Role of body mass index in the relationship between adverse childhood experiences, resilience, and mental health: a multivariate analysis

This cross-sectional study used an extensive survey procedure to assess the sociodemographic and psychosocial environmental factors that contribute to the mental health and health risk behaviors of the general population [19, 20]. The present study was designed and reported according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist. Before the investigation, all on-site investigators were trained, and the working standards were unified.

A cross-sectional study design was used to conduct nationwide sample surveys in Anhui Province, China. We took into account both the sampling method and the partnership. We randomly selected 2 schools according to the geographical distribution and the degree of school cooperation [21‐23]. The study participants were college students sampled from May 2022 to June 2022 in Anhui Province using multi-stage convenience sampling method. The study subjects were class-based, and students were required to complete the questionnaire anonymously on the spot during non-class hours, according to the principle of informed consent. The teacher maintained order in the classroom for approximately 20–30 min, and the quality control personnel answered the questions of the survey subjects on the spot and were responsible for collecting and reviewing the questionnaire. The design and data collection procedures were approved by the Ethics Committee of Anhui Medical University. Further, written informed consent was obtained from the parents or guardians of all students. All methods were performed per relevant guidelines and regulations.

Participants: (1) obtained informed consent from the participants and their guardians; (2) were college students [aged 15–26 years]; (3) had no history of mental illness; and (4) were attending the chosen school.

Participants: (1) Informed consent was not obtained; (2) age between 15 and 26 years; (3) failure to submit the questionnaire; and (4) not college students from the chosen school.

Sex, age, parental education level, residential area, only child status, family economic status, number of friends, and academic records were included as covariates.

Continuous variables were described using mean and standard deviation (SD) and one-way analysis of variance. Categorical variables were described using frequencies and percentages using the Chi-square test. Multivariate logistic regression was conducted to evaluate the relationships among life events, chronotype, mental health, and health risk behaviors, presented as adjusted odds ratios (aORs) with 95% confidence intervals (CIs). Moderate mediation analysis was employed using the PROCESS method to explore the relationships among ACEs, BMI, resilience and mental health [29]. The bootstrap method was used to re-sample 1000 samples, and 95% CI were calculated. All analyses were conducted using SPSS Windows software version 23.0. After preliminary data sorting, missing data were processed using SPSS. Generally, the missing data rate in this study was meager, with the missing rate for each item being less than 1%. Therefore, a multiple imputation method was adopted in SPSS to process missing data at the project level.

Among the 3578 students in the questionnaire included in the analysis, 1833 were male students, and 1745 were female students; the age was (20.53 ± 1.65) years old; 1665 (46.5%) lived in rural areas, 783 (21.9%) lived in towns, and 1130 (31.6%) live in cities. There were 1180 only children (33.0%) and 2398 non-only children (67.0%). Among them, 2100 (58.7%) had junior high school education or below, and 1350 (37.7%) had senior high school education or above. A total of 2457 (68.7%) of mothers had junior middle school education or below, and 968 (27.1%) had senior high school education or above. Among them, 1103 (30.8%) were classified as poor, 2242 (62.7%) as moderate, and 233 (6.5%) as good. A total of 1272 (35.6%) had less than two friends, 1730 (48.4%) had three to five friends, and 576 (16.1%) had more than six friends. The prevalence rates of depression and anxiety symptoms were 37.5% and 11.6%, respectively. The demographic characteristics and mental health assessment results are presented in Tables 1 and 2.

In Table 3, there was a dose‒response relationship between different ACEs and depression (emotional abuse: β = 1.13, physical abuse: β = 1.21, sexual abuse: β = 1.28, emotional neglect: β = 0.97, physical neglect: β = 0.16, ACEs: β = 0.62). After controlling for covariates, the relationships were statistically significant. There was also a negative dose–response relationship between resilience and depression (β=-0.28), and after controlling for covariates, this relationship was significant (β=-0.27). Similar results are presented in Tables 4 and 5.

Moderation analyses were conducted for emotional abuse, resilience, BMI, and mental health. Table 6 presents the results. First, emotional abuse did not correlate with the severity of depression (P > 0.05). Resilience was associated with depression (P < 0.01), emotional abuse× resilience did not significantly correlate the severity of depression, emotional abuse× BMI did not significantly correlate the severity of depression, and resilience× BMI was correlated with depression. There was no three-way interaction effect between emotional abuse, resilience and BMI (P > 0.05). After controlling for education level, marital status, total annual household income, ethnicity, and sex, the results were similar.

Among the anxiety results, there was a three-way interaction effect between emotional abuse and resilience, BMI, and anxiety (P < 0.05). After controlling for education level, marital status, total annual household income, ethnicity, and sex, the results were similar. Table 7 presents the results. Among the pressure results, there was a three-way interaction effect between emotional abuse and resilience, BMI, and anxiety (P < 0.01). After controlling for education level, marital status, total annual household income, ethnicity and sex, the results were similar. Table 8 presents the results of the study.

Moderation analyses were performed with clustering of ACEs and resilience, BMI, and mental health. The results are presented in Tables 9, 10 and 11. First, ACEs did not correlate with the severity of depression (P > 0.05), resilience was associated with depression, and ACEs× resilience did not significantly correlate the severity of depression (P > 0.05). ACEs×BMI did not significantly correlate depression severity, and resilience×BMI did not correlate with depression. There was no three-way interaction effect between the ACEs, resilience, and BMI (P > 0.05). These results bring into correspondence with previous results after controlling for covariates.

Among the anxiety results, there was a three-way interaction effect between ACEs and resilience, BMI, and anxiety (P < 0.05). After controlling for education level, marital status, total annual household income, ethnicity and sex, the results were similar. Among the pressure results, there was a three-way interaction effect between ACEs and resilience, BMI, and pressure (P > 0.05). After controlling for education level, marital status, total annual household income, ethnicity and sex, the results were similar.

The study examined the relationship between ACEs, BMI, resilience and mental health in a sample of current college students. The prevalence rates of depression, anxiety and stress symptoms were 18.6%, 24.5% and 5.8%, respectively. Second, there is an interaction between ACEs, BMI, and mental health, and between ACEs, resilience and mental health. Third, the study also found a relationship between resilience-moderate ACEs and mental health. The results verify the stress buffering effect model of resilience. The stress buffer effect model showed that higher level resilience could cushion the adverse effects of ACEs on emotional problems. Finally, our results also found a three-way interaction between ACEs, BMI, and resilience. Our results support research hypotheses regarding different factors and underlying mechanisms of student mental health. In addition, multiple layers of trauma-informed early childhood interventions are needed to prevent adverse adolescent outcomes associated with ACEs [30].

Similar study demonstrated that 29% of the students were depressed, 27% were anxious, and 24% were stressed; About 67% of students who were anxious were also depressed and 61% of the anxious students were also stressed [31, 32]. Several factors account for this phenomenon [33, 34]. These include daily life stressors and stressors specific to tedious learning environments. Our study had a sex difference in mental health, similar to another study’s findings [35, 36]. ACEs including physical, sexual, or psychological abuse or neglect of a child or children, especially by a parent or other caregiver [24]. Based on social-ecological risk factor research, some studies point out that teenagers with high ACE scores are at a higher risk of psychological, emotional, and behavioral problems. Further, those with lower social-ecological risk and teenagers with medium or high risk in internal and external behavior differences [37]. No single risk factor plays a decisive role in forming an individual’s healthy development, and the effect of intervention targeting only a single risk is significantly reduced. Specifically, for adolescents, an individual risk factor may not threaten mental health development; only when risk factors are cumulative can they significantly impact personal development [14]. Our results demonstrated that cumulative life events were positively related to poor mental health, per ecological system theory. Since the pioneering studies on ecological risk, cumulative risk has gained support in various social and cognitive domains [38, 39]. Based on these findings, we propose that ACEs are also influenced by social ecosystem theory.

Our study further explored the relationship between BMI, resilience, ACEs and mental health. First, favorable resilience is associated with lower HRBs and mental health problems [40]. This is consistent with our findings that there is also an association between BMI and mental health. The study also further sheds light on the potential mechanistic relationship of ACEs to mental health and identifies how and what aspects of ACEs can be targeted through therapeutic interventions [41]. Second, a previous study found that favorable resilience was associated with lower mental health [42‐44]. Resilience is not only closely related to the mental health of children and adolescents, but also plays a well-regulated role in adults, and should play a more prominent role in real-world research and prevention programs [40, 45]. Our analysis was based on hypotheses derived from the psychopathology between ACEs and mental health [46] and resilience as a potential mediator between adverse childhood experiences and prenatal depression [47, 48]. The impact of ACEs on mental health can be reduced by managing current stressors and improving mental resilience in students [49]. These results provide a theoretical basis for studying the moderating effects of the analyses on the significance of resilience, ACEs, BMI, and mental health. Third, the interaction between high-risk ACEs and obesity correlates with adverse mental health symptoms [50]. It also means that adolescents who are exposed to high-risk social environments, combined with disharmony in life and rest, can lead to physical and mental health [51]. This means that social ecological psychology research is a type of process research that further shows that the specific characteristics of social ecology cause a psychological state. This psychological state can also cause the target’s cognition, emotion, or behavior through some mechanism. These results are similar to those of a previous study [52]. Possible mechanisms were explained “pub hypothesis” [53] and the diathesis-stress model [54] or the comprehensive social-ecological diathesis-stress model [55]. Similar research investigations have been conducted to examine the trajectory of multiple childhood adversity and their relationship to adolescent mental health outcomes, as well as the role of good parenting practices as a buffer [56]. Another study found a moderating effect of resilience and concluded that it may reduce the negative effects of neuroticism and enhance the positive effects of extraversion, agreeableness and conscientiousness on depressive symptoms [40]. This further verifies the association between ACEs, BMI, resilience, and mental health. ACE is a complex etiological marker whose effects appear to vary in terms of the type, timing, and severity of abuse, in addition to a wide range of vulnerability and resilience cofactors [50].

We also explored the moderating effect of resilience on BMI. ACEs are also associated between ACEs and obesity [57, 58]. The possible mechanism is that physiological responses to psychosocial stress (manifested in altered levels of neuroendocrine hormones, the development of malignant stress, and increased allostatic load) are hypothesized to accumulate and interact with each other to further slow metabolic processes. As a result, metabolic related diseases, including cardiovascular, immune and nervous system damage, including the occurrence and development of obesity [58‐60]. We tried to tridimensify the correlation between resilience, BMI, and mental health from a social environment perspective (such as the ACEs environment). Identifying factors that promote resilience may be a target of concurrent and retrospective interventions [61]. Our study found that high life event scores were associated with higher resilience and better mental health. Similarly, adolescents who experienced a worse social environment, together with obesity, through decreased resilience, increased adverse mental health. In summary, obesity can moderate mental health problems caused by ACEs mediated by lower resilience. Several researchers have described recent behavioral and neurobiological resilience studies suggesting that adolescence (a period marked by increased plasticity, development of vital neurobiological circuits, and sensitivity to social environments) may be a particularly appropriate time for ELA intervention [62]. Other researchers have also provided insights into the resilience and vulnerability impact of multiple ACEs, highlighting its research value and clinical implications for further understanding of trauma in young people affected by conflict [63]. There is growing evidence that the fundamental mechanisms by which stressors lead to stable changes in behavior include epigenetic changes, which result from interactions with the genome that lead to changes in DNA structure and gene expression [64].

By integrating the diathesis-stress model with a social-ecological framework, we aim to develop a comprehensive approach to understanding mental health that recognizes the complex and dynamic nature of mental health experiences across various settings, such as individuals, families, communities, and schools, and over time. Social-ecological models consider the interconnectedness of the child’s world. Quality-stress models allow an understanding of the complexity of stressors and risk or protective factors that influence participation and intervention in mental and behavioral health [55]. Potential social factors that influence mental health, and the resulting mental health, may be dampened or overridden by obesity. These results corroborate and extend previous research by adopting multiple social environments to portray the whole society of adolescents, together with their physical conditions (in this study, obesity), to explore the moderating role of mental health in resilience and further verify the moderating role of BMI.

It also provides perspectives to investigate further the association between BMI and mental health, resilience, and mental health, in addition to ACEs [65]. Social-ecological risk factors influence adolescent mental health, which may be moderated by obesity and resilience; therefore, we should advocate ameliorating possible ACEs and provide more support at home, school, and other levels.

The strengths of this study are its multilevel design and the large number of college participants. However, this study has some limitations. As a cross-sectional study, it was difficult to observe a causal association between life events and mental health, and longitudinal studies are needed to evaluate the association between variables in the future. All the data used in this study were the results of self-reported questionnaires; therefore, there were questions of subjectivity, validity, and reliability. This study only investigated the results of 1 city, it is not clear how representative this sample may be, and the follow-up survey will be carried out among samples from different regions and cultures across the country. Although this study focused on the association between ACEs and mental health in different fields, it did not cover other fields. Future research should explore the association between ACEs and mental health at various levels across different fields.