1. Introduction
1.1 Background and theoretical framework
The Developmental Origins of Health and Disease (DOHaD) paradigm, grounded in the fetal programming hypothesis (Barker, 1995), posits that the intrauterine environment shapes the developmental trajectory of the fetus, with lasting implications for health across the life course. During gestation, the maternal organism undergoes normative neuroendocrine fluctuations that facilitate physiological adaptation to pregnancy. It is essential to draw a clear ontological and physiological distinction between this adaptive stress response and chronic prenatal distress. Whereas the former constitutes a necessary allostatic response, chronic distress represents a pathological exposure characterised by the failure of the maternal system to restore homeostasis (Van den Bergh et al., 2020; Monk et al., 2019; Gluckman et al., 2008).
The biological mechanism underlying this transmission of adversity involves dysregulation of the maternal hypothalamic–pituitary–adrenal (HPA) axis. Chronic distress induces sustained cortisol secretion that overwhelms the buffering capacity of the placental enzyme 11β-hydroxysteroid dehydrogenase type 2 (11β-HSD2) (Glover, 2014). This downregulation exposes the fetus to supraphysiological glucocorticoid levels, catalysing epigenetic modifications — particularly methylation of the glucocorticoid receptor gene (NR3C1). These structural and functional alterations preferentially affect the development of glucocorticoid receptor-rich regions, including the limbic system and the prefrontal cortex (Monk et al., 2012; Moog et al., 2017; Sandman et al., 2012).
Fetal vulnerability to these neuroendocrine insults exhibits marked temporal sensitivity. Exposure during the first trimester is predominantly associated with alterations in cognitive development and learning, whereas distress in the third trimester tends to manifest as phenotypes of emotional reactivity and difficult temperament (Bergman et al., 2010; Davis & Sandman, 2010; Sandman et al., 2012). To capture this chronic exposure, hair cortisol concentration (cHCC) has emerged as the gold-standard retrospective biomarker, enabling quantification of systemic glucocorticoid accumulation over extended periods and thereby overcoming the limitations of single time-point measures (Caparros-Gonzalez et al., 2017; Russell et al., 2012).
1.2 Rationale and research gaps
Despite the proliferation of studies on prenatal stress, the literature lacks a rigorous synthesis that simultaneously integrates psychological and biological assessment. No existing systematic review or meta-analysis has jointly evaluated maternal distress using validated self-report instruments and HPA-axis biomarkers as dual exposure measures. This omission precludes determining whether variance in neurodevelopmental outcomes is attributable to the subjective perception of distress or to the objective allostatic burden (Huizink et al., 2003; Laplante et al., 2004; Riquelme-Gallego et al., 2023).
Furthermore, prior research has disproportionately prioritised infant outcomes, relegating maternal neuropsychological outcomes in the postpartum period (Kingston et al., 2012). The quality of the mother–infant dyadic interaction — encompassing constructs such as maternal sensitivity and attachment — remains underrepresented in models of fetal programming, despite its critical role as a mediator of infant development (Thiel et al., 2021). The absence of pre-specified moderator analyses comparing measurement modality (self-report versus biomarker) constitutes a methodological gap that this protocol aims to address.
1.3 Objectives
Primary objective:
To quantify the association between chronic prenatal maternal stress (assessed via validated psychometric instruments and/or HPA biomarkers) and neurodevelopmental outcomes in infants at 6, 12, and 24 months of age.
Secondary objectives:
1. To compare the predictive validity of subjective stress assessments (psychometric scales) versus objective biomarker-based assessments (cHCC and related HPA measures) for infant neurodevelopmental outcomes.
2. To evaluate the association between chronic prenatal maternal stress and maternal neuropsychological functioning in the postpartum period.
3. To examine the relationship between chronic prenatal maternal stress and the quality of mother–infant dyadic interaction (sensitivity, responsiveness, attachment).
4. To investigate potential moderating factors including trimester of exposure, sex of the infant, gestational age, and country income level (HIC vs. LMIC).
2. Methods
2.1 Protocol and registration
This methodological protocol was designed in accordance with the PRISMA-P guidelines (Preferred Reporting Items for Systematic review and Meta-Analysis Protocols) (Shamseer et al., 2015). The protocol is pre-registered in the Open Science Framework (OSF; https://doi.org/10.17605/OSF.IO/2Y3RC) and published in Open Research Europe as its formal dissemination record.
2.2 Eligibility criteria (PECOS framework)
Studies will be eligible for inclusion if they meet the following PECOS criteria.
| P
| Pregnant women ≥18 years (any trimester, parity, and nationality) with a formal prenatal stress assessment. No restriction on gestational age or country of origin. |
| E
| Chronic prenatal maternal stress assessed by: (a) validated psychometric tools (PSS-10, EPDS, STAI, PRAQ-R, PDQ, life-event scales) or (b) HPA-axis biomarkers (hair cortisol concentration [cHCC], salivary/serum/plasma cortisol, cortisol awakening response [CAR], urinary catecholamines). |
| C
| Women with low or normative prenatal stress levels on the same measurement instruments (validated scales or biomarker values within reference ranges). |
| O
| Primary: Neurodevelopmental outcomes in infants at 6, 12, and 24 months (Bayley-III, ASQ-3, CBCL, Mullen Scales). Secondary: maternal neuropsychological postpartum outcomes; mother–infant dyadic interaction quality (sensitivity, responsiveness, attachment). |
| S
| Longitudinal observational designs (prospective cohorts) and retrospective designs (case-control). Peer-reviewed articles and preprints. No language or date restrictions. Animal studies, purely clinical/obstetric outcomes, and cross-sectional designs excluded. |
Exclusion criteria: Animal or in vitro studies; participants <18 years; studies assessing stress exclusively in the postpartum period; outcomes limited to obstetric parameters (gestational age, birthweight, Apgar); cross-sectional designs; secondary literature (narrative reviews, editorials).
2.3 Information sources and search strategy
The selection of information sources follows the guidance set out in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins et al., 2022) and the PRESS 2015 Evidence-Based Checklist for Librarian Peer Review of Electronic Search Strategies (McGowan et al., 2016), with the dual aim of maximising both sensitivity and specificity. The seven databases searched encompass biomedical, psychological, nursing, and Latin American scientific literature, thereby ensuring adequate coverage of the heterogeneous disciplinary landscape in which research on prenatal distress, HPA axis dysregulation, and infant neurodevelopment is published. The search queries were constructed by combining controlled vocabulary terms—MeSH headings in PubMed, Emtree terms in Scopus, and thesaurus descriptors in PsycINFO and CINAHL—with free-text terms applied to title and abstract fields. This dual approach is methodologically necessary: reliance solely on controlled indexing terms would systematically exclude studies indexed inconsistently or prospectively, particularly those from non-English or emerging research groups, a recognised limitation that has been extensively documented in the information retrieval literature.
Two distinct rounds of searching were conducted. The first round (November 2025) employed the full primary search string across all seven databases, retrieving a combined total of 2,083 records. The second round (February 2026) was a targeted complementary search, deliberately designed to capture studies on mother–infant interaction, maternal sensitivity, and attachment outcomes that may have been underrepresented in the first round owing to the keyword emphasis on cortisol concentrations and neuropsychological outcomes. This second round contributed an additional 725 records across five databases (PubMed, PsycINFO, Web of Science, Scopus, and EBSCO). Across both rounds, searches yielded a combined total of 2,808 raw records. Following automated deduplication within each database platform prior to export and cross-database deduplication in Rayyan (Ouzzani et al., 2016), the final unique record set submitted for title and abstract screening comprised 1,476 records.
Table 1 presents the complete search record by database, interface, and round.
Table 1. Records retrieved per database across both search rounds.
| Database | Interface/platform | Search 1 date | Records (S1) | Search 2 date | Records (S2) | Total |
|---|
| PubMed/MEDLINE | NCBI | 25 Nov 2025 | 21 | 03 Feb 2026 | 7 | 28 |
| PsycINFO | ProQuest | 26 Nov 2025 | 416 | — | 481* | 897 |
| Web of Science Core Collection | Clarivate Analytics | 27 Nov 2025 | 426 | 05 Feb 2026 | 128 | 554 |
| Scopus | Elsevier | 27 Nov 2025 | 258 | 05 Feb 2026 | 48 | 306 |
| CINAHL | EBSCOhost | 25 Nov 2025 | 7 | — | — | 7 |
| LILACS/BVS | VHL Regional Portal | 27 Nov 2025 | 448 | — | — | 448 |
| EBSCO (Academic Search Ultimate) | EBSCOhost | 27 Nov 2025 | 507 | 05 Feb 2026 | 61 | 568 |
| TOTAL
| | | 2,083
| | 725
| 2,808
|
The primary search string is structured around three conceptual Boolean blocks. Concept block A captures prenatal stress and adversity exposure, incorporating both psychological constructs (perceived stress, psychological distress, anxiety) and physiologically operationalised adversity (perinatal stress, intrapartum stress, maternal stress transfer). Concept block B retrieves neurodevelopmental and neuropsychological outcomes in the offspring and in the mother herself, covering cognitive development, motor milestones, language acquisition, executive function, emotion regulation, and behavioural outcomes. Concept block C functions as a biological pathway precision filter, targeting studies that operationalise HPA axis activity through cortisol measurement—whether salivary, urinary, serum, or hair—or through related glucocorticoid markers. The decision to include Concept block C as a mandatory element of the primary string reflects the conceptual specificity of this review: by requiring all three blocks to co-occur, the strategy retrieves studies that explicitly link prenatal psychosocial adversity to offspring outcomes through a measurable biological mechanism, substantially reducing the retrieval of studies that address behavioural correlates of prenatal stress without any neuroendocrine component.
The second-round search relaxed Concept block B and replaced it with dyadic and relational outcome terms—mother–infant interaction, maternal sensitivity, maternal responsiveness, parent–child interaction, attachment, bonding, and caregiving behaviour—while retaining Concept block A (prenatal stress exposure) and Concept block C (cortisol/HPA markers). This design reflects the recognition that studies examining maternal sensitivity and mother–infant dyadic quality as outcomes of prenatal neuroendocrine dysregulation occupy a partially distinct literature from those focused on infant neurodevelopmental milestones, and would not be reliably retrieved by the primary string alone.
Across all databases, the Boolean NOT operator was applied exclusively to publication types (systematic reviews, meta-analyses, narrative reviews) and methodological designs incompatible with the longitudinal exposure–outcome framework of this review (cross-sectional studies). Critically, NOT was not applied to content terms, in order to prevent inadvertent exclusion of relevant primary studies in which secondary study type terms appear incidentally in title or abstract fields. This conservative application of exclusionary logic is consistent with best-practice recommendations for high-sensitivity systematic review search design.
In addition to the primary database searches, grey literature was systematically canvassed through several complementary channels. Open Grey (www.opengrey.eu) and ProQuest Dissertations & Theses Global were searched to retrieve unpublished reports, conference proceedings, and doctoral dissertations relevant to the review topic. Clinical trial registries—specifically ClinicalTrials.gov and the International Standard Randomised Controlled Trial Number (ISRCTN) registry—were searched to identify ongoing or recently completed trials that may yield eligible data prior to formal publication. To supplement database retrieval, the reference lists of all studies included after full-text review will be manually screened (backward citation tracking), and forward citation tracking will be conducted via Google Scholar and Web of Science for key included studies, in order to capture relevant work not indexed in the primary databases.
Deduplication was performed in two sequential stages. First, within each database platform, automated deduplication was carried out prior to record export to remove intra-database duplicates arising from multiple index versions of the same article. Second, cross-database deduplication was performed in Rayyan (Ouzzani et al., 2016), an AI-assisted web and mobile application purpose-built for systematic review management, which applies smart deduplication algorithms to identify and merge duplicate records across sources. Following completion of both stages, the final set of unique records submitted for title and abstract screening comprised 1,476 records.
The complete search strings for all databases and both search rounds are provided in Supplementary Material S1.
2.4 Study selection process
Study selection will follow PRISMA 2020 guidelines (Page et al., 2021). Two independent reviewers will screen titles and abstracts using the Rayyan platform (Ouzzani et al., 2016). Articles passing the screening stage will be evaluated at full-text level by the same two reviewers. Discrepancies will be resolved by consensus; in cases of persistent disagreement, a senior third reviewer (R.A. Caparrós-González or M.A. Muñoz-Garcia) will adjudicate. Where required data are missing, corresponding authors will be contacted following a standardised two-attempt protocol with a 15–30-day interval.
2.5 Data extraction
Data extraction will be performed independently by two investigators (V.M. Salmeron-Vega and L. Leonardo-Encabo) using a standardised, pilot-tested form. The following variables will be extracted: bibliometric information; sample sociodemographic characteristics; methodological details of exposure assessment (instrument type, trimester of assessment, biomarker type and protocol); neurodevelopmental assessment tools and time points; statistical estimators required for meta-analysis (means, standard deviations, effect sizes, confidence intervals, correlation coefficients); and data relevant to dyadic quality outcomes.
2.6 Risk of bias
The methodological quality and risk of bias of included observational studies will be assessed using the ROBINS-E tool (Risk Of Bias In Non-randomized Studies – of Exposures; Sterne et al., 2019). Seven critical domains will be examined: (1) confounding; (2) selection of participants; (3) classification of the exposure; (4) co-exposures; (5) missing data; (6) measurement of outcomes; and (7) selection of the reported result. Assessments will be conducted independently by two reviewers; discrepancies will be resolved by consensus.
2.7 Statistical synthesis and meta-analytic strategy
Quantitative analyses will be performed in the R programming environment (version ≥4.3.0) using the meta and metafor packages. A random-effects model with the DerSimonian and Laird (1986) estimator will be used to calculate between-study variance (τ2), with the restricted maximum likelihood (REML) method applied in sensitivity analyses. Effect sizes will be standardised using Fisher’s r-to-z transformation for correlations and Hedges’s g for standardised mean differences between groups.
Statistical heterogeneity will be quantified using Cochran’s Q test and the I2 statistic (Higgins et al., 2003), interpreting I2 values of 25%, 50%, and 75% as indicative of low, moderate, and high heterogeneity, respectively. Meta-regression and subgroup analyses have been pre-specified to explore sources of heterogeneity, including: (a) trimester of exposure; (b)
measurement modality (self-report vs. biomarker); (c) neuropsychological domain assessed; (d) infant age at outcome measurement; and (e) country income level (HIC vs. LMIC) (Baroutis et al., 2025).
2.8 Assessment of meta-biases
Publication bias and small-study effects will be evaluated visually through contour-enhanced funnel plots and statistically by Egger’s test, provided that a minimum of 10 studies per outcome are available. The trim-and-fill method will be applied to impute theoretically missing studies and assess the robustness of pooled estimators. The Doi plot and LFK index will be used as a complementary non-asymmetry
test.
2.9 Certainty of the evidence
Overall confidence in the body of evidence will be assessed using the GRADE framework (Grading of Recommendations Assessment, Development and Evaluation). Observational evidence will begin at a ‘low’ certainty rating and may be downgraded for risk of bias, inconsistency, indirectness, imprecision, or publication bias, or upgraded in the presence of large effect sizes (OR > 5 or < 0.2) or a dose–response gradient.
3. Discussion
3.1 Expected contributions
This systematic review will provide the first quantitative synthesis directly contrasting the predictive validity of subjectively perceived maternal stress against objective allostatic burden measured via HPA-axis biomarkers. By integrating the assessment of the mother–infant dyad, the findings will clarify whether alterations in infant neurodevelopment operate independently or are mediated by disruptions in maternal sensitivity and postpartum attachment.
From a clinical and applied perspective, these findings may inform screening strategies in perinatal settings by clarifying whether subjective or biological measures of stress should be prioritised, or whether a combined approach provides greater predictive value. This has direct implications for early detection and targeted intervention in populations at risk.
3.2 Limitations
Considerable methodological heterogeneity is anticipated, arising from variability in the psychometric instruments employed and biomarker collection protocols across primary studies. Furthermore, the observational nature of the available literature limits the capacity to draw definitive causal inferences, requiring cautious interpretation of the pooled effect sizes.
Additionally, variability in biomarker assessment protocols (e.g., hair segment length, timing of sampling, and laboratory procedures) may limit comparability across studies. Publication bias and selective reporting may also influence the observed effect sizes, particularly in studies reporting significant associations. These factors should be considered when interpreting the findings.
3.3 Dissemination
The results of this systematic review will be submitted for publication in a high-impact international peer-reviewed journal in psychoneuroendocrinology or paediatrics. Findings will be presented at relevant national and international scientific conferences. Executive summaries for perinatal health professionals and policy-relevant briefs will be produced to ensure translation of findings into clinical practice.
Ethics and consent
Ethical approval and consent were not required. This is a systematic review and meta-analysis protocol involving secondary analysis of previously published data. No primary data collection involving human participants will be undertaken as part of this study.
Registration
This protocol is published in Open Research Europe. The formal pre-registration is deposited in OSF (https://doi.org/10.17605/OSF.IO/2Y3RC).
Ethics
No ethics committee approval is required for this systematic review, as no primary data are collected. Only published literature will be analysed.
Data availability
No primary data are generated during this protocol phase.
Acknowledgements
The authors acknowledge the institutional support provided by the University of Granada and the Instituto de Investigación Biosanitaria ibs. GRANADA. The authors thank the COST Action CA22114 TREASURE network for intellectual exchange and methodological support.
In accordance with the F1000 AI Policy, the authors disclose that AI-assisted tools (Perplexity Computer, November 2025 – April 2026) were used to support organizational tasks during the preparation of this manuscript, including screening workflow management and document structuring. All scientific content, analytical decisions, and conclusions were developed, reviewed, and approved exclusively by the human authors.
References
- Barker DJP:
Fetal origins of coronary heart disease.
BMJ.
1995; 311(6998): 171–174. PubMed Abstract
| Publisher Full Text
| Free Full Text
- Baroutis G, Sotiropoulou IM, Mantzioros R, et al.:
Prenatal maternal stress and long-term neurodevelopmental outcomes: A narrative review.
J. Perinat. Med.
2025; 53(9): 1159–1171. Publisher Full Text
- Bergman K, Sarkar P, O'Connor TG, et al.:
Maternal stress during pregnancy predicts cognitive ability and fearfulness in infancy.
Biol. Psychiatry.
2010; 67(11): 1026–1032. PubMed Abstract
| Publisher Full Text
| Free Full Text
- Caparros-Gonzalez RA, Romero-González B, Peralta-Ramírez MI, et al.:
Hair cortisol levels, psychological stress and psychopathological symptoms as predictors of postpartum depression.
Psychoneuroendocrinology.
2017; 84: 47–53. Publisher Full Text
- Davis EP, Sandman CA:
The timing of prenatal exposure to maternal cortisol and psychosocial stress is associated with human infant cognitive development.
Child Dev.
2010; 81(1): 131–148. PubMed Abstract
| Publisher Full Text
| Free Full Text
- DerSimonian R, Laird N:
Meta-analysis in clinical trials.
Control. Clin. Trials.
1986; 7(3): 177–188. Publisher Full Text
- Glover V:
Maternal depression, anxiety and stress during pregnancy and child outcome; what needs to be done.
Best Pract. Res. Clin. Obstet. Gynaecol.
2014; 28(1): 25–35. Publisher Full Text
- Gluckman PD, Hanson MA, Cooper C, et al.:
Effect of in utero and early-life conditions on adult health and disease.
N. Engl. J. Med.
2008; 359(1): 61–73. Publisher Full Text
- Higgins JPT, Thomas J, Chandler J, et al.: (
2022). Manual Cochrane para revisiones sistematicas de intervenciones (version 6.3).Cochrane.https://www.training.cochrane.org/Handbook
- Higgins JPT, Thompson SG, Deeks JJ, et al.:
Measuring inconsistency in meta-analyses.
BMJ.
2003; 327(7414): 557–560. PubMed Abstract
| Publisher Full Text
| Free Full Text
- Huizink AC, Robles de Medina PG, Mulder EJ, et al.:
Stress during pregnancy is associated with developmental outcome in infancy.
J. Child Psychol. Psychiatry.
2003; 44(6): 810–818. Publisher Full Text
- Kingston D, Tough S, Whitfield H:
Prenatal and postpartum maternal psychological stress and infant development: A systematic review.
Child Psychiatry Hum. Dev.
2012; 43(5): 683–714. Publisher Full Text
- Laplante DP, Barr RG, Brunet A, et al.:
El estres durante el embarazo afecta el funcionamiento intelectual y linguistico general en ninos pequenos humanos.
Pediatric Research.
2004; 56(3): 400–410. PubMed Abstract
| Publisher Full Text
- McGowan J, Sampson M, Salzwedel DM, et al.:
Revision por pares PRESS de estrategias de busqueda electronica: declaracion de la guia 2015.
Journal of Clinical Epidemiology.
2016; 75: 40–46. PubMed Abstract
| Publisher Full Text
- Monk C, Spicer J, Champagne FA:
Linking prenatal maternal adversity to developmental outcomes in infants: The role of epigenetic pathways.
Dev. Psychopathol.
2012; 24(4): 1361–1376. PubMed Abstract
| Publisher Full Text
| Free Full Text
- Monk C, Lugo-Candelas C, Trumpff C:
Prenatal developmental origins of future psychopathology: Mechanisms and pathways.
Annu. Rev. Clin. Psychol.
2019; 15: 317–344. PubMed Abstract
| Publisher Full Text
| Free Full Text
- Moog NK, Buss C, Entringer S, et al.:
Maternal exposure to childhood trauma is associated during pregnancy with placental-fetal stress physiology.
Neuroscience.
2017; 342: 68–100. PubMed Abstract
| Publisher Full Text
| Free Full Text
- Ouzzani M, Hammady H, Fedorowicz Z, et al.:
Rayyan — a web and mobile app for systematic reviews.
Syst. Rev.
2016; 5(1): 210. PubMed Abstract
| Publisher Full Text
| Free Full Text
- Page MJ, McKenzie JE, Bossuyt PM, et al.:
The PRISMA 2020 statement: An updated guideline for reporting systematic reviews.
BMJ.
2021; 372: n71. Publisher Full Text
- Riquelme-Gallego B, Navarro-Membrives M, Aguilera-Manrique G, et al.:
Prenatal stress and its effects on the neuropsychological development of infants: A systematic review.
J. Reprod. Infant Psychol.
2023. Publisher Full Text
- Russell E, Koren G, Rieder M, et al.:
Hair cortisol as a biological marker of chronic stress: Current status, future directions and unanswered questions.
Psychoneuroendocrinology.
2012; 37(5): 589–601. PubMed Abstract
| Publisher Full Text
- Salmeron-Vega VM, Leonardo-Encabo L, de la Torre-Luque A , et al.:
Maternal Prenatal Stress and Its Neuropsychological Sequelae in Infants and Mother–Infant Dyadic Quality: Study Protocol Materials. [Dataset].
Open Science Framework.
2026a. Publisher Full Text
- Salmeron-Vega VM, Leonardo-Encabo L, de la Torre-Luque A , et al.:
PRISMA-P 2015 Compliance Checklist [Extended Data 1].
Zenodo.
2026b. Publisher Full Text
- Salmeron-Vega VM, Leonardo-Encabo L, de la Torre-Luque A , et al.:
List of Abbreviations [Extended Data 2].
Zenodo.
2026c. Publisher Full Text
- Salmeron-Vega VM, Leonardo-Encabo L, de la Torre-Luque A , et al.:
Supplementary Material S1: Full Boolean Search Strings for All Databases [Extended Data 3].
Zenodo.
2026d. Publisher Full Text
- Sandman CA, Davis EP, Buss C, et al.:
Exposure to prenatal psychobiological stress exerts programming influences on the mother and her fetus.
Neuroendocrinology.
2012; 95(1): 8–21. Publisher Full Text
- Shamseer L, Moher D, Clarke M, et al.:
Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015: Elaboration and explanation.
BMJ.
2015; 349: g7647. Publisher Full Text
- Sterne JAC, Hernán MA, McAleenan A, et al.:
Assessing risk of bias in a non-randomized study.Higgins JPT, Thomas J, Chandler J, et al., editors.
Cochrane Handbook for Systematic Reviews of Interventions.
Cochrane;
2019.
- Thiel F, Eberhard-Gran M, Garthus-Niegel S:
The impact of perinatal life stress on infant temperament and development: A 2-year follow-up cohort study.
J. Dev. Behav. Pediatr.
2021; 42(4): 299–306. PubMed Abstract
| Publisher Full Text
- Van den Bergh BRH, van den Heuvel MI , Lahti M, et al.:
Prenatal developmental origins of behavior and mental health: The influence of maternal stress in pregnancy.
Neurosci. Biobehav. Rev.
2020; 117: 26–64. Publisher Full Text
