Excerpts from article:
“The observation that higher rates of schizophrenia and other nonaffective psychotic disorders are associated with city living is robustly replicated across epidemiologic studies conducted throughout Northern Europe and North America. The weight of this evidence, combined with exhaustive investigations of potential bias, suggest that methodological artifacts are unlikely to explain such associations. Instead, they imply a causal role for underlying environmental factors more common in urban settings. Evidence from northern Europe strongly links social factors, including deprivation, inequality, and social isolation, to raised rates of nonaffective psychotic disorders in urban settings, and while mechanisms through which such factors influence psychosis remain uncharacterized, they may include effects of social stress and possibly exposure to socially distributed biological factors, such as poor nutrition, infections, and/or pollutants.
The first idea is consistent with emerging neurobiological evidence, which has found that healthy volunteers who live or grew up in more urban environments show heightened activity in the amygdala and perigenual anterior cingulate cortex, respectively, after stress. These 2 interconnected brain regions are involved in threat perception and social evaluative stress processing, which may be more demanding in response to city life, and which are associated with psychotic disorder. Nonetheless, a definitive statement about causality (or refutation of it) awaits results of ongoing efforts to produce convergent evidence that elucidates the exact mechanisms through which some or all of these factors account for differential rates of psychotic disorder in some urban populations.
It does not follow, however, that city living should be associated with psychotic disorders in all contexts. Indeed, recent data from southern Europe suggest that the association may not pertain even across European countries, while findings from rural England suggest urbanicity per se need not be the primary social construct driving increased incidence rates in some environments.
New data from the World Health Survey (WHS) published by DeVylder et al further expand our purview and underscore how mistaken it could be to assume that city living has the same meaning across the globe. Across 42 low- and middle-income countries (LMICs), they found no consistent pattern of urban-rural differences in the prevalence of subclinical psychotic experiences (PEs). However, as we expand the epidemiological horizon, we cannot afford to uncritically accept evidence that emerges on urban-rural differences. Studying PEs does not yield direct evidence about variation in psychotic disorders, given that PEs are relatively common, yet to be conclusively associated with polygenic risk for schizophrenia, and associated with risks for several other psychiatric disorders and have low positive predictive value for psychotic disorders. We must be even more cautious about WHS findings on psychotic disorders, which were based on a single-question item asking whether a physician had ever diagnosed survey participants with a psychotic disorder. In LMICs, access to clinical care is often very limited, and the diagnosis, communication, and understanding of the term psychotic disorder is likely to vary widely. Finally, WHS data are measures of prevalence rather than incidence. With respect to psychotic disorders in LMICs, where rural-to-urban migration has strongly contributed to rapid urbanization, urban-rural differences in prevalence could be influenced by reverse causation, if people who developed such disorders were less likely to migrate than their peers or more likely return home after psychosis onset.
The attempt to use the WHS data to study urban-rural differences in psychosis underscores the need for a new, global vista, long overdue in epidemiologic studies of city living and psychosis. Although these findings do not invalidate Northern European findings on urban living and associated risk of psychotic disorders, they do challenge us to investigate whether such gradients exist elsewhere and understand why or why not this may be the case.
Moreover, global investigations of PEs are important in their own right, given that they predict several adverse psychopathologies. In this regard, it is unfortunate that DeVylder et al were unable to contextualize cultural variation between and within these vastly different countries to explore reasons for the apparent heterogeneity in urban-rural differences in PE prevalence, instead favoring a meta-analytic approach which led to pooled, null effects. A multitude of factors may account for why PEs were more apparent in urban Laos, Estonia, and Mali and rural Vietnam, Hungary, and South Africa, which, given the large WHS sample sizes, were unlikely to be chance findings. For example, the social construction of urban environments in each of these countries is likely to vary, following complex historical, sociocultural morphologies that alter both the patterns of exposure to putative social or biological risk factors in such environments, as well as the very meaning of urban or rural living. More fundamentally, we know little about possible cross-cultural variation in endorsement of PEs between or within LMICs, which could have contributed to differential urban-rural patterns in the WHS data.
Despite these limitations, these results press us to build theory and collect in-depth empirical data that reveal the variety of meanings of city living across contexts, and explain how and why observed associations between city living and psychotic disorders differ across these contexts. A recent study of men in Chengdu, China, underscores how profoundly the meaning of city living might differ by sociocultural and historical context, where rapid urbanization has been driven by rural-to-urban labor migration over a single generation. Here, rural-to-urban labor migrants were not found to be at increased risk of severe psychotic symptoms indicative of likely disorder, but men born in more urban areas were. Such data allow us to generate new theories in this context. Perhaps internal migrants are a selectively resilient group relative to men born in Chinese cities, those left behind, and those who returned to rural areas. Perhaps early-life exposure to adversities in urban environments holds greater causative relevance for psychotic disorders than adulthood exposure. Perhaps settlement patterns for labor migrants in Chengdu foster strong bonding social capital, irrespective of material conditions.
Overcoming some of the challenges above will require us to expand carefully conducted incidence studies of psychotic disorder in LMIC settings. This could provide a falsifiable methodology for testing different causative theories in samples across different settings, with different patterns of genetic, social, and biological exposures. Such studies will also be vital for global public mental health, given the seismic shifts in population dynamics anticipated over the next 30 years; by 2050, the United Nations anticipates that two-thirds of the world’s population will live in cities, with 90% of this growth occurring in Africa and Asia. The WHS findings stimulate us to expand our scope of enquiry, and as such, they are important for psychosis research.
As with all public health issues, however, the costs of misinterpreting their results may have negative unintended consequences. Just as we should not expect a universal association between city living and psychotic disorders, nor should we infer from pooled results on PEs across heterogeneous, uncharacterized settings that such associations do not exist. Rather, we should anticipate and seek to understand the causes of heterogeneous associations between city living and psychotic disorders as we develop a global purview of the epidemiology of psychotic disorders in a fast-changing world.”
Kirkbride JB, Keyes KM, Susser E: City Living and Psychotic Disorders – Implications of Global Heterogeneity for Theory Development. JAMA Psychiatry 75(12):1211-1212 (2018).
