| Objective:The metabolic abnormalities could be developed in patients with schizophrenia after the antipsychotic medication,which included weight gain,dyslipidemia,impaired glucose tolerance,and insulin resistance.The metabolic abnormalities could significantly increase the risk of cardiovascular diseases such as myocardial infarction and stroke in patients with schizophrenia,and also deteriorate the cognitive function of theirs.Metformin,as a well-investigated biguanides,has been widely used to treat diabetes.Our previous studies have found that metformin significantly reduced weight gain,insulin resistance,and alleviated dyslipidemia caused by antipsychotic drugs in patients with schizophrenia.Clinical evidence has found that metformin could improve the cognitive function in patients with depression comorbid type 2 diabetes.Meanwhile,results from animal studies have also illustrated that metformin could improve the spatial learning and memory in mice fed with a high-fat diet.In this study,we aimed to identify the characteristics of cognitive impairment in patients with schizophrenia comorbid weight gain in conjunction with neuroimaging,and to investigate the efficacy and potential mechanisms of metformin in alleviating the cognitive impairment in patients with schizophrenia comorbid weight gain.Method:(1)Cross-sectional study: 71 schizophrenia patients with weight gain(WG),41 schizophrenia patients without weight gain(NW),and 50 healthy controls(HC)were recruited in this study.Corresponding demographic and clinical data were then collected.Cognitive function was assessed using the MATRICS Consensus Cognitive Battery(MCCB).Also,the levels of serum lipid,fasting blood glucose(FBG),fasting insulin(INS)and insulin resistance index(IR)were then assessed.We used Pearson correlation analyses to investigate the correlations between metabolic indicators and cognitive function.Meanwhile,resting-state functional magnetic resonance imaging(f MRI)scanning was performed,and the differences in the amplitude of low-frequency fluctuations(ALFF),regional homogeneity(Re Ho),and voxel-wise whole-brain functional connectivity(FC)with dorsolateral prefrontal cortex(DLPFC)subregions and hippocampus as seeds were compared among the three groups.The correlations between cognitive function and corresponding f MRI measurements of brain regions involved in weight gain were also analyzed.(2)Longitudinal study: 72 schizophrenia patients with weight gain were included and randomized assigned to the metformin intervention group(intervention group,n=48)and the blank control group(control group,n=24)at a ratio of 2:1.The duration of the trial was for 24 weeks.The demographic and clinical information,the levels of serum lipid,FBG,INS,IR,the cognitive function assessed by the MCCB,and the f MRI measurements(ALFF,Re Ho and FC)were collected at the baseline,week12 and week 24.The changes in cognitive function and changes in metabolic measurements before and after follow-up between the intervention group and the control group were compared,as well as the group × time interaction effects of f MRI measurements between groups.We further analyzed the association of cognitive function with metabolic indexes and f MRI measurements.Results:(1)Cross-sectional study: The WG group had significantly higher body mass index(BMI),waist circumference,FBG,INS,and IR than the NW group(P<0.05).Results from the ANCOVA showed significant differences in speed of processing,attention/vigilance,verbal learning,visual learning,reasoning and problem solving,social cognition,and composite score among the three groups after controlling for age,sex and education(P<0.001).Post-hoc analyses showed that speed of processing,attention/vigilance,verbal learning,reasoning and problem solving,social cognition,and composite score in the WG group were significantly lower than those in the NW group.We also found that weight was negatively correlated with speed of processing(r=-0.200,P=0.034),and BMI was negatively correlated with verbal learning(r=-0.203,P=0.032)in patients with schizophrenia.In addition,the levels of triglyceride(r=-0.247,P=0.009)and low density lipoprotein cholesterol levels(r=-0.210,P=0.027)were negatively correlated with social cognition,while the level of high density lipoprotein cholesterol were positively correlated with social cognition(r=0.251,P=0.008).Compared with the NW group,the WG group exhibited significantly increased ALFF in the left angular gyrus,and increased Re Ho in the left orbital part of middle frontal gyrus/left orbital part of inferior frontal gyrus(ORBinf)and triangular part of inferior frontal gyrus(IFGtri).The FC between the DLPFC left A46 subregion and anterior cingulate cortex(ACC),between the right A46 subregion and left ORBinf/rectus/right hippocampus/ACC/left precuneus,between the right A9/46 v subregion and medial superior frontal gyrus/ACC were significantly reduced in the WG group in comparison with the NW group.Moreover,the WG group showed significantly decreased FC between the left hippocampus and left superior temporal gyrus/left supramarginal gyrus/right supramarginal gyrus,and decreased FC between the right hippocampus and right middle frontal gyrus/left supramarginal gyrus /left inferior parietal lobule/right inferior parietal lobule than those FCs in the NW group.After Bonferroni correction,the Re Ho values in the left IFGtri were negatively correlated with reasoning and problem solving(r=-0.293,P=0.008),the FC values of the left A46-ACC were positively correlated with reasoning and problem solving(r=0.321,P=0.003),and the FC values of the right A46-ACC were positively correlated with reasoning and problem solving(r=0.357,P<0.001)and composite score(r=0.339,P=0.002)in patients with schizophrenia.(2)Longitudinal study: There were 40 participants completed the 12-week follow-up in the intervention group compared to 23 in the control group,and 39 participants completed the 24-week follow-up in the intervention group compared to 20 in the control group.Linear mixed model(LMM)showed that there was a significant main effect of group in speed of processing,working memory,verbal learning,visual learning,and composite score(P<0.05).The simple effects of group of the LMM showed greater improvement in speed of processing and composite score after 12 weeks of metformin intervention(P<0.001).Meanwhile,greater improvement was observed in speed of processing,working memory,verbal learning,visual learning,and composite score after 24 weeks of intervention(P<0.05).These differences mentioned above still remained significant after excluding patients with an MCCB global deficit score <0.5.Results from the LMM for the metabolic indicators showed there was a significant main effect of group in weight(F=4.923,P=0.030).The simple effect of group of LMM showed that the level of total cholesterol(TC)in the intervention group was significantly lower than that in the control group after 12 weeks of metformin intervention(P=0.023),and weight in the intervention group was significantly lower than in the control group after 24 weeks of intervention(P=0.031).Correlation analyses showed that the decrease of TC from baseline to week 12 was associated with the increase of verbal learning from baseline to week 12(r=-0.342,P=0.014),and with better verbal learning performance at week 12(r=-0.338,P=0.015).The reduction of weight from baseline to week 24 was associated with the increase of attention/vigilance from baseline to week24(r=-0.298,P=0.040),and with better speed of processing(r=-0.309,P=0.033)and verbal learning(r=-0.305,P=0.035)scores at week 24.From baseline to week 12,the intervention group had significant increases in the FC of left A9/46d-right ACC/right middle cingulate cortex(MCC),the FC of left A46-right ACC,and the FC of right A46-right dorsolateral superior frontal gyrus/right middle frontal gyrus/MCC than the control group.Correlation analyses showed that the enhancements in these FCs from baseline to week 12 were associated with the increases of cognitive function from baseline to week 12/week 24,and with better cognitive performance at week 12/week 24.Mediation analyses showed that the reduction in weight from baseline to week 12 significantly mediated metformin-induced improvement in MCCB composite score at week 12(ab=1.441,P=0.048,95% CI: 0.143 to 2.996),and the increase of FC of left A9/46d-right ACC/right MCC from baseline to week 12 significantly mediated metformin-induced improvement in speed of processing at week12(ab=3.850,P=0.015,95% CI: 1.020 to 7.011),and the increase of FC of left A46-right ACC from baseline to week 12 significantly mediated metformin-induced greater improvement in verbal learning from baseline to week 24(ab=4.787,P=0.022,95% CI: 0.957 to 9.296).Conclusion:(1)Compared with schizophrenia patients without weight gain,schizophrenia patients with weight gain displayed more severe cognitive impairment,which correlated with increased weight,higher BMI,and higher level of lipid metabolism.(2)Patients with weight gain showed abnormal brain function in the left angular gyrus,left orbital frontal cortex and left IFGtri,and exhibited decreased voxel-wise FCs with DLPFC subregions and hippocampus as seeds to the brain regions associated with the salience network,reward system,default mode network,and frontoparietal network.Increased Re Ho in the left IFGtri,and decreased FC between DLPFC subregions and ACC might be the underlying pathophysiological mechanisms for more severe cognitive impairment in patients with schizophrenia comorbid weight gain.(3)Metformin was effective in improving cognitive function in patients with schizophrenia comorbid weight gain.In addition,the improvement in cognitive function was associated with reduced weight,and associated with enhanced voxel-wise FCs between DLPFC subregions and ACC/MCC,and also associated with enhanced voxel-wise FCs between DLPFC subregions.All the results indicated that the effect of metformin on weight and FC may be the potential neurobiological mechanism for the improvement of metformin on cognitive function in patients with schizophrenia comorbid weight gain. |
PDF Full Text Request