This study is the first to assess the association of different autistic traits and autism with incontinence/constipation in a large prospective cohort study. We found evidence that difficulties with social-communication and coherence are associated with increased odds of daytime wetting and soiling at age 9 years. We also found that the autism factor-mean-score (indicative of a broad autism phenotype) was associated with daytime wetting and soiling and diagnosed autism was associated with daytime wetting. There was weaker evidence of associations between the autistic traits and bedwetting at age 9, and little evidence of associations with constipation at age 9, except for the autism factor-mean-score. At age 14 years, the associations between the autism factor-mean-score and incontinence/constipation persisted and associations were also found between coherence difficulties, autism diagnosis and soiling. We found less evidence of associations between autistic traits/diagnosed autism and incontinence at age 14, but autistic traits were associated with constipation at age 14 years.
Our findings are consistent with previous studies which have found associations between autism/autistic traits and incontinence/constipation (Cuffman & Burkhart, 2021; Gubbiotti et al., 2019; Niemczyk et al., 2019; Peters et al., 2014), but these studies were either restricted to specialised settings such as paediatric clinics (Cuffman & Burkhart, 2021), had small sample size (Niemczyk et al., 2019), or used a cross-sectional design (Peters et al., 2014). Strengths of our study include the prospective birth cohort design, and the availability of data on diagnosed autism in addition to different autistic traits. The autistic traits we examined have previously been found to have good predictive validity for autism ((Rai et al., 2018) and enabled us to examine if there are differential associations with incontinence/constipation. Other strengths include the availability of data on both parent-reported (in late childhood) and self-reported (in adolescence) incontinence/constipation, and a range of confounders. This large community-based cohort includes majority of the children in the population who do not meet diagnostic criteria for autism. This is important because autistic traits are believed to be on a continuum (Marinopoulou et al., 2023) and means that our findings generalise beyond children with a clinical diagnosis of autism.
Although we adjusted for a range of potential confounders, the possibility of residual and unmeasured confounding cannot be ruled out, which limits our ability to infer causality. This study, like other cohort studies suffered from missing data and attrition, which could potentially cause selection bias since the complete case sample was more socially advantaged than the original cohort. Our estimates, however, should be unbiased provided there are no systematic differences in the rates of the outcomes considered after conditioning on the exposure and confounders included in the model (Hughes et al., 2019), since we accounted for a number of confounders including known predictors of missing data in ALSPAC and childhood incontinence in ALSPAC has been shown to be only weakly socially patterned (Butler & Heron, 2008), we believe this assumption is tenable in this current study.
Compared with the autistic traits, our study had lower statistical power to examine associations between diagnosed autism and incontinence/constipation due to the small number of autism cases in this community-based sample (see supplementary tables 3 and 4). This is reflected in the imprecise estimates with wide confidence intervals.
Another limitation of this study is the use of self-reported questionnaires for autistic traits and incontinence/constipation. However, differential misclassification is unlikely because this is a prospective study, and the outcomes (incontinence/constipation) were not known when parents provided data on the autistic traits. The exception is the coherence subscale of the Children’s Communication Checklist, which was completed at 9 years, which might explain why associations of this autistic trait with incontinence at age 9 were stronger. It is notable, however, that coherence difficulties at age 9 were prospectively associated with soiling and constipation at age 14.
It is possible that the stronger associations of coherence difficulties and social-communication with incontinence may be attributed to the concept of fractionation of component features of autism (Happé & Ronald, 2008). This concept suggests that social and non-social aspects of autism may have distinct causes, and hence may present with different symptoms (Happé & Ronald, 2008). The timing of assessment of autistic traits (sociability and repetitive behaviours) could also have affected the strength of the associations. Compared with social-communication coherence difficulties (which were reported at ages 7 and 9 respectively), the sociability and repetitive behaviour scales were completed when the children were younger (at ages 3 and 5 years), hence certain features might not have been fully developed or expressed. The stronger associations between autistic traits and constipation at age 14 years than at age 9 years could be due to self-reporting of constipation at age 14 years by the study children, compared with the use of parent-reports at age 9 which might have missed constipation (Joinson et al., 2019).
Developmental delays in some children with autism may lead to difficulties with toilet training, which could explain the associations between autistic traits/autism and incontinence. When continence has been achieved, about 30% of children with autistic traits regress (Dalrymple & Ruble, 1992). Stool withholding, which can arise from a previous painful or distressing bowel movement (Dalrymple & Ruble, 1992; Mugie et al., 2011) is more common in children with autism (Dalrymple & Ruble, 1992; Mugie et al., 2011) and this is associated with constipation and soiling (Dalrymple & Ruble, 1992; Mugie et al., 2011). A study found that about 32% of parents of children with autistic traits reported that their child had fears related to toileting (sitting on the toilet, toilet flushing, wiping up bowel movements, being alone in the bathroom, and other fears) (Dalrymple & Ruble, 1992). Children with autistic traits also have neurological and sensory processing difficulties that affect how they perceive and respond to bodily sensations, including those related to bladder and bowel functions (McElhanon et al., 2014; Woodward, 1996).
Previous studies have indicated that children with autistic traits often have restricted diets or preferences for starches, snack foods, and processed foods and avoid certain textures or types of foods including fruits, vegetables, and proteins (Harris et al., 2021; McElhanon et al., 2014). It is well established that diets deficient in fibre increase the risk of constipation (Cummings, 1984; Mugie et al., 2011). Studies have also found evidence that children with autistic traits may have altered microbiome (Dan et al., 2020; Son et al., 2015), which might be a consequence of restricted diet (Son et al., 2015). Alterations of intestinal microbiota may contribute to constipation and constipation-related symptoms (Zhang et al., 2021).
Genetic factors might also explain the associations between autistic traits/autism and incontinence. Studies suggest that autistic traits may result from many different patterns of genetic causality including single genes, combinations of several genes, and large groups of genes (Wei et al., 2021). Consequently, there could be shared genes for autistic traits and incontinence/constipation, that could be a common underlying cause.
This population-based study provides new evidence that children with autistic traits/autism are more likely to experience subsequent problems with incontinence/constipation. Social-communication and coherence difficulties were more strongly associated with incontinence than the other autistic traits. Incontinence and constipation can significantly impact a person's quality of life, leading to embarrassment, social isolation, emotional distress, a loss of confidence, and also has an impact on mental health (Mari et al., 2022; Gordon et al., 2023; Whale et al., 2018; Grzeda et al., 2017). Early assessment and treatment for incontinence and constipation should be considered for children with autistic traits to reduce the risk of incontinence/constipation becoming chronic.