10 January 2007. How common is schizophrenia? The answer to that question has important public health consequences, but a precise number has been hard to come by. A common estimate for lifetime prevalence (LTP) is 1 percent, but recent studies have yielded a range of values from different populations, with a median somewhat lower at 0.4 percent (Saha et al., 2005).
Now, a comprehensive study by Jonna Perala and colleagues at the National Public Health Institute in Helsinki, Finland, pegs the prevalence of schizophrenia among Finnish adults at the 1 percent level. In the study, published in the January issue of the Archives of General Psychiatry, the researchers screened a large representative sample of Finnish adults for psychotic illness using multiple sources of information, including interviews, self-reports, medical exams, and register information. They find that a surprising 3 percent of Finns will experience any psychotic disorder in their lifetime.
In an accompanying commentary, John McGrath of the University of Queensland in Australia calls the work “arguably the most thorough study ever undertaken on the prevalence of psychotic disorders.
“The survey by Perala and colleagues reminds us just how common psychotic disorders are in the general population,” McGrath writes. “While they may be less prevalent than depression or substance abuse, we should probably not mislead people by labeling psychoses as low-prevalence disorders.” This fact lends ever more urgency to getting the most out of future epidemiological studies in this fast-moving field, he argues.
Combining strategies to dig deeper
Different methods of gauging the prevalence of psychotic illness, as well as shifting diagnostic criteria over time, can cause variations in prevalence estimates. Although schizophrenia is the dominant psychotic disorder, a significant proportion of other disorders, including schizoaffective disorder, bipolar disorder, major depressive disorder, substance abuse, and others can feature psychosis. Perala and colleagues sought to combine several screening methods to provide the most reliable estimate of disease, and to figure out the best ways of screening for psychotic illness in a general population.
Their nationally representative sample group consisted of 8,028 Finnish adults 30 years or older who were participating in a larger health study. Based on self-reports, physician exams, a structured interview (the Composite International Diagnostic Interview, or CIDI), and national registers detailing hospitalizations, prescription drug use, and disability, the investigators identified 746 people (9.29 percent) with possible psychotic illness.
These subjects were further evaluated and best-estimate diagnoses made after additional interviews (Structured Clinical Interview, or SCID-I) according to the DSM-IV and by examination of case notes for those with treatment histories. Case notes for those unable to attend the interview were also checked. All together, the researchers were able to assess 92.8 percent of the selected population by one of these methods. Because non-responders might have had more mental illness, the investigators also checked register diagnoses for that group. They found that 35.8 percent of the screen-positive population had any psychotic disorder, and report a lifetime prevalence for the general population of 3.48 percent. For schizophrenia only, the LTP was 0.87 percent among responders, with an increase to 1 percent when nonresponders were included. This value is higher than recorded in many recent studies, but comparable to previous Finnish studies that used register data.
“In this study we were able to overcome many of the methodological problems inherent general population studies of psychotic disorders,” Perala and coauthors write. Screening based on multiple sources gave higher coverage than previous studies that rely on just one method. For example, they say, a common approach that uses the CIDI alone would only have recognized one-quarter of people with psychotic disorders in this study. Registers were the most important and reliable source of information, and had the further advantage of letting the researchers estimate rate of illness in nonresponders, which pushed up prevalence by 12 percent. The use of case notes was also critical for making specific diagnoses. “Our results support previous suggestions that multiple sources of information are essential to estimate the LTP of psychotic disorders,” the authors write.
While the numbers cannot be extrapolated to other populations, they write, “We believe that our prevalence estimates are more accurate than those in previous studies and that screening from nationwide health care registers and use of case note information would have increased the case detection substantially in other general population studies as well.”
Exclusion of younger adults might have skewed some prevalence data, however. In women, the onset of schizophrenia occurs later than in men, which might have affected the sex-specific incidence. In this study, there was no difference between men and women, whereas other work indicates a higher prevalence among men.
A new epidemiology of schizophrenia
In his commentary, McGrath points out recent epidemiological studies are showing that schizophrenia “is not the egalitarian disorder we once thought it was.” Rather than the previous view—that schizophrenia prevalence varied little among different societies and cultures—it is now appreciated that the disease varies from country to country, and between men and women, so that “the epidemiology of schizophrenia is no longer a flat and featureless horizon,” he writes, but instead bears “surprisingly rich contours.”
Because of this, epidemiological studies have an important place in sorting out the environmental and social factors that increase schizophrenia risk, including urban birth, migrant status, prenatal factors, and others. McGrath calls on the schizophrenia research community to “generate a sense of urgency about unraveling the environmental risk factors contributing to the gradients.”
The possibility that growing urbanization all over the world could contribute to higher disease prevalence in the future should “galvanize” the research community, he says. “It is interesting to speculate about what the response would be if the health outcome was cardiovascular disease rather than mental illness. One would predict that the government funding agencies would invest heavily in projects aimed at understanding the mechanism of action linking the variables of interest.” He continues, “The fact that this has not yet happened for schizophrenia is a cause for concern.”
McGrath calls for several changes in epidemiological approaches, from broadening the categories of observation beyond narrow diagnostic criteria to “adding value” to epidemiological studies by collecting biological samples for genetic and biochemical add-on studies. Finally, he calls for epidemiology researchers to “build strong links with molecular, cellular and behavioral neurosciences,” which can provide the biological basis for explaining the causes of the disease.—Pat McCaffrey.
Perala J, Suvisaari J, Saarni SI, Kuoppasalmi K, Isometsa E, Pirkola S, Partonen T, Tuulio-Henriksson A, Hintikka J, Kieseppa T, Harkanen T, Koskinen S, Lonnqvist J. Lifetime Prevalence of Psychotic and Bipolar I Disorders in a General Population.
Arch Gen Psychiatry. 2007 Jan;64(1):19-28. Abstract
McGrath JJ. The surprisingly rich contours of schizophrenia epidemiology. Arch Gen Psychiatry. 2007 Jan;64(1):14-6. Abstract