BackgroundPsychotic experiences (PEs) refer to traits in the general population, such as paranoia, that at the extreme are characteristic of disorders such as schizophrenia. PEs are common in teenagers and are modestly... [ view full abstract ]
Background
Psychotic experiences (PEs) refer to traits in the general population, such as paranoia, that at the extreme are characteristic of disorders such as schizophrenia. PEs are common in teenagers and are modestly heritable. Mechanistic studies demonstrate that PEs can be induced by certain activities and environmental contexts, such as smoking cannabis or by reducing someone’s relative height in social situations. Our aim was to compare the degree of genetic and environmental influences on PEs in 16-year-old twins who were more and less exposed to multiple environmental risk factors.
Methods
Participants in the Twins Early Development Study (TEDS), a community sample, reported on PEs (paranoia, hallucinations, cognitive disorganization, grandiosity, anhedonia) at age 16 and the twins’ parents reported on the twins’ negative symptoms (N = 4542 pairs). An environmental composite was created, based on risk factors including cannabis use, tobacco use, dependent stressful life events and bullying victimization. Twin pairs were categorized as ‘more exposed’ if either or both twins had experienced at least three exposures and as ‘less exposed’ if both twins experienced fewer than three ‘environmental’ exposures. Univariate heterogeneity twin models and Purcell moderation models were run to test for differences in the heritability of PEs in the exposure groups. As an internal control, analyses were re-run on height and autistic traits, for which no differences between exposure groups were predicted. Finally, a univariate liability threshold model was run on the environmental composite.
Results
Heterogeneity twin models showed that heritability could not be equated in the more and less exposed groups for four of the six PE subscales. These four PEs – paranoia, hallucinations, cognitive disorganization and anhedonia --were on average 50% less heritable in the more exposed than the less exposed group (mean heritability = 21% versus 42%, respectively). Lower heritability in the more exposed group was not explained by higher measurement error (assessed using Cronbach’s alphas). Purcell moderation models concurred in showing a pattern of reduced heritability in the exposed group. Significant differences in heritability contingent on exposure were not present for the control variables (autistic traits and height), suggesting the effect was not general to other variables. Negative symptoms was the only subscale that showed the opposite pattern, with significantly higher heritability in the more exposed group. The environmental composite showed significant heritability (46%) and significant shared and nonshared environmental influences (both 27%).
Discussion
These results suggest that psychotic experiences are less heritable when reported by adolescents in the community who have experienced multiple environmental risks. PEs may sometimes result from social environments and these PEs may differ in etiology from PEs that form part of a more heritable neurodevelopmental/neuropsychiatric pathway. If further work supports it, this hypothesized distinction could be useful for both gene-discovery work and for early intervention approaches. It is unlikely that “environmentally-induced” PEs are completely independent of genetic effects, since exposure to environmental risk factors is in itself partly heritable through gene-environment correlation. Future steps include testing for differential genome-wide polygenic score prediction of PEs contingent on environmental exposure.
