Fig. 4: Partial correlation and logistic regression analyses in all the patients with schizophrenia after antipsychotic treatment.

a Significant correlations between fatty acid level changes and PANSS total score improvement. b Significant correlations between fatty acid ratio changes and PANSS total score improvement. c Forest plot of logistic regression of treatment response and fatty acid level changes. Gender, age, chlorpromazine-equivalent dose, number of episodes, and disease course are added to the regression model to control the influence of these confounding factors on antipsychotic responses. Partial Spearman’s rank correlation and binary logistic regression were used.