Skip Navigation



Epidemiologic Reviews Advance Access published online on July 30, 2008
Epidemiologic Reviews, doi:10.1093/epirev/mxn006
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
30/1/84    most recent
mxn006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by March, D.
Right arrow Articles by Susser, E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by March, D.
Right arrow Articles by Susser, E.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Epidemiologic Reviews © The Author 2008. Published by the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.

Article

Psychosis and Place

Dana March1, Stephani L. Hatch2, Craig Morgan3, James B. Kirkbride4, Michaeline Bresnahan1, Paul Fearon5 and Ezra Susser1,6

1 Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY
2 Department of Psychological Medicine, Institute of Psychiatry, King's College London, London, United Kingdom
3 Department of Health Service and Population Research, Institute of Psychiatry, King's College London, London, United Kingdom
4 Department of Psychiatry, Addenbrooke's Hospital, University of Cambridge, Cambridge, United Kingdom
5 Department of Psychological Medicine, Institute of Psychiatry, King's College London, London, United Kingdom
6 New York State Psychiatric Institute, New York, NY

Correspondence to Dana March, Department of Epidemiology, Mailman School of Public Health, Columbia University, 722 West 168th Street, New York, NY 10032 (e-mail: dm2025{at}columbia.edu).

accepted for publication April 29, 2008.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
One important line of epidemiologic inquiry implicating social context in the etiology of psychosis is the examination of spatial variation in the distribution of psychotic illness. The authors conducted a systematic review of evidence from urbanicity and neighborhood studies regarding spatial variation in the incidence of psychosis in developed countries since 1950. A total of 44 studies (20 of urbanicity and 24 of neighborhood) were culled from three databases with similar time frames: Medline (1950–2007), PsychInfo (1950–2007), and Sociological Abstracts (1952–2007). With a special emphasis on social factors potentially relevant to etiology, the authors elucidated contributions, limitations, and issues related to study design, measurement, and theory. Evidence from both arenas supports a possible etiologic role for social context. Studies of urbanicity indicate that early-life exposure may be important; dose-response relations, spatial patterning of schizophrenia, and interactions with other factors may exist. Neighborhood studies indicate heterogeneity in rates, hint at spatial patterning of schizophrenia, and offer intriguing evidence implying more proximal social (as opposed to physical) exposures. The authors encourage the exploration of social pathways engaging theory, methodological advances, and the life-course perspective. They also propose a conceptual shift from studies of spatial variation in outcomes to research addressing the etiologic effect of exposures shaped by place as a reservoir of risk or resilience.

geography • incidence • psychotic disorders • residence characteristics • schizophrenia • social environment


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Accumulating evidence implicates social context in the etiology of psychosis. One important line of epidemiologic research pointing to a potentially causal role of social context is the examination of spatial variation in the distribution of psychotic illness. Such investigations have a rich history, in which social theory featured prominently in the first half of the 20th century. Eclipsed for a period of time, interest in the etiologic role of social context has been revived (1), and a new era of research has begun, stimulated in part by methodological advances and studies of migration and race/ethnicity (2). However, more recent studies have been somewhat disconnected from potentially enriching social theoretical perspectives that shaped earlier investigations. As recent work has indicated (3), integrating methodological advances and social theory with the life-course approach (4, 5) can help in unraveling the meaning of important findings, such as the spatial distribution of psychosis. Toward that end, we conducted a systematic review of evidence regarding spatial variation in psychosis. Here we present our findings and critically discuss their implications.

We first sketch the history and current relevance of classic research investigating the spatial distribution of psychosis, and follow it with a synthesis and interpretation of incidence studies carried out in developed countries since 1950. With an emphasis on social factors potentially relevant to etiology, we elucidate contributions, limitations, and issues related to study design, measurement, and theory. Finally, in order to increase understanding of the etiologic role of social factors, we encourage movement from analyses examining the spatial distribution of outcomes to analyses examining places as exposures—from space to place.

History

Attempts to characterize the spatial distribution of psychosis have produced a rich history of empirical research. In the United States, these efforts can be traced back to the sixth decennial Census, conducted in 1840 (6). Though plagued with methodological problems, notorious errors, and misinterpretations later exposed by Edward Jarvis (7), it constituted the first attempt to enumerate the "insane" in the general population. Subsequently, Jarvis himself conducted a survey (8) marking the beginnings of descriptive psychiatric epidemiology in the United States (9). Later in the 19th century, other researchers put forward environmental conditions such as climate, meteorologic phenomena, and topographic features to explain the purported spatial distribution of insanity (10). Around the turn of the century, however, the emphasis on the physical environment began to shift to the social environment, and classifications of psychotic disorders began to emerge (9, 11).

The first half of the 20th century witnessed several seminal contributions to our understanding of the social and geographic distribution of psychosis. In 1932, Ödegaard (12) demonstrated a doubling of rates of first hospital admission for psychosis among Norwegian migrants to Minnesota, as compared with native Minnesotans and with Norwegians living in Norway. This work provided a foundation for investigations of migration and psychosis by generating several testable hypotheses to explain this pattern of findings, including selective migration, dismissed only recently as a generally applicable explanation (13).

Several years later, Faris and Dunham (14) published their classic ecologic study of psychosis in Chicago, Illinois, Mental Disorders in Urban Areas. This work explored the relation between the spatial distribution of psychosis and social organization, a construct formed from sociologic theories relating the social architecture of urban settings to various social problems, of which mental disorders constituted one set. Faris and Dunham posited social isolation and poor communication among people in various communities as sociologic explanations for mental disorders, which they empirically tested by applying the concentric zone model of urban organization, developed by Park and Burgess (15) a decade earlier (figure 1). In this model, social organization increased with distance from the epicenter. Inner urban zones consisted of the most disorganized and unstable communities, characterized by isolation and poor communication among residents; outer zones were the most organized and stable communities, with the least isolation and better communication. Faris and Dunham hypothesized and found an inverse relation between social organization and rates of schizophrenia: The least socially organized inner urban zones had the highest rates. Similar ecologic patterns were found in other cities in the United States (14, 16).


Figure 1
View larger version (21K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIGURE 1. The concentric zone model of urban organization. Reproduced with permission from Mental Disorders in Urban Areas by R. Faris and H. Dunham (University of Chicago Press, Chicago, Illinois, 1939) (14).

 
While best known for these ecologic findings, Faris and Dunham's Chicago study also examined ethnic density (i.e., proportion of a given ethnic group in a defined area), concentration of immigrants, and social class. Thus, in addition to the findings regarding social organization, Mental Disorders in Urban Areas provided a stimulus for two important lines of investigation. First, it gave rise to research centered on neighborhood composition (e.g., proportions of ethnic minorities and immigrants) and other aspects of the social environment in relation to severe mental illness (17). Second, it invigorated the debate regarding social class and mental illness, taken up in depth by Hollingshead and Redlich (18) a century after it was first noted by Jarvis in 1855 (8).

As research regarding the spatial distribution of psychosis continued to take shape, a prominent area of inquiry was defined by the question of whether lower social class was a cause or consequence of mental illness (19, 20). Two processes by which social class could result from mental illness are selection (the intergenerational process by which individuals are selected into various social positions before and during the prodromal phase of disorder) and drift (the intragenerational process by which mentally ill persons, after onset, ultimately occupy various social positions). While distinct, both have been folded into a general category of selection. Although the social causation-selection debate is still not entirely resolved, for schizophrenia, results overall tend to favor selection as the dominant explanation—that is, people with schizophrenia either drift into or are selected into lower social classes because of disability or discrimination. Such findings created a context wherein the findings of Faris and Dunham, which largely supported social causation, were increasingly viewed as having reached the wrong conclusions about the causal direction between social conditions and psychoses (20, 21). The skepticism extended to their findings regarding urban ecology: Were higher rates of schizophrenia in inner urban areas due to the social disorganization of such areas or to people with schizophrenia drifting or being selected into them? However, the fact that Faris and Dunham catalyzed thinking about the spatial and social patterning of severe mental illness beyond the scope of social selection, drift, and causation remains often overlooked.

Currency

Between the 1960s and the 1990s, with notable exceptions (22), a once-prominent social emphasis in etiologic research receded alongside a sharpening focus on biologic underpinnings. Studies of migration and psychosis in the United States (8, 14, 17) literally migrated to Western Europe (23) and, over time, revived interest in the role of social context in the etiology of psychosis. Interest in the spatial distribution of psychosis itself also expanded, exemplified by interest in urban areas and psychosis (24) and increasingly rigorous comparative studies of urban and rural areas (25, 26) and landmark studies of countries and cultures (27, 28). In a recent review, McGrath et al. (29) concluded that there is heterogeneity in rates. Given the rich history of and current interest in social etiology, a synthesis and critical examination of epidemiologic studies of the spatial distribution of psychosis is perhaps a timely contribution.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
We developed a strategy for systematically identifying studies regarding the spatial distribution of psychosis. The scope of our review comprised English-language literature published between 1950 and 2007. We examined three databases with similar time frames—Medline (1950–mid-November 2007), PsychInfo (1950–2007), and Sociological Abstracts (1952–2007)—to identify relevant literature. A broad range of terms and/or strings used to capture studies of spatial distribution (e.g., area, city, communit*, county, district, environment*, geograph*, locale, metro*, municipal*, neighborhood, place, region*, residence, rural, space, spatial, town, tract, urban*, ward) was combined with terms relating to psychosis (e.g., psychosis, psychotic, schizophrenia). In the initial search, approximately 28,000 references were returned.

We elected to restrict our analysis to studies of the incidence of diagnosed psychosis in developed countries. While prevalence and cross-sectional studies are both important and informative, incidence studies—including administrative incidence based on hospital admission or first contact with health-care services—are often better suited to studying etiology (5). Studies of incidence among developing countries or comparisons between developed and developing countries (27, 28) certainly form an important dimension of the spatial variation of psychosis. However, evidence in and across these settings merits a separate review, accounting for differences attributable to the processes of development itself. Limiting our review to studies conducted in the developed world allowed us to focus sharply on several key issues of current interest in the field.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Of the approximately 28,000 references returned in our search and articles cited in those publications, a total of 44 studies (26, 30–72) met the inclusion criteria. Thirty-nine of the 44 studies were identified with an initial title sift, followed by an abstract sift, and finally, a paper sift. Five additional studies (50–52, 56, 57) were identified in the paper sift by examination of references not returned in our search. Results were separated into two primary domains that constitute the vast majority of research in the developed world regarding the spatial distribution of psychosis: variation by level of urbanicity and variation by neighborhood. While urbanicity is a characteristic of neighborhoods, we considered the two as distinct entities in this review. In table 1, the characteristics and relevant principal findings of studies published in the last decade are categorized by urbanicity and neighborhood and presented according to location.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Findings from comparative epidemiologic studies of psychosis by urbanicity and neighborhood, 1997–2008

 
Results from studies included in this review suggest heterogeneity in the distribution of psychoses by urbanicity and neighborhood. Regarding this heterogeneity, there are two main questions that can be asked. The first deals with the direction of causation: Does exposure to or residing in certain areas increase the risk of psychosis or confer protection from psychosis? The second asks, If so, what factors are culpable? This review responds to these two questions by providing a systematic examination of the nature and quality of data that can address them. Rather than conduct a meta-analysis, which would focus on effect sizes of reported results, we opted to critically evaluate the validity and meaning of the literature.

Variation by urbanicity

In 20 studies, rates of psychosis were examined according to urbanicity (26, 30–48). All studies except the earliest, conducted in the United States by Eaton (30), were carried out in Western Europe: nine in Denmark (34–36, 38, 39, 43–45, 48)—the bulk by members of a research group using the Danish psychiatric registry; three in Sweden (26, 41, 42); three in the United Kingdom (37, 46, 73); two in the Netherlands (32, 33); and one each in Finland (47) and Italy (31). The evidence indicates an association between urban life and rates of psychosis; in most studies, urbanicity is associated with an approximately twofold increase in risk of psychosis, though associations indicating a risk increase as high as fourfold have been reported in early-onset cases (35). In only two (26, 31) of the 20 studies were no significant differences in rates of psychosis found according to urbanicity.

Studies of urbanicity have defined exposure to urban environs in two primary ways: by dichotomies, consisting of urban-rural comparisons, and by degrees, defined as three or more categories of urbanicity according to population density. Two studies included in this review examined urbanicity in terms of the urban-rural dichotomy (31, 37); in parallel, one study classified areas of interest as cities and noncities (73). In another investigation, researchers examined two urban areas, one of which was more urbanized, and a mixed urban-suburban-rural area (46). The majority of investigators defined urbanicity by degrees of population density, defined either as population per square kilometer (32, 33, 42) or, most commonly, number of inhabitants within a defined location (e.g., capital, city, or town) (26, 30, 34–36, 38, 39, 41, 43, 44, 48). Among studies in which urbanicity was defined categorically as opposed to dichotomously, investigators in more than one third reported a dose-response relation between exposure to urban environs and risk of psychosis (30, 32, 35, 36, 39, 42).

The timing of exposure has been examined in the urbanicity literature in some detail—a valuable methodological contribution to the characterization of etiologically relevant social factors. Analyses have been facilitated by the employment of psychiatric case registers linked to population registers and other registers. Although associations between population density and psychosis at illness onset have been reported (31, 36, 37, 42, 46), we focus on exposure prior to illness onset because such findings are less likely to be due to drift across places.

From studies assessing the impact of urbanicity as defined by population density prior to illness onset, there is good evidence that elevated rates in urban areas are not simply the products of drift. In a number of studies examining exposure at birth (32–35, 38, 39, 41 43–45, 47, 48, 73), investigators reported an approximate doubling in risk of psychosis in adulthood. Among these studies, several attempts to further refine understanding of the timing of exposure to urbanicity have been made. A classic study of Swedish conscripts by Lewis et al. (26) examined place of upbringing; men brought up in urban areas were more likely to develop schizophrenia than those brought up in rural areas. In another analysis employing the psychiatric case register in the Netherlands, Marcelis et al. (33) sought expressly to determine whether urbanicity operates before or around the time of illness onset. The highest rates of schizophrenia were seen among persons born in urban areas, regardless of urban residence at illness onset; in accord with other studies assessing risk of urban birth, there was an approximately twofold elevation in incidence. In contrast, persons not exposed to urban areas at birth but exposed around the time of illness onset showed no increase in risk (33). In one study using the psychiatric case register and the civil registration system in Denmark, Pedersen and Mortensen (39) probed the possibility of particular periods of vulnerability prior to illness onset and the role of cumulative exposure by examining urban residence at birth and over 5-year intervals during upbringing. In this study—the lone investigation examining duration of exposure to urbanicity in this review—risk increased with cumulative exposure to the most urbanized areas; persons who resided in the most urbanized areas during the first 15 years of life had the highest (nearly threefold) increase in risk (39). Finally, some of the most recent evidence from the Danish case register indicates that the risk(s) conferred by urban life prior to conception and/or birth may affect offspring born in areas less urbanized than those in which the family previously resided (44), providing evidence counter to drift and selection explanations. Together, these studies indicate that urbanicity confers a risk of psychosis early in the life course; moreover, they flag accumulation and persistence of exposure both across the life course and across generations as areas for further exploration.

The urbanicity effect does not seem entirely attributable to drift or selection. Whether the reported associations are artifacts of health-care service utilization should also be considered. In a number of early studies comparing urban and rural areas, service utilization cannot be dismissed as an explanation for the findings. More recent studies, which have assessed the timing of exposure, permit an evaluation. For instance, Pedersen and Mortensen (39) reported a linear relation between degree of urbanization at birth and risk of schizophrenia, given fixed urbanization at age 15 years, in a large population-based Danish cohort. Since the relation between exposure to increasing urbanicity at birth and psychosis cannot be explained alternatively by service use around the time of illness onset, such findings offer support that the observed associations are not completely explained by the availability of psychiatric services that may differ according to degree of urbanization.

Consistent with the findings of Faris and Dunham (14), two analyses indicated spatial patterning of schizophrenia, though not of affective psychoses. Eaton et al. (35) reported a dose-response relation between degree of urbanization and risk of schizophrenia but not risk of affective psychoses in Denmark. Kirkbride et al. (46) found three times more nonaffective psychoses and the highest rates of schizophrenia in the most urbanized (South East London) of three centers studied in the United Kingdom; this was not the case, however, for affective psychoses. Though these findings are not supported unanimously (32), accumulating evidence from several large, prospective, population-based observational studies (74–77) suggests an absence of spatiality in affective psychoses.

Despite reports of a dose-response relation, findings regarding timing of exposure, and some indication of diagnostic specificity with schizophrenia, little regarding what is etiologically operative about urban life is known. High rates of schizophrenia in two studies of highly urbanized areas in London were partly explained by elevated rates in ethnic minorities (37, 46), consistent with the growing body of literature indicating elevated incidence among ethnic minority and immigrant groups in developed countries (23, 78, 79). However, this does not account for the bulk of the findings in the urbanicity literature.

Several interactions reported in the urbanicity literature may offer clues to etiology. Persons with familial liability seem to be particularly at risk (80). van Os et al. provided evidence of an additive interaction between urbanicity and family history of schizophrenia and severe mental illness in the Netherlands (81) and in Denmark (43). Winter birth has also been reported to be a modifier of the urbanicity effect (73), which might indicate gestation as a period of vulnerability to factors associated with urban life, such as infection, toxins, or nutrition. Furthermore, more robust effects have been reported in early-onset cases (32, 35, 82) and in males (42, 82), though contrary reports (34) exist with respect to sex.

Other factors have been examined as potential mediators of the urbanicity effect. Neither obstetric complications (35, 41) nor childhood socioeconomic circumstances (41) appear to mediate the risk of psychosis conferred by urbanicity. Likewise, according to one report (45), traffic-related pollution in urban areas does not seem to mediate the association. However, these few reports cannot be considered to rule out such explanations entirely; more work addressing potential mediators is much needed in the urbanicity literature.

In sum, the urbanicity literature reviewed here creates a pattern of findings that does not seem attributable to either reverse causation or service utilization. The timing, consistency, and dose-response relation reported in some studies provide evidence for an etiologic effect of urbanicity. Overall, the studies of urbanicity concerned with causation support a hypothesis that exposure at some point early in the life course could play a role in the etiology of psychosis, yet the explanation for the apparently psychotogenic effect of urban life remains unknown. This literature demonstrates the need for theory to inform investigations of mediators of the urbanicity effect, ideally employing competing hypotheses. As is discussed below, further research might more directly measure and examine both physical (e.g., toxins, infections, and nutrition) and social exposures, as well as characteristics of and changes in urban social context itself over time, and their respective impact on the development of psychosis in a diverse range of settings.

Variation by neighborhood

In 24 studies, investigators directly compared the incidence of psychosis among neighborhoods and/or examined specific neighborhood-level characteristics in relation to the incidence of psychosis (49–72). Nineteen studies (49, 53–55, 58–72) were returned in our database search, and five additional studies were included because they were referenced (50–52, 56, 57) in one of the articles returned. Most studies were conducted in Western Europe: 11 presented results from studies conducted in England (50, 51, 54, 58–61, 63, 66, 68, 69), four from Germany (55–57, 71), three from the Netherlands (62, 65, 72), and one each from Scotland (64) and Sweden (67). Four studies were carried out in the United States (49, 52, 53, 70).

Like studies of urbanicity, neighborhood studies have relied on administrative geographic units. Neighborhood boundaries in this literature have been formed most commonly by community analogs such as districts, electoral wards, municipalities, and tracts. In very few studies (69, 72) were neighborhoods constructed on the basis of similarity of social characteristics, although these, too, relied on administrative units as bases. One of the limitations of exposure definition is thus the fact that administrative units serve as crude proxies for neighborhoods as they are defined and experienced in the lives of their constituents.

Compared with studies of urbanicity, far fewer neighborhood studies have systematically examined the timing of exposure. Thus, as a group, neighborhood studies are more vulnerable to interpretations of drift across places. Only two studies (52, 58) measured neighborhood both at birth and at illness onset. The most complete assessment of timing was made by Dauncey et al. (58), who also examined neighborhood during childhood and adolescence to address questions regarding lifelong residential mobility. A third study conducted by Veling et al. (72) measured neighborhood at illness onset, and also accounted for drift by determining whether cases were residing in their parental home at first treatment contact. All other studies assessed neighborhood at first treatment contact and could not entirely dismiss drift as an explanation for variation in rates of psychosis. To date, no neighborhood studies have directly examined the duration of exposure in relation to psychosis. In order to determine whether exposure to or living in certain neighborhoods is important in the etiology of psychosis, the timing and duration of exposure must be systematically examined; absence of this systematic evaluation constitutes a primary limitation of the neighborhood literature.

Investigators in a number of studies reported spatial variation in psychoses at the neighborhood level. Aside from drift and selection, which cannot be eliminated conclusively as an explanation in many analyses, what can explain the reported variation in rates? The neighborhood literature, which has relied more on social theory to inform analyses, offers some of the most intriguing evidence that the variation may be due to group-level social processes. One signal, ripe for further investigation in a range of settings, comes from research regarding neighborhood composition; such investigations lie at the crossroads of spatial variation, immigration, and ethnicity. In three studies included in this review, researchers reported a significant protective effect of ethnic density (i.e., percent composition of persons of a given ethnicity in a geographic area) (53, 63, 72), consistent with the initial findings of Faris and Dunham (14). An early study of Italian Americans conducted by Schwartz and Mintz (53) in Boston, Massachusetts, neighborhoods found that rates of schizophrenia among Italian Americans were lower in communities with higher proportions of Italian-American residents. Two more recent studies provided even stronger evidence that the protective effect conferred by ethnic density is both social and group-level. In the African-Caribbean population in South East London, Boydell et al. (63) found an inverse linear relation between ethnic density and rates of schizophrenia that remained even after adjustment for neighborhood-level deprivation. A multilevel analysis conducted by Veling et al. (72) in The Hague, Netherlands, also found that ethnic density conferred a significant protective effect, even in the most deprived areas. Recently, investigators in several studies have examined other neighborhood-level social structural correlates of ethnicity, such as ethnic fragmentation (i.e., segregation), and have found direct relations with psychosis (64, 68).

Other findings also suggest that the nature of neighborhood social ties may be related to the distribution of psychosis. The earliest relevant report included in this review is that of Jaco (49), who found direct relations between several indicators of social isolation in communities in Texas and high rates of schizophrenia, which accords with other evidence (83). More recently, social capital—community-level characteristics that foster participation for mutual benefit—has been tested by several investigators (65, 67, 68). In two (67, 68) of the three studies, researchers found significant inverse relations between social capital and incidence of psychosis; in one report (68), this included both schizophrenia and other nonaffective psychoses. These results remained significant after adjustment for individual-level variables and neighborhood-level deprivation.

It is unclear whether, at the neighborhood level, population density is related to risk of psychosis. A classic study conducted by Hare (51) found a significant relation between population density and psychosis in neighborhoods in Bristol, United Kingdom, while a more recent study by Kirkbride et al. (66) in South East London did not. Temporality could not be established, and thus ambiguity at the level of neighborhood remains.

A number of other neighborhood characteristics have been examined in relation to psychosis. The most commonly examined in these studies was socioeconomic deprivation, the factor most subject to drift across places. In nine studies (52, 58–61, 64, 65, 67, 71), researchers directly examined the association with deprivation; overall, the studies suggested a relation between neighborhood-level deprivation and incidence of psychosis. One early study in the United States found the overwhelming majority of schizophrenia incidence in neighborhoods that were in the lowest half of their socioeconomic categories (52). Another investigation in Nottingham, United Kingdom, found the rate of schizophrenia in the most deprived neighborhoods to be nearly three times that in the least deprived areas (58). Two other recent studies conducted in Sweden (67) and Scotland (64) found a linear relation between neighborhood-level deprivation and rates of psychosis. In another study conducted in Nottingham, Croudace et al. (61) demonstrated an excess incidence (i.e., a higher proportion than expected) of psychosis in neighborhoods with above-average social deprivation, which suggests a possible threshold effect. Only one analysis of data collected in the Netherlands failed to find an association between neighborhood-level deprivation, defined at first contact, and incidence of schizophrenia (65). Together, these findings suggest that neighborhood deprivation is related to heterogeneity in rates of psychosis. However, these results are still subject to a drift interpretation, because people might have moved into deprived neighborhoods before first contact for treatment.

As suggested by studies addressing socioeconomic deprivation, some findings (e.g., ethnic density, social isolation) in the neighborhood literature may be more prominent for schizophrenia; for other psychoses, the association is less clear. Some neighborhood studies also hint at spatial patterning of schizophrenia. Kirkbride et al. (61) recently demonstrated significant neighborhood variation in incidence rates of schizophrenia, but not affective psychoses, in the concentrated urban area of South East London, and they reported that the observed variation was not due to individual-level factors, which accords with some urbanicity studies (35, 46). Consistent with the findings of Faris and Dunham (14), investigators in several earlier studies reported the highest concentration of schizophrenia to be in the inner city (54, 56, 57); stability in spatial patterning has been reported in Mannheim, Germany (56), though this remains relatively unexplored in other locations. As with other findings in the literature reviewed here, drift across places is a plausible explanation for the findings suggesting spatial patterning at the neighborhood level.

To summarize, the great majority of these studies report heterogeneity in incidence rates of psychosis at the neighborhood level in developed countries. Among the most intriguing findings in the neighborhood literature are those reported from studies of ethnic density. These findings indicate an interaction between the individual context and the social context, and they offer the possibility of a group-level social explanation for spatial variation, as opposed to an effect mediated by purely material factors, such as toxins. Moreover, these findings are also the least vulnerable to drift and selection interpretations, which pose barriers to causal inference in the neighborhood literature reviewed here. If drift or selection were an explanation, one might expect an increase in psychosis rates in high-ethnic-density, socioeconomically deprived areas; however, studies of ethnic density show a fairly consistent protective effect, which is more consistent with social causation. The measurement of exposure—both timing and duration—constitutes one of the primary limitations of neighborhood-level studies of psychosis incidence. While the use of administrative boundaries to define neighborhoods enables this sort of research, such boundaries do not necessarily capture neighborhoods as they are experienced and shaped by their inhabitants. One of the strengths of the neighborhood studies reviewed here is the employment of social theory to define constructs that form the basis for the definition of covariates. Social context itself has been measured in greater detail; a corollary strength is that social groups have been examined within neighborhoods, which helps shape our etiologic knowledge about social processes themselves.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In developed countries, spatial variation in rates of psychosis appears to be common in studies of both urban areas and neighborhoods. However, several differences exist between these two domains. The most basic distinction is that urbanicity is usually a broad and undifferentiated construct, which permits only general conclusions about the etiologic role of urban living; the neighborhood literature, in contrast, consists of studies of specific contextual factors—narrower constructs, such as ethnic density and social capital. Another notable difference is the statistical approach employed to model variation. In studies of urbanicity, variation is often modeled as a few parameters in the mean specification, which permits estimation of the direction of the variation. The mean is not specified in neighborhood studies in which all variation is modeled—except those that specify a variable (e.g., percent minority composition)—thus, estimation of its direction is prohibited (see Torrey et al. (48) for a comparison). Finally, the findings from each arena are distinct. Urbanicity studies point to early life as a period of elevated risk. Evidence largely from neighborhood investigations indicates that both risk and protective social factors at the group and individual levels may be etiologically relevant. Though they both suggest that social context might play a role in etiology, each domain is subject to methodological limitations, particularly inconsistent use of designs that permit the determination of causal direction and eliminate selection and drift as potential explanations for observed spatial variation.

An interesting finding noted here and observed consistently for nearly 70 years is the spatiality of nonaffective psychoses and the apparent absence of spatiality of affective psychoses. In this review, we have observed that this finding holds both between settings (46) and within settings (14, 66). While both disorders have been associated with ethnic minority status (78), putative explanations for factors associated with nonaffective psychosis, specifically schizophrenia—such as ethnic density (63, 72) and societal-level discrimination (84)—are inherently spatial. Still, whether these two broad sets of disorders are etiologically distinct per their respective spatial distributions remains to be conclusively established. It is possible that unspecified factors at other levels (e.g., the family) may explain the distinction between affective and nonaffective psychoses. Given the phenomenological overlap with affective psychoses, studies of common mental disorders may offer clues: After accounting for individual- and family-level variation, researchers have consistently failed to observe significant neighborhood-level variation (85, 86). Understanding the full specification of effects across multiple levels of organization in psychosis research may offer parallel insights with respect to spatial variation.

Social pathways: from space to place

The findings highlighted in this review beg further characterization of social and contextual factors, as we simply do not know what processes produce the observed associations, or at what level. In studies of the spatial variation of psychosis, two distinct questions are often conflated, which complicates the determination of causality. The first question, whether there is spatial heterogeneity of the outcome, is a descriptive question that is particularly important for establishing the demand for psychiatric services. Processes of group-level selection and individual-level drift probably contribute to spatial heterogeneity of psychosis, though each has been assessed largely in cross-sectional studies. The second, whether exposure either to or in certain conditions (e.g., urban environs) produces variation in the outcome, is the central etiologic question. Answering this question is challenging and requires understanding of the conditions produced by broader social processes; characterization of specific social and contextual factors and knowledge of the interrelations between factors that form social processes is essential. The multilevel processes responsible for spatial variation in psychosis operate alongside those that produce conditions resulting in relevant exposures, and may overlap. As a result, disentangling extant evidence to answer the etiologic question is difficult.

However, this inherently challenging problem and the limitations of the studies reviewed here can be minimized if not overcome. One possibility for future research is to focus on key social pathways—the social processes by which exposures of interest are created and maintained—that may affect the development of psychosis. Social pathways comprise cascades of social processes across multiple levels that produce a variety of conditions in a given place, which shape exposures more proximal to the individual (87).

Place, then, is broadly defined as a reservoir of risk or resilience. As a reservoir, place consists of the natural and built environment, physical structures, and material resources that shape experience within a designated geographic location. The physical and social architecture of place both shapes and reflects relationships among individual inhabitants, social groups, and social structures and institutions. Place thus harbors the dynamics of power, varies by culture, and changes over time. Given these characteristics, the study of social pathways requires consideration of place, since both risk and protective factors accumulate therein.

Social pathways might operate etiologically in two ways, which are not mutually exclusive. First, social processes might create conditions that ultimately result in physical exposures. For instance, as a number of authors have hypothesized, the association between urban birth and psychosis could be attributable in part to exposure to infection during gestation. Large-scale social processes of migration into and urbanization in a range of social, political, and economic contexts could produce a variety of conditions, including crowding, which could increase person-to-person contact and thereby alter the dynamics of infectious disease transmission. Therefore, a pregnant woman living in crowded conditions is more likely to be exposed to infections, such as influenza, that have been associated with adult-onset psychosis in offspring (88–91). Similarly, living in a deprived neighborhood could increase risk of psychosis through exposure to pollutants or toxins (45). Structural discrimination and its associated social processes relegate members of minority groups and/or lower social classes to living in neighborhoods with unhealthy conditions—for example, proximity to highways, landfills, and other potential sources of pollutants and toxins (92–94).

Second, social processes might operate directly, in the absence of physical exposures. Studies of ethnic density suggest social processes resulting in exposures more proximal to the individual that are social, as opposed to physical, in nature. Ethnic density could confer a protective effect with respect to psychosis through group-level social cohesion that provides a buffer against interpersonal discrimination and social exclusion, which may be etiologically important (84). Ethnic density might also partially provide a buffer against structural discrimination, since its protective effect seems to hold even in the most deprived neighborhoods (63, 72). Ultimately, whether social pathways result in individual-level exposure that is physical or social in nature, social processes at some level are etiologically important.

Research strategies

Understanding social pathways, and therefore relevant exposures, requires systematic investigation of social processes and the conditions they create in a given place at a particular period in time. Integration of the life course perspective (95) to better understand both critical and sensitive periods of exposure (96) is also necessary (97). For example, a number of urbanicity studies indicate that early life may be the critical time of exposure with respect to psychosis, meaning that urbanicity—or whatever exposure is truly operative—cannot have an etiologic effect later in life. However, urbanicity or the factor(s) responsible for the observed associations may also operate as a precipitating factor closer to illness onset (98), and thus early life may be a sensitive period—a period during which risk is simply greater. In contrast, little is known about the relevant timing of exposure in the neighborhood investigations of psychosis reviewed here. For instance, studies of ethnic density have examined exposure around the time of illness onset (63, 72), although the observed protective effect might be exerted much earlier in life, either during gestation or during childhood and adolescence.

Although determining the relevant timing of exposure poses an enormous challenge, it is possible with longitudinal birth cohort data that can be linked to area characteristics, such as population density or neighborhood ethnic density, over the life course. Using this type of data would permit assessments of whether changes in context itself matter, especially for persons living for extended periods in one place that changes over time. It would also facilitate assessments of whether residential mobility matters. However, two circumstances must be considered: Contexts change at various rates, and people move, both within and between contexts, at different rates. Examining prospective population data in urban areas before and after the influence of key social processes, such as urban development or renewal and changing racial or ethnic composition, could enrich our understanding of the relative impact of these processes.

Both risk and protective social factors in psychosis research could be examined more rigorously over the life course with the assistance of theory, much as the study of complicated factors in developmental psychopathology has been (99). Classical studies of psychosis tested theories to better understand etiology. For example, Faris and Dunham (14) examined social organization, while Jaco (49) and others (83) explored social isolation as potential explanations for psychosis. In recent studies, such as those that have examined the relation between social capital and psychosis (65, 67, 68, 100), investigators have begun to employ social theory to understand risk and protective factors. This could be explored further to distinguish among the group-level construct (101), the individual-level conceptualization (102), and other seemingly distinct constructs (e.g., social fragmentation and social isolation).

If grounded in theory and integrated with the life-course perspective, the study of social pathways could capitalize on conceptual and methodological advances to help investigators determine whether and how certain places act as reservoirs of risk for—or protection from—psychosis. We could determine the conditions under which the exposures ultimately produced by social pathways might give rise to different predictions; both between- and within-location and social group strategies might be employed gainfully (103), and differential spatial distributions of affective and nonaffective psychoses could be probed further. Moving beyond the question of whether spatial heterogeneity exists in the outcome to the more difficult question of whether place-based social processes are etiologically relevant would effectively constitute a conceptual shift from space to place. Such a shift could offer a dual advantage: By understanding how social processes at various levels of organization create the conditions in a given place that shape individual-level exposures, whether physical or social, we could better characterize etiologically relevant exposures and uncover potential points of public health intervention.

Once we have a better handle on what social processes result in etiologically relevant exposures, we can systematically integrate this with investigations of the distribution of psychoses across places to better understand spatial variation in the outcome. In this way, we will be better able to understand how processes of social causation operate with processes of social selection and drift to produce observed distributions of various psychoses across social strata, geographic locations, and time periods.

The study of social pathways should not be viewed as an alternative to studying biologic pathways. Rather, the two are complimentary. One particularly accessible example is the well-documented association between paternal age at conception and schizophrenia in offspring (104–111). Societal norms shape paternal age at conception; links between paternal age and de novo mutations have been put forward to explain the reported relation to schizophrenia (105, 109, 112).

Studying how various social processes create certain conditions in a given place—essentially, how exposures become emplaced—could be a fruitful strategy for understanding social causes, whether they result in more proximal physical or social exposures. Furthermore, investigation of the nature and timing of exposure to putative social causes could point to potential biologic mechanisms, including gene-environment interactions (80), de novo mutations (112), and epigenetics (113), among others. Understanding how what is emplaced ultimately becomes embodied could facilitate the integration of social and biologic knowledge, and thus elucidation of the full range and sequence of causes of psychosis and potential points of intervention.


   ACKNOWLEDGMENTS
 
This work was supported in part by a training grant (T32-MH-13043) from the National Institute of Mental Health (D. M.). S. L. H., C. M., and P. F. were supported by the NIHR Biomedical Research Centre for Mental Health, The South London and Maudsley NHS Foundation Trust.

Conflict of interest: none declared.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Allardyce J, Boydell J. Review: the wider social environment and schizophrenia. Schizophr Bull (2006) 32:592–8.[Abstract/Free Full Text]
  2. Fearon P, Morgan C. Environmental factors in schizophrenia: the role of migrant studies. Schizophr Bull (2006) 32:405–8.[Abstract/Free Full Text]
  3. Morgan C, McKenzie K. Society and psychosis. Fearon P, ed. (2008) Cambridge, United Kingdom: Cambridge University Press.
  4. Eaton WW. The sociology of mental disorders. (2000) New York, NY: Praeger Publishers.
  5. Susser E, Schwartz S, Morabia A, et al. Psychiatric epidemiology: searching for the causes of mental disorders. (2006) New York, NY: Oxford University Press.
  6. Cohen P. A calculating people: the spread of numeracy in early America. (1982) Chicago, IL: University of Chicago Press.
  7. Jarvis E. Insanity among the colored population of the free states. J Insanity (1852) 8:268–82.
  8. Jarvis E. Insanity and idiocy in Massachusetts: report of the Commission in Lunacy. (1971) Cambridge, MA: Harvard University Press. (1855).
  9. Grob G. The origins of American psychiatric epidemiology. Am J Public Health (1985) 75:229–36.[Abstract/Free Full Text]
  10. White W. Geographical distribution of insanity in the United States. J Nerv Ment Dis (1903) 30:257–79.[CrossRef]
  11. Grob G. Origins of DSM-I: a study in appearance and reality. Am J Psychiatry (1991) 148:421–31.[Abstract/Free Full Text]
  12. Ödegaard Ö. Emigration and insanity. Acta Psychiatr Neurol Scand. (1932) 4(Suppl):1–206.
  13. Selten JP, Cantor-Graae E, Slaets J, et al. Ödegaard's selection hypothesis revisited: schizophrenia in Surinamese immigrants to the Netherlands. Am J Psychiatry (2002) 159:669–71.[Abstract/Free Full Text]
  14. Faris R, Dunham H. Mental disorders in urban areas. (1939) Chicago, IL: University of Chicago Press.
  15. Park R, Burgess E, McKenzie R. The city. (1925) Chicago, IL: University of Chicago Press.
  16. Schroeder CW. Mental disorders in cities. Am J Sociol (1942) 48:40–7.[CrossRef]
  17. Malzberg B. Social and biological aspects of mental disease. (1940) Utica, NY: New York State Hospitals Press.
  18. Hollingshead A, Redlich F. Social class and mental illness. (1958) New York, NY: John Wiley and Sons, Inc.
  19. Goldberg E, Morrison S. Schizophrenia and social class. Br J Psychiatry (1963) 109:785–802.[Abstract/Free Full Text]
  20. Dunham H. Community and schizophrenia: an epidemiological analysis. (1965) Detroit, MI: Wayne State University Press.
  21. Kohn M. Social class and schizophrenia: a critical review. J Psychiatr Res (1968) 6:155–73.[CrossRef]
  22. Dohrenwend B, Levav I, Shrout P, et al. Socioeconomic status and psychiatric disorders: the causation-selection issue. Science (1992) 255:946–52.[Abstract/Free Full Text]
  23. Cantor-Graae E, Selten J. Schizophrenia and migration: meta-analysis and review. Am J Psychiatry (2005) 162:12–24.[Abstract/Free Full Text]
  24. Freeman H. Schizophrenia and city residence. Br J Psychiatry (1994) 164:39–50.[Web of Science]
  25. Torrey EF. Schizophrenia and civilization. (1980) New York, NY: Jason Aronson Publishers.
  26. Lewis G, David A, Andreasson S, et al. Schizophrenia and city life. Lancet (1992) 340:137–40.[CrossRef][Web of Science][Medline]
  27. Sartorius N, Jablensky A, Korten A, et al. Early manifestations and first-contact incidence of schizophrenia in different cultures. Psychol Med (1986) 16:909–28.[Web of Science][Medline]
  28. Jablensky A, Sartorius N, Ernberg G, et al. Schizophrenia: manifestations, incidence, and course in different cultures. A World Health Organization ten-country study. Psychol Med Monogr (1992) 20(Suppl):1–97.
  29. 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) 2:13–34.[CrossRef][Medline]
  30. Eaton WW. Residence, social class, and schizophrenia. J Health Soc Behav (1974) 15:289–99.[CrossRef][Web of Science][Medline]
  31. Thornicroft G, Bisoffi G, De Salvia D, et al. Urban-rural differences in the associations between social deprivation and psychiatric service utilization in schizophrenia and all diagnoses: a case-register study in Northern Italy. Psychol Med (1993) 23:487–96.[Web of Science][Medline]
  32. Marcelis M, Navarro-Mateu F, Murray RM, et al. Urbanization and psychosis: a study of 1942 –1978 birth cohorts in the Netherlands. Psychol Med (1998) 28:871–9.[CrossRef][Web of Science][Medline]
  33. Marcelis M, Takei N, van Os J. Urbanization and risk for schizophrenia: does the effect operate before or around the time of illness onset? Psychol Med (1999) 29:1197–203.[CrossRef][Web of Science][Medline]
  34. Mortensen PB, Pedersen CB, Westergaard T, et al. Effects of family history and place and season of birth on the risk of schizophrenia. N Engl J Med (1999) 340:603–8.[Abstract/Free Full Text]
  35. Eaton WW, Mortensen PB, Frydenberg M. Obstetric factors, urbanization and psychosis. Schizophr Res (2000) 43:117–23.[CrossRef][Web of Science][Medline]
  36. Schelin E, Munk-Jorgensen P, Olesen A, et al. Regional differences in schizophrenia incidence in Denmark. Acta Psychiatr Scand (2000) 101:293–9.[CrossRef][Web of Science][Medline]
  37. Allardyce J, Boydell J, van Os J, et al. Comparison of the incidence of schizophrenia in rural Dumfries and Galloway and urban Camberwell. Br J Psychiatry (2001) 179:335–9.[Abstract/Free Full Text]
  38. Pedersen CB, Mortensen PB. Family history, place and season of birth as risk factors for schizophrenia in Denmark: a replication and reanalysis. Br J Psychiatry (2001) 179:46–52.[Abstract/Free Full Text]
  39. Pedersen CB, Mortensen PB. Evidence of a dose-response relationship between urbanicity during upbringing and schizophrenia risk. Arch Gen Psychiatry (2001) 58:1039–46.[Abstract/Free Full Text]
  40. van Os J, Hanssen M, Bijl RV, et al. Prevalence of psychotic disorder and community level of psychotic symptoms. Arch Gen Psychiatry (2001) 58:663–8.[Abstract/Free Full Text]
  41. Harrison GL, Fouskakis D, Rasmussen F, et al. Association between psychotic disorder and urban place of birth is not mediated by obstetric complications or childhood socioeconomic position: a cohort study. Psychol Med (2003) 33:723–31.[CrossRef][Web of Science][Medline]
  42. Sundquist K, Frank G, Sundquist J. Urbanisation and incidence of psychosis and depression. Br J Psychiatry (2004) 184:293–8.[Abstract/Free Full Text]
  43. van Os J, Pedersen CB, Mortensen PB. Confirmation of synergy between urbanicity and familial liability in the causation of psychosis. Am J Psychiatry (2004) 161:2312–14.[Abstract/Free Full Text]
  44. Pedersen CB, Mortensen PB. Are the cause(s) responsible for urban-rural differences in schizophrenia risk rooted in families or in individuals? Am J Epidemiol (2006) 163:971–8.[Abstract/Free Full Text]
  45. Pedersen CB, Mortensen PB. Urbanization and traffic related exposures as risk factors for schizophrenia. BMC Psychiatry (2006) 6:1–7.[CrossRef][Medline]
  46. Kirkbride JB, Fearon P, Morgan C, et al. Heterogeneity in incidence rates of schizophrenia and other psychotic disorders: findings from the 3-center ÆSOP Study. Arch Gen Psychiatry (2006) 63:250–8.[Abstract/Free Full Text]
  47. Haukka J, Suvisaari J, Lonnqvist J. Regional variation in the incidence of schizophrenia in Finland: a study of birth cohorts born from 1950 to 1969. Psychol Med (2001) 31:1045–53.[CrossRef][Web of Science][Medline]
  48. Torrey EF, Mortensen PB, Pedersen CB, et al. Risk factors and confounders in the geographical clustering of schizophrenia. Schizophr Res (2001) 49:295–9.[CrossRef][Web of Science][Medline]
  49. Jaco EG. The social isolation hypothesis and schizophrenia. Am Sociol Rev (1954) 19:567–77.[CrossRef]
  50. Hare E. Family setting and the urban distribution of schizophrenia. J Ment Sci (1956) 102:753–60.[Web of Science][Medline]
  51. Hare EH. Mental illness and social conditions in Bristol. J Ment Sci (1956) 102:349–57.[Web of Science][Medline]
  52. Lapouse R, Monk MA, Terris M. The drift hypothesis and socioeconomic differentials in schizophrenia. Am J Public Health (1956) 46:978–86.[Free Full Text]
  53. Schwartz DT, Mintz NL. Ecology and psychosis among Italians in 27 Boston communities. Soc Probl (1963) 10:371–4.[CrossRef][Web of Science]
  54. Giggs J, Cooper J. Ecological structure and the distribution of schizophrenia and affective psychosis in Nottingham. Br J Psychiatry (1987) 151:627–33.[Abstract]
  55. Maylath E, Weyerer S, Häfner H. Spatial concentration of the incidence of treated psychiatric disorders in Mannheim. Acta Psychiatr Scand (1989) 80:650–6.[CrossRef][Web of Science][Medline]
  56. Weyerer S, Häfner H. The stability of the ecological distribution of the incidence of treated mental disorders in the city of Mannheim. Soc Psychiatry Psychiatr Epidemiol (1989) 24:57–62.[CrossRef][Web of Science][Medline]
  57. Weyerer S, Häfner H. The high incidence of psychiatrically treated disorders in the inner city of Mannheim. Soc Psychiatry Psychiatr Epidemiol (1992) 27:142–6.[Web of Science][Medline]
  58. Dauncey K, Giggs J, Baker K, et al. Schizophrenia in Nottingham: lifelong residential mobility of a cohort. Br J Psychiatry (1993) 163:613–19.[Abstract/Free Full Text]
  59. Harrison J, Barrow S, Creed F. Social deprivation and psychiatric admission rates among different diagnostic groups. Br J Psychiatry (1995) 167:456–62.[Abstract/Free Full Text]
  60. Boardman AP, Hodgson RE, Lewis M, et al. Social indicators and the prediction of psychiatric admission in different diagnostic groups. Br J Psychiatry (1997) 171:457–62.[Abstract/Free Full Text]
  61. Croudace TJ, Kayne R, Jones PB, et al. Non-linear relationship between an index of social deprivation, psychiatric admission prevalence and the incidence of psychosis. Psychol Med (2000) 30:177–85.[CrossRef][Web of Science][Medline]
  62. van Os J, Driessen G, Gunther N, et al. Neighbourhood variation in incidence of schizophrenia. Br J Psychiatry (2000) 176:243–8.[Abstract/Free Full Text]
  63. Boydell J, van Os J, McKenzie K, et al. Incidence of schizophrenia in ethnic minorities in London, United Kingdom: ecological study into interactions with environment. BMJ (2001) 323:1–4.[Abstract/Free Full Text]
  64. Allardyce J, Gilmour H, Atkinson J, et al. Social fragmentation, deprivation, and urbanicity: relation to first-admission rates for psychoses. Br J Psychiatry (2005) 187:401–6.[Abstract/Free Full Text]
  65. Drukker M, Krabbendam L, Driessen G, et al. Social disadvantage and schizophrenia: a combined neighbourhood and individual-level analysis. Soc Psychiatry Psychiatr Epidemiol (2006) 41:595–604.[CrossRef][Web of Science][Medline]
  66. Kirkbride JB, Fearon P, Morgan C, et al. Neighbourhood variation in the incidence of psychotic disorders in Southeast London. Soc Psychiatry Psychiatr Epidemiol (2007) 42:438–45.[CrossRef][Web of Science][Medline]
  67. Lofors J, Sundquist K. Low-linking social capital as a predictor of mental disorders: a cohort study of 4.5 million Swedes. Soc Sci Med (2007) 64:21–34.[CrossRef][Web of Science][Medline]
  68. Kirkbride JB, Morgan C, Fearon P, et al. Neighbourhood-level effects on psychoses: re-examining the role of context. Psychol Med (2007) 37:1413–25.[Web of Science][Medline]
  69. Giggs J. Mental disorder and ecological structure in Nottingham. Soc Sci Med (1986) 23:945–61.[CrossRef][Web of Science][Medline]
  70. Jaco EG. Social factors in mental disorders in Texas. Soc Probl (1957) 4:322–8.[CrossRef][Web of Science]
  71. Löffler W, Häfner H. Ecological pattern of first admitted schizophrenics in two German cities over 25 years. Soc Sci Med (1999) 49:93–108.[CrossRef][Web of Science][Medline]
  72. Veling W, Susser E, van Os J, et al. Ethnic density of neighborhoods and incidence of psychotic disorders among immigrants. Am J Psychiatry (2008) 165:66–73.[Abstract/Free Full Text]
  73. Takei N, Sham PC, O'Callaghan E, et al. Schizophrenia: increased risk associated with winter birth and city birth—a case-control study in 12 regions within England and Wales. J Epidemiol Community Health (1995) 49:106–7.[Free Full Text]
  74. Wainwright N, Surtees P. Area and individual circumstances and mood disorder prevalence. Br J Psychiatry (2004) 185:227–32.[Abstract/Free Full Text]
  75. Scully PJ, Owens JM, Kinsella A, et al. Schizophrenia, schizoaffective and bipolar disorder within an epidemiologically complete, homogeneous population in rural Ireland: small area variation in rate. Schizophr Res (2004) 67:143–55.[CrossRef][Web of Science][Medline]
  76. Mortensen PB, Pedersen CB, Melbye M, et al. Individual and familial risk factors for bipolar affective disorders in Denmark. Arch Gen Psychiatry (2003) 60:1209–15.[Abstract/Free Full Text]
  77. Pedersen CB, Mortensen PB. Urbanicity during upbringing and bipolar affective disorders in Denmark. Bipolar Disord (2006) 8:242–7.[CrossRef][Web of Science][Medline]
  78. Fearon P, Kirkbride JB, Morgan C, et al. Incidence of schizophrenia and other psychoses in ethnic minority groups: results from the MRC ÆSOP Study. Psychol Med (2006) 36:1541–50.[CrossRef][Web of Science][Medline]
  79. Veling W, Selten JP, Veen N, et al. Incidence of schizophrenia among ethnic minorities in the Netherlands: a four-year first-contact study. Schizophr Res (2006) 86:189–93.[CrossRef][Web of Science][Medline]
  80. Krabbendam L, van Os J. Schizophrenia and urbanicity: a major environmental influence—conditional on genetic risk. Schizophr Bull (2005) 31:795–9.[Abstract/Free Full Text]
  81. van Os J, Hanssen M, Bak M, et al. Do urbanicity and familial liability coparticipate in causing psychosis? Am J Psychiatry (2003) 160:477–82.[Abstract/Free Full Text]
  82. 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.[CrossRef][Web of Science][Medline]
  83. Kohn M, Clausen J. Social isolation and schizophrenia. Am Sociol Rev (1955) 20:265–73.[CrossRef]
  84. Veling W, Selten JP, Susser E, et al. Discrimination and the incidence of psychotic disorders among ethnic minorities in the Netherlands. Int J Epidemiol (2007) 36:761–8.[Abstract/Free Full Text]
  85. Weich S, Holt G, Twigg L, et al. Geographic variation in the prevalence of common mental disorders in Britain: a multilevel investigation. Am J Epidemiol (2003) 157:730–7.[Abstract/Free Full Text]
  86. Propper C, Jones K, Bolster A, et al. Local neighbourhood and mental health: evidence from the UK. Soc Sci Med (2005) 61:2065–83.[CrossRef][Web of Science][Medline]
  87. March D, Morgan C, Bresnahan M, et al. Conceptualising the social world. In: Society and psychosis—Morgan C, McKenzie K, Fearon P, eds. (2008) Cambridge, United Kingdom: Cambridge University Press. 41–57.
  88. Takei N, van Os J, Murray RM. Maternal exposure to influenza and risk of schizophrenia: a 23-year study from the Netherlands. J Psychiatr Res (1995) 29:435–45.[CrossRef][Web of Science][Medline]
  89. Yolken R, Torrey EF. Viruses, schizophrenia, and bipolar disorder. Clin Microbiol Rev (1995) 8:131–45.[Abstract]
  90. Brown AS, Schaefer CA, Wyatt RJ, et al. Maternal exposure to respiratory infections and adult schizophrenia spectrum disorders: a prospective birth cohort study. Schizophr Bull (2000) 26:287–95.[Abstract/Free Full Text]
  91. Brown AS, Begg MD, Gravenstein S, et al. Serologic evidence of prenatal influenza in the etiology of schizophrenia. Arch Gen Psychiatry (2004) 61:774–80.[Abstract/Free Full Text]
  92. Hurley A. Environmental inequalities: class, race, and industrial pollution in Gary, Indiana, 1945–1980. (1995) Raleigh, NC: University of North Carolina Press.
  93. Sze J. Noxious New York: the racial politics of urban health and environmental justice. (2006) Cambridge, MA: MIT Press.
  94. Pellow DN. Garbage wars: the struggle for environmental justice in Chicago. (2004) Cambridge, MA: MIT Press.
  95. Kuh D, Ben-Shlomo Y. A life course approach to chronic disease epidemiology: tracing the origins of ill-health from early to adult life. (2004) New York, NY: Oxford University Press.
  96. Ben-Shlomo Y, Kuh D. A life course approach to chronic disease epidemiology: conceptual models, empirical challenges and interdisciplinary perspectives. Int J Epidemiol (2002) 31:285–93.[Free Full Text]
  97. Wheaton B, Clarke P. Space meets time: integrating temporal and contextual influences on mental health in early adulthood. Am Sociol Rev (2003) 68:680–706.[CrossRef]
  98. Weiser M, van Os J, Reichenberg A, et al. Social and cognitive functioning, urbanicity, and schizophrenia. Br J Psychiatry (2007) 191:320–4.[Abstract/Free Full Text]
  99. Rutter M, Sroufe A. Developmental psychopathology: concepts and challenges. Dev Psychopathol (2000) 12:265–96.[CrossRef][Web of Science][Medline]
  100. Kirkbride JB, Boydell J, Ploubidis GB, et al. Testing the association between the incidence of schizophrenia and social capital in an urban area. Psychol Med (2008) 38:1083–94.[Medline]
  101. Putnam R. Making democracy work: civic traditions in modern Italy. (1993) Princeton, NJ: Princeton University Press.
  102. Bourdieu P. Outline of a theory of practice. (1992) Cambridge, United Kingdom: Cambridge University Press.
  103. Rutter M. Ethnicity and causal mechanisms. Tienda M, ed. (2005) Cambridge, United Kingdom: Cambridge University Press.
  104. 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]
  105. Malaspina D, Corcoran C, Fahim C, et al. Paternal age and sporadic schizophrenia: evidence for de novo mutations. Am J Med Genet (2002) 114:299–303.[CrossRef][Web of Science][Medline]
  106. Brown AS, Schaefer CA, Wyatt RJ, et al. Paternal age and risk of schizophrenia in adult offspring. Am J Psychiatry (2002) 159:1528–33.[Abstract/Free Full Text]
  107. Dalman C, Allebeck P. Paternal age and schizophrenia: further support for an association. Am J Psychiatry (2002) 159:1591–2.[Abstract/Free Full Text]
  108. Byrne M, Agerbo E, Ewald H, et al. Parental age and risk of schizophrenia: a case-control study. Arch Gen Psychiatry (2003) 60:673–8.[Abstract/Free Full Text]
  109. Zammit S, Allebeck P, Dalman C, et al. Paternal age and risk for schizophrenia. Br J Psychiatry (2003) 183:405–8.[Abstract/Free Full Text]
  110. El-Saadi O, Pedersen CB, McNeil TF, et al. Paternal and maternal age as risk factors for psychosis: findings from Denmark, Sweden and Australia. Schizophr Res (2004) 67:227–36.[CrossRef][Web of Science][Medline]
  111. Sipos A, Rasmussen F, Harrison G, et al. Paternal age and schizophrenia: a population based cohort study. BMJ (2004) 329:1070–4.[Abstract/Free Full Text]
  112. McClellan J, Susser E, King M. Maternal famine, de novo mutations, and schizophrenia. JAMA (2006) 296:582–4.[Free Full Text]
  113. Mill J, Tang T, Kaminsky Z, et al. Epigenomic profiling reveals DNA-methylation changes associated with major psychosis. Am J Hum Genet (2008) 82:696–711.[CrossRef][Web of Science][Medline]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
30/1/84    most recent
mxn006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by March, D.
Right arrow Articles by Susser, E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by March, D.
Right arrow Articles by Susser, E.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?