Moving houses and increased risk of non-affective psychotic disorders in children and adolescents

by Ashild Kummen

Whereas last week we talked about the association between psychotic disorders and moving to a different country (which you can read about here), another one of our recent papers, published in last month’s issue of JAMA Psychiatry, found that, before age 20, the greater the distance moved within a country, the greater the risk to be diagnosed with a non-affective psychotic disorder (such as schizophrenia).

It is not uncommon to move house during childhood and adolescence, for example due to financial problems or a change of parental employment in the household. These time periods are important for social development and may be a central stressor when migrating is the breaking up of social networks. Moving during childhood may be associated with changing schools, meaning young people could be faced with the challenges of fitting in with a new peer group. Social isolation and lack of peer relationships have been previously found to be associated with psychotic disorders (Winsper, Wolke, Bryson, Thompson, & Singh, 2016), and consistent with this, research has also found a link between psychotic disorders and moving houses during childhood and adolescence (Paksarian, Eaton, Mortensen, & Pedersen, 2015).

Until now, it has remained unclear whether this impact extended beyond mid-adolescence, and into moves in late adolescence and adulthood. We hypothesized that there would be a positive relationship between amount of times moved during childhood and adolescence and later risk of non-affective psychotic disorders. Additionally, we predicted that the greater the distance moved, the greater this risk would be, being strongest when distances moved were most likely to indicate a breakup of social networks (we set this to be 30 km, a distance we hypothesised would likely involve changing schools) .

Our study included 1.4 million participants, all Swedish residents born between 1982 and 1995. They were followed from birth to age 29. We were interested in four time periods: early childhood (0-6 years), childhood (7-15 years), adolescence (16-19 years) and early adulthood (20-29 years). Data on non-affective psychotic diagnoses were collected, as well as additional information such as sex, family income, educational attainment, and parental history of mental illness from the Swedish registers.

As predicted, residential mobility was associated with non-affective psychosis. For all time periods, a change in residence significantly increased the risk, often by number of times moved. For example; moving once over the time period of 16-19 years old increased the risk with 28% but moving four times or more doubled the risk. The pattern was different in early adulthood, where increased risk was notable only when the amount of times moved was 3 or more. Also consistent with our predictions, greater distance moved led to greater risk. However, this was only true during childhood and adolescence, as early adulthood yielded a decrease in risk compared to the general Swedish population (see graph).

During childhood and adolescence, a change in residence is likely initiated by guardians. Our results could be explained in that the breakup of social networks when moving disrupts the developing ability to make new friends and maintain friendships. This could lead to increased social isolation, which again makes one vulnerable to mental illness including psychotic disorders. However, past 20 years old, moving is likely a personal choice by the individual. We found that longer distances moved predicted a lower likelihood of being diagnosed with a non-affective psychotic disorder past 20 years old. This could be explained by the fact that the personal choice made reflects health and independence as the individual moves out from their parent’s house and enters work or attends university. Meanwhile, it would be more difficult for individuals vulnerable to non-affective psychotic disorders to move out and enter work or university, due to symptoms revolved around cognitive deficits as well as the strong association between psychotic disorders and social defeat (Dickson, Laurens, Cullen, & Hodgins, 2012; Selten, van der Ven, Rutten, & Cantor-Graae, 2013).

Our paper published last week found that for all ages investigated, including childhood, adolescence and early adulthood, migrating to a different country increased the risk of being diagnosed with a psychotic disorder. This paper, as well as the one on moving houses within a country, support the idea that migration involving disruption of social networks appear to increase risk, especially for international migrants. This might be particularly difficult for visible minorities. Greater differences, both visible and not, involves greater difficulty for a child to fit in with a peer group.  A good network of social support, including both family and peer-relationships, is crucial as a preventive measure against psychotic disorders. Hence, it is important that children and adolescents receive the necessary attention and support when migrating, both internationally and within a country, to minimize risks of a future diagnosis.


Paksarian, D., Eaton, W. W., Mortensen, P. B., & Pedersen, C. B. (2015). Childhood residential mobility, schizophrenia, and bipolar disorder: a population-based study in Denmark. Schizophr Bull, 41(2), 346-254.

Selten, J. P.,van der Ven, E., Rutten, B. P. F., & Cantor-Graae E. (2013).  The social defeat hypothesis of schizophrenia: an update. Schizophr Bull, 39 (6), 1180-1186.

Winsper, C., Wolke, D., Bryson, A., Thompson, A., & Singh, S. P. (2016). School mobility during childhood predicts psychotic symptoms in late adolescence. J Child Psychol Psychiatry, 57 (8), 957-966.

Dickson, H., Laurens, K. R., Cullen, A. E., & Hodgins, S. (2012). Meta-analyses of cognitive and motor function in youth aged 16 years and younger who subsequently develop schizophrenia. Psychol Med, 42(4), 743-755.

Migrants at elevated risk for psychotic disorders, but not for non-psychotic bipolar disorder – new PsyLife paper

by Ashild Kummen

Image result for migrationMigration can be a difficult process, involving a multitude of stressors. This may make some migrants more vulnerable to mental health problems, and previous research have found migrants to be at an elevated risk for psychotic disorders, especially with visible minorities, such as people from black Caribbean and African backgrounds in the UK . In regard to bipolar disorders, however, the research is more mixed, and enquiries have rarely distinguished between bipolar disorders with and without psychotic symptoms.

A new paper from our Ph.D. student Jennifer Dykxhoorn, published today in Psychological Medicine, sheds light on whether the increased risk of psychotic disorders extends to bipolar disorder with and without psychosis. In her paper, Jen hypothesised that migrants were at elevated risk for schizophrenia and affective psychoses (such as bipolar disorder with psychotic features), but not for bipolar disorders without psychotic symptoms. Longitudinal data from almost 1.8 million Swedish residents born between 1982-1996 were collected, with differences by migrant status, age-at-migration and region of origin investigated. Consistent with previous findings , we predicted that migration during early childhood would increase the risk of being diagnosed with a psychotic disorder.


Compared with the Swedish-born population, migrants and their children were at elevated risk for all psychotic disorders. This included schizophrenia and schizoaffective disorder, where migrants and their children were, on average, twice as likely to be diagnosed. First-generation migrants were also at double the risk for other types of non-affective psychoses. Rates of bipolar disorder with psychosis were also elevated in migrants and their children by 42% and 22%, respectively, compared to the Swedish-born population. In stark contrast,  migrants were up to 40% less likely to be diagnosed with non-psychotic bipolar disorders than the Swedish-born population. For second-generation migrants, risk of non-psychotic bipolar disorder was similar to Swedish-born rates.

While we predicted that migrating during early childhood would elevate risk of psychotic disorders, our findings did not support this. Instead, there was a pattern of increased risk at all ages of immigration to Sweden for psychotic disorders.. By contrast, all ages at immigration were associated with lower risk for bipolar disorder, except for those who migrated during infancy, where incidence rates were closer to the Swedish-born population.

In accordance with previous research by the PsyLife team , African migrants were at highest risk for all psychotic outcomes, including being five times more likely to be diagnosed with schizophrenia. In total, migrants from all origins (except other Nordic origin) were at elevated risk for all psychotic disorders, compared to the Swedish-born population.

Possible explanations

Migrants are exposed to several difficulties relating to the process of migrating and the minority status. Jen’s paper found that the incidence rates for psychotic disorder and bipolar disorders (without psychotic symptoms) is largely different, with higher risk for psychotic disorders and lower risk for bipolar disorders compared to the Swedish-born population. This could suggest that the stressors of migrating and post-migration life have a specific effect on developing psychotic disorders. This is consistent with some research on the neurodevelopmental origins of psychotic and non-psychotic bipolar disorders, conveying these origins are distinct. For example, patients with schizophrenia and bipolar disorders with psychotic features are observed to have premorbid cognitive deficits, which have not been consistently found in non-psychotic affective disorders (Reichenberg et al., 2002; Trotta et al., 2014). Our research suggests further work is now required to test whether certain environmental factors impact on specific neurobiological pathways to influence the occurrence of certain types of mental health problem.


PsyLife references

Other references

  • Reichenberg, A., Weiser, M., Rabinowitz, J., Caspi, A., Schmeidler, J., Mark, M., Kaplan, Z, & Davidson, M. (2002). A population-based cohort study of premorbid intellectual, language and behavioural functioning in patients with schizophrenia, schizoaffective disorder, and nonpsychotic bipolar disorder. Psychiatry: Interpersonal and Biological Processes, 159, 2027-2035.
  • Trotta, A., Murray, R. M & MacCabe, J. H. (2014). Do premorbid and post-onset cognitive functioning differ between schizophrenia and bipolar disorder? A systematic review and meta-analysis. Psychological Medicine, 45, 381-394.

Incidence and risk factors of psychotic disorders in older people

by Ashild Kummen

Whilst it is known that the first episode of a psychotic disorder usually occurs during adolescence or early adulthood (Kessler et al., 2007), there is a considerable amount of people also experiencing the onset in old age, termed very late-onset of schizophrenic-like psychosis (VLOSLP).Over a year, between 0.1- 0.5% of the population past 65 years old are diagnosed with or have an existing diagnosis of schizophrenia (Howard, Rabins, Seeman & Jeste, 2000).

Robust research in regard to the variance in incidences of VLOSLP and its associated risk factors is important to inform public mental health in hopes of improving interventions for affected individuals. Yet, there is a limited amount of epidemiological research on VLOSLP. The studies that do investigate an onset of a psychotic disorder past 60 or 65 years old, are mostly cross-sectional studies of small samples with limited generalisability. Furthermore, their findings lack in replication. In tackling these limitations, Jean Stafford, one of our PhD students in the PsyLife team has conducted a systematic review on the research there is on incidence rates and different risk factors for VLOSLP. She also conducted a longitudinal cohort study looking into risk factors identified in previous research, including gender, age and sensory impairment as well as unexplored areas such as social economic status, migrant status, social isolation and trauma.

The systematic review and meta-analysis found 41 papers (dated between 1960 and March 2016) that looked at incidence cases of patients diagnosed with a psychotic disorder past 65 years old, not including cases related to dementia, organic and drug-induced psychoses. The majority of these papers were rated of high and average quality, and 25 were included for quantitative analysis. The pooled incidence rate showed that, every year, 7.5 out of 100 000 people over 65 years old were diagnosed with schizophrenia. Moreover, psychoses related to an affective disorder had a yearly rate of 30.9 new cases per 100 000 people. Incidence rates for non-affective psychoses differed largely between studies, and a pooled estimate was not possible. However, the cohort study offers some clarification as it investigated only non-affective psychotic disorders.

The cohort study consisted of over 3 million participants past age 60, all Swedish residents born between 1920 and 1949. Large-scale data collection was possible as every resident in Sweden is given a national identification number that is recorded within a range of health and administrative services. This revealed an incidence rate for non-affective psychotic disorders of 37.66 per 100 000 people every year.

Stafford also investigated the influence of gender and age. The systematic review found reports that women were more likely to be diagnosed than men past age 65 within all branches of psychotic disorders analysed (non-affective, affective psychoses and schizophrenia). The cohort study supported this, with participants diagnosed with a non-affective psychotic disorder past 60 years old being 60% female. In regard to age, the systematic review revealed previous research has been mixed. However, the cohort study revealed a significant increase with age, with differing patterns between men and women. Past 80 years old, women were at risk for VLOSLP at an accelerated rate.

The cohort study also collected data on other possible risk factors, such as migrant status, income, family and sensory impairments. Migrants from Africa, North America and Europe were at elevated risk compared to the rest of the population. This is an association also found in younger populations , possibly explained by the stressors encompassed in migration. Furthermore, low income at age 60 significantly predicted a higher future risk of VLOSLP. In an investigation of family history, it was found that participants with children diagnosed with a psychotic disorder were twice as likely to have VLOSLP. Looking at social isolation; participants with no children and/or no partner 2 years before exiting the study were at elevated risk. Experiencing the death of a partner 2 years before also led to increased risk, whilst the death of a child before age 18 had no association and the death of a child at infancy had weak evidence for elevated risk. These could be indicators of the influence of trauma. Contradictory to previous research (Cooper & Porter, 1976; Cooper et al. 1974), participants with sensory impairments were less likely have VLOSLP compared to the general Swedish population. However, this result could reflect an under-detection of psychiatric disorders in people with physical health issues (Roberts, Roalfe, Wilson & Lester, 2007)

In conclusion, these results indicate that people with VLOSLP were more likely to have social disadvantages, such as lower income and social isolation. This could be interpreted two ways; either there is causal link where the stressors of inequalities lead to increased risk of a psychotic disorder, or that those with premorbid symptoms of a psychotic disorder will subsequently suffer disadvantages due to lower functioning, although VLOSLP patients have been shown to have higher premorbid functioning (Castle et al., 1997). Older women are at particularly prominent risk, indicating this is a group deserving of more clinical recognition.

Full version of the systematic review can be found here, and the cohort study here.


PsyLife References:


Other references:

  • Castle, D.J., Wessely, S., Howard, R., et al. (1997). Schizophrenia with onset at the extremes of adult life. Int J Geriatr Psychiatry, 12, 712–717.
  • Cooper, A. F., Curry, A. R., Kay, D. W., Garside, R. F. & Roth, M. (1974). Hearing loss in paranoid and affective psychoses of the elderly. Lancet 2, 7885, 851–54.
  • Cooper, A. F. & Porter, R. (1976). Visual acuity and ocular pathology in the paranoid and affective psychoses of later life. Journal of Psychosomatic Research 20, 107–114.
  • Howard, R., Rabins, P. V., Seeman, M. V., & Jeste, D. V. (2000). Late-onset schizophrenia and very-late onset schizophrenia-like psychosis: An international consensus. American Journal of Psychiatry, 157, 172–178.
  • Kessler, R. C., Amminger, G. P., Aguilar-Gaxiola, S., Alonso, J., Lee S., & Utsun, T. B. (2007). A controlled family study of late-onset non-affective psychosis (late paraphrenia). The British Journal of psychiatry, 170, 511-514.
  • Roberts, L., Roalfe, A., Wilson, S., & Lester, H. (2007). Physical health care of patients with schizophrenia in primary care: a comparative study. Fam Pract, 24, 34–40.