Smoking, symptoms, and quality of life in patients with psychosis, siblings, and healthy controls: a prospective, longitudinal cohort study

Background: The self-medication hypothesis postulates that the high prevalence of smoking in patients with psychosis can be explained by the ameliorating effect of smoking on symptoms. However, there are few large prospective studies testing this hypothesis. We aimed to examine the multi-cross-sectional and prospective associations of changes in smoking behaviour with symptoms and quality of life. Methods: In this prospective cohort study we recruited patients with a non-affective psychosis (n=1094), unaffected siblings (n=1047), and healthy controls (n=579). Patients aged between 16 and 50 years and diagnosed with a non-affective psychosis according to DSM-IV were recruited by clinicians from four university medical centres and 36 associated mental health-care institutions in the Netherlands and Belgium between Jan 13, 2004, and March 6, 2014. Smoking status and number of cigarettes per day were assessed at baseline, and at 3-year and 6-year follow-up using the Composite International Diagnostic Interview (CIDI). Symptom frequency was self-rated with the Community Assessment of Psychotic Experience (CAPE), and quality of life was assessed by the WHO Quality of Life (WHOQOL) schedule. Multiple linear mixed-effects regression analyses were done accounting for multiple confounders. Findings: At baseline, 729 (67%) of 1094 of patients smoked (mean 17·5 cigarettes per day, SD 8·8) compared with 401 (38%) of 1047 siblings and 145 (25%) of 579 healthy controls. Multi-cross-sectional results of linear mixed-effects analyses showed that smoking in patients and siblings was associated with more frequent positive symptoms (estimate 0·14, SE 0·02, p<0·0001 in patients; 0·03, 0·01, p=0·0019 in siblings), negative symptoms (0·15, 0·03, p<0·0001 in patients; 0·09, 0·02, p<0·0001 in siblings), and depressive symptoms (0·12, 0·03 p<0·0001 in patients; 0·08, 0·02 p<0·0001 in siblings) and lower quality of life (−0·59, 0·11, p<0·0001 in patients; −0·31, 0·09, p=0·0002 in siblings) than non-smokers. In controls, smoking was associated with significantly higher frequency of subclinical positive symptoms (0·03, 0·01, p=0·0016) and depressive symptoms (0·05, 0·03, p=0·0432) than in participants who did not smoke. Patients who started to smoke during follow-up showed a significant increase in self-reported symptoms, particularly positive symptoms (0·161, 0·077, p=0·0381), whereas smoking cessation was not associated with changes in symptoms or quality of life compared with those who showed no change in smoking behaviour. Similar results were obtained for the changes in the number of cigarettes smoked. Interpretation: Our findings do not empirically support the self-medication hypothesis. The absence of long-term symptomatic relief from smoking should encourage clinicians to help patients with psychosis to quit smoking. Funding: Dutch Health Research Council, Lundbeck, AstraZeneca, Eli Lilly, Janssen Cilag, Academic Psychiatric Center of the Academic Medical Center, GGZ inGeest, Arkin, Dijk en Duin, GGZ Rivierduinen, Erasmus Medical Center Amsterdam, GGZ Noord Holland Noord, University Medical Center Groningen, Lentis, GGZ Friesland, GGZ Drenthe, Dimence, Mediant, GGNet Warnsveld, Yulius Dordrecht, Parnassia Psycho-medical Center, Maastricht University Medical Center, GGzE, GGZ Breburg, GGZ Oost-Brabant, Vincent van Gogh voor Geestelijke Gezondheid, Mondriaan, Virenze riagg, Zuyderland GGZ, MET GGZ, Universitair Centrum Sint-Jozef Kortenberg, CAPRI University of Antwerp, PC Ziekeren Sint-Truiden, PZ Sancta Maria Sint-Truiden, GGZ Overpelt, OPZ Rekem, University Medical Center Utrecht, Altrecht, GGZ Centraal, and Delta.