Association between prenatal maternal anxiety and/or stress and offspring’s cognitive functioning: a systematic review and meta-analysis


 The relationship between prenatal maternal stress and/or anxiety and children’s cognitive development is inconclusive. This systematic review and meta-analysis aims to examine the relationship between the different domains of stress, anxiety and children’s cognitive outcomes during the first 18 years of life. Five databases were searched for all observational studies investigating the association between symptoms of prenatal maternal anxiety and/or stress and children’s cognitive outcomes. Of 7,004 articles identified, 21 met the inclusion criteria and 11 provided usable data that were further analysed quantitatively. A weak negative association was found between prenatal stress and/or anxiety exposure and the offspring’s language development. Associations varied based on the type of prenatal psychological exposure (i.e., trait anxiety, state anxiety, perceived stress, stress response). Specifically, maternal prenatal exposure to stress but not anxiety was found to correlate with children’s general intellectual skills. Moreover, trait anxiety but not the combined exposure to trait and state anxiety was correlated with attention in the offspring. While exposure to stress and/or anxiety at any stage of pregnancy was found to be associated with adverse cognitive outcomes in the offspring, there was a trend that timing of exposure may be associated with distinct cognitive outcomes. Findings support the need for screening and interventions to prevent or minimise mental health problems in pregnant women in order to optimise child development. Findings also have implications for conceptualising prenatal stress and anxiety in future research, as well as investigating timing and cumulative effects of prenatal stress and/anxiety exposure.


Introduction
Many women do not receive appropriate mental health checks during pregnancy, despite the high rate of maternal stress and anxiety in the prenatal period which affects up to 78% and 18-22% of pregnant women worldwide respectively 1,2,3,4,5 . Studies have reported that stress and anxiety are associated with negative health and neurobehavioural outcomes in the offspring 6,7,8,9 , while systematic reviews have found prenatal maternal stress/anxiety to be related to infant (0-2 years) cognitive development 10,11,12,13 . Studies examining prenatal stress and anxiety typically pooled these exposure factors, which is a limitation if the nature of the association with cognition varies across these different maternal psychological constructs. Thus, future research should 1) separates the concepts of prenatal anxiety and prenatal stress, and 2) examines a range of cognitive outcomes beyond infancy.
Stress can be operationalised and measured as perceived stress or a stress response 14 . Perceived stress is an emotional reaction characterised by feelings or thoughts that arise from individuals' appraisal of life stressors and vary based on several factors, including one's social support, personality traits and coping behaviours 14,15,16 . A stress response is most commonly characterised as biological changes in cortisol levels triggered by a perceived stressful event. It is mediated by factors such as personality traits and cognitive appraisal processes 17,18 .

Page 3/32
Anxiety is also a multi-dimensional construct, comprising state and trait anxiety. State anxiety refers to situation-speci c and transient negative emotion characterised by psychological and physiological reactions, such as irritability, di culties concentrating, palpitations and a reduction in serotonin release 19,20,21 . Trait anxiety refers to an innate and enduring characteristic of a person which predisposes them to experience more intense degrees of fear and worry across many situations that most people would not nd threatening 22 . Pregnancy-related anxiety has also been described, which refers to worries speci cally related to pregnancy, such as giving birth or bearing a handicapped child 23 .
While related, perceived stress, stress response, trait anxiety, state anxiety and pregnancy-related anxiety constitute unique psychological constructs with distinct neurobiological underpinnings, causes and outcomes 14,22,24,25 . Accordingly, these different psychological constructs are assessed using different methods 16,26,27 .
Studies have shown a lack of correlation between 1) the different aspects of stress, types of anxiety, and 2) these constructs and health and birth outcomes 14,22,28,29 . As such, exposure to these different constructs may have a differential impact on the cognitive development of the child. For instance, research has shown that prenatal exposure to stress but not anxiety has a negative impact on attentional control in childhood, whilst prenatal exposure to trait but not state anxiety has been associated with poorer speech-language development 30,31 .
This review aimed to examine the relationship between prenatal maternal stress and anxiety and a range of key cognitive domains (i.e., general intellectual skills, attention, language, learning, memory, executive functions) in infancy, childhood, and adolescence (birth until 18 years of age). Unlike previous reviews, we will also separately examine the in uence of the different aspects of stress and anxiety. We hypothesised that there would be an association between maternal prenatal stress and anxiety and offspring's cognitive domains, but expected the association with cognition would differ for the different psychological constructs of perceived stress, stress response, trait anxiety, state anxiety and pregnancy-related anxiety.

Methods
Protocol and search strategy The study protocol was registered on PROSPERO (CRD42020185906). A search strategy was conducted for papers published before the date of the search (May 2020) on PsycINFO, MEDLINE, Web of Science Core Collection, CINAHL and EMBASE databases. The search strategy included key-terms related to the prenatal period, maternal anxiety, maternal stress, childhood and cognitive development (see supplement materials). The literature search was limited to humans and studies published in English. There were no criteria for years of publication. Additional articles were identi ed using reference list from previous meta-analyses and review articles.

Inclusion and exclusion criteria
Prospective and retrospective observational studies investigating the association between symptoms of prenatal maternal anxiety and/or stress and children's cognitive outcomes were considered for inclusion. Measures of prenatal maternal anxiety and/or stress included validated physiological measures and self-report questionnaires. The outcome measures were validated age-standardised measures of general intelligence, learning, memory (long-term free recall, long-term recognition), language (receptive, expressive), speed of information processing, attention (span, divided, switching, sustained, selective), executive functions (working memory, inhibition, planning, organisation, self-monitoring, decision making, cognitive exibility, word generation) and academic skills (reading, maths, writing). Both direct assessment and parent and teacher-rated questionnaires were included if relevant.
Studies were excluded for the following reasons: (i) the study was a treatment or clinical trial, case-control study, review, comment, letter, thesis or book, (ii) individual cognitive skills were not assessed separately, (iii) the offspring were older than 18 years old or speci cally selected based on a previous diagnosis (e.g. autism and ADHD), (iv) anxiety and/or stress was examined in the postnatal period, (v) a measure of anxiety and/or stress was not included, (vi) measures of anxiety and stress were combined in the analyses, (vii) children's cognitive function was assessed using experimental measures only, and (iix) the study involved non-human participants.
When the same cognitive domain was measured at multiple time points in the same cohort or when different measures of a same cognitive domain were examined, preference was given to the largest sample size. When two timepoints were of similar sample size, the timepoint where children were the eldest was given preference. If the above two criteria were the same, the cognitive outcome evaluated by the most valid or widely used measure in the eld was used 32,33 . For the overall analysis, when a study reported multiple stress and/or anxiety exposures, preference was given to the to the exposure that was the most comparable to those included in the analysis.

Study selection
All the references retrieved using the strategy search outlined above were downloaded and deduplicated into Endnote before being transferred, stored and managed into Covidence. Titles and abstracts were screened independently by two reviewers, GD and SV. Full texts of potentially relevant articles were obtained and independently assessed for inclusion by the two reviewers using the previously described criteria. In case of disagreement, the reviewers discussed eligibility. When disagreement could not be resolved, PJA made the nal decision. Ninety-nine percent agreement was reached after the rst screen of titles / abstracts, and 91% after the rst screen of full texts. After discussion, 100% agreement was reached by the three reviewers.

Data extraction
A customised extraction excel sheet adapted from the Cochrane Review handbook was used to guide extraction. Reviewers (GD and SV) extracted relevant information independently and discussion was organised in case of discrepancy. When available, extracted information included: Independent variable: Measures used to assess maternal anxiety and/or stress during pregnancy; anxiety levels during pregnancy; stress levels during pregnancy; trimester(s) of exposure.
Outcome variables: Tasks; Performance on cognitive tasks; unadjusted and adjusted effect sizes Confounding factors: Covariates considered and adjusted for in analysis.
Methodology: Participants' demographic information; Participant attrition rate; study design; times of measurement, sample size.

Bias and quality assessment
The quality of the studies selected was assessed by GD using a modi ed version of the Scottish Intercollegiate Guideline Network appraisal tool (SIGN) criteria for cohort studies (https://www.sign.ac.uk/sign-50). SIGN is a widely used tool which appraises studies by examining six domains: study design, withdrawals and drop-outs, potential for selection bias, measurement of outcomes and exposure factors, confounders and report of statistical analyses, and blinding. Using a scoring algorithm adapted from the Quality Assessment Tool for Quantitative Studies (http://www.ephpp.ca/tools.html), each of these domains were rated as being of low, moderate or strong quality, and an overall study quality rating was made based on the quality of each domain. Studies encompassing four or more strong and no weak domains were rated as strong; those with less than four strong and no more than one weak domain were rated as moderate; and those with more than two weak domains were rated as weak.

Meta-analytic procedures
Using Comprehensive Meta-Analysis Version 3, a series of standard meta-analyses of correlations estimated the pooled associations between prenatal maternal anxiety and/or stress and children's cognitive function.
Due to the variability in characteristics and sample sizes across studies, the more conservative randomapproach was used 34 . Separate meta-analyses were performed when at least two studies provided an effect for an outcome 35 . Results were strati ed by type of exposure (i.e., stress and anxiety) and cognitive domain (i.e., general intellectual function, language, attention, learning and memory and working memory). While pregnancy-related anxiety was found to constitute a valid and reliable construct 36,37 , the scales used to measure this psychological construct assess the nature of their anxiety or worries rather than the type and severity of symptoms of anxiety 30,36,37,38 . As such, studies examining this construct were not included in the main analyses examining the relationship between all types of anxiety and/or stress combined and children's later cognitive outcomes to maximise consistency. When enough data was available, results were strati ed by type of anxiety and stress exposure (i.e., perceived stress, stress response, trait anxiety and state anxiety) and timing of exposure (trimester 1, 2 or 3). Although the original intent was to assess associations for different developmental periods, this was not possible due to the small number of eligible studies which examined stress and/or anxiety at different timepoints during pregnancy. Consequently, analyses were performed combining children from all age groups to increase power and were also performed for the infancy period only.
A measure of effect size (coe cient of correlation r) was calculated for each study. A correlational effect size r of .05 was considered as very small, .10 as small, .30 as medium, .30 as large, and .50 or greater as large 39 .
The statistical signi cant level used was α = 0.05 . Both I 2 statistic and Tau squared were computed to assess for heterogeneity of effect sizes 34 . I 2 re ects the percentage of variability in the correlational effect size that is attributed to heterogeneity rather than sampling error. A value of 25% or below can be interpreted as a small amount of heterogeneity, while values of 50% and 75% or above represent moderate and high heterogeneity, respectively 40 . Tau squared represents the variance of the true correlational effect size across studies.

Study Selection, Characteristics, and Quality
The PRISMA ow diagram displayed in Figure 1 details the selection strategy and resulting outcomes. Our electronic search of ve databases yielded 7,001 articles after duplicates were removed and three articles were further identi ed from the references of the retrieved articles. Of these, 21 met our eligibility criteria and 11 had su cient data that enabled their quantitative analysis. Characteristics of studies included in the systematic reviews are reported in Tables 2a and 2b. Thirteen studies examined the association between stress and children's cognitive outcomes and 13 investigated the in uence of anxiety. Sample sizes ranged from 29 to 5,768 in studies examining prenatal stress (mean = 632) and from 46 to 3,298 in those assessing prenatal anxiety (mean = 1,092). Of the eligible stress studies, the majority measured maternal prenatal perceived stress (n = 8). Of the eligible anxiety studies, n = 3 measured state anxiety, n = 5 measured trait anxiety, n = 4 measured trait and state anxiety combined, and n = 2 measured pregnancy related anxiety. Anxiety and stress were predominantly assessed during the third trimester only, while 4 studies measured maternal prenatal mental health across multiple trimesters. Children were aged 0 to 9 years old. In relation to child cognition, most studies assessed offspring cognition during infancy, i.e., 0-2 years old (75% and 58% of studies investigating stress and anxiety, respectively), with only nine studies examining child cognition past the age of 2 years. Five studies were from the USA, ve from the UK, three from Canada, two from Greece, and one from the Netherlands, Poland, China, Germany, Spain and New-Zealand.
In relation to study quality, 1 study was classi ed as strong, 7 as moderate and 3 as weak (Table 1). Few studies reported information on whether the outcome assessment was made blind to the exposure status (n = 5), which was needed to reach an overall strong quality rating. A notable risk of bias was also identi ed, related to inappropriate adjustment of missing data when attrition rate was high. As attrition was found to be associated with poorer mental health status, there is a risk of selective sampling bias that should be taken into account when interpreting the results 41 . Notably, all studies addressed a speci c research question, clearly de ned their outcome of interest, and used reliable and valid methods of measurement.

Meta-analytic Results
Maternal prenatal mixed anxiety and stress (Tables 3 and e1) Exposure to prenatal maternal stress and/or anxiety was signi cantly associated with lower language skills (r = -0.10 [95% CI, -0.19; -0.01], I 2 = 47.02%), albeit with a small effect size ( Figure 2). Nonsigni cant negative associations were also observed between prenatal exposure to maternal stress and/or anxiety and the general intellectual and attentional skills of the offspring. Similar ndings were found when these associations were examined during the infancy period only. Heterogeneity varied from moderate but non-signi cant (47.02%) to high and signi cant (82.27%) across studies. Results could not be strati ed by timing of exposure as most studies measured prenatal maternal stress/anxiety in the 3 rd trimester only.
Maternal prenatal anxiety (Tables 3 and e2) No associations were found between maternal prenatal anxiety and offspring's general intellectual functions, sustained attention, expressive language, receptive language and working memory. to three studies (n = 186 to 403) for the secondary. Heterogeneity ranged from low to high, with I 2 ranging from 0.00% to 89.01%.
Maternal prenatal stress (Tables 3 and e3) A weak adverse relationship was found between maternal prenatal stress and offspring's general intellectual skills (r = -0.26 [95% CI, -0.41; -0.10, I 2 = 87.43%; Figure 4). In contrast, no evidence for an association was found between prenatal exposure to stress and children's attention, learning and memory abilities. For perceived stress, a weak negative association between maternal prenatal stress and children's general intellectual skills remained but was nonsigni cant (r=-.25 [95%CI -.50; .04], I 2 = 76.63%). Heterogeneity varied substantially across studies, ranging from 0.00% to 87.20%. Sample sizes were generally small, with two to three studies (n = 309 to 428) for the primary analyses and two studies (n = 205) for the secondary analysis.
Covariates and Timing (Tables 4a and 4b) Twenty studies performed regression analyses, adjusting for a range of covariates including gestational age, With regard to timing, most studies measured maternal prenatal stress and/or anxiety during the third trimester and only one study assessed these constructs in the rst trimester. Only one study examined stress and/or anxiety at different timepoints during pregnancy and reported usable data 30 . As such, associations between prenatal maternal stress and/or anxiety and children's cognitive functions for different developmental periods could not be examined.

Discussion
Two key ndings emerged from this meta-analysis examining the association between maternal prenatal stress and anxiety with different aspects of offspring's cognitive development from infancy through adolescence. Firstly, there was a weak negative association between the exposure to stress and anxiety (combined) during pregnancy and the offspring's language development. Meaningful relationships did not extend to other cognitive skills, such as general intellectual abilities and attention. Previous systematic reviews and meta-analyses that focused only on developmental outcomes in infants reported adverse negative associations between prenatal stress and/or anxiety exposure and offspring's cognitive functions, but the strength of the associations reported has varied across reviews 11,13,56 . Different factors may contribute to these disparities across reviews, including distinct selection criteria around the type of prenatal exposure and outcome measures.
The second key nding of this meta-analysis was some evidence for differential associations between the different types of exposure and children's cognitive function. Exposure to stress but not anxiety was found to weakly correlate with children's general intellectual skills. Furthermore, when only including studies which looked at trait anxiety, a signi cant negative correlation was found with attention skills in the offspring.
However, when including studies that looked at both trait and state anxiety, this association disappeared. The present ndings provide some support for perspectives suggesting that state anxiety, trait anxiety, stress response and perceived stress constitute distinct constructs that might be differentially associated with children's cognitive outcomes 14,18,22 . Based on this preliminary evidence, future research should attempt to examine these prenatal types of exposure independently.
Although the original intent was to assess associations between prenatal maternal stress and/or anxiety at different developmental periods and children's cognitive functions, this was not possible due to the small number of eligible studies which analysed these psychological constructs at different timepoints. Pregnancy involves key processes that arise in a timely manner 57,58,59,60,61 . For instance, trimester 1 involves the formation of the neural tube and early neurogenesis, while trimester 2 includes neural migration and some axonal/dendritic development and trimester 3 largely consists of synaptogenesis and early myelination 61 . Also speci c brain structures have unique developmental trajectories and peak period of rapid growth and differentiation 61,62 . For example, the cerebellum's peak period of development starts around week 4 of gestation and proceeds throughout childhood, the thalamus develops from week 5 to week 10 of pregnancy, and the striatum from week 7 to week 19 of pregnancy 63 . Consequently, exposure to stress and/or anxiety at different timepoints during pregnancy is likely to differentially affect different neural networks which, in turns, may be associated with different cognitive outcomes in children. The literature on different teratogens, such as depression, nicotine, marijuana and cocaine, supports this assumption, revealing distinct cognitive outcomes in rodents and children exposed to teratogens at different timepoints during pregnancy 7,64 . Similar ndings might also arise following prenatal exposure to stress and/or anxiety. For instance, Polanska and colleagues 54 revealed that exposure to perceived stress during the second semester was associated with lower general intellectual skills but not language in children, and Koutra and colleagues 46 reported that exposure to trait anxiety during the third trimester was associated with attention di culties but not with other cognitive domains in the offspring. Accordingly, future research should aim to evaluate the in uence of timing in the associations between maternal prenatal stress and/or anxiety and children's later cognitive outcomes by examining these psychological constructs at different timepoints during pregnancy.
Numerous confounders should be considered when evaluating the association between prenatal stress/anxiety and later cognitive outcome, however this was implausible in our analyses due to the extensive variability in covariates across studies. Some of the above associations between stress and/or anxiety and developmental outcomes in childhood weaken when covariates were included in the analyses. For instance, Koutra and colleagues 46 reported that the association between prenatal exposure to trait anxiety and children's lower attention skills weaken once covariates, such as maternal age at delivery, maternal education, child's sex and prematurity, were controlled for. In another study by Polanska and colleagues 54 , a negative association between perceived stress and infants' general intellectual skills persisted after adjusting for several covariates, including active/passive smoking and alcohol consumption during pregnancy, maternal education, child sex, prematurity and breastfeeding duration. Agreeing on a consistent set of confounders to include in future studies by an expert panel would help to overcome the issue.
The present meta-analysis has several limitations that need to be taken into account when interpreting the results. One important consideration is that our analyses were correlational. As such, causality between maternal prenatal stress and/or anxiety and adverse children's cognitive outcomes cannot be established, and the role of covariates in this relationship remains unknown. It is likely that factors that were not taken into account in the meta-analysis (such as children's postnatal stress) played an important role in the relationship between prenatal maternal stress and/or anxiety and children's later cognitive functions. Another limitation of this study was the collapsing of participants across all age groups in the main analyses. We had hoped to assess the association across different developmental stages but this was not possible given the small number of eligible studies and large number of outcomes of interest. A third limitation of this body of work is that many of the studies pooled had limited samples and the pooled sample sizes were small. Accordingly, some of our analyses have lacked power, thereby potentially overestimating some of our effect sizes 65 .
Fourth, low to high heterogeneity was found across studies, potentially limiting our ability to detect meaningful relationships 66 . For example, nonsigni cant associations of small to medium effect sizes were sometimes found in analyses with high heterogeneity, such as between prenatal exposure to anxiety and children's working memory (r=-.24). Thus, more studies are required to increase con dence in the ndings that emerged from this systematic review and meta-analysis. It would also be interesting to understand why such a high heterogeneity was observed and address contributing factors in future research. One reason might be the wide range of measures used to assess stress response. Stress response was measured using blood, saliva and hair cortisol samples collected both in the morning and evening. The reliability as well as maternal and children's outcomes associated with these distinct measurement modes and times, however, were previously found to vary 67,68,69 . Therefore, an expert panel should come out with recommendations around which measurement method best assesses stress response to increase consistency in future research and provide more comparable results. Fifth, given the small number of studies found, we could not examine regional differences, even though studies were pulled from different regions of the world. Due to the small number of studies identi ed, we could not examine the effect of timing of prenatal stress and/or anxiety exposure either. Lastly, the current study focused speci cally on the cognitive outcomes of offspring. It is possible that the effects on offspring may be greater for other outcome domains such as emotional, behavioural and social outcomes. Therefore, future meta-analyses should also aim to separately examine the in uence of stress and anxiety on these domains.
In summary, our ndings support the current consensus that prenatal maternal stress and/or anxiety is negatively, albeit weakly, associated with offspring cognitive function. These ndings are of public health signi cance and support the need for screening and interventions to prevent or minimise mental health problems in pregnant women and optimise child development. Our ndings also have implications for conceptualising prenatal stress and anxiety in future research, as well as investigating the timing and cumulative effects of prenatal stress and/anxiety exposure. Finally, our ndings support the need for future studies to examine the relationship between maternal prenatal stress and/or anxiety and children's domains of cognitive function, in particular in children older than 2 years old. Tables  Note. GIF = general intellectual skills; RL=Receptive language; EL=Expressive language; T1, T2, T3=Trimesters 1, 2, 3   Association between Trait anxiety and sustained attention Association between Stress and General intellectual functions

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Supplementtables.pdf