Mental health and social difficulties of late-diagnosed autistic children, across childhood and adolescence

Harsh Parenting and Trajectories of Emotional and Behavioural Difficulties in Autistic Children

Autistic children show higher rates of co-occurring emotional and behavioural difficulties compared to other children in the general population. However, the environmental factors which contribute to the development of emotional (internalising) and behavioural (externalising) difficulties in autistic individuals are poorly understood. This study sought to investigate the association between harsh parenting (smacking, shouting, telling off) and the trajectories of emotional and behavioural difficulties in autistic children from ages 3 to 7. A sample comprising of 349 autistic children participated from the UK's Millennium Cohort Study. Associations between harsh parenting and child emotional and behavioural difficulties were modelled using multilevel growth curve models. In autistic children, harsh parenting was associated with total emotional and behavioural difficulties and behavioural (but not emotional) problems concurrently. Moreover, harsh parenting was not associated with changes in emotional and behavioural difficulties over time. Harsh parenting may have an important role in externalising problems in young autistic children indicating the need for public health strategies which educate parents on its effects.

Who, when, where, and why: A systematic review of "late diagnosis" in autism

An autism diagnosis can be a critical milestone toward effective and affirming support. Despite the sharp increase in the number of studies focused on late diagnosis over the last 15 years, there remains no consensus as to what constitutes a late diagnosis of autism, with cutoffs ranging from infancy to middle adulthood. This preregistered systematic review evaluated (a) the field's current quantification of late diagnosis in autism, (b) how the threshold for late diagnosis varies as a function of demographic and population factors, and (c) trends over time. Of the 11,697 records retrieved, N = 420 articles met inclusion criteria and were extracted. Articles spanned 35 years (1989-2024) and included participants from every continent except Antarctica. Only 34.7% of included studies provided a clear threshold for "late diagnosis" (n = 146/420). Late diagnosis cutoffs averaged 11.53 years (range = 2-55 years; median = 6.5 years) with a bimodal distribution (3 and 18 years). The threshold for late diagnosis varied by participant location, F(5,140) = 10.4, p < 0.0001, and sample age, F(5,140) = 20.1, p < 0.0001. Several key rationales for age determinations emerged, including access to services, considerations for adult diagnoses, and data driven approaches. What authors consider to be a "late" diagnosis of autism varies greatly according to research context. Justifications for a specific late-diagnosis age cutoff varied, underscoring the need for authors to contextualize their conceptualizations.

Time is of the essence: Age at autism diagnosis, sex assigned at birth, and psychopathology

LAY ABSTRACT

Previous research has shown that girls/women are diagnosed later than boys/men with autism. Individuals who are diagnosed later in life, especially girls/women, have greater anxious and depressive symptoms. Previous research has been limited due to narrow inclusionary criteria for enrollment in studies. The present study uses two samples-one clinic-based, large "real-world" sample and another research-based sample with strict criteria for autism diagnosis-to understand the relationships between diagnostic age, sex assigned at birth, and symptoms of anxiety/depression. In both samples, those who were diagnosed later had greater anxious/depressive symptoms, and anxiety was not predicted by sex. In the clinic-based but not research-based sample, those assigned female at birth were diagnosed later than those assigned male at birth. In the clinic-based sample only, individuals assigned female at birth and who were later diagnosed experienced greater symptoms of anxiety/depression compared to those assigned male who benefited from earlier diagnostic timing. Within the research-based sample, those assigned female at birth had greater depressive symptoms than those assigned male. These findings highlight the importance of timely identification of autism, especially for girls/women who are often diagnosed later. Community-based samples are needed to better understand real-world sex-based and diagnostic age-based disparities in mental health.

Twenty-four-month effortful control predicts emerging autism characteristics

Longitudinal research can assess how diverging development of multiple cognitive skills during infancy, as well as familial background, are related to the emergence of neurodevelopmental conditions. Sensorimotor and effortful control difficulties are seen in infants later diagnosed with autism; this study explored the relationships between these skills and autism characteristics in 340 infants (240 with elevated familial autism likelihood) assessed at 4-7, 8-10, 12-15, 24, and 36 months. We tested: (1) the relationship between parent-reported effortful control (Rothbart's temperament questionnaires) and sensorimotor skills (Mullen Scales of Early Learning), using random intercept cross-lagged panel modelling; (2) whether household income and maternal education predicted stable individual differences in cognition; (3) sensorimotor and effortful control skills as individual and interactive predictors of parent-reported autism characteristics (Social Responsiveness Scale) at 3 years, using multiple regression; and (4) moderation of interactions by familial likelihood. Sensorimotor skills were longitudinally associated with effortful control at the subsequent measurement point from 12-15 months. Socioeconomic status indicators did not predict stable between-infant differences in sensorimotor or effortful control skills. Effortful control skills were longitudinally related to 3-year autism characteristics from the first year of life, with evidence for an interaction with sensorimotor skills at 24 months. Effects of effortful control increased with age and were particularly important for infants with family histories of autism. Results are discussed in relation to different theoretical frameworks: Developmental Cascades and Anterior Modifiers in the Emergence of Neurodevelopmental Disorders. We suggest a role for 24-month effortful control in explaining the emergent autism phenotype. RESEARCH HIGHLIGHTS: Sensorimotor skills longitudinally predicted effortful control from 12-15 months onward but effortful control did not longitudinally predict sensorimotor skills during infancy. Measures of effortful control skills taken before the age of 1 predicted continuous variation in autism characteristics at 36 months, with associations increasing in strength with age. Effortful control (measured at 12-15 and 24 months) was a stronger predictor of 36-month autism characteristics in infants with elevated familial likelihood for autism. The relationship between 24-month sensorimotor skills and 36-month autism characteristics was stronger in infants with weaker effortful control skills.

Gender, assigned sex at birth, and gender diversity: Windows into diagnostic timing disparities in autism

LAY ABSTRACT

Later autism diagnosis is associated with risk for mental health problems. Understanding factors related to later autism diagnosis may help reduce mental health risks for autistic people. One characteristic associated with later autism diagnosis is female sex. However, studies often do not distinguish sex assigned at birth and gender identity. Gender diversity may be more common in autistic relative to neurotypical people, and autism is more common in gender-diverse populations. We studied age at autism diagnosis by sex assigned at birth, gender identity, and gender diversity (gender-diverse vs cisgender) status, separately. We studied three separate autistic samples, each of which differed in how they were diagnosed and how they were recruited. The samples included 193 persons (8.0-18.0 years) from a research-recruited academic medical center sample; 1,550 people (1.3-25.4 years) from a clinic-based sample; and 244 people (18.2-30.0 years) from a community-enriched sample. We found significant differences in the clinic-based and community-enriched samples. People assigned female sex at birth were diagnosed with autism significantly later than people assigned male at birth. People of female gender were diagnosed significantly later than people of male gender. Gender-diverse people were diagnosed significantly later than cisgender people. Sex assigned at birth, gender identity, and gender diversity may each show unique relationships with age of autism diagnosis. Differences in how autistic people are diagnosed and recruited are important to consider in studies that examine sex assigned at birth or gender identity. More research into autism diagnosis in adulthood is needed.

Determining unmet need: clinical relevance of suspected neurodivergence in first-episode psychosis

AIMS AND METHOD

We explored the prevalence of autism and attention-deficit hyperactivity disorder in first-episode psychosis. Through service evaluation involving 509 individuals, detailed analyses were conducted on neurodevelopmental traits and patterns of service utilisation.

RESULTS

Prevalence of neurodivergence in first-episode psychosis was 37.7%. Neurodivergent individuals used urgent mental health services more frequently (Mann-Whitney U = 25925, Z = -2.832, P = 0.005) and had longer hospital stays (Mann-Whitney U = 22816, Z = -4.886, P ≤ 0.001) than non-neurodivergent people. Neurodivergent people spend more than twice as long in mental health hospitals at a time than the non-neurodivergent people (Mann-Whitney U = 22 909.5, Z = -4.826, P ≤ 0.001). Mediation analysis underscored indirect impact of neurodivergence on hospital stay durations through age at onset of psychosis and use of emergency services.

CLINICAL IMPLICATIONS

Prevalence of neurodevelopmental conditions in first-episode psychosis is underestimated. Neurodivergent individuals show increased utilisation of mental health services and experience psychosis earlier. Early assessment is crucial for optimising psychosis management and improving mental health outcomes.

Evaluation of the Behavioral Effect of Psychostimulants in Children with Autism Spectrum Disorder: A Cross-Sectional Study

BACKGROUND

 Autism spectrum disorder (ASD) is often accompanied by comorbid conditions such as attention deficit hyperactivity disorder and epilepsy. In this context, patients are often treated with psychostimulants in an attempt to control behavioral symptoms. This study aims to understand the behavioral effects of psychostimulants in children with ASD and investigate if interictal epileptiform discharges on electroencephalogram (EEG) can act as a modifying factor in this behavior.

METHODS

 Sixty-eight patients with ASD who were being accompanied in the Department of Child and Adolescent Psychiatry of the Centro Hospitalar Universitário de São João and had previously done an EEG assessment answered a questionnaire regarding their behavioral response to psychostimulants.

RESULTS

 In total, 47.4% of patients reported improved agitation, 56.1% enhanced concentration, and 8.8% improved sleep. Conversely, 28.1% experienced worsened agitation, 15.8% worsened concentration, and 17.5% worsened sleep. The remaining reported no alterations. The age of diagnosis correlated significantly with improved agitation, with a higher diagnosis age being associated with a higher probability of improvement. Extended-release methylphenidate and genetic variations were significantly associated with worsening of agitation. Regarding speech, 86% exhibited no changes, while 14% showed alterations, mostly, 87.5%, characterized as negative. For other behavioral alterations, 45.6% reported negative changes, 3.5% reported positive changes, and 50.9% reported no additional alterations. Female gender was significantly associated with other negative behavioral changes. A significant correlation was found between treatment duration and the probability of improvement in agitation, concentration, and other behavioral changes.

Pourcain B, Robinson EB, Bourgeron T, Baron-Cohen S, Børglum AD, Martin HC, Warrier V. An axis of genetic heterogeneity in autism is indexed by age at diagnosis and is associated with varying developmental and mental health profiles

There is growing recognition that earliest signs of autism need not clearly manifest in the first three years of life. To what extent is this variation in developmental trajectories associated with age at autism diagnosis? Does the genetic profile of autism vary with age at autism diagnosis? Using longitudinal data from four birth cohorts, we demonstrate that two different trajectories of socio-emotional behaviours are associated with age at diagnosis. We further demonstrate that the age at autism diagnosis is partly heritable (h2 SNP = 0.12, s.e.m = 0.01), and is associated with two moderately correlated (rg = 0.38, s.e.m = 0.07) autism polygenic factors. One of these factors is associated with earlier diagnosis of autism, lower social and communication abilities in early childhood. The second factor is associated with later autism diagnosis, increased socio-emotional difficulties in adolescence, and has moderate to high positive genetic correlations with Attention-Deficit/Hyperactivity Disorder, mental health conditions, and trauma. Overall, our research identifies an axis of heterogeneity in autism, indexed by age at diagnosis, which partly explains heterogeneity in autism and the profiles of co-occurring neurodevelopmental and mental health profiles. Our findings have important implications for how we conceptualise autism and provide one model to explain some of the diversity within autism.

Females evaluated for autism: characteristics and co-occurring and differential DSM-5 diagnoses

Objective: While sex differences in autism spectrum disorder (ASD) have been identified in areas such as neurocognitive functioning, behavior patterns, and diagnostic criteria, less work has focused on differences within females referred for ASD evaluation, including those who did not go on to receive a diagnosis. This study examined psychological and behavioral characteristics and co-occurring and differential DSM-5 diagnoses between pediatric female participants who received an ASD diagnosis (ASD+) and those who did not (ASD-). Method: Data on cognitive functioning, adaptive functioning, internalizing symptoms, externalizing symptoms, and ADOS-2 scores were analyzed among 137 3- to 20-year-old patients. The sample was divided into two age groups (ages 3-8 and ages 9-20) for analyses of between-group differences (ASD+ vs. ASD-) and predictors of group membership. Results: Females in the ASD+ group were significantly younger, had lower cognitive scores, lower internalizing and externalizing symptoms, and had higher Autism Diagnostic Observation Schedule-2 (ADOS-2) scores than those in the ASD- group. ADOS-2 scores were also the only significant predictor of ASD group membership across age groups. The ASD+ group had a higher percentage of intellectual disability while the ASD- group had higher percentages of anxiety disorder, attention-deficit/hyperactivity disorder, and disruptive behavior disorders. Conclusions: Psychological and behavioral presentations among females referred for ASD evaluation varied with age and ASD diagnostic groups. These results highlight potential female differences in ASD referrals and identification of ASD and the need to improve care for females in consideration of demographic factors.

Potential mechanisms underlying the association between feeding and eating disorders and autism

There is a reliable association between autism and Feeding and Eating Disorders. Concerningly, where these two conditions co-occur, clinical outcomes of Feeding and Eating Disorders are significantly worse, and treatment less effective, than when the Feeding and Eating Disorders occur in neurotypical individuals. Problematically, the reason for the association between autism and Feeding and Eating Disorders is poorly understood, which constrains advances in clinical care. This paper outlines several possible mechanisms that may underlie the observed association and suggests ways in which they may be empirically tested. Mechanisms are split into those producing an artefactual association, and those reflecting a genuine link between conditions. Artefactual associations may be due to conceptual overlap in both diagnostic criteria and measurement, Feeding and Eating Disorders causing transient autistic traits, or the association being non-specific in nature. A genuine association between autism and Feeding and Eating Disorders may be due to common causal factors, autism directly or indirectly causing Feeding and Eating Disorders, and Feeding and Eating Disorders being a female manifestation of autism.

Case report: Diagnostic challenges in an adolescent case of autistic catatonia

Catatonia is a complex neuropsychiatric syndrome involving a constellation of psychomotor disturbances including catalepsy, waxy flexibility, stupor, mutism, negativism, agitation, posturing, stereotypes, mannerisms, grimacing, echolalia, and echopraxia. Catatonia occurs in several conditions including psychotic, affective and neurodevelopmental disorders such as autism spectrum disorder (ASD). ASD is a neurodevelopmental disorder characterized by persistent deficits in communication, social interaction, restricted interests, repetitive behaviours and sensory sensitivities. Catatonia can occur in response to life stressors such as extreme fear or threat, interpersonal conflict, tragic events or following significant loss. Those with ASD may be particularly vulnerable to the negative impact of stressors and the link between catatonia and ASD is being increasingly recognized. The overlapping features of catatonia and ASD make it difficult to differentiate often resulting in delayed or missed diagnosis. Catatonia in ASD remains a significant clinical challenge; it is difficult to diagnose and can pose debilitating difficulties for those affected. Catatonia is a treatable condition and prompt recognition is vital in securing the best possible outcome. We report a complex and unique case of a 15-year-old boy who presented with severe cognitive and functional decline with a background history of significant bullying and deterioration in his mental state. This case posed a diagnostic conundrum leading to a diagnosis of underlying ASD, anxiety and trauma.

Dropping the mask: It takes two

LAY ABSTRACT

In some situations, autistic people feel pressure to change their social behaviour by camouflaging. In other situations, autistic people feel they don't need to change their social behaviour. Instead, they feel they can socialise in ways that feel authentic or true to themselves. Past research has tended to focus on autistic people's experiences of camouflaging rather than their experiences of authenticity. In this study, we asked autistic people what it is like for them when they can socialise in ways that feel authentic or true to themselves. Autistic people described authentic-feeling socialising as more free, spontaneous and open than camouflaging. In supportive environments, this kind of socialising had more positive and less negative consequences than camouflaging. Autistic people felt that having self-awareness and acceptance of their own social needs and being around autistic and nonautistic people who were accepting and understanding helped them to socialise in authentic-feeling ways. Autistic people also spoke about communication behaviours they felt nonautistic people should use to help overcome misunderstandings and create autism-friendly social environments. These findings suggest it is helpful for autistic people to have access to supportive and accepting social environments in which they feel able to socialise in ways that feel authentic to them. In creating such social environments, it is important to focus on nonautistic people's knowledge and attitude towards autistic people and also their ability to use helpful communication behaviours.

Does camouflaging predict age at autism diagnosis? A comparison of autistic men and women

It is frequently reported that females are likely to receive an autism diagnosis at a later age than their male counterparts, despite similar levels of autistic traits. It has been suggested that this delay in diagnosis may in part reflect the propensity of females, more than males, to engage in camouflaging behaviors that reduce the appearance of autism-related traits. This article presents two studies which examined the relationship between gender/sex, camouflaging, and age at diagnosis in two samples of (cis-gender) autistic adults. Study 1 included data from three online samples including 242 autistic men and 570 autistic women aged 18-75 years. Study 2 included data from a longitudinal population-based sample including 24 autistic men and 35 autistic women aged 20-24 years. Camouflaging was measured with the self-report Camouflaging Autistic Traits Questionnaire (CAT-Q). Overall, the results showed that, on average, females were diagnosed later than males. There was a stronger relationship between camouflaging and age at autism diagnosis (AaD) for females, compared with males. Within sample one, there was a significant camouflaging-by-sex interaction; high-camouflaging females had a later AaD. The role of autistic traits and changes in attitudes towards female autism and camouflaging need further exploration. These findings highlight the need for greater clinician and key stakeholder awareness and understanding of camouflaging behavior, particularly for females, during the diagnostic process.

Exploring camouflaging by the Chinese version Camouflaging Autistic Traits Questionnaire in Taiwanese autistic and non-autistic adolescents: An initial development

LAY ABSTRACT

Camouflaging is a coping strategy used by some autistic and other neurodivergent people to fit in neurotypical social contexts. The self-reported Camouflaging Autistic Traits Questionnaire has been validated for use in research with adults in some Western societies, but not in non-Western cultural-ethnic groups. We translated Camouflaging Autistic Traits Questionnaire into traditional Chinese and examined the use of this measure in Taiwanese adolescents via both self-report and caregiver-report in 100 autistic and 105 non-autistic adolescents. Both self-reported and caregiver-reported Chinese version Camouflaging Autistic Traits Questionnaire were composed of two factors (i.e. a "compensation-masking" subscale and an "assimilation" subscale). Both adolescent self-reported and caregiver-reported Chinese version Camouflaging Autistic Traits Questionnaire total score and subscales were reliable in measurement, and they highly correlated with each other. Taiwanese autistic adolescents were more likely to camouflage than non-autistic adolescents, especially on assimilation. Female autistic adolescents showed higher assimilation than male autistic adolescents. Higher camouflaging, especially assimilation, was associated with higher stress in autistic and non-autistic adolescents alike. Both self-reported and caregiver-reported Chinese version Camouflaging Autistic Traits Questionnaire were reliable and offered meaningful information to help us understand the social coping experiences of autistic and non-autistic adolescents.

Identifying Risk Factors Associated with Repeated Referrals Within a Pediatric Navigation Program

Approximately 1-in-5 children have a diagnosed mental, behavioral, and/or developmental disorder or delay by age 8 in the United States. Children with such conditions often require complex, complicated diagnostic and specialty care, making them susceptible to repeated referrals and ongoing unmet healthcare needs. Patient navigation programs (PNPs) are designed to integrate care from primary care providers to community-based services, using trained navigators to help patients and their families manage referrals and connect with referred services. This study examines factors associated with repeated referrals to an active PNP to inform ongoing referral patterns and adaptations to standard navigation support within a large healthcare system in South Carolina (SC). Data is sourced from the inception of the PNP in 2017 through 2022, including 15,702 referrals. Overall, 71.07% had no repeated referrals. Children who are older, diagnosed with attention deficit disorder(s), behavioral concerns, depression, multiple referral needs, and insured by Medicaid were found to be most susceptible to repeated referrals. Conversely, children who are non-Hispanic Black, were referred at a well-child visit, and are primarily insured by private insurance or Tricare were least likely to have repeated referrals. Children who are insured by Medicaid are more likely to be younger, identify as non-Hispanic Black, Hispanic, or another race/ethnicity, and have multiple needs at time of initial referral, identifying a potentially compounded risk for those who hold multiple risk factors to experiencing repeated referrals. Findings may inform adaptations to this PNP model to adjust navigator protocol for at-risk populations and equitably optimize referral-to-service connection.

Psychiatric conditions in autistic adolescents: longitudinal stability from childhood and associated risk factors

Autistic people experience high rates of co-occurring psychiatric diagnoses. Current prevalence estimates vary considerably due to an over-reliance on clinical cohorts and the longitudinal stability of diagnoses from childhood into adolescence is poorly understood. This study aims to provide prevalence rates of co-occurring DSM-5 psychiatric diagnosis for autistic adolescence and investigate, for the first time, the stability of diagnoses from childhood. Using a longitudinal stratified sample of autistic youth (N = 77; 13-17 years; 60% male), selected from a larger community-derived sample of those with pre-existing autism diagnoses (N = 277) weighted prevalence estimates of emotional (anxiety, depression), behavioural (oppositional and conduct disorders) and ADHD diagnoses were calculated based on semi-structured psychiatric interview. Prediction of adolescent psychiatric diagnosis based on childhood diagnostic status, sex, childhood IQ (both assessed at age 4-10 years) was tested. Emotional and behavioural disorders in adolescence were particularly prevalent, and significantly predicted by childhood disorder status. Attention-deficit/hyperactivity-disorder (ADHD) was prevalent but not predicted by childhood ADHD diagnosis. Neither sex nor childhood IQ predicted diagnostic outcomes. Autistic youth have high levels of co-occurring psychiatric conditions, which are broadly persistent across childhood and adolescence. Emotional disorders are particularly prevalent and remain persistent from childhood to adolescence. Greater diagnostic variability was found for ADHD with more adolescents moving across diagnostic thresholds.

Non-suicidal self-injury and its relation to suicide through acquired capability: investigating this causal mechanism in a mainly late-diagnosed autistic sample

BACKGROUND

Non-suicidal self-injury (NSSI) has been linked with a higher risk of suicide attempts in autistic and non-autistic people. In the general population, NSSI may confer acquired capability for suicide by eroding one's fear and avoidance of pain and death. The present study aimed to explore acquired capability as the mediator of increased suicide risk conferred by NSSI in autistic and non-autistic adults.

METHODS

Autistic and non-autistic adults (n = 314, n = 312) completed an online survey exploring lifetime suicide attempts, experience with NSSI, and acquired capability for suicide. We explored relationships between lifetime incidence of NSSI and lifetime suicide attempts via three facets of acquired capability (pain tolerance, reduced fear of death, and mental rehearsal of suicide). In self-harming participants (224 autistic and 156 non-autistic), we explored whether particular types and features of NSSI might be especially associated with capability and through that with suicide: namely engagement in scratching, cutting, and self-hitting, and engaging in more numerous forms of NSSI.

RESULTS

While a higher frequency of NSSI was associated with all three facets of acquired capability, only reduced fear of death and mental rehearsal of suicide mediated an indirect relationship with lifetime suicide attempts. NSSI also directly predicted more numerous suicide attempts. Autistic people tended towards reduced fear of death and mental rehearsal regardless of NSSI status. Among self-harming autistic and non-autistic participants, cutting and an increased number of NSSI behaviours were associated with lifetime suicide attempts directly and indirectly via acquired capability. In both groups, self-hitting was associated with lifetime suicide attempts only via acquired capability.

LIMITATIONS

Our cross-sectional methodology negates inferences of directionality. While we controlled for age, our samples were poorly matched, with the autistic group two times older on average. The autistic sample, predominantly late-diagnosed, female and highly qualified, were unrepresentative of the whole autistic community.

CONCLUSIONS

Our data suggest that acquired capability, as measured herein, is an incomplete explanation for the association between NSSI and suicide risk. A broader construct with stable and transient facets may offer greater explanatory power, but it is probable that other variables explain or provide additional means through which this association arises.

The relevance of the interpersonal theory of suicide for predicting past-year and lifetime suicidality in autistic adults

Background

While there are known risk factors for suicidality in autistic adults, these are often unconnected from theoretical frameworks that might explain why risk is elevated and guide clinical interventions. The present study investigated the relevance of constructs from the Interpersonal Theory of Suicide (ITS), including perceived burdensomeness, thwarted belongingness and acquired capability for suicide, and explored mechanisms through which certain risk factors (relationship status, age at diagnosis) might elevate suicide risk.

Methods

Autistic adults (n = 314) completed an online study including measures of depression, anxiety and constructs from the ITS. Linear and multinomial regression analysis disentangled contributions of ITS variables from effects of depression and anxiety for past-year suicide ideation, past-year and lifetime suicide attempts. Mediation analyses examined associations between risk factors and these suicide outcomes via mechanisms proposed by the ITS.

Results

Past-year suicide ideation was associated with burdensomeness, mental rehearsal of suicide plans (a facet of acquired capability), and depression. Greater feelings of burdensomeness, and reduced fear of death, marked out participants who had attempted suicide in comparison to those who had experienced suicide ideation in the past year. Relationship status was indirectly associated with past-year suicide ideation via the mediators of depression and burdensomeness, and was associated with past-year attempts via its effect on ideation. Age at diagnosis was unrelated to any variables.

Limitations

Cross-sectional research is insensitive to causality and temporal dynamics, which is likely why interaction hypotheses from the ITS were unsupported. Normative measures may be invalid in autistic samples. There was no control group. The autistic sample was unrepresentative of the whole population, particularly autistic people with intellectual disabilities, ethnic/racial minorities, and gender minorities.

Conclusions

Perceived burdensomeness and acquired capability appear potentially important to suicide in autistic people, and may mediate the effects of some risk factors. Future research should explore the temporal dynamics of suicide trajectories in longitudinal, prospective designs.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13229-022-00495-5.