Abstract
Prenatal antidepressant exposure has been associated with increased risk for neurodevelopmental disorders in childhood, including autism spectrum disorder (ASD). The current study utilized multi-cohort data from the Environmental influences on Child Health Outcomes (ECHO) program (N=3129) to test for this association, and determine whether the association remained after adjusting for maternal prenatal depression and other potential confounders. Antidepressants and a subset of selective serotonin reuptake inhibitors (SSRIs) were examined in relation to binary (e.g., diagnostic) and continuous measures of ASD and ASD related traits (e.g., social difficulties, behavior problems) in children 1.5 to 12 years of age. Child sex was tested as an effect modifier. While prenatal antidepressant exposure was associated with ASD related traits in univariate analyses, these associations were statistically non-significant in models that adjusted for prenatal maternal depression and other maternal and child characteristics. Sex assigned at birth was not an effect modifier for the prenatal antidepressant and child ASD relationship. Overall, we found no association between prenatal antidepressant exposures and ASD diagnoses or traits. Discontinuation of antidepressants in pregnancy does not appear to be warranted on the basis of increased risk for offspring ASD.
Keywords: antidepressants, prenatal, neurodevelopment, autism spectrum disorder
Approximately 5% to 10% of women in the United States report antidepressant use (primarily selective serotonin reuptake inhibitors [SSRIs]) during pregnancy (Austin, 2006; Huybrechts et al., 2013 ).Recent studies show SSRI use during pregnancy may be increasing (Bérard et al., 2016; Molenaar et al., 2020). The prevalence of diagnosed autism spectrum disorder (ASD) has also been rising in the United States, leading to a search for environmental exposures that influence risk for this outcome (Chiarotti & Venerosi, 2020). In 2011, a large case-control study exploring the association between prenatal SSRIs and later child neurodevelopmental outcomes linked prenatal SSRI exposure with slightly greater risk for ASD (Croen et al., 2011). This initial report was followed by additional studies and reviews substantiating the link between antidepressant exposures and ASD (Andrade, 2017a; Gentile, 2015; Healy et al., 2016; Kobayashi et al., 2016; Mezzacappa et al., 2017; Rai et al., 2017).
A major concern in the literature on prenatal antidepressant exposure is how to best control for the potentially confounding influence of maternal depression during pregnancy. Prenatal studies often use observational designs that do not control for either prenatal affective disturbance or medication for its treatment. Not all antidepressant exposure studies have incorporated measures of maternal prenatal depression (e.g., Rai et al., 2013). Other studies have inferred maternal prenatal depression status based solely on diagnoses from medical records (e.g., historical diagnoses of Major Depressive Disorder), which may not reflect the actual symptomatology experienced during pregnancy (Brown et al., 2017; Castro et al., 2016; Viktorin et al., 2017). Meta-analyses of the association between maternal antidepressant use during pregnancy and child neurodevelopmental outcomes indicate that confounding by indication is a significant problem in the literature (Andrade, 2017b). Women who are prescribed antidepressants are more likely to have a diagnosis of depression, which in itself carries risk for offspring neurodevelopmental problems. There is a need for studies that measure and control for maternal depression when examining the association between antidepressant medication use in pregnancy and child autism-related outcomes (Rommel et al., 2020).
Considerable heterogeneity of functioning exists among individuals who receive a diagnosis of ASD (Masi et al., 2017). Epidemiological analyses suggest that recent increases in the rates of ASD in the United States are primarily attributed to an increase in mild ASD cases that do not involve intellectual disability (Chiarotti & Venerosi, 2020). Studies that measure social functioning and language deficits have noted poorer performance in these outcome measures for children whose mothers reported antidepressant use in pregnancy (Brandlistuen et al., 2015; Skurtveit et al., 2014; Smearman et al., 2020).
Preclinical studies have also noted changes in social behavior in rodents exposed to SSRIs during the perinatal period, providing experimental evidence for causal associations between social behavior and SSRIs (Gemmel et al., 2018; Houwing et al., 2019). Studies also suggest that biological sex might act as an effect modifier of prenatal SSRI exposure, with poorer neurodevelopmental outcomes typically noted in males relative to females (Leussis et al., 2021). The aims of the current study are to examine associations between prenatal antidepressant exposures and child ASD diagnoses and traits, to test whether these associations are stronger for male offspring relative to female offspring, and to assess whether any noted prenatal medication and ASD associations remain significant when maternal depression and other relevant confounds are controlled.
The current study examines data from the Environmental influences on Child Health Outcomes (ECHO) program, which combines longitudinal data across ongoing child cohort studies throughout the United States (Gillman & Blaisdell, 2018). Use of the ECHO-wide data set allows for a diverse and nationally representative sample, in contrast to the majority of previous studies where samples were limited to one geographic area within the country (Molenaar et al., 2020). We completed secondary analyses of data from 22 unique cohorts (including community cohorts, and case-control cohorts) in the ECHO-wide consortium to test the following hypotheses: (1) Prenatal exposure to any antidepressant medication, and to the subset of SSRI medications in particular, will be associated with a childhood ASD diagnosis and autism-related traits. (2) These hypothesized associations will persist when controlling for potentially confounding maternal and child characteristics, including prenatal maternal depression. (3) Child sex will act as an effect modifier in the association between prenatal antidepressant exposure and child ASD-related outcomes, with males being more susceptible to the exposure than females.
Methods
Participants
This study included mother-child dyads from the ECHO-wide Cohort with available data on prenatal exposure to antidepressants as well as measures of ASD and/or autism-related traits collected when the child was between 1 and 12 years of age (N=3129). Table 1 shows demographic and prenatal characteristics of women in the sample who took antidepressants during pregnancy (approximately 5% of the sample) compared with those who did not take antidepressants during pregnancy.
Table 1.
Mothers without any antidepressant use | Mothers with any antidepressant use | P-value | |
---|---|---|---|
(N=2966) | (N=163) | ||
| |||
Depression diagnosis? during pregnancy | |||
No | 1771 (59.7%) | 13 (8.0%) | <0.001 |
Yes | 170 (5.7%) | 88 (54.0%) | |
Missing | 1025 (34.6%) | 62 (38.0%) | |
Age at the birth of the first child | |||
Mean (SD) | 30.7 (5.51) | 31.3 (5.23) | 0.172 |
Median [Min, Max] | 31.0 [15.0, 50.0] | 31.0 [16.0, 45.0] | |
Missing | 18 (0.6%) | 0 (0%) | |
Race category | |||
White | 2093 (70.6%) | 133 (81.6%) | <0.001 |
Black | 282 (9.5%) | <5 (<3.1%) | |
Asian | 135 (4.6%) | 0 (0%) | |
American Indian or Alaska Native | 16 (0.5%) | 0 (0%) | |
Multiple Race | 34 (1.1%) | 0 (0%) | |
Native Hawaiian or other Pacific Islander | 162 (5.5%) | 8 (4.9%) | |
Other Race | 88 (3.0%) | 0 (0%) | |
Missing | 156 (5.3%) | <20 (<12.3%) | |
Ethnicity | |||
non-Hispanic | 2485 (83.8%) | 134 (82.2%) | 0.307 |
Hispanic | 360 (12.1%) | 14 (8.6%) | |
Missing | 121 (4.1%) | 15 (9.2%) | |
Highest education a | |||
Less than high school | 104 (3.5%) | <5 (<3.1%) | 0.225 |
High school degree | 292 (9.8%) | 10 (6.1%) | |
Some college without degree | 792 (26.7%) | 49 (30.1%) | |
Bachelor’s degree | 903 (30.4%) | 50 (30.7%) | |
Master’s degree or above | 836 (28.2%) | 51 (31.3%) | |
Missing | 39 (1.3%) | <5 (<3.1%) | |
Marital status b | |||
Married | 2438 (82.2%) | 133 (81.6%) | 0.375 |
Not married | 445 (15.0%) | <30 (< 18.4%) | |
Missing | 83 (2.8%) | <5 (<3.1%) | |
Income of the household, 5 categories | |||
<$30,000 | 354 (11.9%) | 17 (10.4%) | 0.938 |
$30,000–$49,999 | 227 (7.7%) | 13 (8.0%) | |
$50,000–$74,999 | 238 (8.0%) | 14 (8.6%) | |
$75,000–$99,999 | 153 (5.2%) | 10 (6.1%) | |
$100,000 or more | 317 (10.7%) | 17 (10.4%) | |
Missing | 1677 (56.5%) | 92 (56.4%) | |
Mode of delivery for the first child | |||
vaginal | 1763 (59.4%) | 83 (50.9%) | 0.64 |
cesarean | 935 (31.5%) | 48 (29.4%) | |
Missing | 268 (9.0%) | 32 (19.6%) | |
Gestational age at birth | |||
Mean (SD) | 35.7 (5.56) | 37.0 (4.31) | <0.001 |
Median [Min, Max] | 38.0 [22.0, 43.0] | 39.0 [23.0, 42.0] | |
Missing | 11 (0.4%) | 0 (0%) | |
Prenatal tobacco/nicotine use | |||
No | 2612 (88.1%) | 130 (79.8%) | 0.045 |
Yes | 264 (8.9%) | 22.0 (13.5%) | |
Missing | 90.0 (3.0%) | 11.0 (6.7%) | |
Prenatal alcohol use | |||
No | 2373 (80.0%) | 125 (76.7%) | 0.206 |
Yes | 421 (14.2%) | 29 (17.8%) | |
Missing | 172 (5.8%) | 9 (5.5%) | |
Prenatal illicit drug use | |||
No | 2305 (77.7%) | 119 (73.0%) | 0.683 |
Yes | 32 (1.1%) | <5 (3.1%) | |
Missing | 629 (21.2%) | <45 (27.6%) | |
Prenatal marijuana use | |||
No | 2386 (80.4%) | 122 (74.8%) | 0.816 |
Yes | 99 (3.3%) | <5 (<3.1%) | |
Missing | 481 (16.2%) | <40 (<24.5%) | |
Pre-pregnancy BMI, kg/m2 | |||
Mean (SD) | 27.1 (6.99) | 29.8 (8.30) | <0.001 |
Median [Min, Max] | 25.2 [15.6, 73.1] | 28.1 [15.8, 60.9] | |
Missing | 401 (13.5%) | 26 (16.0%) |
Some college without degree also includes associate degree and trade school. Master’s degree or above also includes (professional) doctoral degrees.
Married stands for Married or living with a partner; Not married includes widowed, separated, divorced, single, never married, and partnered (boyfriend or girlfriend) but not living together.
preterm gestational category is defined as gestational age at birth from 22 to 36 weeks.
P-values were given from the t-test for continuous variables or chi-square test for categorical variables.
Measures
The current study utilized ECHO prenatal exposure and child neurodevelopmental measures as detailed below. Data were collected according to the ECHO data collection protocol (https://echochildren.org/wp-content/uploads/2021/05/ECHO-wide-Cohort-Data-Collection-Protocol-Version2.0_24FEB2021_clean.pdf).
Maternal Prenatal Antidepressant and SSRI Use
Information on maternal prenatal antidepressant medication use was obtained from two sources in the ECHO database: the Maternal Medical Record Abstraction form (MMRA), and the Pregnancy Medical Conditions and Interventions (PMCI) questionnaire completed by the mothers. The MMRA was completed once for each pregnancy included in ECHO. Information was gathered for the period from 4 weeks prior to the last menstrual period (LMP) before the pregnancy through 8 weeks postpartum. All medications listed in the medical records were recorded with notation of timing before, during, or after pregnancy. The PMCI was completed by the biological mother for each pregnancy resulting in a child (or children) being enrolled in an ECHO cohort. The questionnaire, completed during the child’s infancy, early childhood, or middle childhood life stages, asked about the mother’s health, medical conditions she had before or during pregnancy, tests or procedures that were done, and medical complications that occurred during pregnancy, labor, and/or delivery. Mothers also reported whether they had taken medication for depression during their pregnancy. A detailed list of antidepressants was provided that included generic and brand names as well as options for adding additional medications that were not specified. The most common type of antidepressant medications used by mothers in this study was SSRIs (~72%), and the most common SSRI medications were paroxetine and sertraline.
Binary Measures of ASD
ASD diagnosis was obtained from several sources including established standardized diagnostic instruments, such as the Autism Diagnostic Observation Schedule (ADOS; Lord et al., 2000), parental or other caregiver report of an ASD diagnosis, or a diagnosis extracted from medical or educational records. A subset of mothers also completed the Modified Checklist for Autism in Toddlers-Revised with Follow-up (M-CHAT-R/F; Robins et al., 2014). This 20-item checklist has demonstrated reliability and validity as a screener for ASD in typically developing toddlers (ages 16 to 30 months old). Children were categorized as above and below the cut-off for probable ASD according to their score on the M-CHAT-R/F. These bivariate measures of ASD (i.e., diagnoses and M-CHAT-R/F screener results) were examined as separate outcomes in this study.
Continuous Measures of Child Autism-Related Traits
Mothers completed the Social Responsiveness Scale, Second Edition (SRS-2; Constantino & Gruber, 2012) about their child. The SRS-2 is a 65-item scale that evaluates the severity of ASD symptoms occurring in day-to-day scenarios. The preschool version was used for children aged 31 months through 47 months, and the school-age version was used for children from ages 4 to 12 years. The overall raw score was used—higher scores reflect more autism-related traits. Mothers also completed the Preschool-Age Child Behavior Checklist (preschool CBCL; Achenbach & Rescorla, 2000) in reference to their children aged 18 months to five years of age. The DSM-5 Autism Spectrum Problems (ASP) subscale of the preschool CBCL assesses the occurrence of 12 child behaviors over the previous 2 months that were rated by international clinical experts as “very consistent” with ASD, such as avoiding eye contact and not getting along with others (Achenbach, 2014). Higher scores indicate more ASD-related behaviors. Mothers also completed the School-Age Child Behavior Checklist (school-age CBCL; Achenbach, 2001) in reference to their children aged 6 to 12 years of age. t-scores on the Withdrawn, Social Problems, and Thought Problems subscales were summed to create a school-age CBCL ASD-related behavior score for each child. This score has been shown to discriminate between children with and without ASD, particularly among high-functioning youth (Biederman et al., 2010).
Covariates
We used a directed acyclic graph (see Supplemental Figure 1) to decide what covariates to include in analyses based on conceptual and empirical justification from the existing literature. Maternal prenatal depression was included as a covariate and was considered present if the maternal medical record indicated this as a condition, or if the mother self-reported this condition on the PMCI questionnaire. Additional covariates included the following maternal characteristics: age, education, race/ethnicity (non-Hispanic White versus others), marital status, prenatal substance use (tobacco, alcohol, marijuana), and pre-pregnancy BMI (obtained from medical record abstractions); as well as the child characteristics of age and sex.
Statistical Analysis
Observations in this study conformed to assumptions of independence. In cases of multiple sibling participation, one sibling was randomly selected for inclusion in the study. In cases of multiple administrations of the same maternal reported measure of ASD, the measure completed at the youngest child age (closest to the prenatal exposure) was selected for inclusion in the analyses. General antidepressant use and SSRI use were examined in separate analyses, as were ASD and autism-related traits. Medical record and self-report indicators of prenatal exposure to antidepressants were examined separately (in univariate models) and combined (in univariate and multivariate models). Multiple imputation chained equations models (van Buuren & Groothuis-Oudshoorn, 2011) were used to account for the missingness in covariates, including child’s age at visit for each outcome variable, child’s sex, prenatal depression diagnosis, maternal age, maternal race/ethnicity (non-Hispanic White versus others), maternal education level, maternal marital status, prenatal tobacco use, prenatal alcohol use, prenatal marijuana use, and pre-pregnancy BMI. Regression estimates were pooled based on the combination of 25 imputations using the Rubin’s rules (Rubin, 2004). Data were not imputed for prenatal medications or ASD outcomes, so analyses had varying sample sizes (as reported in the results tables).
Mixed-effects models were used to perform univariate and multivariate analyses, adjusting for cohort effects by including cohort as a random intercept in the models. Linear mixed-effects models were used for continuous outcomes (using the “lmer” function from the “lme4” R package), examining coefficients, confidence intervals (CIs), and P-values from these models. Mixed-effects logistic models were used for binary outcomes (using the “glmer” function from the “lme4” R package), examining odds ratios, CIs, and P-values. Models were fitted with maximum likelihood estimators. Wald 95% CIs were constructed, and P-values were derived from the Wald z-test.
Univariate analyses examined prenatal antidepressant and SSRI exposure as the predictor and child ASD measures as the outcome. The interaction term of child sex and each prenatal antidepressant and SSRI exposure was tested to assess whether sex was a significant effect modifier. Finally, multivariate analyses were performed by including all covariates in the model and assessing for independent effects of prenatal antidepressant or SSRI exposures on each ASD outcome. All analyses were conducted in R software version 4.1.0.
Results
Our hypothesis that prenatal exposures to antidepressants (and the subset of SSRI medications) would be associated with child ASD and autism-related traits was partially supported in univariate analyses. As can be seen in Table 2, neither SSRI use nor general antidepressant use in pregnancy was associated with ASD diagnoses or with M-CHAT-R/F cut-off scores. However, analyses that included continuous measures of ASD-related traits revealed that any prenatal antidepressant use (and self-reported antidepressant use in particular) predicted higher child SRS-2 total raw scores. In addition, prenatal SSRI use (and self-reported SSRI use in particular) predicted higher child preschool CBCL-ASP t-scores, and self-reported prenatal antidepressant use was significantly associated with higher child school-age CBCL sum t-scores for ASD.
Table 2.
Associations (exposed vs. unexposed groups for outcomes) | N-Total | N-Exposed | Coefficient/OR | LCL | UCL | P-value |
---|---|---|---|---|---|---|
SRS total raw score, N=1346 | ||||||
Any antidepressant | 1339 | 76 | 6.4 | 1.0 | 11.9 | 0.02 |
SSRI | 1214 | 50 | 6.5 | −0.3 | 13.3 | 0.06 |
Medical reported any antidepressant | 118 | 8 | 11.4 | −12.0 | 34.9 | 0.34 |
Medical reported SSRI | 120 | 7 | 14.3 | −11.4 | 39.9 | 0.28 |
Self-reported any antidepressant | 1221 | 68 | 5.7 | 0.2 | 11.2 | 0.04 |
Self-reported SSRI | 1094 | 43 | 5.1 | −1.7 | 12.0 | 0.14 |
CBCL DSM-5 Autism t score for preschool children, N=1202 | ||||||
Any antidepressant | 1199 | 63 | 1.0 | −0.7 | 2.6 | 0.25 |
SSRI | 1072 | 38 | 2.2 | 0.1 | 4.4 | 0.04 |
Medical reported any antidepressant | 59 | 6 | 1.1 | −3.7 | 6.0 | 0.65 |
Medical reported SSRI | 58 | 5 | 2.0 | −3.3 | 7.2 | 0.47 |
Self-reported any antidepressant | 1140 | 57 | 1.0 | −0.7 | 2.7 | 0.25 |
Self-reported SSRI | 1014 | 33 | 2.4 | 0.1 | 4.7 | 0.04 |
CBCL Autism Sum t score for school age children, N=546 | ||||||
Any antidepressant | 546 | 35 | 5.0 | −0.3 | 10.3 | 0.07 |
SSRI | 477 | 22 | 0.9 | −5.7 | 7.6 | 0.78 |
Medical reported any antidepressant | 59 | <5 | NA | NA | NA | NA |
Medical reported SSRI | 59 | <5 | NA | NA | NA | NA |
Self-reported any antidepressant | 487 | 32 | 5.5 | 0.4 | 10.6 | 0.03 |
Self-reported SSRI | 418 | 19 | 1.2 | −5.3 | 7.7 | 0.73 |
ASD diagnosis, N=2384 | ||||||
Any antidepressant | 2377 | 98 | 1.3 | 0.6 | 2.7 | 0.51 |
SSRI | 2323 | 76 | 1.8 | 0.7 | 4.2 | 0.20 |
Medical reported any antidepressant | 515 | 21 | 1.2 | 0.4 | 3.4 | 0.76 |
Medical reported SSRI | 512 | 15 | 1.9 | 0.5 | 7.1 | 0.32 |
Self-reported any antidepressant | 1862 | 77 | 1.4 | 0.5 | 4.1 | 0.49 |
Self-reported SSRI | 1811 | 61 | 1.7 | 0.5 | 5.6 | 0.40 |
M-CHAT, N=834 | ||||||
Any antidepressant | 831 | 41 | 1.3 | 0.4 | 4.0 | 0.64 |
SSRI | 761 | 23 | 1.0 | 0.2 | 4.5 | 0.98 |
Medical reported any antidepressant | 42 | <5 | NA | NA | NA | NA |
Medical reported SSRI | 42 | <5 | NA | NA | NA | NA |
Self-reported any antidepressant | 789 | 37 | 1.7 | 0.5 | 5.5 | 0.35 |
Self-reported SSRI | 719 | 19 | 1.5 | 0.3 | 7.6 | 0.60 |
N, number of subjects; LCL, lower confidence limit; UCL, upper confidence limit; OR, odds ratio.
SRS total raw score, CBCL DSM-5 Autism t score, CBCL Autism Sum t score are continuous outcomes. The estimated coefficient of exposure are given in the third column. ASD diagnosis and M-CHAT are binary outcomes. The estimated OR for each binary outcome variable comparing the exposed and unexposed groups are given in the third column.
NA, not applicable. It represents that corresponding model did not converge or the estimations were too extreme to be true.
Child sex was also explored as an effect modifier of the association between prenatal antidepressant/SSRI exposure and child ASD outcomes. No exposure-child sex interaction terms were statistically significant. Table 3 presents the univariate associations separated by sex.
Table 3.
Associations (exposed vs. unexposed groups for outcomes) | Child’s sex | N | Coefficient/OR | LCL | UCL | P-value |
---|---|---|---|---|---|---|
SRS-2 total raw score, N=1346 | ||||||
Any antidepressant | Male | 732 | 6.8 | −1.4 | 15.0 | 0.12 |
Female | 607 | 5.8 | −3.5 | 15.1 | 0.30 | |
Interaction | −1 | −11.8 | 9.8 | 0.12 | ||
SSRI | Male | 664 | 7.2 | −2.6 | 17.0 | 0.19 |
Female | 550 | 4.5 | −7.9 | 16.9 | 0.66 | |
Interaction | −2.7 | −16.5 | 11.1 | 0.19 | ||
Medical reported any antidepressant | Male | 70 | 15.6 | −17.8 | 49.0 | 0.51 |
Female | 48 | 5.0 | −38.4 | 48.4 | 0.96 | |
Interaction | −10.6 | −58.3 | 37 | 0.5 | ||
Medical reported SSRI | Male | 71 | 20.5 | −18.1 | 59.0 | 0.42 |
Female | 49 | 5.6 | −39.1 | 50.3 | 0.95 | |
Interaction | −14.9 | −66.6 | 36.8 | 0.41 | ||
Self-reported any antidepressant | Male | 662 | 5.6 | −2.7 | 13.8 | 0.25 |
Female | 559 | 5.7 | −3.6 | 15.0 | 0.31 | |
Interaction | 0.1 | −10.7 | 11 | 0.25 | ||
Self-reported SSRI | Male | 593 | 5.3 | −4.5 | 15.2 | 0.40 |
Female | 501 | 3.5 | −9.2 | 16.3 | 0.78 | |
Interaction | −1.8 | −15.9 | 12.3 | 0.4 | ||
CBCL DSM-5 ASP t-score for preschool children, N=1202 | ||||||
Any antidepressant | Male | 653 | 1.1 | −1.2 | 3.5 | 0.49 |
Female | 546 | 0.5 | −2.4 | 3.5 | 0.90 | |
Interaction | −0.6 | −3.9 | 2.7 | 0.49 | ||
SSRI | Male | 577 | 2.5 | −0.5 | 5.6 | 0.13 |
Female | 495 | 1.4 | −2.5 | 5.4 | 0.66 | |
Interaction | −1.1 | −5.5 | 3.3 | 0.13 | ||
Medical reported any antidepressant | Male | 33 | −2.9 | −12.2 | 6.3 | 0.73 |
Female | 26 | 3.5 | −3.4 | 10.4 | 0.45 | |
Interaction | 6.4 | −3.7 | 16.5 | 0.73 | ||
Medical reported SSRI | Male | 32 | −2.4 | −15.4 | 10.6 | 0.90 |
Female | 26 | 3.5 | −3.4 | 10.4 | 0.45 | |
Interaction | 5.9 | −7 | 18.8 | 0.9 | ||
Self-reported any antidepressant | Male | 620 | 1.4 | −1.1 | 3.8 | 0.39 |
Female | 520 | 0.2 | −3.0 | 3.3 | 0.99 | |
Interaction | −1.2 | −4.7 | 2.3 | 0.39 | ||
Self-reported SSRI | Male | 545 | 2.7 | −0.4 | 5.9 | 0.10 |
Female | 469 | 1.0 | −3.7 | 5.7 | 0.86 | |
Interaction | −1.7 | −6.7 | 3.2 | 0.1 | ||
CBCL autism sum t-score for school-age children, N=546 | ||||||
Any antidepressant | Male | 301 | 5.6 | −2.6 | 13.8 | 0.23 |
Female | 245 | 4.2 | −4.7 | 13.1 | 0.50 | |
Interaction | −1.4 | −12 | 9.2 | 0.23 | ||
SSRI | Male | 268 | 1.4 | −8.5 | 11.3 | 0.94 |
Female | 209 | 0.1 | −11.8 | 12.0 | 1.00 | |
Interaction | −1.3 | −14.9 | 12.3 | 0.94 | ||
Medical reported any antidepressant | Male | 44 | NA | NA | NA | NA |
Female | 15 | NA | NA | NA | NA | |
Interaction | NA | NA | NA | NA | ||
Medical reported SSRI | Male | 44 | NA | NA | NA | NA |
Female | 15 | NA | NA | NA | NA | |
Interaction | NA | NA | NA | NA | ||
Self-reported any antidepressant | Male | 257 | 6.5 | −1.5 | 14.6 | 0.14 |
Female | 230 | 4.7 | −3.4 | 12.8 | 0.35 | |
Interaction | −1.9 | −11.9 | 8.2 | 0.14 | ||
Self-reported SSRI | Male | 224 | 1.9 | −8.2 | 12.1 | 0.89 |
Female | 194 | 0.4 | −10.3 | 11.1 | 1.00 | |
Interaction | −1.6 | −14.5 | 11.3 | 0.89 | ||
ASD diagnosis, N=2384 | ||||||
Any antidepressant | Male | 1304 | 1.1 | 0.4 | 3.0 | 0.94 |
Female | 1073 | 1.5 | 0.3 | 8.8 | 0.85 | |
Interaction | 1.3 | 0.2 | 7.7 | 0.94 | ||
SSRI | Male | 1274 | 2.2 | 0.7 | 6.7 | 0.24 |
Female | 1049 | 0.8 | 0.1 | 9.2 | 0.97 | |
Interaction | 0.4 | 0.03 | 3.9 | 0.24 | ||
Medical reported any antidepressant | Male | 325 | 1.1 | 0.3 | 4.2 | 0.98 |
Female | 190 | 1.3 | 0.1 | 22.9 | 0.98 | |
Interaction | 1.2 | 0.1 | 18.9 | 0.98 | ||
Medical reported SSRI | Male | 321 | 2.4 | 0.4 | 14.7 | 0.51 |
Female | 191 | 1.3 | 0.1 | 23.0 | 0.98 | |
Interaction | 0.5 | 0.03 | 10.9 | 0.51 | ||
Self-reported any antidepressant | Male | 979 | 1.2 | 0.3 | 5.0 | 0.93 |
Female | 883 | 1.7 | 0.2 | 17.3 | 0.84 | |
Interaction | 1.4 | 0.1 | 14.7 | 0.93 | ||
Self-reported SSRI | Male | 953 | NA | NA | NA | NA |
Female | 858 | NA | NA | NA | NA | |
Interaction | NA | NA | NA | NA | ||
M-CHAT-R/F, N=834 | ||||||
Any antidepressant | Male | 451 | 1.5 | 0.3 | 6.7 | 0.80 |
Female | 380 | 0.9 | 0.1 | 10.0 | 1.00 | |
Interaction | 0.6 | 0.1 | 7.2 | 0.8 | ||
SSRI | Male | 412 | NA | NA | NA | NA |
Female | 349 | NA | NA | NA | NA | |
Interaction | NA | NA | NA | NA | ||
Medical reported any antidepressant | Male | 22 | NA | NA | NA | NA |
Female | 20 | NA | NA | NA | NA | |
Interaction | NA | NA | NA | NA | ||
Medical reported SSRI | Male | 22 | NA | NA | NA | NA |
Female | 20 | NA | NA | NA | NA | |
Interaction | NA | NA | NA | NA | ||
Self-reported any antidepressant | Male | 429 | 1.8 | 0.4 | 8.3 | 0.65 |
Female | 360 | 1.4 | 0.1 | 15.8 | 0.94 | |
Interaction | 0.8 | 0.1 | 9.3 | 0.65 | ||
Self-reported SSRI | Male | 390 | NA | NA | NA | NA |
Female | 329 | NA | NA | NA | NA | |
Interaction | NA | NA | NA | NA |
ASD, autism spectrum disorder; CBCL, Child Behavior Checklist; LCL, lower confidence limit; M-CHAT-R/F, Modified Checklist for Autism in Toddlers-Revised with Follow-up; NA, not applicable; OR, odds ratio; SRS-2, Social Responsiveness Scale, Second Edition; SSRI, selective serotonin reuptake inhibitors; UCL, upper confidence limit
SRS-2 total raw score, CBCL DSM-5 Autism t-score, and CBCL autism sum t-score are continuous outcomes. The estimated coefficients of exposure are given in the third column. ASD diagnosis and M-CHAT-R/F are binary outcomes. The estimated OR for each binary outcome variable comparing the exposed and unexposed groups are given in the third column. NA represents that the corresponding model did not converge or the estimations were too extreme to be true.
Multivariate regression analyses did not support our second hypothesis that prenatal medication and child ASD outcome relationships would hold when controlling for child age, child sex, maternal age, maternal race/ethnicity, maternal marital status, maternal education, prenatal substance exposures, pre-pregnancy BMI, and prenatal maternal depression (see Tables 4 and 5). Neither prenatal antidepressant use nor SSRI use predicted any of the ASD outcome measures when these potential confounders were controlled.
Table 4.
Models | Coefficient/OR | LCL | UCL | P-value |
---|---|---|---|---|
SRS total raw ~ any antidepressant, N=1339 | ||||
Any prenatal antidepressant use: Yes vs. No | 2.5 | −4.1 | 9.1 | 0.45 |
Age at visit for the outcome, yrs | 0.5 | −0.2 | 1.3 | 0.17 |
Child’s sex: Female vs. Male | −4.3 | −6.7 | −1.9 | 0.001 |
Prenatal depression diagnosis: Yes vs. No | 2.9 | −1.9 | 7.7 | 0.24 |
Maternal age at delivery, yrs | 0.0 | −0.2 | 0.2 | 0.98 |
Maternal Non-Hispanic white vs. others | −3.1 | −5.9 | −0.2 | 0.03 |
Maternal above vs. less than high school education | −9.6 | −13.5 | −5.6 | <0.001 |
Maternal not married vs. married | 8.7 | 5.1 | 12.3 | <0.001 |
Prenatal tobacco use: Yes vs. No | 4.1 | −0.6 | 8.7 | 0.09 |
Prenatal alcohol use: Yes vs. No | 0.3 | −3.4 | 4.1 | 0.86 |
Prenatal marijuana use: Yes vs. No | 1.2 | −5.9 | 8.4 | 0.74 |
Pre-pregnancy BMI, kg/m2 | 0.3 | 0.2 | 0.5 | <0.001 |
CBCL DSM-5 Autism t score for preschool children ~ any antidepressant, N=1199 | ||||
Any prenatal antidepressant use: Yes vs. No | −0.4 | −2.4 | 1.6 | 0.69 |
Age at visit for the outcome, yrs | 0.0 | −0.5 | 0.5 | 0.89 |
Child’s sex: Female vs. Male | −0.9 | −1.6 | −0.2 | 0.02 |
Prenatal depression diagnosis: Yes vs. No | 1.6 | 0.0 | 3.1 | 0.05 |
Maternal age at delivery, yrs | 0.0 | −0.1 | 0.1 | 0.90 |
Maternal Non-Hispanic white vs. others | −0.7 | −1.5 | 0.1 | 0.11 |
Maternal above vs. less than high school education | −1.2 | −2.3 | −0.2 | 0.02 |
Maternal not married vs. married | 1.6 | 0.5 | 2.6 | 0.003 |
Prenatal tobacco use: Yes vs. No | 0.7 | −0.5 | 2.0 | 0.26 |
Prenatal alcohol use: Yes vs. No | −1.1 | −2.2 | −0.1 | 0.04 |
Prenatal marijuana use: Yes vs. No | 1.0 | −0.7 | 2.6 | 0.24 |
Pre-pregnancy BMI, kg/m2 | 0.0 | 0.0 | 0.1 | 0.16 |
CBCL Autism Sum t score for school age children~ any antidepressant, N=546 | ||||
Any prenatal antidepressant use: Yes vs. No | 5.0 | −2.0 | 12.0 | 0.16 |
Age at visit for the outcome, yrs | −0.1 | −1.5 | 1.3 | 0.89 |
Child’s sex: Female vs. Male | −3.2 | −5.8 | −0.7 | 0.01 |
Prenatal depression diagnosis: Yes vs. No | −1.4 | −7.1 | 4.3 | 0.64 |
Maternal age at delivery, yrs | −0.1 | −0.3 | 0.1 | 0.42 |
Maternal Non-Hispanic white vs. others | −0.5 | −3.6 | 2.7 | 0.76 |
Maternal above vs. less than high school education | −0.5 | −5.6 | 4.7 | 0.86 |
Maternal not married vs. married | 7.1 | 2.8 | 11.4 | 0.001 |
Prenatal tobacco use: Yes vs. No | 2.1 | −2.5 | 6.8 | 0.37 |
Prenatal alcohol use: Yes vs. No | 3.6 | −0.3 | 7.5 | 0.07 |
Prenatal marijuana use: Yes vs. No | 12.1 | 0.1 | 24.1 | 0.05 |
Pre-pregnancy BMI, kg/m2 | 0.4 | 0.2 | 0.7 | <0.001 |
ASD diagnosis~ any antidepressant, N=2377 | ||||
Any prenatal antidepressant use: Yes vs. No | 1.0 | 0.4 | 2.3 | 0.97 |
Age at visit for the outcome, yrs | 1.2 | 1.1 | 1.4 | 0.01 |
Child’s sex: Female vs. Male | 0.4 | 0.3 | 0.6 | <0.001 |
Prenatal depression diagnosis: Yes vs. No | 1.3 | 0.8 | 2.2 | 0.26 |
Maternal age at delivery, yrs | 1.0 | 1.0 | 1.0 | 0.32 |
Maternal Non-Hispanic white vs. others | 1.2 | 0.9 | 1.7 | 0.29 |
Maternal above vs. less than high school education | 0.9 | 0.6 | 1.4 | 0.66 |
Maternal not married vs. married | 1.3 | 0.8 | 1.9 | 0.25 |
Prenatal tobacco use: Yes vs. No | 1.4 | 0.9 | 2.2 | 0.19 |
Prenatal alcohol use: Yes vs. No | 0.6 | 0.3 | 1.2 | 0.17 |
Prenatal marijuana use: Yes vs. No | 0.8 | 0.4 | 1.7 | 0.63 |
Pre-pregnancy BMI, kg/m2 | 1.0 | 1.0 | 1.0 | 0.48 |
MCHAT~ any antidepressant, N=831 | ||||
Any prenatal antidepressant use: Yes vs. No | 1.4 | 0.3 | 5.6 | 0.64 |
Age at visit for the outcome, yrs | 0.8 | 0.5 | 1.3 | 0.34 |
Child’s sex: Female vs. Male | 0.5 | 0.4 | 0.8 | 0.01 |
Prenatal depression diagnosis: Yes vs. No | 0.8 | 0.3 | 2.1 | 0.67 |
Maternal age at delivery, yrs | 1.02 | 0.99 | 1.1 | 0.23 |
Maternal Non-Hispanic white vs. others | 1.0 | 0.6 | 1.5 | 0.88 |
Maternal above vs. less than high school education | 0.4 | 0.2 | 0.6 | <0.00 |
Maternal not married vs. married | 1.9 | 1.1 | 3.1 | 0.01 |
Prenatal tobacco use: Yes vs. No | 1.7 | 0.9 | 3.1 | 0.09 |
Prenatal alcohol use: Yes vs. No | 1.0 | 0.4 | 3.0 | 0.94 |
Prenatal marijuana use: Yes vs. No | 0.7 | 0.3 | 1.5 | 0.32 |
Pre-pregnancy BMI, kg/m2 | 1.01 | 0.99 | 1.04 | 0.32 |
ASD, autism spectrum disorder; BMI, body mass index; CBCL, Child Behavior Checklist; LCL, lower confidence limit; M-CHAT-R/F, Modified Checklist for Autism in Toddlers-Revised with Follow-up; OR, odds ratio; SRS-2, Social Responsiveness Scale, Second Edition; UCL, upper confidence limit
Table 5.
Models | Coefficient/OR | LCL | UCL | P-value |
---|---|---|---|---|
SRS total raw ~ SSRI, N=1214 | ||||
Prenatal SSRI use Yes vs. No | 0.9 | −6.8 | 8.6 | 0.82 |
Age at visit for the outcome, yrs | 0.4 | −0.5 | 1.2 | 0.39 |
Child’s sex: Female vs. Male | −4.4 | −7.0 | −1.8 | 0.001 |
Prenatal depression diagnosis: Yes vs. No | 4.2 | −0.8 | 9.2 | 0.10 |
Maternal age at delivery, yrs | 0.0 | −0.2 | 0.3 | 0.87 |
Maternal Non-Hispanic white vs. others | −3.0 | −6.1 | 0.1 | 0.06 |
Maternal above vs. less than high school education | −10.2 | −14.4 | −6.0 | <0.001 |
Maternal not married vs. married | 8.6 | 4.7 | 12.4 | <0.001 |
Prenatal tobacco use: Yes vs. No | 2.5 | −2.2 | 7.3 | 0.30 |
Prenatal alcohol use: Yes vs. No | 0.2 | −3.9 | 4.2 | 0.94 |
Prenatal marijuana use: Yes vs. No | 2.2 | −5.4 | 9.7 | 0.57 |
Pre-pregnancy BMI, kg/m2 | 0.4 | 0.2 | 0.6 | <0.001 |
CBCL DSM-5 Autism t score for preschool children ~ SSRI, N=1072 | ||||
Prenatal SSRI use Yes vs. No | 0.4 | −2.1 | 2.9 | 0.76 |
Age at visit for the outcome, yrs | 0.01 | −0.6 | 0.6 | 0.98 |
Child’s sex: Female vs. Male | −0.8 | −1.6 | −0.1 | 0.04 |
Prenatal depression diagnosis: Yes vs. No | 1.9 | 0.2 | 3.5 | 0.03 |
Maternal age at delivery, yrs | −0.02 | −0.1 | 0.1 | 0.70 |
Maternal Non-Hispanic white vs. others | −0.7 | −1.6 | 0.2 | 0.13 |
Maternal above vs. less than high school education | −1.2 | −2.3 | −0.1 | 0.04 |
Maternal not married vs. married | 1.5 | 0.3 | 2.6 | 0.01 |
Prenatal tobacco use: Yes vs. No | 0.7 | −0.6 | 2.0 | 0.29 |
Prenatal alcohol use: Yes vs. No | −1.2 | −2.4 | −0.1 | 0.04 |
Prenatal marijuana use: Yes vs. No | 1.2 | −0.6 | 2.9 | 0.20 |
Pre-pregnancy BMI, kg/m2 | 0.04 | −0.01 | 0.1 | 0.15 |
CBCL Autism Sum t score for school age children ~ SSRI, N=477 | ||||
Prenatal SSRI use Yes vs. No | 1.1 | −6.8 | 9.0 | 0.79 |
Age at visit for the outcome, yrs | −0.6 | −2.2 | 1.0 | 0.48 |
Child’s sex: Female vs. Male | −3.0 | −5.8 | −0.2 | 0.04 |
Prenatal depression diagnosis: Yes vs. No | −2.7 | −8.4 | 3.0 | 0.35 |
Maternal age at delivery, yrs | −0.1 | −0.3 | 0.2 | 0.50 |
Maternal Non-Hispanic white vs. others | 0.5 | −3.0 | 4.1 | 0.77 |
Maternal above vs. less than high school education | −1.0 | −6.6 | 4.5 | 0.71 |
Maternal not married vs. married | 6.9 | 2.3 | 11.5 | 0.003 |
Prenatal tobacco use: Yes vs. No | 1.9 | −2.9 | 6.7 | 0.44 |
Prenatal alcohol use: Yes vs. No | 3.6 | −0.6 | 7.9 | 0.09 |
Prenatal marijuana use: Yes vs. No | 16.1 | 2.9 | 29.3 | 0.02 |
Pre-pregnancy BMI, kg/m2 | 0.5 | 0.3 | 0.7 | <0.001 |
ASD diagnosis ~ SSRI, N=2323 | ||||
Prenatal SSRI use Yes vs. No | 1.5 | 0.6 | 3.5 | 0.38 |
Age at visit for the outcome, yrs | 1.2 | 1.1 | 1.4 | 0.004 |
Child’s sex: Female vs. Male | 0.4 | 0.3 | 0.6 | <0.001 |
Prenatal depression diagnosis: Yes vs. No | 1.2 | 0.8 | 1.8 | 0.49 |
Maternal age at delivery, yrs | 1.0 | 1.0 | 1.0 | 0.29 |
Maternal Non-Hispanic white vs. others | 1.3 | 1.0 | 1.8 | 0.08 |
Maternal above vs. less than high school education | 0.9 | 0.6 | 1.3 | 0.62 |
Maternal not married vs. married | 1.2 | 0.9 | 1.8 | 0.24 |
Prenatal tobacco use: Yes vs. No | 1.5 | 0.9 | 2.3 | 0.09 |
Prenatal alcohol use: Yes vs. No | 0.6 | 0.3 | 1.2 | 0.18 |
Prenatal marijuana use: Yes vs. No | 0.9 | 0.5 | 1.7 | 0.72 |
Pre-pregnancy BMI, kg/m2 | 1.0 | 1.0 | 1.0 | 0.76 |
MCHAT ~ SSRI, N=761 | ||||
Prenatal SSRI use Yes vs. No | 1.0 | 0.2 | 5.8 | 0.99 |
Age at visit for the outcome, yrs | 0.8 | 0.5 | 1.2 | 0.26 |
Child’s sex: Female vs. Male | 0.5 | 0.3 | 0.8 | 0.01 |
Prenatal depression diagnosis: Yes vs. No | 0.8 | 0.3 | 1.9 | 0.56 |
Maternal age at delivery, yrs | 1.0 | 1.0 | 1.1 | 0.12 |
Maternal Non-Hispanic white vs. others | 0.8 | 0.5 | 1.4 | 0.50 |
Maternal above vs. less than high school education | 0.4 | 0.2 | 0.6 | <0.001 |
Maternal not married vs. married | 1.9 | 1.2 | 3.2 | 0.01 |
Prenatal tobacco use: Yes vs. No | 1.7 | 0.9 | 3.2 | 0.08 |
Prenatal alcohol use: Yes vs. No | 1.1 | 0.4 | 3.4 | 0.83 |
Prenatal marijuana use: Yes vs. No | 0.7 | 0.3 | 1.5 | 0.33 |
Pre-pregnancy BMI, kg/m2 | 1.0 | 1.0 | 1.0 | 0.28 |
ASD, autism spectrum disorder; BMI, body mass index; CBCL, Child Behavior Checklist; LCL, lower confidence limit; M-CHAT-R/F, Modified Checklist for Autism in Toddlers-Revised with Follow-up; OR, odds ratio; SRS-2, Social Responsiveness Scale, Second Edition; SSRI, selective serotonin reuptake inhibitors; UCL, upper confidence limit
To check whether missingness on our maternal prenatal depression variable was related to ASD outcomes, univariable mixed logistic regression models (adjusting for the cohort effect) were performed, including maternal prenatal depression as a dependent variable and each ASD outcome as a predictor. Results showed that missingness on maternal prenatal depression was unrelated to all ASD outcomes with the exception of ASD diagnoses. Cases with missing data for prenatal depression, compared to those without missing data, had significantly higher odds of receiving ASD diagnoses (p=.0032). Since ASD outcomes were included in the imputation models, this difference should be addressed through our statistical procedures.
Sensitivity analyses were performed to test whether imputation strategies, cohort study design (case control for ASD versus community cohort), self-report versus medical record source of prenatal medication data, or timing of maternal self-report prenatal medication use substantially impacted study findings. Results from sensitivity analyses confirmed the primary findings in all cases (see supplementary tables).
Discussion
The current study expands upon the previous literature by utilizing data from the ECHO-wide Cohort, a national cohort of mother-child pairs, to examine the association between prenatal antidepressant/SSRI exposures and both binary (diagnostic) and continuous (behavior and symptom scales) ASD-related outcomes.
One of the methodological constraints in observational studies of prenatal antidepressant exposures and child health outcomes is confounding by indication. Specifically, women who are prescribed and/or continue the use of antidepressants in pregnancy are more likely to suffer from more severe depression. These women may also systematically differ from others in terms of demographic or other health characteristics that might confound the association between prenatal antidepressant exposures and child ASD-related behaviors. In the current study, statistical control for these potential confounds rendered initial univariate associations between prenatal exposure and child ASD-related behavior associations statistically non-significant. Our study adds to the growing consensus in the field that prenatal antidepressant exposure is not causally associated with ASD, as evidenced by results from well-controlled observational studies, as well as sibling control designs that account for maternal depression and/or genetic risk (Ames et al., 2021; Vega et al., 2020).
The current study examined ASD outcomes as measured by both bivariate/diagnostic indicators and continuous measures of ASD traits. Importantly, the continuous maternal report measures capture behaviors that are often associated with ASD but are not necessarily unique to the disorder. Recent studies have found associations between prenatal antidepressant exposure and child neurodevelopmental problems outside of the realm of ASD, including affective disorders (Rommel et al., 2020), attention deficit hyperactivity disorder (Clements et al., 2015), and pragmatic language deficits (Smearman et al., 2020). It is possible that symptomatology overlap between ASD and other neurodevelopmental disorders has led to noted associations between prenatal antidepressant exposures and ASD outcomes in previous research.
We found that maternal race was associated with antidepressant use during pregnancy in the ECHO cohort, with medication rates being higher for White participants. This finding is consistent with survey reports of racial disparities in mental health treatment for pregnant women in the U.S. (Salameh et al., 2019), and it may also reflect racial differences in preference for treatment with pharmacotherapy in the perinatal period (Goodman et al., 2013). The significant associations noted between marital status and ASD-related outcomes are consistent with previous studies demonstrating higher divorce rates in families with children with ASD (Hartley et al, 2010), and likely reflect the financial and emotional stressors associated with raising a child with a disability (Karst & Van Hecke, 2012).
We examined whether child biological sex was an effect modifier in the association between prenatal antidepressant exposure and child ASD-related outcomes. Contrary to what has been seen in some preclinical (Leussis et al., 2021) and human (Harrington et al., 2014) studies, and contrary to our hypothesis, biological sex did not interact with prenatal antidepressant or SSRI exposures to predict ASD-related outcomes in this multi-cohort analysis. It is possible that other untested risk or protective factors might significantly influence the prenatal medication exposure and child neurodevelopmental outcome associations. As larger cohort studies and meta-analyses are becoming more common in this research area, tests of other potential effect modifiers might be useful to inform clinical decision-making strategies (e.g., discontinuation of antidepressant use) for obstetricians and pregnant women in their care.
The results of our study were relatively consistent for any antidepressant use versus the subset of SSRIs. We did not have sufficient power to assess the potential impact of specific medications or the timing of exposures during pregnancy. More fine-grained analyses of these indicators would be informative for both theoretical and clinical purposes. If deleterious impacts of prenatal antidepressant use are medication-specific or trimester-specific, analyses that combine all antidepressant use in pregnancy into a single predictor may mask these differences.
When left untreated, maternal prenatal depression has been associated with a number of adverse child outcomes including birth complications and higher rates of psychopathology (Barker et al., 2011). Treatment decisions for pregnant women are multifaceted and require weighing a number of differential risks. Findings from the current study of a large, multi-cohort population suggest that prenatal antidepressant exposure does not increase the risk for ASD and autism-related traits, once maternal prenatal depression and other demographic confounders are controlled. These results suggest that the discontinuation of antidepressant use in pregnancy to decrease risk for ASD may not be warranted, particularly if it is not clinically indicated by maternal depression status or history.
Supplementary Material
Acknowledgments
The authors wish to thank our ECHO colleagues; the medical, nursing, and program staff; and the children and families participating in the ECHO cohorts. We also acknowledge the contribution of the following ECHO program collaborators:
ECHO Components—Coordinating Center: Duke Clinical Research Institute, Durham, North Carolina: Smith PB, Newby KL; Data Analysis Center: Johns Hopkins University Bloomberg School of Public Health, Baltimore, Maryland: Jacobson LP; Research Triangle Institute, Durham, North Carolina: Parker CB; Person-Reported Outcomes Core: Northwestern University, Evanston, Illinois: Gershon R, Cella D; ECHO Awardees and Cohorts: Albert Einstein College of Medicine, Bronx, New York: Aschner J; Icahn School of Medicine at Mount Sinai, New York, NY: Teitelbaum SL; Stroustrup A; Cohen Children’s Medical Center, Northwell Health: Stroustrup A; Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio: Merhar S; Children’s Hospital and Clinic Minnesota, Minneapolis, MN: Lampland A; University of Buffalo, Buffalo, NY: Reynolds A; University of Florida, College of Medicine, Jacksonville, FL: Hudak M; University of Rochester Medical Center, Rochester, NY: Pryhuber G; Vanderbilt Children’s Hospital, Nashville, TN: Moore P; Wake Forest University School of Medicine, Winston Salem, NC: Washburn L; Memorial Hospital of Rhode Island, Providence RI: Deoni S; New York State Psychiatric Institute, New York, NY: Duarte C; University of Puerto Rico, San Jaun, PR: Canino G; University of Wisconsin, Madison WI: Gern J; University of California Davis Mind Institute, Sacramento, CA: Hertz-Picciotto I; University of Washington, Department of Environmental and Occupational Health Sciences, Seattle, WA: Karr C; Children’s Mercy, Kansas City, MO: Carter B; Emory University, Atlanta, GA: Marsit C; Helen DeVos Children’s Hospital, Grand Rapids, MI: Pastyrnak S; Kapiolani Medical Center for Women and Children, Providence, RI: Neal C; Los Angeles Biomedical Research Institute at Harbour-UCLA Medical Center, Los Angeles CA: Smith L; Wake Forest University School of Medicine, Winston Salem, NC: Helderman J; Pennsylvania State University, University Park, PA: Lyall K; Johns Hopkins Bloomberg School of Public Health Kennedy Krieger Institute, Baltimore, MD: Landa R; University of Pittsburgh Medical Center, Magee Women’s Hospital, Pittsburgh, PA: Simhan H; Harvard Pilgrim Health Care Institute, Boston, MA: Oken E; Michigan State University, East Lansing, MI: Kerver J; Henry Ford Health System, Detroit, MI: Barone, C; Michigan Department of Health and Human Services, Lansing, MI: McKane, P; University of Michigan, Ann Arbor, MI: Elliott, M; University of Illinois, Beckman Institute, Urbana, IL: Schantz S; New York School of Medicine, New York, NY: Trasande L; Icahn School of Medicine at Mount Sinai, New York, NY: Wright R; Boston Children’s Hospital, Boston MA: Bosquet-Enlow M; University of California, San Francisco, San Francisco CA: Bush N; University of Minnesota, Minneapolis, MN: Nguyen R; University of Rochester Medical Center: Rochester, NY: Barrett E
Funding
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Research reported in this publication was supported by the Environmental influences on Child Health Outcomes (ECHO) program, Office of the Director, National Institutes of Health, under Award Numbers U2COD023375 (Coordinating Center), U24OD023382 (Data Analysis Center), U24OD023319 (PRO Core), UH3OD023318 (Dunlop/Brennan), UH3OD023347 (Lester), UH3OD023279 (Elliott), UH3OD023313 (Deoni), UH3OD023286 (Oken), UH3OD023320 (Aschner), UH3OD023365 (Hertz-Picciotto), UH3OD023271 (Karr), UH3OD023272 (Schantz), UH3OD023285 (Kerver), UH3OD023249 (Stanford), UH3OD023337 (Wright), UH3OD023282 (Gern), UH3OD023342 (Lyall), UH3OD023305 (Trasande), UH3OD023328 (Duarte), UH3OD023349 (O’Connor), UH3OD023244 (Hipwell).
Footnotes
Conflicts of Interest
None noted
Ethics Approval
Properly constituted Institutional Review Boards – either the ECHO single IRB or the ECHO cohort’s local IRB – are accountable for compliance with regulatory requirements for the ECHO-wide Cohort Data Collection Protocol at participating cohort sites. Governing IRBs review ECHO protocols and all informed consent/assent forms, HIPAA authorization forms, recruitment materials, and other relevant information prior to the initiation of any ECHO-wide Cohort Data Collection Protocol-related procedures or activities. ECHO Cohort Investigators (or their designated study personnel) obtain written informed consent or parent’s / guardian’s permission along with child assent as appropriate, for ECHO-wide Cohort Data Collection Protocol participation and for participation in their specific cohorts. The work of the ECHO Data Analysis Center is approved through the Johns Hopkins Bloomberg School of Public Health Institutional Review Board.
References
- Achenbach T (2001). Child Behavior Checklist for Ages 6–18. ASEBA, University of Vermont. [Google Scholar]
- Achenbach TM (2014). DSM-Oriented guide for the Achenbach System of Empirically Based Assessment (ASEBA). Center for Children, Youth and Families, The University of Vermont. [Google Scholar]
- Achenbach T, & Rescorla L (2000). Child Behavior Checklist for Ages 1 1/2–5 ASEBA, University of Vermont. [Google Scholar]
- Ames JL, Ladd-Acosta C, Fallin MD, Qian Y, Schieve LA, DiGuiseppi C, Lee LC, Kasten EP, Zhou G, Pinto-Martin J, Howerton EM, Eaton CL, & Croen LA (2021). Maternal psychiatric conditions, treatment with selective serotonin reuptake inhibitors, and neurodevelopmental disorders. Biological Psychiatry, 90(4), 253–262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Andrade C (2017a). Antidepressant exposure during pregnancy and risk of autism in the offspring, 1: Meta-review of meta-analyses. The Journal of Clinical Psychiatry, 78(8), 905. [DOI] [PubMed] [Google Scholar]
- Andrade C (2017b). Antidepressant exposure during pregnancy and risk of autism in the offspring, 2: Do the new studies add anything new? The Journal of Clinical Psychiatry, 78(8), e1052–e1056. [DOI] [PubMed] [Google Scholar]
- Austin MP (2006). To treat or not to treat: maternal depression, SSRI use in pregnancy and adverse neonatal effects. Psychological Medicine, 36(12), 1663–1670. [DOI] [PubMed] [Google Scholar]
- Barker ED, Jaffee SR, Uher R, & Maughan B (2011). The contribution of prenatal and postnatal maternal anxiety and depression to child maladjustment. Depression and Anxiety, 28(8), 696–702. [DOI] [PubMed] [Google Scholar]
- Bérard A, Boukhris T, & Sheehy O (2016). Selective serotonin reuptake inhibitors and autism: Additional data on the Quebec Pregnancy/Birth Cohort. American Journal of Obstetrics and Gynecology, 215(6), 803–805. [DOI] [PubMed] [Google Scholar]
- Biederman J, Petty CR, Fried R, Wozniak J, Micco JA, Henin A, … Faraone SV (2010). Child behavior checklist clinical scales discriminate referred youth with autism spectrum disorder: A preliminary study. Journal of Developmental & Behavioral Pediatrics, 31(6), 485–490. [DOI] [PubMed] [Google Scholar]
- Brandlistuen RE, Ystrom E, Eberhard-Gran M, Nulman I, Koren G, & Nordeng H (2015). Behavioural effects of fetal antidepressant exposure in a Norwegian cohort of discordant siblings. International Journal of Epidemiology, 44(4), 1397–1407. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brown HK, Ray JG, Wilton AS, Lunsky Y, Gomes T, & Vigod SN (2017). Association between serotonergic antidepressant use during pregnancy and autism spectrum disorder in children. JAMA, 317(15), 1544–1552. [DOI] [PubMed] [Google Scholar]
- Castro VM, Kong SW, Clements CC, Brady R, Kaimal AJ, Doyle AE, Robinson EB, Churchill SE, Kohane IS, & Perlis RH (2016). Absence of evidence for increase in risk for autism or attention-deficit hyperactivity disorder following antidepressant exposure during pregnancy: A replication study. Translational Psychiatry, 6(1), e708. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chiarotti F, & Venerosi A (2020). Epidemiology of autism spectrum disorders: A review of worldwide prevalence estimates since 2014. Brain Sciences, 10(5), 274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Clements CC, Castro VM, Blumenthal SR, Rosenfield HR, Murphy SN, Fava M, Erb JL, Churchill SE, Kaimal AJ, Doyle AE, Robinson EB, Smoller JW, Kohane S & Perlis RH (2015). Prenatal antidepressant exposure is associated with risk for attention-deficit hyperactivity disorder but not autism spectrum disorder in a large health system. Molecular Psychiatry, 20(6), 727–734. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Constantino JN, & Gruber CP (2012). Social Responsiveness Scale, Second Edition (SRS-2). Torrance, CA: Western Psychological Services. [Google Scholar]
- Croen LA, Grether JK, Yoshida CK, Odouli R, & Hendrick V (2011). Antidepressant use during pregnancy and childhood autism spectrum disorders. Archives of General Psychiatry, 68(11), 1104–1112. [DOI] [PubMed] [Google Scholar]
- Gemmel M, Bögi E, Ragan C, Hazlett M, Dubovicky M, van den Hove DL, Oberlander TF, Charlier TD, & Pawluski JL (2018). Perinatal selective serotonin reuptake inhibitor medication (SSRI) effects on social behaviors, neurodevelopment and the epigenome. Neuroscience & Biobehavioral Reviews, 85, 102–116. [DOI] [PubMed] [Google Scholar]
- Gentile S (2015). Prenatal antidepressant exposure and the risk of autism spectrum disorders in children. Are we looking at the fall of Gods? Journal of Affective Disorders, 182, 132–137. [DOI] [PubMed] [Google Scholar]
- Gillman MW, & Blaisdell CJ (2018). Environmental influences on Child Health Outcomes, a research program of the National Institutes of Health. Current Opinions in Pediatrics, 30(2):260–262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Goodman SH, Dimidjian S, & Williams KG (2013). Pregnant African American women’s attitudes toward perinatal depression prevention. Cultural Diversity and Ethnic Minority Psychology, 19(1), 50. [DOI] [PubMed] [Google Scholar]
- Harrington RA, Lee L-C, Crum RM, Zimmerman AW, & Hertz-Picciotto I (2014). Prenatal SSRI use and offspring with autism spectrum disorder or developmental delay. Pediatrics, 133(5), e1241–e1248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hartley SL, Barker ET, Seltzer MM, Floyd F, Greenberg J, Orsmond G, & Bolt D (2010). The relative risk and timing of divorce in families of children with an autism spectrum disorder. Journal of Family Psychology, 24(4), 449. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Healy D, Le Noury J, & Mangin D (2016). Links between serotonin reuptake inhibition during pregnancy and neurodevelopmental delay/spectrum disorders: A systematic review of epidemiological and physiological evidence. International Journal of Risk & Safety in Medicine, 28(3), 125–141. [DOI] [PubMed] [Google Scholar]
- Houwing DJ, Heijkoop R, Olivier JD, & Snoeren EM (2019). Perinatal fluoxetine exposure changes social and stress-coping behavior in adult rats housed in a seminatural environment. Neuropharmacology, 151, 84–97. [DOI] [PubMed] [Google Scholar]
- Huybrechts KF, Palmsten K, Mogun H, Kowal M, Avorn J, Setoguchi-Iwata S, & Hernández-Díaz S (2013). National trends in antidepressant medication treatment among publicly insured pregnant women. General Hospital Psychiatry, 35(3), 265–271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Karst JS, & Van Hecke AV (2012). Parent and family impact of autism spectrum disorders: A review and proposed model for intervention evaluation. Clinical child and family psychology review, 15(3), 247–277. [DOI] [PubMed] [Google Scholar]
- Kobayashi T, Matsuyama T, Takeuchi M, & Ito S (2016). Autism spectrum disorder and prenatal exposure to selective serotonin reuptake inhibitors: A systematic review and meta-analysis. Reproductive Toxicology, 65, 170–178. [DOI] [PubMed] [Google Scholar]
- Leussis MP, Thanos JM, Powers A, Peterson E, Head JP, McGovern NJ, Malarkey FJ, & Drake A (2021). Sex differences in long-term behavioral alterations, especially anxiety, following prenatal fluoxetine exposure in C57BL/6 mice. Pharmacology Biochemistry and Behavior, 211, 173293. [DOI] [PubMed] [Google Scholar]
- Lord C, Risi S, Lambrecht L, Cook EH, Leventhal BL, DiLavore PC, Pickles A, & Rutter M (2000). The autism diagnostic observation schedule-generic: A standard measure of social and communication deficits associated with the spectrum of autism. Journal of Autism and Developmental Disorders, 30(3), 205–223. [PubMed] [Google Scholar]
- Masi A, DeMayo MM, Glozier N, & Guastella AJ (2017). An overview of autism spectrum disorder, heterogeneity and treatment options. Neuroscience Bulletin, 33(2), 183–193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mezzacappa A, Lasica P-A, Gianfagna F, Cazas O, Hardy P, Falissard B, … Gressier F (2017). Risk for autism spectrum disorders according to period of prenatal antidepressant exposure: A systematic review and meta-analysis. JAMA Pediatrics, 171(6), 555–563. [DOI] [PubMed] [Google Scholar]
- Molenaar NM, Lambregtse-van den Berg MP, & Bonsel GJ (2020). Dispensing patterns of selective serotonin reuptake inhibitors before, during and after pregnancy: A 16-year population-based cohort study from the Netherlands. Archives of Women’s Mental Health, 23(1), 71–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rai D, Lee BK, Dalman C, Golding J, Lewis G, & Magnusson C (2013). Parental depression, maternal antidepressant use during pregnancy, and risk of autism spectrum disorders: Population based case-control study. BMJ (Clinical Research Ed.), 346, f2059. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rai D, Lee BK, Dalman C, Newschaffer C, Lewis G, & Magnusson C (2017). Antidepressants during pregnancy and autism in offspring: Population based cohort study. BMJ, 358, j2811. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Robins DL, Casagrande K, Barton M, Chen CMA, Dumont-Mathieu T, & Fein D (2014). Validation of the modified checklist for autism in toddlers, revised with follow-up (M-CHAT-R/F). Pediatrics, 133(1), 37–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rommel AS, Bergink V, Liu X, Munk-Olsen T, & Molenaar NM (2020). Long-term effects of intrauterine exposure to antidepressants on physical, neurodevelopmental, and psychiatric outcomes: a systematic review. The Journal of Clinical Psychiatry, 81(3), 10661. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rubin DB (2004). Multiple Imputation for Nonresponse in Surveys. Vol. 81. New York, NY: John Wiley & Sons Inc. [Google Scholar]
- Salameh TN, Hall LA, Crawford TN, Staten RR, & Hall MT (2019). Racial/ethnic differences in mental health treatment among a national sample of pregnant women with mental health and/or substance use disorders in the United States. Journal of psychosomatic research, 121, 74–80. [DOI] [PubMed] [Google Scholar]
- Skurtveit S, Selmer R, Roth C, Hernandez-Diaz S, & Handal M (2014). Prenatal exposure to antidepressants and language competence at age three: Results from a large population-based pregnancy cohort in Norway. BJOG: An International Journal of Obstetrics & Gynaecology, 121(13), 1621–1631. [DOI] [PubMed] [Google Scholar]
- Smearman EL, Hendrix CL, Winiarski DA, Johnson KC, Smith AK, Ousley OY, Stowe ZN, Newport DJ, & Brennan PA (2020). School-age social behavior and pragmatic language ability in children with prenatal serotonin reuptake inhibitor exposure. Development and Psychopathology, 32(1), 21–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- van Buuren S & Groothuis-Oudshoorn K (2011). mice: Multivariate Imputation by Chained Equations in R. Journal of Statistical Software, 45(3), 1–67. [Google Scholar]
- Vega ML, Newport GC, Bozhdaraj D, Saltz SB, Nemeroff CB, & Newport DJ (2020). Implementation of advanced methods for reproductive pharmacovigilance in autism: A meta-analysis of the effects of prenatal antidepressant exposure. American Journal of Psychiatry, 177(6), 506–517. [DOI] [PubMed] [Google Scholar]
- Viktorin A, Uher R, Reichenberg A, Levine SZ, & Sandin S (2017). Autism risk following antidepressant medication during pregnancy. Psychological Medicine, 47(16), 2787–2796. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.