Comparative predictive efficacy of atherogenic indices on metabolic syndrome in patients with schizophrenia

Schizophrenia patients endure high risks of metabolic syndrome and related cardiovascular mortality. Evidence on comparing detective power among atherogenic indices of the metabolic syndrome in schizophrenia patients with antipsychotics treatment is still lacking.

We recruited 128 schizophrenia patients and collected blood samples to determine plasma levels of fasting glucose, total cholesterol, triglycerides (TG), high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol. Five components of metabolic syndrome were assessed. Atherogenic indices, such as atherogenic index of plasma (AIP), atherogenic coefficient (AC), Castelli's risk index-I (CRI-I) and Castelli's risk index-II (CRI-II), were calculated. The area under the receiver operating characteristics curve (AUC) and regression analysis were adopted to compare the detective power of each atherogenic index for metabolic syndrome. The optimal cutoff points using maximization of Youden's index and the positive likelihood ratios were calculated.

51 (39.8 %) had metabolic syndrome. AIP (0.2 ± 0.2 vs. 0.6 ± 0.2), AC (2.5 ± 0.9 vs. 3.4 ± 0.9), CRI-I (3.5 ± 0.9 vs. 4.4 ± 0.9,) and CRI-II (2.1 ± 0.7 vs. 2.6 ± 0.7) were higher in the group with metabolic syndrome (all p < 0.001). AIP had the highest AUC (0.845, 95 % CI: 0.770, 0.920). The optimal cut-off point of AIP to predict metabolic syndrome was 0.4 with the corresponding sensitivity 83.7 %, specificity 80.3 %, and positive likelihood ratio 4.2. Regression analysis revealed that only AIP significantly correlated with the metabolic syndrome (p < 0.001).

Among atherogenic indices, only AIP has superior discrimination for detecting metabolic syndrome in schizophrenia with antipsychotics treatment.