Quoting from article:
At baseline, 66.6% of the psychosis patients smoked, compared with 38.3% of the siblings and 25.2% of the control subjects. The authors identified a significant association between smoking and lower performance in tasks related to processing speed, working memory, and/or problem solving in all three groups. For processing speed and problem solving, the deficits were dose-dependent; that is, the more cigarettes smoked per day, the lower the average performance.
Assessments at follow-up visits showed that patients who quit smoking had significant improvements in processing speed compared with those who did not quit smoking; patients who quit gained almost five points on the Wechsler Adult Intelligence Scale digit-symbol coding task. This association was not found in siblings or control subjects.