The Relationship Among Small Molecular Metabolites In Serum With Parkinson’s Disease And Non-motor Symptom Scores Of The Disease

ObjectivesWe aim to screen the potential biomarkers of Parkinson’s disease in small molecule metabolites of serum,after that,to determine the diagnostic value of Parkinson’s disease and to further explore the relationship between these serum indexes and the non-motor symptom scores of Parkinson’s disease.MethodsCase and control group respectively were derived from Parkinson’s disease patients of a Third Grade&Class hospital and healthy subjects of a physical examination center in Dalian from December,2015 to June,2017.Finally,275 participants were analyzed by using exclusion and inclusion criteria in the present study.including 90 patients with Parkinson’s disease and 185 healthy subjects.We collected several patients’ information,such as demographics,concentration of small molecule metabolites and personal medical history,and evaluated the non-symptoms scores(MMSE,HAMA,MADRS and PSQI)of Parkinson’s disease.In entire cohort,we collected blood samples of 42 PD and 31 healthy subjects by convenience sampling to detect the serum amino acid metabolites using the pre-column derivatization with phenyl isocyanate of High Performance Liquid Chromatography.ResultsCompared to the control group,the univariate results showed that the PD group was significantly associated with decreased serum uric acid and total cholesterol(Uric acid:PD vs.Con,310.96±60.81μmol/L vs.351.85±89.42μmol/L,P=0.047;Total cholesterol:PD vs.Con,5.06±0.99mmol/L vs.5.65±1.26mmol/L,P=0.001),and the PD group was significantly associated with increased serum phenylalanine and homocysteine(Phenylalanine:PD vs.Con,8.60±4.55μmol/L vs.4.27±1.92μmol/L,P<0.001;Homocysteine:PD vs.Con,12.9μmol/L(10.5,16.8)vs.9.58μmol/L(8.17,12.13),P<0.001).After that,we put the significantly univariate factors into the multivariable logistic regression model.The result showed that the serum phenylalanine(OR=1.484,P＝0.007)and total cholesterol(OR=0.310,P=0.036)were related with Parkinson’s disease adjusted for the age and gender,while the homocysteine level(OR=1.052,P=0.085)was just close to statistical significance.We further analyzed the area under the ROC curve of the influence factors of P<0.100 in the multivariable logistic regression model,respectively,serum phenylalanine of area under the curve was 0.806±0.058(P<0.001);homocysteine of area under the curve was 0.756±0.065(P=0.001);total cholesterol of area under the curve was 0.638±0.073(P=0.067)among which serum phenylalanine is the best indicator for Parkinson’s disease,nevertheless,in the combined model,the combination of three small molecule metabolites of area under the curve was 0.915±0.041(P<0.001).In women with Parkinson’s disease,serum phenylalanine was positively correlated with disease duration(r=0.593,P=0.042)and positively correlated with HAMA scores(r=0.587,P=0.027),serum homocysteine was negatively correlated with MMSE scores(r=-0.552,P=0.018).Nevertheless,in aged>63 years old of patients with Parkinson’s disease,the serum total cholesterol was negatively correlated with depression scores(r=-0.509,P=0.031),negatively correlated with sleep score(r=-0.459,P=0.048)and similarly,negatively correlated with the disease duration(r=-0.613,P=0.015).Conclusion(1)The serum phenylalanine,homocysteine and total cholesterol were related with Parkinson’s disease.(2)The diagnostic value of phenylalanine was the highest in single detection,nevertheless,in the combined model,the combination of three small molecule metabolites was the highest diagnostic value.(3)In the lower age group,with the increase of homocysteine level,the patients’ cognitive performance was worse;as the level of phenylalanine raised,the patients’ anxious performance was worse.In the high age group,depression and sleep quality were better with total cholesterol level rising.