The Diagnostic and Statistical Manual of Mental Disorders, fifth edition1 characterizes autism spectrum disorder (ASD) as persistent deficits in social communication and social interaction across multiple contexts, including social-emotional reciprocity; nonverbal communicative behaviors used for social interaction; restricted, repetitive patterns of behavior, interests, or activities such as stereotyped or repetitive motor movements; inflexible adherence to routines or ritualized patterns of verbal and nonverbal behavior; highly restricted and fixated interests; and hyper- or hyporeactivity to sensory input that become apparent in the early developmental period, resulting in significant functional impairment. ASDs are rated in severity to reflect variations in symptoms from person to person and are often considered an umbrella diagnosis for what was previously referred to as “Asperger’s Syndrome, Autism, and Pervasive Developmental Disorder Not Otherwise Specified.” The rationale for combining the conditions is to reflect their shared genetic etiology despite the heterogeneous phenotypic presentation of the disorder that often resulted in inconsistent diagnoses.2
The most recent data from the Autism and Developmental Disabilities Monitoring Network, an active surveillance system in the United States that provides estimates of the presence of ASDs among 8-year-old children (the peak prevalence age), indicates about 1 in 68 children have been identified with ASDs. This disorder is equally prevalent in racial, ethnic, and socioeconomic groups, but it is almost 5 times more prevalent in boys (1 in 42) than in girls (1 in 189). These findings are notable for the steady increase in prevalence from the year 2000 when the rate was 1 in 150, a growth that is attributed to genetic and environmental factors.3
Risk Factors for Autism Spectrum Disorders
ASDs can be reliably diagnosed by age 2 years, but prevention of this disorder is complex due to the many risk factors noted in numerous studies. Twin studies demonstrate that less than 70% are concordant for the disorder, suggesting a possibility for nonheritable influence on the development of ASDs. Many studies have most consistently shown a higher risk of ASDs in children born to older parents, born prematurely or at a low birth weight, and in children born to families who already have at least one child with ASD.2 Further studies suggest ASDs’ variation in phenotype could be due to multiple risk factors including genetic mutations, deletions, viral infections, and encephalitis, resulting in genetic defects or brain inflammation from an immature blood brain barrier, a defective placenta, or a premature birth.4
Another review of prenatal and perinatal risk factors for autism, per the Diagnostic and Statistical Manual, fourth edition, text revision (DSM-IVTR)5 criteria, found an increased risk of the disorder due to advanced paternal age, maternal age over 33 years, prematurity, and low birth weight in addition to intrapartum hypoxia measured by an Apgar score of less than 7. The advance in maternal age was also associated with obstetric complications resulting in prematurity, hypoxia, and low birth weight.6 Although there is a correlation between all of these conditions and ASDs, they do not consistently demonstrate a link to the disorder.
A comprehensive meta-analysis of 60 perinatal and neonatal risk factors by Gardener et al.7 found that factors associated with autism (now considered a severe form of ASD) were abnormal presentation, umbilical cord complications, fetal distress, birth injury or trauma, multiple birth, maternal hemorrhage, summer birth, low birth weight, small for gestational age, congenital malformation, low 5-minute Apgar score, feeding difficulties, meconium aspiration, neonatal anemia, ABO blood type or Rh incompatibility, and hyperbilirubinemia. Factors not associated with severe ASD risk included anesthesia, assisted vaginal delivery, post-term birth, high birth weight, and head circumference. A prenatal risk factor meta-analysis by the same researchers included advanced maternal and paternal age at birth, maternal gestational bleeding, gestational diabetes, being first born versus third born or later, maternal prenatal medication use, and maternal birth abroad, which suggests the prenatal conditions resulting in or operating with perinatal and neonatal complications that increased the risk for ASDs. The factors with the strongest evidence against a role in severe ASD risk included previous fetal loss and maternal pre-eclampsia, proteinuria, hypertension, and swelling.7
The Serotonin Theory of Autism Spectrum Disorders
Advanced parental age, gestational diabetes, birth order, many obstetric complications resulting in low birth weight, and hypoxia often remain nonmodifiable risk factors associated with the development of ASDs, but the notable finding of the influence of medication has prompted a field of debate because it may be a clinically controllable variable. Particular interest is given toward selective serotonin reuptake inhibitors (SSRIs) and the development of ASDs, due to the serotonin (5-hydroxytryptamine [5-HT]) theory of neurodevelopment. Because SSRI use shows no evidence of gross structural neuroteratogenic effects in infants, they are often considered safe for antenatal use and have been increasingly used for management of maternal depression during pregnancy. Antidepressant use among mothers between 3 months gestation and before the end of their pregnancy increased from 2.5% in 1998 to 8.1% in 2005.8 More recent research suggests that 7% to 13% of pregnant women use antidepressants.9
There are three important neurodevelopmental factors relating to the influence of 5-HT on fetal brain development. The first is the placenta, a major source of 5-HT with the assistance of maternally supplied essential amino acid tryptophan; the second is the fetal forebrain, which accumulates 5-HT within the first trimester; and the third is maternal-placental-fetal interactions that result in genetic mutations throughout gestation.10 Because SSRIs cross the placental and blood-brain barriers, they have the potential to alter fetal central 5-HT signaling by blocking the 5-HT transporter 5-HTT, thereby raising extra-cellular 5-HT levels. 5-HT is widely distributed within the central nervous system and acts as a modulatory neurotransmitter by regulating cognition, attention, emotion, learning, sleep, arousal, and stress responses in a mature brain. During neurodevelopment, 5-HT regulates cell division, differentiation, migration, growth cone elongation, myelination, synaptogenesis, and dendritic pruning.
As a result, the serotonergic hypothesis suggests that during neurodevelopment, auto-inhibitory feedback restrains the maturation of the 5-HT system through negative feedback signaling from a high 5-HT volume and decreasing 5-HT signaling later in life, creating an increased risk of depression. However, reduced 5-HT levels due to maternal depression are also associated with neurobehavioral disturbances, which suggests SSRI treatment during pregnancy could potentially have beneficial effects on fetal neurodevelopment by counteracting other factors that reduce fetal central 5-HT-ergic tone, but such exposure may create the influence of other factors that increase fetal 5-HT-ergic tone through genetic polymorphisms.
Regardless, the central serotonergic auto-feedback hypothesis is not a singular construct and requires further examination of functional alterations in specific raphe subregions and downstream effects on postsynaptic structures that contribute to the complexity of behavioral outcomes.11
Both the presence and absence of 5-HT, conditions affected by SSRI exposure and maternal depression respectively, influence molecular, neuroanatomical, and functional aspects of development and are clinically apparent. A higher than expected rate of depression and anxiety are reported in families of children with ASDs, but if 5-HT levels are important, a lack of treatment with an SSRI is definitively linked to maternal weight gain, underutilization of health care, substance use, impaired psychosocial functioning, increased risk of preterm birth, lower birth weight, fetal growth restriction, and pre-eclampsia, with behavioral effects such as increased irritability, less activity and attentiveness, elevated cortisol, decreased peripheral dopamine levels, and lower vagal tone.8
In conjunction with this theory, researchers have demonstrated higher rates of ASDs in children exposed inutero to serotonergic drugs, including cocaine and possibly alcohol. Studies of serum serotonin levels in autism in the 1960s and 1970s demonstrated hyperserotonemia to about 50% over normal levels in one-third of individuals with autism. Animal research has shown hyperserotonemia produces a reduced drive for social attachment through the inhibition of social distress, but peripheral measurements are confounded by the fact that all 5-HT found in blood is manufactured in the gut prior to platelet absorption. However, hyperserotonemia is not apparent in all individuals with severe ASD, and it has been found in a variety of conditions such as schizophrenia, Huntington’s disease, and severe intellectual disability.12 In 2008, an evaluation of the Swedish Multi-Generation Register and Hospital Discharge Register by Daniels et al.13 found parents of children with autism per DSM-IV-TR criteria were more likely to have been hospitalized for a mental disorder, particularly schizophrenia, than parents of control subjects. Depression and personality disorders were also common among mothers of children with severe ASD. The positive association between any parental psychiatric disorder and the child’s diagnosis of autism was present regardless of the timing of the parent’s diagnosis relative to the child’s diagnosis, although mild cases of mental illness were not evaluated in this review.13
Despite positive studies of correlation, two studies failed to show an inverse relationship between serum 5-HT levels and verbal expressive ability, and so the role of hyperserotonemia in severe ASD remains unknown. These findings suggest most reports of serum 5-HT are complicated by factors beyond the diagnosis of autistic disorder, but similar studies of blood, urine, and cerebrospinal fluid for dopamine, endogenous opioids, glycine, gamma-aminobutyric acid, cortisol, oxytocin, and norepinephrine did not find significant differences in children with and without the severe form of ASD. The majority of neuro-chemical research in autism research has been contradictory, with limited subject and comparison groups. However, central 5-HT synthesis and genetic variations remain the most promising area for future neurochemical research of ASDs’ etiology, as central 5-HT can now be recognized by the fifth week of gestation.12
SSRI Use and Obstetric Complications
The growth in antidepressant use during pregnancy as well as ASD prevalence has prompted research on the link between 5-HT and obstetric risk factors. Maternal stress, often linked to an absence of 5-HT, has established genetic and environmental effects from pregnancy into infancy through disruption of neurobehavioral development, reduced birth weight, and increased prematurity. This has been thought to be due to exposure to increased levels of adrenal hormones, which adversely influence glucocorticoid receptors in the developing fetal brain and result in disruption of stress responses. Newborns of depressed mothers may show more irritability, greater right frontal electroencephalographic activation, and may have reduced dopaminergic levels compared with controls.
A 2006 study by Oberlander et al.14 found neonates born to non–SSRI-exposed depressed mothers and SSRI-exposed depressed mothers were similar in gestational ages, birth weights, incidence of feeding problems, convulsions, and jaundice, suggesting depression severity rather than SSRI exposure contributed to adverse neonatal outcomes. However, prenatal SSRI exposure was associated with reduced birth weight for gestational age and an increased incidence of neonatal respiratory distress, which suggested a compounding effect of SSRI exposure and depression in comparison to depression alone. Although not exploring ASDs as an outcome, this study did present respiratory distress, which may lead to hypoxia, as a risk factor associated with SSRI use for the development of ASDs. This study notes the difficulty of studying the effects of SSRI exposure independent of depression because SSRI use occurs in the context of maternal mental illness.
A 2012 population-based case control study of a large cohort group with a small percentage of mothers with depression taking SSRIs found that untreated maternal depression was associated with slower rates of fetal body and head growth, whereas depressed mothers treated with SSRIs had fewer depressive symptoms and their fetuses had no delay in body growth but had delayed head growth and were at increased risk for preterm birth, which is an established risk factor for ASD.15
SSRI Exposure and Developmental Delay
The early diagnosis of ASDs through failure to meet developmental milestones prompts the question of whether neonatal developmental delay may also provide diagnostic clues. A 2004 study of 17 cases and 16 controls found that women who use SSRIs during pregnancy have healthy, full-birth weight newborn infants with shorter mean gestational ages, increased tremulousness, disrupted rapid eye movement sleep, and few changes in behavioral states within the first day of life.16 Although the infants were not evaluated further, the study postulated these behaviors could be due to neonatal withdrawal syndrome or serotonin toxicity, and longitudinal evaluation of the effects of SSRIs as related to ASDs or other delays were not observed.
In 2010, Pedersen et al.17 examined developmental milestone changes in infants of mothers with SSRI use during pregnancy through parental self-report. Children with second- or third-trimester exposure to antidepressants were able to sit up 15.9 days and walk 28.9 days later than children of women not exposed to antidepressants, but still were within the normal range of development. Fewer children with second- or third-trimester exposure to antidepressants were able to sit without support at 6 months of age, and fewer were able to occupy themselves at 19 months of age. None of the other milestones measured showed statistically significant associations with antidepressant exposure. Again, as there was no longitudinal follow-up beyond age 19 months or specific evaluation for development of ASDs, the effects of antidepressants in this study may have a permanent or reversible effect. Further confounding occurred due to reporter bias and uncertain depression severity.17
A review of studies between 1973 and 2010 that focused on neurodevelopmental disorders following in-utero exposure to antidepressants found no effect on development.18 However, two cohort studies found significant psychomotor retardation between ages 6 and 40 months for infants exposed to fluoxetine, paroxetine, sertraline, or other antidepressants. A large Danish study16 of a birth registry comparing psychomotor development of 415 children exposed to antidepressants in-utero, 489 mothers with depression who did not receive antidepressants while pregnant, and 81,042 children whose mothers were not depressed and had not been exposed to psychotropic medication found, on average, that children exposed to anti-depressants during the third trimester learned to walk or sit up slightly later than unexposed children, but no link between ASDs and SSRI exposure was confirmed.
SSRI Exposure and Autism Spectrum Disorder Development
Following the dearth of findings in risk factors and developmental delay, several studies evaluated the link between SSRIs and ASDs more directly. Researchers at Kaiser Permanente compared medical records of mothers who gave birth to children with ASDs to controls and found women who used SSRIs at any time during the 12 months before delivery were twice as likely as other women to give birth to children with ASD.19 A population based case-control study by Rai et al.20 determined a heightened risk for maternal depression and severe ASD is conferred through SSRI use, although it is not the only source because a number of medications have serotonergic activity and because severe depression could not be ruled out as a source. The study also considered alternate explanations, such as alcohol or drug use during pregnancy not severe enough to require health services, that are plausible.20 Gidaya et al.,21 in their review of Denmark’s health registers, found an odds ratio (OR) of 2.0 for ASDs in the offspring of mothers with SSRI exposure and an OR of 1.9 when compared to mothers with untreated depression. However, they also found children with ASDs were more likely than controls to be male, have older parents, and have a mother with history of depression.21
Although multiple studies have linked SSRI use to the development of ASDs, particularly in the first trimester when extraembryonic sources of 5-HT influence very early development of forebrain circuitry related to ASDs and mood disorders, many of these studies caution abstinence from antidepressants in mothers with depression due to study limitations preventing appropriate evaluation of depression severity and its impact on the development of ASDs.22 There were further variations in the number of case-control populations, diagnostic criteria for ASDs (including the International Classification of Diseases23 and DSM-IV-TR5), and extensive studies including chart review of registries, medical records, or self-report with limited patient contact. Confirmed risk factors for the development of ASDs, such as advanced parental age and an existing child with ASD, were controlled for in some studies.
Hviid et al.24 conducted a large population-based and registry-based prospective cohort study of SSRI exposure before and during pregnancy as well as diagnosis of ASDs in children in Denmark. Researchers included only single births in a 10-year period, following children from birth to age 5–10 years, and excluded conditions associated with an increased risk of ASDs while incorporating confounders into their analysis. The study found that in comparison with no SSRI exposure, children of women exposed to SSRIs both before and during pregnancy were more likely to receive a diagnosis of ASD according to a crude analysis (rate ratio 1.62; 95% confidence interval [CI], 1.23–2.13), but after adjusting for psychiatric diagnosis and other confounders the association was no longer apparent (rate ratio 1.20; 95% CI, 0.90–1.61). The prevalence of pregnancy-related use of SSRIs and ASDs in this study was very low (0.97% and 0.62%, respectively) compared to the United States (5.6% and 1.14%, respectively). The length of follow-up also varied between 5 and 10 years, which may have resulted in an erroneous diagnosis of ASDs after the age of 5 years. This study could not rule out a relative risk up to 1.61, and therefore the association warrants further study in this regard.22,24
Conclusions and Clinical Recommendations
SSRIs do cause certain limitations at birth that are linked to risk factors for ASDs, but not all of these risk factors independently, or even in some cases combined, always result in the disorders. The inconsistencies and contradictions between studies reflect variations in diagnostic criteria, methodology (including inadequate sample sizes with innumerable difficult-to-control factors), and poor follow-up. Although the serotonergic theory has proven to be consistent in animal models, the link between SSRIs and ASDs is complicated by the influence of depression severity and other ASD risk factors. The shared message of all research endeavors exploring this complex association remains an urge to clinicians to cautiously weigh the risks and benefits of beginning an SSRI in a woman suffering from mental illness who may have difficulty committing to psychotherapy, as untreated depression remains a concerning and established factor in poor long-term behavioral and physical outcomes for children.
- American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th edition. Washington, DC: American Psychiatric Publishing; 2014.
- Autism Spectrum Disorder. American Psychiatric Association. Available at: http://www.dsm5.org/Documents/Autism%20Spectrum%20Disorder%20Fact%20Sheet.pdf. Published 2013. Accessed January 26, 2015.
- Autism Spectrum Disorder Data & Statistics. Centers for Disease Control and Prevention. Updated March 24, 2014. Available at: http://www.cdc.gov/ncbddd/autism/data.html. Accessed January 26, 2015.
- Ratajczak HV. Theoretical aspects of autism: causes--a review. J Immunotoxicol. 2011;8(1):68–79. doi:10.3109/1547691X.2010.545086 [CrossRef]
- American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th edition, text revision. Washington, DC: American Psychiatric Publishing; 2000.
- Kolevzon A, Gross R, Reichenberg A. Prenatal and perinatal risk factors for autism: a review and integration of findings. Arch Pediatr Adolesc Med. 2007;161(4):326–333. doi:10.1001/archpedi.161.4.326 [CrossRef]
- Gardener H, Spiegelman D, Buka SL. Perinatal and neonatal risk factors for autism: a comprehensive meta-analysis. Pediatrics. 2011;128(2):344–355. doi:10.1542/peds.2010-1036 [CrossRef]
- Alwan S, Reefhuis J, Rasmussen SA, Friedman JM. Patterns of antidepressant medication use among pregnant women in a United States population. J Clin Pharmacol. 2011;51(2):264–270. doi:10.1177/0091270010373928 [CrossRef]
- Autism spectrum disorders revisited. Harvard Mental Health Letter. October2011.
- Harrington RA, Lee LC, Crum RM, Zimmerman AW, Hertz-Picciotto I. Serotonin hypothesis of autism: implications for selective serotonin reuptake inhibitor use during pregnancy. Autism Res. 2013;6(3):149–168. doi:10.1002/aur.1288 [CrossRef]
- Oberlander TF, Gingrich JA, Ansorge MS. Sustained neurobehavioral effects of exposure to SSRI antidepressants during development: molecular to clinical evidence. Clin Pharmacol Ther. 2009;86(6):672–677. doi:10.1038/clpt.2009.201 [CrossRef]
- Lam KS, Aman MG, Arnold LE. Neuro-chemical correlates of autistic disorder: a review of the literature. Res Dev Disabil. 2006;27(3):254–289. doi:10.1016/j.ridd.2005.03.003 [CrossRef]
- Daniels JL, Forssen U, Hultman CM, et al. Parental psychiatric disorders associated with autism spectrum disorders in the off-spring. Pediatrics. 2008;121(5):e1357–e1362. doi:10.1542/peds.2007-2296 [CrossRef]
- Oberlander TF, Warburton W, Misri S, Aghajanian J, Hertzman C. Neonatal outcomes after prenatal exposure to selective serotonin reuptake inhibitor antidepressants and maternal depression using population-based linked health data. Arch Gen Psychiatry. 2006;63(8):898–906. doi:10.1001/archpsyc.63.8.898 [CrossRef]
- El Marroun H, Jaddoe VW, Hudziak JJ, et al. Maternal use of selective serotonin reuptake inhibitors, fetal growth, and risk of adverse birth outcomes. Arch Gen Psychiatry. 2012;69(7):706–714. doi:10.1001/archgenpsychiatry.2011.2333 [CrossRef]
- Zeskind PS, Stephens LE. Maternal selective serotonin reuptake inhibitor use during pregnancy and newborn neurobehavior. Pediatrics. 2004;113(2):368–375. doi:10.1542/peds.113.2.368 [CrossRef]
- Pedersen LH, Henriksen TB, Olsen J. Fetal exposure to antidepressants and normal milestone development at 6 and 19 months of age. Pediatrics. 2010;125(3):e600–608. doi:10.1542/peds.2008-3655 [CrossRef]
- SSRI antidepressants: altered psychomotor development following exposure in utero?Prescrire Int. 2013;22(135):43–44.
- Croen LA, Grether JK, Yoshida CK, Odouli R, Hendrick V. Antidepressant use during pregnancy and childhood autism spectrum disorders. Arch Gen Psychiatry. 2011;68(11):1104–1112. doi:10.1001/archgenpsychiatry.2011.73 [CrossRef]
- Rai D, Lee BK, Dalman C, Golding J, Lewis G, Magnusson C. Parental depression, maternal antidepressant use during pregnancy, and risk of autism spectrum disorders: population based case-control study. BMJ. 2013;346:f2059. doi:10.1136/bmj.f2059 [CrossRef]
- Gidaya NB, Lee BK, Burstyn I, Yudell M, Mortensen EL, Newschaffer CJ. In utero exposure to selective serotonin reuptake inhibitors and risk for autism spectrum disorder. J Autism Dev Disord. 2014;44(10):2558–2567. doi:10.1007/s10803-014-2128-4 [CrossRef]
- Ostuzzi G, Barbui C. Autism spectrum disorders: weighing the risk of SSRI exposure in pregnancy. Epidemiol Psychiatr Sci. 2014;23(3):231–233. doi:10.1017/S2045796014000286 [CrossRef]
- World Health Organization. International Classification of Diseases, version 10. Geneva, Switzerland: World Health Organization; 1994.
- Hviid A, Melbye M, Pasternak B. Use of selective serotonin reuptake inhibitors during pregnancy and risk of autism. N Engl J Med. 2013;369(25):2406–2415. doi:10.1056/NEJMoa1301449 [CrossRef]