Several studies have suggested that nicotine could have beneficial effects on cognitive functioning.

To explore the association between smoking and cognitive functioning in bipolar disorder and schizophrenia.

Observational study. We analyzed data from 184 participants in 3 groups (61 healthy controls, 47 euthymic patients with bipolar-I disorder, 76 patients with clinically stable schizophrenia) assessed by a neuropsychological battery. Statistical analysis was performed comparing subgroups of smokers and non-smokers data. Both groups were demographic and clinically comparable.

Smoking rate was higher in participants with bipolar disorder and schizophrenia (X 2 = 26.8, p = 0.001). Bipolar Group: neuropsychological performance was similar in smokers (n = 28) and non-smokers (n = 19) except perseverative errors of the Wisconsin test, in which smokers performed significantly worse (t = −2.1, p = 0,03). Schizophrenia group: smokers (n = 46) performed significantly better than non-smokers (n = 30) in the verbal fluency test (t = −2.46, p = 0.046), finger tapping right (t = −2.19; p = 0.03) and immediate (t = −2.84, p = 0.006) and delayed recall (t = −2.59, p = 0.01) of Rey figure test.

This is the first study comparing these clinical groups. The global cognitive function in euthymic state was similar in both groups despite of smoking status. However, smoking could be associated with a worse executive function in bipolar disorder. In patients with schizophrenia smoking was associated with better performance in visual memory, verbal fluency and motor speed, but not attention.