Objectives: High rates of psychotic disorders and psychotic symptoms have been found in urban environments but reports for bipolar affective illness have been inconsistent, possibly due to failure to stratify for comorbid psychotic symptoms. It was hypothesised, therefore, that any effect of urbanicity on the bipolar phenotype would be moderated by comorbid psychotic symptoms.
Methods: In a random, representative population cohort of 7049 adults with no history of non-affective psychotic disorder, the cumulative incidence of bipolar and psychotic symptoms and syndromes, assessed with the CIDI, was examined over five levels of population density of place of residence. Similarly, the degree of comorbidity between broadly and narrowly defined bipolar phenotypes on the one hand, and the dichotomous presence of broadly (17.2%) and narrowly defined (3.8%) psychotic symptoms on the other, was examined as a function of population density of place of residence.
Results: The rate of bipolar disorder, however defined, was progressively higher in more urbanised areas. However, in models of bipolar phenotypes, a strong interaction between comorbid psychosis and level of urbanicity was apparent, indicating that the greater the degree of psychotic comorbidity, the greater the effect size of the urban environmental factor. For bipolar disorder without psychosis, no effect of urbanicity was apparent.
Conclusions: The results suggest differential environmental causal effects on affective and cognitive dimensions of bipolar psychopathology that are nevertheless strongly comorbid within the same categorically defined disorder, possibly due to the effect of shared genetic risk factors.