Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Apr 1.
Published in final edited form as: Autism. 2023 Aug 30;28(4):975–984. doi: 10.1177/13623613231188876

Maternal obesity and diabetes during pregnancy and early autism screening score at well-child visits in standard clinical practice

Sarah A Carter 1, Jane C Lin 1, Ting Chow 1, Mayra P Martinez 1, Jasmin M Alves 2,3, Klara R Feldman 4, Chunyuan Qiu 4, Kathleen A Page 2, Rob McConnell 5, Anny H Xiang 1
PMCID: PMC10902177  NIHMSID: NIHMS1914736  PMID: 37646431

Abstract

Early intervention can reduce disability in children diagnosed with autism spectrum disorder (ASD). Screening for ASD in young children identifies those at increased likelihood of diagnosis who may need further support. This study assessed in utero exposure to maternal obesity and diabetes and offspring performance on the Quantitative Checklist for Autism in Toddlers (QCHAT-10), an ASD screening questionnaire administered between 18–24 months at well-child visits. This retrospective cohort study included 65,433 singletons born in a single healthcare system. Demographic data, maternal obesity, Type 1/Type 2 (T1D/T2D) and gestational diabetes (GDM) information, and QCHAT-10 score in children 12–30 months old were extracted from electronic medical records. Negative binomial models were used to estimate incidence ratio ratios (IRR) of associations between maternal obesity and diabetes exposure and continuous offspring QCHAT-10 scores. Maternal obesity, T1D/T2D (IRR:1.13,1.06–1.21) and GDM ≤26 weeks (IRR:1.10,1.05–1.17) were associated with one-unit increases in QCHAT-10 scores. Relationships with obesity and GDM≤26 weeks remained after mutual adjustment and excluding children with ASD diagnoses. No associations were reported for GDM>26 weeks. Maternal obesity and diabetes were associated with higher QCHAT-10 scores in children 12–30 months old, suggesting these exposures in pregnancy may be associated with a range of early childhood behavior.

Keywords: autism spectrum disorders, early screening tools, maternal obesity, maternal diabetes

Lay summary

Early intervention and treatment can help reduce disability in children diagnosed with autism spectrum disorder (ASD). Screening for ASD in young children identifies those at increased likelihood of diagnosis who may need further support. Previous research has reported that exposure to maternal obesity and diabetes during pregnancy is associated with higher likelihood of ASD diagnosis in children. However, little is known about whether these maternal conditions are associated with how very young children score on ASD screening tools. This study examined associations between exposure to maternal obesity and diabetes during pregnancy and offspring scores on the Quantitative Checklist for Autism in Toddlers (QCHAT-10), an ASD screening questionnaire administered between 18–24 months at well-child visits. A higher score on the QCHAT-10 suggests a higher likelihood of ASD; children with scores 3 or greater are referred to developmental pediatricians for evaluation. Our study found that children of mothers with obesity or diabetes during pregnancy had higher scores than children whose mothers did not have these conditions. Associations with maternal obesity and gestational diabetes diagnosed at or before 26 weeks of pregnancy were also present in children who did not have later ASD diagnoses, suggesting that exposure to these conditions during early pregnancy may be associated with a broad range of social and behavioral abilities. Identifying associations between maternal health conditions and early QCHAT-10 screening scores could influence future screening and provision of support for children of mothers with these conditions.

Introduction

Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental disorder characterized by impaired communication and social behaviors and repetitive movements(Maenner et al., 2013). While ASD etiology is multifactorial(Geschwind, 2011; Hallmayer et al., 2011; Wang, Geng, Liu, & Zhang, 2017), in utero exposure to maternal inflammatory conditions, such as obesity and diabetes, has been associated with ASD diagnosis in children(Li et al., 2016; Xiang et al., 2015). Maternal obesity and diabetes may contribute to increased inflammation and disruptions in cytokine and chemokine expression(Smith, Li, Garbett, Mirnics, & Patterson, 2007), leading to divergent fetal brain development which presents in young children. These impairments are typically noticeable in early childhood(Lord, Elsabbagh, Baird, & Veenstra-Vanderweele, 2018), with severity of symptoms varying between individuals(Masi, DeMayo, Glozier, & Guastella, 2017). Early intervention and treatment can reduce disability for children diagnosed with autism spectrum disorder (ASD)(Kodak & Bergmann, 2020; Landa, Gross, Stuart, & Faherty, 2013; Oono, Honey, & McConachie, 2013; Sacrey, Bennett, & Zwaigenbaum, 2015). The rate of ASD diagnosis is currently rising(Prevention, 2022), meaning an increasing number of children need a range of support and intervention. Therefore, it is important to screen and diagnose early.

Several early screening tools have helped identify young children with increased likelihood of ASD diagnosis. In 1999, the Modified Checklist for Autism in Toddlers (MCHAT), a 23-point questionnaire with telephone follow-up, was developed to screen for potential autism traits in toddlers(Robins, Fein, Barton, & Green, 2001). By 2008, the Quantitative Checklist for Autism in Toddlers (QCHAT-25), containing 25 parent-report questions on a Likert scale assessing attention, social interaction, language development, repetitive behaviors, and social communication(Allison et al., 2008), was being used to identify young children who may be on the developmental pathway toward ASD(Allison, Auyeung, & Baron-Cohen, 2012). However, the MCHAT and QCHAT-25 have not been used to screen for potential ASD at the population level in everyday clinical practice; this may be due to the amount of time it takes to complete these screening questionnaires(Allison et al., 2008).

To make screening easier to conduct at population level, a shorter version with 10 questionnaires (QCHAT-10) was developed in 2012 and is more widely administered to children between 18–24 months at standard well-child visits with pediatricians(Allison et al., 2012). A QCHAT-10 score of ≥3 is considered abnormal; children with a score of ≥3 should be referred to a developmental pediatrician for further assessment for potential ASD diagnosis(Allison et al., 2012). The QCHAT-10 has been validated across several populations, including the United Kingdom, Italy, and Singapore(Magiati et al., 2015; Raza et al., 2019; Roman-Urrestarazu et al., 2021; Ruta et al., 2019), with reported sensitivity of 0.91 and specificity of 0.89 in a cohort with known ASD diagnoses(Allison et al., 2012). As a continuous variable, the QCHAT-10 score measures the distribution of ASD-related traits in the general population(Allison et al., 2008). While children with high QCHAT-10 scores are more likely to be diagnosed with ASD, those with scores in the upper distribution that do not meet abnormal criteria or receive an ASD diagnosis may also present with social or behavioral deficits which warrant recognition, support, and intervention(Tavassoli et al., 2018).

To our knowledge, the relationship between potential risk factors associated with ASD diagnosis and early QCHAT-10 screening of toddlers in the general population has not been studied. This study examined associations between maternal obesity and diabetes during pregnancy and offspring’s QCHAT-10 score administered at standard 18–24 month clinical well-child appointments, a particularly important question as rates of both maternal obesity and diabetes and ASD diagnosis are rising(Driscoll & Gregory, 2020; Gregory & Ely, 2022), and previous studies in the same study population have shown that maternal obesity and diabetes were associated with ASD diagnosis(Xiang, 2017; Xiang et al., 2015). Results will help identify children with specific prenatal exposure risk factors, aiding in the promotion of appropriate screening in children of mothers with obesity or diabetes during pregnancy, and helping secure further support for children with differing social and behavioral capabilities.

Materials and Methods

Study Population

This retrospective, population-based birth cohort study included 65,433 mother-child pairs of singletons born in Kaiser Permanente Southern California (KPSC) hospitals between January 1, 2014 and December 31, 2017 who had QCHAT-10 screening during a standard well-child visit between 12 and 30 months old (eFigure 1). Children were followed through KPSC electronic medical records (EMR) from birth until age 5 years. Those who had a diagnosis of ASD before QCHAT-10 administration were excluded. The KPSC healthcare system serves a diverse population of 4.5 million members throughout Southern California, with member social and demographic data reflecting regional census tracts(Koebnick et al., 2012). Maternal social and demographic information and pregnancy health data as well as child’s QCHAT-10 score and ASD diagnosis were extracted from KPSC’s comprehensive electronic medical records (EMR) system. This study was approved by KPSC Institutional Review Boards, with individual subject consent waived. There was no community involvement in the study.

eFigure 1:

eFigure 1:

Open in a new tab

Derivation of the sample

Outcomes: QCHAT-10 score

The primary outcome of interest was QCHAT-10 score administered during standard well-child visits between 12 and 30 months. QCHAT-10 scores ranged from 0 (lowest) to 10 (highest) and reflected the number of social and behavioral questionnaire answers indicative of autism(Allison et al., 2012). The QCHAT-10 score cut point for potential later autism diagnosis is ≥3; children with scores 3 or greater are referred for further evaluation(Allison et al., 2012). QCHAT-10 scores were treated both as a continuous variable (with scores from 0 to 10) and a categorical variable as abnormal (scores ≥3) or normal.

To assess QCHAT-10 as a screening tool for ASD, we extracted ASD diagnoses after QCHAT-10 for these children. ASD diagnoses were identified in EMR by ICD-9 codes 299.0, 299.1, 299.8, 299.9, or ICD-10 codes F84.0, F84.3, F84.5, F84.8, F84.9 recorded by clinical practitioners at ≥ two separate clinical visits, as reported in previous work using this cohort(Coleman et al., 2015; Jo, Eckel, Wang, et al., 2019; Xiang, Wang, Martinez, Getahun, et al., 2018; Xiang et al., 2015).

Exposures during Pregnancy

The primary exposures of interest were maternal obesity and diabetes during pregnancy. Maternal body mass index (BMI) was calculated using height and weight measurements on the date closest to the last menstrual period (LMP) from a window of 6 months before LMP until 3 months after recorded in EMR(Xiang et al., 2019). There were 189 mothers with unknown BMI; as BMI was an exposure of interest in this study, their information was imputed using the median from the study cohort and included in data analyses. Maternal pre-pregnancy BMI was categorized into five groups: 1) No overweight or obesity (BMI <18.5–24.9 kg/m2); 2) Overweight (25.0–29.9 kg/m2); 3) Obese Class 1 (30.0–34.9 kg/m2); 4) Obese Class 2 (35.0–39.9 kg/m2); and 5) Obese Class 3 (≥40.0 kg/m2). Diabetes exposure during pregnancy was determined by ICD-10 codes and included pre-existing diabetes [Type 1 (T1D) or Type 2 (T2D)] and gestational diabetes mellitus (GDM), which have been associated with likelihood of ASD diagnosis in KPSC cohorts(Xiang, Wang, Martinez, Page, et al., 2018; Xiang et al., 2015). GDM was categorized by diagnoses at 1) ≤26 and 2) >26 gestational weeks, as previous work using this cohort has reported differences in ASD likelihood by timing of GDM diagnosis(Xiang et al., 2015).

Covariates

Covariates included to control for potential confounding were child’s sex and maternal age, parity, self-reported race/ethnicity and education, census tract-level household income at child’s first birthday, and maternal history of comorbidity (≥1 diagnosis of heart, lung, kidney, liver disease, or cancer); all have been previously associated with likelihood of ASD diagnosis in KPSC study populations(Jo, Eckel, Chen, et al., 2019; Xiang et al., 2015). Birth year was included to account for increasing ASD diagnosis trends over the study period(Jo, Eckel, Wang, et al., 2019).

Statistical analyses

Maternal and child characteristics by normal and abnormal QCHAT-10 scores were presented using descriptive statistics, with median and interquartile range (IQR) for continuous variables and total number (N) and proportion (%) for categorical variables. Statistically significant differences in characteristics between children with QCHAT-10 scores ≥3 and children with scores <3 were compared using Wilcoxon rank sum test for continuous variables and Pearson’s Chi-squared test for categorical variables.

For continuous scores of QCHAT-10 (scores from 0 to 10), because the variance was larger than the mean, negative binomial regression models were used to assess associations between exposure to maternal obesity and diabetes and one-unit increases in child’s continuous QCHAT-10 score(Green, 2021). Results were presented as incidence rate ratios (IRRs) and 95% confidence intervals (95% CI) of offspring QCHAT-10 score associated with each maternal exposure relative to the reference group. Analysis began by assessing maternal obesity or diabetes exposure separately in a single exposure model followed by including both simultaneously in the same model to assess independent associations. For the categorical abnormal vs normal QCHAT-10 scores, logistic regression was used with similar analytical steps and results were presented as odds ratios (ORs) and 95% CI. Data analyses were repeated after excluding children who were later diagnosed with ASD to reveal relationships between maternal metabolic conditions and QCHAT-10 scores in typically developing children. Finally, potential interaction of maternal obesity and diabetes exposure on offspring continuous QCHAT-10 score was assessed using interaction terms in negative binomial regression models. Birth year, maternal age, self-reported race/ethnicity, education, parity, census tract household income (per $10k) at child’s first birthday, history of comorbidity, and child’s sex were included as covariates in all models.

Statistical significance was set at p<0.05. All statistical analyses were performed in STATA (version 13.0).

Results

Among the 65,433 children included in this study, 18,830 (28.7%) children had mothers with maternal pre-pregnancy BMI ≥30kg/m2 and 7,757 (11.9%) had mothers with diabetes during pregnancy. Overall, 3806 (5.8%) children had QCHAT-10 scores of ≥3. Among children with abnormal QCHAT-10 scores, 788 (20.7%) were diagnosed with ASD by age 5. In contrast, among children with normal QCHAT-10 scores, 568 (0.92%) had a diagnosis of ASD by age 5. Table 1 displays maternal and child characteristics by abnormal/normal QCHAT-10 score. More male children (67.2%) were in the abnormal group than in the normal QCHAT-10 group (49.9%). While median maternal age at delivery was approximately 31 years in both groups, greater proportions of mothers reporting Hispanic race/ethnicity, Asian/Pacific Islander, Black, and Other Ethnicity, respectively, had children with QCHAT-10 scores ≥3. Similar trends in parity were present in both groups, with most women having at least one other child. A larger proportion of high school or unknown educational qualifications were present in mothers of children with abnormal QCHAT-10 scores, and median household income was lower in this group. Children with abnormal QCHAT-10 scores had a greater proportion of mothers with histories of comorbidities and had higher percentages of mothers with Class 1 (16.8% vs 16.2%), Class 2 (10.5% vs 7.7%), and Class 3 (7.6% vs 4.6%) obesity than children with normal QCHAT-10 scores. This trend was also present for diabetes, with mothers of children in the abnormal group having more preexisting diabetes (6.3% vs 5.7%), GDM diagnosed at ≤26 gestational weeks (4.9% vs 4.0%), and GDM diagnosed at >26 weeks (3.2% vs 2.2%) than those of children with normal QCHAT-10 scores. No significant interaction of maternal obesity and diabetes exposure on offspring continuous QCHAT-10 score was reported.

Table 1:

Cohort characteristics by child’s QCHAT-10 score*

Normal (QCHAT-10 <3) N = 61,627 Abnormal (QCHAT-10 ≥3) N = 3,806
Child characteristics
Male, N (%) 30,767 (49.9%) 2,556 (67.2%)
     
Maternal characteristics
Age, Median (IQR) 31.2 (7.7) 31.3 8.5)
Race/Ethnicity, N (%)
 White 13,951 (22.6%) 605 (15.9%)
 Black 4,084 (6.6%) 316 (8.3%)
 Hispanic 33,368 (54.2%) 2,193 (57.6%)
 Other 1,849 (3.0%) 119 (3.1%)
 API 8,375 (13.6%) 573 (15.1%)
Parity, N (%)
 0 17,221 (27.9%) 1,130 (29.7%)
 1 21,286 (34.5%) 1,180 (31.0%)
 2 13,974 (22.7%) 875 (23.0%)
 Unknown 9,146 (14.8%) 621 (16.3%)
Education, N (%)
 High School/Unknown1 17,116 (27.7%) 1,297 (34.1%)
 Some College 19,610 (31.8%) 1,331 (35.0%)
 College & Post Graduate 24,901 (40.4%) 1,178 (31.0%)
Household income, US dollars, Median (IQR) 66,696 (21,145) 63,246 (36,450)
History of comorbidity2, N (%) 12,577 (20.4%) 876 (23.0%)
     
Maternal metabolic conditions in pregnancy
Maternal obesity
No overweight or obesity (BMI <18.5–24.9 kg/m2) 25,836 (41.9%) 1,400 (36.8%)
Overweight (BMI 25.0–29.9 kg/m2) 18,288 (29.7%) 1,079 (28.4%)
Obese Class 1 (BMI 30.0–34.9 kg/m2) 9,970 (16.2%) 641 (16.8%)
Obese Class 2 (BMI 35.0–39.9 kg/m2) 4,722 (7.7%) 398 (10.5%)
Obese Class 3 (BMI >40.0 kg/m2) 2,811 (4.6%) 288 (7.6%)
     
Maternal diabetes
Preexisting diabetes, N (%) 1,352 (2.2%) 121 (3.2%)
GDM ≤ 26 weeks, N (%) 2,456 (4.0%) 188 (4.9%)
GDM > 26 weeks, N (%) 3,487 (5.7%) 240 (6.3%)
Open in a new tab
*

Tests for difference in each characteristic between the two QCHAT-10 groups were statistically significant at p<0.001, except for maternal age (p=0.494)

1

High school and Unknown educational qualifications were combined due to small Unknown sample size

2

Maternal comorbidity was defined as ≥=1 diagnosis of heart, lung, kidney, liver disease, or cancer.

Table 2 presents associations between maternal obesity and diabetes and QCHAT-10 scores as a continuous outcome (scores from 0 to 10). When analyzed as a single exposure, significant associations were observed for maternal obesity where children exposed to maternal Obese Class 1 (IRR: 1.07, 95% CI 1.04–1.10), Obese Class 2 (1.16, 1.11–1.21), and Obese Class 3 (1.25, 1.19–1.32) had significantly greater risk of higher QCHAT-10 scores compared to children of mothers without overweight or obesity during pregnancy. Risk increased with severity of maternal obesity (from obese Class 1 to Class 2 to Class 3). No significant associations were found for maternal overweight exposure (1.02, 0.99–1.05). For diabetes during pregnancy, significant associations were observed for preexisting diabetes (1.13, 1.06–1.21) and GDM diagnosed at ≤26 gestational weeks (1.10, 1.05–1.17). GDM diagnosed at >26 gestational weeks was not associated with QCHAT-10 scores (1.04, 0.99–1.09) (Table 2). Including both maternal obesity and diabetes in the same model did not affect results for maternal obesity but reduced associations with maternal diabetes. However, both maternal obesity (Class 1, Class 2, Class 3) and preexisting diabetes and GDM diagnosed at ≤26 gestational week remained significantly associated with continuous QCHAT-10 scores (Table 2).

Table 2:

Incidence rate ratios (95% CI) for one-unit increases in QCHAT-10 score associated with maternal obesity and diabetes during pregnancy1

Relative increase in QCHAT-10 score (IRR 95% CI)
Single exposure *
Obesity 2
Underweight (BMI >18.5kg/m2) 1.06 (0.98–1.14)
Overweight (BMI 25–29.9kg/m2) 1.03 (0.99–1.05)
Obese Class 1 (BMI 30–34.9kg/m2) 1.07 (1.04–1.11)
Obese Class 2 (BMI 35.0–39.9 kg/m2) 1.16 (1.12–1.21)
Obese Class 3 (BMI >40.0 kg/m2) 1.26 (1.19–1.32)
   
Diabetes 3
Preexisting diabetes 1.13 (1.06–1.21)
GDM ≤ 26 weeks 1.10 (1.05–1.17)
GDM > 26 weeks 1.04 (0.99–1.09)
   
Multiple exposure, mutually adjusted +
Obesity
Underweight (BMI >18.5kg/m2) 1.06 (0.98–1.14)
Overweight (BMI 25–29.9kg/m2) 1.02 (0.99–1.05)
Obese Class 1 (BMI 30–34.9kg/m2) 1.07 (1.03–1.10)
Obese Class 2 (BMI 35.0–39.9 kg/m2) 1.15 (1.11–1.20)
Obese Class 3 (BMI >40.0 kg/m2) 1.24 (1.18–1.30)
   
Diabetes
Preexisting diabetes 1.08 (1.00–1.16)
GDM diagnosed at ≤ 26 gestational weeks 1.07 (1.01–1.13)
GDM diagnosed at > 26 gestational weeks 1.02 (0.97–1.07)
Open in a new tab
1

From negative binomial regression adjusted for birth year, maternal age, parity, maternal race/ethnicity, maternal educational qualifications, history of comorbidity, income (per 10k), and child’s sex

2

Reference category is Normal BMI (18.5–24.9 kg/m2)

3

Reference category is No Diabetes

*

Not adjusted for the other maternal pregnancy exposure

+

Simultaneously adjusted for the other maternal pregnancy exposure

Table 3 reports results of logistic regression models when QCHAT-10 was analyzed as a binary categorical outcome (abnormal vs normal). Similar to results for continuous QCHAT-10 scores (Table 2), in utero exposure to maternal Class 1, Class 2 and Class 3 obesity was associated with greater odds of having abnormal QCHAT-10 scores. In single exposure models, the ORs (95% CI) were (1.13, 1.03–1.25) for Obese Class 1, (1.46, 1.30–1.65) for Obese Class 2, and (1.75, 1.53–2.01) for Obese Class 3. Maternal overweight (1.06, 0.97–1.15) was not associated with abnormal QCHAT-10 scores. Maternal preexisting diabetes (1.35, 1.11–1.64) and GDM at ≤26 weeks (1.19, 1.02–1.39) were significantly associated. Mutually adjusting for maternal obesity and diabetes did not affect the association with maternal obesity but reduced the association with maternal diabetes.

Table 3:

Odds ratios (95% CI) for associations between maternal obesity and diabetes during pregnancy and QCHAT-10 scores ≥31

OR (95% CI)
Single exposure *
Obesity 2
Underweight (BMI >18.5kg/m2) 0.99 (0.78–1.27)
Overweight (BMI 25–29.9kg/m2) 1.06 (0.97–1.15)
Obese Class 1 (BMI 30–34.9kg/m2) 1.13 (1.03–1.25)
Obesity Class 2 (BMI 35.0–39.9 kg/m2) 1.46 (1.30–1.65)
Obesity Class 3 (BMI >40.0 kg/m2) 1.75 (1.53–2.01)
   
Diabetes 3
Preexisting diabetes 1.35 (1.11–1.64)
GDM ≤ 26 weeks 1.19 (1.02–1.39)
GDM > 26 weeks 1.09 (0.95–1.24)
   
Multiple exposure, mutually adjusted +
Obesity
Underweight (BMI >18.5kg/m2) 0.99 (0.78–1.27)
Overweight (BMI 25–29.9kg/m2) 1.05 (0.97–1.14)
Obese Class 1 (BMI 30–34.9kg/m2) 1.12 (1.02–1.24)
Obesity Class 2 (BMI 35.0–39.9 kg/m2) 1.44 (1.28–1.63)
Obesity Class 3 (BMI >40.0 kg/m2) 1.71 (1.49–1.97)
   
Diabetes
Preexisting diabetes 1.18 (0.97–1.43)
GDM ≤ 26 weeks 1.08 (0.93–1.27)
GDM > 26 weeks 1.03 (0.90–1.19)
Open in a new tab
1

Results were from logistic regression adjusted for birth year, maternal age, parity, maternal race/ethnicity, maternal educational qualifications, history of comorbidity, income (per 10k), and child’s sex

2

Reference category is Normal BMI (18.5–24.9 kg/m2)

3

Reference category is No Diabetes

*

Not adjusted for the other maternal pregnancy exposure

+

Simultaneously adjusted for the other maternal pregnancy exposure

Repeating data analysis excluding 1,356 children who later had ASD diagnosis by age 5 showed that point estimates were slightly smaller for all exposures when compared to the full cohort including all children (Table 4 vs Table 2). However, direction and significance largely remained for nearly all risk factors identified from the full cohort analyses. IRRs for increase in QCHAT-10 scores associated with Obese Class 1 were (1.05, 95% CI 1.02–1.08), Obese Class 2 (1.13, 1.08–1.18), and Obese Class 3 (1.20, 1.14–1.26) after adjustment for maternal diabetes. Preexisting diabetes (1.14, 1.06–1.22) and GDM at ≤26 weeks (1.09, 1.04–1.15) were associated with increases in QCHAT-10 scores; these associations remained significant after adjusting for maternal obesity (Table 4).

Table 4:

Incidence rate ratios (95% CI) for one-unit increases in QCHAT-10 score associated with maternal obesity and diabetes during pregnancy, excluding children with later ASD diagnoses1,2

Continuous QCHAT-10 score
IRR (95% CI)
Single exposure *
Obesity 3
Underweight (BMI >18.5kg/m2) 1.06 (0.98–1.14)
Overweight (BMI 25–29.9kg/m2) 1.01 (0.99–1.04)
Obese Class 1 (BMI 30–34.9kg/m2) 1.05 (1.02–1.09)
Obese Class 2 (BMI 35.0–39.9 kg/m2) 1.13 (1.09–1.18)
Obese Class 3 (BMI >40.0 kg/m2) 1.20 (1.14–1.26)
   
Diabetes 4
Preexisting diabetes 1.14 (1.06–1.22)
GDM ≤ 26 weeks 1.09 (1.04–1.15)
GDM > 26 weeks 1.04 (0.99–1.09)
   
Multiple exposure, mutually adjusted +
Obesity
Underweight (BMI >18.5kg/m2) 1.06 (0.98–1.14)
Overweight (BMI 25–29.9kg/m2) 1.01 (0.99–1.04)
Obese Class 1 (BMI 30–34.9kg/m2) 1.05 (1.02–1.08)
Obesity Class 2 (BMI 35.0–39.9 kg/m2) 1.12 (1.08–1.17)
Obesity Class 3 (BMI >40.0 kg/m2) 1.19 (1.13–1.25)
   
Diabetes
Preexisting diabetes 1.09 (1.02–1.17)
GDM ≤ 26 weeks 1.06 (1.01–1.12)
GDM > 26 weeks 1.02 (0.98–1.07)
Open in a new tab
1

Adjusted for birth year, maternal age, parity, maternal race/ethnicity, maternal educational qualifications, history of comorbidity, income (per 10k), and child’s sex

2

1,356 children with later ASD diagnoses were excluded

3

Reference category is Normal BMI (18.5–24.9 kg/m2)

4

Reference category is No Diabetes

*

Not adjusted for the other maternal pregnancy exposure

+

Simultaneously adjusted for the other maternal pregnancy exposure

Discussion

In this multi-ethnic birth cohort study with QCHAT-10 data derived from standard well-child visits in clinical practice, maternal obesity and diabetes were associated with higher scores on QCHAT-10 screening in toddlers. This was observed for both QCHAT-10 as a continuous measure of scores and categorized as abnormal vs normal scores. Children exposed to maternal obesity and diabetes in utero had increased risk of higher QCHAT-10 scores, relative to children who were not exposed to either condition. The greatest associations were in children of mothers with pre-pregnancy BMIs >40 kg/m2 (Class 3 Obesity), with IRRs significant in children of mothers with Class 2 Obesity, and Class 1 Obesity, respectively. Higher QCHAT-10 scores were also found in children of mothers with preexisting diabetes and GDM diagnoses ≤26 weeks pregnant. Increased risk of abnormal QCHAT-10 score was associated with maternal obesity and diabetes separately; while mutual adjustment attenuated risk associated with diabetes, it did not change the significance of risk for maternal obesity, indicating that results for maternal obesity were independent of maternal diabetes, but that obesity may be a mediating factor for diabetes. These associations remained after excluding those who went on to have a diagnosis of ASD by age 5, suggesting maternal obesity and diabetes may contribute to a broader spectrum of atypical neurodevelopment in children. Our results concerning the association between maternal obesity and diabetes during pregnancy and elevated QCHAT-10 screening scores in toddlers during standard well-child visits are consistent with previous research showing maternal obesity and diabetes impacted the likelihood of ASD in children(Chen, Zhao, Dalman, Karlsson, & Gardner, 2021; Edlow, 2017; Sanchez et al., 2018; Xiang, 2017; Xiang et al., 2015).

Several validation studies have supported use of QCHAT-10 to screen for potential later ASD diagnosis(Magiati et al., 2015; Raza et al., 2019; Roman-Urrestarazu et al., 2021; Ruta et al., 2019), with QCHAT-10 scores of ≥3 indicating the need for further ASD screening(Allison et al., 2012). In our cohort, 788 (20.7%) of children with QCHAT-10 scores ≥3 were diagnosed with ASD by age 5. Among children with normal QCHAT-10 scores, 568 (0.92%) had a diagnosis of ASD by age 5. The QCHAT-10 screening tool, and its relationship to early childhood neurodivergence, is essential for providing early intervention to children in need, especially considering developmental trajectories in children with ASD may be altered as early as six months old(Landa et al., 2013). Associations between maternal metabolic conditions and early QCHAT-10 screening scores could influence the screening and provision of support for exposed children with a range of neurodevelopmental presentations by identifying children at increased risk due to maternal exposures in utero. Recent research has reported that in California, lower socioeconomic status (SES) is associated with higher rates of ASD diagnosis(Winter, Fountain, Cheslack-Postava, & Bearman, 2020), a reversal of previous trends in which higher SES increased likelihood of ASD diagnosis(M. D. King & Bearman, 2011). This reversal could be due to increased universal screening at standard care visits and care access for all children with potential deficits. In our study, mothers with fewer educational qualifications reported more traits associated with ASD on QCHAT-10 screenings. Potential relationships between maternal SES and early ASD screening scores are an avenue for future research.

To our knowledge, this is the first study to report that maternal obesity and diabetes in utero are associated with higher QCHAT-10 screening scores in toddlers at standard clinical well-child visits at the populational level. Previous studies focusing on the Social Responsive Scale (SRS), a quantitative screening tool for autistic traits, have reported associations between indicators of maternal obesity and increased screening tool scores in children. Increased levels of leptin in offspring cord blood were associated with higher SRS at 8–9 years old in 762 children in a research cohort(Iwabuchi et al., 2021). Higher ω−6 polyunsaturated fatty acid status in mothers at approximately 20 weeks pregnant was associated with higher SRS score at age 6 years among 4,624 children(Steenweg-de Graaff et al., 2016). Increases in SRS scores have also been reported after exposure to maternal diabetes in 6,778 children across 40 cohorts; however, maternal pre-pregnancy obesity was not adjusted (Lyall et al., 2022). Maternal obesity in pregnancy is a risk factor for developing gestational diabetes(Chu et al., 2007); thus, considering the influence of concurrent maternal obesity and diabetes on health outcomes is critical. While others have reported interactions of maternal obesity and diabetes on offspring health outcomes(Ijas et al., 2019; Kong, Nilsson, Gissler, & Lavebratt, 2019), significant interactions of maternal obesity and diabetes on offspring QCHAT-10 scores were not found in our study. This may be due to our categorization of maternal obesity and diabetes by severity; to our knowledge, ours is the first study to consider associations of early ASD screening scores with maternal obesity by class. Taken together, our study and others suggest that exposure to maternal obesity and diabetes in utero may induce changes in offspring neurodevelopment which can be captured by early screening in toddlers in standard clinical well-child visits. Further study on the interaction of maternal obesity and diabetes on early ASD screening scores is warranted.

Brain development, and its influence on behavior, is multifactorial; ASD itself is a spectrum of heterogeneous behaviors that vary between individuals. Neurodevelopmental disorders are distinct from other neurological or psychiatric conditions in that they are typically early-onset, lifelong conditions that occur more frequently in boys and tend to present with other neurological and psychiatric comorbidities(Thapar, Cooper, & Rutter, 2017). There is also considerable overlap in the symptoms of several neurodevelopmental disorders such as ASD, ADHD, and learning and communication disorders (B. H. King, 2016; Thapar et al., 2017), and many neurodevelopmental conditions are associated with the same environmental exposures, like maternal inflammatory conditions and air pollutant exposure(Carter et al., 2022; Oudin et al., 2019; Xiang, Wang, Martinez, Getahun, et al., 2018; Xiang et al., 2015). This study reported that maternal obesity and diabetes exposure remained significantly associated with abnormal QCHAT-10 scores after the exclusion of children was ASD. Our finding supports previous research indicating that maternal metabolic conditions may impact not only likelihood of ASD diagnosis but may also be associated with divergent brain development more generally(Chen et al., 2021; Edlow, 2017; Guo et al., 2020; Sanchez et al., 2018). This information is especially important given that rates of maternal obesity and diabetes are rising(Driscoll & Gregory, 2020; Gregory & Ely, 2022) and neurodevelopmental disorders may now impact up to 20% of the general population(Patel & Merrick, 2020).

Strengths and Limitations

An important strength of this study is our use of a large, representative birth cohort including social demographics, maternal obesity and diabetes diagnosis during pregnancy, and continuous follow-up of children from birth, allowing access to QCHAT-10 scores and later clinical diagnosis of ASD through EMR. KPSC continuity of care reduces ascertainment bias for both exposures and outcomes. Use of KPSC EMR allowed us to demonstrate novel associations between maternal obesity and diabetes exposure during pregnancy and QCHAT-10 scores in toddlers and permitted us to investigate this association including and excluding children with later ASD diagnoses. This study bridges the gap between maternal exposure work done on ASD and questions about the utility of QCHAT-10 screening in identifying behaviors across the spectrum of child development.

As this is an observational study, it has limitations. The findings presented here do not establish a causal link between maternal obesity and diabetes exposures and toddler QCHAT-10 scores. We did not have access to paternal data and parental genetic information and were therefore unable to control for paternal factors and genetic contributions to early childhood behavior and QCHAT-10 screening score. QCHAT-10 questionnaires rely on parental report of childhood behavior; there may be further parental influences on scoring unavailable in the social and demographic data present in EMR. Other potential residual confounders, such as environmental exposures and postnatal risk factors, were not adjusted for in analyses.

Conclusions

Maternal obesity and diabetes exposure during pregnancy were both associated with higher QCHAT-10 scores in early childhood. These associations were strongest in children of mothers with a greater degree of obesity, and in children of mothers with preexisting diabetes and GDM diagnoses at or before 26 weeks pregnant. Associations remained significant after the exclusion of children later diagnosed with ASD, indicating that maternal metabolic exposures in pregnancy are associated with a range of divergent early childhood behavior captured by QCHAT-10 screening. Effective treatment and screening of maternal obesity and diabetes during pregnancy are needed to address the potential impact of these conditions across generations. Future research should assess associations between other known risk factors for ASD, such as maternal hypertensive disorders and environmental exposures during pregnancy, and QCHAT-10 screening scores in childhood.

Acknowledgements:

The authors thank KPSC patients for helping us improve care using information collected via our electronic health record systems.

Footnotes

Conflict of interests: We declare no actual or potential competing interests.

References

  1. Allison C, Auyeung B, & Baron-Cohen S (2012). Toward brief “Red Flags” for autism screening: The Short Autism Spectrum Quotient and the Short Quantitative Checklist for Autism in toddlers in 1,000 cases and 3,000 controls [corrected]. J Am Acad Child Adolesc Psychiatry, 51(2), 202–212 e207. doi: 10.1016/j.jaac.2011.11.003 [DOI] [PubMed] [Google Scholar]
  2. Allison C, Baron-Cohen S, Wheelwright S, Charman T, Richler J, Pasco G, & Brayne C (2008). The Q-CHAT (Quantitative CHecklist for Autism in Toddlers): a normally distributed quantitative measure of autistic traits at 18–24 months of age: preliminary report. J Autism Dev Disord, 38(8), 1414–1425. doi: 10.1007/s10803-007-0509-7 [DOI] [PubMed] [Google Scholar]
  3. Carter SA, Rahman MM, Lin JC, Shu YH, Chow T, Yu X, … Xiang, A. H. (2022). In utero exposure to near-roadway air pollution and autism spectrum disorder in children. Environ Int, 158, 106898. doi: 10.1016/j.envint.2021.106898 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chen S, Zhao S, Dalman C, Karlsson H, & Gardner R (2021). Association of maternal diabetes with neurodevelopmental disorders: autism spectrum disorders, attention-deficit/hyperactivity disorder and intellectual disability. Int J Epidemiol, 50(2), 459–474. doi: 10.1093/ije/dyaa212 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chu SY, Callaghan WM, Kim SY, Schmid CH, Lau J, England LJ, & Dietz PM (2007). Maternal obesity and risk of gestational diabetes mellitus. Diabetes Care, 30(8), 2070–2076. doi: 10.2337/dc06-2559a [DOI] [PubMed] [Google Scholar]
  6. Coleman KJ, Lutsky MA, Yau V, Qian Y, Pomichowski ME, Crawford PM, … Croen LA. (2015). Validation of Autism Spectrum Disorder Diagnoses in Large Healthcare Systems with Electronic Medical Records. J Autism Dev Disord, 45(7), 1989–1996. doi: 10.1007/s10803-015-2358-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Driscoll AK, & Gregory ECW (2020). Increases in Prepregnancy Obesity: United States, 2016–2019. NCHS Data Brief(392), 1–8. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/33270551 [PubMed] [Google Scholar]
  8. Edlow AG (2017). Maternal obesity and neurodevelopmental and psychiatric disorders in offspring. Prenat Diagn, 37(1), 95–110. doi: 10.1002/pd.4932 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Geschwind DH (2011). Genetics of autism spectrum disorders. Trends Cogn Sci, 15(9), 409–416. doi: 10.1016/j.tics.2011.07.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Green JA (2021). Too many zeros and/or highly skewed? A tutorial on modelling health behaviour as count data with Poisson and negative binomial regression. Health Psychol Behav Med, 9(1), 436–455. doi: 10.1080/21642850.2021.1920416 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Gregory EC, & Ely DM (2022). Trends and Characteristics in Gestational Diabetes: United States, 2016–2020. Natl Vital Stat Rep, 71(3), 1–15. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/35877134 [PubMed] [Google Scholar]
  12. Guo D, Ju R, Zhou Q, Mao J, Tao H, Jing H, … Dai, J. (2020). Association of maternal diabetes with attention deficit/hyperactivity disorder (ADHD) in offspring: A meta-analysis and review. Diabetes Res Clin Pract, 165, 108269. doi: 10.1016/j.diabres.2020.108269 [DOI] [PubMed] [Google Scholar]
  13. Hallmayer J, Cleveland S, Torres A, Phillips J, Cohen B, Torigoe T, … Risch, N. (2011). Genetic heritability and shared environmental factors among twin pairs with autism. Arch Gen Psychiatry, 68(11), 1095–1102. doi: 10.1001/archgenpsychiatry.2011.76 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Ijas H, Koivunen S, Raudaskoski T, Kajantie E, Gissler M, & Vaarasmaki M (2019). Independent and concomitant associations of gestational diabetes and maternal obesity to perinatal outcome: A register-based study. PLoS One, 14(8), e0221549. doi: 10.1371/journal.pone.0221549 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Iwabuchi T, Takahashi N, Nishimura T, Rahman MS, Harada T, Okumura A, … Tsuchiya, K. J. (2021). Associations Among Maternal Metabolic Conditions, Cord Serum Leptin Levels, and Autistic Symptoms in Children. Front Psychiatry, 12, 816196. doi: 10.3389/fpsyt.2021.816196 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Jo H, Eckel SP, Chen JC, Cockburn M, Martinez MP, Chow T, … Xiang, A. H. (2019). Associations of gestational diabetes mellitus with residential air pollution exposure in a large Southern California pregnancy cohort. Environ Int, 130, 104933. doi: 10.1016/j.envint.2019.104933 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Jo H, Eckel SP, Wang X, Chen JC, Cockburn M, Martinez MP, … McConnell, R. (2019). Sex-specific associations of autism spectrum disorder with residential air pollution exposure in a large Southern California pregnancy cohort. Environ Pollut, 254(Pt A), 113010. doi: 10.1016/j.envpol.2019.113010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. King BH (2016). Psychiatric comorbidities in neurodevelopmental disorders. Curr Opin Neurol, 29(2), 113–117. doi: 10.1097/WCO.0000000000000299 [DOI] [PubMed] [Google Scholar]
  19. King MD, & Bearman PS (2011). Socioeconomic Status and the Increased Prevalence of Autism in California. Am Sociol Rev, 76(2), 320–346. doi: 10.1177/0003122411399389 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kodak T, & Bergmann S (2020). Autism Spectrum Disorder: Characteristics, Associated Behaviors, and Early Intervention. Pediatr Clin North Am, 67(3), 525–535. doi: 10.1016/j.pcl.2020.02.007 [DOI] [PubMed] [Google Scholar]
  21. Koebnick C, Langer-Gould AM, Gould MK, Chao CR, Iyer RL, Smith N, … Jacobsen, S. J. (2012). Sociodemographic characteristics of members of a large, integrated health care system: comparison with US Census Bureau data. Perm J, 16(3), 37–41. doi: 10.7812/tpp/12-031 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kong L, Nilsson IAK, Gissler M, & Lavebratt C (2019). Associations of Maternal Diabetes and Body Mass Index With Offspring Birth Weight and Prematurity. JAMA Pediatr, 173(4), 371–378. doi: 10.1001/jamapediatrics.2018.5541 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Landa RJ, Gross AL, Stuart EA, & Faherty A (2013). Developmental trajectories in children with and without autism spectrum disorders: the first 3 years. Child Dev, 84(2), 429–442. doi: 10.1111/j.1467-8624.2012.01870.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Li M, Fallin MD, Riley A, Landa R, Walker SO, Silverstein M, … Wang X (2016). The Association of Maternal Obesity and Diabetes With Autism and Other Developmental Disabilities. Pediatrics, 137(2), e20152206. doi: 10.1542/peds.2015-2206 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lord C, Elsabbagh M, Baird G, & Veenstra-Vanderweele J (2018). Autism spectrum disorder. Lancet, 392(10146), 508–520. doi: 10.1016/S0140-6736(18)31129-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lyall K, Ning X, Aschner JL, Avalos LA, Bennett DH, Bilder DA, … Environmental Influences On Child Health Outcomes, O. (2022). Cardiometabolic Pregnancy Complications in Association with Autism-Related Traits as Measured by the Social Responsiveness Scale in ECHO. Am J Epidemiol doi: 10.1093/aje/kwac061 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Maenner MJ, Schieve LA, Rice CE, Cunniff C, Giarelli E, Kirby RS, … Durkin, M. S. (2013). Frequency and pattern of documented diagnostic features and the age of autism identification. J Am Acad Child Adolesc Psychiatry, 52(4), 401–413 e408. doi: 10.1016/j.jaac.2013.01.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Magiati I, Goh DA, Lim SJ, Gan DZ, Leong JC, Allison C, … group, G. w. (2015). The psychometric properties of the Quantitative-Checklist for Autism in Toddlers (Q-CHAT) as a measure of autistic traits in a community sample of Singaporean infants and toddlers. Mol Autism, 6, 40. doi: 10.1186/s13229-015-0032-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Masi A, DeMayo MM, Glozier N, & Guastella AJ (2017). An Overview of Autism Spectrum Disorder, Heterogeneity and Treatment Options. Neurosci Bull, 33(2), 183–193. doi: 10.1007/s12264-017-0100-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Oono IP, Honey EJ, & McConachie H (2013). Parent-mediated early intervention for young children with autism spectrum disorders (ASD). Cochrane Database Syst Rev(4), CD009774. doi: 10.1002/14651858.CD009774.pub2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Oudin A, Frondelius K, Haglund N, Kallen K, Forsberg B, Gustafsson P, & Malmqvist E (2019). Prenatal exposure to air pollution as a potential risk factor for autism and ADHD. Environ Int, 133(Pt A), 105149. doi: 10.1016/j.envint.2019.105149 [DOI] [PubMed] [Google Scholar]
  32. Patel DR, & Merrick J (2020). Neurodevelopmental and neurobehavioral disorders. Transl Pediatr, 9(Suppl 1), S1–S2. doi: 10.21037/tp.2020.02.03 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Prevention C f. D. C. a. (2022). Data & Statistics on Autism Spectrum Disorder Retrieved from https://www.cdc.gov/ncbddd/autism/data.html [Google Scholar]
  34. Raza S, Zwaigenbaum L, Sacrey LR, Bryson S, Brian J, Smith IM, … Garon N (2019). Brief Report: Evaluation of the Short Quantitative Checklist for Autism in Toddlers (Q-CHAT-10) as a Brief Screen for Autism Spectrum Disorder in a High-Risk Sibling Cohort. J Autism Dev Disord, 49(5), 2210–2218. doi: 10.1007/s10803-019-03897-2 [DOI] [PubMed] [Google Scholar]
  35. Robins DL, Fein D, Barton ML, & Green JA (2001). The Modified Checklist for Autism in Toddlers: an initial study investigating the early detection of autism and pervasive developmental disorders. J Autism Dev Disord, 31(2), 131–144. doi: 10.1023/a:1010738829569 [DOI] [PubMed] [Google Scholar]
  36. Roman-Urrestarazu A, Yanez C, Lopez-Gari C, Elgueta C, Allison C, Brayne C, … Baron-Cohen, S. (2021). Autism screening and conditional cash transfers in Chile: Using the Quantitative Checklist (Q-CHAT) for early autism detection in a low resource setting. Autism, 25(4), 932–945. doi: 10.1177/1362361320972277 [DOI] [PubMed] [Google Scholar]
  37. Ruta L, Chiarotti F, Arduino GM, Apicella F, Leonardi E, Maggio R, … Muratori, F. (2019). Validation of the Quantitative Checklist for Autism in Toddlers in an Italian Clinical Sample of Young Children With Autism and Other Developmental Disorders. Front Psychiatry, 10, 488. doi: 10.3389/fpsyt.2019.00488 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Sacrey LA, Bennett JA, & Zwaigenbaum L (2015). Early Infant Development and Intervention for Autism Spectrum Disorder. J Child Neurol, 30(14), 1921–1929. doi: 10.1177/0883073815601500 [DOI] [PubMed] [Google Scholar]
  39. Sanchez CE, Barry C, Sabhlok A, Russell K, Majors A, Kollins SH, & Fuemmeler BF (2018). Maternal pre-pregnancy obesity and child neurodevelopmental outcomes: a meta-analysis. Obes Rev, 19(4), 464–484. doi: 10.1111/obr.12643 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Smith SE, Li J, Garbett K, Mirnics K, & Patterson PH (2007). Maternal immune activation alters fetal brain development through interleukin-6. J Neurosci, 27(40), 10695–10702. doi: 10.1523/JNEUROSCI.2178-07.2007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Steenweg-de Graaff J, Tiemeier H, Ghassabian A, Rijlaarsdam J, Jaddoe VW, Verhulst FC, & Roza SJ (2016). Maternal Fatty Acid Status During Pregnancy and Child Autistic Traits: The Generation R Study. Am J Epidemiol, 183(9), 792–799. doi: 10.1093/aje/kwv263 [DOI] [PubMed] [Google Scholar]
  42. Tavassoli T, Miller LJ, Schoen SA, Jo Brout J, Sullivan J, & Baron-Cohen S (2018). Sensory reactivity, empathizing and systemizing in autism spectrum conditions and sensory processing disorder. Dev Cogn Neurosci, 29, 72–77. doi: 10.1016/j.dcn.2017.05.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Thapar A, Cooper M, & Rutter M (2017). Neurodevelopmental disorders. Lancet Psychiatry, 4(4), 339–346. doi: 10.1016/S2215-0366(16)30376-5 [DOI] [PubMed] [Google Scholar]
  44. Wang C, Geng H, Liu W, & Zhang G (2017). Prenatal, perinatal, and postnatal factors associated with autism: A meta-analysis. Medicine (Baltimore), 96(18), e6696. doi: 10.1097/MD.0000000000006696 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Winter AS, Fountain C, Cheslack-Postava K, & Bearman PS (2020). The social patterning of autism diagnoses reversed in California between 1992 and 2018. Proc Natl Acad Sci U S A, 117(48), 30295–30302. doi: 10.1073/pnas.2015762117 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Xiang AH (2017). Association of Maternal Diabetes With Autism in Offspring. JAMA, 317(5), 537–538. doi: 10.1001/jama.2016.20122 [DOI] [PubMed] [Google Scholar]
  47. Xiang AH, Chow T, Mora-Marquez J, Martinez MP, Wang X, Yu W, … Schneider, D. I. (2019). Breastfeeding Persistence at 6 Months: Trends and Disparities from 2008 to 2015. J Pediatr, 208, 169–175 e162. doi: 10.1016/j.jpeds.2018.12.055 [DOI] [PubMed] [Google Scholar]
  48. Xiang AH, Wang X, Martinez MP, Getahun D, Page KA, Buchanan TA, & Feldman K (2018). Maternal Gestational Diabetes Mellitus, Type 1 Diabetes, and Type 2 Diabetes During Pregnancy and Risk of ADHD in Offspring. Diabetes Care, 41(12), 2502–2508. doi: 10.2337/dc18-0733 [DOI] [PubMed] [Google Scholar]
  49. Xiang AH, Wang X, Martinez MP, Page K, Buchanan TA, & Feldman RK (2018). Maternal Type 1 Diabetes and Risk of Autism in Offspring. JAMA, 320(1), 89–91. doi: 10.1001/jama.2018.7614 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Xiang AH, Wang X, Martinez MP, Walthall JC, Curry ES, Page K, … Getahun, D. (2015). Association of maternal diabetes with autism in offspring. JAMA, 313(14), 1425–1434. doi: 10.1001/jama.2015.2707 [DOI] [PubMed] [Google Scholar]

RESOURCES