Generalizability of findings from four clinical cohort studies and a general population study to patients with bipolar I disorder in outpatient treatment in the Netherlands

Abstract Background Little attention has been paid to the generalizability of cohort studies in bipolar disorder (BD) to patient with BD in everyday clinical practice. Methods A sample of patients with bipolar I disorder (BD-I) treated at a Dutch outpatient clinic for BD were compared with Dutch participants with BD-I of four clinical cohort studies, and participants with BD-I in a general population study in the Netherlands, on sociodemographic and clinical characteristics. Results On many variables participants from the outpatient sample matched with those of the included studies. However, compared with participants of several of the clinical cohort studies, these outpatients were significantly younger, had an earlier age of onset of mood symptoms, and had a shorter duration of illness. Compared with participants in the general population study, outpatients had significant higher levels of education and less often lived together or were married. One cohort study reported much lower comorbidity rates of alcohol use disorders, drug use disorders, and anxiety disorders than in the outpatient sample. In contrast, comorbidity rates were higher in the population study. Limitations Due to methodological differences between studies, comparisons between several variables was limited, and for some variables data was lacking. Conclusions Our findings suggest that many findings from cohort studies and general population study in BD-I are generalizable to everyday clinical practice, especially mood disorder outpatient centers. However, differences between samples indicate some selection and referral bias.