Abstract

Autistic individuals experience significantly elevated rates of childhood trauma, self-harm and suicidal behaviour and ideation (SSBI). Is this purely the result of negative environmental experiences, or does this interact with genetic predisposition? In this study we investigated if a genetic predisposition for autism is associated with childhood trauma using polygenic scores (PGS) and genetic correlations in the UK Biobank (105,222 < N < 105,638), and tested potential mediators and moderators of the association between autism, childhood trauma and SSBI. Autism PGS were significantly associated with childhood trauma (max R2 = 0.096%, P < 2 × 10−16), self-harm ideation (max R2 = 0.108%, P < 2 × 10−16), and self-harm (max R2 = 0.13%, P < 2 × 10−16). Supporting this, we identified significant genetic correlations between autism and childhood trauma (rg = 0.36 ± 0.05, P = 8.13 × 10−11), self-harm ideation (rg = 0.49 ± 0.05, P = 4.17 × 10−21) and self-harm (rg = 0.48 ± 0.05, P = 4.58 × 10−21), and an over-transmission of PGS for the two SSBI phenotypes from parents to autistic probands. Male sex negatively moderated the effect of autism PGS on childhood trauma (β = −0.023 ± 0.005, P = 6.74 × 10−5). Further, childhood trauma positively moderated the effect of autism PGS on self-harm score (β = 8.37 × 10−3 ± 2.76 × 10−3, P = 2.42 × 10−3) and self-harm ideation (β = 7.47 × 10−3 ± 2.76 × 10−3, P = 6.71 × 10−3). Finally, depressive symptoms, quality and frequency of social interactions, and educational attainment were significant mediators of the effect of autism PGS on SSBI, with the proportion of effect mediated ranging from 0.23 (95% CI: 0.09–0.32) for depression to 0.008 (95% CI: 0.004–0.01) for educational attainment. Our findings identify that a genetic predisposition for autism is associated with adverse life-time outcomes, which represent complex gene-environment interactions, and prioritizes potential mediators and moderators of this shared biology. It is important to identify sources of trauma for autistic individuals in order to reduce their occurrence and impact.

Highlights

  • Autistic individuals have elevated rates of self-harm and suicidal behaviour and ideation (SSBI) [1,2,3,4,5,6]

  • While there are a few methods that improve the variance explained of the polygenic scores (PGS) compared to clumping and thresholding [76,77,78,79], we decided not to use these as: (1) The increase in variance explained is minimal compared to clumping and thresholding, with one study showing no statistically significant difference in variance explained [80]; (2) The current study investigates covariance rather than variance, and it is unclear if other methods improve the covariance explained; (3) The large sample size of the testing dataset (UK Biobank) used in the current study makes using methods such as LDPred [76] computationally inefficient and impractical; and 4

  • We investigated if autism PGS are associated with childhood trauma

Read more

Summary

1234567890();,: 1234567890();,: Introduction

Autistic individuals have elevated rates of self-harm (with or without suicidal intent) and suicidal behaviour and ideation (SSBI) [1,2,3,4,5,6]. Polygenic scores (PGS) derived from genome-wide association studies (GWAS) of autism, which represent the underlying genetic predisposition for autism, currently explain a small proportion of the variance in autism (2.5% of the total variance compared to a SNP heritability (h2SNP) of 11%) They are a useful index of the genetic predisposition for autism and have been associated with a number of traits in the general population: social and communication difficulties in childhood [46], autistic traits [47], and cognitive aptitude [48]. The h2SNP for suicidal attempt for individuals with an autism diagnosis was 9.6% (95% CI: 1.1–18.1%) [58], suggesting a small but significant common genetic component for suicide attempt in autistic individuals The heritability of both childhood trauma and SSBI represent complex gene-environment effects. We identified 150 autistic individuals using the UK Biobank data field 20544 (‘Mental Health problems ever diagnosed by a professional’)

Participants
Results
Discussion
Compliance with ethical standards
Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call