Autism spectrum disorder (ASD) is a neurodevelopmental disorder marked by impairments in social interaction and communication, and restricted and repetitive interests and behaviors. While the etiology of ASD is not well understood, studies suggest contributions from both genetic and environmental factors (McGuinn et al., 2020; Hallmayer et al., 2011; Colvert et al., 2015).

Later-born siblings of children with ASD are more likely to be diagnosed with ASD than later-born siblings of children not diagnosed with ASD (Miller et al., 2019). Both sibling concordance rates and the results from genetic linkage analyses are consistent with a polygenic mode of transmission (Hallmayer et al., 2011; Colvert et al., 2015; Constantino et al., 2010; Nordenbæk et al., 2014). Additionally, epidemiologic studies have reported associations between ASD and early life air pollution exposure (McGuinn et al., 2020; Flores-Pajot et al., 2016; Volk et al., 2020) and maternal health conditions, such as autoimmune conditions, infertility disorders, obesity, and pregnancy complications (e.g., hypertension), all of which can recur in subsequent pregnancies (Croen et al., 2019; Schieve et al., 2017; Matias et al., 2021; Cordero et al., 2019). Thus, both genetic and non-genetic environmental and maternal factors may explain the high rates of ASD within sibling groups (Croen et al., 2019; Schieve et al., 2017; Matias et al., 2021; Cordero et al., 2019).

Previous studies that evaluated the health and development of siblings of individuals with ASD report familial clustering of ASD (Miller et al., 2019; Jokiranta-Olkoniemi et al., 2016; Ozonoff et al., 2011; Shephard et al., 2017). Studies also report other (non-ASD) developmental disabilities in siblings of children with ASD, most notably attention-deficit/hyperactivity disorder (ADHD) and/or ADHD-related symptomatology (Lin et al., 2021; Miller et al., 2019; Septier et al., 2019; Ghirardi et al., 2018; Chien et al., 2017; Shephard et al., 2017; Jokiranta-Olkoniemi et al., 2016). Using medical records data from two large US health care systems. Miller et al. (2019) found that later-born siblings of children with ASD or ADHD were at elevated risk for the same disorder, and also of being diagnosed with the other disorder (Miller et al., 2019). In a case-control study conducted at three clinical centers in France, ADHD was found to be 3–4 times more frequent in relatives of children with ASD than in relatives of a control group of unaffected children (Septier et al., 2019). Studies based on registry-based data in Sweden (Ghiradi et al., 2018) and Finland (Jokiranta-Olkoniemi et al., 2016) reported a strong co-aggregation between ASD and ADHD among individuals from the same family. Chien et al. (2017) assessed sibling pairs enrolled through a clinical center in Taiwan and found that siblings of children with ASD had higher than expected levels of hyperactivity and impulsivity, and reduced levels of attention (Chien et al., 2017). A subsequent longitudinal study using Taiwan’s National Health Insurance Research Database also found that unaffected siblings of children with ASD were more likely to develop ADHD compared with a control group (7.8% vs. 2.3% in controls; Lin et al., 2022).

Smaller case series with in-depth developmental assessments support the findings from larger studies, demonstrating a high proportion of ASD case siblings have ADHD and anxiety symptomatology (Shephard et al., 2017). The purported mechanism for ADHD in siblings of ASD cases is that the two conditions partially share both genetic and other biological causes (Miller et al., 2019; Musser et al., 2014; Oerlemans et al., 2015).

Studies of other non-ADHD developmental outcomes in siblings include the longitudinal study from Taiwan, which reported increased diagnoses of speech or language disorder, developmental coordination disorder, intellectual disability (ID), anxiety disorder, and depression among siblings of individuals with ASD (Lin et al., 2022). A nested case-control study within the Finish registry reported increased diagnoses of learning disorders, coordination disorders, ID, oppositional defiant disorder, and emotional disorders among siblings of children with ASD (Jokiranta-Olkoniemi et al., 2016). Various small clinical studies demonstrate that siblings of children with ASD manifest cognitive, language, or learning difficulties (Gamliel et al., 2009; Georgiades et al., 2013; Piven et al., 1990; Fombonne et al., 1997; Folstein et al., 1999). Additionally, a meta-analysis that assessed numerous studies of developmental and psychological outcomes reported that overall, siblings of individuals with ASD were significantly more likely to have internalizing behavioral problems and psychological problems than comparison groups; however, the authors note the effect size was small and that across studies inconsistent outcome measures were used and there was variation in the findings for individual studies, with numerous studies reporting no significant differences between sibling groups (Shivers et al., 2019).

In sum, while the co-aggregation of ASD and ADHD among sibling groups has been fairly well-established, non-ADHD developmental and psychological outcomes in siblings has been less well studied. Moreover, studies to date have been hampered to varying extents by limitations including small non-generalizable clinical samples, lack of/limited comparison groups, non-standardized measures of ASD and other DD diagnoses among both index children and siblings, and passive outcome ascertainment methods based on clinical diagnoses. Indeed, one of the most robust studies to date, which used a longitudinal design, did not prospectively evaluate study participants with standardized developmental assessments, but instead relied on diagnoses in children’s medical records; given the timeframe for This study was more than 20 years ago, the presence of DDs among siblings was likely underestimated and potentially biased given parents who already had one child with ASD might have differentially sought clinical developmental evaluations for their other children (Lin et al., 2022).

We examined ASD and other developmental, psychiatric, and neurological conditions among the siblings of children with or without ASD who were included in the Study to Explore Early Development (SEED), a large case-control study of ASD conducted in several US states. In addition to its large size, SEED’S population-based methodology enrolled a much more racially diverse sample than most previous studies assessing this topic. SEED also collected a much larger number of developmental, psychiatric, and neurologic conditions among siblings than most previous studies. Additionally, SEED has several unique features that add to the robustness of this study. ASD case classification in index children did not rely on past diagnoses but instead included children with symptomatology without a diagnosis as all children’s final case status was based on an in-person standardized developmental assessment. SEED included two control groups – children from the general population, and children with non-ASD developmental conditions – that provided data to help disentangle familial clustering that was specific to ASD from that common to developmental disorders generally. Thus, this study adds to the current understanding of the extent of familial clustering of developmental diagnoses in children, and the findings inform the comprehensive family care needs for children with developmental disorders.

For the analysis, we excluded children with only younger siblings or no siblings (n = 2,379), as well as children whose parent did not respond to the questions on conditions in siblings (n = 3) or responded “don’t know” for all conditions (n = 13). Our analytic sample included 738 index children with ASD, 1,186 index children with another DD, and 1,039 index children in the POP group.

When compared to POP group children, children with ASD or another DD were less likely to have been born to non-Hispanic white mothers, mothers with a bachelor’s degree or higher, and in high income families. Over 70% of index children in all groups were 48 months of age or older. ASD and DD group children were more likely to be male (82.0% and 66.0%) than POP group children (53.8%; p-values < 0.05). Over 75% of mothers in all groups were born in the United States and more than 85% reported English as their primary language, although mothers of POP children were more likely than mothers of ASD or DD children to be born in the United States and to speak English. Across study groups, most index children had only 1 older sibling, with a descending percentage having 2 or 3 or more older siblings (Table 1).

Index children in the ASD group were significantly more likely (p < 0.05) than index children in the POP group to have one or more older siblings with the following conditions: ASD, developmental delay, ADHD, ID, SID, learning disabilities, speech/language delays, and/or psychiatric diagnoses (Table 2). Index children in the DD group were significantly more likely than index children in the POP group to have one or more older siblings with all of the same conditions except psychiatric diagnoses. Index children in the ASD group were significantly more likely than index children in the DD group to have one or more older siblings with two conditions—ASD and developmental delay.

Overall, most results were consistent between unadjusted and adjusted analyses. The adjusted prevalence ratios (aPRs) comparing study groups (ASD, DD, POP) on whether index children had one or more older siblings with a developmental, psychiatric, or neurologic condition are shown in Fig. 1. The aPRs indicate that children with ASD were significantly more likely than children in the POP group to have one or more older siblings with ASD, developmental delay, ADHD, ID, SID, speech/language delays, and/or a psychiatric diagnosis, with point estimates ranging from 1.4 (for ADHD) to 3.7 (for ID and SID). Children in the DD group were also significantly more likely than children in the POP group to have an older sibling with most of the aforementioned conditions, with the exception of a psychiatric diagnosis, and most aPR point estimates were attenuated relative to those comparing ASD to POP, ranging from 1.4 (ADHD) to 2.2 (SID). Children with ASD were significantly more likely than children with other DDs to have an older sibling with an ASD (aPR = 1.9), developmental delay (aPR = 1.4), SID (aPR = 1.6), or a psychiatric diagnosis (aPR = 1.9).

The number of index children with one or more older siblings with parent-reported diagnosis of ASD is reported in Supplemental Table 1. Among index children with only one older sibling, we found that the percentage of older siblings with ASD varied significantly by index child study group—54 (11.1%) for the ASD group, 47 (6.9%) for the DD group, and 25 (3.7%) for the POP group. Among index children with two or more older siblings the likelihood of having an older sibling with ASD was even higher, particularly for the ASD study group. Among ASD index children with two older siblings, the percentage who had one or more older siblings with ASD was 20%.

When index children who had an older sibling with ASD were excluded from analyses, there were fewer significant differences between study groups in the percentage having an older sibling with a psychiatric or most developmental conditions (Supplemental Table 2). Index children with ASD remained significantly (p < 0.05) more likely than index children in the POP group to have one or more older siblings with ADHD (13.7% vs. 10.5%) or speech/language delays (13.3% vs. 9.2%). Index children in the DD group were significantly more likely than index children in the POP group to have an older sibling with developmental delay (6.1% vs. 3.5%), ADHD (16.4% vs. 10.5%), learning disabilities (7.8% vs. 4.2%), and/or speech/language delays (14.8% vs. 9.2%). Additionally, index children in the DD group were more likely than index children in the ASD group to have an older sibling with learning disabilities (7.8% vs. 4.6%). Among the three study groups, the most common conditions seen in older siblings without ASD were developmental delays, ADHD, learning disabilities, and speech/language delays.

Previous SEED analyses examined the prevalence of co-occurring developmental conditions and symptomatology in SEED index children and found that over half of the children in the ASD case group had one or more behavioral, developmental, medical, or psychiatric condition other than ASD (Wiggins et al., 2015a, 2022). Here, we compared developmental, psychiatric, and neurologic conditions among older siblings of children with ASD, children with other DDs, and children from the general population. We found that older siblings of children with ASD or with other DDs were more likely to have developmental conditions than older siblings of children in the POP comparison group. Older siblings of children with ASD were also more likely to have psychiatric conditions than either older siblings of children in the POP group or older siblings in the DD group. However, there was little variation between study groups in the three neurologic conditions we assessed — epilepsy and vision and hearing problems. Overall, our findings suggest that developmental disorders cluster in families. Thus, once a child is diagnosed with ASD or other DDs, enhanced monitoring of developmental and psychiatric problems in siblings might be warranted. While the current analysis also provides some insight into heritability, the data should be interpreted cautiously in that regard, given the small sample size of some conditions found in older siblings and that we were limited to using parent-reported data on conditions in siblings.

The rate of ASD among older siblings of index children with ASD in our study (i.e., the likelihood that ASD was also diagnosed in one or more older siblings if the index child had ASD) was 14%, which is within the range of the prevalence of recurrence (10–20%) reported in other studies (Miller et al., 2019; Jokiranta-Olkoniemi et al., 2016; Ozonoff et al., 2011; Sandin et al., 2014; Grønborg et al., 2013), although to be clear, our study did not measure recurrence of ASD, but rather what would be considered “pre-occurrence” in families. For index children with only one older sibling, siblings in the ASD group were 3 times more likely to be diagnosed with ASD compared to older siblings in the POP group and 1.6 times more likely compared to older siblings in the DD group. Although few children in the study group had more than one older sibling with ASD, we found that the likelihood of an index child with ASD having one or more older siblings with ASD nearly doubled (from 11 to 20%) when the number of older siblings increased from one to two. These findings could be due to genetic risks of ASD but also to non-genetic maternal risks equally common across pregnancies of the index child and their siblings.

Our finding of increased prevalence of diagnosed SID in older siblings of children with ASD supports previous results from several large, population-based samples that have indicated a high prevalence of sensory features in children with ASD (Wiggins et al., 2015a, 2022; Kirby et al., 2022). Assessments of sensory traits in the broader autism phenotype have found that parents and siblings of individuals with ASD show higher levels of self-reported sensory traits relative to the general population (Robertson & Baron-Cohen, 2017). Atypical sensory experience is estimated to occur in as many as 90% of individuals with ASD and to affect every sensory modality, and the latest diagnostic criteria for ASD include sensory sensitivities as a diagnostic feature (Robertson & Baron-Cohen, 2017). Notably, in our supplemental analyses in which we excluded older siblings who had an ASD diagnosis, SID was no longer differential between the siblings of ASD, DD, and POP group index children. These findings support the supposition that SID may be specific to the ASD phenotype.

Previous clinical studies have reported lower cognitive abilities among siblings of children with ASD (Georgiades et al., 2013; Piven et al., 1990; Folstein et al., 1999); however, other studies have reported no association, (Szatmari et al., 1993; Pilowsky et al., 2007; Ozonoff et al., 1993) or higher verbal intelligence quotient levels among siblings of children with ASD (Fombonne et al., 1997; Pilowsky et al., 2003). Here, we report a higher percentage of ID in older siblings of children with ASD or with other DDs; however as with SID, this finding was no longer evident when siblings with an ASD diagnosis were excluded from analysis, suggesting it was not an independent association.

As with ID, previous studies of psychiatric disorders (anxiety, depression, and bipolar disorder) in siblings of children with ASD have reported inconsistent findings (Lin et al., 2022; Lovell & Wetherell, 2016; Quintero & McIntyre, 2010; Sullivan et al., 2012; Selten et al., 2015). In the current study, index children in the ASD group were significantly more likely than children in either the POP or DD groups to have a sibling with a psychiatric condition. However, as with SID and ID, this differential was no longer observed in the sub-sample that excluded older siblings with an ASD diagnosis, which again, suggests that the high level of ASD recurrence in families explained the psychiatric condition finding. Nonetheless, in our and other studies, sample sizes limited the assessment of individual psychiatric conditions and, thus, negative findings should be cautiously interpreted.

In sum, our and previous reports support the possibility of cross-aggregation of ASD and various other conditions including SID, ID, and psychiatric conditions. However, in our study, we conducted supplemental analyses to further disentangle the underlying reasons developmental and psychiatric conditions were more prevalent in siblings. Those findings indicate that the positive findings initially observed for SID, ID, and psychiatric conditions likely reflect ASD clustering within many of those families, rather than independent familial clustering between ASD and the aforementioned conditions.

Nonetheless, we also found that several non-ASD conditions, including ADHD, developmental delay, and speech/language delays, were significantly higher in the older siblings of children with ASD than siblings of POP group children, even after the exclusion of siblings with a diagnosis of ASD. Thus, our finding of clustering of ASD with some non-ASD developmental conditions appears to be robust and not explained by the known high risk for ASD clustering within families. However, we found the same conditions were increased in siblings of children in the (non-ASD) DD group. Thus, the findings might not be specific to ASD co-aggregation.

Still, our findings that children with ASD were more likely than children in the POP group to have one or more older siblings with ADHD align with numerous previous studies and lend support to the hypothesis of some shared biological pathways between ASD and ADHD (Miller et al., 2019; Septier et al., 2019; Lin et al., 2022; Jokiranta-Olkoniemi et al., 2016). In addition to co-diagnoses of ASD and ADHD, studies suggest that many children with ADHD have symptomology that includes social and communication difficulties and, likewise, children with ASD commonly exhibit hyperactivity and attention difficulties (Mulligan et al., 2009). An etiologic link between ADHD and ASD symptoms has been supported by family and twin studies (Nijmeijer et al., 2009; Mulligan et al., 2009; Reiersen et al., 2008; Musser et al., 2014) and genetic analyses have documented shared copy number variants (Martin et al., 2014; Williams et al., 2012; Musser et al., 2014) and candidate genes (Zhou et al., 2023) between the two conditions, although the specific genetic findings vary across studies.

The composite findings from this study have implications for service needs of families. Children with ASD use more health services and have more difficulties accessing needed care than children in the general population (Kogan et al., 2018). Because of the multiple types of services needed by children with ASD, care coordination is challenging, and parents are less likely to report that their child’s care meets all of the criteria for a medical home than the parents of children without ASD. Particular difficulties include getting needed referrals and access to specialist care, such as mental health services (Kogan et al., 2018). When more than one child is affected with ASD or another DD, there is even greater cumulative health and family resource needs and increased caregiving burden. The data we report here, that nearly 40% of the older siblings of children with ASD had one or more developmental, psychiatric, or neurologic condition, provides a sense of how often families are caring for multiple children with developmental and mental health conditions and associated specialized service needs. Moreover, our data underscore the importance of monitoring siblings of children diagnosed with ASD or another DD, such that they can be identified and begin interventions as early as possible.

Limitations to this study should be considered in interpreting these findings. Given the young age of the index children in our sample, we excluded younger siblings from the analysis as they may not yet have had sufficient time for symptom manifestation and diagnosis for many developmental, psychiatric, and neurologic conditions examined (Soke et al., 2018). Thus, we may have underestimated the magnitude of clustering among siblings for some conditions. Additionally, some conditions, such as ADHD, may not be diagnosed until school-age (Soke et al., 2018); thus, while we may have limited misclassification by including only index children with older siblings in our study sample, it is possible that some older siblings who have ADHD were not yet diagnosed by the time of their family’s SEED participation. Conversely, having a child with an ASD or other developmental diagnosis may prompt families to pro-actively seek an evaluation and obtain a diagnosis for their other children. That is, for parents who have raised a child with ASD or another developmental condition, there may be increased sensitivity to mental and developmental symptoms found in other siblings (Lin et al., 2022). As parents’ awareness about developmental disorders increases through treating their child’s ASD or other DD, they may more readily seek help for siblings than parents who do not have a child with ASD or other DD diagnosis (Jokiranta-Olkoniemi et al., 2016). These parents also may more likely recall their other children’s health conditions than parents who do not have a child with a developmental disability. However, parents were provided with a preparation guide before the interview so they could collect information to reduce recall bias. Additionally, it’s also possible that some parents with more than one child with DD/ID, or ASD might not have participated in SEED due to time constraints caused by their high caregiving burden. However, a previous SEED analysis in one study site compared children who did and did not participate in the study and found little impact on several key birth characteristics or outcomes, suggesting the final study sample was robust (Schieve et al., 2018). We did not have the data to determine the genetic and non-genetic environmental contributions to the familial clustering of developmental disabilities that we observed in this study. Additionally, sample size constraints precluded us from adjusting for sex of the sibling when examining associations for each condition; given some conditions are more commonly diagnosed in males (Maenner et al., 2020; Arnett et al., 2015), our findings may have been influenced by unmeasured confounding. Finally, siblings’ diagnoses were reported by parents, and we did not have access to medical records or conduct study developmental assessments on siblings to confirm parent reports; nonetheless, question verbiage specified that conditions reported should have been diagnosed by a health care provider.

Despite these limitations, the SEED sample had many strengths, including recruitment methods that resulted in a diverse sample of children in all three study groups that better represented the racial, ethnic, and socioeconomic diversity in the US than previous studies on this topic (DiGuiseppi et al., 2016). While our sample size was not sufficient to examine all conditions of interest in siblings of index children, it is nonetheless one of the largest studies to date with detailed data on developmental, psychiatric, and neurologic conditions reported in siblings of study group children.

Our results suggest that increased monitoring and screening for ASD and other DDs may be warranted when an older sibling has a diagnosis or symptomatology of ASD or another DD. Future research could explore the contributions of both genetic and non-genetic environmental factors to the clustering of developmental conditions within families (e.g., Matias et al., 2021). Our results also suggest the need to consider the impacts when multiple children are diagnosed with conditions within one family, i.e., the implications of how both family burden and resource needs might increase. Additionally, this study demonstrates that having more than one child affected by a disability is not uncommon and, thus, highlights the importance of health care and other service providers considering the total family needs when treating a child with ASD or other DD.