American Journal of Epidemiology Advance Access originally published online on May 4, 2006
American Journal of Epidemiology 2006 163(11):979-981; doi:10.1093/aje/kwj170
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Invited Commentary |
Invited Commentary: Taking the Search for Causes of Schizophrenia to a Different Level
1 Department of Epidemiology, Joseph L. Mailman School of Public Health, Columbia University, New York, NY
2 Center for History and Ethics of Public Health, Department of Sociomedical Sciences, Joseph L. Mailman School of Public Health, Columbia University, New York, NY
Correspondence to Dr. Ezra Susser, Department of Epidemiology, Joseph L. Mailman School of Public Health, Columbia University, 722 West 168th Street, 15th Floor, New York, NY 10032 (e-mail: ess8{at}columbia.edu).
Received for publication February 6, 2006. Accepted for publication February 13, 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT References |
|---|
In recent years, epidemiologists have established major variations in the incidence of schizophrenia and have begun to investigate the causes of these variations. The report by Pedersen and Mortensen (Am J Epidemiol 2006;163:971–8) in this issue of the Journal examines the contribution of family-level factors to the urban-rural difference in the incidence of schizophrenia. Their results suggest that familial life in urban environments confers some effect that persists after families move to rural settings. Taking these findings together with those of previous studies, it appears that factors operating at the level of the social context, the family, and the individual may all contribute to the urban-rural difference in schizophrenia incidence. This work exemplifies an integrative, multilevel approach to epidemiologic research that employs principles central to eco-epidemiology and other, similar frameworks.
causality; cities; epidemiologic methods; family characteristics; risk factors; schizophrenia; urbanization
Epidemiologic research on the causes of schizophrenia is flourishing (1
). A robust series of studies has demonstrated substantial variations in incidence that may provide clues to etiology (for a comprehensive review, see McGrath et al. (2)). The incidence of schizophrenia is elevated among persons raised in urban areas, among offspring of older fathers, and among several immigrant groups in Western Europe. Yet another series of schizophrenia studies illuminates its developmental origins (1). Two successive natural experiments indicated increased incidence after early prenatal exposure to famine (3, 4). Some of the genes implicated in schizophrenia play a role in neurodevelopment (5, 6) and may interact with environmental factors, such as prenatal nutritional deficiency (7).
Much of what has been accomplished in the study of schizophrenia is the fruit of the labor of previous generations of epidemiologists, who bequeathed to us three valuable legacies. They established 1) prospective pregnancy/birth cohorts that continue to yield results 30–50 years later in life-course studies (8), 2) national registry systems for large-scale studies of health and disease, and 3) methods for studying the incidence and course of mental disorders in the community. If we use this inheritance wisely, we can expect more discoveries in the coming decade. It is therefore important for psychiatric epidemiologists to be well-positioned to mine these scientific treasures (9). We echo calls for a more creative and assertive stance in investigations of schizophrenia (10). Epidemiology has reached the point at which it is feasible to study causes at various levels of organization, during specific periods of development, and in specific historical moments (11).
Progress depends partially on the questions we ask. Our best-developed designs—cohort and case-control studies—focus on identifying factors that cause variation in disease risk among individuals within a population. It is equally important, however, to identify factors that cause variation in disease rates across populations and over time. These may be characteristics of groups that do not vary among individuals within a population (12). The family represents an intermediate level of organization between the individual and the community that is often overlooked.
The work of Pedersen and Mortensen in this issue of the Journal (13) exemplifies the possibilities for schizophrenia research. Their starting point is observed variation in the incidence of schizophrenia across social contexts—higher incidence in urban areas than in rural areas. They and other investigators have previously recognized the need to consider the urban context itself (e.g., the nature of social ties) (14, 15), the urban context in conjunction with individual risk factors (e.g., cannabis use) (16), the duration and timing of exposure to urban environs at successive points in development (17), and historical time trends (18). This study, though, poses a new and intriguing question: Is the increased incidence of schizophrenia in urban areas related to factors best understood at the level of the family?
The Danish national psychiatric registry provided Pedersen and Mortensen with an empirical data set of unparalleled scope for the study of schizophrenia outcomes. Their design rested upon use of urban birth of the nearest older sibling as an indicator of family-level urban factors. Their results showed that among persons born and raised in rural areas, risk of schizophrenia varied according to the place of birth of the nearest older sibling: It was higher if the older sibling had been born in an urban area (13). Apparently, familial life in urban environments confers some effect that persists after families move to rural settings. One plausible hypothesis is that toxic chemicals, accumulated and retained in the body following exposure over extended periods of time, can be transmitted from mother to fetus during gestation. This and other hypotheses put forward in Pedersen and Mortensen's paper are testable. They set the stage for further work in which family-level factors of interest are measured directly.
Of course, such results require caution and replication, as well as more direct analytical approaches (19, 20). We should also retain interest in factors existing at other levels, as it is unlikely that family-level factors explain all of the urban-rural difference in schizophrenia. In Pedersen and Mortensen's study (13), among persons who had been born and raised in urban areas, disease risk did not vary with the older sibling's place of birth. Moreover, in previous studies, disease risk varied with the duration of a person's childhood exposure to urban life (17). Taken together, these and other results suggest that social context, familial factors, and individual-level factors may all be important.
The originality of Pedersen and Mortensen's investigation and its broader import for research strategies in schizophrenia merits appreciation. These researchers are gradually teasing apart the contributions of different factors at various levels and time points to the urban-rural difference in schizophrenia incidence. This stance corresponds well with what we have referred to elsewhere as "eco-epidemiology" (9, 11, 21) and other proposed frameworks with similar features (22–24).
An eco-epidemiologic strategy encourages epidemiologists to take successive steps to integrate knowledge of factors at various levels of organization in order to formulate further cogent questions as the next crucial step for investigation. It is an iterative approach, in which sharply focused studies are conceived within a broader framework, and therefore does not require single, all-encompassing studies. Pursuit of findings on one level of organization leads to questions on other levels. With attempts to grasp the full range and sequence of causes, the gaps in our knowledge of schizophrenia and other illnesses will surely diminish.
Eco-epidemiology entails consideration of "upstream" causes that have been the focus of research on urban living and schizophrenia as well as "downstream" molecular and genetic causes. Current research on the association between paternal age at conception and schizophrenia (25), first suggested over 30 years ago (26), provides an illustration of investigating a downstream cause. Genetic and epigenetic studies are under way to illuminate the causal pathway, focusing largely on de novo genetic events in the male germ line which are associated with increased paternal age (27). In this phase of investigation, human molecular genetics and animal studies play an important role. The distribution of paternal age at conception, however, is partly an ecologic phenomenon: Societal factors influence the age of fathers when children are conceived. In a later phase, therefore, the paternal-age association with schizophrenia will lead to questions at such higher levels of organization as sociocultural context, at which point sociologic and demographic studies will play an important role. Once we better understand how such upstream factors act in conjunction with downstream factors, we can turn to implications for prevention.
What stands in the way of adopting such a strategy? As specialized researchers within institutions that foster narrow bands of scientific expertise, we often find it difficult to synthesize the full spectrum of findings for any given disease. For instance, genetic causes are very important in schizophrenia, but we have not observed much variation in incidence that could be attributable to population differences in genetic ancestry. The disease is found worldwide, and population isolates with very high incidence have not been identified. Schizophrenia does, however, exhibit marked variations in incidence, such as the urban-rural difference. The juxtaposition of these findings presents a puzzling picture that may have implications for etiology. One plausible hypothesis is that genetic causes include recurring de novo germ-line mutations which do not persist across many generations because of reduced fertility in schizophrenia. Such genetic causes need not vary much according to genetic ancestry, because vulnerable points for de novo mutations in the human genome are shared by all populations. This hypothesis has important implications for epidemiologic studies: We should investigate exposures that might plausibly increase the rate of de novo mutations, such as paternal age at conception (28), periconceptional exposure to famine, and cumulative parental toxic exposures prior to conception. Yet epidemiologists and geneticists alike have not usually integrated their observations, and as a result, they have formulated few hypotheses that might help resolve the apparent discrepancies.
Despite difficulties, there is growing recognition among epidemiologists of the importance of integrating the various elements of our branching discipline, bringing together those of us who study chronic and infectious diseases, physical and mental health, genetic and social causes, the individual life course, and historical time trends. Eco-epidemiology, still embryonic in its development, has this goal. We envision an epidemiology that subsumes these branches and conceives of both individuals and populations as systems, in order to understand the complex dynamics between causes on multiple levels. It should encourage understanding of how, across time, macro-level phenomena can manifest at the micro level and how micro-level phenomena can manifest at the macro level.
Our hope is that this way of thinking about epidemiology finds wider acceptance in the field, and that ultimately an integrative, multilevel epidemiology ceases to require a label for the sake of differentiation. The study by Pedersen and Mortensen (13) provides an example of the type of research it would engender. Epidemiologists are, quite literally and quite fortunately, taking the search for causes of schizophrenia to a different level.
|
ACKNOWLEDGMENTS |
|---|
For their insightful comments on earlier drafts, the authors thank Dr. Michaeline Bresnahan, Sarah Conover, Dr. Mary-Claire King, and Dr. Mervyn Susser.
Conflict of interest: none declared.
|
References |
|---|
|
TOPABSTRACTReferences |
|---|
- Murray R, Jones P, Susser E, et al. The epidemiology of schizophrenia. New York, NY: Cambridge University Press, 2003.
- McGrath J, Saha S, Welham J, et al. A systematic review of the incidence of schizophrenia: the distribution of rates and the influence of sex, urbanicity, migrant status, and methodology. BMC Med 2004;Apr 28:13. (Electronic article).
- Susser E, Lin S. Schizophrenia after prenatal exposure to the Dutch Hunger Winter of 1944–1945. Arch Gen Psychiatry 1992;49:983–8.
[Abstract/Free Full Text] - St Clair D, Xu M, Wang P, et al. Rates of adult schizophrenia following prenatal exposure to the Chinese famine of 1959–1961. JAMA 2005;294:557–62.
[Abstract/Free Full Text] - Millar J, Brown J, Maule J, et al. A long-range restriction map across 3Mb of the chromosome 11 breakpoint region of a translocation linked to schizophrenia: localization of the breakpoint and the search for neighbouring genes. Psychiatr Genet 1998;8:175–81.[Web of Science][Medline]
- Owen M, Williams N, O'Donovan M. The molecular genetics of schizophrenia: new findings promise new insights. Mol Psychiatry 2004;9:14–27.[CrossRef][Web of Science][Medline]
- Lewis S, Zammit S, Gunnell D, et al. A meta-analysis of the MTHFR C677T polymorphism and schizophrenia risk. Am J Med Genet B (Neuropsychiatr Genet) 2005;135B:2–4.[Medline]
- Susser E, Terry MB. A conception-to-death cohort. Lancet 2003;361:797–8.[CrossRef][Web of Science][Medline]
- Susser E, Schwartz S, Morabia A, et al. Psychiatric epidemiology: searching for the causes of mental disorders. New York, NY: Oxford University Press, 2006.
- McGrath J. Invited commentary: gaining traction on the epidemiologic landscape of schizophrenia. Am J Epidemiol 2003;158:301–4.
[Free Full Text] - Susser E. Eco-epidemiology: thinking outside the black box. Epidemiology 2004;15:519–20.[CrossRef][Web of Science][Medline]
- Schwartz S, Diez-Roux A. Causes of incidence and causes of cases—a Durkheimian perspective on Rose. Int J Epidemiol 2001;30:435–9.
[Free Full Text] - Pedersen CB, Mortensen PB. Are the cause(s) responsible for the urban-rural differences in schizophrenia risk rooted in families or in individuals? Am J Epidemiol 2006;163:971–8.
[Abstract/Free Full Text] - Faris R, Dunham H. Mental disorders in urban areas: an ecological study. Chicago, IL: University of Chicago Press, 1939.
- van Os J, Driessen G, Gunther N, et al. Neighbourhood variation in incidence of schizophrenia. Evidence for person-environment interaction. Br J Psychiatry 2000;176:243–8.
[Abstract/Free Full Text] - Lewis G, David A, Andreasson S, et al. Schizophrenia and city life. Lancet 1992;340:137–40.[CrossRef][Web of Science][Medline]
- Pedersen C, Mortensen P. Evidence of a dose-response relationship between urbanicity during upbringing and schizophrenia risk. Arch Gen Psychiatry 2001;58:1039–46.
[Abstract/Free Full Text] - Pedersen CB. No evidence of time trends in the urban-rural differences in schizophrenia risk among five million people born in Denmark from 1910 to 1986. Psychol Med 2006;36:211–19. (Epub 2005 Nov 23).[CrossRef][Web of Science][Medline]
- Matte T, Bresnahan M, Begg M, et al. The influence of birth weight variation between and within sibships on IQ at age 7. BMJ 2001;11:310–14.
- Begg M, Parides M. Separation of individual-level and cluster-level covariate effects in regression analysis of correlated data. Stat Med 2003;22:2591–602.[CrossRef][Web of Science][Medline]
- Susser M, Susser E. Choosing a future for epidemiology: II. From black box to Chinese boxes and eco-epidemiology. Am J Public Health 1996;86:674–7.
[Abstract/Free Full Text] - Krieger N. Epidemiology and the web of causation: has anyone seen the spider? Soc Sci Med 1994;39:887–903.[CrossRef][Web of Science][Medline]
- McMichael AJ. Prisoners of the proximate: loosening the constraints on epidemiology in the age of change. Am J Epidemiol 1999;149:887–97.
[Abstract/Free Full Text] - Davey Smith G. The uses of ‘Uses of Epidemiology’. Int J Epidemiol 2001;30:1146–55.
[Free Full Text] - Malaspina D, Harlap S, Fennig S, et al. Advancing paternal age and the risk of schizophrenia. Arch Gen Psychiatry 2001;58:361–7.
[Abstract/Free Full Text] - Hare E, Moran P. Raised paternal age in psychiatric patients: evidence for the constitutional hypothesis. Br J Psychiatry 1979;134:169–77.
[Abstract/Free Full Text] - Crow J. The high spontaneous mutation rate: is it a health risk? Proc Natl Acad Sci U S A 1997;94:8380–6.
[Abstract/Free Full Text] - Crow JF. The origins, patterns and implications of human spontaneous mutation. Nat Rev Genet 2000;1:40–7.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Welham, M. Isohanni, P. Jones, and J. McGrath The Antecedents of Schizophrenia: A Review of Birth Cohort Studies Schizophr Bull, May 1, 2009; 35(3): 603 - 623. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Bresnahan, M. D Begg, A. Brown, C. Schaefer, N. Sohler, B. Insel, L. Vella, and E. Susser Race and risk of schizophrenia in a US birth cohort: another example of health disparity? Int. J. Epidemiol., August 1, 2007; 36(4): 751 - 758. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||