Losing a family member in childhood associated with psychotic illness

The researchers say that the risks are highest for children who have experienced a suicide in the 'nuclear family' (brothers, sisters, parents).

Previous studies have concluded that the risk of adult disease can be influenced by genetics, lifestyle and environmental experience. There is also evidence that maternal psychological stress adversely affects the development of the fetus.

Population studies have so far provided weak support for an association between prenatal maternal psychological stress and later psychosis. Researchers from the UK, US and Sweden therefore set out to examine the association between deaths in the family as a form of severe stress to the individual and subsequent psychosis. Data were taken from Statistics Sweden and the Swedish National Board of Health and Welfare and children born between 1973 and 1985 in Sweden.

Definitions of psychosis were: non-affective psychosis (including schizophrenia) and affective psychosis (bipolar disorder with psychosis and unipolar depression with psychosis).

Exposure periods were divided into 'any exposure' (all pre and postnatal); 'any prenatal' (prior to birth) and 'any postnatal' (birth up to 13 years of age) and further subdivided by trimester (first, second, third) and by three year periods in childhood between birth and 13 years of age (0-2.9 years; 3-6.9 years and 7-12.9 years). If more than one exposure occurred during the study period, priority was given to the earliest exposure.

Death was categorised into suicide, fatal injury / accident and others (such as cancers and cardiac arrests).

Models were adjusted for year of birth, child sex, maternal and paternal age, maternal and paternal nationality, parental socioeconomic status and history of any psychiatric illness in the family.

The final number of children included in the study was 946,994. Altogether, 321,249 (33%) children were exposed to a family death before the age of 13. Of individuals exposed to any death during the study period, 1323 (0.4%) developed a non-affective psychosis while 556 (0.17%) developed an effective psychosis. 11,117 children were exposed to death from suicide, 15,189 from accidents and the majority, 280,172 to deaths due to natural causes.

No increased risk of psychosis was seen following exposure in any prenatal period. Postnatally, an increased risk of 'all psychosis' was associated with deaths in the nuclear family and risk increased the earlier in childhood the death occurred.

Risks associated with exposure to suicide were higher compared with exposure to deaths from accidents which in turn were higher than risks associated with other deaths from natural causes.

The largest risk was seen in children exposed ages 0-3 years and risks reduced as age of exposure increased.