| ObjectiveTo analyse the correlation of blood homocysteine(Hcy)level and carotid intima-media thickness(cIMT)with the severity of white matter lesions(WML)and their combined predictive value for WML.Methods and materials1.Object of studyWe retrospectively collected consecutive patients admitted to the department of neurology,the Third Affiliated Hospital of Southern Medical University from January 2019 and March 2021.2.Data collection2.1 Clinical dataIncluding name,gender,age,with or without hypertension,with or without diabetes,with or without coronary heart disease,with or without smoking history,with or without drinking history.2.2 Blood index dataAll patients harvest fasting venous blood in the next morning.The fasting venous blood was sent to our hospital laboratory to complete the routine blood,clotting function,liver and kidney function,blood lipids,serum uric acid,blood Hcy and other tests,adopted Beckman and au5800 automatic biochemical analyzer.Collecting indicators include total cholesterol,triglycerides,high density lipoprotein,low density lipoprotein,Hcy,uric acid and so on.2.3 Imaging data collectionAll patients enrolled in the hospital were examined by color Doppler flow imaging to measure the carotid intima-media wall thickness.All patients were accepted 1.5T or 3.0T MRI examination,including T1MRI(TR/TE:2045/20ms),T2MRI(TR/TE:4000/107ms),T2-FLAIR sequence(TR/TE:9000/120ms,reversal time:2500ms).Then two clinically experienced neurologists using double-blind method based on the amount of Fazekas Table to evaluate the severity of white matter hyperintensity(WMH)[1].3.Statistical analysisUsing SPSS 25.0 statistical software to analysis.The measurement data of normal distribution were expressed as mean as mean ± standard deviation,and two independent samples t-test were used to compare between groups.If it does not obey the normal distribution,use the median(lower quartile,upper quartile)to describe it,and use the nonparametric test.The enumeration data expressed as a percentage,two groups were compared by chi-square test.Binary logistic regression was used to analyze the independent risk factors of moderate and severe WMH.The predictive value of relevant indicators was drawn by plotting the receiver operating curve(ROC curve).It was considered that P<0.05 was significant.Results1.Clinical dataIt was included 427 patients in this study,with an average age of 62.83±11.478 years.There were 306 male patients(71.66%)and 121 female patients(28.34%).185 patients(43.33%)had hypertension and 242 patients(56.67%)had hypertension.69 patients(16.16%)were diabetic and 358 patients(83.84%)were non-diabetic.There were 32 patients(7.49%)with coronary heart disease and 395 patients(92.51%)without coronary heart disease.123 patients(28.81%)had a history of smoking and 304 patients(71.19%)had no history of smoking.There were 42 patients(9.84%)with drinking history and 385 patients(90.16%)without drinking history.258(60.42%)were mild WMH and 169(39.58%)were moderate-severe WMH.2.According to the clinical characteristics of homocysteine groupingHcy was divided into four grades according to blood level by four percentile method.There were significant differences in gender,age,hypertension and uric acid among four groups,while diabetes,coronary heart disease,smoking history,drinking history,total cholesterol,triglyceride,high-density lipoprotein,low density lipoprotein,and very low density lipoprotein were no significant.3.Analysis of clinical characteristics after grouping according to the thickness of left-sided carotid intima-mediaAccording to the intima-media thickness of the left carotid artery,there were 247 patients in the normal group,145 patients in the intimal thickening group and 35 patients in the plaque formation group.There were significant differences in age,gender,hypertension,coronary heart disease,smoking history,uric acid and Hcy among the three groups,while diabetes,drinking history,total cholesterol,triglyceride,high-density lipoprotein,low density lipoprotein and very low density lipoprotein were no significant.4.Analysis of clinical characteristics after grouping according to the thickness of right-sided carotid intima-mediaAccording to the intima-media thickness of the right carotid artery,there were 248 patients in the normal group,147 patients in the intimal thickening group and 32 patients in the plaque formation group.There were significant differences in age,gender,hypertension,coronary heart disease,smoking history,total cholesterol,triglyceride,very low density lipoprotein and Hcy among the three groups,while diabetes,alcohol consumption,uric acid,high density lipoprotein and low density lipoprotein were no significant.5.Compare the difference between left and right-sided cIMTIn all age groups,the left-sided(0.992mm)and the right-sided(0.991mm)were similar with the mean value of cIMT,and there was no significant difference between the two groups.Taking 60 years old as the cut-off point,the differences of cIMT between the left and right were compared:in patients under 60 years old,the average value of left-sided cIMT(0.946mm)was greater than that of right-sided cIMT(0.937mm),and there was no significant difference between the two groups;Among patients over 60 years old,the average value of right-sided cIMT(1.037mm)was greater than that of left-sided cIMT(1.031mm).Between the two groups,it was no significant difference.6.Univariate analysis of risk factors for paraventricular white matter hyperintensity(PVWMH)in the mild group and moderate-severe groupThere were 273 patients in the normal/mild PVWMH group and 154 patients in the moderate-severe PVWMH group.Compared with the normal/mild PVWMH group,the age of moderate-severe PVWMH is older,and the proportion of men is higher.The prevalence of hypertension,diabetes and coronary heart disease and Hcy levels were higher.Total cholesterol,triglyceride and very low density lipoprotein levels were lower.Besides,the severity of PVWMH was not associated with uric acid,high-density lipoprotein and low density lipoprotein.It was no significant difference in smoking history and drinking history.According to the results of univariate analysis,the independent variables of P<0.1 were included in the binary logistic regression model.After adjusting for confounding factors,age,hypertension and diabetes were found to be independent risk factors for moderate-severe PVWMH.7.Univariate analysis of risk factors for deep white matter hyperintensity(DWMH)in the mild group and moderate-severe groupThere were 335 patients in the normal/mild DWMH group and 92 patients in the moderate-severe DWMH group.Compared with the normal/mild DWMH group,the age of moderate-severe DWMH is older,and the proportion of men is higher.The prevalence of hypertension and coronary heart disease was higher,triglyceride,very low density lipoprotein and homocysteine levels were higher.Besides,the severity of DWMH was not associated with uric acid,total cholesterol,high-density lipoprotein and low density lipoprotein.It was no significant difference in diabetes and smoking and drinking history.According to the results of univariate analysis,the independent variables of P<0.1 were included in the binary logistic regression model.After adjusting for confounding factors,age,hypertension,Hcy and right-sided cIMT were found to be independent risk factors for moderate-severe DWMH.8.The results of univariate analysis for the mild and moderate-severe WMHAmong 427 WMH patients,258 cases(60.42%)were mild WMH and 169 cases(39.58%)were moderate and severe WMH.Compared with mild WMH,moderate-severe WMH was older,the prevalence of hypertension,diabetes and coronary heart disease were higher,total cholesterol,triglyceride and very low density lipoprotein were lower,while homocysteine levels was higher and IMT was thicker.In addition,we also found the severity of WMH has nothing to do with uric acid,high-density lipoprotein,low-density lipoprotein,smoking history and drinking history.9.The results of binary logistic regression analysis of risk factors for moderate-severe WMHAccording to the results of univariate analysis,the independent variables of P<0.1,were included in the binary logistic regression model.After adjusting confounding factors,age,hypertension,diabetes,Hcy and the right-sided cIMT were found to be independent risk factors for moderate-severe WMH.Hcy and right-sided cIMT were divided into four grades by quartiles and enter the logistic equation as a rank variable.After adjusting confounding factors,the increase of Hcy level and the thickening of right-sided cIMT are increase the risk of WMH.In patients with WMH,the risk of developing moderate-severe WMH in those with the highest levels of Hcy and the right-sided cIMT were 8.609 times and 3.167 times higher than those with the lowest levels,respectively.10.Plot the receiver operating curve of Hcy and right-sided cIMT between moderate-severe WMHThe area under the Hcy curve was 0.696(95%CI 0.642~0.749,P<0.001).The best cut-off value determined by Youden index was 13.10 for Hcy with the sensitivity of 69.23%and the specificity of 77.52%.The area under the right-sided cIMT curve was 0.695(95%CI 0.644~0.746,P<0.001).The best cut-off value determined by Youden index was 1.05 for ride-sided cIMT with the sensitivity of 59.76%and the specificity of 69.77%.Hcy and the right-sided cIMT were entered the binary logistic equation as a rank variable to obtain the joint prediction probability and draw the ROC curve.The area under the ROC curve was 0.781(95%CI 0.735~0.827,P<0.001).When the yoden index was the largest,the predictive sensitivity and specificity for moderate-severe WMH were 69.23%and 77.52%.ConclusionsAge,hypertension,diabetes,Hcy and right-sided cIMT were the independent risk factors for moderate-severe WMH in patients with WML.High level Hcy and thickened right-sided cIMT provide a certain predictive value for moderate-severe WMH.Hcy combined with right-sided cIMT can improve the predictive value of WMH. |