Differential effects of cholesterol levels on cognition according to body mass index in Parkinson's disease.

BACKGROUND: Cholesterol is an essential component of the neuronal cell membrane and is crucial for neuronal function; however, the role of cholesterol levels in Parkinson's disease (PD) is debatable. This study investigated the complex relationship between total cholesterol (TC) levels, body mass index (BMI), and cognition in patients with PD. METHODS: This study included 321 drug-naïve patients with PD who underwent dopamine transporter (DAT) imaging and baseline neuropsychological tests. Multivariate linear regression and Cox regression models were used to investigate the effect of TC levels on the composite score of each cognitive domain and dementia conversion after adjusting for covariates, respectively. Interaction analyses were performed to examine the interaction effect between TC levels and BMI on baseline cognition and dementia conversion. RESULTS: TC levels and cognition showed no significant relationship after adjusting for potential confounders. A significant interaction effect between TC levels and BMI was observed in frontal/executive function and dementia conversion. Further analyses showed that TC levels were positively associated with frontal/executive function in the under-/normal weight group (ß = 0.205, p = 0.013), whereas a negative relationship existed between TC levels and frontal/executive function in the obese group (ß = - 0.213, p = 0.017). Cox regression analyses also showed the differential effects of TC levels on dementia conversion according to BMI (under-/normal weight group: hazard ratio [HR] = 0.550, p = 0.013; obese group: HR = 2.085, p = 0.014). CONCLUSIONS: This study suggests a cross-over interaction between TC levels and BMI on cognitive symptoms in PD.

Different effect of hypo- and hypermetabolism on cognition in dementia with Lewy bodies: are they coupled or independent?

Patients with dementia with Lewy bodies (DLB) show widespread brain metabolic changes. This study investigated whether brain hypo- and hypermetabolism in DLB have differential effects on cognition. We enrolled 55 patients with DLB (15 prodromal DLB [MCI-LB] and 40 probable DLB) and 13 healthy controls who underwent 18F-fluorodeoxyglucose positron emission tomography and detailed neuropsychological tests. Metabolic indices reflecting associated changes in regional cerebral glucose metabolism were calculated as follows: index(-) for hypometabolism [DLB-hypo] and index(+) for hypermetabolism [DLB-hyper]. The effects of DLB-hypo or DLB-hyper on cognitive function were assessed using a multivariate linear regression model. Additionally, a linear mixed model was used to investigate the association between each index and the longitudinal cognitive decline. There was no correlation between DLB-hypo and DLB-hyper in the disease group. The multivariate linear regression model showed that DLB-hypo was associated with language, visuospatial, visual memory, and frontal/executive functions; whereas DLB-hyper was responsible for attention and verbal memory. There was significant interaction between DLB-hypo and DLB-hyper for verbal and visual memory, which was substantially affected by DLB-hyper in relatively preserved DLB-hypo status. A linear mixed model showed that DLB-hypo was associated with longitudinal cognitive outcomes, regardless of cognitive status, and DLB-hyper contributed to cognitive decline only in the MCI-LB group. The present study suggests that DLB-hypo and DLB-hyper may be independent of each other and differentially affect the baseline and longitudinal cognitive function in patients with DLB.