerged as protective, while seed oil, pork meat, fish, fish canned, and tomatoes canned showed partial protective effects. In Southern Italy, the SMR of PD was significantly lower than in the North due to a lower consumption of causative foods and a higher consumption of protective ones. Conclusions In 2016, the PD death rate in South of Italy was significantly lower than that in the North. The food pattern that emerged in Southern regions was also significantly different: a lower consumption of causative foods and higher consumption of protective ones. Using data on food expenditure and quantities enabled us to track the correlation with PD SMR on an annual basis.