Early Life Stress Effects on Children’s Biology, Behavior, and Health: Evidence, Mediators, Moderators, and Solutions

Abstract

This review synthesizes and critiques research on early life adversity and stress effects on multidomain health outcomes in child samples to fill a gap in the literature that has largely focused on adults. Prioritizing evidence from meta-analytic and systematic reviews as well as findings from (quasi-)experimental or large prospective longitudinal studies, we integrate interdisciplinary findings to characterize patterns of evidence for stress associations with child outcomes, including mental, physical, and positive health; academic, social, and justice system–related domains; and intermediary phenotypes that may predict disease, including biomarkers. We note cohesive evidence for sensitive periods of susceptibility to stress exposure and describe key mediators and moderators of stress effects, especially family-level factors. Then we highlight interventions targeting malleable factors that hold promise for ameliorating the effects of stress on children. Leveraging a developmental lens, we conclude with field-wide limitations and propose future directions for stress and health research that centers child development.

INTRODUCTION

While it was once difficult to convince lay audiences and healthcare providers alike that stress affected mental and physical health, it is now commonly understood. Today, numerous thoughtful, comprehensive reviews outline the well-documented links between stress and wide-ranging health outcomes, particularly psychological ones. This science has made its way into popular culture and spawned an extensive industry promoting stress-reduction products, techniques, and classes to promote wellness. Notably, the US Surgeon General issued a call to action in 2024 for a focus on parenting and research in underserved populations, citing drastic increases in parenting stress and caregiver and child mental health problems (see https://www.hhs.gov/surgeongeneral/priorities/parents/index.html). The impressive tide shift in the understanding of the stress–health connection presents opportunities to harness change in our social environment to improve health. However, as is often the case in the health sciences, the bulk of the evidence and accompanying solution-oriented insights are derived from adult samples (e.g., Bourassa & Sbarra 2024) and rarely considers adults’ roles within families (e.g., parenting, caregiving stress).

Simultaneously, at a much smaller scale, scientists, clinicians, and policymakers have extended this examination and solution generation to earlier stages of development. Although general concepts of stress activation can readily be overlaid across the life course, there are distinct differences across developmental periods. Overgeneralization has prevented clarity regarding when stress effects can be detected and whether young developing brains and bodies are more sensitive and/or resilient to such exposures—in which manners, for how long, and to what end. Although many now embrace the concept of early life stress affecting child and family outcomes, rigorous evidence is lacking for many pieces of those myriad developmental puzzles. In this review, we synthesize and critique the evidentiary basis for stress effects on multidomain child health outcomes, drawing from empirical evidence observed in child samples to fill this critical gap.

Background: Adversity and Stress

Stress is a universal human experience. Across history, humankind has contended with macro-level stress from wars, famines, genocides, pandemics, and climate events as well as individual-level interpersonal and intrafamilial stressors, including domestic violence, financial instability, and divorce. Evolution equipped us with a remarkable set of perceptive, cognitive, physiological, and behavioral responses, which act together to optimize adaptation to perceived threats, aid coping and survival, and facilitate the passing on of genes (for reviews, see Doom & Gunnar 2013, McEwen 2007, O’Connor et al. 2021).

Western society–based understanding that stress affects health has roots in the nineteenth and twentieth centuries (for review, see McEwen 2007). Later, concepts such as the flight-or-fight response (to acute stress) and allostatic load (in response to chronic stress) developed for articulating how stressors can influence patterns of biologically based stress responsiveness, with the potential to ultimately affect the structure, function, and rate of aging and disease across the life course (McEwen 2007). Truly developmental theoretical models of stress physiology, articulating how systems interact with and respond to the environment to generate typical and atypical development across the lifespan, have more recently become prominent (Doom & Gunnar 2013). For example, the concept of the biological embedding of early life adversity was introduced to explain how population-level well-being, learning, behavior, and physical health outcomes fall along a socioeconomic gradient (see Hertzman 2012 for review). Integrated with psychological theories and evidence, this framework expanded to include how a broad range of early social environments and experiences “get under the skin” to affect health, with childhood effects progressing to broader manifestations across health outcomes throughout the life course (see Boyce et al. 2021). These theoretical framings with catchy names, and the readily translatable concept of toxic stress (Shonkoff et al. 2012), have served to educate the public on the importance of prevention to reduce the risk of problems. However, they are often understood or used in a manner that fails to acknowledge the vast variability in adversity’s effects on child health, potentially leading to problematically labeling children as doomed and promoting fear and stigma.

Life course developmental theories expand upon those prevailing models of dysregulation framing to emphasize how early stress can initiate the development of adaptive strategies with evolutionary value (e.g., early puberty), as they promote survival and reproduction under adverse conditions, despite their potential costliness to the organism. The adaptive calibration model (see Ellis & Del Giudice 2019 for review and details), the biological sensitivity to context theory (see Boyce & Ellis 2005 for review and details), and the biological embedding framework help to explain how stress response systems adapt, in positive and negative ways, to environmental conditions over time, beginning early in life, with major variability across persons in susceptibility to environmental influences. Similarly, the developmental origins of health and disease (DOHaD) framework posits that experiences during key developmental periods, particularly the prenatal period, can have long-lasting effects on health and well-being, even into adulthood (Barker 2007), through, for example, the intergenerational transmission of stress. Intergenerational stress theories have evolved substantially and include considerations of micro- and macro-level social adversities (for review, see Bush 2024). Although most empirical evidence for stress exposure effects comes from the psychological literature, it is important to acknowledge and include in this review the newer bodies of literature from a range of fields addressing the social determinants (or social drivers) of health (Adler et al. 1994), which operate in part through psychological stress pathways (see the sidebar titled Social Determinants of Health Versus Early Life Adversity Frameworks).

SIDEBAR 1: SOCIAL DETERMINANTS OF HEALTH VERSUS EARLY LIFE ADVERSITY FRAMEWORKS.

Capturing the health effects of early life stress requires multidisciplinary, intersectional frameworks. Largely distinct from psychology frameworks, SDoH research—central to epidemiology, public health, and medicine fields—typically focuses on factors like socioeconomic indicators, food security, healthcare access, and neighborhood quality influence health. Psychology frameworks examining early life stress and adverse events, however, are more commonly focused on capturing acute event types (e.g., maltreatment, death of a loved one, divorce), interpersonal dynamics (parental psychopathology, negative parenting), or individual perceptions (e.g., severity of distress after natural disaster, experiences of discrimination) (for review, see Epel et al. 2018). Given its epidemiological and public health origins, SDoH typically reflect indicators accessible through brief surveys, medical records, or census/public data set linkages, whereas stress research often requires more nuanced inquiry into stress, biopsychosocial mechanisms, and coping. Proposed mechanisms of SDoH effects extend beyond stress-related biosocial factors and include nutrition, educational and health services, environmental capacity to support physical fitness, and opportunities for upward mobility. Fields increasingly interact to conduct multidisciplinary science, however, and most SDoH literature emphasizes activation of stress pathways in response to social drivers such as racism and neighborhood disadvantage (Priest et al. 2024, Shonkoff et al. 2021), while psychological research increasingly includes macro-level social determinants.

A key evolution in stress–health frameworks has been the acknowledgment that the severity and frequency of stress exposures have not been equally distributed across or within populations, with factors of power, privilege, and oppression manipulating their distribution. In particular, colonialism and chattel slavery reaped centuries of unimaginable trauma and enduring inequities (Nagata et al. 2024) upon much of the global population. While the legal forms of these traumas mostly abated in the nineteenth century, their effects reverberate among the nations and cultural groups affected. Research centering communities exposed collectively to historical trauma has shown intergenerational effects on physical and mental health among descendants of people who experienced the Holocaust, Japanese-American internment, settler colonization, and chattel slavery (for reviews, see Bush 2024, Nagata et al. 2024). Through historical trauma and current day policies, certain communities are persistently more exposed to stress. For example, in the United States, neighborhoods with higher concentrations of formerly enslaved or colonized groups (e.g., African Americans, Native Americans) have experienced decades of divestment because of redlining and myriad discriminatory policies—thus, their residents are exposed to more environmental (pollution, noise), health (less access to healthcare and nutritious foods), and community (crime, substandard housing) stressors as well as reduced opportunities for growth/wellness (lack of high-quality schools or green spaces) (Acevedo-Garcia et al. 2020). These policies affect child and family health through ecological systems that include the workplace, school, healthcare system, and beyond (Bronfenbrenner 1994) and require scientific efforts and solution generation informed by those with these intersectional lived experiences (Geronimus et al. 2023).

Stress Measurement Considerations

Measuring stress and its effects is complex. Epel and colleagues (2018) provide an analysis of the challenges and potential solutions, highlighting that stress is experienced on multiple levels (including social, psychological, and physiological), with few agreed-upon measures. They argue that a common language and more precise multilevel models are required for scientific advances. In child and family research, for example, given low base rates for most individual adverse event types (especially at younger ages), most studies lump exposures together, not discerning among acute, event-based, daily, and chronic stressors and/or the developmental periods or timescales during which those events occurred. Some of the tremendous variability found in how individuals respond, psychologically and physiologically, to such grossly assessed stress exposures is likely due to differences in qualities and severity of adversity (e.g., single incident of mild abuse versus chronic severe abuse, which are often both coded equivalently as maltreatment exposure)—see the sidebar titled Stress Measurement Complexities for more details. Findings from the evidence reviewed below should be considered in light of these complexities.

SIDEBAR 2: STRESS MEASUREMENT COMPLEXITIES.

Stress is measured with highly variable methods, complicating comparisons of findings across studies. Given that many stressors are theorized (or documented) to operate through similar mechanisms to affect health, stress exposure research often aggregates experiences across domains of adversity. The prevailing unitary risk lumping approach includes cumulative risk scores (originating in child psychology; Sroufe & Rutter 1984) and Adverse Childhood Experience (ACE) scores counts (originating later in adult medicine; Felitti et al. 1998). Evidence suggests that the accumulation of various risk factors, regardless of their specific type, increases the likelihood of negative outcomes. Cumulative risk has been found to robustly predict dozens of developmental and disease outcomes, although other researchers find value in distilling experiences into core dimensions that share common features and associations with outcomes, such as threat and deprivation, social versus nonsocial stressors, or predictable versus unpredictable risk. Across approaches, researchers also sometimes (particularly in samples with lower base rates of stress) create dichotomized indicators to test threshold models (where risk increases substantially after a certain number of risk factors is reached, such as 4+ ACEs cutoffs) versus linear models (where risk for problems increases in a relatively steady fashion as risk factors accumulate).

Some of the strongest approaches to demonstrating stress exposure effects come from experimental and quasi-experimental study designs that allow for causal inference, which is rightfully rare in stress research, given ethical concerns. Research leveraging natural experiments to capture extreme adversity exposures has been particularly informative. For example, the Bucharest Early Intervention Project (BEIP) compares outcomes for children reared in Romanian orphanages characterized by extreme to either remain in institutionalized care or receive high-quality foster care with intensive training and support, including financial, for caregivers (see Wade et al. 2022 for a review). This enables researchers to assess differences in child outcomes based on exposure to neglect and adversity and timing of environmental remediation. Others have leveraged randomized controlled trials (RCTs) of foster care placements, with and without intensive caregiver training and support or relative to community samples, to test how variations in caregiving quality affect a range of family and child outcomes (for review, see Kirby et al. 2025).

The Current Review

In this review, we synthesize and critique the evidentiary basis for stress effects on multidomain child health outcomes, drawing from empirical evidence observed in child samples to fill a critical gap. Aligned with the broad stress frameworks leveraged across meta-analyses, systematic reviews, and large prospective longitudinal initiatives we prioritized for review, our conceptual definition of stress exposure encapsulates traumatic events and poverty-related stress, as it is difficult to disentangle frequently co-occurring exposures.

EFFECTS OF EARLY LIFE STRESS ON MULTIDOMAIN CHILD HEALTH AND WELL-BEING OUTCOMES

Mental Health Outcomes

Early life stress is robustly, consistently, and enduringly associated with mental health problems across the lifespan (Juwariah et al. 2022, Kim & Royle 2025, Peverill et al. 2021) at great cost to society via healthcare costs to treat conditions, disruptions to quality of life and work, and other factors (Bellis et al. 2019). Much of the research linking adversity exposure to mental health problems leverages retrospective study designs, wherein adult participants recall childhood adversity exposures (see, e.g., Kim & Royle 2025 for an umbrella review of meta-analyses inclusive of adults). In recent years, through well-characterized small samples and collective national efforts, research collecting repeated measures of stress exposure and tracking multi-level outcomes across development has generated evidence that early life stress exposure relates to higher levels of mental health problems during childhood [see, for example, the Avon Longitudinal Study of Parents and Children (ALSPAC) (Major-Smith et al. 2022) and the Environmental influences on Child Health Outcomes (ECHO) program (Bush et al. 2020)]. Systemic and meta-analytic review evidence from international and population-based studies demonstrates that adversity exposure (characterized by traumatic event exposure and/or individual- and neighborhood-level poverty) is positively associated with diverse forms of youth psychopathology (Juwariah et al. 2022, Peverill et al. 2021, Visser et al. 2021). Empirical data from large cohort studies suggest that exposure to individual stressors (e.g., from a Welsh cohort of 191,035 children; Lowthian et al. 2021) and cumulative adversity exposure (e.g., from the US-based National Survey of Children’s Health, with a cohort of 98,732 children; Walker et al. 2022) are associated with greater risk of having psychiatric diagnoses. Some evidence suggests effects may be larger for child psychopathology associations with traumatic events compared to associations with low socioeconomic status (SES) (Gur et al. 2019).

Findings from randomized trials, such as the BEIP, show causal relations between early adversity exposure and higher general psychopathology (i.e., sometimes measured with a single dimension construct called the p-factor, which represents the overlap or shared variation between various mental health problems) during childhood and, more markedly, adolescence (Wade et al. 2022). Further evidence supporting the causal influence of stress exposure on child mental health can be found in mechanism-designed RCTs that target stress exposure and stress-mitigating strategies. For example, interventions that reduced maltreatment exposure improved child mental health (Baldwin et al. 2023). Consistent findings spanning these multiple methodologies increase confidence in the causal relationship between stress exposure and childhood problems.

Robust evidence from large empirical studies also demonstrates that positive associations between stress exposures and mental health problems persist at syndrome-specific levels across childhood, including internalizing problems such as anxiety and depression (Gur et al. 2019, Lowthian et al. 2021, Walker et al. 2022) and externalizing problems such as attention-deficit/hyperactivity disorder (ADHD) and antisocial behavior (Björkenstam et al. 2018, Lowthian et al. 2021). Meta-analytic evidence suggests the effects of poverty-linked exposures may be larger for externalizing compared to internalizing disorders (Peverill et al. 2021).

Early life adversity is also associated with high-risk behaviors in children, including nonsuicidal self-injury, suicidality, and substance use (Juwariah et al. 2022). Findings from an epidemiological assessment of the behavior of US adolescents (N = 9,937) indicate that early life adversity (particularly threat-based) is also associated with multiple sexual risk indicators (e.g., younger sexual initiation; Thomas et al. 2023).

Developmental Delay and Disorder

Early life stress exposure also relates to higher levels of neurodevelopmental disorders, such as developmental delay (DD) (Lowthian et al. 2021, Oh et al. 2018), intellectual disability (Leonard et al. 2022), and autism spectrum disorder (Hartley et al. 2024). Importantly, empirical evidence from numerous well-controlled studies indicates that autism has biological and genetic origins and is not caused by poor parenting practices or inadequate nurturing (Lord et al. 2022).

Cross-national evidence suggests socioeconomic factors—such as material deprivation and lower levels of cognitive stimulation—are clear risk factors for developmental delay (Salhi et al. 2021). This may be due in part to the associations between stressors (e.g., individual- and community-level poverty or experiences of discrimination) and risk for preterm birth and small-for-gestational-age at birth, which are strong predictors of developmental challenges (Lord et al. 2022). Children with DD and related conditions are at particularly elevated risk for future trauma exposure (Jones et al. 2012), which may lead to continued transactional processes of risk and outcome escalating over time.

Child Neuropsychological Health (Executive Functions)

Stress exposure impairs executive functioning (EF), a term encompassing a set of cognitive processes that involve regulation of goal-directed behavior, including processes like inhibitory control, working memory, and cognitive flexibility (Miyake et al. 2000) that are conceptualized to lie upstream of transdiagnostic mental health problems (Zelazo 2020). Meta-analytic evidence identifies a consistent small negative effect of stressors on childhood cognition and EF, with deprivation (Hedges’ g = −0.43) having slightly stronger associations relative to threat (Hedges’ g = −0.27; Johnson et al. 2021). More nuanced examinations suggest that considering exposure and outcome domain is important. For example, SES appears to be negatively associated with task-assessed cognitive functioning (e.g., EF and episodic memory) (Gur et al. 2019), whereas associations identified between traumatic event exposure and neurocognition within a highly stressed, low-income sample are less clear (Cohodes et al. 2020).

General Child Functioning

Exposure to early life stress is also related to impairment across general aspects of child daily functioning, including social, academic, and juvenile justice–related domains.

Social functioning.

Two recent systematic reviews summarized early life stress associations with social functioning. Wang et al. (2024) examined 92 studies on early life stress and peer relationships and found mixed associations, depending on the relationship aspects examined. While early life stress was generally negatively associated with relationship quality and social status within a peer group, early life stress was not related to peer support, and associations with susceptibility to peer influence were mixed (though few studies have examined this). Merrin et al. (2024) examined 51 studies and found a consistent pattern of positive associations between early life stress and bullying perpetration and victimization. Across reviews, whereas cumulative early life stress scores tended to be consistently associated with social functioning, only some individual types of early adversity were predictive, suggesting that the operationalization of stress exposure impacts findings.

While much of the research we reviewed considers stress exposure to be early life trauma or poverty, a growing area of research investigates peer-based stressors, such as bullying and relational victimization (e.g., gossip and exclusion). Peer-based stress is associated with differences in brain development, which in turn may mediate the development of psychopathology (Quinlan et al. 2020). Such stressors hold particular potential for harm during middle childhood and into adolescence, during which the importance of peer groups increases and youths are undergoing rapid biological and behavioral skills changes (Dahl et al. 2018).

Academic functioning.

Early life stress is also associated with children’s academic achievement. In a meta-analysis of 20 studies, stress exposure was positively associated with poor academic achievement, grade repetition, and special education support (Qu et al. 2024). Associations between cumulative early life stress exposure and poor academic achievement followed a dose–response pattern, and subtypes of early life stress were independently related to poor academic achievement. Further, a recent population-based study found that, in addition to predicting differences in IQ and academic achievement, early life stress predicts larger discrepancies between the two (Schuurmans et al. 2022), revealing its likely role in preventing children from reaching their school and career potential.

Justice system involvement.

Early life stress is also related to juvenile justice system involvement. Indeed, childhood maltreatment is one of the most frequently studied risk factors for youth violence and offending (Yohros 2023), and a systematic review of 10 studies found consistent positive associations between early adversity exposure and justice system contact (Graf et al. 2021). In addition, two systematic reviews and meta-analyses confirmed the association between cumulative early adversity and juvenile recidivism (Astridge et al. 2023, Yohros 2023).

Positive Health Outcomes

Although the evidence base is small, emerging emphasis on assessing positive health (see Bodryzlova & Moullec 2023) is worth highlighting. For example, in national samples, exposure to fewer adverse family experiences was related to better flourishing (mental well-being) (N = 51,156 children and adolescents; Ignacio et al. 2022).

Physical Health

In general, stress-related predictors of mental health also appear to impact physical health. Many infer that the consistent, strong evidence of effects of early life stress on adult health outcomes (e.g., Bellis et al. 2019) can be generalized to children. However, most diseases studied do not typically manifest at clinical levels during childhood, particularly early in development. Systematic review evidence indicates that adversity is associated with pediatric asthma, infection, somatic complaints, and sleep disruption (Oh et al. 2018). Novel evidence demonstrating that moving from high-poverty to low-poverty neighborhoods related to improved pediatric asthma symptoms (Pollack et al. 2023) shows that adversity effects also occur at the level of the neighborhood. A meta-analysis of research on access to neighborhood opportunities parallels these findings, demonstrating that children living in neighborhoods with limited access to education, healthcare, and other resources are at higher risk of pediatric mortality and all-cause emergency department use (Tyris et al. 2025). Meta-analytic evidence suggests stress specific to experiencing racism relates to higher levels of physical health problems during childhood, including obesity, asthma, and somatic symptoms (Priest et al. 2024). Finally, institutionalized care is strongly negatively associated with physical growth, with associations falling on a dose–response continuum, such that longer duration under institutionalization translates to worse outcomes (van IJzendoorn et al. 2020).

Intermediary Pediatric Phenotypes of Stress Exposure: Biomarkers

A recent review in this journal (O’Connor et al. 2021) highlights the psychobiological processes operating between stress and health outcomes, integrating pediatric and adult literatures. Frequent and/or enduring stress system activation, especially during developmental sensitive periods (see below), such as childhood, can lead to lasting physiological changes that allocate resources to support short-term adaptation but may increase the risk of long-term health consequences. Scientists posit that these biological changes may be mechanisms accounting for how stress results in immediate and long-term disease (Hertzman 2012, McEwen 2007) [see the relevant National Institute of Mental Health mechanistic framework called Research Domain Criteria (RDoC); Insel et al. 2010], resulting in the prioritization of research to discover and refine markers that identify people, including children, at high risk of health problems.

Hypothalamic-pituitary-adrenal axis and cortisol.

For decades, scientists have striven to parse the complex associations between stress exposure and hypothalamic-pituitary-adrenal (HPA) axis activity via the stress-responsive hormone cortisol. In child populations, cortisol concentrations are used as one indicator of a cascade of HPA axis–mediated hormonal processes in response to a stressor (Doom & Gunnar 2013). Cortisol concentrations assayed from saliva, blood, and urine are prevalent in child research and can indicate consistency with or deviations from the normal fluctuation observed in cortisol across the day due to its diurnal rhythm of output. Experts argue that, whereas acute stressors can increase cortisol output, chronic adversity can result in hypocortisolism, or lower overall cortisol output and flattened diurnal rhythms, which may increase risk for health problems, like externalizing problems (e.g., Chong et al. 2022, Koss & Gunnar 2018) Findings from a meta-analysis of 129 studies across specimen types confirm these assertions, documenting that childhood adversity exposure associates with higher afternoon cortisol levels, flatter diurnal slope, more blunted reactivity to acute stress, and higher hair cortisol concentrations (HCC) in children aged 0–17 (Niu et al. 2025). Notably, effect sizes for stress and HCC were approximately twice as large (~0.10) as stress associations with other specimen-type measures (~0.04–0.05), in line with scientific thinking that HCC may be a better marker of chronic exposure to stress across a sustained period (i.e., 3 months) relative to more dynamic measures of cortisol (e.g., salivary) (Koss & Gunnar 2018). Further, when subtypes of exposure were examined in a meta-analysis (Niu et al. 2025), threat-based stress related to heightened reactivity before middle childhood but blunted reactivity in older adolescents, whereas deprivation was associated with blunted reactivity across development. Threat subtype and child developmental stage moderation of associations with hair were also found but appeared more nuanced. Specifically, the estimate shifted from negative (where greater levels of threat related to lower HCC) to positive (where greater levels of threat related to higher HCC) at around age 4. Statistically significant positive effects were evident by approximately age 10. In a meta-analysis of a smaller, distinct group of studies examining racism-related stress and pediatric health, exposure to racism was associated with elevated salivary cortisol in six studies (Priest et al. 2024).

The autonomic nervous system.

The autonomic nervous system (ANS) innervates most organs and coordinates bodily responses (e.g., fight, flight, freeze) to stressors via sympathetic nervous system (SNS) activation and parasympathetic nervous system (PNS) withdrawal. SNS activity at rest, in reaction to a stressor, or in recovery (or lack thereof) after stressors have abated is typically assessed via indicators such as pre-ejection period (PEP) and skin conductance. PNS activity across contexts is typically assessed via respiratory sinus arrhythmia (RSA) and heart rate variability (HRV). Systematic reviews indicate that adversity exposure relates to blunted SNS and PNS responsivity to stressors and lower levels of resting PNS activity (e.g., lower RSA) (Holochwost et al. 2021, Young-Southward et al. 2020). However, a meta-analytic investigation of PNS in a recent study of over 55 mostly pediatric samples reported no overall association between childhood adversity exposure and an index of HRV at rest or in reaction to stress. Rather, sample characteristics (e.g., clinical diagnoses, severity of stress exposure) explained significant heterogeneity in associations. Groups characterized by higher degrees of risk (e.g., risk of psychopathology, adversity subtypes coded as more severe) exhibited lower resting HRV (Wesarg et al. 2022), whereas older samples and samples reporting more temporally distal child adversity exposures displayed less vagal withdrawal; both patterns of physiology suggest less optimal functioning and risk for later problems. Isolating a type of stress exposure, one meta-analysis found evidence that racism related to higher systolic and diastolic blood pressure (Priest et al. 2024). Finally, a meta-analysis of ANS activation during a standardized stressful caregiver–infant interaction task found that infants from families exposed to higher stress (e.g., of lower SES) did not exhibit RSA recovery post stressor, which can predict higher risk of disorder and disease. Infants at lower risk of stress exposure exhibited RSA recovery (Jones-Mason et al. 2018). We were unable to identify any meta-analyses on pediatric SNS indicators.

Inflammation.

Inflammatory processes, assessed in part with pro- and anti-inflammatory cytokines measurable in blood and saliva, underpin stress reactions and are closely linked with learning, memory, and myriad health-relevant processes (Quinones et al. 2020), eliciting great interest in their capacity to forecast transdiagnostic health outcomes (e.g., Danese & Lewis 2017). Neuroimmune models of psychopathology, particularly depression, underscore the causal role inflammation plays in increasing risk (Nusslock et al. 2024). Although relatively few pediatric studies link inflammation to future psychopathology, a meta-analysis of a subset of three studies in adolescence indicates that inflammation is positively associated with future depressive symptoms (Colasanto et al. 2020). A meta-analysis of samples ranging across the lifespan (Chiang et al. 2022) indicated that childhood stress related to higher concentrations of proteins indicative of elevated inflammation (e.g., IL-6). Notably, the strength of the positive associations found in childhood increased across development, with larger associations found with adulthood inflammation. Racism-related stress exposure is also associated with markers of elevated systemic inflammation (Priest et al. 2024).

Brain development, structure, and function.

Brain imaging technological advances have allowed insight into the association of early life stress exposure with children’s brain development. Findings drawn from studies using structural magnetic resonance imaging (sMRI) methods that measure the size, shape, and volume of brain structures show that early life adversity relates to smaller gray matter volumes, particularly in the amygdala, medial prefrontal cortex, and hippocampus as well as lower cortical surface areas (McLaughlin et al. 2019, Rakesh & Whittle 2021). Relative smaller gray matter volumes may mediate stress associations with prospective health problems (Kim et al. 2022). Adversity exposure is also positively associated with higher levels of amygdala activity [via functional magnetic resonance imaging (fMRI)] in response to task-evoked threat (McLaughlin et al. 2019). However, inferences drawn from longitudinal fMRI research should be considered with particular caution given low consistency within individual-level fMRI assessments repeated over short periods of time (see Flournoy et al. 2024).

Epigenetic processes.

Stress exposure results in epigenetic changes, supporting bodily adaptations to contend with the environment and support resilience (Merrill et al. 2025). Epigenetic processes involve molecular changes [e.g., histone modifications and changes to DNA methylation (DNAm) levels] that can result in enduring modifications to genetic activity, and therefore cellular functions, without altering the underlying DNA sequences. Importantly, epigenetic changes are heritable and posited as a biological mechanism for the intergenerational transmission of stress effects (Monk et al. 2019). Further, prenatal stress exposure appears to influence newborn epigenetic markers (Kotsakis Ruehlmann et al. 2023).

Although this body of research is relatively nascent, epigenome-wide studies indicate that a range of stress exposure subtypes (low SES, family adversity, etc.) may differentially influence DNAm levels at unique sites along the epigenome (Bush et al. 2018). It remains unclear, however, what health outcomes these epigenetic changes engender. Candidate gene approaches offer insight into the directionality of adversity and DNAm associations for specific gene loci. Importantly, stress and candidate-gene DNAm associations rarely replicate. Research on NR3C1, a gene closely related to HPA axis functioning, provides an exception (Merrill et al. 2025), with most studies identifying positive associations between child maltreatment exposure and NR3c1 methylation (for review, see Parade et al. 2021), which may engender a phenotype of heightened stress reactivity (Merrill et al. 2025).

Biological aging (epigenetic clocks, telomere length, pubertal timing).

Accelerated biological aging is a concept used to describe phenotypes in which physiological marker patterning outpaces typically expected values for chronological age and can be a useful intermediary phenotype for disease risk. In child samples, stress exposure is associated with accelerated aging measured with the epigenetic clock (Musci et al. 2023) and telomere length (Coimbra et al. 2017, Colich et al. 2020) measures. Some evolutionarily framed science indicates that early pubertal timing may occur in the context of threat-centric stress exposure to enhance reproductive capacity (Colich et al. 2020, Gur et al. 2019).

Allostatic load.

Allostatic load is a theoretical framework capturing the cumulative physiological and psychological costs of individuals’ adaptations to chronic stress or repeated exposure to stressors, assessed with a composite measure of biomarkers (e.g., cortisol, blood pressure) (McEwen 2007). In the only study to our knowledge examining pediatric allostatic load, higher allostatic load was observed among children exposed to 1–3 adversity subtypes relative to those exposed to 0 (de la Rosa et al. 2023). Systematic review of 38 studies in children suggests indicators of allostatic load are positively associated with future physical and mental health problems (e.g., Lucente & Guidi 2023), suggesting the value of further studying allostatic load in stress research.

Other markers with less evidence.

The pediatric evidence base assessing biomarkers linking stress exposures to disease states is diffuse and heterogenous, characterized by rapid discovery and evolving technology. While some biomarker domains are still in their early days even in studies of adults, their normative or expected values across the population and clinical relevance during earlier stages of human development, such as childhood, are even more unclear, impeding translation to early life. For example, novel biomarkers from the adult literature are increasingly examined for linkages to adversity in pediatric samples (Shonkoff et al. 2022), including mitochondrial function (Zhao et al. 2023) and the microbiome (Querdasi et al. 2023). Although some have found compelling evidence for associations between child stress exposure and genetic expression-linked cellular activity in young adult samples (e.g., Cole et al. 2020), we could not find reviews or meta-analyses of these associations in pediatric samples. One related empirical study, however, finds evidence for prenatal maternal stress associations with fetal placental transcriptomics (Baker et al. 2024).

SENSITIVE PERIODS FOR STRESS EXPOSURE EFFECTS

Stress exposure during key developmental windows of rapid brain and body development is particularly harmful for child health and lifelong health trajectories (Knudsen 2004, Shonkoff et al. 2021). Conversely, these windows also hold greater potential for resilience and are poised to receive greater benefits from effective interventions and preventions (Masten et al. 2021).

Prenatal

The prenatal period is a sensitive period for stress exposure effects on child health (Bush 2024). Framed by prenatal programming and DOHaD theories (Barker 2007), maternal stress exposure that results in elevated stress system responses warns the fetus to anticipate stress in the postnatal environment and to adapt its development to optimize later survival. Numerous mechanisms appear to account for this intergenerational transmission, including upregulated maternal HPA axis activity resulting in downstream influences on epigenetic regulation of glucocorticoid-relevant genes (Monk et al. 2019) and placental transcriptomics (Baker et al. 2024). This process is thought to result in more stress-attuned infant phenotypes (Hartman & Belsky 2018); however, mismatch with the actual environment, enduring costly activation of physiology, or compensatory behaviors and biological processes in other domains can result in disease states. Indeed, stress exposure during pregnancy is positively associated with immediate (e.g., early birth timing, lower birthweight; Ding et al. 2021) and later child mental and physical health problems (for review, see Bush 2024).

Importantly, in addition to influencing fetal development and child health trajectories, maternal brains undergo remarkable changes during pregnancy, with this plasticity potentially supporting factors beyond the physical aspects of gestation, such as the cognitive, emotional, and behavioral transition to parenthood (Thomason & Hendrix 2024). Collectively, the stress calibration theories noted above suggest that excessive stress exposure during the prenatal period may durably influence both fetal (eventually, child) and maternal stress system responsiveness (Gunnar & Howland 2022).

Infancy and Early Childhood

Rapid synapse proliferation occurs throughout infancy and into toddlerhood (Blakemore & Choudhury 2006, Kolb et al. 2012), eventually transitioning into neural development characterized by myelination and synapse pruning. This rapid development results in highly plastic brains sensitive to environmental inputs, including stressors. Public health and early life resource programs developed to improve outcomes for young children and their families, such as Zero to Three and The First 1000 Days, are built upon the foundation that early childhood comprises a key time for child development, both for increasing inputs of beneficial influences and for minimizing adversity exposure (van IJzendoorn et al. 2020). Evidence drawn from large cohorts with repeated measurements of adversity exposure and child outcomes provides more precision, suggesting that adversity exposure before age 3 (relative to later childhood exposure) is related to more differences in DNAm at stress-sensitive CpG sites and worse externalizing problems (Dunn et al. 2019, 2020). However, review evidence suggests nuance in how dosage and timing associate with health outcomes. For example, relative to children exposed to institutional environments only up to 6 months of age, children exposed to institutional environments from 6 to 24 months of age were more vulnerable to the consequences of deprivation- and institutionalization-linked stress exposure (van IJzendoorn et al. 2020).

Middle Childhood

As children age into middle childhood, it is posited that their sensitivity to stress is somewhat attenuated relative to earlier periods, as the pace of brain development slows. Although exposure to stress during this period likely continues to have a notable impact on brain development (Knudsen 2004), empirical evidence regarding the influence of stress exposure during middle childhood is relatively scarce. Given increasing interactions with environments external to the home (e.g., school) and rapidly developing meaningful relationships with peers and nonfamilial adults, social stress (e.g., related to bullying and relational victimization) appears to be detrimental during this time, particularly as children approach puberty (Dahl et al. 2018).

Adolescence

Pubertal processes that typically occur across preadolescence and adolescence may activate another particularly sensitive period to stress exposure (Dahl et al. 2018). Pubertal hormones play an active role in brain organization (Gunnar & Howland 2022), and the brain is still undergoing rapid development (e.g., myelination and pruning) relative to adulthood. As such, exposure to stress may durably influence adolescent physiology, brain function, and endocrine and immune functioning, resulting in increased risk of psychopathology (Laricchiuta et al. 2023). Notably, stress systems that have previously slowed in their developmental responsiveness to inputs may recalibrate during puberty (Gunnar & Howland 2022). Evidence suggests that adolescents previously exposed to severe early life deprivation and then transitioned to supportive caregiving contexts retune their HPA axis functioning to better align with healthy functioning (e.g., from flattened or hypo-responsiveness to more typical curvilinear reactivity to stress; Gunnar et al. 2019).

FAMILY HEALTH FACTORS MEDIATE STRESS EXPOSURE EFFECTS ON CHILD OUTCOMES

Long-standing theoretical models across disciplines, such as ecological systems theory (Bronfenbrenner 1994), the family stress model (Masarik & Conger 2017), and biopsychosocial models of pediatric disease states (e.g., chronic pain), emphasize the critical role caregivers play as intermediaries between stressful environments and child health. Aligned with these models, the American Academy of Pediatrics recently coined the term “early relational health” and disseminated guidelines aimed at promoting caregiver health, positive parenting behaviors, and healthy caregiver–child relationships in order to facilitate child transdiagnostic health (Garner & Yogman 2021), a concept illustrated in Figure 1. Below, we highlight evidence on caregiver- and family-level factors that, while often framed as risk factors for child health problems, are often outcomes of other adversity exposures operating as mediators, positioning them as intervention inlets ripe with two-generation potential.

Figure 1.

Graphical illustration of the multilevel, longitudinal process of early life adversity and stress effects on child mental and physical health and well-being. This graphic highlights only some of the key factors involved in these processes and is not meant to be exhaustive. Abbreviations: ANS, autonomic nervous system; HPA, hypothalamic-pituitary-adrenal.

Caregiver Mental Health

Caregiver mental health is foundational to healthy child development and can be deeply affected by caregiver stress exposure. Decades of extensive psychological research, synthesized in numerous meta-analyses, show that maternal depression, both pre- and postnatal, is a clear risk factor for negative developmental and mental health outcomes in children (Goodman et al. 2020, Monk et al. 2019, Racine et al. 2024). Other relatively underexplored caregiver mental health factors also associate with child outcomes, including schizophrenia and bipolar disorder (for a meta-analysis, see Rasic et al. 2014). While the preponderance of research examining parental psychopathology as a child adversity factor focuses on its relation to child socioemotional and mental health problems, meta-analytic findings indicate it is also associated with child physical health problems (e.g., asthma, obesity; Pierce et al. 2020). Further, although the majority of previous caregiver studies have focused on maternal mental health, paternal mental illness is also negatively associated with child socioemotional and mental health problems (for an umbrella review, see Scarlett et al. 2024). Accordingly, reflecting these associations and in line with decades of developmental psychopathology research on family risk factors and child health (see Sroufe & Rutter 1984), the originally popularized Adverse Childhood Experiences (ACE) screener (Felitti et al. 1998) included two caregiver factors—mental illness and substance abuse—as key contributors to child health outcomes, although both are likely influenced by broader context adversities (see Figure 1). Moreover, these associations may apply to professional caregiving contexts, with early life stress potentially influencing child health via its effects on early childhood educators (ECEs), teachers, and other home- and center-based childcare providers (for review, see Corr et al. 2014). For example, early evidence indicates that ECEs’ exposure to early adversity was negatively associated with quality of behaviorally coded teacher–child interactions (Rancher & Moreland 2023).

Whereas most evidence underscores a unidirectional, caregiver-driven path from parental health factors to child outcomes, as early as 50 years ago Sameroff (1975) argued for the importance of considering transactional, bidirectional associations between child behavior and parental mental health. Repeated-measure, prospective longitudinal research demonstrates that, while maternal histories of maltreatment can operate through maternal psychopathology occurring in pregnancy and postnatal periods to affect child developmental and mental health problems (for review, see Bush 2024), child health also may drive caregiver mental health (e.g., Roubinov et al. 2022a).

Caregiving Behaviors and Caregiver–Child Relationships

The child–caregiver attachment relationship is a foundational socioemotional context that influences multifinal child developmental outcomes via the shaping of caregiver and child thoughts, feelings, and behaviors. Parental sensitivity (the caregiver-driven process through which parents identify and respond to their child’s cues and needs) reciprocally transacts with attachment quality, resulting from attachment quality and enhancing it (Ainsworth 1989). As well articulated in the family stress model (Masarik & Conger 2017), stressful contexts hinder secure attachment, positive parenting, and, subsequently, child and caregiver health. Although research typically focuses on maternal–child dynamics, both maternal and paternal parental sensitivity influence attachment quality (Madigan et al. 2024), indicating that, in two-caregiver family structures, both influence child health.

Meta-analytic evidence indicates that caregiving behaviors relate to child mental health. For example, parental sensitivity was negatively associated with offspring’s internalizing and, more strongly so, externalizing problems (Cooke et al. 2022). Accordingly, numerous evidence-based interventions target parenting behaviors to promote child mental health, and parenting interventions are the first-line treatment for numerous early childhood psychiatric problems. Recent interdisciplinary research has expanded the potential scope of influence of positive parenting behaviors. RCTs provide causal evidence that enhancing caregiver–child early relational health promotes numerous positive developmental outcomes (e.g., receptive vocabulary, cognitive flexibility) in young children (Kirby et al. 2025). Building from cross-species evidence that effective parenting promotes healthy physiological development in offspring (Gunnar et al. 2015), a small but growing body of intervention research demonstrates that supporting early relational health may “get under the skin” to positively influence child biomarkers indicative of concurrent and future mental and physical health (for review, see Sullivan et al. 2024).

MODERATORS AND INTERVENTIONS THAT TARGET THEM

While evidence generally supports significant population- and sample-level average effects of early life adversity exposure on child and family outcomes, there is profound heterogeneity in effects (Boyce et al. 2021, Gee 2021), and resilience in one or many outcome domains is common (Masten et al. 2021). Most importantly, it is evident that the findings reviewed here are not necessarily predictive of a particular individual’s response to adversity exposure—a point beautifully illustrated by Baldwin et al. (2021), who showed that ACE counts are poor predictors of particular individuals’ physical and mental health, despite compelling effects on the population average. Multilevel moderating factors spanning a child’s bioecological context (Bronfenbrenner 1994) and networked, multisystem protective factors that work in concert (Masten et al. 2021) explain variability in stress effects and transdiagnostic resilience from the level of the child to society. Here, we highlight key bodies of evidence to illustrate the importance of considering the myriad factors at play, emphasizing modifiable factors and multi-programs, interventions, and policies that target them.

Child-Level Factors

Although child sex is a default consideration as a potential moderator for most researchers, an analysis of the studies reviewed yielded very mixed results. While traumatic event exposure is associated with greater levels of psychopathology for all children, some evidence suggests girls may be more sensitive to its effects. Girls evidence greater effect sizes compared to boys, particularly for internalizing symptoms (Gur et al. 2019) and biological measures such as cortisol (Niu et al. 2025). Further, other reviews find sex modifies the effects of other moderators (e.g., a 3-way interaction of sex × adversity × physiologic reactivity; Chong et al. 2022). The effects of sex may contribute to variability in response during periods in which hormones fluctuate widely and sex differences are most pronounced, such as during fetal development (Monk et al. 2019) and puberty (Dahl et al. 2018, Gunnar & Howland 2022).

An extensive body of evidence shows that individual differences in children’s reactivity and regulation—whether indicated by variations in temperament, by stress physiology, or by the genetics underlying those phenotypes—can influence whether and how adversity affects mental and physical health. Although historically thought to be risk factors, as illustrated by work drawn from biological sensitivity to context (Boyce & Ellis 2005) and differential susceptibility (Ellis et al. 2011) theories, highly reactive phenotypes have been shown to operate for better or for worse, depending on context. Specifically, while children with highly reactive phenotypes demonstrate worse outcomes in the face of adversity, they also show superior outcomes in advantaged environments or when they receive intervention in comparison with low-reactivity phenotypes. For example, a systematic review of 46 studies investigating linkages between trauma exposure, externalizing behaviors, and biomarkers (Chong et al. 2022) found consistent evidence for nonreciprocal activation of the PNS/SNS systems operating as an exacerbator of risk (moderated effects of violence), while high and low physiological reactivity promoted adaptive responses to environmental challenges. Others have demonstrated that a range of susceptibility factors, including genetic, temperament, and physiologic phenotypes, predict variation in children’s response to therapeutic interventions, with features often thought to make some vulnerable to risk actually promoting better outcomes to positive environments (for review, see de Villiers et al. 2018).

Family-Level Factors

A range of family-level factors have been found to either exacerbate or buffer children from the health and behavior risks associated with exposure to early life environmental adversity.

Caregiver mental health.

Caregiver mental health is a well-studied moderator of the association between early life adversity and child health outcomes. Indicators of prenatal and postnatal caregiver mental health moderate effects.

Psychosocial prenatal intervention.

Enhancing access to psychosocial prenatal intervention and prevention may decrease stress exposure effects on both children and birthing people. Notably, pregnancy constitutes a window of opportunity during which pregnant individuals may have frequent contact with healthcare services, which may support easier access to embedded behavioral health preventions. Interventions targeting stress and stress-linked psychopathology, such as cognitive-behavioral or mindfulness interventions, have the capacity to decrease maternal psychopathology (e.g., depression; Roubinov et al. 2022b) and support infant health outcomes (Noroña-Zhou et al. 2022). However, programs should be tailored to fit variations in cultural needs to ensure an equitable impact on pregnant people and their infants (Ponting et al. 2020).

Postnatal caregiver mental health interventions.

Interventions that target postnatal parent health also ameliorate stress effects and promote health. Parental well-being in the United States is currently in a state of crisis (Office of the Surgeon General 2024), with 48% or parents reporting their perceived stress is overwhelming (relative to 26% of nonparent adults). Parents, rather than children, are often the frontline recipient of stress exposure, and children’s health exists within the family and dyadic contexts. Thus, supporting caregivers is a high priority for preventing mental health problems in children and families.

Caregiver–child dyadic relationship and parenting.

Positive characteristics of caregiving quality, such as parental sensitivity, are a potent protective factor, buffering against stress effects on child mental health problems (see Gee 2021 for review). Review findings demonstrated that parental insensitivity or nonresponsivity to experimentally induced infant distress associated with poorer infant ANS regulation, highlighting the moderating role of parent sensitivity in infant’s stress response, prior to the onset of disorders (Jones-Mason et al. 2018). Intergenerational factors in the family system, such as maternal exposure to early childhood adversity, also influence offspring’s mental and physical health, either directly (Moog et al. 2023) or as an amplifier of pregnancy stress effects (Bush 2024).

Thus, interventions that address both parent and child health may be particularly impactful in reducing the harms of stress exposure (Roubinov et al. 2023). Dyadic interventions that target the parent–child relationship are also especially effective (e.g., child-parent psychotherapy, attachment & biobehavioral catchup; Dozier & Bernard 2023), given their capacity to bolster relational health as well as individual child and parent health. Dyadic interventions have also been found to enhance parenting for foster caregivers, bolstering support for a subgroup of children with higher adversity exposure (Dalgaard et al. 2022, Pears et al. 2012). Indeed, parenting behaviors, a key foundation to early relational health, are conceptualized as a primary mechanism of change in many child mental health interventions (e.g., in the case of externalizing disorders; Forehand et al. 2014). In our recent review, we found that interventions that act upon the parent–child relationship (dyadic interventions) improve child biomarkers (Sullivan et al. 2024b). As these biomarkers lie upstream of the mental and physical health outcomes attributed to adversity exposure (Merrill et al. 2025), their improvement implies long-reaching transdiagnostic and multisystemic influences of dyadic intervention on child health (evident at the cellular level).

Neighborhood Characteristics

It is well established that zip code matters more for one’s health than one’s genetic code (Graham 2016), and neighborhood environments have the potential to offset risks as well, with frameworks evolving from an exclusive consideration of disadvantage to a focus on opportunity. A review of 87 studies, with a subset focused on childhood outcomes, found that community-level resources spanning indicators of social, natural, and built environments had a range of associations with child physical health, well-being, and academic outcomes and might have the capacity to buffer adversity exposures, although data were too limited for a cohesive determination (Banyard et al. 2025). High-quality neighborhoods, indicated by geospatially linked indicators of socioeconomic, educational, and health opportunities, have been demonstrated to buffer the effects of low family SES on children’s physical health and daily cortisol output (Roubinov et al. 2018). Experimental evidence highlights the causal role the neighborhood plays in facilitating child health, with RCTs that assign children to opportunities for better neighborhood environments demonstrating improvements in childhood health and young adult achievement (see Shonkoff et al. 2021 for review).

School Context

A range of studies suggest the potential of school and peer contexts to buffer children from the effects of stress. For example, student–teacher relationship closeness has been found to buffer children from the risk for high physiologic activation (Roubinov et al. 2017) or increased oppositional behavior in response to harsh parenting (Roubinov et al. 2020). Better peer relationships have also been shown to be protective (Roubinov et al. 2020). Given these documented effects on child health and the relative ease of disseminating interventions to large groups of children in school settings, school-based preventions and interventions, including mindfulness-based stress reduction and holistic socioemotional curricula, may be an effective, prudent way to buffer the effects of stress on child health, particularly in settings with high degrees of community-level adversity (for review, see Fisher et al. 2020).

Social Policies

Powerful, evidence-based policies and programs can be implemented to optimize child and family health and development for those facing adversity, including cash transfer programs, increased minimum wage, community-based doula coverage, paid family and medical leave, expanded public health insurance (e.g., Medicaid) eligibility, access to early childcare/education programs (e.g., Early Head Start), comprehensive screening for social needs paired with connection to supportive programs, evidence-based home visiting programs, and access to early interventions services (see Figure 2). Further, programs that reduce child poverty likely lead to reductions in child adversity exposure (e.g., in maltreatment; Kim et al. 2024, Maguire-Jack et al. 2022). While limited research has examined how these policies directly ameliorate stress effects on child health, it is reasonable to conjecture beneficial effects on both stress and health given that many of these health-promoting policies also address fundamental causes of family-level stress (e.g., family income strain, lack of healthcare access or pregnancy support, birthing parent postpartum recovery time) (see the sidebar titled Policies Related to Assessing and Addressing Stress Exposure in Early Life).

Figure 2.

Exemplar solutions for addressing the effects of early life adversity and stress child mental and physical health and well-being. This graphic was derived from collated evidence and content provided by the Prenatal-to-3 Policy Impact Center (https://pn3policy.org). The figure is meant to highlight only some of the evidence-based promising solutions and is not meant to be exhaustive.

SIDEBAR 3: POLICIES RELATED TO ASSESSING AND ADDRESSING STRESS EXPOSURE IN EARLY LIFE.

In response to the volume of evidence documenting early life stress effects on health, some states have begun to create or modify programs and policies to address those harms. For example, California developed a statewide adverse childhood experience awareness campaign (see https://www.acesaware.org/) focused on widespread public health messaging and provider and citizen education about the health effects of adversity. This program, based in the office of the California Surgeon General, also developed policy and allocated funding to support reimbursement to primary care providers for screening for adverse childhood experience exposure in the state’s Medicaid (MediCal) patients. Importantly, mental health services for MediCal-supported youth can now be provided based on a child’s experience of trauma and risk of impairment rather than as a consequence of a formal diagnosis. Moreover, acknowledging the connection between children and caregivers’ health, MediCal also implemented family therapy and dyadic care benefits, which make it easier for youth with positive Adverse Childhood Experience (ACE) scores and their caregivers to receive treatment that bolsters caregiver–child relationships and health (Margolis 2022). These programs have been shown to be feasible and acceptable, although much remains to be learned about the costs and benefits to patients and care systems (Austin et al. 2024). Evaluations and replication efforts in other states are underway.

Multilevel Approaches to Intervention

Inequities in who is most exposed to stress, and the extent to which children and families are impacted by those stressors, are multifaceted problems requiring multifaceted solutions and universal approaches to intervention. For example, an RCT showed that children from low-income families in hospital emergency departments who were provided with in-person support for addressing social needs were less likely to be hospitalized (Pantell et al. 2020). Further, building from compelling cost-effectiveness and health promotion evidence, jurisdictions in Australia and in the state of Ohio are taking a multilevel prevention approach to promote positive parenting, such as via the Triple P program (encapsulating universal communication strategies, group-level programming, and intensive single-family interventions; Prinz 2019, Sanders 2023). These programs are promising and may pave the way to more universal application.

SUMMARY OF EVIDENCE TO DATE

We endeavored to provide a balanced overview of the extensive bodies of literature on early life stress, including systematic reviews, meta-analyses, umbrella reviews, large prospective longitudinal studies, and experimental paradigms. In sum, there is consistent evidence indicating that early life adversity exposures and experiences of stress are associated with higher levels of multidomain health problems in children. Moreover, this body of evidence strongly supports a dose–response association, with greater numbers of stressors or higher levels of stress severity relating incrementally to more health problems. The most compelling evidence was found for mental health problems, particularly internalizing and externalizing subdomains, with strong (albeit drawn from fewer studies) support for associations with neurodevelopmental disorders and neuropsychological functioning. While these bodies of literature are smaller to date, there is also consistent evidence for negative early life stress effects on social functioning and peer relationships, academic performance, and juvenile justice involvement as well as children’s physical health outcomes, including asthma, infection, somatic complaints, sleep disruption, obesity, growth, and mortality.

Exposure to adversity seems to associate with biological indicators of potential future disease states in child samples, supporting arguments that short-term adaptation may disrupt long-term health (Boyce et al. 2021, McEwen 2007). Early life stress associations with stress-relevant biomarkers, conceptualized either as outcomes in themselves or as mediators of stress effects on health, are fairly robust for measures of cortisol: Adversity exposure is associated with flatter diurnal slopes, blunted stress reactivity, and higher levels in indicators of chronic stress exposure (i.e., HCC). The findings for stress and pediatric ANS are less consistent, with evidence indicating that stress and PNS indicator associations emerge only when accounting for large variation in methods and socioeconomic backgrounds of the samples leveraged, whereas evidence for pediatric SNS indicators requires more study before the direction of effects can be confidently claimed. Stress exposures seem to associate with higher levels of inflammatory marker concentrations, with increasing effect sizes across the life span, although consistent aggregated effects are only available in a handful of proteins. Evidence in childhood for stress effects on newer emerging measures, such as epigenome-wide methylation or epigenetic age acceleration, is suggestive and requires meta-analysis. Notably, though, systematic review evidence suggests consistency in the direction of effects for measures of telomere length and epigenetic age acceleration, whereby more stress relates to more telomere length attrition and higher levels of epigenetic age acceleration. Promising findings from RCT and quasi-experimental intervention designs support a causal inference of stress effects on these child biomarkers, but effects must be replicated to enhance confidence.

Stress exposure and elevated distress states associated with unmet needs during sensitive periods in development can tune stress physiology in long-lasting ways, potentially resulting in strategies that may help cope with stress in the short term but increase long-term risks of disease in childhood and beyond (Gunnar & Howland 2022, McEwen 2007). Experimental nonhuman animal models, with the capacity to rigorously control environments and exposure timing across development, identify clear early life sensitive periods. However, rigorously isolating sensitive periods of stress exposure in humans is complicated, and we found that few studies were appropriately powered with sufficient repeated measurements to comparatively measure variability in stress exposure severity and type across development and to assess how such variability associates differentially with child outcomes. This is further complicated by the potential for stressors to have latent or sleeper effects, such that resultant phenotypes or risk factors are not apparent until years after the stress occurs (Hertzman 2012). Strikingly, despite 75% of studies reporting the presence of at least one sensitive period, a systematic review of 118 studies examining associations between child maltreatment exposure and psychiatric disorders failed to find consistent evidence for sensitive periods for most studied outcomes (Schaefer et al. 2022).

Rare exemplar studies with the capacity to precisely define sensitive periods exist, however. Neighborhood-level interventions for children exposed to deprivation, for example, have been shown to be most beneficial to children when they are provided before the age of 13 (see Shonkoff et al. 2021 for review). Dunn and colleagues (2019) used the ALSPAC cohort to show that stress exposure before the age of 3 (relative to stress occurring between the ages of 3 and 7) was associated with relatively more changes in CPG sites (DNAm). As such, it is important to advance precise sensitive period research in large prospective longitudinal studies. However, this future research need not limit the progression of policy to intervene upon stress exposures to promote child health and recovery.

Our review also highlights family health factors, such as caregiver mental health, parent–child dyadic relationship quality, and caregiving quality as key domains related to child outcomes. Thus, these factors have the potential to mediate effects of contextual stressors, such as socioeconomic stress and structural racism, on child outcomes. However, actual prospective longitudinal evidence for mediation in this manner is quite limited, only existing for child developmental and mental health outcomes but less so for physical health outcomes (likely because consideration of caregiver factors remains limited in medicine).

Across all literatures reviewed, even in the context of well-documented sample-level average associations between stress exposure and child health outcomes, we found that heterogeneity of effects predominated. Although consistent moderation patterns were not discernable across the literatures, effects were often amplified by exposure to higher adversity (often captured via sociodemographic factors) and attenuated by sensitive caregiving or higher-quality school and community environments. Particularly across the biomarker literature, most meaningful stress–biomarker associations emerged in context (i.e., not in the aggregate). This variability may be due in part to evolutionarily intended stress-system plasticity, wherein developing biology may benefit from being sensitive to external cues in order to prepare for success in the forthcoming expected environment.

It is imperative to stay aware that variation in response to stress can change over time, due to biological embedding processes or improvements in environmental supportive capacities. For example, despite the clearly deleterious effects of stress exposure during early childhood, evidence from institutionalization research shows that children also display rapid recovery upon leaving institutionalized environments, particularly in terms of physical health and brain development (van IJzendoorn et al. 2020). Stress exposure can even lead to posttraumatic growth, wherein some not only exhibit recovery following trauma but also experience personal growth in a manner that extends beyond their pre-trauma functioning. Whereas average effects suggest that adversity harms health and biobehavioral functioning, there are powerful factors, interventions, and policies we can tap into to enhance children’s resilience to stress effects (Masten et al. 2021), and the most just and economically beneficial approaches target root causes of harm rather than expecting individuals to promote their own resilience after exposure.

LIMITATIONS ACROSS THE EVIDENCE REVIEWED AND FUTURE PRIORITIES

Despite the wealth of evidence reviewed, several limitations warrant highlighting to chart a course for future endeavors. Although we prioritized rigorous longitudinal studies, most available research reviewed was observational or cross-sectional and/or used retrospective report of early life stress. Disentangling stress and child health associations needs more prospective, longitudinal studies repeatedly tracking children across development, utilizing richly assessed multidomain stress exposures and well-characterized child outcomes.

A better understanding of both normative developmental trajectories and adaptive developmental trajectories that arise in the context of stress exposure, paying attention to when such exposure occurs, is critical. Exemplar research on the development of infant cortisol patterns illustrates the necessity of characterizing normative developmental patterns: In a recent meta-analysis, researchers demonstrated that cortisol rhythms emerge gradually and stabilize at around 6–9 months of age (Kervezee et al. 2025). Therefore, a lack of diurnal rhythm before 6 months appears to be typical, in contrast to patterns observed in older children for whom flattened rhythms may be a developmentally atypical—albeit physiologically reasonable—adaptation to chronic adversity exposure (Niu et al. 2025).

Adding even more complexity, adversity may simultaneously accelerate and decelerate development, with variations in developmental tempo varying concurrently across developmental domains (for review, see Roubinov et al. 2021). Whereas biological age acceleration may forecast disease, other forms of advanced development, like walking or taking before same-age peers, are widely lauded as positive indicators.

While we should prioritize science with the capacity to highlight early phenotypes/mechanisms of effects, rather than waiting for disease manifestation, advances in our ability to identify stress-related biomarkers are still in their early days and have limited interpretability for public health stakeholders. For some biomarkers, activation directly translates to health outcomes, such as higher stress-linked systemic inflammation associating with future disease states. Other biomarkers may be stress responsive, but we lack evidence for concurrent or later clinical utility of these responses. As understanding of the roles of these biological factors during childhood increases, we will gain capacity to utilize biomarkers as screening criteria to indicate exposure to adversity, to identify which children have concerning levels of physiological reactions that should be prioritized for support, and to assess whether children are recovering or responding to treatment (Shonkoff et al. 2022). When interpreted transparently and used judiciously, objectively assessed health-relevant biomarkers of stress effects can be particularly compelling to policymakers. To many, biological measures are more readily understood as being objective and beyond one’s control, potentially addressing the tendency for mental health symptom reports to result in unfair stigma and victim blaming. Biomarkers also avoid concern about bias associated with caregiver reports of child well-being and have greater transdiagnostic relevance. Interdisciplinary demonstrations of stress-targeting psychosocial intervention effects on trauma-responsive biomarkers associated with child health outcomes (see Sullivan et al. 2024 for an example) have the potential to demonstrate to scientists, clinicians, families, and insurance and policy stakeholders the transdiagnostic impact of psychosocial treatments on preventing or addressing stress-related disease progression when it is most impactful—in childhood and across generations (Bush 2024, Shonkoff et al. 2022).

Parsing heterogeneity in associations remains a key mission for preventing harmful deterministic framing and for advancing health. The preponderance of data suggests significant average-level stress outcome effects, and yet many individuals are resilient after adversity exposure, and meta-analytic effects were characterized by significant variability that merits interrogation. Further, a lack of main effects (e.g., in some biomarker outcomes) may not mean that there are no effects but instead may call for enhanced model complexity and precision, leveraging better measurement (see Epel et al. 2018 for recommendations) with an eye toward moderators. More nuanced approaches can inform precision medicine approaches for identification of risk and solution generation. This variability-explaining task is essential for parsing intervention effects, too (see Fisher et al. 2020).

Multidisciplinary, participant- and community-engaged approaches bridging psychology, pediatrics, and public health are critical to run these studies, which require novel funding structures as well as tolerance of differences in field-specific vernaculars, statistical preferences, and publishing practices. Several prospective longitudinal consortia that track development starting in pregnancy (e.g., the ECHO program) embody these values and hold great potential in this regard. Further, we need training opportunities that support the development of interdisciplinary scientists, savvy at field cross-talk and science communication, who can translate findings for and advocate to stakeholders, and ideally many of these scientists would come from the communities targeted, bringing lived experience and insights to the work.

Further, future research initiatives need to collect data from generalizable populations as well as subpopulations with unique experiences. While stress is a universal experience that cuts across race and class, meriting greater efforts to prevent exposure for all, stress exposure is inequitably distributed. Historically marginalized groups (e.g., gender-diverse youth, children of color) are more likely to be exposed to stress and to contend with stress-linked health problems with fewer supportive resources (Priest et al. 2024, Tankersley et al. 2021). Therefore, there is high value in inclusive research and social justice efforts that center on the margins and bring the perspectives of the populations that experience the greatest levels of stress to the forefront (Geronimus et al. 2023). However, the child health and development literatures are often deeply rooted in whiteness and Eurocentric cultural values. To address this limitation, throughout this review we intentionally highlighted findings derived from large, sociodemographically diverse samples (though often difficult to find) and prioritized those using equity-centering frameworks. Additionally, while some have recently expanded the field’s maternally focused family research to include fathers, most work still fails to capture the diverse kin networks and cultural practices and perspectives that children are often raised within. Accordingly, most evidence lacks a consideration of critical contexts or nuances in which children develop, including diverse aspects of family- and culture-bound strengths that interventions can tap into.

Finally, to date, only a few (albeit quite compelling) instances of human studies permitting causal inferences exist. We should endeavor to better leverage mechanistic intervention and experimental designs to rigorously and ethically shed light on causality links between stress and child health and to ameliorate them (e.g., Baldwin et al. 2023, Sullivan et al. 2024a).

CONCLUSION

Robust evidence shows that early life adversity and stress have a transdiagnostic impact on child health, spanning mental and physical health and affecting intermediatory biobehavioral markers. The influence of stress on child development and health begins early in life, and the effects can be cumulative, with greater potential for harm when exposures accumulate, occur during sensitive periods of development, and lack accompanying protective factors. Thankfully, exposure is not deterministic, and considerable variation in effects suggests the value of precision medicine approaches to screening and treatment. Further, supportive family, classroom, and community environments are powerful buffers and offer multifaceted opportunities for resilience.

Preventing traumatic or chronic stress exposures is ideal. However, it is clear that early intervention is effective after stress has occurred and can initiate shifts in developmental trajectories away from impairment or disease that have the capacity to cascade and compound across time, creating profound individual, public health, and economic benefits (Heckman 2006). Amplifying and extending this hope-bolstering science are critical tasks, as is translating it to stakeholders in positions to implement such efforts. Multilevel problems require multilevel solutions, ranging from broad, universal policies supporting family wellness to individual-level precision medicine approaches. By reviewing the evidence and highlighting its strengths and limitations, we endeavor to empower others to acknowledge the potentially enduring, pervasive harm from stress and to advance the just goal of improving the health and wellness of all children, families, and future generations.

SUMMARY POINTS.

1. The volume and quality of science on stress and health is rapidly expanding, but research and summaries thereof often focus on adult samples, failing to consider how early life stress influences health during infancy, childhood, and adolescence—unique periods of biological, behavioral, and social development that pave the foundation for adult health. Sufficient empirical and aggregate empirical (e.g., meta-analytic) evidence exists for a synthesis focusing on pediatric samples.

2. Rigorous aggregated evidence from pediatric samples consistently shows that stress is positively associated with multidomain poorer health and general functioning outcomes. These outcomes include mental and physical health, developmental outcomes, executive functioning, social and academic functioning, justice system involvement, and positive health. The strength of associations typically follows dose–response gradients, with exposure to more stressors relating to incrementally worse levels of problems.

3. While the volume of rigorous evidence is smaller than the one derived from investigations of mental health outcomes, exposure to adversity also clearly associates with biological indicators of potential future disease states in child samples. The strongest evidence suggests that early adversity is associated with flatter diurnal cortisol slopes, blunted cortisol stress reactivity, higher hair cortisol concentrations, higher inflammatory marker concentrations, shortened telomere length, and higher levels of epigenetic age acceleration.

4. Stress occurring earlier in the life span (e.g., in sensitive periods such as infancy and toddlerhood) may more durably program physiology relative to stress later in life, translating to enduring negative health effects of stress exposure. However, optimally designed methodological investigation across the life course is rare, and there is evidence for sensitive periods for various exposure-outcome relationships across development, suggesting that adversity experienced throughout childhood and adolescence is of concern.

5. Associations between stress and child outcomes are highly variable, particularly at the level of individuals. Factors spanning individual- to macro-level moderators breakdown this variability. These moderators include, but are not limited to, children’s individual differences in reactivity and regulation capacity, family sociodemographic factors, parenting behaviors and other caregiver-level factors, school and neighborhood contexts, and social policies. These moderators present opportunities for prevention and solutions to promote child health in the context of stress exposure.

6. Family-level factors, particularly caregiver health, are key mediators of stress effects on child health outcomes. As such, stressors that disrupt caregiver wellness are likely to impact children, and resources that promote caregiver functioning and wellness can have downstream positive effects on child health and well-being.

7. Stress is a common human experience, and although not deterministically, the effects of traumatic or chronic stress on children and families are found in samples across all levels of socioeconomic status, across sexes, race/ethnicities, cultures, and regions. Population-level prevention of traumatic experiences is likely to benefit society broadly, although some communities bear disproportionate risk and equity-focused efforts are likely to have the greatest society social and economic impact.

8. When the prevention of stress exposure is infeasible, early interventions attenuate stress effects on child health and shift lifelong trajectories away from disease states, promoting resilience. These programs include caregiver–child psychosocial treatment, prenatal and postnatal caregiver mental health treatment, school-based programs, and social policies addressing the fundamental causes of family- and community-level stress.

FUTURE ISSUES.

1. Advancing research on stress and health associations in children requires appropriately powered, prospective longitudinal studies repeatedly assessing children across development with richly assessed multidomain stress exposures and well-characterized child outcomes.

2. Multidisciplinary, participant- and community-engaged approaches bridging psychology, pediatrics, and public health are critical for advancing understanding in this realm. This requires novel funding structures as well as tolerance of differences in field-specific vernaculars, statistical preferences, and publishing practices.

3. We need to better understand both normative developmental trajectories and variability in adaptive developmental trajectories that arise in the context of stress, with attention toward measuring precisely when that stress exposure occurs to refine our detection of exposure, potential resilient coping, and risk for later impairment.

4. Pediatric stress biomarker research is still in its early days. Our capacity to leverage stress-related biomarkers to predict and understand health outcomes remains limited and requires more investment and attention.

5. Despite average negative associations with health, many children exposed to adversity exhibit healthy outcomes. Research should endeavor to document associations between stress and child health and identify factors that account for variability in those linkages, which is critical for solution generation and identification of those in greatest need.

6. Strategic studies, leveraging mechanistic intervention and quasi-experimental designs, are key to rigorously and ethically illuminating causality links between stress and child health and to ameliorate them.

7. As policymakers embrace the social, population health, and economic wisdom of developing and enforcing policies to prevent and address early life adversity, it will be critical to evaluate and document the improved outcomes and cost savings associated with those changes for children, families, and communities.

Glossary

Allostatic load

theoretical framework and measurable physiological variable capturing the cumulative physiological costs (“wear and tear”) of individuals’ adaptations to chronic stress

Biological embedding

theoretical framework explaining how major stressors “get under the skin” and influence physiological processes, potentially leaving lasting mark on individuals

Toxic stress

high levels of uncontrollable stress with the potential to harm child health unless offset by buffers (e.g., high-quality caregiving)

Biological sensitivity to context

theory suggesting individuals vary in their susceptibility to positive and negative environmental influences, depending on their biological reactivity to stress

Developmental origins of health and disease (DOHaD)

a theoretical framework positing that early life, including prenatal, experiences have long-lasting effects on health

Intergenerational transmission of stress

a theoretical framework describing how traumatic and adverse experiences of one generation influences the health of subsequent generations

Social determinants (or social drivers) of health

framework examining nonmedical factors, such as socioeconomic status, environmental quality, and healthcare access, which can drive health outcomes

Historical trauma

collective trauma inflicted on a group of people who share a specific identity, transmitted across generations through biological and cultural processes

Positive health

health measured based on features of well-being, meaning and purpose in life, and reserve in capacity more than on the absence of disease

Sensitive period

a time-period of heightened biological system plasticity when exposures (e.g., to stress or certain chemicals) have particularly impactful effects on development

Early relational health

the quality of a child’s social and emotional connection to a primary caregiver, potentially conferring enhanced well-being

Attachment

quality of a child’s relationship with a primary caregiver, typically categorized as secure or insecure