Using lipidomic profiling to explore antipsychotic effects in schizophrenia

Lipid disturbances have been described in patients with schizophrenia and bipolar disorder. Additionally, the use of several antipsychotics is associated with weight gain, lipid disturbances, glucose dysregulation and the development of metabolic syndrome. However, the degree to which metabolic alterations occur in treatment with different antipsychotics is still unclear. In this study, we aim to characterize the lipidomic profiles in patients with schizophrenia; to identify antipsychotic- related changes and differences in the lipidomic profiles of schizophrenia patients during treatment; and to identify lipidomic profiles that are associated with clinical outcomes in schizophrenia patients. One hundred and three participants of the BeStInTro study, with a diagnosis within schizophrenia-spectrum were included and evaluated. Patients were treated with 3 different types of antipsychotics (Amisulpride, Olanzapine and Aripiprazole) and were clinically evaluated in 4 different time points all over a year: At the baseline, after 1 week, 2 weeks and 1 year of treatment. General lipids such as HDL LDL, total cholesterol and triglycerides were measured. Serum samples were used for the identification and quantification of lipid species. Lipidomic profiles were determined by liquid chromatography and mass spectrometry using the Complex Lipids Targeted Panel by Metabolon Inc. We used principal component analysis to investigate the effect of variables over the lipidomic profiles. Linear mixed models were used to study the relationship between changes over time in lipid profiles and treatment. Sex, medication, time and the interaction between medication and time were included as fixed effects. Body mass index was used as a covariable in the models. Amisulpride treatment group and the baseline appointment were chosen as references for the analyses. Fourteen main classes of lipids in serum (including 992 lipid species) were identified/quantified. After applying quality control criteria, we obtained 104 lipid species. Using principal component analysis, we observed an influence of the sex and BMI in the lipid levels, but no effect of technical bias related to storage time or batch effects. Using correlation analysis, we confirmed expected correlations between blood lipids (cholesterol and triglycerides) and some of the lipid classes such as Cholesteryl Esters (CE) and Triacylglycerols (TAG) and Diacylglycerols (DAG) at the baseline. Negative correlations between symptoms, and measurements of Ceramides (CER), Dihydroceramide (DCER), Lysophosphatidylcholine (LPC), Lysophosphatidylethanolamine (LPE), and phosphatidylethanolamine (PE) were also observed at baseline but not in the follow up assessments. In a preliminary analysis we observed a discrete effect of the treatment over time in lipid species from 10 different lipid classes, with small estimation coefficients. These differences will be further evaluated. We observed an effect of sex and BMI over lipidomic profiles in accordance to findings of previous studies with large cohorts. Similarly, to our results, some lipid species from Lysophosphatidylcholine (LPC) and Lysophosphatidylethanolamine (LPE) classes were positively associated with the reduction rates of PANSS positive subscore.