1. Introduction

Adverse childhood experiences (ACEs) have been associated with a broad spectrum of deleterious health outcomes throughout life. The prevalence of childhood violence and the resulting substantial economic losses—estimated at hundreds of billions of dollars annually due to healthcare costs and productivity losses—highlight the profound burden of early-life adversity.^1,2 These broad societal effects suggest that ACEs result in acute trauma and trigger a cascade of long-term psychological and biological dysregulation.^3,4 Although the effects of ACEs on cardiovascular⁵ and metabolic⁶ health are well established, their widespread systemic burden suggests that they compromise other complex physiological systems, particularly the gastrointestinal tract, through similar stress-mediated pathways.^7,8 However, the association between ACEs and digestive diseases remains insufficiently characterized.

The biological embedding of childhood adversity occurs through neuroendocrine alterations, including dysregulation of the hypothalamic–pituitary–adrenal (HPA) axis and epigenetic modifications.^2,9 Chronic stress exposure induces persistent inflammatory responses that contribute to accelerated disease pathogenesis later in life.^10,11 Current evidence further demonstrates a robust dose–response relationship between ACEs and depression, with each additional ACE associated with a 20%–50% increase in depression risk and exposure to ≥ 4 ACEs associated with a 2.3–2.65-fold increase in the odds of depression.¹² These psychological disturbances are not merely comorbid conditions but key mediators through which early-life trauma contributes to subsequent physical pathology.

Digestive disorders are particularly sensitive to stress and are regulated by the autonomic nervous system and neuroendocrine pathways.^13,14 Stress-induced disruptions in the brain–gut axis can impair gastrointestinal motility, increase intestinal permeability, and enhance visceral sensitivity, contributing to chronic pain and various gastrointestinal conditions.^15,16 Although previous studies have associated ACEs with specific functional conditions such as irritable bowel syndrome (IBS),^17,18 their associations with a broader spectrum of gastrointestinal and esophageal diseases remain inadequately explored, particularly in developing nations like China.

This study addresses these knowledge gaps through a secondary analysis of data from the China Health and Retirement Longitudinal Study (CHARLS).¹⁹ Using data from 5145 middle-aged and older Chinese adults, we aimed to comprehensively examine the associations between ACEs and multiple digestive disorders. We utilized a lifespan stress accumulation model to examine whether depression serves as a modifiable mediator in the pathway linking ACEs to digestive disorders, thereby providing clinically actionable evidence for interventions targeting high-risk aging populations.

2. Methods

2.1. Ethical approval

Ethical approval for CHARLS was granted by the Institutional Review Board of Peking University (Approval Number: IRB00001052-11015).²⁰ All participants provided written informed consent.

2.2. Sample and study design

For this study, we used the data from CHARLS, an ongoing, nationally representative longitudinal survey initiated in 2011. CHARLS includes participants from 150 counties or districts and 450 villages or urban communities across 28 provinces in China. The survey employs a multistage stratified sampling design based on probability-proportional-to-size principles. To date, five major waves of surveys (in 2011, 2013, 2015, 2018, and 2020) and several special surveys (in other years) have been conducted under the CHARLS project. The comprehensive research framework and details have been described previously. In the present study, we used data on ACEs from the life-history survey conducted between June 1, 2014, and December 31, 2014, and the outcome data from the subsequent main survey, conducted between July 1, 2015, and September 30, 2015.  The study was conducted and reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist.

A successful matching process was conducted, which identified 18,680 participants from the 20,452 individuals recruited in the 2014 CHARLS life-history survey for ACE assessment and the 21,097 individuals recruited in the 2015 follow-up survey for additional health data collection. Participants were required to have completed both surveys. They were excluded if they had missing ACE data (n = 11,328), were aged < 45 years or had missing age data (n = 956), or had missing depression data (n = 281). They were also excluded if they lacked information on lifestyle covariates and education (n = 960) or had extreme body mass index (BMI) values (BMI < 10 or BMI > 100, n = 10), as these values may indicate severe comorbidities, result in data-entry errors, or are unlikely to be representative of the general population. The final sample chosen for analysis comprised 5145 participants. The flowchart of the participant selection procedure is shown in Fig. 1. To assess potential selection bias, an attrition analysis was performed by comparing the baseline characteristics between the final sample chosen for analysis (n = 5145) and the excluded participants (n = 13,535), including those with missing data on ACEs or covariates. Differences between groups were evaluated using the chi-square test for categorical variables and the Wilcoxon rank-sum test for continuous variables. Additionally, standardized mean differences (SMDs) were calculated, with an SMD < 0.1 indicating a negligible difference between groups.

2.3. Measurements

2.3.1. ACEs

The cumulative ACE score, ranging from 0 to 12, was calculated by summing 12 dichotomized childhood adversity indicators derived from the original ACEs,²¹ expanded ACEs,²² and more recently identified ACEs^23,24 (Supplemental Table 1). These indicators included parental separation, physical abuse, domestic violence, emotional neglect (low affection or supervision), household mental illness, substance abuse, incarcerated household members, residence in an unsafe neighborhood, bullying, parental death, parental disability, and sibling death. Based on responses to the ACE questionnaire, each ACE indicator was coded as 0 (absent) or 1 (present), and the scores were summed to generate a cumulative ACE score for each participant.

2.3.2. CESD-10 score

The 10-item Center for Epidemiologic Studies Depression Scale (CESD-10)²⁵ has been widely used in CHARLS to assess depressive symptoms among older adults. The CESD-10 score demonstrates high internal consistency (Cronbach’s α = 0.69–0.89) and validity and has been validated through confirmatory factor analysis in Chinese middle-aged and older populations.²⁶ In CHARLS, the scale is administered through face-to-face interviews using a 4-point Likert scale (ranging from 0 = never to 3 = most of the time), with a cutoff score of ≥ 10 indicating clinically significant depressive symptoms.

2.3.3. Digestive diseases

Digestive disease status was self-reported and defined as (1) an affirmative response to the question “Have you been diagnosed with a stomach or other digestive disease (excluding a tumor or cancer) by a doctor?”, or (2) self-reported receipt of current treatment for a digestive disease or its complications, based on an affirmative response to the question “Are you currently receiving any of the following treatments for digestive diseases or their complications?”, with treatment status serving as evidence of a diagnosis. During follow-up, participants were asked to reconfirm their initial disease reports, and analytical adjustments were made to mitigate potential recall bias.

2.3.4. Control variables

According to established covariate-selection frameworks in life-course epidemiology and previous studies on digestive diseases,^27–29 the following baseline characteristics were included: age, sex, marital status (single or married/partnered), highest educational attainment (no formal education, elementary school, middle school, or high school and above), residence (urban or rural), smoking status (never smoked or ever smoked), drinking status (never consumed alcohol or ever consumed alcohol), and BMI calculated from self-reported height and weight.

2.4. Statistical analysis

Sample characteristics were summarized using means ± standard deviations for continuous variables and percentages for categorical variables. Spearman’s correlation was used to assess the bivariate associations among key variables. To evaluate the mediating role of depressive symptoms in the association between ACEs and digestive diseases,³⁰ a multivariable mediation analysis was performed within the counterfactual framework using the mediation package in R. The average causal mediation effect (ACME; indirect effect) and average direct effect (ADE) were estimated, with robust 95% confidence intervals constructed via a nonparametric bootstrap procedure with 1,000 resamples.³¹ During the variable screening process, ACE components with near-zero variance or highly skewed distributions (prevalence > 99% or < 5 positive cases) were excluded from the individual component analyses to improve model stability and reduce the risk of spurious correlations. We used R version 4.5.1 for all statistical analyses.

3. Results

3.1. Participant characteristics

We analyzed the data of 5145 middle-aged and older adults who participated in the CHARLS study. Attrition analysis was performed to compare the participants included in the final analysis (n = 5145) with those excluded because of missing data or extreme BMI values (n = 13,535). Although the participants in the included group were older (median age: 64 versus 57 years, SMD = 0.582) and had a slightly different educational distribution, no significant differences were observed in the primary outcome of digestive disease (33.2% versus 32.0%, P = 0.133, SMD = 0.026) or mental health status (CESD-10 score, SMD = 0.094[1] ) (Supplemental Table 2). The mean age of the participants in the entire sample was 63.89 ± 8.49 years, and the sex distribution was approximately balanced (49.1% female, 50.9% male). Most participants were married (84.9%) and resided in rural areas (66.2%). Educational attainment was generally low, with only 9.7% of participants reporting education beyond high school and nearly half (47.3%) being in the lowest educational category. Participants had a median ACE score of 2 (interquartile range (IQR): 2–3), indicating moderate exposure to childhood adversity. The median depressive symptom score was 7 (IQR: 3–12), and the prevalence of digestive diseases was 33.2% (Table 1). The prevalence of individual ACEs is presented in Supplemental Table 3.

3.2. Associations between childhood adversity, depression, and digestive diseases

Correlation analyses revealed significant associations between the key study variables. ACE scores showed a small but significant positive correlation with depressive symptoms (r = 0.14, P < 0.001), indicating that individuals with greater childhood adversity tended to report more depressive symptoms later in life. Similarly, ACE scores were positively correlated with digestive diseases (r = 0.12, P < 0.001), suggesting that early-life adversity may contribute to digestive health problems in adulthood. The strongest correlation was observed between depressive symptoms and digestive diseases (r = 0.54, P < 0.001), highlighting the close relationship between psychological distress and digestive disorders in this population.

The associations between childhood adversity, depressive symptoms, and digestive diseases were further clarified using multivariable logistic regression analyses. In Model 1, adjusted for age, sex, educational attainment, BMI, and drinking status, a higher ACE total score was significantly associated with increased odds of digestive diseases (odds ratio (OR) = 1.11, 95% confidence interval (CI): 1.06–1.16, P < 0.001). Among the covariates, increasing age was modestly associated with lower odds of digestive disease (OR = 0.99, P < 0.05), and men had substantially lower odds compared with women (OR = 0.67, P < 0.001). Higher BMI was also inversely associated with digestive diseases (OR = 0.97, P < 0.001), whereas drinking status was associated with a higher risk (OR = 1.15, P < 0.05). With respect to education, individuals with middle school attainment had significantly lower odds of digestive disease compared with those with no formal education (OR = 0.72, 95% CI: 0.60–0.85, P < 0.001), while an education level of elementary school and high school or above showed no statistically significant associations. In Model 2, after further adjustment for depressive symptoms (measured by the CESD-10 score), they were found to be strongly and positively associated with digestive diseases (OR = 1.05, 95% CI: 1.04–1.06, P < 0.001). Importantly, the association between ACE total score and digestive diseases remained statistically significant but was attenuated (OR decreased from 1.11 to 1.07, P < 0.05), suggesting partial mediation by depressive symptoms. A similar attenuation was observed for educational attainment: the protective association of middle school education was reduced but remained significant (OR = 0.79, 95% CI: 0.66–0.94, P < 0.05) (Table 2).

In the individual component analysis, household mental illness was excluded because its high prevalence (99.11%) resulted in an insufficient size of the control group, rendering statistical estimates noninformative. Similarly, parental separation was excluded due to its low frequency (n = 1), which could lead to extreme ORs and unstable CIs. The remaining 10 ACE components were used to assess specific associations with digestive health. Notably, parental disability showed the strongest association (OR = 1.41, 95% CI: 1.22–1.62), followed by emotional neglect (OR = 1.19, 95% CI: 1.05–1.35) and bullying (OR = 1.19, 95% CI: 1.01–1.40). Other ACE indicators, such as sibling death and incarcerated household members, were not significantly associated with digestive diseases (Fig. 2).

3.3. Mediating role of depressive symptoms in the ACE–digestive disease pathway

The results of the multivariable causal mediation analysis, using the final regression models and bootstrap estimations, support the pathways linking early-life adversity, psychological distress, and adult digestive health. In the total effect model adjusted for age, sex, educational attainment, BMI, and drinking status, higher ACE total scores were significantly associated with an increased risk of digestive diseases (B = 0.023, 95% CI: 0.013–0.034, P < 0.001). In the mediator model that further adjusted for the same covariates, childhood adversity was robustly and positively associated with depressive symptoms (B = 0.723, standard error (SE) = 0.070, P < 0.001), supporting a strong exposure–mediator relationship. When depressive symptoms were included in the fully adjusted outcome model, CESD-10 score was independently associated with a higher risk of digestive diseases (B = 0.011, SE = 0.001, P < 0.001); the direct association between ACE total score and digestive diseases remained statistically significant but was attenuated (B = 0.016, P = 0.003), indicating partial mediation. A formal causal mediation analysis using nonparametric bootstrapping with 1,000 replications confirmed the statistical significance of both the indirect and direct pathways. The ACME was 0.0078 (95% CI: 0.0056–0.0101, P < 0.001), while the ADE was 0.0156 (95% CI: 0.0057–0.0259, P = 0.004). The total effect of childhood adversity on digestive diseases was 0.0233 (95% CI: 0.0135–0.0339, P < 0.001). Overall, depressive symptoms accounted for approximately 33.3% of the total association between ACEs and adult digestive diseases (proportion mediated = 0.333, 95% CI: 0.216–0.585, P < 0.001) (Fig. 3).

To evaluate the consistency of the mediating effect across demographic strata, we performed stratified mediation analyses based on sex, age group, and residence. The indirect effect of ACEs on digestive diseases through depressive symptoms remained significant across all subgroups (for all ACMEs, P < 0.001), confirming the robustness of the psychological pathway (Table 3). Notably, the proportion of the total effect mediated by depression varied across subgroups, with a larger mediated proportion observed among women (40.9%) than among men (26.4%). In the female subgroup, the ADE was not significant (P = 0.240), suggesting near-complete mediation. Furthermore, participants aged ≥ 65 years exhibited a substantially higher mediation proportion (55.1%) than those aged 45–64 years (18.7%). For the older group, the direct effect of ACEs on digestive health was no longer significant (P = 0.272). Finally, the mediation effect was slightly higher among rural (33.0%) than among urban (27.0%) residents (Table 3).

4. Discussion

To our knowledge, this is the first study to investigate the mediating role of depressive symptoms in the relationship between ACE scores and susceptibility to digestive diseases in a population-based cohort of older Chinese adults. Participants included in our study were generally older than those excluded (Supplemental Table 2). Because digestive disorders become more prevalent with age, we adjusted for age as a continuous covariate in all multivariate models to minimize the potential selection bias. Our analyses showed that ACE scores were significantly associated with the risk of digestive diseases through both direct and indirect pathways. Depression significantly mediated this association, accounting for 33.3% of the total effect (95% CI: 21.6%–58.5%; ACME = 0.0078, SE = 0.0011, P < 0.001). This relationship may be attributed to the neurobiological mechanisms linking early-life adversity to long-term alterations in stress-response systems,^32,33 which can contribute to the development of digestive pathology later in life.^6,7 Reverse causality (digestive diseases → ACEs) is unlikely because ACEs are fixed childhood exposures, supporting the temporal plausibility of the observed mediation pathway. However, longitudinal studies are needed to confirm the trajectory from depressive symptoms to clinically diagnosed digestive disease. Each one-unit increase in ACE score was associated with a 2.4% increase in the risk of digestive diseases (B = 0.023, SE = 0.005, P < 0.001; 95% CI: 0.013–0.034) after adjustment for demographic and lifestyle covariates, including age, sex (B = −0.087, SE = 0.016, P < 0.001; SE = 0.016, P < 0.001 for men compared with women), educational attainment, BMI (B = −0.007, SE = 0.002, P < 0.001), and alcohol consumption.

The association identified in our preliminary analysis (r = 0.54) likely reflects the complex, bidirectional dynamics of the gut-brain axis, in which psychological distress and digestive disorders operate within a continuous feedback loop. Depressive symptoms may induce dysregulation of the HPA axis and cortisol-induced alterations in gut motility, resulting in increased visceral sensitivity. In contrast, chronic gastrointestinal discomfort and microbiome dysbiosis can generate proinflammatory signals and transmit them to the central nervous system via the vagus nerve, inducing neuroinflammation and “sickness behaviors” that mirror clinical depression. In addition to these physiological mechanisms, this strong correlation may be amplified by subjective self-report bias, as individuals experiencing psychological distress may perceive somatic symptoms more acutely (somatization), while chronic physical illness can fundamentally bias a respondent’s psychological perception. Accordingly, the transition to multivariate adjustments in Model 2 aimed to isolate the unique variance of these domains and adjust for this inherent bidirectional confounding.

The mediation pathway operates through psychoneuroimmunological mechanisms linking childhood adversity, depression, and digestive disease. ACE scores were positively associated with depressive symptom severity in a dose-dependent manner (B = 0.723, SE = 0.070, P < 0.001 per ACE unit), corresponding to an approximately 0.72-point increase in CESD-10 score for each additional ACE. Depressive symptoms were, in turn, positively associated with the risk of digestive diseases (B = 0.011, SE = 0.001, P < 0.001 per 1-point increase in the CESD-10 score), consistent with the proposed role of brain-gut axis dysregulation. This pattern translated into clinically significant risk differences, with individuals exhibiting clinically significant depressive symptoms (CESD-10 ≥ 10) experiencing an 11% greater digestive disease burden than those without depression, despite the comparable ACE exposure. The depression pathway operates primarily through corticotropin-releasing factor (CRF)-mediated increments in intestinal permeability, vagal nerve dysfunction impairing gut motility, and depression-associated health behaviors disrupting microbial diversity.^34–36 The mediation proportion of 33.3% was statistically significant (P < 0.001), highlighting the role of depression as a critical translational mechanism converting psychological distress into gastrointestinal pathophysiology.

Nevertheless, the persistent direct effect after adjustment for depression (B = 0.016, SE = 0.005, P = 0.004; accounting for 66.7% of the total effect) indicates complementary biological mechanisms independent of mood disturbances, including epigenetic modifications in glucocorticoid receptor genes altering the HPA axis sensitivity, permanent alterations in gut-associated lymphoid tissue development, and microbiome-immune crosstalk established during critical developmental windows.^13,37,38 This interpretation is supported by preclinical evidence demonstrating that early-life stress permanently alters enteroendocrine cell density and intestinal stem cell function through DNA methylation changes that persist until late adulthood.^2,28,39,40

Notably, women consistently exhibited a 9.2%–11.3% higher age-adjusted risk of digestive disease across models (P < 0.001). This finding is consistent with previous evidence of sex differences in stress responses and visceral pain processing.^41,42 Higher BMI was associated with a 5.0%–6.0% lower risk of digestive disease per unit increase (P < 0.001). This unexpected association may reflect factors such as greater nutritional reserves among older adults with sarcopenia or differences in adipose tissue-derived anti-inflammatory adipokine production.⁴³ These findings also suggest that cognitive reserve primarily buffers the risk of digestive diseases through pathways associated with psychological resilience.^44,45

The observed association between ACEs and digestive disorders may reflect alterations in the gut-brain-microbiota axis, although the underlying mechanisms are likely to differ across disease subtypes. Early-life adversity has been associated with dysfunction of the prefrontal-limbic-brainstem circuitry, in which reduced top-down regulation by the prefrontal cortex may contribute to increased amygdala reactivity and subsequent activation of sympathetic stress-response pathways.⁴⁶ In functional disorders such as IBS, ACEs contribute to dysregulation of the HPA axis, sensitization of the enteric nervous system, and vagal dysfunction, resulting in increased gut motility and enhanced visceral hypersensitivity.^47,48 In contrast, organic conditions such as inflammatory bowel disease and peptic ulcer disease are more strongly driven by immune-inflammatory pathways, including increased intestinal permeability and mucosal injury.⁴⁹

Methodologically, this study is strengthened by its nationally representative sample (N = 5,145), comprehensive covariate adjustment, and bootstrap validation of mediation effects. However, the interpretation of the findings must account for the cohort-specific nature and historical context of the CHARLS population. The participants in this study (born predominantly before 1970) experienced unique socioeconomic transitions in China, where early-life adversity often co-occurred with systemic nutritional scarcity. Such experiences may lead to distinct epigenetic and physiological adaptations compared with those in younger cohorts. Additionally, improvements in healthcare systems and social security in recent decades may have influenced the relationship between ACEs and digestive diseases, as improved living conditions could either mitigate the long-term effects of early-life trauma or, conversely, reveal latent vulnerabilities as life expectancy increases.⁵⁰ Finally, changes in health awareness should be taken into account. With the improvement in mental health literacy in China in recent decades, participants may be more likely to recognize and report depressive symptoms and digestive complaints than in the past.⁵¹ This increasing awareness may influence the observed magnitude of the mediation effect, suggesting that the relationship among ACEs, depression, and digestive diseases reflects not only biological processes but also broader sociocultural changes.

From a clinical perspective, these findings support the reconceptualization of certain functional gastrointestinal disorders as potential manifestations of developmental trauma.^52,53 Gastroenterology practices should consider implementing standardized ACE screening protocols using validated instruments such as the World Health Organization Adverse Childhood Experiences International Questionnaire, whereas integrated treatment models may incorporate trauma-focused cognitive behavioral therapy to manage depression among patients with high ACE exposure and persistent digestive symptoms.⁵⁴ Community-level prevention efforts should prioritize school-based resilience programs for children exhibiting indicators of early adversity, potentially interrupting the pathway linking ACEs, depression, and digestive disease before long-term biological changes become established. For research translation, future studies should evaluate multispecies psychobiotics that may simultaneously influence mood and gut barrier function in those with ACE. Ultimately, this study demonstrates that childhood adversity is associated with an increased risk of digestive disease through partially distinct but interconnected pathways, with depression serving as a clinically modifiable mediator that accounts for approximately one-third of the observed association. The estimated effect sizes provide actionable thresholds for clinical intervention; in particular, participants with ACE scores ≥ 4 and CESD-10 scores ≥ 10 may represent a subgroup that could benefit from multidisciplinary assessment and management.

Nevertheless, several limitations of this study should be acknowledged. First, selection bias may be present, as the attrition analysis indicated that participants with complete data tended to be older than those who were excluded. However, the similar prevalence of digestive diseases and comparable mental health profiles observed between the included and excluded groups suggest that this attrition is unlikely to have substantively biased the primary associations. To mitigate the effects of age, education and demographic discrepancies were adequately adjusted for in all multivariate models. Second, reverse causality (digestive diseases → ACEs) is unlikely because ACEs occur during childhood, supporting the temporal plausibility of the mediation pathway, although longitudinal studies are needed to confirm this relationship. Third, unadjusted cognitive function may have affected the recall accuracy of childhood experiences, warranting the inclusion of this covariate in future studies. Additionally, the status of the digestive disease was self-reported and lacked objective clinical confirmation (e.g., endoscopy), potentially leading to misclassification. Future longitudinal studies should incorporate assessments of stress-related biomarkers to better characterize the ACE–depression–digestive disease pathway over time. Meanwhile, mechanistic studies should investigate ex vivo intestinal organoid responses to physiological stress among individuals with high ACE exposure to further elucidate the biological mechanisms linking psychological trauma and digestive dysfunction.

5. Conclusion and implications

Using CHARLS data, we demonstrated that ACE scores are associated with digestive disease risk through two pathways: a direct pathway and an indirect pathway mediated by depressive symptoms, which accounted for 33.3% of the total effect. The depression-mediated component suggests that interventions targeting mental health may help reduce the risk of digestive diseases among individuals with ACE. These findings align with the brain–gut axis model, which proposes that childhood adversity contributes to digestive dysfunction through both psychological and biological pathways. The results highlight the potential value of integrated mental health and gastroenterological care as well as early-life interventions aimed at reducing long-term disease risk. By bridging developmental neuroscience and gastroenterology, this study supports a lifespan perspective on digestive health in which early adversity, psychological distress, and digestive dysfunction represent interconnected domains that may benefit from integrated biopsychosocial approaches.