Introduction

Autism is a common neurodevelopmental condition characterized mainly by social impairments, deficits in both language and nonverbal communication abilities, and repetitive, fixed interests or behaviors (Lord et al., 2018). The prevalence of autism ranges from approximately 0.6–1.0% of the general population in the United States and Europe, with a male-to-female ratio ranging from 3:1 to 4:1 (Solmi et al., 2022; Talantseva et al., 2023). Reports from the World Health Organization indicate a clear increase in the prevalence (39.3%) and Disability-Adjusted Life Year (38.7%) of autism between 1990 and 2019 (Li et al., 2022).

Studies have increasingly reported a coaggregation of autism, schizophrenia, bipolar disorder, major depressive disorder, and substance and alcohol use disorders within families (Daniels et al., 2008; Davis et al., 1992; Wang et al., 2022). Shayestehfar et al. revealed that parents of autistic children were more likely to receive a diagnosis of schizophrenia compared with those of nonautistic children (Shayestehfar et al., 2023). Davis et al. and Daniels et al. discovered an association between parental substance and alcohol use disorders, such as cocaine use disorder, and the likelihood of offspring autism (Daniels et al., 2008; Davis et al., 1992). Our previous study indicated that both paternal and maternal depression were significantly associated with the likelihood of autism in the children (Chen et al., 2020). Furthermore, Sipsock et al. discovered that the severity of autism was associated with cumulative numbers of major psychiatric disorders, including schizophrenia, major affective disorders, and substance and alcohol use disorders, of their first-degree relatives (Sipsock et al., 2021).

The early identification and diagnosis of autism, ideally between 18 and 24 months of age, is crucial for enhancing the overall outcomes in autistic children. Despite the ability of experienced physicians to detect autism in children before the age of 2 years, delayed diagnoses are prevalent, particularly in autistic children who exhibit normal verbal skills (Feng-lei Zhu et al., 2023). A meta-analysis of studies from 40 countries revealed that the age at which autism was diagnosed varied widely, spanning from 38 to 120 months, with an average of approximately 60 months (van’t Hof et al., 2021). Wei et al. reported that the mean age of autism diagnosis was approximately 5.07 ± 3.10 years in Taiwan (Wei et al., 2021). Leng et al. discovered that up to 16% of autistic children were diagnosed after starting school, and also suggested that children from immigrant families commonly received delayed diagnosis (Leng et al., 2023). Durkin et al. highlighted that race and socioeconomic background can influence the age at which autism is diagnosed in the United States (Durkin et al., 2017). These studies suggest that the variance in the age of autism diagnosis reflects the biopsychosocial nature of autism.

Data are scarce on the relationship between the age of autism diagnosis in children and severe mental disorders of their parents. This study aimed to investigate the association between severe parental mental disorders, namely schizophrenia, bipolar disorder, major depressive disorder, and substance and alcohol use disorders, and the age of autism diagnosis using data from the Taiwan National Health Insurance Research Database. Given evidence that the average diagnosis age of autism ranges between 2 and 5 years (Levy et al., 2009; Wei et al., 2021), we hypothesized that parental severe mental disorders may be associated with the delayed diagnosis (≥ 6 years) of autism.

Methods

Data Source

The Taiwan National Health Insurance Research Database (NHIRD), which comprises healthcare data from > 99.7% of Taiwan’s population, was audited and released by the Taiwan National Health Research Institute for scientific research upon a formal application. The insurance claim information of the subjects is anonymous to maintain privacy. Comprehensive information on insured subjects is included in the database, such as demographic data, clinical visit dates, disease diagnoses, and prescriptions, between 2000 and 2011. The diagnostic codes used were based on the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM). The National Health Insurance Research Database has been extensively used in many Taiwanese epidemiologic studies (; Chen et al., 2016; Zhang et al., 2021). The Institutional Review Board of our Hospital reviewed and approved this study protocol.

Inclusion Criteria for Autistic Individuals

The autistic group was defined as autistic people (single offspring) born between 1990 and 2010 who had received an autism diagnosis (ICD-9-CM codes: 299.0, 299.8, and 299.9) from board-certified psychiatrists at least twice. Non-autistic individuals were paired in a ratio of 10:1 with autistic individuals based on their birth year, sex, family income, and level of urbanization after those who had an autism diagnosis (ICD-9-CM code: 299) anytime in the database were excluded. Autistic individuals were categorized into three subgroups based on age at autism diagnosis: < 6 years, 6–11 years, and ≥ 12 years. Parental severe mental disorders were assessed between autistic and non-autistic individuals. Severe mental disorders include schizophrenia, bipolar disorder, major depressive disorder, alcohol use disorder, and substance use disorder. As a proxy for healthcare accessibility in Taiwan, income level (levels 1–3 per month: < 19,100 new Taiwan dollars [NTD], 19,100 ~ 42,000 NTD, and > 42,000 NTD) and urbanization level of residence (levels 1–5, most to least urbanized) were evaluated (Liu et al., 2006).

Statistical Analysis

For between-group comparisons, the independent t-test was used for continuous variables, and Pearson’s χ2 test was used for nominal variables. Logistic regression models with adjustments for birth-year, sex, family income, and level of urbanization were performed to investigate the association between age at autism diagnosis (< 6 years, 6–11 years, and ≥ 12 years) and parental severe mental disorders (schizophrenia, bipolar disorder, major depressive disorder, alcohol use disorder, and substance use disorder). We further assessed the effects of age at autism diagnosis on the likelihoods of paternal and maternal psychiatric disorders, separately. Finally, because current findings support greater growth in language and adaptive skills when children are identified by age 4, we additionally assessed associations between parental severe mental disorders and autism diagnosis age < 4 years, as well as autism diagnosis age 4–5 years. SAS 9.2 (SAS Institute, Cary, NC, USA) was used for all statistical analyses. All tests were two-tailed, and p < 0.05 was considered statistically significant.

Community Involvement

The Taiwan NHIRD comprises healthcare data from > 99.7% of Taiwan’s population, including autistic individuals. The medical records of autistic individuals who sought medical consultation and help were included in the NHIRD. The insurance claim information of autistic and nonautistic individuals is anonymous to maintain privacy.

Results

In all, 22,859 autistic individuals and 228,590 non-autistic individuals were enrolled in the present study, with a mean age of 12.43 ± 4.62 years and a male predominance (84.6%) (Table 1). More than half of autistic individuals (51.2%) were diagnosed in preschool (< 6 years); 34.4% at the age of 6–11 years; and 14.4% at the age of ≥ 12 years (Table 1). Parents of autistic individuals were more likely to be diagnosed with any severe mental disorder, including schizophrenia (1.1 vs. 0.6%, p < 0.001), bipolar disorder (0.9 vs. 0.6%, p < 0.001), major depressive disorder (8.0 vs. 4.9%, p < 0.001), and substance use disorder (1.6 vs. 1.3%, p = 0.002), than did parents of non-autistic individuals (Table 1). The prevalence of alcohol use disorder did not differ (p = 0.722) between parents of autistic and non-autistic individuals (Table 1).

Table 1 Demographic characteristics between autistic and non-autistic individuals
Full size table

There were substantial variations in the odds ratios (ORs) for parental schizophrenia, bipolar disorder, and major depressive disorder among autistic children aged < 6 years, those aged 6–11 years, and those aged ≥ 12 years (Table 2). ORs in parental diagnoses of schizophrenia (2.14, 95% CI 1.62–2.82) and substance use disorder (1.57, 1.24–1.99) were highest in the age at autism diagnosis ≥ 12 years group (Table 2). The highest likelihoods of parental bipolar disorder (OR 1.98, 95% CI 1.61–2.44) and major depression (1.86, 1.72–2.02) were noted in the age at autism diagnosis 6–11 years group (Table 2). The risk of parental alcohol use disorder did not differ between the three autism groups after adjusting for demographic data (Table 2).

Table 2 Risks of parental severe mental disorders between autistic and non-autistic individuals
Full size table

The subanalyses stratified by fathers and mothers showed consistent findings of associations between parental severe mental disorders and age at autism diagnosis (Table 3). Specifically, ORs in paternal schizophrenia (2.15, 95% CI 1.44–3.21), maternal schizophrenia (2.27, 1.58–3.27), and maternal substance use disorder (2.10, 1.48–2.98) were highest in the age at autism diagnosis ≥ 12 years group (Table 3). The likelihoods of maternal bipolar disorder (OR 2.20, 95% CI 1.70–2.84) and major depression (2.06, 1.89–2.27) were highest in the age at autism diagnosis 6–11 years group (Table 3). Finally, Fig. 1 showed the distribution of the age of autism diagnosis based on the presence vs. absence of specific parental severe mental disorder. Finally, sub-analyses stratified by autism diagnosis age < 4 years and 4–5 years showed similar risks of parental severe mental disorders between the two subgroups (Tables 2 and 3).

Table 3 Risks of paternal or maternal severe mental disorders between autistic and non-autistic individuals
Full size table
Fig. 1
figure 1

Parental severe mental disorders on the distribution of autism diagnosis (n = 22,859)

Full size image

Discussion

Our findings partially supported our hypothesis that parental severe mental disorders, namely schizophrenia, bipolar disorder, major depressive disorder, and substance use disorder, were associated with the delayed diagnosis (≥ 6 years) of autism. Furthermore, given that autism was diagnosed in children < 6 years old in most cases, parental schizophrenia and substance use disorder were associated with the most delayed diagnosis of autism (≥ 12 years), whereas parental bipolar disorder and major depression were associated with a less delayed diagnosis of autism (6–11 years).

Diagnosing autism requires experienced practitioners and thorough observation of various signs by primary caregivers (Levy et al., 2009). Amin et al. found more profound autistic symptoms, measured by the parent-reported Autism Quotient-Child questionnaire, in children of parents with severe mental disorders (i.e., schizophrenia) than did those of parents without severe mental disorders (Amin & Salah EL-Deen, 2021), which may logically lead to an early identification of autism. However, we discovered that parental severe mental disorders, especially schizophrenia and substance use disorders, were associated with a later diagnosis of autism in the offspring. We speculate that the compromised mental health of the parents made it more difficult for them to learn about autism and other neurodevelopmental issues and to closely monitor the neurodevelopment of their children. Evidence has shown that compared with healthy parents, parents with severe mental disorders, such as schizophrenia and major depressive disorder, exhibit poorer parenting skills, including active involvement, monitoring, and supervision (Healy et al., 2016; Kahng et al., 2008; Oyserman et al., 2000; Rabha et al., 2021; Riordan et al., 1999). Goodman et al. revealed that parental behaviors toward their children (mean age 2 years) in the schizophrenia group were associated with reduced responsiveness and stimulation, less affectionate involvement, and a poorer child-rearing environment compared with behaviors in the control group (Goodman, 1987). Kahng et al. demonstrated that the nurturance exhibited by parents with severe mental disorders improved as their symptoms became less severe (Kahng et al., 2008). The hypothesis that parents’ overall and cognitive function may affect the timing of autism diagnosis in their children and the evidence that patients with schizophrenia and substance use disorder possess more functional deficits than those with bipolar disorder and major depression (Eslami Shahrbabaki et al., 2022; Ramey & Regier, 2019; Zhu et al., 2019) may explain our findings that delayed diagnosis of autism was mostly associated with parental schizophrenia and substance use disorder, followed by parental bipolar disorder and major depressive disorder. However, further clinical studies would be required to validate the above hypothesis of the association between parental severe mental disorders and delayed autism diagnosis.

A group of autistic children being raised by parents with severe mental illness are also at risk for co-occurring mental health challenges. Evidence has shown the common mental comorbidities, such as attention deficit hyperactivity disorder (ADHD), anxiety disorders, and mood disorders, with autism, which may also be related to the delayed diagnosis of autism (Gunnarsdottir et al., 2018; Karunakaran et al., 2020; Wei et al., 2021). Our previous study assessed the sequence of autism and ADHD diagnoses and found that ADHD comorbidity may delay autism identification in autistic children with ADHD (Wei et al., 2021). Emotional and behavioral disturbances stemming from the comorbidities may mask or overlap the symptoms of autism, resulting in a delayed diagnosis (Gunnarsdottir et al., 2018; Karunakaran et al., 2020).

This study has several limitations. First, in Taiwan, the NHIRD provides highly affordable and inexpensive medical services to the public. People can directly access the specialists, including psychiatrists, without the need for a referring note. However, the prevalence of autism and parental severe mental disorders may be still underestimated because only those who sought medical and mental health consultations would be included in the database. In addition, the present study used the ICD-9-CM codes for the identification of autism and severe mental disorders. The lack of a semi-structured interview to evaluate psychopathology likely led to the omission of some mental disorders, including autism. However, the severe mental disorders in parents and autism in offspring were diagnosed by board-certified psychiatrists, which enhances diagnostic validity. Second, our study was a cross-sectional study, which limited our ability to clarify the temporality between parental severe mental disorder diagnoses and offspring autism diagnoses. One possibility was that the mental diagnoses given to the parents after the birth of their autistic children may suggest that the parents’ mental disorders, such as depression, are a consequence of raising an autistic child (Picardi et al., 2018). Chien et al. reported that parental severe mental disorders both before children’s birth and after children’s birth were associated with offspring autism risk (Chien et al., 2022).

To clarify the aforementioned temporality, further clinical longitudinal studies would be required. Fourth, the database failed to provide some information, including information on disease severity, environmental factors, and cultural factors. Therefore, we could not examine the effects of these factors in the present study. Finally, further investigation would be needed to verify whether our findings can extend to people in Western countries because social attitudes toward diagnosing autism differs between Taiwan and Western countries (Babik & Gardner, 2021).

In conclusion, our study established that delayed diagnosis of autism was mostly associated with parental schizophrenia and substance use disorder, followed by parental major affective disorders. These findings inform clinicians and public health officers about the importance of early monitoring of neurodevelopmental conditions among the offspring of parents with severe mental disorders. Compared with cases of delayed diagnoses, the timely detection of such conditions enables prompt treatment with tailored interventions as well as the full utilization of public support systems, leading to superior overall outcomes for children.