Up to 20% of mothers experience antenatal depression and approximately 30% of these women are treated with a serotonergic antidepressant during pregnancy [1,2]. Selective serotonin reuptake inhibitors (SSRI) are the most common serotonergic medications prescribed [3-7]. They readily cross the placenta and the fetal blood-brain barrier, potentially altering serotonin (or 5-hydroxytryptamine [5-HT]) signaling in the fetal brain [8-14], and such exposure has been paradoxically reported to be associated with an increased risk for anxiety, attention, and behavioral disorders in children of mothers with depression treated with an SSRI during pregnancy [15-17]. Importantly, childhood behaviors have been associated with altered indices of stress regulation [18-21], raising critical questions about whether prenatal exposure to serotonergic psychotropic medications alters stress reactivity, thereby contributing to an increased risk for behavioral disturbances.

Long before 5-HT becomes a neurotransmitter in the mature brain, it plays a role as a neurodevelopmental signal regulating cell growth and function [22,23]. In the fetal brain, 5-HT and its receptors are overexpressed and widespread in regions where they are absent in adults, pointing to a time-dependent specificity to 5-HT expression during development [24,25]. Early 5-HT alterations, either via pharmacological, genetic, or other manipulations, potentially alter these processes via the presynaptic, membrane-bound serotonin transporter protein (5-HTT), the target of SRI antidepressants. 5-HTT is a key regulator of brain 5-HT [26,27].

Serotonin is central to the development and function of two key stress response systems—the locus-coeruleus-norepinephrine (autonomic nervous system [ANS]) and the hypothalamic-pituitary-adrenal (HPA) systems [28-30], which may illustrate sites affected by prenatal exposure to serotonin reuptake inhibitors on stress responses [31-33]. The relationship between 5-HT and stress reactivity is bidirectional; stressors appear to alter 5-HT metabolism as well as bias how one copes with subsequent stressful challenges [32,34]. ANS activation leads to a rapid “flight or fight” response, in turn leading to increased cardiac activity (heart rate) and release of catecholamines (norepinephrine) [35]. Central to our understanding of how prenatal exposure to serotonergic medications such as antidepressants influences early brain development is understanding the diverse roles the neurotransmitter 5-HT plays in early brain development, stress regulation, and mental health [23,28,36]. Prenatal maternal mood disturbances, the very disorders that lead to antidepressant treatment, have also been shown to shape the development of the HPA axis [37,38].

5-HT and cardiovascular/autonomic stress regulation are highly interrelated via links between reflex control of parasympathetic outflow to the heart and other organs that involve central 5-HT1A receptors located in the vicinity of preganglionic vagal neurons. Further, 5-HT3 receptors are implicated in afferent regulation of central sympathetic and parasympathetic tone [39]. The development and function of the HPA stress response and the serotonergic regulatory systems are highly interrelated and exquisitely sensitive to the effects of early adverse experience [40,41]. Serotonin influences how an individual copes with subsequent social stressors and plays a role in mediating the effects of adverse experience [42]. Early differences in maternal care alter central 5-HT levels that change HPA axis stress function, reflected as an altered capacity to regulate stress responses [40,43].

Considering the importance of serotonin in neurodevelopment, it is conceivable that early changes to 5-HT, secondary to prenatal serotonergic medication exposure or maternal mood disorders, could have developmental consequences [26,44] and may modify the formation and function of key stress regulatory systems such as the ANS and HPA axis in ways that may affect subsequent responses to stress challenges and may have life-long implications for the offspring’s health and behavior. Understanding relationships between stress response systems and children’s behavior may provide essential insight into the developmental origins of physiological processes that contribute to disrupted behavior. Taken together, it is possible that prenatal exposure to serotonergic psychotropic medications used to manage mood disturbances during pregnancy could alter stress reactivity/regulation in offspring.

Our study aims to assess the impact of prenatal exposure to serotonergic medications, and distinguish these effects from the impact of maternal mood disturbances on neonatal, infant, childhood, and adolescent indices of stress regulation.

We will adhere to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines for reporting systematic reviews [45] and have used the PRISMA for systematic review protocols (PRISMA-P) [46]. Our research protocol was designed a priori, defining methods for searching the literature, including and examining articles, and extracting and analyzing data.

We will assess physiological outcomes and how they relate to each other. Specifically, our outcome includes the following physiological outcomes: cortisol, heart rate variability, salivary amylase, pupillary size, C-reactive protein (CRP), and immunological biomarkers (cytokines, chemokines, lymphokines, IL-6, etc).

We developed our search strategy with the consultation of a librarian and will search the literature by population and intervention (Table 1). The following databases will be searched: MEDLINE (Ovid), Embase (Ovid), CINAHL (EBSCOhost), PsycINFO (EBSCOhost), and ClinicalTrials.gov. We will use both Medical Subject Headings (MeSH) terms and keywords. Our search strategy for the MEDLINE database is outlined in Multimedia Appendix 1. We will adjust the MeSH terms and keywords used to accommodate the different databases’ requirements and limitations. We will limit our search to studies published in English or French. We will not limit for year of publication and will include original studies of all study types.

We will use the following keywords and MeSH terms: Pregnancy Trimesters/ or Pregnancy/ or Pregnancy Trimester, Third/ or pregnancy.mp. or Pregnancy Trimester, First/ or Pregnancy, or Pregnancy Trimester, Second/, pregnant.mp. or Pregnant Women/, or gestation.mp. or perinatal.mp. or prenatal.mp. or pregnant* AND Depression/ or depressive disorder/ or depressive disorder, major/ or depressive disorder, treatment-resistant/ or Anxiety Disorders/ or Anxiety/ or anxiety.mp., or Mood Disorders/, AND antidepressants.mp. or Antidepressive Agents/, serotonergic drugs.mp. or Serotonin Agents/, selective serotonin reuptake inhibitors.mp. or Serotonin Uptake Inhibitors/, citalopram.mp. or Citalopram/, escitalopram.mp. or Citalopram/, fluoxetine.mp. or Fluoxetine/, fluvoxamine.mp. or Fluvoxamine/, paroxetine.mp. or Paroxetine/, sertraline.mp. or Sertraline/, “Serotonin and Noradrenaline Reuptake Inhibitors”/ or Serotonin Uptake Inhibitors/ or serotonin-norepinephrine reuptake Inhibitor.mp., desvenlafaxine.mp., or Desvenlafaxine Succinate/, duloxetine.mp. or Duloxetine Hydrochloride/, levomilnacipran.mp. or Levomilnacipran/, venlafaxine.mp. or Venlafaxine Hydrochloride/, Antipsychotic Agents/ or Second generation antipsychotics.mp., aripiprazole.mp., or Aripiprazole/, brexpiprazole.mp., olanzapine.mp., or Olanzapine/, quetiapine.mp. or Quetiapine Fumarate, risperidone.mp. or Risperidone/, Serotonin Modulators.mp., trazodone.mp., or Trazodone/, vilazodone.mp. or Vilazodone Hydrochloride/, vortioxetine.mp. or Vortioxetine/, tricyclic antidepressants.mp. or Antidepressive Agents, Tricyclic/, amitriptyline.mp. or Amitriptyline/, clomipramine.mp. or Clomipramine/, desipramine.mp. or Desipramine/, doxepin.mp. or Doxepin/, imipramine.mp. or Imipramine/, nortriptyline.mp. or Nortriptyline/, trimipramine.mp. or Trimipramine/ or buspirone.mp. or Buspirone/.

We will conduct a manual search of the journals Psychoneuroendocrinology, Early Human Development, Neuroscience, and Neuroscience & Biobehavioral Reviews, as well as a forward and backward citation search through Google Scholar [47] on all included papers to locate additional papers that may have been missed in our literature search.

Using our predetermined selection criteria, two authors (EZZ and AL) will independently screen all retrieved papers at level 1 (title and abstract) for inclusion in the study using Covidence, a screening and data extraction tool for systematic reviews [48]. Once a list of studies is determined, the selected papers will be reviewed at level 2 (full text) to select a final list of review studies. Screening questions can be found in Multimedia Appendix 2. At any point, authors will meet to discuss discrepancies and a third author (SH) will be consulted if disagreement occurs.

Two authors (EZZ and AL) will independently extract the following information from all included studies: year of publication, country, sample size, study design, study setting, trimester of pregnancy, drug exposure class, drug exposure generic name, drug dose, cortisol level (diurnal), cortisol level (stress challenge), heart rate variability, salivary amylase, pupillary size, CRP, cytokines, chemokines, lymphokines, IL-6, maternal depression, maternal depression diagnosis method, maternal anxiety, and maternal anxiety diagnosis method. Authors will meet to discuss discrepancies and a third author (SH) will be consulted if disagreement occurs. If needed, we will contact study corresponding authors for unpublished or missing data.

To assess the methodological quality of the included studies and their risk of bias, we will use different checklists and scales. Two authors (EZZ and AL) will independently screen each included paper according to its methodology. We will use the Newcastle-Ottawa Scale for observational cohort studies [49], the modified Newcastle-Ottawa Scale for observational cross-sectional studies [50], and the Cochrane risk-of-bias tool for randomized controlled trials and experimental studies [51]. We will use funnel plots to assess for publication bias [52]. In case of a publication bias, we will use the trim-and-fill method. We will remove (“trim”) the studies that give rise to the funnel plot’s asymmetry and then impute (“fill”) the suggested missing studies based on the bias-corrected overall estimate [53].

We will pool studies based on their reported outcome and will present the characteristics of included studies both descriptively and in a table. Where possible, we will pool reported levels of cortisol, heart rate variability, salivary amylase, pupillary size, and CRP. We will calculate the Cochrane Q test (chi-square) and Higgins I² score to assess the statistical heterogeneity of effect size estimates across our included studies before running a meta-analysis. When meta-analysis is possible, we will calculate pooled mean differences for continuous data and perform a random-effects meta-analysis for dichotomous data [54]. If meta-analysis is not possible, we will conduct a narrative synthesis of the data. When a sufficient number of studies are found, we will perform subgroup analysis and assess outcomes measured by drug class, dose, trimester of exposure, and child’s age and gender. We will consider method of assessment for pooling. We will perform all data analysis using Stata (version 15; StataCorp LLC).