Background: The self-medication hypothesis is commonly put forward to explain the high prevalence of smoking in psychiatric patients. However, studies supporting the self-medication hypothesis have most often been carried out on chronic patients stabilized by antipsychotics.
Aim: Given that antipsychotics tend to erase psychiatric symptoms, the present study was undertaken on acutely ill patients usually receiving no medications, or on whom medications are ineffective.
Methods: Participants were 492 consecutively hospitalized patients. They were evaluated the day of their hospitalization with the Brief Psychiatric Rating Scale (BPRS, 18 items). Urinary cotinine and creatinine were measured the morning following their hospitalization. The urinary cotinine/creatinine ratio and the cotinine/creatinine/number of cigarettes smoked per day ratio (nicotine extraction index) were calculated for each patient.
Results: The positive symptoms subscale of the BPRS significantly correlated with smoking, whereas other BPRS subscales did not. In patients with mood disorder, the nicotine extraction index correlated with the positive symptoms, activation and hostility subscales, but not with the negative symptoms subscale. Analyses of individual BPRS items using the cotinine/creatinine ratio measure showed that smoking is positively associated with "unusual thought content" and "grandiosity" items and negatively associated with "guilt feeling", "depressed mood" and "motor retardation". Analyses of individual BPRS items using the nicotine extraction index showed a positive association only with "unusual thought content" and "grandiosity" items. Patients with schizophrenia extract more nicotine from cigarettes than other patients.
Conclusion: In acutely ill psychiatric patients, smoking is linked with positive symptoms and not with negative symptoms.