Selective serotonin reuptake inhibitors and preeclampsia: a quality assessment and meta-analysis

Abstract

Serotonin modulates vascular, immune, and neurophysiology and is dysregulated in preeclampsia. Despite biological plausibility that selective serotonin reuptake inhibitors (SSRIs) prevent preeclampsia pathophysiology, observational studies have indicated increased risk and providers may be hesitant. The objective of this meta-analysis and quality assessment was to evaluate the evidence linking SSRI use in pregnancy to preeclampsia/gestational hypertension. PubMed was searched through June 5, 2020 manually and using combinations of terms: “preeclampsia”, “serotonin”, and “SSRI”. This review followed MOOSE guidelines. Inclusion criteria were: 1) Observational cohort or population study, 2) exposure defined as SSRI use during pregnancy, 3) cases defined as preeclampsia or gestational hypertension, and 4) human participants. Studies were selected that addressed the hypothesis that gestational SSRI use modulates preeclampsia and/or gestational hypertension risk. Review Manager Web was used to synthesize study findings. Articles were read and scored (Newcastle-Ottawa Quality Assessment Scale) for quality by two independent reviewers. Publication bias was assessed using a funnel plot and the Egger test.

Of 179 screened studies, nine were included. The pooled risk ratio (random effects model) was 1.43 (95% CI: 1.15-1.78, P<0.001; range 0.96-4.86). Two studies were rated as moderate quality (both with total score of 6); others were high quality. Heterogeneity was high (I2=88%) and funnel asymmetry was significant (p<0.00001).

Despite evidence for increased preeclampsia risk with SSRIs, shared risk factors and other variables are poorly controlled. Depression treatment should not be withheld due to perceived gestational hypertension risk. Mechanistic evidence for serotonin modulation in preeclampsia demonstrates a need for future research.

Introduction

Preeclampsia is a severe hypertensive complication of pregnancy associated with multiple health risks within and beyond pregnancy.[1–3] The incidence of preeclampsia is high and continues to grow (now 3-10% worldwide), making it a significant public health concern.[4] Despite a robust literature on the immune, neurological, and cardiovascular mechanisms underlying preeclampsia,[5–8] there is no cure for this often-devastating gestational disorder.[9] Recent work has found that preeclampsia is associated with dysregulation of serotonergic systems, which are also involved in neuropsychiatric function.[10–12] Studies of mood disorders further point to a link between preeclampsia and serotonin dysfunction. There is a significant bidirectional interaction between risk for preeclampsia and mood disorders, particularly in the peripartum period. Perinatal or postpartum mood and anxiety disorders (PMAD) are a significant and growing public health concern.[13, 14] Mood disorders (e.g., depression, dysthymia) are associated with an approximately three-fold increase in risk for preeclampsia,[15, 16] and preeclampsia is a significant risk factor for PMAD and is associated with increased depression severity.[2, 17] Women who are diagnosed with preeclampsia are at 2.4 times higher odds of developing PMAD,[18] and gestational hypertensive disease is significantly more common in women with psychiatric illness.[19] For example, women with depression have significantly increased rates of preeclampsia and gestational hypertension (1.28 and 1.23 odds, respectively).[20] Furthermore, ongoing debates in the field biological psychiatry about the role of serotonin in major depression underscore the need to understand non-nervous system mechanisms of SSRI action. Inflammatory and vascular mechanisms, for example, which are implicated in depression and gestational hypertension, are explored here as potential substrates for SSRI efficacy.

The objective of this meta-analysis and quality assessment was to review the existing literature, including its limitations, linking gestational SSRI use to preeclampsia or gestational hypertension risk. The research question considered was whether gestational SSRI use altered the risk for hypertensive disorders of pregnancy. Additionally, we examined potential sources of bias and confounds in the existent literature. Because SSRIs are used in mood disorder patients, who are themselves at increased risk for preeclampsia/gestational hypertension, this and psychiatric symptom severity are significant potential sources of confounding bias. We propose future directions to improve the application of serotonin therapeutics to hypertensive disorders of pregnancy and the use of serotonin biomarkers for preeclampsia treatment and detection. This work has important implications for current psychiatric and obstetrics care, as well as for future development of targeted, mechanistic therapeutics.

Materials and Methods

Eligibility criteria, information sources, search strategy

This review and meta-analysis were performed according to the Meta-analysis Of Observational Studies in Epidemiology (MOOSE) guidelines.[21] We identified published English-language observational studies of preeclampsia risk or gestational hypertension given SSRI use in pregnancy. Studies were identified via an electronic search of PubMed through June 5, 2020 with no starting date, language, or location restrictions. Combinations of the following search terms were used to identify relevant studies: “preeclampsia,” “gestational hypertension,” “serotonin,” and “SSRI.” Manual searching was also used. Resultant titles and abstracts were evaluated for relevance. Abstracts and conference proceedings were excluded.

The population of interest was pregnant women. The exposure group was women who were exposed to SSRIs (all types) during pregnancy. The comparison group was those who were unexposed to SSRIs. The outcome of interest was gestational hypertension or preeclampsia.

The protocol for this meta-analysis and review was registered prospectively to the International Prospective Register of Systematic Reviews (CRD42022300867) on February 4, 2022.

Study selection

A single reviewer (S.B.G.), in consultation with the senior author (M.K.S.), determined study eligibility after reading the abstracts and titles of all 179 screened studies (via Pubmed or manual search). Inclusion criteria were: 1) Observational cohort or population study, 2) exposure defined as SSRI use during pregnancy, 3) cases defined as preeclampsia or gestational hypertension, and 4) inclusion of human participants. Abstract and title review revealed that 161 of these 179 studies were not primary data papers (e.g., review, case study), were conference proceeding abstracts, were animal studies, or were not focused on SSRI use in pregnancy and preeclampsia and/or gestational hypertension outcomes. Based on abstract and title screening, a subset of 18 studies appeared to meet inclusion criteria. One reviewer then read the full text of these 18 studies to determine final inclusion of nine studies. Studies were included if they addressed the hypothesis that gestational SSRI use modulates preeclampsia and gestational hypertension risk. The study selection process is outlined in Figure 1.

Figure 1: PRISMA Flow diagram.

Flow diagram of studies assessing risk for preeclampsia or gestational hypertension following gestational SSRI use. PRIMSA, preferred reporting items for systematic reviews and meta-analyses.

Data extraction

Data from the final nine eligible studies were extracted and analyzed. Data on study protocol and statistical methods, study confounders/covariates, and risk ratio (adjusted odds ratios or relative risks with confidence interval) were extracted.

Assessment of risk bias

Articles were read and scored for quality by two independent, blinded reviewers (SBG and BMS) using the Newcastle-Ottawa Quality Assessment Scale.[22] The Newcastle-Ottawa Scale assigns stars to each study along quality criteria, which are selection of study groups, comparability of study groups, and ascertainment of exposure and/or outcome of interest. Total scores range from 0-9, with 0-3 considered low quality, 4-6 considered moderate quality, and 7-9 considered high quality. Reviewers agreed on all scores, so a tiebreaker (by the senior author) was not necessary.

Data synthesis

Data were pooled using a random effects model. Generic inverse variance was calculated to determine an adjusted pooled risk ratio for the association between SSRI use in pregnancy and gestational hypertension or preeclampsia. A summary estimate of the risk ratio was tabulated manually from adjusted relative risks and odds ratios, which were pooled per the rare disease assumption (Excel, Microsoft). Risk ratios (odds ratios and relative risk) for each study and the pooled risk ratio (± 95% confidence intervals) were displayed in a forest plot.

Heterogeneity between studies was assessed by the I² statistic, with >75% considered high heterogeneity.[23] Publication bias was assessed using a funnel plot and the Egger test given large differences between the sizes of treatment and control groups across studies. P<0.05 was considered statistically significant. Review Manager Web (version 5.4.1) was used to perform analyses and to generate plots.[24]

Results

Study selection and characteristics

Of 179 total English-language studies identified via manual and database search and evaluated for inclusion, the full text of 18 studies was reviewed and nine were included in the final analysis (Figure 1). Study sample size ranged from 1,171 to 1,062,190 participants in the non-SSRI user control group, and from 31 to 14,979 participants in the SSRI user group. Among the final nine included studies, two[25, 26] evaluated gestational hypertension risk while seven[11, 27–32] evaluated preeclampsia risk. The studies were published between 2009 and 2020.

Risk of bias

All studies had final total scores of between 6 and 9 on the Newcastle-Ottawa Quality Assessment Scale (Table 1). Two were rated as moderate quality (both with total score of 6) [25, 30], while all others were rated as high quality. Study limitations, including a failure to adjust adequately for confounding variables, are summarized in Table 2.

Table 1:

Newcastle-Ottawa Quality Assessment Scale

Reference	Newcastle-Ottawa Quality Assessment
	Selection	Comparability	Outcome	Total score
(De Vera & Bérard, 2012)	★★★★		★★	6
(Palmsten, Setoguchi, Margulis, Patrick, & Hernández-Díaz, 2012)	★★★★	★★	★★★	9
(Palmsten, Chambers, Wells, & Bandoli, 2020)	★★★★	★	★★★	8
(Avalos, Chen, & Li, 2015)	★★★★	★	★★★	8
(Zakiyah et al., 2018)	★★★★	★	★★★	8
(Toh et al., 2009)	★★★★	★	★	6
(Reis & Kallen, 2010)	★★★★	★	★★	7
(Lupattelli, Wood, Lapane, Spigset, & Nordeng, 2017)	★★★	★★	★★	7
(Palmsten et al., 2013)	★★★	★	★★★	7

Table 2:

Summary of studies of SSRIs and associated gestational hypertension or preeclampsia outcomes including study limitations.

Ref	Design	Outcome	Risk Ratio ± 95% CI	Sample size		Limitations	Co-medication reporting, analyses
				Non-SSRI controls	SSRI exposed
(De Vera & Bérard, 2012)	nested case-control study	GH	aOR 1.60±1.00-2.55	1171	31	Sev, Dose, Hyp, NS	Adjusted for co-medications through 1 year pre-pregnancy.
(Palmsten, Setoguchi, Margulis, Patrick, & Hernánde z-Díaz, 2012)	retrospective cohort study	PE	aRR 1.22±0.77-1.54	65392	3169	NS	SNRI and TCA monotherapy and polytherapy assessed. Adjusted for benzodiazepines, anticonvulsants, antipsychotics.
(Palmsten, Chambers , Wells, & Bandoli, 2020)	retrospective cohort study	PE	aRR 1.28±0.95-1.74	4949	424	Sev, NS	Sensitivity analysis restricted to SSRIs, adjusted for benzodiazepines, antipsychotics.
(Avalos, Chen, & Li, 2015)	retrospective cohort study	PE	aRR 1.34±1.00-1.81	16402	1262	Sev, Dose, n, SSRI, Comed, NS	SNRI, NDRI, SARI, TCA, and other antidepressant monotherapy and polytherapy assessed.
(Zakiyah et al., 2018)	retrospective cohort study	GH	aOR 2.07±1.25-3.44	27481	394	Sev, SSRI, Hyp	MAOI, TCA, non-MAOI, RIMA, and other antidepressants assessed. Excluded users of antihypertensives or antidiabetics. Adjusted for co-medications.
(Toh et al., 2009)	retrospective cohort study	PE	aRR 4.86±2.70-8.76	5532	199	Sev, Dose, SR-med, SR-pree, n, SSRI, Co-med	Assessed co-medications through 2 months pre-pregnancy. Adjusted for non-SSRI antidepressant use.
(Reis & Kallen, 2010)	population-based database review	PE	aOR 1.50±1.33-1.69	1062190	14979	Sev, Dose, SSRI, Comed, Hyp	TCAs, MOAIs, and SNRIs assessed. Other co-medications reported.
(Lupattelli, Wood, Lapane, Spigset, & Nordeng, 2017)	retrospective cohort study	PE	aRR 0.96±0.64-1.45	5093	652	Dose, NS	SNRIs, TCAs, and other antidepressants assessed. Polytherapy excluded. Adjusted for co-medications (analgesics, psychotropics).
(Palmsten et al., 2013)	retrospective population study	PE	aRR 1.00±0.93-1.07	59219	19000	SSRI, Hyp, NS	SNRI, TCA, and other antidepressant monotherapy and polytherapy assessed. Adjusted for co-medications (anticonvulsants, benzodiazepines).

Abbreviations: Sev, did not account or anxiety/depression severity; Dose, did not account for SSRI dose; SR-med, utilized self-report medication data; SR-pree, utilized self-report preeclampsia data; n, small exposure group; SSRI, SSRI type not specified or controlled for; Comed, co-medication not controlled for; Hyp, no adjustment for superimposed preeclampsia and/or multiple preeclampsia risk factors (e.g., obesity, diabetes, smoking, race); BMI, Body mass index; NS, nonsignificant finding; Ref, reference number; aRR, adjusted relative risk; aOR, adjusted odds ratio; CI, confidence interval; GH, gestational hypertension; PE, preeclampsia; TCA, tricyclic antidepressant; SSRI, selective serotonin reuptake inhibitor; SNRI, serotonin-norepinephrine reuptake inhibitor; NDRI, norepinephrine-dopamine reuptake inhibitor; SARI, serotonin antagonist and reuptake inhibitor; MAOI, monoamine oxidase inhibitors; RIMA, reversible inhibitors monoamine oxidase A

Synthesis of results

Summary results are given in Table 2 and in Figure 1 (forest plot). Of the nine studies assessed, three reported a statistically significant effect of gestational SSRI use on increased preeclampsia[30, 31] or gestational hypertension[26] risk. Four studies reported a nonsignificant increased risk of preeclampsia[27–29] or gestational hypertension[25] with gestational SSRI use. The pooled RR for all nine included studies was 1.43 (95% CI: 1.15-1.78, P<0.001) and risk ratios ranged from 0.96-4.86 (Figure 2).

Figure 2. Forest plot of studies of the association between gestational SSRI use and gestational hypertension or preeclampsia.

Red rectangles represent odds ratio or relative risk reported in each study. Note that some CIs deviate very slightly from reported CIs due to RevMan automated calculations. Black lines represent 95% confidence interval (CI) for each study. Tips of black diamond represent 95% CI for the pooled OR and center of diamond represents pooled risk ratio across all nine studies.

Study limitations were also evaluated. Limitations included a failure to account for anxiety/depression severity,[25, 26, 28–31] SSRI dose,[11, 25, 29–31] no adjustment for superimposed preeclampsia and/or multiple preeclampsia risk factors (e.g., obesity, diabetes, smoking, race),[25, 26, 31, 32] nonsignificant increase in preeclampsia or gestational hypertension among SSRI users,[11, 25, 27–29, 32] no subgroup analysis by SSRI type,[26, 29–32] no control for comedication use,[29–31] use of self-report for preeclampsia[30] or medication use,[30] and a small exposure group [30] (Table 2). The most common limitation was a failure to account for anxiety/depression severity, which may independently drive risk for gestational hypertension/preeclampsia.

Heterogeneity and funnel asymmetry

Study heterogeneity was high (I²=88%). Despite this, adjusted effect estimates were in the same direction with the exception of two studies.[11, 32] High heterogeneity was therefore likely due to a high degree of overlap between confidence intervals. The test for funnel asymmetry was significant (p<0.00001) for the outcomes of interest (preeclampsia and gestational hypertension), suggesting that publication bias was a concern. A funnel plot depicts somewhat asymmetrical distribution of studies (Figure 3). The asymmetry noted on the plot can be associated with the relatively small numbers of subjects in each study upon which the conclusions were based.[33]

Figure 3. Funnel plot of studies associating gestational SSRI use with preeclampsia or gestational hypertension.

The test for funnel asymmetry was significant (p<0.00001) for the outcomes of interest (preeclampsia and gestational hypertension), suggesting that publication bias was a concern. A funnel plots depicts a somewhat symmetrical distribution of studies.

Discussion

This meta-analysis reveals a significantly increased pooled risk for preeclampsia in women who take SSRIs during pregnancy. However, the strength of the available evidence for this association is poor. Studies were confounded and limited by high heterogeneity and significant asymmetry for outcomes of interest. Findings may be due to the relatively small size of cohorts from which study conclusion were drawn. Furthermore, despite evidence for increased preeclampsia risk with SSRIs, most studies do not account for risk factors shared between mood disorders and hypertension or for underlying risk factors shared by depression and preeclampsia (Table 2). Providers and patients should not over-attribute gestational hypertension risks to antidepressant use in pregnancy—it is likely that benefits outweigh medication risks in many cases.

SSRIs represent one of the largest medical advances of the last several decades. More than one in every 10 Americans uses an antidepressant, with SSRIs being the common of these.[34] Use of SSRIs in pregnancy is common, generally well-tolerated, and safe.[35, 36] Despite this, many women forgo SSRI treatment in pregnancy due to concerns about impacts on their babies and pregnancies.[37] Healthcare providers themselves also often endorse concerns about psychotropic medication use in pregnancy.[38] Despite the fact that use of SSRIs to treat depression remains the clinical standard world-wide and is cost-effective, ongoing discussions have cast doubt on the serotonin hypothesis of depression. [39–41] Misunderstanding about the role of serotonin modulation in psychiatric diseases abounds. However, SSRI discontinuation or untreated anxiety or depression in pregnancy are associated with increased risk of adverse child health (e.g., preterm birth, low birth weight, difficult temperament, etc.[42, 43]), maternal health (e.g., suicide, cardiovascular disease, etc.[44]), and dyad outcomes (e.g. disrupted attachment, poor prenatal care, breastfeeding difficulties, etc.[43, 45, 46]). Given the health impacts of untreated depression in pregnancy, it is critical to clarify real risks associated with SSRI use in pregnancy so that balanced clinical decisions can be made.

One major limitation in the literature linking SSRIs to preeclampsia risk are differing baseline psychiatric differences between patients who do and do not discontinue SSRI use in pregnancy and postpartum. The most common limitation across the studies evaluated here was a failure to account for psychiatric disease severity. For example, more severe maternal depression (χ² = 6.4, p<.05), increased functional impairments [(t)=2.46, p<0.05), and increased psychiatric history complexity (χ² = 4.0, p<.05) are associated with increased use of antidepressants.[47] By failing to control for psychiatric disease severity or SSRI dose, as a number of studies reviewed here have done,[10, 25] findings could in fact be interpreted to mean that increased preeclampsia risk is linked to increased depression severity, as reported previously,[48] rather than SSRI use itself. To address this, future work might longitudinally assess PMAD symptoms and determine symptom control by SSRIs at the individual level within pregnancy, rather than relying on group-wise, baseline scores and diagnoses relative to standardized measures. This may be particularly relevant to pregnancy and peripartum, when stress reactivity, endocrine signaling, and mood and psychology are labile. [49, 50]

Women who use SSRIs in pregnancy may have increased preeclampsia risk simply because they also have more severe depression, which is an independent preeclampsia risk factor.[2] In fact, when dimensions of maternal mental health including psychiatric care, depression incidence, and co-medication are accounted for, relative risk for preeclampsia among SSRI users in mid-pregnancy is not significant (RR 1.16, 95% CI 0.92-1.45).[27] Similarly, after accounting for lifetime events of major depression in patients, SSRI use in early and mid-pregnancy does not significantly increase preeclampsia risk.[11] Additionally, there are a variety of risk factors that are common to mood disorders and preeclampsia that could drive associations. For example, both disorders are associated with metabolic and cardiovascular disease.[51–54] The question of SSRI efficacy for preeclampsia in patients without comorbid psychopathology therefore warrants further study.

The developmental programming literature also supports the conclusion that SSRI-related risk in epidemiology is often related to the underlying attributes of the study population rather than SSRI medications themselves. Studies have previously demonstrated that psychiatric disease symptom severity and not SSRI use is linked to adverse pregnancy and developmental programming outcomes. For example, neurodevelopmental disorders in offspring including ADHD and autism are linked to maternal depression and psychiatric symptoms but not to antidepressant exposure, such that when analyses are adjusted for psychiatric diagnoses, the association between SSRI exposure and neurodevelopmental outcome is no longer significant or is much diminished.[55–57]

Additional significant limitations of the literature linking SSRI use to preeclampsia risk include a lack of information on co-medication, dose and/or SSRI type, exposure misclassification, psychiatric disorder severity, use of self-report data, and limited sample sizes. For instance, Toh et al. utilized self-reported antidepressant use and preeclampsia history rather than medical or dispensation records, which may have resulted in a misclassification of patient SSRI exposure and thus an overestimation of risk,[30] as noted previously.[27] Use of non-SSRI anti-depressant medications may also have impacted gestational hypertension risk in some studies. For instance, SNRIs are known hypertensive agents, though impacts on blood pressure are typically modest (e.g., 9% increase, clinical significance in less than 6% of patients).[58, 59] Clinical considerations for the use of anti-depressant drugs often include management of cardiovascular risk, particularly in obstetrics patients.[60] Limitations in other studies are annotated in Table 1.

Rather than increasing obstetric risk, SSRIs may in fact be a viable therapeutic avenue. As we review here, targeting serotonergic systems [e.g., via selective serotonin reuptake inhibitors (SSRIs), serotonin receptor-specific modulators, etc.] may offer promise for improved preeclampsia therapeutics. Repurposing of SSRIs for use in non-psychiatric indications, for example in viral infection, is not unprecedented.[61, 62] SSRI action on nervous system, vascular, and immune mechanisms makes it a promising candidate for preeclampsia therapy.

Serotonergic dysregulation is an area of significant mechanistic overlap between mood disorders and preeclampsia, disorders which share notable clinical comorbidity.[63, 64] For instance, both mood disorders and preeclampsia are associated with significant serotonin-mediated vascular dysfunction. Vascular hyporeactivity to serotonin has been reported previously in the uterine and other arteries of preeclamptic patients,[65, 66] while endothelial dysfunction biomarkers are enhanced in depression patients in a serotonin-dependent manner.[67] Platelet aggregation in preeclamptic hyperserotonemia[68] may exacerbate spiral artery dysfunction in preeclampsia pathogenesis. Depression similarly features platelet reactivity and aggregation disruptions, which are modified by SSRI use.[67, 69, 70] Serotonin also plays a significant role in preeclampsia and mood disorder immune mechanisms, including T cell dysregulation, kynurenine pathway metabolism, and cytokine production abnormalities, as was recently reviewed.[12]

Serotonin modulates preeclampsia mechanisms

Recent debates in the field of psychiatry have ignited around the mixed evidence linking brain serotonin dysfunction to depression. SSRI efficacy, however, is undoubtedly high for the treatment of depression and anxiety.[39] This implores the exploration of addition, non-canonical, neuronal mechanism of SSRI action, as we discuss and implicate in preeclampsia pathogenesis here. In fact, evidence suggests that chronic SSRI use reduces cardiovascular disease risk.[71] These drugs may be preeclampsia-protective via action on the vascular, platelet, immune, and central nervous system serotonergic mechanisms that are disrupted in preeclampsia.

Vascular hyporeactivity to serotonin has been reported in the preeclamptic uterine, placental, and other vessels, though underlying mechanism(s) remain largely unclear.[65, 66, 72, 73] Animal models offer some insights. SSRI administration in sheep transiently decreases uterine blood flow and fetal oxygen saturation, which recover then remain unchanged over a longer-term period of eight days.[74] Chronic SSRI administration in dogs over a 14-day period potentiates pulmonary vasoconstriction in response to endothelin-1, which is elevated in preeclampsia and may mediate the renin-angiotensin system suppression, hypertension, and the proteinuria associated with preeclampsia.[75] However, in patients with cardiovascular morbidity, SSRI use is associated with decreased blood pressure reactivity to stress.[76] Changes in vasoreactivity in response to chronic SSRI administration in human pregnancy require further examination.

In preeclampsia, endothelial cell damage and platelet aggregation cause enhanced local serotonin release, contributing to systemic platelet-free hyperserotonemia and further vascular disease.[77, 78] It is possible that the platelet aggregation and activation abnormalities that occur in depression make it a risk factor for preeclampsia.[69, 70, 79, 80] These abnormalities may be corrected by SSRI use: among chronic SSRI users, platelet serotonin content and platelet aggregation are decreased relative to non-users.[71] Similarly, decreased aggregation has been reported with in vitro SSRI treatment of human platelets[81] and in SSRI-users (versus non-users) undergoing antiplatelet therapy for myocardial infarction.[82] Likewise, ex vivo studies have demonstrated inhibited platelet aggregation and activation in the presence of an SSRI.[83, 84] Chronic SSRI depletion of platelet serotonin[85] and downregulation of platelet SERT[86, 87] may drive platelet effects. These results demonstrate potential biological means by which SSRI treatment in the context of preeclampsia may decrease platelet aggregation and consequential impacts on vascular health and physiology.

Timing may be another critical factor underlying the impacts of SSRI use on pregnancy health and outcomes. Evidence suggests differential impacts of SSRIs on early versus late pregnancy.[30, 68] For instance, antidepressant and/or anxiolytic exposure before the 16^th week of pregnancy significantly increases risk for preeclampsia, with risk increased more than three-fold if SSRI use is extended further into pregnancy (aOR 3.41 ± 0.21-12.34 95%CI).[88] Neonatal impacts of SSRI exposure also differ relative to the timing of this exposure. For example, NICU admission rates (e.g., for pulmonary hypertension, hypoglycemia, etc.) are significantly increased among babies whose mothers are SSRI-treated during late (16.5%) versus early (10.8%) pregnancy.[89] This may reflect physiological differences between early placentation processes and later-term, established vasculature, as well as fetal developmental stage. The timing of SSRI treatment should therefore be accounted for in future work.

In addition to vascular and platelet abnormalities, preeclampsia involves immune disruptions including systemic pro-inflammation and decreased metabolism of tryptophan by indolamine-2,3-dioxygenase (IDO).[90] Serotonin manipulation in other pro-inflammatory animal models decreases inflammation,[91] suggesting some promise for the application of SSRI treatment to pro-inflammation in preeclampsia. In humans, the expression of IDO and select cytokines (e.g., pro-inflammatory interleukin-6), which are generally increased in MDD patients, is attenuated with eight-weeks of SSRI treatment.[92, 93] In patients with a first episode generalized anxiety disorder, 12 weeks of treatment with SSRIs similarly decreases serum pro-inflammatory cytokine levels.[94] Depression during peripartum is similarly pro-inflammatory, systemically and in the brain.[95–97] In vivo, SSRI treatment blocks upregulation of IDO in response to a pro-inflammatory insult in brain cells and brain macrophages, called microglia, as well as ex vivo in murine and human peripheral blood mononuclear cells.[98] IDO expression and cytokine production are regulated by the JAK/STAT1,[99] P38/MAPK,[100] and TLR4/NF-κB signaling pathways,[101] and further work is needed to determine precisely how or if SSRIs interact with these pathways to regulate pro-inflammation in the context of preeclampsia. However, none of the studies in the present met-analysis included measures of chronic inflammation. Future work may pair psychiatric, obstetric, and peripheral inflammatory measures to determine inflammatory mediator interactions with SSRI efficacy and cardiovascular pathology.

In addition to manipulating underlying biological mechanisms, is also possible that SSRIs counter preeclampsia by treating underlying central nervous system disruptions or mood disorders and their associated behavioral symptoms, which themselves increase preeclampsia risk.[16] SSRIs have direct impacts on hypothalamic arginine vasopressin (AVP) expression, which is disrupted in preeclampsia and a biomarker of the disease,[7, 102] as well as the glucocorticoid stress response regulated by AVP.[103] More functionally, sedentary behavior, which is associated with preeclampsia,[104] is also increased in depression.[105] Relatedly, SSRI treatment has been shown to decrease fatigue among depressed individuals,[106] though its known side effects include fatigue and weight gain in some patients. Life stress is similarly associated with preeclampsia risk[107] and depression,[108] while SSRIs blunt stress reactivity.[107] Together, these associations between preeclampsia and depression demonstrate that, while there are clear biological mechanisms by which SSRIs may directly mitigate preeclampsia physiology, central nervous system and behavioral impacts likely also play an important, protective role.

Additional evidence supporting a role for serotonin modulation in the treatment of preeclampsia is derived from studies of ketanserin, a HTR₂ antagonist. HTR₂ is expressed in the nervous system, platelets, fibroblasts, monocytes, and elsewhere. The clinical efficacy of ketanserin for the treatment of preeclampsia was first indicated in 1984.[109] In a randomized controlled trial, ketanserin decreased preeclampsia and severe hypertension in pregnant women and non-significantly decreased perinatal mortality.[110] In patients with gestational hypertension, ketanserin successfully decreases systemic blood pressure but does not change placental perfusion (measured via Doppler flow) or fetal outcomes.[111, 112] In spontaneously hypertensive rats, it ameliorates systolic hypertension and decreases placental blood flow.[113] In 5-HTP-treated rats, ketanserin rescues preeclampsia phenotypes such as plasma volume constriction, proteinuria and intrauterine growth restriction.[114] Multiple reviews have further detailed the promise of HTR₂ antagonism via ketanserin medication for the treatment of preeclampsia.[111, 115, 116] Despite its potential benefits for preeclampsia, ketanserin is also associated with cardiovascular side-effects including dysrhythmias, and approximately one-third of ketanserin users experience prolongation of the QT interval.[117] These side effects have limited its clinical application, and ketanserin is not FDA approved.[118] However, targeting of HTR₂ signaling may remain an appropriate strategy for the mitigation of preeclampsia risk.

Conclusions

In the present review, we detailed interactions between serotonin dysregulation and the pathogenesis of preeclampsia, including impacts on vascular, platelet, immune, and nervous system biology. Animal and in vitro models may be leveraged to better understand serotonin targets in the context of this biology to develop improved therapeutics. We found that SSRI use in pregnancy may contribute to preeclampsia risk, though the literature supporting this conclusion is significantly confounded. In fact, despite a long-standing focus on the brain serotonin modulation impacts of SSRIs for the treatment of depression, mounting evidence suggests alternative mechanisms (e.g., vascular, immune) for SSRI efficacy. These same mechanisms may be implicated in gestational hypertensive disorders, and there is biological plausibility for serotonin modulation in the treatment for preeclampsia, especially in subsets of patients with serotonin dysregulation. To clarify whether these therapeutics are themselves preeclampsia protective, it is necessary to study SSRI impacts on hypertension and preeclampsia endophenotypes in psychiatrically healthy patients. Such study may reveal additional treatment targets, as well as dispel provider and patient concerns about the use of SSRIs in pregnancy. Future research advances will clarity this critical issue and promote mechanistic efforts to determine and prevent the deleterious impacts of serotonin dysregulation and preeclampsia on mothers and babies.

Highlights.

• Reviewed studies linking SSRIs to preeclampsia(PE)/gestational hypertension (GH)

• SSRIs may modulate pathoetiology of hypertensive disorders of pregnancy

• Pooled SSRI-associated PE/GH risk ratio was 1.43 (95% CI: 1.15-1.78)

• Studies demonstrate bias and poorly control for cofactors (dose, comorbidities)

• Future studies should utilize model organisms or SSRIs in healthy participants