Disordered eating attitudes in hospitalized schizophrenia patients of Guangdong, China: prevalence, associated factors, and the exploratory mediating role of insight in the depression-eating attitude link

Abstract

This study aimed to examine disordered eating attitudes and their associated factors among hospitalized patients with schizophrenia. The primary objective was to identify sociodemographic, clinical, and psychological factors associated with eating attitudes. The secondary objective was to explore, as an exploratory analysis, whether a statistically significant indirect association through insight was present between depressive symptoms and eating attitudes within the framework of the Common Sense Model of Self-Regulation, while acknowledging that the cross-sectional design does not allow causal inference. In this cross-sectional study, 300 Chinese long-term hospitalized patients with schizophrenia were recruited by convenience sampling method in South China from October 2022 to May 2023. The patients completed a questionnaire, including socio-demographics, Eating Attitudes Test-26, Insight and Treatment Attitudes Questionnaire, Hamilton Rating Scale for Depression, and Hamilton Rating Scale for Anxiety. The factors associated with eating attitudes were analyzed by multiple linear regression, and the exploratory mediating role of insight between depressive symptoms and eating attitudes was examined through the Bootstrap mediating effect. Given the positive skewness of the EAT-26 scores, a sensitivity analysis was additionally performed using a generalized linear model with a normal distribution, identity link function, and robust covariance estimator to assess the robustness of the primary regression findings. 22% of patients with schizophrenia had a score of 20 or above. Regarding the primary findings, univariate analysis showed that gender (P = 0.022), smoking (P = 0.008), weekly exercise frequency (P < 0.001), and diabetic status (P = 0.005) were independently associated with the eating attitudes of hospitalized schizophrenia patients. Regression showed eating attitudes were related to smoking (P = 0.003), while insight (P < 0.001), frequency of exercise (P < 0.001) and diabetes status (P = 0.018) were negatively related to the Eating Attitudes Test score. Depression was significantly correlated with an increasing Eating Attitudes Test score (P = 0.003). The final regression model explained 33.7% of the variance in eating attitudes (R² = 0.337). The sensitivity analysis using a generalized linear model with a normal distribution, identity link function, and robust covariance estimator yielded broadly consistent findings, with the same key variables remaining statistically significantly associated with eating attitudes. In the exploratory mediation analysis, depressive symptoms showed a significant total association with eating attitudes (total effect β = 0.35, P < 0.001), including a direct association (direct effect β = 0.21, P = 0.005) and a statistically significant indirect association through insight (indirect effect β = 0.15, P < 0.001). The mediating effect accounted for 41.4% of the total effect. However, this mediation finding should be interpreted as a statistical association rather than evidence of a causal pathway because of the cross-sectional study design. Hospitalized schizophrenia patients exhibit a high prevalence of disordered eating attitudes. Smoking, diabetes status, exercise frequency, insight and depression were independently associated with eating attitudes, and these findings were broadly robust to an alternative generalized linear modeling approach with robust covariance estimation. Exploratory mediation analysis identified a statistically significant indirect association through insight between depressive symptoms and eating attitudes. These results highlight the importance of integrating emotional, cognitive, and behavioral factors in the clinical management of schizophrenia. • Approximately 22% of hospitalized schizophrenia patients exhibited abnormal eating attitudes, indicating a high prevalence of disordered eating in this population. • Primary findings showed that smoking, diabetes status, exercise frequency, insight, and depression were independently associated with eating attitudes. Sensitivity analysis using a generalized linear model with robust covariance estimation yielded findings broadly consistent with the primary regression analysis. • As a secondary exploratory finding, mediation analysis suggested that insight statistically mediated the association between depression and eating attitudes, accounting for 41.4% of the total effect; however, this result should be interpreted cautiously because the cross-sectional design does not establish causal relationships.