Prenatal psychological distress and 11 β -HSD2 gene expression in human placentas: Systematic review and meta-analysis

Background: The placenta acts as a buffer to regulate the degree of fetal exposure to maternal cortisol through the 11-Beta Hydroxysteroid Dehydrogenase isoenzyme type 2 ( 11-β HSD2 ) enzyme. We conducted a systematic review and meta-analysis to assess the effect of prenatal psychological distress (PPD) on placental 11-β HSD2 gene expression and explore the related mechanistic pathways involved in fetal neurodevelopment. Methods: We searched PubMed, Embase, Scopus, APA PsycInfo ® , and ProQuest Dissertations for observational studies assessing the association between PPD and 11-β HSD2 expression in human placentas. Adjusted regression coefficients ( β ) and corresponding 95% confidence intervals (CIs) were pooled based on three contextual PPD exposure groups: prenatal depression, anxiety symptoms, and perceived stress. Results: Of 3159 retrieved records, sixteen longitudinal studies involving 1869 participants across seven countries were included. Overall, exposure to PPD disorders showed weak negative associations with the placental 11-β HSD2 gene expression as follows: prenatal depression ( β (cid:0) 0.01, 95% CI 0.05 – 0.02, I2 = 0%), anxiety symptoms ( β (cid:0) 0.02, 95% CI 0.06 – 0.01, I2 = 0%), and perceived stress ( β (cid:0) 0.01 95% CI 0.06 – 0.04, I2 = 62.8%). Third-trimester PPD exposure was more frequently associated with lower placental 11-β HSD2 levels. PPD and placental 11-β HSD2 were associated with changes in cortisol reactivity and the development of adverse health outcomes in mothers and children. Female-offspring were more vulnerable to PPD exposures. Conclusion: The study presents evidence of a modest role of prenatal psychological distress in regulating placental 11-β HSD2 gene expression. Future prospective cohorts utilizing larger sample sizes or advanced statistical methods to enhance the detection of small effect sizes should be planned. Additionally, controlling for key predictors such as the mother ’ s ethnicity, trimester of PPD exposure, mode of delivery, and infant sex is crucial for valid exploration of PPD effects on fetal programming.


Introduction
The placenta serves as a channel for transferring oxygen and vital nutrients to the developing fetus and discarding waste.It helps safeguard fetal development by regulating collateral substances, toxins, and hormones that come into reach with the in-utero environment and fetal circulation (Ruffaner-Hanson et al., 2022).This includes crucial chemicals in maternal circulation, such as cortisol (Galbally et al., 2021;Osborne et al., 2018).Cortisol is a bioactive form of glucocorticoids in humans involved in maintaining homeostasis and adapting the body to different physiological challenges (Lee et al., 2015;Ruffaner-Hanson et al., 2022).The production and secretion of cortisol are mediated by the hypothalamic-pituitary-adrenal (HPA) axis in response to both internal and external stressors, including disruptions in the sleep-wake cycle (Lightman et al., 2020).
Cortisol acts on a broad spectrum of tissues and organ systems, including the brain, liver, kidneys, heart, pancreas, muscles, and adipose tissues, serving a variety of critical functions beyond its role in regulating stress response (Wandja Kamgang et al., 2023).Among these include the modulation of hemodynamic changes and underlying inflammatory processes, immune response regulation, glucose metabolism, and muscle growth (Lightman et al., 2020;Shields et al., 2015;Wandja Kamgang et al., 2023).This prompted investigations into cortisol levels as an indicator of various health conditions.For example, evidence suggests that greater cortisol concentrations are associated with heightened susceptibility to poor cardiovascular outcomes (Barugh et al., 2014;Mohd Azmi et al., 2021).Disturbances in cortisol secretion have also been linked to the pathophysiology of mental health disorders and other cognitive and behavioral deficits (Dziurkowska and Wesolowski, 2021;Lightman et al., 2020;Murphy et al., 2022).
Research shows that elevations in maternal prenatal cortisol levels correspond to increases in both fetal and offspring cortisol levels (Agapaki et al., 2022;Osborne et al., 2018;Zijlmans et al., 2015).Physiological increase in cortisol during pregnancy plays a pivotal role in fetal organogenesis.However, sustained exposure to elevated cortisol concentrations can have a detrimental impact on the fetal programming of various physiological systems (Baibazarova et al., 2013;Buss et al., 2012;Davis and Sandman, 2012;Lewis et al., 2015;Zijlmans et al., 2015).Therefore, a close monitoring and intricate balance of cortisol secretion remain crucial for promoting maternal and fetal well-being (Rinne et al., 2023).
Distress refers to heightened levels of emotional discomfort that are not confined to a medically defined mental disorder (American Psychiatric Association, 2013; Middleton and Shaw, 2000), while psychological distress is recognized as a multifaceted construct often ascribed to "non-specific" group of symptoms of depression, anxiety, and various forms of stress (APA, 2013;Drapeau et al., 2012;Viertiö et al., 2021;Wu et al., 2020).Research shows heightened comorbidity of depression, anxiety, and stress during pregnancy (Cena et al., 2021;Obrochta et al., 2020;Premji et al., 2020).Pregnant women are at higher risk of psychosocial issues, with pooled global data demonstrating the prevalence of antenatal depression to be as high as 65% and 17% in developing and high-income nations, respectively (Mareckova et al., 2020).A growing body of literature demonstrates that PPD could drive long-term adverse effects on the offspring's functional, neurological, and behavioral development (Davis et al., 2007;Fekadu Dadi et al., 2020;Mareckova et al., 2020;Osborne et al., 2018;Turney, 2011).The risk of diagnosing any mental or behavioral disorder was up to two-fold higher in children born to pregnancies complicated by antenatal depression, anxiety, or perceived stress (Leis et al., 2014;Tuovinen et al., 2021).
Limited empirical evidence is available on the underlying mechanisms and channels through which maternal PPD could affect in-utero fetal development and offspring health outcomes.One biological paradigm proposes a link between early life exposure during critical windows of development (i.e., fetal and early infancy phases) and long-term health complications, namely the Developmental Origins of Health and Disease (DoHAD) model (Kubota et al., 2015;O'Donnell and Meaney, 2016).The DoHAD concept explores the influences related to human experiences during critical perinatal windows of development, including exposures in the in-utero environment.These include exposures mediating in-utero epigenetic modifications, which involve changes in placental gene expression without altering the underlying DNA sequence, including DNA methylation and histone modification (Heindel et al., 2017;Holme and Sitras, 2020;Paquette et al., 2016).The manifestation of PPD has been linked to such epigenetic changes.
For instance, maternal depression during pregnancy has been shown to increase glucocorticoid levels in the fetus (i.e.cortisol), which contributed to DNA methylation of genes involved in early neurogenesis and altered placental glucocorticoid signaling pathways (Kubota et al., 2015;Mansell et al., 2016;Provençal et al., 2020).As a result, offspring exposed to increased glucocorticoid levels in utero may have an exaggerated response to stress exposures in adulthood.Epigenetic abnormalities at an early stage can also be manifested in women who experience psychological distress before or during pregnancy and can be inherited by future generations (Bush et al., 2023;Galbally et al., 2019;Kubota et al., 2015;Richetto and Meyer, 2021).A decade of research suggests that epigenetic processes mediated the adverse effects of prenatal PPD on altered in-utero programming, including fetal central nervous system, metabolic functions, and physical development (Paquette et al., 2014;Ciesielski et al., 2015;Monk et al., 2016;Wu et al., 2020).This included the DNA methylation of major stress response genes, including the 11-β HSD2.
Most findings related to the molecular mechanisms mediating adverse fetal neurodevelopment are based on investigations from animal studies.Particularly, the downregulation of the 11-β HSD2 gene has been well-established in animal models (Galbally et al., 2021;Grégoire et al., 2020;Jensen Peña et al., 2012;Ruffaner-Hanson et al., 2022;Welberg et al., 2005).Yet, there is currently no summary of evidence on the effect of prenatal psychological distress on the 11-β HSD2 gene expression in human studies.Therefore, this review aimed to summarize findings on the association between prenatal psychological distress and the expression and regulation of the 11-β HSD2 gene in human placentas.The research also aims to explore the mechanistic pathways through which the placental expression of 11-β HSD2 can influence fetal neurodevelopment in pregnant women with PPD disorders.Summarizing findings from population-based cohort studies that investigate the role of epigenetic changes could help identify targets for early-life interventions that can prevent the development of diseases (Felix and Cecil, 2019).

Methods
This systematic review and meta-analysis was conducted according to the guidance on conducting systematic reviews and meta-analyses of observational studies of etiology (COSMOS-E) and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) Guidelines (Dekkers et al., 2019;Page et al., 2021).

Search strategy
We searched PubMed, Embase, Scopus, APA PsycInfo®, and Pro-Quest Dissertations for observational research, including cohort, casecontrol, cross-sectional, and registry-based analysis studies (Fig. 1).We included studies that reported on the relationship between PPD disorders, including depression, anxiety, pregnancy trait anxiety, posttraumatic stress disorder, and placental expression of the 11-β HSD2 gene in humans.Studies that reported on stress exposure solely due to states of illness, such as cardiometabolic disorders or obesity, or life events, such as the death of a loved one or socioeconomic challenges, were excluded.We excluded case reports, case series, reviews, and abstracts with insufficient outcome data on the relationship between PPD and 11-β HSD2 placental gene expression.The search ran from the 22nd to the 25th of February 2023, with no restrictions on language or year of publication.The detailed search strategy for each database can be found in Supplementary Table 1.

Study selection and data extraction
Retrieved records were imported into the Rayyan web tool (htt ps://www.rayyan.ai/),and duplicates were removed.The titles and abstracts were independently screened by three reviewers (A.T., S.E., and F.E.), followed by full-text screening.Conflicts were resolved by consulting a fourth reviewer (G.B.).The three reviewers independently extracted data into a standardized, pre-piloted form.We extracted data on study characteristics, sample sizes, participants' characteristics, exposure measures, statistical analyses used to assess the relationship between PPD and 11-β HSD2 placental expression, and the modeled confounders, if any.All effect estimates of correlation or association (i.e., β-regression, Pearsons's correlation r, or Spearman's rho coefficients) were extracted along with corresponding 95% confidence intervals (CIs), standard errors, and p-values as applicable for the systematic review.Only adjusted regression coefficients were used for the metaanalysis.Any disagreements or discrepancies were resolved via discussion and consensus after consulting expert reviewers (G.B., T.W., and M. H.). Study authors were contacted when needed for additional information or clarifications.

Methodological quality assessment
Included studies were assessed for methodological quality using the MethodologicAl STandards for Epidemiological Research (MASTER) scale (Stone et al., 2021).The MASTER scale assesses the broad mechanisms by which systematic error (bias) may compromise the internal validity of studies, i.e., design deficiencies, selection bias, information bias, and analysis deficiencies.The scale scores studies by assigning a score of 1 for the presence and 0 for the absence of 36 unique safeguards aligned with seven methodological standards.These are equitable recruitment, retention, ascertainment, consistent implementation, unbiased prognosis, comprehensive analysis, and established temporal precedence (Stone et al., 2021).Total scores were averaged, and records were ranked based on the highest-scoring study, demonstrating the propensity towards bias of a study compared to the pool of studies included in the review.Methodological quality was assessed by at least two reviewers for each study.

Data synthesis
The outcome of the study selection process is demonstrated using the PRISMA 2020 flow diagram (Fig. 1).We provide a descriptive summary of all included studies in tables and in the text.Study characteristics and findings were described qualitatively.Additionally, meta-analyses of the association between PPD and placental 11-β HSD2 gene expression were conducted using the Quality Effects Model (QEM) (Doi and Thalib, 2008).The QEM incorporated bias adjustment using a score from the MASTER scale for each study.
Analysis was stratified by the context of the three commonly identified PPD disorders across eligible studies: prenatal depression, anxiety symptoms, and perceived stress.Despite the well-established comorbidity of depression, anxiety, and stress symptomatology during pregnancy, each represents a unique facet of PPD and may have varying implications on maternal and fetal health outcomes (Obrochta et al., 2020).Consequently, independent analyses were conducted for each construct to elucidate their differential impacts on placental 11-β HSD2 gene expression.Adjusted regression coefficients and their corresponding 95% CIs were pooled.One study provided supplementary raw data, which we used to calculate adjusted effect estimates (EEs) and 95% CIs via multiple linear regression analysis (i.e., Zhang et al., 2020).Heterogeneity across studies was assessed using the I2-Statistic; an I2 Statistic value of 25%, 50%, and 75% indicated low, moderate, and high heterogeneity, respectively (Higgins et al., 2003).Results are displayed in forest plots.
Sub-group analysis was planned based on pre-determined variables (i.e., trimester of PPD measurement and infant sex).However, the number of studies was insufficient (<2 studies per subgroup) to conduct a meaningful analysis (Cuijpers et al., 2021;Higgins et al., 2020).Instead, the two dimensions were descriptively summarized in the text.The Doi plot and Luis Furuya-Kanamori (LFK) index were used to visualize and quantify asymmetry of study effects, respectively, to assess publication bias (Furuya-Kanamori et al., 2018;Furuya-Kanamori and Doi, 2021).LFK indexes below ± 1, above ± 1 and less than ± 2, and above ± 2 indicated no, minor, and major asymmetry, respectively.Data were analyzed using the "metan" commands on STATA/SE version 18.0.The meta-analysis was conducted by A.T. and T.C.

Study characteristics
All included records (n=16) were prospective longitudinal cohort studies published between 2011 and 2022.Most studies were conducted in Australia (n=4), Sweden (n=3), and the United States (n=3).The total sample size from all 16 studies is 1869 participants (N=24 to N=303).Prenatal depression, anxiety symptoms, and perceived stress were the most frequently investigated PPD conditions.A wide range of instruments was used to measure these exposures across studies, of which the Edinburgh Pre/Postnatal Depression Scale (EPDS) was the most utilized tool (n=12), followed by the State and Trat Anxiety Inventory (STAI) (n=8) and the Perceived Stress Scale (PSS) (n=4).Three studies investigated PPD as an aggregate scale, combining different PPD measures (St-Pierre et al., 2018;Togher et al., 2017;Zhang et al., 2020).
Study characteristics are summarized in Supplementary Table 2.

Participants characteristics
The time of participant recruitment and follow-up periods varied greatly across studies (i.e., early gestation versus mid-to-late gestation versus one day prior to delivery).Not all studies, however, reported a full spectrum of participant characteristics.The average age of mothers across the studies ranged from 27 ±6.0 to 35 ±5 years (n=15), while the average BMI ranged from 22.8 ±3.5 to 26.5 ±4.95 kg/m 2 (n=9).Around 33.4-94.4% of participants were married during data collection (n=8).The average gestational age at birth across studies ranged from 38.4 ±1.2 to 40.43 ±1.2 weeks (n=15), and the average birth weight ranged from 3268 ±594 to 3712 ±469 grams (n=11).The predominance of the offspring sex varied across studies, whereby the prevalence of female newborns ranged from 43% to 61% (n=12).Zhang et al., (2020) achieved the highest MASTER score (=27).Hence, it served as the reference study against which all other records were ranked.Table 1 shows the total scores and relative ranking of all 16 studies.Ponder et al., 2011, Galbally et al., 2022;Jahnke et al., 2021;Martinez et al., 2020 andHellgren et al., 2016 achieved the lowest MASTER scores, mainly due to failure to adjust for confounders (Galbally et al., 2022;Hellgren et al., 2016;Jahnke et al., 2021;Martinez et al., 2020;Ponder et al., 2011), information biases (Galbally et al., 2022;Hellgren et al., 2016;Ponder et al., 2011), and selection biases (Galbally et al., 2022;Jahnke et al., 2021;Martinez et al., 2020;Ponder et al., 2011).Studies that scored higher were more likely to identify and adjust for important confounders, have higher sample sizes, report on appropriate cut-offs for measuring exposure variables, and address relevant design limitations.A more detailed scoring on individual safeguards across studies can be found in Supplementary Table 3.

Maternal and offspring cortisol outcomes
Five studies reported outcomes on maternal and infant cortisol levels and reactivity (Galbally et al., 2022(Galbally et al., , 2021;;Hellgren et al., 2016;Jahnke et al., 2021;Togher et al., 2018).Table 3 summarizes the types of samples utilized and the direction of findings.Significant findings on cortisol-related outcomes are reported in both directions, entailing both an increase and a decrease in cortisol levels or reactivity secondary to PPD exposure or placental 11-β HSD2 gene expression.

Fig. 2. Forest plot of adjusted effect sizes for the association between prenatal depression, anxiety symptoms, and perceived stress and the placental 11-β HSD2 gene expression.
A) The effect of prenatal depression summarized from five studies, B) The effect of prenatal anxiety summarized from four studies, and C) The effect of prenatal perceived stress summarized from four studies using the Quality Effects Model.The MASTER rank is the study's MASTER score relative to the highest-scoring study in this analysis (i.e., Zhang et al., 2020)-95% CI: 95% confidence interval.
prenatal depression on lowering 11-β HSD2 placental levels achieved significance only in female-stratified samples (β= − 1.74).Similarly, a substantial increase in 11-β HSD2 gene methylation was observed only in the female offspring of depressed mothers (β= 6.74) (Jahnke et al., 2021).Togher et al., (2018) reported a sex-specific reduction in birth weights of male infants exposed to second and third trimester maternal anxiety (β= − 0.43) and stress disorders (β= − 0.40), respectively.In contrast, the same study found that prenatal anxiety throughout pregnancy predominantly reduced the birth weight of female newborns (β= − 0.50).Meanwhile, Martinez et al., (2020)) demonstrated a more pronounced upregulation in 11-β HSD2 expression in male placentas of PPD-exposed samples.Moreover, Galbally et al., 2022 found a higher risk of emotional disorders in female children who demonstrated low 11-β HSD2 expression and high cortisol reactivity.

Discussion
In this review of 16 longitudinal studies, we found that exposure to prenatal psychological distress was weakly associated with placental 11β HSD2 gene expression.In overall analyses, prenatal depression, anxiety symptoms, and perceived stress were associated with small decreases (Betas of − 0.02 to − 0.01) in placental 11-β HSD2 levels.
The effect of prenatal perceived stress showed moderate heterogeneity (I 2 =62.8%), indicating that other study-specific factors may have influenced the impact of stress exposure on placental 11-β HSD2 levels.Likewise, despite the absence of statistical heterogeneity (I 2 =0%) for the effect of prenatal depression and anxiety symptoms, clinical and methodological sources of heterogeneity evident across the included studies may have influenced the variability observed across effect sizes.Those include inconsistencies in defining the target population, using distinct tools to measure the exposure or outcome, and variations in individuals' genetic markup.For example, most studies (N=14) administered selfreported tools to assess PPD, subjecting analyses to potential measurement errors (Capron et al., 2018;Edvinsson et al., 2019;Galbally et al., 2022Galbally et al., , 2021;;Jahnke et al., 2021;Martinez et al., 2020;O'Donnell et al., 2012;Reynolds et al., 2015;Seth et al., 2015;St-Pierre et al., 2018;Togher et al., 2018Togher et al., , 2017;;Zhang et al., 2020Zhang et al., , 2018)).The two studies that captured PPD solely using medical records or standardized diagnostic interviews showed no difference in placental 11-β HSD2 levels between exposure groups (Hellgren et al., 2016;Ponder et al., 2011).
The periods of measuring PPD exposures were also inconsistent across studies, suggesting potential variations in 11-β HSD2 enzyme regulation in response to acute versus prolonged psychological distress in pregnancy (Welberg et al., 2005).For example, Seth et al., (2015), Reynolds et al., (2015), St-Pierre et al., (2018) and Togher et al., (2018) assessed PPD at multiple time points throughout pregnancy.Meanwhile, O'Donnell et al., (2012) and Capron et al., (2018) measured PPD one day before cesarean section, which may have influenced women's mental states in anticipation of the surgery.Measurement error, especially when non-differential, is known to bias results towards the null, possibly explaining the modest effect sizes observed (Jurek et al., 2008(Jurek et al., , 2005)).The small scale of effects is a prevalent observation in environmental epigenetic research particularly linked to "exposure-associated differences" (Breton et al., 2017, p. 519).Nonetheless, some studies have reasoned that these findings reflect a true minimal effect (Bollati et al., 2011;Huen et al., 2014;Lee et al., 2015;Michels et al., 2011); the impact of measurement error cannot be overlooked.
Methodological differences in placental 11-β HSD2 gene expression measurements were also observed.The time to placental sample collection varied across the 16 studies, ranging from zero to up to 3 hours after birth.Investigations show deterioration in RNA integrity and distortion in gene expression with delays in placental sample collection time, suggesting a cap at 2 hours post-delivery (Freedman et al., 2021;Wolfe et al., 2014).Similarly, factors such as infant sex, mode of delivery, and sampling sites have been shown to influence gene expression levels (Chatterjee et al., 2022;Janssen et al., 2015).Only two studies included in the meta-analyses adjusted for infant sex and mode of delivery (Reynolds et al., 2015;Zhang et al., 2020), while none controlled for the sampling site (Table 2).Additionally, the publication bias across all three meta-analyses, as evident by the asymmetry in the Doi plots and LFK indexes, indicates a potential distortion in the overall effect sizes, warranting caution in interpreting our results.
The absence of prior meta-analytical syntheses limits the ability to compare our findings directly against a broader body of evidence.The inconsistency of findings reported by the individual research records in our analysis could explain the small effect sizes observed.For instance, three of the five studies included in our meta-analysis for the effect of prenatal depression reported an upregulation in 11-β HSD2 gene expression (Capron et al., 2018;Reynolds et al., 2015;Togher et al., 2018).Similar trends were demonstrated by one and two studies included in our meta-analyses for the effect of prenatal anxiety and perceived stress, respectively (Togher et al., 2018;Zhang et al., 2020).This is similar to findings reported on prenatal exposure to different forms of psychosocial distress (Bulka et al., 2023;2020;Lund et al.,  et al., 2020;Paquette et al., 2014;Ruffaner-Hanson et al., 2022).The magnitude of effects may also be confounded by other clinical factors and gene-environment interactions (Dieckmann and Czamara, 2024;Sosnowski et al., 2018;Szyf, 2019).Genetic variations among individuals may influence the placental response to maternal PPD, contributing to variability in results (Anaya et al., 2023;Paquette et al., 2014).Another explanation could be ascribed to the failure to account for relevant effect modifiers by individual research records, including the regulation of other placental genes, such as the glucocorticoid receptor (NR3C1), the serotonin transporter (SLC6A4), and the mineralocorticoid receptor (NR3C2) genes, hence concealing the real effect of PPD on 11-β HSD2 gene expression (Capron et al., 2018;Mansell et al., 2016;Ponder et al., 2011;Stroud et al., 2016).Furthermore, findings from Capron et al., 2018 (Supplementary Table 2) show significant downregulation of placental 11-β HSD2 gene expression solely in Caucasian mothers (β= − 0.381; P= 0.008), indicating that ethnicity could be an important variable in predicting the association between PPD and placental 11-β HSD2 gene expression.
We also observed trimester-specific variations in the effect of PPD on the fetoplacental barrier.Specifically, exposure to prenatal depression, anxiety, and perceived stress in the third trimester was reported in a relatively higher number of studies (n=7) associated with the downregulation of placental 11-β HSD2 levels compared to 5 studies reporting the outcome in earlier trimesters.This could suggest that the timing of exposure to PPD during pregnancy significantly influences the molecular processes in the placenta.This is consistent with earlier studies that demonstrated significant behavioral and emotional abnormalities in offspring who were specifically exposed to maternal PPD in the third trimester (Davis et al., 2007;O'Connor et al., 2003;Werner et al., 2007).However, the majority of studies included in our review measured PPD at a single time point during pregnancy (n=11), yielding only preliminary conclusions on the influence of third-trimester PPD exposure on placental function and child outcomes.
Our review also explores the associations between PPD, placental 11β HSD2 gene expression, and cortisol levels.Varying cortisol response patterns in both mothers and infants were associated with placental 11-β HSD2 gene regulation and adverse outcomes in offspring exposed to PPD.The effect of prenatal psychopathology on placental function was previously explored by Glover et al., 2009, who found significant correlations between cortisol concentrations in maternal blood and amniotic fluid that were more pronounced in anxious women (Glover et al., 2009), supporting the potential role of PPD in fetal reprogramming.The significance of cortisol reactivity as an underlying mechanism linking prenatal PPD and neurobehavioral development in children is vastly discussed in the literature (Baibazarova et al., 2013;Braithwaite et al., 2016;Buss et al., 2012;Kortesluoma et al., 2022).However, the use of different types of biological samples (e.g., blood, saliva, hair, and urine), cortisol measurement protocols, and outcome measures across the individual studies included in our review (Table 3) entails careful consideration in the interpretation of the mixed findings observed.This is consistent with an earlier systematic review which attributed the varied associations between maternal stress, cortisol, and child outcomes to inconsistencies related to cortisol assessment methods (Zijlmans et al., 2015).
Respectively, saliva sampling has become the gold standard for assessing short-term or acute cortisol reactivity, owing to its noninvasive nature and capacity to provide a more precise measurement of bioactive-free cortisol compared to blood sampling (Harville et al., 2007;McGuinn et al., 2022;Oaks et al., 2022;Ryan et al., 2016;Stalder et al., 2016).However, external factors may influence fluctuations in saliva cortisol concentrations, including sampling time, circadian rhythm, physical exercise, underlying infections, and medication use (Bozovic et al., 2013;Kortesluoma et al., 2021;Lee et al., 2015;Pritchard et al., 2017).Such fluctuations may vary considerably within a given day and across different days, undermining the reliability of single time-point measurements in accurately depicting systemic stress exposure (Lee et al., 2015).Repeated cortisol measures (i.e., ≥3-5 samples throughout the day over the course of ≥2 consecutive days) at standardized time points are recommended to provide a more reliable interpretation of stress exposure over a defined period, while minimizing between (sampling time)-and within (day-to-day)-person variability (Harville et al., 2007;McGuinn et al., 2022;Oaks et al., 2022;Orta et al., 2019;Rinne et al., 2023;Ryan et al., 2016;Stalder et al., 2016).Selecting and reporting adequate cortisol outcome measures is also suggested, particularly cortisol awakening response, diurnal slope, and area under the curve (McGuinn et al., 2022;Ryan et al., 2016).
Recent prenatal research also suggests the role of hair cortisol as a more inclusive measure for detecting prolonged stress exposure throughout pregnancy (Agapaki et al., 2022;Galbally et al., 2019;Mustonen et al., 2018).This is underpinned by the slow growth rate of hair samples, stable cortisol levels, standardized collection methods across participants, greater practicality, and lower risk of attrition (Adam and Kumari, 2009;Khoury et al., 2023;Murphy et al., 2022;Mustonen et al., 2018).Still, hair samples may fail to reflect the day-to-day changes in the effect of PPD on cortisol levels, which can be addressed by collecting repeated salivary cortisol samples (Khoury et al., 2023;Orta et al., 2019).Measuring hair cortisol at specific time windows may also be unreliable when the scales used to measure PPD do not represent stress at comparable periods (Khoury et al., 2023).For instance, a single measurement on the EPDS scale at 36 weeks gestation evaluates the severity of depressive symptoms over the past 7 days preceding assessment (i.e. from 35 to 36 weeks gestation).Meanwhile, at a growth rate of one centimeter per month (Khoury et al., 2023;Murphy et al., 2022), a 3-centimeter hair sample collected at 36 weeks reflects cumulative cortisol levels over three months (i.e., from 24 to 36 weeks gestation).Variations in cortisol reactivity patterns and PPD scores across trimesters delineate the importance of capturing and accounting for gestational age (in weeks) at both sample collection and PPD assessment (Cherak et al., 2018;McGuinn et al., 2022).
Six included records further demonstrated strong associations between prenatal depression, anxiety, and perceived stress exposure and a range of adverse maternal and child health outcomes.Similar findings have been well established in recent systematic reviews and metaanalyses (Bussières et al., 2015;Lima et al., 2018;Madigan et al., 2018;Manzari et al., 2019), underscoring the importance of addressing maternal psychosocial well-being during pregnancy.Sex-specific effects of PPD exposure were also explored.Specifically, we observed lower placental 11-β HSD2 expression and a higher risk of negative child outcomes in female-stratified samples (Galbally et al., 2022;Jahnke et al., 2021;Togher et al., 2018).In parallel, a systematic review of 21 observational studies observed a higher vulnerability of female offspring to maternal stressors, in which they were more susceptible to HPA axis programming (Carpenter et al., 2017).
To our knowledge, this is the first review synthesizing findings on the association between prenatal psychological distress and the expression of the placental 11-β HSD2 gene in human pregnancies.No restrictions were applied on the language or date of publications, in addition to exploring the grey literature, ensuring a comprehensive review of the available evidence.The decision to include only studies with adjusted regression coefficients in the meta-analyzes was made to ensure the robustness and comparability of the findings, allowing the control for potential confounding variables, and providing a more accurate estimation of the association under investigation.We also evaluated publication bias using highly sensitive methods, such as the Doi plot and the LFK indexes, given the small number of studies used in our analysis (n<10) (Furuya-Kanamori et al., 2018).
Despite our careful analysis, several limitations should be acknowledged.First, the small number of studies available for meta-analysis (≤5) may have limited the accurate estimation of between-study heterogeneity, resulting in biased effect estimates (von Hippel, 2015).Additionally, small study effects influenced our analysis, as evident by the Doi plots and LFK indexes.However, the pooled results did not detect significant changes in placental 11-β HSD2 gene expression.Second, all records included in our review are observational studies, which are inherently characterized by methodological variations (Metelli and Chaimani, 2020), potentially introducing variability into our results.Third, we excluded studies investigating PPD exposures solely attributed to traumatic life events.This could have influenced the understanding of the relationship between the temporality of PPD and the outcomes studied.Moreover, the sample sizes in half of the included studies were relatively small (<100), potentially affecting the precision of our estimates (Capron et al., 2018;Jahnke et al., 2021;Martinez et al., 2020;O'Donnell et al., 2012;Reynolds et al., 2015;Seth et al., 2015;St-Pierre et al., 2018;Togher et al., 2018).
Lastly, despite the weak effect sizes observed in our analyses, other findings in our review show strong associations between PPD exposure and adverse outcomes in offspring.Earlier research demonstrated that even small changes in gene expression could have negative consequences on infant growth and neurobehavioral outcomes (Green et al., 2015;Kappil et al., 2015), suggesting the potential role of PPD in mediating adverse child effects through epigenetic processes observed in our review.Mitigating long-term health adversity in future generations has potential implications for early identification, prevention, and management of psychological distress in pregnancy.Our findings further underscore the need for a larger evidence base of well-conducted study designs.Large-scale prospective studies should be planned with meticulous attention paid to estimating adequate sample sizes for appropriate effect size while minimizing the margin of error.For example, Campagna et al., 2021, Serdar et al., 2021and Yin et al., 2022 provide guidance regarding the importance of sample size and its relationship to effect size, including planning for sufficient sample size to detect smaller changes in differentially expressed genes (DEGs) (Campagna et al., 2021;Serdar et al., 2021;Yin et al., 2022).
However, in the absence of larger sample sizes, research proposes the application of advanced analysis methods to improve the robustness of DEGs identification in small datasets, such as the "limma" model and the joint "meta-analysis, SVM-RFE (support vector machines-recursive feature elimination), and permutation test (MSPJ)" method (Campagna et al., 2021;Smyth, 2005;Vasiliu et al., 2015;Yin et al., 2022).Controlling for a comprehensive set of relevant covariates could also drive robust statistical analysis.Those could include infant sex, trimester of PPD exposure (i.e., gestational age at PPD assessment), mother's ethnicity, mode of delivery, and expanded genetic profiles.Future research could also benefit from exploring interaction effects (Aschard, 2016), such as PPD exposure and gene expression, on offspring health outcomes, which could warrant more biologically plausible conclusions.Still, standard analyses of observational data often overlook uncertainties beyond random error, leading to incomplete judgments about biases (Greenland, 2005).Future studies should consider multiple-bias analysis by systematically incorporating key uncertainties to enhance the reliability of research and policy insights (Greenland, 2021;Lash et al., 2014).
Finally, it is important to note that 93% of the included studies were conducted in high-income Western countries.This is rather predictable, given the magnitude of resources required for biological sampling, sequencing, and quantification of gene expression and regulation.However, with reference to the pronounced burden of PPD disorders in low-and middle-income countries compared to developed nations (Mareckova et al., 2020), international funding bodies must remobilize sufficient resources for such critical investigations in developing countries.

Conclusion
Our review found weak evidence for the association between prenatal depression, anxiety, and perceived stress and the expression of the 11-β HSD2 gene in human placental tissues.Still, it highlights the implications of PPD exposure on fetal stress-response systems and longterm disease development.Given the costly and time-intensive nature of genetic sequencing, which limits routine clinical applications, identifying target biomarkers to assess adverse outcomes in the offspring of mothers with PPD could facilitate more tailored investigations and timely interventions.Future investigations involving temporal precedence in causal inference, such as prospective cohorts, should aim to improve the power of detecting small to moderate effects by utilizing proper statistical techniques or larger sample sizes.Future research should also control for significant predictors including the mother's ethnicity, trimester of PPD exposure, mode of delivery, and infant sex, enhancing the validity of exploring the effects of PPD exposure on fetal reprogramming.

Fig. 1 .
Fig. 1.PRISMA 2020 flow diagram of the study selection process for the systematic review.Three reports were conference abstracts and were not retrieved for full-text screening after contacting the authors.Sixteen studies were included in the systematic review.*All records imported into Rayyan software; APA: American Psychiatry Association; n¼ number of records.

Table 2
Data for meta-analyses of the effect of PPD on 11-β HSD2 placental expression.
Zhang et al., 2020)the study's MASTER score relative to the highest-scoring study in the analysis (i.e.,Zhang et al., 2020).BMI: body mass index; CI: confidence interval; combined: trimesters 2 and 3; EE: effect estimate; EPDS: the Edinburgh Pre/Postnatal Depression Scale; CES-D: the Centre for Epidemiological Studies Depression Scale; STAI: State-Trait Anxiety Inventory; PSS: Perceived Stress Scale; LEQ: The Life Events Questionnaire; UK: United Kingdom; USA: United States of America.

Table 3
Summary of findings on cortisol-related outcomes (N=5).