Objective - Patients presenting with psychotic or mood disorders present with neuropsychological deficits such as executive and memory disturbance. Deficits of these functions have also been reported in patients presenting with alcohol use or substance use disorders. A large percentage of patients with non-affective psychotic or mood disorders present with a comorbid substance use disorder. These subjects are often a priori excluded from most neuropsychological studies. However, using such an exclusion criterion may induce a selection bias linked to the high prevalence of this dual diagnosis. It is therefore necessary to further explore the impact of substance abuse on neuropsychological performance in subjects with psychotic or mood disorders. Method - Patients consecutively hospitalized for a non-affective psychotic disorder or a mood disorder were included. A standardised method was used to collect information on addictive behaviour, clinical and social characteristics. DSM IV diagnoses, including those of substance use, were made using a structured diagnostic interview and all other available clinical and historical information collected during the hospital stay. Memory performance was tested using the Batterie d'Efficience Mnesique 84 (Battery of memory efficiency 84 items, BEM 84). Executive abilities were explored using the Wisconsin Card Sorting Test (WCST) and the Stroop test. ANCOVAs with cannabis use disorder or alcohol use disorder as main factor were used to examine associations with neuropsychological test scores. Results - We have included 77 patients fulfilling the diagnostic criteria for non-affective psychotic disorders (schizophrenia, schizoaffective disorder, delusional disorder, other psychotic disorder, n = 35) or mood disorders (n = 42). Among these patients, 273 % presented with a lifetime history of alcohol abuse/dependence (current prevalence : 14.3 %) and 23.4 % presented with a lifetime history of cannabis abuse/dependence (current prevalence: 11.7 %). We have assessed the specific impact of alcohol and cannabis use on neuropsychological performance. No significant differences on memory and executive performance were found between patients presenting with and without a lifetime history of alcohol abuse/dependence. These results were not modified after adjustment for potential confounding factors (age, gender, educational level, age at onset, diagnosis, current versus past addictive behaviour). Patients with a lifetime history of cannabis abuse/dependence had significantly higher (i.e. better performance) general BEM 84 score (F = 3.89, df = 1, p = 0.05), higher complex figure delayed recall scores (F = 6.62, df = 1, p = 0.01) and higher recognition scores (F = 3.9, df = 1, p = 0.05) than patients presenting without a lifetime history of cannabis use. After adjustment on covariables (age, gender, educational level, age at onset, diagnosis, current versus past addictive behaviour), the differences on memory performance between the two groups were no longer significant, the differences found before adjustment were mainly explained by the confounding effect of age. Patients presenting with a lifetime history of cannabis abuse/dependence had significantly lower interference scores on the Stroop test than subjects without cannabis use (F = 5.67, df = 1, p = 0.02). This finding was not modified after adjustment for confounding factors. Information on substance use was collected by interviewing the patient and was completed by using all other available source of information, but no urine testing was performed. Thus, substance use could have been underestimated or unrecognized in some patients. We did not distinguish patients who presented with substance abuse from those who presented with dependence because there were few of the latter. Distinguishing these two populations would be of interest because dependence may have a more deleterious effect than abuse in neuropsychological performances. Finally, we did not included normal control subjects so we can not assess if our cohort present with memory and executive deficits compared to normal subjects. Conclusion - Comorbid alcohol or cannabis abuse/dependence has limited effects on memory and executive abilities in subjects with psychotic or mood disorder. The only significant difference between subjects with and without a dual diagnosis was that subjects with cannabis use disorder performed poorly on the Stroop test. No other significant difference in executive and memory performance was found after adjustment for confounding factors. Since there is a high prevalence of a comorbid substance use disorder in subjects with psychotic or mood disorder, the exclusion of these patients in neuropsychological studies may not be systematically justified.
