© 2017 American Medical Association. All rights reserved.
Supplementary Online Content
Viktorin A, Uher R, Kolevzon A, Reichenberg A, Levine SZ, Sandin S. Association of
antidepressant medication use during pregnancy with intellectual disability in offspring. JAMA
Psychiatry. Published online July 12, 2017. doi:10.1001/jamapsychiatry.2017.1727
eTable 1. List of Medications Considered in the Study
eTable 2. Psychotropic Drugs During Pregnancy
eTable 3. Detailed Information About Parental Education Levels
eTable 4. Psychiatric Diagnosis Codes Included in the Study
eTable 5. Clinical Subsample Subject Characteristics
eFigure 1. Medication Exposure Definition
eFigure 2. Kaplan-Meier Failure Estimate of Intellectual Disability
eFigure 3. Relative Risks of Intellectual Disability in Children of Mothers With a Single
Dispensation of Any Antidepressant During Pregnancy
eFigure 4. Relative Risks of Intellectual Disability in Children of Mothers With Any Number of
Dispensations of Any Type Antidepressant During Pregnancy
eFigure 5. Plotted Schoenfeldt Residuals for Each Covariate With 4 Degrees of Freedom
eFigure 6. Relative Risks of Mild to Moderate Intellectual Disability
eFigure 7. Relative Risks of Severe Intellectual Disability
eFigure 8. Relative Risks of Intellectual Disability Without Comorbid Autism Spectrum Disorder
eFigure 9. Results From Bootstrapped Analyses
eFigure 10. Sex-Specific and Sex-Combined Relative Risks of Intellectual Disability
eFigure 11. Relative Risks of Intellectual Disability in Children of Mothers Treated With
Antidepressants During Pregnancy, Including Gestational Age and Birth Weight
eFigure 12. Trimester-Specific Relative Risks of Intellectual Disability
eFigure 13. Duration-Dose-Response Effect on the Relative Risk of Intellectual Disability
eFigure 14. Effects of Potential Misclassification in Unexposed Children
eFigure 15. Relative Risks of Intellectual Disability Stratified on Maternal Polypharmacy
eFigure 16. Percentage of Mothers With Medication During Pregnancy and a Psychiatric Diagnosis
Before Child Birth
eFigure 17. Relative risks of Intellectual Disability Due to Exposure to Any Antidepressant During
Pregnancy, Where All Antidepressant Treated Mothers Were Assumed to Have Depression
This supplementary material has been provided by the authors to give readers additional information about their work.
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eTable 1. List of Medications Considered in the Study*
Abbreviations: ATC, Anatomical Therapeutic Chemical Classification System.
* List of medications with drug names and ATC codes considered in the study.
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eTable 2. Psychotropic Drugs During Pregnancy*
Abbreviations: ATC, Anatomical Therapeutic Chemical Classification System.
* The sample consists of the mothers with at least two dispensations overlapping pregnancy with
either antidepressants (N=3,982) or other psychotropic drugs (N=1,626). The table presents the
number of mothers with at least one dispensation overlapping pregnancy with the specific
psychotropic drugs considered in the study (see eTable 1). Any drug listed in eTable 1 not in this
table list was not observed in the sample.
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eTable 3. Detailed Information About Parental Education Levels
Full Population sample
Antidepressant medication during
pregnancy
Characteristics, N (%) Unexposeda Exposed
b Uncertain
c
Maternal education level
Compulsory school shorter than 9 years 8,179 (4.7) 107 (2.7) 73 (3.1)
Compulsory school 9 years 13,587 (7.9) 620 (15.6) 382 (16.1)
Upper secondary school up to 2 years 19,079 (11.1) 673 (16.9) 387 (16.3)
Upper secondary school 3 years 49,860 (28.9) 1,119
(28.1)
671 (28.2)
Post-secondary education less than 3 years 21,037 (12.2) 486 (12.2) 301 (12.7)
University 3 years 59,315 (34.4) 952 (23.9) 552 (23.2)
Doctoral education 1,589 (0.9) 25 (0.6) 13 (0.6)
Paternal education level
Compulsory school shorter than 9 years 7,732 (4.5) 123 (3.1) 103 (4.3)
Compulsory school 9 years 16,333 (9.5) 506 (12.7) 282 (11.9)
Upper secondary school up to 2 years 35,168 (20.4) 1,013
(25.4)
582 (24.5)
Upper secondary school 3 years 46,771 (27.1) 1,101
(27.7)
680 (28.6)
Post-secondary education less than 3 years 24,063 (13.9) 500 (12.6) 280 (11.8)
University 3 years 40,007 (23.2) 706 (17.7) 431 (18.1)
Doctoral education 2,572 (1.5) 33 (0.8) 21 (0.9)
Abbreviations: N, Number. SD, standard deviation. NA, Not applicable. a Children of mothers with no antidepressant dispensation with a medication period overlapping
pregnancy. b Children of mothers with at least two antidepressant dispensations with medication periods
overlapping pregnancy (eFigure 1 example A1 and example A2). c Children of mothers with a single antidepressant dispensation with a medication period
overlapping pregnancy.
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eTable 4. Psychiatric Diagnosis Codes included in the Study*
Abbreviations: ICD, International Classification of Diseases. ADHD, attention deficit hyperactivity
disorder. ASD, autism spectrum disorder. NA, not available.
*Note: * at the end of a diagnosis code indicates all sub-categories.
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eTable 5. Clinical Subsample Subject Characteristics
Clinical sub-samplea
Antidepressant medication during pregnancy
Characteristics, No (%) Unexposedb Exposed
c Uncertain
d
Number of offspring 4,976 (62.0) 2,372 (29.6) 673 (8.4)
Offspring with intellectual disability 36 (0.7) 27 (1.1) 6 (0.9)
Offspring with severe intellectual disability 5 (0.1) 3 (0.1) 0 (0.0)
Offspring with intellectual disability without
autism spectrum disorder
28 (0.6) 18 (0.8) 6 (0.9)
Maternal use of psychotropic medication other
than antidepressants during pregnancy
229 (4.6) 448 (18.9) 71 (10.6)
Paternal psychotropic medication during
pregnancy
348 (7.0) 286 (12.1) 64 (9.5)
Birth year
2006 1,912 (38.4) 937 (39.5) 254 (37.7)
2007 3,064 (61.6) 1,435 (60.5) 419 (62.3)
Maternal psychiatric diagnosis 4,976 (100) 2,372 (100) 673 (100)
Paternal psychiatric diagnosis 1,192 (24.0) 590 (24.9) 167 (24.8)
Maternal age (years) at delivery
© 2017 American Medical Association. All rights reserved.
c Children of mothers with at least two antidepressant dispensations with medication periods
overlapping pregnancy (eFigure 1 example A1 and example A2). d Children of mothers with a single antidepressant dispensation with a medication period
overlapping pregnancy. e Details of mothers’ and fathers’ education within 7 levels is provided in Supplement eTable 3.
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 1. Medication Exposure Definition*
Abbreviations: Disp., dispensation.
* Exposure to medication during pregnancy was established in both parents using dispensation
dates, number of pills dispensed, and the DDD - the assumed average maintenance dose per day for
a drug used for its main indication in adults, provided by the World Health Organization (WHO).
For antidepressants, medication was assumed to cover the time from the dispensation date until an
end date based on number of pills dispensed, assuming one pill was consumed per day. For any
other psychotropic drug, the medication period was calculated based on the dispensation dates and
the number of pills dispensed divided by the medication-specific DDD. To be classified as using a
medication during pregnancy, a continuous pattern of dispensations was required according to
either a) at least two dispensations during pregnancy (example A in Figure), or b) a single
dispensation (one-time dispensation) during pregnancy, and a second previous dispensation
covering the pregnancy (example B in Figure). However, offspring to mothers with a one-time
dispensation was not included in the main results but analyzed and presented separately (N=2,370;
eFigure 3), and in combination with mothers having at least two dispensations with medication
periods in pregnancy (N=6,361; eFigure 4).
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 2. Kaplan-Meier Failure Estimate of Intellectual Disability*
* Plotted Kaplan-Meier failure estimate of intellectual disability in the children of the cohort,
showing a gradual increase in number of children receiving their first diagnosis of intellectual
disability over follow-up. The children are followed from birth in 2006 and 2007 and throughout
2014 when aged 7 to 8.
0.0
%0
.5%
Pre
va
lence
0 1 2 3 4 5 6 7 8
Analysis time (years)
Kaplan-Meier failure estimate
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 3. Relative Risks of Intellectual Disability in Children of Mothers With a Single
Dispensation of Any Antidepressant During Pregnancy*
Abbreviations: N, number of children. CI, confidence interval.
* The full sample consists of 179,007 children born during 2006 and 2007, of which 873 had been
diagnosed with intellectual disability, and 3,982 had were born to a mother with antidepressant
treatment during pregnancy. The figure presents relative risks of intellectual disability and two-
sided 95% confidence intervals in children of mothers with a single dispensation of an
antidepressant drug overlapping the pregnancy (N=2,379), compared with unexposed children. a A clinically relevant sub-sample consisting of 8,021 children, of which 69 had been diagnosed
with intellectual disability, and 2,372 had were born to a mother with antidepressant treatment
during pregnancy. All mothers, both medicated and non-medicated, had at least one diagnosis of
depression or an anxiety disorder prior child birth (eTable 4). Thereby, the offspring of mothers
with medication during pregnancy is contrasted with offspring of mothers that may share similar
underlying factors. b Analyses not adjusted for covariates.
c Analyses adjusted for birthdate, maternal and paternal age, the father's psychotropic medication
overlapping the pregnancy, and maternal and paternal education level at child birth. d
Analyses adjusted for the factors listed in c, and for any maternal diagnoses of depression prior
child birth (yes/no) (see eTable 4 for specific diagnosis codes). e Analyses adjusted for the factors listed in
c, and for maternal and paternal diagnoses prior child
birth of specific psychiatric disorder subgroups in either the mother and/or father's life time
(yes/no), including depression, anxiety disorders, substance use disorder, bipolar disorder,
compulsive disorder, attention deficit hyperactive disorder (ADHD), autism spectrum disorder,
intellectual disability, schizophrenia, and 'other psychiatric diagnosis' (see eTable 4 for specific
diagnosis codes).
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 4. Relative Risks of Intellectual Disability in Children of Mothers With Any Number of
Dispensations of Any Type Antidepressant During Pregnancy*
Abbreviations: N, number of children. CI, confidence interval.
* The full sample consists of 179,007 children born during 2006 and 2007, of which 873 had been
diagnosed with intellectual disability, and 3,982 had were born to a mother with antidepressant
treatment during pregnancy. The figure presents relative risks of intellectual disability and two-
sided 95% confidence intervals in children of mothers with any number of dispensations of an
antidepressant drug overlapping the pregnancy (N=6,361), compared with unexposed children. a A clinically relevant sub-sample consisting of 8,021 children, of which 69 had been diagnosed
with intellectual disability, and 3,045 had were born to a mother with any number of antidepressant
dispensations during pregnancy. All mothers, both medicated and non-medicated, had at least one
diagnosis of depression or an anxiety disorder prior child birth (eTable 4). Thereby, the offspring of
mothers with medication during pregnancy is contrasted with offspring of mothers that may share
similar underlying factors. b Analyses not adjusted for covariates.
c Analyses adjusted for birthdate, maternal and paternal age, the father's psychotropic medication
overlapping the pregnancy, and maternal and paternal education level at child birth. d
Analyses adjusted for the factors listed in c, and for any maternal diagnoses of depression prior
child birth (yes/no) (see eTable 4 for specific diagnosis codes). e Analyses adjusted for the factors listed in
c, and for maternal and paternal diagnoses prior child
birth of specific psychiatric disorder subgroups in either the mother and/or father's life time
(yes/no), including depression, anxiety disorders, substance use disorder, bipolar disorder,
compulsive disorder, attention deficit hyperactive disorder (ADHD), autism spectrum disorder,
intellectual disability, schizophrenia, and 'other psychiatric diagnosis' (see eTable 4 for specific
diagnosis codes).
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 5. Plotted Schoenfeldt Residuals for Each Covariate With 4 Degrees of Freedom
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 6. Relative Risks of Mild to Moderate Intellectual Disability*
Abbreviations: N, number of births to treated mothers. SSRI, selective serotonin re-uptake inhibitor.
CI, confidence interval.
* Relative risks of mild to moderate intellectual disability and two-sided 95% confidence intervals
in children to mothers with antidepressant or other psychotropic medication during pregnancy
compared with children to mothers without any psychotropic medication during pregnancy. The
sample consists of 179,007 children born during 2006 and 2007, of which 767 had been diagnosed
with mild to moderate intellectual disability. a Analyses not adjusted for covariates.
b Analyses adjusted for birthdate, maternal and paternal age, the father's psychotropic medication
overlapping the pregnancy, and maternal and paternal education level at child birth. c Analyses adjusted for the factors listed in
d, and for any maternal diagnosis of depression prior
child birth (yes/no) (see eTable 4 for specific diagnosis codes). d Analyses adjusted for the factors listed in
d, and maternal and paternal diagnoses prior child birth
of specific mental illness subgroups (yes/no), including depression, anxiety disorders, substance use
disorder, bipolar disorder, compulsive disorder, attention deficit hyperactive disorder (ADHD),
autism spectrum disorder, intellectual disability, schizophrenia, and 'other psychiatric diagnosis'
(see eTable 4 for specific diagnosis codes).
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 7. Relative Risks of Severe Intellectual Disability*
Abbreviations: N, number of births to treated mothers. SSRI, selective serotonin re-uptake inhibitor.
CI, confidence interval.
* Relative risks of severe intellectual disability and two-sided 95% confidence intervals in children
to mothers with antidepressant or other psychotropic medication during pregnancy compared with
children to mothers without any psychotropic medication during pregnancy. The sample consists of
179,007 children born during 2006 and 2007, of which 106 had been diagnosed with severe
intellectual disability. a Analyses not adjusted for covariates.
b Analyses adjusted for birthdate, maternal and paternal age, the father's psychotropic medication
overlapping the pregnancy, and maternal and paternal education level at child birth. c Analyses adjusted for the factors listed in
d, and for any maternal diagnosis of depression prior
child birth (yes/no) (see eTable 4 for specific diagnosis codes). d Analyses adjusted for the factors listed in
d, and maternal and paternal diagnoses prior child birth
of specific mental illness subgroups (yes/no), including depression, anxiety disorders, substance use
disorder, bipolar disorder, compulsive disorder, attention deficit hyperactive disorder (ADHD),
autism spectrum disorder, intellectual disability, schizophrenia, and 'other psychiatric diagnosis'
(see eTable 4 for specific diagnosis codes).
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 8. Relative Risks of Intellectual Disability Without Comorbid Autism Spectrum Disorder*
Abbreviations: N, number of births to treated mothers. SSRI, selective serotonin re-uptake inhibitor.
CI, confidence interval.
* Relative risks of intellectual disability without comorbid autism spectrum disorder and two-sided
95% confidence intervals in children to mothers with antidepressant or other psychotropic
medication during pregnancy compared with children to mothers without any psychotropic
medication during pregnancy. The sample consists of 179,007 children born during 2006 and 2007,
of which 513 had been diagnosed with intellectual disability without comorbid autism spectrum
disorder. a Analyses not adjusted for covariates.
b Analyses adjusted for birthdate, maternal and paternal age, the father's psychotropic medication
overlapping the pregnancy, and maternal and paternal education level at child birth. c Analyses adjusted for the factors listed in
d, and for any maternal diagnosis of depression prior
child birth (yes/no) (see eTable 4 for specific diagnosis codes). d Analyses adjusted for the factors listed in
d, and maternal and paternal diagnoses prior child birth
of specific mental illness subgroups (yes/no), including depression, anxiety disorders, substance use
disorder, bipolar disorder, compulsive disorder, attention deficit hyperactive disorder (ADHD),
autism spectrum disorder, intellectual disability, schizophrenia, and 'other psychiatric diagnosis'
(see eTable 4 for specific diagnosis codes).
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 9. Results From Bootstrapped Analyses*
Abbreviations: N, number of births to treated mothers. SSRI, selective serotonin re-uptake inhibitor.
CI, confidence interval.
* Relative risks of intellectual disability and two-sided 95% confidence intervals in children to
mothers with antidepressant or other psychotropic medication during pregnancy compared with
children to mothers without any psychotropic medication during pregnancy. The sample consists of
179,007 children born during 2006 and 2007, of which 873 had been diagnosed with intellectual
disability without comorbid autism spectrum disorder. Bootstrapping was performed with 100
repetitions. a Mothers not using antidepressant, but other psychotropic medication during pregnancy.
b Relative risk and 95% CI for 1) any antidepressant (top in black), 2) SSRI antidepressants (2
nd in
red), 3) other antidepressants (3rd
in green), and 4) other psychotropic drugs (bottom in yellow),
repeated for each level of adjustment. c Analyses not adjusted for covariates.
d Analyses adjusted for birthdate, maternal and paternal age, the father's psychotropic medication
overlapping the pregnancy, and maternal and paternal education level at child birth. e Analyses adjusted for the factors listed in
d, and any maternal diagnosis of depression prior child
birth (yes/no) (see eTable 4 for specific diagnosis codes). f Analyses adjusted for the factors listed in
d, and maternal and paternal diagnoses prior child birth
of specific mental illness subgroups (yes/no), including depression, anxiety disorders, substance use
disorder, bipolar disorder, compulsive disorder, attention deficit hyperactive disorder (ADHD),
autism spectrum disorder, intellectual disability, schizophrenia, and 'other psychiatric diagnosis'
(see eTable 4 for specific diagnosis codes).
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 10. Sex-Specific an Sex-Combined Relative Risks of Intellectual Disability*
Abbreviations: N, number of births to treated mothers. RR, relative risk. CI, confidence interval.
* Sex-combined and sex-specific relative risks of intellectual disability and two-sided 95%
confidence intervals in children to mothers with antidepressant medication during pregnancy
compared with children to mothers without any psychotropic medication during pregnancy. The
sample consists of 179,007 children born during 2006 and 2007, of which 90,090 (51.5%) were
males and 84,935 (48.5%) females. There were 873 children diagnosed with intellectual disability,
of which 545 (62.4% of cases; 0.3% of all children) where male, and 328 (37.6% of cases; 0.2% of
all children) female. a Analyses not adjusted for covariates.
b Analyses adjusted for birthdate, maternal and paternal age, the father's psychotropic medication
overlapping the pregnancy, and maternal and paternal education level at child birth. c Analyses adjusted for the factors listed in
c, and for any maternal diagnosis of depression prior
childbirth (yes/no) (see eTable 4 for specific diagnosis codes). d Analyses adjusted for the factors listed in
c, and for maternal or paternal diagnoses prior child birth
of specific psychiatric disorder subgroups (yes/no), including depression, anxiety disorders,
substance use disorder, bipolar disorder, compulsive disorder, attention deficit hyperactive disorder
(ADHD), autism spectrum disorder, intellectual disability, schizophrenia, and 'other psychiatric
diagnosis' (see eTable 4 for specific diagnosis codes).
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 11. Relative Risks of Intellectual Disability in Children of Mothers Treated With
Antidepressants During Pregnancy, Including Gestational Age and Birth Weight *
Abbreviations: RR, relative risk. CI, confidence interval.
* Relative risks of intellectual disability and two-sided 95% confidence intervals in children to
mothers with antidepressant medication during pregnancy compared with children to mothers
without antidepressant medication during pregnancy. The analyses examine the role of gestational
age and birth weight in the causal pathway.
a Results from Model 4 in Figure 1.
b Model 4 in Figure 1, with additional adjustment for gestational age in days (RR: 0.98; 95% CI
0.98-0.98; p-value < 0.001). All children had gestational age information.
c Model 4 in Figure 1, with additional adjustment for birth weight in grams (RR: 1.00; 95% CI 1.00-
1.00; p-value < 0.001). There were 271 children without birth weight information. In these children,
we used the cohort mean birth weight stratified on offspring sex.
d Model 4 in Figure 1, with additional adjustment for gestational age in days and birth weight in
grams. Gestational age and birth weight were closely correlated in the cohort (Pearson’s correlation
coefficient 0.66; p-value
© 2017 American Medical Association. All rights reserved.
eFigure 12. Trimester-Specific Relative Risks of Intellectual Disability*
Abbreviations: N, number of children. CI, confidence interval. RR, relative risk.
* Analyses of the association between antidepressant exposure at specific trimesters and intellectual
disability in the offspring. The analyses allowed exposure in several trimesters, and correspond to
Model 4 in the main analysis (Figure 1), adjusting for all included covariates.
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 13. Duration-Dose-Response Effect on the Relative Risk of Intellectual Disability*
Abbreviations: N, number of children. CI, confidence interval. RR, relative risk.
* The full sample consists of 179,007 children born during 2006 and 2007, of which 873 had been
diagnosed with intellectual disability, and 3,982 had were born to a mother with antidepressant
treatment during pregnancy. The figure presents relative risks of intellectual disability and two-
sided 95% confidence intervals in children of mothers with at least two dispensations of
antidepressant drugs overlapping the pregnancy, compared with unexposed children. a A clinically relevant sub-sample consisting of 8,021 children, of which 69 had been diagnosed
with intellectual disability, and 2,372 had were born to a mother with antidepressant treatment
during pregnancy. All mothers, both medicated and non-medicated, had at least one diagnosis of
depression or an anxiety disorder prior child birth (eTable 4). Thereby, the offspring of mothers
with medication during pregnancy is contrasted with offspring of mothers that may share similar
underlying factors. b Analyses adjusted for the mean difference of antidepressant exposure duration from the mean,
separately in the exposed and unexposed children. This was done by creating an additional
covariate that reflect the deviation from the mean number of days with antidepressant treatment
during pregnancy. Thus, in exposed children, the mean number of days exposed to antidepressants
was estimated at 156 (Table 1), which was set to 0. Children with fewer days of antidepressant
exposure during pregnancy than 156 would get a negative number reflecting the difference in
duration from the mean, and vice versa for children with longer duration. All children unexposed to
maternal antidepressant treatment during pregnancy would have the value 0. For each additional 30
days (approximately a month) of exposure longer than the mean duration of exposure (156 days),
the RR was estimated at 1.01 (95% CI: 0.89-1.15; P-value: 0.88). c Analyses adjusted for duration of medication exposure (explained in
b), birthdate, maternal and
paternal age, the father's psychotropic medication overlapping the pregnancy, and maternal and
paternal education level at child birth. d
Analyses adjusted for the factors listed in c, and for any maternal diagnosis of depression prior
childbirth (yes/no) (see eTable 4 for specific diagnosis codes). e Analyses adjusted for the factors listed in
c, and for maternal or paternal diagnoses prior child birth
of specific psychiatric disorder subgroups (yes/no), including depression, anxiety disorders,
substance use disorder, bipolar disorder, compulsive disorder, attention deficit hyperactive disorder
(ADHD), autism spectrum disorder, intellectual disability, schizophrenia, and 'other psychiatric
diagnosis' (see eTable 4 for specific diagnosis codes).
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 14. Effects of Potential Misclassification in Unexposed Children*
Abbreviations: N, number of children. RR, relative risk. CI, confidence interval.
* To estimate the potential effect of misclassification in children born to mothers without
antidepressant medication in pregnancy, but that may have medication prior conception, the sample
was restricted to children conceived after January 1 2006, to allow six months follow-up prior
conception (the Swedish Prescribed Drug Register began follow-up at July 1 2005). The full sample
consists of 125,257 children, of which 598 had been diagnosed with intellectual disability, and
2,949 were born to a mother with antidepressant treatment during pregnancy. Antidepressant
unexposed children had to be born to a mother without any antidepressant medication in pregnancy,
however, medication in the six months prior conception was not considered. In a second sample of
123,964 children, antidepressant medication in the six months prior conception was considered and
any children defined as unexposed had to be born to a mother with neither antidepressant
medication during pregnancy, nor in the six months preceding pregnancy. This excluded 1,293
children. a Analyses not adjusted for covariates.
b Analyses adjusted for birthdate, maternal and paternal age, the father's psychotropic medication
overlapping the pregnancy, and maternal and paternal education level at child birth. c Analyses adjusted for the factors listed in
c, and for any maternal diagnosis of depression prior
childbirth (yes/no) (see eTable 4 for specific diagnosis codes). d Analyses adjusted for the factors listed in
c, and for maternal or paternal diagnoses prior child birth
of specific psychiatric disorder subgroups (yes/no), including depression, anxiety disorders,
substance use disorder, bipolar disorder, compulsive disorder, attention deficit hyperactive disorder
(ADHD), autism spectrum disorder, intellectual disability, schizophrenia, and 'other psychiatric
diagnosis' (see eTable 4 for specific diagnosis codes).
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 15. Relative Risks of Intellectual Disability Stratified on Maternal Polypharmacy*
Abbreviations: N, number of children. RR, relative risk. CI, confidence interval.
* The full sample consists of 179,007 children born during 2006 and 2007, of which 873 had been
diagnosed with intellectual disability, and 3,982 had were born to a mother with antidepressant
treatment during pregnancy. The figure presents relative risks of intellectual disability and two-
sided 95% confidence intervals in analyses stratified on maternal polypharmacy, defined as having
any dispensation with a non-antidepressant psychotropic medication with a medication period
overlapping pregnancy. There were 3,059 children of mothers without any dispensation of a non-
antidepressant psychotropic medication in pregnancy, and 923 children of mothers with at least one
dispensation with a non-antidepressant psychotropic medication with a medication period
overlapping pregnancy. a Analyses not adjusted for covariates.
b Analyses adjusted for birthdate, maternal and paternal age, the father's psychotropic medication
overlapping the pregnancy, and maternal and paternal education level at child birth. c Analyses adjusted for the factors listed in
c, and for any maternal diagnosis of depression prior
childbirth (yes/no) (see eTable 4 for specific diagnosis codes). d Analyses adjusted for the factors listed in
c, and for maternal or paternal diagnoses prior child birth
of specific psychiatric disorder subgroups (yes/no), including depression, anxiety disorders,
substance use disorder, bipolar disorder, compulsive disorder, attention deficit hyperactive disorder
(ADHD), autism spectrum disorder, intellectual disability, schizophrenia, and 'other psychiatric
diagnosis' (see eTable 4 for specific diagnosis codes).
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 16. Percentage of Mothers With Medication During Pregnancy and a Psychiatric Diagnosis
Before Child Birth*
Abbreviations: SSRI, selective serotonin re-uptake inhibitor. ADHD, attention-deficit hyperactive
disorder. ASD, autism spectrum disorder.
* Mothers with medication during pregnancy (at least two dispensations with medication periods
overlapping pregnancy, see eFigure 1; N=5,608) with either SSRI antidepressants (N=3,178), non-
SSRI antidepressants (N=804), or non-antidepressant psychotropic medications (N=1,626), and the
percentage of these mothers having a diagnosis in the Swedish Patient Register.
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021
© 2017 American Medical Association. All rights reserved.
eFigure 17. Relative Risks of Intellectual Disability Due to Exposure to Any Antidepressant
During Pregnancy, Where All Antidepressant Treated Mothers Were Assumed to Have
Depression*
Abbreviations: N, number of children. CI, confidence interval. RR, relative risk.
* The full sample consists of 179,007 children born during 2006 and 2007, of which 873 had been
diagnosed with intellectual disability, and 3,982 had were born to a mother with antidepressant
treatment during pregnancy. The figure presents relative risks of intellectual disability and two-
sided 95% confidence intervals in children of mothers with at least two dispensations of
antidepressant drugs overlapping the pregnancy, compared with unexposed children. Among
mothers with antidepressant medication during pregnancy, 25% lack any psychiatric diagnosis in
the Swedish Patient Register. The Swedish Patient Register lack primary care information, and
antidepressant medication is accompanied with a diagnosis in the Swedish healthcare system.
Therefore, it is possible that the lack of a psychiatric diagnosis in 25% of the mothers with
antidepressant medication cause residual confounding. In an attempt to estimate the size of the
potential residual confounding by lack of psychiatric diagnosis from the primary care, in this
analysis, all mothers with antidepressant medication was also assumed having suffered depression.
Model 3 and Model 4 in the full sample shows the estimated relative risks if all antidepressant
treated mothers were also assumed having suffered depression. a A clinically relevant sub-sample consisting of 8,021 children, of which 69 had been diagnosed
with intellectual disability, and 2,372 had were born to a mother with antidepressant treatment
during pregnancy. All mothers, both medicated and non-medicated, had at least one diagnosis of
depression or an anxiety disorder prior child birth (eTable 4). Thereby, the offspring of mothers
with medication during pregnancy is contrasted with offspring of mothers that may share similar
underlying factors. b Analyses not adjusted for covariates.
c Analyses adjusted for birthdate, maternal and paternal age, the father's psychotropic medication
overlapping the pregnancy, and maternal and paternal education level at child birth. d
Analyses adjusted for the factors listed in c, and for any maternal diagnosis of depression prior
child birth (yes/no) (see eTable 4 for specific diagnosis codes). e Analyses adjusted for the factors listed in
c, and for maternal and paternal diagnoses prior child
birth of specific psychiatric disorder subgroups (yes/no), including depression, anxiety disorders,
substance use disorder, bipolar disorder, compulsive disorder, attention deficit hyperactive disorder
(ADHD), autism spectrum disorder, intellectual disability, schizophrenia, and 'other psychiatric
diagnosis' (see eTable 4 for specific diagnosis codes).
Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 04/18/2021