Sociocultural Factors in the Development of Schizophrenia

In a recent review of the relevant research, Jane Boydell, Jim van Os and Robin Murray (2004), in their “Is there a role for social factors in a comprehensive developmental model for schizophrenia?” (contained in an excellent new volume “Neurodevelopment and Schizophrenia” edited by Matcheri Keshavan, James Kennedy and Robin Murray in 2004 for Cambridge University Press), noted:

“In the 1950s and 1960s, there was much extravagant discussion of the role of social factors in the etiology of schizophrenia. However, there was little scientific basis to this speculation, and it was swept away by the demonstration that people with schizophrenia showed abnormalities of brain structure on computed tomographic scans (Johnstone et al 1976) [it soon became apparent to those of us who were studying both areas of neuroscience research, schizophrenia and the effects of profound fear/stress/anxiety/trauma/social isolation on CNS structure/function, that there was a substantial overlap between the two areas of scientific inquiry]. A decade later, the neurodevelopmental model of schizophrenia was proposed, and it subsequently became the dominant etiological and pathogenetic model (Murray and Fearon, 1999; Murray et al., 1992). As a result of these two developments, researchers have come to regard schizophrenia as a brain disease, and social factors have been largely ignored as putative etiological agents.

It is increasingly clear, however, that the neurodevelopmental model, an essentially neurological concept, does not explain all the available data about schizophrenia. One consequence has been a revival during the 1990s, particularly in Europe [unfortunately, in the USA, economic and political concerns, the latter increasingly seeming to be driven by the former, as well as reductionistic models lacking integration of different scientific domains, wittingly or unwittingly, have indirectly served as roadblocks to the development of more comprehensive and scientific models of the schizophrenic disorders], of research into the role of social factors as causal agents in schizophrenia” (p.224).

Boydell et al (2004) pointed to the neurobiological effects of isolation rearing and social stress in animals. For example, rats raised in isolation demonstrated structural and physiological differences from controls in the hippocampi. Isolation-raised rats demonstrate anxiety, learning deficits (analogue of working memory, hypofrontality, etc.?), sensory changes, dopaminergic dysfunction, etc. (see the excellent research of Myron Hofer on the psychobiology of developmental loss and separation). In terms of human development, social relationship experiences may alter prefrontal neural systems which mediate emotional self-regulation (Lyons et al 2002). The early social environment impacts on various levels of psychobiological and neurobiological development. The early social environment has been demonstrated to induce synaptic changes that may be indicative of, and perhaps the cause of, alterations in behavioral and cognitive functioning (Ovtscharoff and Braun 2001). There is evidence that the early social environment can mediate the establishment of neural networks that regulate a child’s response to stress and emotional self-control (DiPietro 2000).

Boydell et al (2004) have identified the following broad categories in which social factors have been implicated in the initiation and course of the schizophrenias: family factors (mother-child relationship, unwantedness, family communication deviance, dysfunctional family environment, communal upbringing, early parental loss, expressed emotion, childhood abuse, etc.); an urban effect (city birth, city upbringing, etc.); social isolation (during childhood, moving schools in adolescence, in young adult life, at time of onset, migration and ethnic minority status, discrimination, unemployment, etc.); life events (socioeconomic factors, deprivation, inequality, etc.); interaction between social and other etiological factors (gene-environment interaction, social factors and cognitive processing, social causation versus social selection, etc.). I shall summarize the main research findings in each category.

Family Factors

Mother-child relationship

The British 1946 cohort study followed all children born in 1 week, a total of 5,362 subjects. By age 43, there were 30 cases of schizophrenia. One of the most powerful risk factors for later schizophrenia, was the quality of the mother-child relationship at age 4, as rated by health visitors. A poor relationship (e.g., rated as less understanding--perhaps in line with the research constructs of Fonagy and colleagues, e.g., ‘mentalization,’ and its relevance to borderline phenomena) carried a sixfold increase in risk for later schizophrenia. Similarly, the Dunedin study which followed 1000 children from birth to age 26 with comprehensive medical and psychiatric assessments, demonstrated that the mothers of children who developed schizophreniform disorders were rated as having poorer attitudes and behavior towards their children at age 3. The British 1946 and the Dunedin studies are particularly interesting since they are population-based, as opposed to high-risk studies. However, is the increased risk a function of impaired caregiving or is impaired caregiving a function of a genetic diathesis in the parent or a reaction to a disturbed child? Some research studies have suggested that impaired relatedness is an independent variable which accounts for the variance apart from the latter two variables.


Myrhrman et al (1996) used the Northern Finland 1966 Birth Cohort (11,017 individuals) followed through age 28. When mothers were 6 or 7 months pregnant, they were asked by a midwife if their pregnancies were wanted, mistimed but wanted, or unwanted. The risk of developing schizophrenia was considerably raised for the unwanted children even after adjusting for sociodemographic, pregnancy (including depression), and perinatal variables.

Family Communication deviance

A study by Wahlberg et al (1997, 2000) found a significant interaction between high genetic risk for schizophrenia and communication deviance in the adoptive parents. The high risk but not the control adoptees showed greater evidence of thought disorder if their adoptive parents showed communication deviance. This demonstrates genetic control of sensitivity to the environment or environmental control of gene expression.

Dysfunctional family environment

In the Finnish Adoptive Family Study by Tienari and colleagues the risk of developing a schizophrenia spectrum disorder was higher in those adopted-away offspring of schizophrenic parents who were exposed to a dysfunctional adoptive family-rearing environment. The most recent data from this study suggests that the adopted-away offspring may have lower risk than children who remain with their parent with schizophrenia. Similarly, the Danish-American adoption studies suggested that the high-risk adopted-away offspring had a lower chance of developing schizophrenia than those who remained with their biological parents (Rosenthal et al 1971), suggesting the importance of the caregiving environment as neuroprotective.

Communal upbringing

In the Israeli High Risk Study (Mirsky et al 1985), 46 children with high genetic risk, half of whom were raised communally on a kibbutz (where their parents lived) and half in family homes, were compared with controls. High risk children were significantly more likely to develop a schizophrenic or affective disorder if they were raised communally as opposed to being raised in a family home. Similar results were observed in the Copenhagen High Risk Project (public care institutions as opposed to a kibbutz).

Early parental loss

Agid et al (1999) observed a significant correlation between parental loss before the age of 8, particularly loss of the mother through death, and development of schizophrenia. Similarly, Mallett et al (2002) found that separation from both parents in childhood discriminated UK African-Caribbean individuals with schizophrenia from Afro-Caribbean controls.

Expressed emotion

Social overstimulation and high-expressed emotion are associated with relapse in schizophrenia, a finding that has been replicated many times over. This does not have to refer only to family settings, but can be extended to other social situations, e.g., inpatient wards.

Childhood abuse

Many researchers have documented a correlation between childhood traumas (sexual, physical and emotional abuse) and later psychosis (Read et al 2001). In a recent study, childhood abuse was a significant predictor of hallucinations, even in the absence of adult abuse (Read et al 2003). Patients with a diagnosis of bipolar disorder and PTSD who also report a history of childhood abuse have a high frequency of delusions and hallucinations. It has been suggested that, of all diagnostic categories, psychosis displays the strongest associations with childhood abuse.

Boydell et al (2004) noted: “The experience of abuse may create a biological (Read et al 2001) or psychological (Garety et al 2001) vulnerability for the development of psychotic symptoms, including subclinical psychotic experiences such as low-grade delusional ideation and isolated auditory hallucinations (Johns and van Os 2001)” (p.229).

In both clinical and non-clinical populations (Ross and Joshi 1992), the diagnostic group with the highest rate of childhood abuse consistently reported the most Schneiderian symptoms (first rank symptoms of schizophrenia).

One difficulty with the above studies is that they are cross-sectional in nature and have a potential bias in terms of self-report. In a recent three year longitudinal study of a general population sample of 4,045 subjects from age 18-64 with no previous psychotic symptoms, baseline reported childhood abuse predicted development of clinically relevant positive psychotic symptoms associated with need for care (Janssen et al 2004). This association remained even after adjusting for a range of risk factors and demographic variables.


The urban effect

City birth

Arguing against the prevalence of the ‘downward-drift’ theories, many studies have shown strong effects for urban birth and/or urban upbringing. One of the most impressive was conducted by Mortensen et al (1999) with a Danish population-based cohort of 1.75 million subjects. These resulted in strong effects for urban birth. There was a clear dose-response relationship for urbanicity in that the larger the town of birth, the greater the risk. A family history of schizophrenia did not explain or effect the results. The investigators calculated a 34.6% population attributable fraction for urban birth compared with 9% and 7% respectively, for having a mother or father with schizophrenia.

City upbringing

Pedersen and Mortensen (2001) used a comprehensive Danish national registration system that accurately recorded changes in residence to demonstrate that risk of developing schizophrenia increased with the number of years that an individual lived in an urban area and with an increasing degree of urbanization.

Social drift and social residue theories

The social drift hypothesis cannot explain the results of the above research studies. In addition, Dauncey et al (1993) investigated where patients lived 5 years prior to their first hospital admission in Nottingham , UK and were unable to find support for systematic geographic drift. In the Danish study by Mortensen et al (1999), the investigators ruled out downward drift as a viable explanation for urban excess. A family history of schizophrenia did not explain or effect the urban-rural difference in the Mortensen or other studies (van Os et al 2002). This is also relevant for the social residue theory (that those at greater risk are ‘left behind’ in an area as it becomes less desirable because they do not have the resources to move out).

Possible explanations

Possible etiologic factors hypothesized to explain urban excess have been: infectious agents, lead pollution, and social factors (e.g., social isolation, deprivation, and adverse life events). Boydell et al (2004) noted:

“...while there is no direct evidence that the urban excess is caused by social factors, the fact that its effect is so widespread across the population is compatible with the notion that part of the excess risk represents a psychological reaction to factors in the wider social environment. certainly some of the major differences between urban and rural areas are to do with social cohesion and support” (p.231).

Social factors can, with some vulnerable individuals, act in concert with genetic risk.

Social Isolation

During childhood

Many clinicians from Bleuler onwards have noted that individuals who develop schizophrenia have had relatively solitary backgrounds including fewer social networks (parenthetically, severity of symptomatology is correlated with size of social network and I am aware of one neuroimaging study which correlated atrophic neural regions with size of social network) and friends. The British 1946 cohort study (Jones et al 1994) observed that preference for solitary play at age 4 or 6 was associated with later schizophrenia. In addition, self-reported anxiety at age 13 and teacher-rated anxiety at age 15 both showed linear associations with later risk for developing schizophrenia.

Moving schools in adolescence

A research finding emerging from the Danish study of Pedersen and Mortensen (2001) was that change in municipality and therefore school, increased the risk for later schizophrenia but not change of address within the municipality. Geographic moved during the early teen years had the greatest risk and the more moves, the greater the risk factor (this was a finding I observed in my patients at a long-term state psychiatric hospital-I theorized that these changes and uprootings created discontinuities in the self leading to a lack of cohesion of the self-I believe this was a factor in the life of Kay Redfield Jamison,, albeit unacknowledged by her as playing a role in her illness).

In young adult life

The Swedish conscript study also looked at the role of premorbid personality for later development in 50087 individuals. The investigators found that there was a significantly increased risk of later developing schizophrenia in young men who felt they were more sensitive than their peers, had fewer than two close friends,, preferred small groups, and did not have a girl friend. Once again, Boydell et al (2004) raised the possibility of a gene-environment interaction: “Until proven otherwise, it is wise to consider that both may be true: individuals with a schizoid or schizotypal personality may be less able to make social relationships, and then the social isolation may propel them further toward frank psychosis” (pp.232-233).

At time of onset

Hare (1956) observed that social isolation, as measured by proportion of single person households in a geographic area, was associated with higher rates of schizophrenia. This finding was not accounted for by downward drift. One theory is that disruption of social networks decreases a person’s capacity to cope with psychosocial stress and may increase the risk for developing schizophrenia.

Van Os et al (2000) demonstrated that single people had a slightly higher risk of developing psychosis if they lived in a neighborhood with fewer single persons compared with a neighborhood with many other single people (theoretically, this would increase the person’s sense of social exclusion, loneliness and isolation).Jablensky and Cole (1997) demonstrated that marriage had a protective effect and that this was not simply a function of better adjusted males being able to marry.

Migration and ethnic minority status

As early as 1932, Odegard (1932) observed that Norwegian persons migrating to the USA had an increased risk of developing psychosis. It was thought that the cultural and geographic disruption resulted in paranoia and alienation. This is a robust finding (i.e., migration as a risk factor for psychosis) and has been demonstrated amongst many migrant groups (Castle et al 1991;Harrison et al 1988, 1997; King et al 1994; Selten et al 1998, 2001; van Os et al 1996;etc). Research methodological issues such as ‘category fallacy’ (cultural variation in diagnosis), misdiagnosis, inaccurate estimation of the denominator, etc., have been overcome in studies which still demonstrate an excess of psychosis in migrant groups even when moving to similar cultures (Bruxner et al 1997). Even when controlling for urbanicity (i.e., excess urban effect-since many migrants settle in urban areas), studies demonstrate an excess migration effect (van Os et al 2001).

African-Caribbeans in the UK

Much research has investigated the increased rates of psychosis in the African-Caribbean population in the UK and the Netherlands . Genetic diathesis cannot be the sole explanation since the increased risk is not shared by the population of origin in the Caribbean (Bhugra et al 1997; Mahy et al 1999). In addition, the risk for second-generation siblings is much greater than for their first-generation equivalents (Hutchinson et al 1996; Sugarman and Crawford 1994). Selective migration has been ruled out in a Surinamese population which migrated to the Netherlands (Selten et al 2002). Other variables which have been ruled out include neurodevelopmental insults secondary to obstetric complications and viral infections, as well as the role of substance misuse (McGuire et al 1995; Selten et al 1997).

The second generation

There are many studies demonstrating higher rates of schizophrenia in the children of migrants. This has been observed in Greenland (Mortensen et al 1999), in the USA (Malzberg 1969), and in the UK with the Afro-Caribbean population (Harrison et al 1988). In the Yemenite Jewish population which migrated to Israel , Weingarten and Orren (1983) observed a high prevalence of schizophrenia among the offspring (the adults did not integrate into Israeli society and their lifestyle was considered primitive). Since the entire population migrated to Israel , selective migration cannot account for the findings.

Boydell et al (2004) concluded:

“There is no satisfactory explanation as to why there are higher rates of psychosis in children of migrants, but the range of countries and circumstances in which this phenomena has been described is suggestive of a socially induced phenomena” (p.235).


Psychosis has also been attributed to such factors as racism (overt and institutionalized), social isolation, and reduced social networks. Boydell et al (2001) demonstrated that incidence rates of schizophrenia increased in ethnic minorities as the proportion of ethnic minorities in the locality fell, suggesting that social experience (isolation, discrimination, etc.) contributed to development of the illness. Janssen et al (2003) measured subjective experiences of discrimination and subsequent development of psychotic illness 3 years afterwards. Experience of discrimination strongly predicted for the development of delusional ideation. Janssen et al (2002) also demonstrated that the effect of ethnic minority status on psychosis was no longer significant when controlling for the experience of discrimination. Karlsen and Nazroo (2002) studied the experience of ethnic harassment and discrimination among a UK representative sample of 5000 persons from ethnic minorities. They demonstrated significantly increased OR (odds ratio) for a range of health problems and particularly psychosis. Mallett et al (1998) demonstrated that one of the main distinguishing characteristics of first-onset patients of Caribbean origin with psychosis in London was that they lived alone and also had been separated from their mothers at an early age.


Bhugra et al (1997) observed high unemployment in persons of Caribbean origin first presenting with schizophrenia in the UK . An analogous effect was not observed in Tinidad (Bhugra et al 2000). Perhaps the social milieu in Trinidad is more tolerant of those at risk of developing schizophrenia and that there is greater stigmatization in London .

Life events

Brown and Birley (1968) observed an excess of life events three weeks prior to an episode of schizophrenia. Prospective studies have also demonstrated an association between life events and relapse into psychosis (Malla et al 1990; Ventura et al 1989). Bebbington (2000), using a case-control study in London, found a significant excess of life events in the three months prior to onset of schizophrenia {I believe similar research findings have been documented in affective illness-perhaps the model of Post et al on ‘kindling’ is relevant here-we see more life events earlier in the onset and progression of the disease than in later episodes-theoretically, because later episodes do not require as much stimulation to cross threshold into active psychosis). There is evidence that it is not so much major life events that precipitate psychotic relapse in vulnerable individuals, but what psychoanalysts call narcissistic injuries of one kind or another encountered in daily life.

Part of this sensitivity may be rooted in exposure to earlier major life events (Myrin-Germays et al 2003), suggesting “synergistic environment-environment interactions” (Boydell et al 2004, p. 237). Boydell et al (2004) speculated that:

“Stress may be generated in part by underlying personality traits, whose genetic contribution may overlap with that of schizophrenia. Alternatively, individuals with vulnerability to schizophrenia may be more sensitive to the effects of stress, while sensitivity to stress is also determined by the degree of prior exposure to stress, possibly including stress in early life (Janssen et al 2004)” (p. 237).

Socioeconomic factors, deprivation, and inequality

Several research studies have found a relation between economic deprivation and incidence rates of psychosis, prevalence and admission rates of schizophrenia (Croudace 2000). Recent studies have implicated inequality as a risk factor. Boydell et al (2003) found a positive correlation between incidence rates of schizophrenia and degree of inequality in deprived areas in London , after adjusting for such variables as age, sex, absolute deprivation and ethnicity.

Interaction between social and other etiological factors

Gene-environment interactions

Boydell et al (2004) noted:

“There is an apparent paradox in that schizophrenia appears highly heritable [as I have noted in previous writings epigenetics and transcriptomes need to be carefully studied in schizophrenia-epigenetics refers to the regulation of gene expressions that are controlled by heritable but potentially reversible changes in DNA methylation and/or chromatin structure] and yet...many environmental and social factors, appear to play a role. The predictive power of each of these environmental low (i.e., most people exposed to each of the risk factors remain well and never develop the illness). A unifying explanation seems to be that the environmental factors operate upon genetic risk. There are many forms that this interaction could take: synergistic, additive, multiplicative (van Os and Sham 2003)” (pp.237-238).

The Finnish and Danish adoption studies, as well as the Israeli High Risk Study, demonstrated that environmental factors operate on genetic factors to increase risk for developing schizophrenia. As I pointed out in previous publications, ‘environmental’ factors can alter the genome, as well as override it. Bolton and Hill (1996) noted that intentionality (beliefs, goal-directed plans, fears, etc.) pervades biological systems to the molecular level. This is borne out in neuroscience research in which psychogenic stress was genotoxic in various body cells. Fishman and colleagues (1996) reporting in the International Journal of Neuroscience demonstrated in rats that behavioral, psychogenic stress can result in DNA damage and chromosome aberrations. They noted:”behavioral stress can induce genotoxic damage on at least two levels, chromosomal and molecular, and in at least two cell types, bone marrow and leukocytes” (p. 224).

Social factors and cognitive processing

Boydell et al (2004) suggested:

“An important implication of any social theory is that the effects of social factors may operate by impacting not only on brain development but also on psychological processes that may contribute to the symptoms of schizophrenia...People with persecutory delusions selectively attend to threatening information, tend to jump to conclusions, attribute negative events to external causes, and have difficulty in understanding others’ intentions, motivations, and states of mind (Blackwood et al 2001). It is plausible that people with this cognitive style would be even more likely to develop delusions when subjected to social adversity, than if they lived in a more benign social milieu. Attributional style, in particular, has been identified as a pathway through which discrimination and racial harassment could lead to psychosis (Sharpley et al 2001)” (p. 238).

I highly recommend a new volume on the cognitive/psychological factors operative in the initiation and mediation of psychotic symptoms: “Cognitive Therapy for Psychosis: A Formulation-Based Approach” by Anthony Morrison, Julia Renton, Hazel Dunn, Steve Williams & Richard Bentall, published in 2004 by Brunner-Routledge.

Social causation versus social selection

There has been an ongoing debate between those who view social factors as influencing the development of schizophrenia (social causation) and those who understand individuals at risk to choose adverse social environments (social selection). A number of the research findings just described cannot be accounted for by social selection theorists. Boydell et al (2001) found that non-white minorities to be at higher risk when they were in a smaller minority even though they lived in an area of higher social



Boydell et al (2004) concluded:

“It is now clear, however, that, in order to understand the causes of schizophrenia, the role of the social environment cannot be continued to be ignored. In saying this, we are not proposing an oppositional social instead of biological approach, which we consider as futile as arguing whether poverty or mycobacteria cause tuberculosis! Rather, we suggest that both social and biological factors need to be studied as well as their interaction.

We need to recognize that (i) social factors can impact on brain development, (ii) some social factors give rise to psychological vulnerabilities, and (iii) many social factors act over the life course, creating developmental liabilities...It is possible that the social environment creates psychological vulnerabilities that act additively to the risk function in combination with genetic or non-genetic neurodevelopmental impairments...

The challenge for schizophrenia researchers in the coming decade is first distinguish those candidate social factors that do contribute to schizophrenia risk from those that do not and, second, to identify the interplay between these factors, genetic susceptibility, and their respective effects on, and interactions with, brain development” (pp. 239-240).

The DSM, which Dr. Karl Menninger once called the contemporary equivalent of the Malleus Maleficarum (the 16th century witch-hunters' manual) is supposedly atheoretical. Much of the psychiatry practiced in the US, and much of the popular ‘scientific’ media, centers around the concept of schizophrenia as a brain disease, in fact, this has been increasingly extended to all sorts of human problems including anxiety, depression, substance misuse, eating disorders, obsessive ruminations, etc. As a neurobiologically-informed clinician who abhors the artificial split between mind and brain events, I have no doubt that there are definite neurobiological, as well as neurogenetic-especially through epigenetic regulation of gene expression- correlates of human experience. What I strongly object to is the systematic marginalization of the social and psychological experiences of those persons who have been diagnosed with a severe mental illness and the uncoupling of brain events from the social surround-in effect, the brain is an organ which is significantly altered by that which it allows us to experience, in the external or internal ‘worlds.’ Pardon the geeky metaphor, but it is a case in which the ‘hardware’ is significantly altered by the ‘software’ which it runs, more specifically, the hardware is inherently designed to do so.

Neuroplasticity may be the non-specific neural diathesis underlying many schizophrenic conditions. May the current incarnations of the medieval Malleus Maleficarum, the neo-Kraepelinian mythology and Jasperian view of incomprehensibility be replaced by a deeper understanding of the intentionality involved in symptoms, an understanding which can reach to the meaningfulness of apparently bizarre symptomatology and not solely hammer away at it with polyneuropsychopharmacology (I have no qualms about the rational and non-reductionistic use of psychopharmacological agents-ideally, when agents are used, it should be in support of psychotherapeutic approaches) but instead dares to allow an ongoing mutual intersubjectivity which seeks to understand and bear the messages, appeals and revelations encoded in the annihilation anxieties and psychotic symptomatology of the human-being.

Brian Koehler PhD
New York University
80 East 11th Street #339
New York NY 10003

Contact Us | Website Privacy Policy | Copyright ©2005-10 ISPS-US. All rights reserved. | Webmaster Susan Dansker Design