7 June 2012.An exhaustive review and subsequent meta-analyses of nearly 60 years’ worth of literature have allowed researchers at the University of Cambridge and King’s College London to draw an extraordinarily detailed picture of the incidence of schizophrenia and other psychotic disorders in England. The study, published online March 22 in PLoS One, is the largest of its kind ever conducted, and reveals that the overall incidence of first-onset schizophrenia and related psychoses has remained stable since 1950 at about 32 cases per 100,000 persons per year.
When the researchers examined the incidence of psychosis over this same period by age, sex, and ethnicity/geography, their findings were in concert with the international literature and with individual studies conducted in the U.K. When the data were analyzed by ethnicity, certain minority populations were found to have higher rates of psychosis compared to a baseline population of white British persons, especially black Caribbean and African migrants and their offspring, in which the incidence of psychotic disorders was “notably and consistently raised,” according to the authors.
The study’s first author, James Kirkbride of Cambridge, told SRF that he believes the findings will prove useful in shaping healthcare delivery in the U.K. “If we know how rates vary according to demographic factors like age, sex, and urbanicity of your environment, and also by ethnicity, then we can build reliable models based on the population structure of different regions to predict the expected incidence of psychosis over a given period,” Kirkbride said.
In the first step of the new study, the first of four commissioned by the U.K. Department of Health, Kirkbride and colleague Tim Croudace, working in Peter Jones’s group at Cambridge, combed through more than 5,000 abstracts related to first-onset psychosis that had been identified in a wide-ranging, systematic search of both the published literature and “grey literature” such as research reports commissioned by government agencies. After several rounds of inclusion criteria were applied, 83 publications that described patients from age 16 to 64 made the cut for subsequent analyses (publications addressing prevalence rather than incidence are being analyzed separately; see below).
“We knew roughly what papers we were looking for at the outset of the project. We could probably have written down beforehand half of those from our own knowledge,” Kirkbride said. “But it was the other half—the ones you didn’t realize contained incidence data—that were the interesting ones.” In the end, the process took over two years. “It was an ordeal. But we did it well, applying a systematic approach. We did it properly.”
Using novel statistical techniques that will be described in a forthcoming paper by study author Daniel Jackson, the researchers examined the incidence of psychotic disorders overall, as well as across time, by gender and age, by geography, and by the reported methodological quality of the individual studies making up the dataset. That there was no discernible variation in the pooled incidence of psychoses over time or when the reported study quality was taken into account is “reassuring,” said Kirkbride, who adds that the group’s findings in these realms dovetail with the international literature as well. But “one of the most beautiful parts of doing the work,” he said, was finding that the pooled incidence of diagnostic subcategories of psychosis—the incidence of affective psychoses such as bipolar disorder and psychotic depression was about half that of non-affective psychoses—lined up well with the group’s a priori assumptions.
“What we believed was that you should see pooled rates of all psychosis as highest overall, with rates of non-affective psychosis next highest, and then within the nonaffective psychoses, higher rates of schizophrenia than other types of non-affective disorder, and then the affective psychoses,” Kirkbride explained. “And when we mapped that out, the gradient conformed to exactly what we’d expect.”
In terms of gender and age, the findings for nonaffective psychoses conformed with previous work in British populations (e.g., Häfner et al., 1993), showing a “peak incidence for men and women in their twenties, declining thereafter for both sexes with a smaller, secondary peak in incidence for women from midlife” for non-affective psychoses, according to the researchers.
Among Kirkbride and colleagues’ more significant findings was its further support for, as they write, “one of the most frequently replicated and yet still controversial public health challenges in contemporary psychiatric epidemiology”: higher rates of psychosis among ethnic minorities. Compared to the baseline white British population, the researchers calculated pooled risk ratios for schizophrenia of 5.6 for the population of black Caribbean origin; 4.7 for those of black African origin; and a smaller, but still notable, risk ratio of 2.4 for South Asian groups.
These differences have been variously ascribed by researchers to genetic factors, the stresses of migration and relocation, urban living, and the effects of discrimination against minorities. While genetic factors cannot be completely ruled out, they are unlikely to be a primary cause of these differences, Kirkbride said. Research has shown no raised rates of incidence of psychosis in these groups’ countries of origin when compared to baseline groups (e.g., Hickling and Rodgers-Johnson, 1995; Bhugra, 1996; Mahy et al., 1999). On the other hand, these differences are unlikely to be explained solely by the stresses of migration, because they persist in subsequent generations among these groups (e.g., Coid et al., 2008). Kirkbride suggested a complex interplay of genetic risk factors and stressors related to migration (for first-generation minorities), discrimination (for subsequent generations), and urban environments.
One study examined by the authors, on which both Kirkbride and Croudace were investigators reported an increased incidence of substance-induced psychoses between 1979 and 1999, largely attributed to an increase in the use of cannabis, “but it’s hard to say whether that increase is genuine,” according to Kirkbride. “It started from a very low base, a tiny rate of substance-abuse-related psychosis in the 1970s, from about 1 in 100,000 to up to 3 or 4 cases per 100,000. So although it’s a three- or four-fold increase, you’re still talking about very small numbers.” Nonetheless, citing other recent work from Jones’s team ( Hickman et al., 2007),the authors advise continued attention to substance-induced psychoses, “given that model projections suggest any link between cannabis use and psychosis will begin to translate into tangible changes in incidence over the next decade.”
In addition to the statistical paper already mentioned, two other publications, one on the prevalence of psychosis and another offering an economic perspective on the assembled data, will be published over the next year or two.
Kirkbride said that this work, though labor-intensive, has benefited greatly from “a long tradition of psychiatric epidemiology in the U.K.,” particularly case registers devised during the 1970s and 1980s that recorded all patients in particular regions presenting with psychosis to any kind of mental health services offered by the National Health Service. “I think our data will allow policy makers to see is that there is quite a clear, varied and replicable epidemiology of psychotic disorder,” which should allow them to plan and commission mental health services—at least for psychosis—in a much more precise, ‘bespoke’ fashion based on the sociodemographic and environmental characteristics of different population catchment areas.”—Peter Farley.
Kirkbride JB, Errazuriz A, Croudace TJ, Morgan C, Jackson D, Boydell J, Murray
RM, Jones PB. Incidence of schizophrenia and other psychoses in England, 1950-2009: a systematic review and meta-analyses. PLoS One. 2012;7(3):e31660. Epub 2012 Mar 22.