Study Of Relationship Between FNIRS Cerebral Hemodynamic Response And Clinical Symptoms And Cognitive Function In Patients With Major Depressive Disorder

BackgroundMajor depressive disorder（MDD）is a common clinical mental illness,usually accompanied by cognitive impairment,somatic symptoms,behavioral changes,and more than half of patients with MDD with anxiety symptoms.Research shows that after the patients’emotional symptoms improve,their residual cognitive impairment can still lead to continuous impairment of work and social function,recurrence of depressive symptoms,and serious cases can lead to self injury and suicide.The evaluation of cognitive function of patients with MDD should be included in the whole course of treatment management,and the recovery of cognitive function should be regarded as an important goal of treatment.It is very important to choose appropriate and standardized cognitive function measurement tools.Research shows that some objective cognitive tests can reliably reflect the degree of cognitive impairment of patients with MDD.Many functional near-infrared spectroscopy（f NIRS）studies based on cognitive tasks have found that decreased activation of frontotemporal cortex is a biomarker for the auxiliary diagnosis and differential diagnosis of MDD.However,the relationship between clinical symptoms,cognitive impairment and frontotemporal cortical function in patients with MDD needs to be further clarified.This study explored the relationship between f NIRS measurement of frontotemporal cortical function,cognitive function and clinical symptoms in patients with MDD,and the potential of f NIRS as an objective measurement method to identify cognitive impairment and clinical symptoms.Objectives1.Analyze changes in cognitive function in patients with MDD.2.Analysis of hemodynamic features of the frontotemporal lobe in patients with MDD during verbal fluency text（VFT）.3.Analyze the correlation between the activation characteristics of frontotemporal lobe and the severity of depressive symptoms,anxious symptoms and cognitive impairment in patients with MDD during VFT.4.To explore the value of frontotemporal activation characteristics during cognitive tasks in distinguishing MDD and healthy controls（HC）.Methods1.Subjects:40 patients with major depressive disorder were selected as MDD group,and 40 healthy controls matched by age,gender and years of education were included in the HC group.2.Evaluation of clinical symptoms and cognitive function:Hamilton depression scale（HAMD）,Beck depression inventory（BDI）,Hamilton anxiety scale（HAMA）and Beck anxiety inventory（BAI）were used to evaluate the severity of clinical symptoms.The cognitive function of all subjects was evaluated by digit symbol substitution text（DSST）,digita span text（DST）,semantic verbal fluency task（SVF）and phonological verbal fluency task（PVF）.3.Near infrared data collection:the activation characteristics of prefrontal and bilateral temporal lobes of all subjects during VFT were collected by Hitachi ETG-4100 functional near-infrared spectroscopy.The system software obtains the quality index by calculating the average near-infrared data of multiple channels in the frontal and temporal lobes to judge whether each channel is an invalid（artifact）channel or an effective channel.The artifact channel is displayed in black.Near infrared data with more than 6 effective channels in frontal lobe and bilateral temporal lobe were included.The f NIRS brain activation characteristics of each subject were recorded,including prefrontal lobe integral value,prefrontal lobe centroid value,temporal lobe integral value and temporal lobe centroid value.4.Statistical analysis:Using SPSS 24.0 statistical software,the independent sample t test was used when the measurement data conformed to the normal distribution,and the Mann-Whitney U test was used when the measurement data did not conform to the normal distribution.Chi-square test was used for enumeration data.Taking the years of education,age and the PVF score as covariates,and group and gender as fixed factors,the multivariate general linear model was used to compare the differences between the two groups.Pearson correlation/Spearman correlation was used to analyze the correlation between HAMD-24 score and each factor score,HAMA score and each factor score,BDI score,BAI score,DSST score,DST score,SVF score,PVF score and prefrontal lobe integral value,prefrontal lobe centroid value,bilateral temporal lobe integral value and bilateral temporal lobe centroid value.Multivariate logistic regression was used to screen the factors that can distinguish the MDD group from HC group.The selected factors were analyzed by receiver operating characteristic curve（ROC）,the area under the curve（AUC）was calculated.The optimal cut-off value for diagnosis was calculated according to the Youden index,and the sensitivity and specificity of this value to distinguish MDD group from HC group were analyzed.Results1.There was no significant difference between these two groups in total DST score（t=0.190,P=0.850）and SVF score（t=-1.605,P=0.113）.The scores of DSST（t=-5.641,P<0.001）and PVF（t=-3.653,P<0.001）in MDD group were significantly lower than those in HC group.2.The integral value of prefrontal lobe in MDD group（t=-2.582,P=0.012）was lower than that in HC group,the difference was statistically significant.There was no significant difference in centroid value of prefrontal lobe（t=1.521,P=0.132）,integral value of bilateral temporal lobe（t=-1.834,P=0.070）and centroid value of bilateral temporal lobe（t=-0.479,P=0.634）between the two groups.3.Taking the years of education,age and total PVF scores as covariates,group and gender as fixed factors,the multivariate general linear model was used to compare the differences of near-infrared data between the two groups.It was found that there was significant difference between the integral value of prefrontal lobe（F=5.578,P=0.021）and bilateral temporal（F=4.076,P=0.047）,and there was no significant difference between the centroid value of prefrontal lobe（F=0.465,P=0.498）and bilateral temporal lobe（F=0.060,P=0.807）.4.In MDD group,the prefrontal centroid value was positively correlated with HAMD retardant factor score（r=0.327,P=0.039）,and the bilateral temporal lobe integral value was negatively correlated with HAMD despair factor score（r_s=-0.351,P=0.026）.There was no significant correlation between HAMD-24 total score and other cofactor score,HAMA total score and each factor score,BDI score,BAI score,DSST score,DST score,SVF score,PVF total score and the four f NIRS characteristics in MDD group（P>0.05）.5.Take the group（whether it is patients with MDD）as the dependent variable,four factors including DSST score,PVF total score,prefrontal lobe integral value and bilateral temporal lobe integral value as the independent variable（P<0.1）.Forward:conditional method was used for logistic regression analysis.Logistic regression equation:logitp=2.604-0.008x prefrontal lobe integral value-0.201x PVF total score.The model chi square value was 18.006,P=0.000,the significance of Hosmer lemeshow test was 0.105,the prediction accuracy was 73.8%,the sensitivity was 72.5%,and the specificity was75.0%.The AUC of the two groups was 0.678（95%CI:0.560～0.795）.According to youden index,when the best diagnostic limit of prefrontal integral value is 67.90mmol.mm,the sensitivity is 80%and the specificity is 52.5%.Conclusions1.Patients with MDD have some degree of impaired cognitive function compared to healthy controls.2.Compared with healthy controls,the activation of prefrontal lobe and temporal lobe hemodynamics during the VFT in patients with major depressive disorder is lower,indicating that the changes of prefrontal and temporal hemodynamics may be a biomarker of major depressive disorder.3.Temporal lobe integral value and prefrontal lobe centroid value can specifically reflect the target symptom groups corresponding to the HAMD despair factor and HAMD retardant factor for patients with major depressive disorder.Decreased hemodynamic activation of the temporal lobe may be a pathophysiological marker of symptoms of incapacitation,inferiority complex,and feelings of hopelessness in patients with MDD.4.Prefrontal lobe integral value and PVF total score are good predictors to distinguish patients with major depressive disorder from healthy controls,the results are highly reproducible and facilitate widespread clinical application.