Cognitive insight in patients with depression and non-suicidal self-injury

Cognitive insight, a higher-order insight encompassing metacognitive processes, is directly linked to patients' help-seeking behavior and disease self-management. Non-suicidal self-injury (NSSI) and depression are closely comorbid major global public health problems. Existing studies show depressed patients with NSSI have executive dysfunction, which may further impair metacognitive abilities. However, research on cognitive insight in this population remains scarce to date. This study aimed to investigate cognitive insight deficits in these patients, to provide evidence for precise clinical evaluation, individualized treatment, and risky behavior intervention.

We enrolled 68 depressed patients with NSSI, 40 without NSSI, and 32 healthy controls. All completed the Beck Cognitive Insight Scale (BCIS); depressed patients also completed the Ottawa Self-Injury Inventory (OSI) and 24-item Hamilton Depression Rating Scale (HAMD-24). Participants were grouped by NSSI status, further divided into mild and severe NSSI subgroups, stratified by age into adolescent and adult subgroups. Data were analyzed via chi-square, t-test, ANOVA, Pearson correlation, and stepwise multiple linear regression.

In the total cohort, the severe NSSI subgroup showed lower self-reflectiveness and composite index, and higher total HAMD-24 score, cognitive disturbance and despair factors versus the mild subgroup (all P < 0.05). Composite index was negatively correlated with total HAMD-24 score (r = -0.366, P < 0.05). Regression revealed that cognitive disturbance factor independently predicted self-reflectiveness (β = -0.326, P < 0.05), and total HAMD-24 score predicted composite index (β = -0.377, P < 0.05). In adolescents, the severe NSSI subgroup showed lower self-reflectiveness and composite index scores, and higher total HAMD-24, cognitive disturbance, and diurnal variation scores versus the mild subgroup (all P < 0.05). BCIS were negatively correlated with total HAMD-24 score and diurnal variation (r = -0.514, -0.559, -0.505, -0.538; all P < 0.05). Regression showed total HAMD-24 score independently predicted self-reflectiveness (β = -0.734, P < 0.05), and diurnal variation factor predicted composite index (β = -0.583, P < 0.05). In adults, no significant between-group differences were found in BCIS (P > 0.05).

This study demonstrates that patients with depression and severe NSSI show impaired cognitive insight, which correlates significantly with depression severity. Notably, this association shows a clear age-dependent pattern, present only in adolescent patients. For adolescents with depression and NSSI, targeted interventions for depressive symptoms, cognitive impairment, and circadian abnormalities may improve cognitive insight and reduce NSSI risk.