M88. EVIDENCE FOR INFLAMMATION AS A PUTATIVE SHARED MECHANISM FOR INSULIN RESISTANCE AND SCHIZOPHRENIA

Benjamin Perry,Stephen Burgess,Hannah Jones,Stanley Zammit,Rachel Upthegrove,Amy Mason,Felix Day,Isobel Stewart,Claudia Langenberg,Nicholas Wareham,Peter Jones,Golam Khandaker
DOI: https://doi.org/10.1093/schbul/sbaa030.400
2020-04-01
Schizophrenia Bulletin
Abstract:Abstract Background Insulin Resistance (IR) predisposes to cardiometabolic disorders, which are common in schizophrenia and are associated with excess morbidity and mortality. The mechanisms of association remain unknown. We aimed 1) To use genetic data to examine the direction of association between IR and related cardiometabolic risk factors, and schizophrenia; 2) To examine whether inflammation could be a shared mechanism for IR and schizophrenia. Methods We used two-sample uni-variable Mendelian randomization (MR) to examine whether genetically-predicted IR-related cardiometabolic risk factors (Fasting insulin (FI), high-density lipoprotein (HDL), triglycerides (TG), low-density lipoprotein, fasting plasma glucose, glycated haemoglobin, leptin, body mass index, glucose tolerance and type 2 diabetes) may be causally associated with schizophrenia. We used the most recent summary statistics for genetic variants associated with schizophrenia and IR-related cardiometabolic risk factors from publicly-available large genome-wide association studies (GWAS). We used bi-directional MR to examine direction of association. To examine whether inflammation could be a shared mechanism for IR and schizophrenia, we first conducted a sensitivity analysis by performing MR using only cardiometabolic genetic variants that were also associated with inflammation, at genome-wide significance. Second, we used multi-variable MR (MVMR) to examine associations between cardiometabolic risk factors and schizophrenia after adjusting for genetically-predicted levels of C-reactive protein. Results In analyses using all associated genetic variants, genetically predicted levels of leptin were associated with risk of schizophrenia (OR=2.54 per SD increase in leptin; 95% CI, 1.02–6.31). In analyses using inflammation-related variants, genetically predicted levels of FI (OR=2.76 per SD increase in FI; 95% C.I., 1.31–6.17), TG (OR=2.90 per SD increase in TG; 95% C.I., 1.36–6.17), and HDL (OR=0.56 per SD increase in HDL; 95% C.I., 0.37–0.83) were associated with schizophrenia. The associations completely attenuated in MVMR analyses controlling for CRP. There was no evidence of an association between genetically-predicted schizophrenia liability and cardiometabolic factors. Discussion The IR phenotype of FI, TG and HDL could be associated with schizophrenia over and above common sociodemographic and lifestyle factors. This association is likely explained by a common inflammatory mechanism. Interventional studies are required to test whether inflammation could represent a putative therapeutic target for the treatment and prevention of cardiometabolic disorders in schizophrenia.
psychiatry
What problem does this paper attempt to address?