Trajectory classes of violent behavior and their relationship to lipid levels in schizophrenia inpatients

To characterize the trajectory patterns of violence in schizophrenia inpatients, examine the relationships between the violence trajectories and baseline clinical features and lipid levels, and generate a model to predict the more violent trajectories.

In a sample of 107 consecutively admitted patients with schizophrenia spectrum disorders, violent behavior was weekly rated using the Violence Scale. The patients' blood levels of total cholesterol (TC), triglycerides (TG), low-density lipoprotein cholesterol, and high-density lipoprotein cholesterol were measured at admission. A trajectory analysis was used to classify the patients' longitudinal courses in violence, and the correlates of these trajectories were assessed using multinomial logistic regression analyses. A stepwise logistic regression was used to select the best predictor variables for the more violent trajectories.

Four violence trajectories of inpatients were obtained: class 1 (no violence, 37.4%), class 2 (low-leveling off, 39.2%), class 3 (high-falling sharply, 10.3%), and class 4 (high-falling slowly, 13.1%). Although the relationship between decreasing TC and TG levels and increased violence in the trajectory classes did not reach statistical significance, a decreasing trend in the proportion of high dichotomized-TG levels was significantly associated with more violence in the trajectory classes (p = 0.04). A five-variable model consisting of female gender, early onset, higher scores of positive symptoms, lower scores of negative symptoms, and low dichotomized-TC levels had a predictive accuracy of 0.85 (95% CI = 0.72-0.97).

Distinct violence trajectories exist in schizophrenia inpatients, and the more violent trajectories can be predicted using baseline clinical features and lipid levels.