Abstract
BACKGROUND: Patients with schizophrenia are unaware of their cognitive impairments. Misperception of cognitive impairment is an important factor associated with real-world functional outcomes in patients with schizophrenia. The patient's family member plays a crucial role in detecting patients' cognitive impairments when the patients are unaware of their own cognitive impairments. Previous studies have reported that not only the patient's subjective rating, but also the patient's family members' rating of their cognitive impairment may not be precise. However, it is unclear why family ratings are inaccurate, and which factors impact family ratings. This study investigated whether family ratings differed significantly from the patients' subjective ratings of the patients' cognitive impairments and sought to determine the reason for the differences between the family ratings and the patients' neurocognitive performances. We investigated the relationship between patients' subjective ratings, family ratings for patients' cognitive impairments, neuropsychological performance, and other aspects, including premorbid IQ and clinical symptoms. METHOD: We evaluated 44 patients with schizophrenia for cognitive function using neuropsychological tests; in addition, both the patients and their families rated the patients' cognitive impairments through questionnaires. We used the Mann-Whitney U test to examine whether the family ratings differed significantly from the patients' self-reported ratings of their cognitive impairment. We conducted multiple regression analysis and structural equation modeling to determine why the patients' subjective ratings and the family ratings were not definitively associated with the patients' neurocognitive performances. We performed multiple regression analysis with a stepwise method with neurocognitive performance, premorbid IQ, positive symptoms, and negative symptoms as independent variables and family ratings of patients' cognitive impairments as dependent variables. RESULTS: We found that the family ratings differed significantly from the patients' subjective self-reported ratings of their cognitive impairments. Our results showed that the premorbid IQ of patients is the strongest predictor of family ratings. Furthermore, among the neurocognitive domains, only the processing speed of patients was associated with family ratings. CONCLUSIONS: We found that the family ratings were not consistent with the patients' subjective self-reported ratings and the family ratings were most affected by the patients' premorbid intellectual abilities. These results suggest that the families' current assessments of the patients' current cognitive impairments were affected by the patients' premorbid intellectual ability rather than the patients' current neurocognitive performance. Patients' processing speed predicted family ratings; however, family members' ratings were not related to verbal learning/memory, executive function, and language of patients. Therefore, our findings highlight that patients' family ratings may differ from patients' subjective ratings, results of performance-based neuropsychological tests, and clinician ratings.