Abstract
Objective:
Children with higher Adverse Childhood Experiences (ACEs) experience more severe parent-reported ADHD. Positive childhood experiences (PCEs) help to build resilience and mitigate the impact of ACEs on ADHD. Prior studies have measured the two constructs as independent factors, but no research has examined their combined influence on children with ADHD. The first aim was to categorize children with different levels of parent-reported ADHD severity into classes based on shared characteristics of ACE and PCE promoters. The second aim was to examine the relationship between the classes and ADHD severity.
Method:
Participants included children 6–17 years with data on the 2019 National Survey of Children’s Health ADHD severity questionnaire (n=19,715; weighted n =49,149,269). Latent Class Analysis (LCA) identified subgroups of children experiencing patterns among PCE promoters and ACEs, which were measured as independent variables in an adjusted ordinal regression model to estimate their composite effects on ADHD severity.
Results:
Utilizing LCA, one class belonging to children with low ACEs and high PCE promoters (class 1) and another belonging to children with high ACEs and low PCE promoters (class 2) were identified. Class 2 was 2.2 times more likely to have more severe ADHD (aOR 2.2; 95% CI: 1.8, 2.6).
Conclusion:
Findings suggest ACEs and PCE promoters don’t operate independently; children with high ACEs had low PCE promoters and had more severe parent-reported ADHD. Clinicians should consider actively screening for the presence of ACEs and PCEs in all children, especially those with high ADHD severity, and build strong alliances with families.
INTRODUCTION
Adverse Childhood Experiences (ACEs) and Health Outcomes
Adverse childhood experiences are stressful events that impact health negatively in a strong, cumulative, and dose-responsive manner.1 The original ACE study linked events such as abuse (sexual, physical, and psychological) and household dysfunction (substance abuse, mental illness, mother treated violently, and criminal behavior in household) to several leading disease risk factors contributing to earlier demise in adulthood.1,2 Adults with 4+ ACEs face higher chances of diabetes, lung diseases, and poor self-rated health. Also, ACEs double the likelihood of smoking, alcoholism, suicide attempts, and drug use.1,2,3
The profound sequelae of ACEs may be explained by the physiological effects of stress.4,5 It first activates hypothalamic-pituitary-adrenocortical axis and sympathetic-adrenomedullary systems, which release stress-hormone-secreting, inflammatory cytokines in proportion to the intensity, duration, and frequency of stress exposure.4,5 Based on the toxic stress model, there are three types of stress responses—positive, tolerable, and toxic which depend on the interaction of the event, level of threat perception, and degree of social support buffering.4,5 Positive stress (e.g. first school day) can foster resilience with adequate support, while tolerable stress (e.g. injury or natural disaster), if minimized with protective factors, allows for full recovery.4 Toxic stress, however, results from chronic, severe stress without adequate support, leaving scars of trauma that are tangible, significant, and permanent.4 Approximately 60% of adults have had at least one ACE by age 18, 2,5,6,7 with 25% experiencing minimally three.7 Studies have consistently linked ACEs to behavioral disorders, including attention-deficit/hyperactivity disorder (ADHD).7–11
Adverse Childhood Experiences (ACEs) and Attention-Deficit/Hyperactivity Disorder (ADHD)
ADHD is a neurodevelopmental disorder characterized by inattention, hyperactivity, and impulsivity, leading to impairments in various aspects of life.12 Individuals with ADHD experience more health problems, psychiatric illness, physical injuries, strained relationships, and academic struggles.13–15 Physiologically, toxic stress disrupts brain development in regions linked to ADHD; this relationship may also be bidirectional.7 Recent studies demonstrated that earlier ACE exposure is associated with a higher likelihood of an ADHD diagnosis by middle childhood,16 and ACEs before age 5 and between 5–9 years are independently associated with parent-reported ADHD diagnosis.14 Furthermore, parents are more likely to report more severe ADHD in children with ACE scores of 2 or more.17
Resilience and Positive Childhood Experiences (PCEs)
Previously seen as solely innate, 18,23 resilience is now understood as a lifelong, complex interplay between intrinsic qualities (i.e., grit) and extrinsic influences (i.e., family support) that help shape, model, and reinforce the child’s capacity to flourish.19,20,21 Therefore, it manifests in a child as a continuously evolving set of skills and abilities acquired throughout lifetime, that is successfully implemented and demonstrated during hardships.4,18 Recognizing that innate qualities are less mutable, focus has shifted to identifying external factors that promote resilience.4 Most notably, safe, stable, nurturing relationships are currently regarded by the American Academy of Pediatrics as “the single most important factor in buffering adversity and building resilience”.4, 10,20,26 Protective factors such as PCEs (stronger family relationships, supportive caregivers, and community ties) may mitigate the negative effects of ACEs on ADHD symptoms.27–29
According to the resilience framework, however, some adversity may help to develop resilience.4,18,22–23 Malhi et al claimed adversity is the “most significant environmental factor that is an essential prerequisite” for developing resilience18,23, and that resilience often represents the “positive end on [the] continuum of adaptation.”23 Many studies have operationalized PCEs and ACEs as independently opposing constructs,27, 28,29 however, PCEs may both co-occur with and operate independently from ACEs.26 No studies to date have investigated patterns characterized by differing levels of ACEs and PCEs, and if contextualizing the child’s life by their hardships and protective factors may help predict parental perception of ADHD severity. The first aim was to categorize children with different levels of parent-reported ADHD severity into classes based on their shared characteristics of ACE and PCEs via Latent Class Analysis (LCA). The second aim was to examine the relationship between the classes and parent-reported ADHD severity.
METHODS
Study Design
This retrospective study utilized data from the 2019 National Survey of Children’s Health (NSCH), the largest national-level survey on the health and well-being of children 0–17 years and families.30 The NSCH is conducted annually by the U.S. Census Bureau and is sponsored by the Maternal and Child Health Bureau (MCHB) of the Health Resources and Services Administration (HRSA).30 The Data Resource Center (DRC) includes more than 300 questionnaire items pertaining to the overall health and healthcare of children, adolescents, and families in a single household, as perceived by the parent/guardian (Child and Adolescent Health Measurement Initiative, 2019).30 The 2019 data were gathered between June 2019 to January of 2020, totaling 29,433 responses.
Participants
We included children ages 6–17 years with complete data on parent-reported ADHD severity. Though the survey item included children of 3–17 years of age, our study deliberately limited the range to best gauge the presence and impact of PCEs, since numerous items were inapplicable for children under 5 years, as they may be too young, for example, to engage in volunteer work or seek out non-family adult mentors. The authors, however, recognize there is no absolute age cut-off for when PCEs become more impactful in building resilience and combatting the effects of adversity on ADHD symptoms.
Measures
Predictor variables.
ACEs:
Per the 2019 NSCH, parents answered questions pertaining to 9 individual ACEs: Hard to cover basics on family’s income, parent/guardian divorced or separated, parent/guardian died, parent/guardian served time in jail, saw or heard parents or adults slap, hit, kick punch one another in the home, was a victim of violence or witnessed violence in neighborhood, lived with anyone who was mentally ill, suicidal, or severely depressed, lived with anyone who had a problem with alcohol or drugs, treated or judged unfairly due to race/ethnicity. All items had dichotomous “yes/no” responses, except for the question, “hard to cover basics, like food or housing, on family’s income” - For this item, responses “somewhat often” and “very often” were coded as “yes”, whereas “never” and “rarely” were coded as “no”.
PCEs:
Cognizant that PCEs may exist on multiple levels of influence in a child’s life24, the authors attempted to best represent PCEs corresponding to the diathesis-stress model, by measuring variables that promote resilience on family, school, and community levels.4 The variables are as follows:
Family resilience:
Measures were based on the responses to the question: “When your family faces problems, how often are you likely to do each of the following? Talk together about what to do, work together to solve our problems, know we have strengths to do draw on, and stay hopeful even in difficult times”. Responses were “all or most of the time to 0–1 items”, “all or most of the time to 2–3 items”, and “all or most of the time to all 4 items”. If the parents reported “all of the time” or “most of the time”, to 2–3 or all 4 items, the child was deemed to have some/a lot of family resilience.
Family shares ideas:
Parents were asked, “How well can you and this child share ideas or talk things that really matter?”, with the response choices “very well”, “somewhat well”, and “not very well or not well at all”. Children whose parents responded “very well” were considered to have this PCE.
Supportive neighborhoods:
To answer the question, “Does this child live in a supportive neighborhood”, caregivers responded to the statements, “People in this neighborhood help each other out”, “We watch out for each other’s children in this neighborhood”, and “When we encounter difficulties, we know where to go for help in our community”, with the choices, “definitely agree”, “somewhat agree”, “somewhat disagree”, and definitely disagree”. Responses to all three statements were included, and children were considered to live in supportive neighborhoods, categorized as “yes” group, if the parents reported “definitely agree” to at least one of the items above and “somewhat agree” or “definitely agree” to the other two items.
Safe neighborhoods:
Parents were asked, “How much do you agree that this child is safe in your neighborhood?”. Response choices were “definitely agree”, “somewhat agree”, “somewhat or definitely disagree”. Responses of “definitely agree” were categorized as the child living in a safe neighborhood.
Safe school:
The survey asked, “How much do you agree that this child is safe at school?”. Parents answered with the choices, “definitely agree”, “somewhat agree”, and “somewhat or definitely disagree”. Responses of “definitely agree” were considered as the child being safe at school.
After school:
Afterschool activities were defined as sports team/lessons and clubs/organizations after school or on weekends, and other organized activities (i.e., music, dance, language). Survey asked, “During the past 12 months, did the child participate in any organized activities or lessons, after school or on weekends?”. Responses were dichotomized as “yes” or “no”.
Community volunteer:
The survey asked, “During the past 12 months, did this child participate in any type of community service or volunteer work at school, church, or in the community?”. Responses were dichotomized as “yes” or “no”.
Outcome Variable.
ADHD severity:
To best gauge parent-reported ADHD severity, parents were asked, “Would you describe this child’s current [ADD/ADHD] as mild, moderate or severe?”. Responses included “does not currently have condition”, “current condition, rated mild”, and “current condition, rated moderate or severe”. The 2019 NSCH codebook states all information gathered is based on parent report. The variable was operationalized as an ordinal variable for the statistical analysis, taking values 0, 1, and 2 corresponding to the three possible responses to the question.
Covariates.
The following demographic variables were considered as potential covariates based on their relationships between ACEs, PCEs and/or ADHD severity. Thus, all models were adjusted for the following: age, sex (male or female), race/ethnicity (Hispanic, non-Hispanic white/black/Asian, and Other/Multi-race), highest level of education of any adult in household (less than high school, high school or GED, some college or technical school, and college degrees or higher), and federal poverty level (0–99% FPL, 100–199% FPL, 200–399% FPL, and 400% FPL or greater).
Statistical Analyses
The NSCH data are collected through a complex, multistage sampling study design that ensures the sample is representative of children across all states in the U.S. The weights are applied to adjust for different probabilities of selection, non-response, and state-specific characteristics (i.e., household size, poverty threshold, education level of the respondent, race, ethnicity, special health care needs status, and national age and sex distributions.30 Prevalence proportions of all variables were adjusted for stratified weighted sampling fractions based on NSCH data guide via package “survey” in R 4.2.2, to account for the complex sample design. Chi-square tests and Kruskal-Wallis rank-sum tests were utilized to examine the bivariate relationships between ADHD severity and each demographic covariate. Analyses of parent-reported ADHD severity versus ACEs and PCEs were done in two steps. In step 1, each ACE and PCE was fitted in an ordinal regression model (proportional odds model) adjusting for demographic covariates to assess their relationship with ordinal outcome, parent-reported ADHD severity.
While the above analyses may elucidate the effect of each 9 ACEs and 7 PCE promoters on parent-perceived ADHD severity, authors wanted to investigate the collective impact of all 16 factors on a child’s wellbeing. In step 2, because certain ACEs and PCE promoters may be intertwined with potential for significant multicollinearity, the authors opted not to measure them concurrently in a single model to determine how they cluster together instead of independently. There’s also no known validated composite measure to group ACEs and PCEs together. Thus a, Latent Class Analysis (LCA) model was utilized to identify meaningful classifications or subgroups of children that share underlying characteristics based on the presence of PCE promoters and individual ACEs. We fit a total of 4 models with one through four latent classes. Model fit was evaluated based on Akaike’s Information Criterion (AIC), Bayesian Information Criterion (BIC), and the sample-size adjusted BIC. The parsimony of a latent class model, as well as interpretability of the latent classes were also considered in model selection. The class variable generated by the final LCA model was then treated as an independent variable in an adjusted ordinal regression model to estimate the effect of these ACEs and PCEs compositely on parent-reported ADHD severity.
Multivariate results are reported as model-based estimates, with corresponding 95% confidence limits and p-value, adjusting for covariates. We defined statistical significance using a two-sided p < 0.05. All statistical analyses were conducted in R Studio 4.2.2.
RESULTS
The final dataset consisted of 19,175 (weighted N = 49,149,269) children and adolescents (Table 1), where 4.2% demonstrated mild ADHD and 5.9% moderate/severe. The children’s median age was 12 (IQR: 9–14), and 51% were male. Additionally, 25.4% of children identified as Hispanic, 50% white (non-Hispanic), 13.8% Black (non-Hispanic), 4.7% Asian (non-Hispanic), and 6% other/multi-race (non-Hispanic). Furthermore, 48.4% of parents reported to have at least a college degree and 18.3% reported to fall between 0–99% household federal poverty line (FPL), 50.6% between 100–399%, and 31.1% on 400% FPL or higher. Overall, the more severe ADHD group of children had higher proportions of males than females (p < 0.001), higher proportion of white (non-Hispanic) children were in mild and moderate/severe ADHD groups when compared to other race/ethnicities (p < 0.001), and higher level of parent education was inversely related to parent-reported ADHD severity (P <0.001).
Table 1.
Caregiver Report of Children’s Demographic Characteristics, Stratified by Reported Current Diagnosis of ADHD
| Characteristic | Overall (N = 49,149,269) |
Without ADHD (N = 44,187,159) |
Mild ADHD (N = 2,060,606) |
Moderate/Severe ADHD (N = 2,901,504) |
P* |
|---|---|---|---|---|---|
| Total Adolescents (N = 49,149,269) |
89.9 | 4.2 | 5.9 | - | |
| Age (years) | 12.0 (9.0, 14.0) | 11.0 (9.0, 14.0) | 13.0 (10.0, 15.0) | 12.0 (9.0, 14.0) | <0.001 |
| Sex | <0.001 | ||||
| Female | 49.0 | 51.2 | 33.2 | 25.3 | |
| Male | 51.0 | 48.8 | 66.8 | 74.7 | |
| Race and Ethnicity | <0.001 | ||||
| Hispanic | 25.4 | 26.3 | 19.8 | 17.2 | |
| White, non-Hispanic | 50.0 | 48.9 | 58.6 | 60.1 | |
| Black, non-Hispanic | 13.8 | 13.7 | 15.7 | 15.1 | |
| Asian, non-Hispanic | 4.7 | 5.1 | 2.0 | 1.2 | |
| Other/Multi-racial, non-Hispanic | 6.0 | 6.1 | 3.9 | 6.4 | |
| Highest Education of Adult in Family | <0.001 | ||||
| < High school or High school/GED | 29.8 | 30.0 | 25.8 | 28.7 | |
| Some college or technical school | 21.8 | 21.1 | 23.9 | 31.3 | |
| College degree or higher | 48.4 | 48.9 | 50.3 | 40.0 | |
| Federal Poverty Level | 0.047 | ||||
| 0–99 FPL | 18.3 | 18.0 | 16.8 | 23.4 | |
| 100–199 FPL | 21.3 | 21.2 | 20.9 | 23.3 | |
| 200–399 FPL | 29.3 | 29.4 | 28.7 | 29.0 | |
| 400 FPL or greater | 31.1 | 31.4 | 33.6 | 24.3 |
All N and proportions were weighted to represent the US population ages 6–17.
Wilcoxon Rank Sum and Chi-square tests
ACEs and ADHD
Exposure to each ACE significantly increased the odds of parents reporting more severe ADHD in their child (Table 2). Children exposed to parental divorce (aOR 1.89 (1.6, 2.3), p <0.001), parental death (aOR 1.87; 95% CI: 1.3, 2.8), parental incarceration (aOR 2.19; 95% CI: 1.7, 2.8), parental mental illness (aOR 2.38; 95% CI: 1.9, 3.0), parental substance use (aOR 1.66; 95% CI: 1.3, 2.1), and financial insecurities (aOR 1.9; 95% CI: 1.5, 2.4) demonstrated higher odds of having more severe ADHD. The exception was racial/ethnic mistreatment (aOR 1.52; 95% CI: 1.1, 2.2), where the percentage was lower for mild ADHD (5.4%) compared to no ADHD (6.2%); however, the percentage of moderate/severe ADHD (9.8%) was significantly higher overall. The odds of having more severe parent-reported ADHD were the highest for children with exposure to domestic violence (aOR 2.64; 95% CI: 2, 3.5) or neighborhood violence (aOR 2.89; 95% CI: 2.1, 4.0), suggesting that child’s behaviors are significantly impacted by phenomena both inside and outside the home.
Table 2.
Associations between Caregiver Report of Children’s ACEs and PCE Promoters and ADHD Severity
| Characteristic | Overall | Without ADHD (N = 44,187,159) |
Mild ADHD (N = 2,060,606) |
Moderate/Severe ADHD (N = 2,901,504) |
P values* | aOR (95% CI)‡ |
|---|---|---|---|---|---|---|
| ACEs | ||||||
| Hard to cover basics on family’s income (ACE1)† | 16.1 | 15.0 | 19.3 | 29.7 | <0.001 | 1.9 (1.5, 2.4) |
| P/G divorced or separated (ACE3)† | 28.7 | 27.1 | 35.6 | 47.8 | <0.001 | 1.89 (1.6, 2.3) |
| P/G died (ACE4)† | 4.0 | 3.7 | 6.9 | 7.2 | <0.001 | 1.87 (1.3, 2.8) |
| P/G served time in jail (ACE5)† | 8.9 | 8.0 | 13.9 | 19.3 | <0.001 | 2.19 (1.7, 2.8) |
| Saw or heard parents or adults slap, hit, kick punch one another in the home (ACE6)† | 6.8 | 5.9 | 11.8 | 16.9 | <0.001 | 2.64 (2, 3.5) |
| Was a victim of violence or witnessed violence in neighborhood (ACE7)† | 5.4 | 4.5 | 11.2 | 13.9 | <0.001 | 2.89 (2.1, 4.0) |
| Lived with anyone who was mentally ill, suicidal, or severely depressed (ACE8)† | 10.0 | 8.9 | 14.1 | 24.3 | <0.001 | 2.38 (1.9, 3.0) |
| Lived with anyone who had a problem with alcohol or drugs (ACE9)† | 10.9 | 10.1 | 13.9 | 19.5 | <0.001 | 1.66 (1.3, 2.1) |
| Treated or judged unfairly due to race/ethnicity (ACE 10)† | 6.4 | 6.2 | 5.4 | 9.8 | 0.021 | 1.52 (1.1, 2.2) |
| PCE promoters | ||||||
| Family resilience (some/a lot)† | 93.1 | 93.4 | 93.6 | 87.8 | <0.001 | 0.6 (0.5, 0.8) |
| Family share ideas (very well)† | 64.9 | 67.5 | 48.7 | 36.5 | <0.001 | 0.35 (0.3, 0.4) |
| Supportive neighborhood (yes)† | 55.0 | 55.8 | 55.9 | 42.6 | <0.001 | 0.69 (0.6, 0.8) |
| Safe neighborhood (definitely agree)† | 64.0 | 64.3 | 65.5 | 57.2 | 0.024 | 0.79 (0.7, 0.9) |
| Safe school (definitely agree)† | 70.1 | 70.7 | 71.0 | 61.1 | <0.001 | 0.8 (0.7, 0.9) |
| After school activity (yes)† | 79.8 | 80.8 | 74.4 | 68.2 | <0.001 | 0.55 (0.4, 0.7) |
| Community volunteer (yes)† | 43.4 | 43.9 | 42.7 | 36.7 | 0.019 | 0.79 (0.7, 0.9) |
All N and proportions were weighted to represent the US population ages 6–17.
Chi-square tests
Missing data resulting in lower N (~1–4).
Adjusted logistic regression: OR = Odds Ratio, CI = Confidence Interval. Models adjusted for child’s age, sex, race, adult’s education, and family income (FPL).
PCE Promoters and ADHD
Our findings showed an inverse relationship between PCE promoters and parent-reported ADHD severity (Table 2). Children with strong family resilience (aOR 0.6; 95% CI: 0.5, 0.8) where sharing ideas (aOR 0.35; 95% CI: 0.3, 0.4) is the norm showed significantly lower odds of having more severe parent-reported ADHD. Similarly, children living in safe (aOR 0.79; 95% CI: 0.7, 0.9) and supportive (aOR 0.69; 95% CI: 0.6, 0.8) neighborhoods demonstrated lower odds of parent-reported ADHD symptoms. Children attending safe schools (aOR 0.8; 95% CI: 0.7, 0.9) with commitments to extracurricular activities (aOR 0.55; 95% CI: 0.4, 0.7) and community volunteer work (aOR 0.79 (0.7, 0.9) were reported to have lower odds of having more severe parent-reported ADHD symptoms.
Latent Class Analysis (LCA)
Recognizing the potential multicollinearity between ACE and PCE variables, LCA was conducted to identify groups based on shared characteristics, with the goal of measuring the relationships with parent-reported ADHD severity (Table 3). During the model building process, our priorities were as follows: 1) examine the parsimony, interpretability, and utility of the latent classes, and 2) determine the balance between fit and parsimony using information criteria. Table 4 presents LCA diagnostic criteria for different class models. We recognize that the AIC, BIC, and aBIC decrease as the number latent classes increase, with the 4-class model having the lowest values. However, the difference between the 3-class model and 2-class model was relatively of a small magnitude (percent change of 2.6), which was similar for the 4-class model and 2-class model (percent change of 3.1 – 3.2). From a conceptual perspective, groups yielded by the 3-class and 4-class models were smaller and tended to overlap in characteristics. These complications can hinder our understanding of who each class is fully representing and reduce our ability to discern distinct patterns of risk factors and their correlations to the outcomes. Per Collins et al12, it is often reasonable to simplify the model, by reducing the number of classes/parameters, to achieve balance between fit and parsimony, without sacrificing conceptual appeal. The distinct groups presented in the 2-class model offer much easier interpretability, clearer communication of findings, and insights that are actionable for healthcare providers. Therefore, we decided to present the 2-class model results.
Table 3.
Caregiver Report of Children’s ACEs, PCE Promoters, and ADHD Severity, Stratified by Classes/Subgroups yielded by Latent Class Analysis (Complete Case, weighted N = 44,055,708)
| Characteristic | Class 1 – Low ACEs and High PCEs (weighted N = 27,254,8391) |
Class 2 – High ACEs and Low PCEs (weighted N = 16,800,869) | P values |
|---|---|---|---|
| ACEs | |||
| Hard to cover basics on family’s income (ACE1) | 5.1 | 33.8 | <0.001 |
| P/G divorced or separated (ACE3) | 14.9 | 50.4 | <0.001 |
| P/G died (ACE4) | 2.3 | 6.8 | <0.001 |
| P/G served time in jail (ACE5) | 0.8 | 21.5 | <0.001 |
| Saw or heard parents or adults slap, hit, kick punch one another in the home (ACE6) | 0.3 | 17.4 | <0.001 |
| Was a victim of violence or witnessed violence in neighborhood (ACE7) | 0.5 | 13.5 | <0.001 |
| Lived with anyone who was mentally ill, suicidal, or severely depressed (ACE8) | 2.6 | 22.3 | <0.001 |
| Lived with anyone who had a problem with alcohol or drugs (ACE9) | 1.3 | 26.4 | <0.001 |
| Treated or judged unfairly due to race/ethnicity (ACE 10) | 2.5 | 12.8 | <0.001 |
| PCE promoters | |||
| Family resilience (some/a lot) | 97.7 | 85.8 | <0.001 |
| Family share ideas (very well) | 74.3 | 49.5 | <0.001 |
| Supportive neighborhood | 74.4 | 23.6 | <0.001 |
| Safe neighborhood (definitely agree) | 83.4 | 32.0 | <0.001 |
| Safe school (definitely agree) | 87.0 | 42.1 | <0.001 |
| After school activity | 87.6 | 68.7 | <0.001 |
| Community volunteer | 50.6 | 31.8 | <0.001 |
| ADHD severity | <0.001 | ||
| No ADHD | 92.4 | 85.5 | |
| Mild ADHD | 3.9 | 4.8 | |
| Moderate/Severe ADHD | 3.7 | 9.7 |
Table 4.
LCA Diagnostic Criteria for Different Class Models
| Number of Latent Classes | Number of Parameters Estimated | Predicted Class Membership (%) | Class Population Shares (%) | AIC | BIC | aBIC |
|---|---|---|---|---|---|---|
| 1 | 20 | 100 | 100 | 279734.9 | 279892.1 | 279701.7 |
| 2 | 41 | 67, 33 | 66, 34 | 263277.5 | 263599.8 | 263202.1 |
| 3 | 62 | 59, 29, 12 | 58, 29, 13 | 256366.2 | 256853.6 | 256248.9 |
| 4 | 83 | 55, 30, 9, 6 | 54, 30, 10, 6 | 254783.9 | 255436.4 | 254624.5 |
Abbreviations: AIC Akaike Information Criterion, BIC Bayesian Information Criterion, aBIC sample-size adjusted BIC.
Thus, we found that one class was composed of children with low ACES and high PCE promoters (class 1), and another composed of children with high ACEs and low PCE promoters (class 2). Overall, children in class 1 live under conditions demonstrating higher family resilience (98% vs 86%), higher likelihood that families share ideas (74% vs 50%), more supportive (74% vs 24%) and safer neighborhoods (83% vs 32%), safer school environments (87% vs 42%), and a higher likelihood of community and after school involvement. Furthermore, they experience less racism (2% vs 13%), parental separation (15% vs 50%), parental death (2% vs 7%), and this pattern is consistent for all nine ACEs. Conversely, children in class 2 live under conditions demonstrating high prevalence of each ACE (i.e., domestic violence (17% vs 0.3%)) and low prevalence of each PCE (i.e., community volunteer (32% vs 51%)). All p-values were < 0.001. In the ordinal regression model adjusted for demographics, children in class 2 have 2.2 times the odds of having parents rate their ADHD symptoms as more severe (aOR 2.2; 95% CI: 1.8, 2.6) compared to children in class 1.
DISCUSSION
We discovered via Latent Class Analysis that one group of children lived in conditions surrounded by low ACEs and high PCE promoters (class 1), and another group belonged in an environment demonstrating high ACEs and low PCE promoters (class 2). Furthermore, children in class 2 were 2.2 times more likely to have more severe parent-reported ADHD, when compared to class 1.
Our study introduces a novel process to frame and operationalize positive childhood experience and adverse childhood experiences. Rather than pitting PCEs and ACEs against each other as independent, opposing factors,27–29 our study provides a framework to investigate the simultaneous patterns in which PCE promoters and ACEs may exist in a child’s life, based on the framework that adversity may help to demonstrate and build resilience.18, 23 Since our findings suggest children experiencing more adverse experiences have less PCE promoters, clinicians should strive to build strong alliances with families to promote safe, stable nurturing relationships, and build child and family strengths to help mitigate the influence of ACEs.20,26 As such, clinicians have the unique opportunity to promote targeted PCEs that are feasible and meaningful.
Our study also suggests that contextualizing the child’s living conditions wholistically, with the negative experiences and the positive supports buffering them inside and outside of the home, may predict how the parents would perceive their ADHD severity. Hence, clinicians should consider universally screening for ACEs and PCEs to mitigate the downstream effects of ADHD severity. Thus, clinicians have the unique opportunity to practice relational health by promoting PCEs to build resilience in the effort to combat the negative effects of ACEs.20,26
One strength of the study is that the NSCH utilizes a weighted sample size to generalize the results to the children across the U.S..30 Another strength was that ADHD severity was categorized on a 3-point in NSCH, instead of a 2-point scale, which improved statistical power. Dichotomizing the severity variable into presence/absence of ADHD, might have oversimplified the range of severity/level experienced by individuals diagnosed with ADHD underestimating the results. Finally, we introduced a novel way of simultaneously defining, measuring, and categorizing patterns among promoters of PCEs in relation to ACEs via Latent Class Analysis.
There were also several limitations to our study. The current study was cross-sectional, which limited our ability to draw causal relationships or longitudinal conclusions amongst the variables, or how the two classes may have evolved over time. Due to the retrospective nature of the study, report of ACEs, PCE promoters and ADHD severity may have been affected by recall bias. Variables were derived from the NSCH, so it was not possible to introduce new variables of interest. Also, categorization of the two ACE and PCE classes prioritized parsimony and easy interpretation, however, it limited our ability to evaluate nuances and more complex relationships among additional classes. The cut-points used for collapsing categories of PCE promoters were developed through a combination of literature review and expert consensus in the field to ensure that we captured a range of experiences without over- or underestimating and that the cut-points align with real-world applications. They were chosen to best capture the most relevant and impactful indicators within the scope of our study. Future studies may want to consider more than two classes to evaluate nuances and more complex relationships among variables. Finally, relying on parent report makes it difficult to ascertain the child’s true diagnoses and their subjective experiences.
Conclusion
Our findings suggest ACEs and PCE promoters do not operate independently; children with high ACEs have low PCE promoters were more likely to have more severe parent-reported ADHD.
Clinicians have the unique opportunity to build strong alliances with family to better understand the nature of ACEs to promote relevant PCEs to combat negative effects of ACEs to help build resilience. Clinicians should consider actively screening for the presence of ACEs and PCEs in all children, especially in those with parental or provider concern for ADHD symptoms. Identification of PCE promotion opportunities could help to build strong alliances with families especially amongst children with high ADHD severity.
Acknowledgments
This work was supported by the Biostatistics and Data Management Core at The Saban Research Institute, Children’s Hospital Los Angeles and by grant UL1TR001855 from the National Center for Advancing Translational Science (NCATS) of the U.S. National Institutes of Health. This research was also supported by the Maternal and Child Health Bureau, Children’s Hospital Los Angeles Developmental-Behavioral Pediatrics Training Program grant T77MC25732. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the Maternal and Child Health Bureau.
Footnotes
Disclosure statement: I certify that no party has a conflict of interest. I certify that all financial and material support for this research and work are clearly identified in the title page of the manuscript.
REFERENCES
- 1.Jacob G, Van Den Heuvel M, Jama N, et al. Adverse childhood experiences: Basics for the paediatrician. Paediatr Child Health. 2019;24(1):30–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Felitti VJ, Anda RF, Nordenberg D, et al. Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults. The Adverse Childhood Experiences (ACE) Study. Am J Prev Med 1998;14(4):245–258. [DOI] [PubMed] [Google Scholar]
- 3.Holman DM, Ports KA, Buchanan ND, et al. The association between adverse childhood experiences and risk of cancer in adulthood: A systematic review of the literature. Pediatrics 2016;138:81–91. [Google Scholar]
- 4.Feldman HM, Elias ER, Blum NJ et al. Developmental-Behavioral Pediatrics. In: Murray M, Keder R. Trauma, Resilience, and Child Development. Section 3, Ch 15. 5th ed. Philadelphia, PA: Elsevier; 2023: 153–162. [Google Scholar]
- 5.Shonkoff JP, Garner AS; Committee on Psychosocial Aspects of Child and Family Health; Committee on Early Childhood, Adoption, and Dependent Care; Section on Developmental and Behavioral Pediatrics. The lifelong effects of early childhood adversity and toxic stress. Pediatrics. 2012;129(1):e232–e246. doi: 10.1542/peds.2011-2663 [DOI] [PubMed] [Google Scholar]
- 6.Merrick MT, Ford DC, Ports KA, et al. Prevalence of adverse childhood experiences from the 2011–2014 behavioral risk factor surveillance system in 23 states. JAMA Pediatrics. 2018;172, 2011–2014. [Google Scholar]
- 7.Lugo-Candelas C, Corbeil T, Wall M, et al. ADHD and risk for subsequent adverse childhood experiences: understanding the cycle of adversity. J Child Psychol Psychiatry. 2021;62(8): 971–978 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Hughes K, Bellis MA, Hardcastle KA, et al. The effect of multiple adverse childhood experiences on health: A systematic review and meta-analysis. Lancet Public Health. 2017; 2(8): e356–e366. [DOI] [PubMed] [Google Scholar]
- 9.Fuller-Thomson E, Mehta R, Valeo A, et al. Establishing a link between attention deficit disorder/attention deficit hyperactivity disorder and childhood physical abuse. Journal of Aggression, Maltreatment & Trauma. 2014; 188–198. [Google Scholar]
- 10.Flaherty EG, Thompson R, Litrownik AJ, et al. Effect of Early Childhood Adversity on Child Health. Arch Pediatr Adolesc Med 2006; 160 (12): 1232–1238. [DOI] [PubMed] [Google Scholar]
- 11.Bethell C, Gombojav N, Solloway M, et al. Adverse Childhood Experiences, Resilience and Mindfulness-Based Approaches: Common Denominator Issues for Children with Emotional, Mental, or Behavioral Problems. Child Adolesc Psychiatr Clin N Am 2016; 25 (2): 139–156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Froehlich T, Schumacher JB, Attention-Deficit/Hyperactivity Disorder. In: Augustyn M, Zuckerman B. Zuckerman, eds. Parker Handbook of Developmental and Behavioral Pediatrics for Primary Care, 4th edition. Wolters Kluwer; 2019: 126–139. [Google Scholar]
- 13.Collins LM and Lanza ST (2010) Latent Class and Latent Transition Analysis: With Applications in the Social, Behavioral and Health Sciences. Wiley, New York. [Google Scholar]
- 14.Jimenez ME, Wade R Jr, Schwartz-Soicher O, et al. Adverse Childhood Experiences and ADHD Diagnosis at Age 9 Years in a National Urban Sample. Acad Pediatr 2017;17(4):356–361. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Walker CS, Walker BH, Brown DC, et al. Defining the role of exposure to ACEs in ADHD: Examination in a national sample of US children. Child Abuse Negl 2021;112:104884. [DOI] [PubMed] [Google Scholar]
- 16.Hunt TKA, Slack KS, Berger LM. Adverse childhood experiences and behavioral problems in middle childhood. Child Abuse Negl 2017;67:391–402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Brown NM, Brown SN, Briggs RD, et al. Associations Between Adverse Childhood Experiences and ADHD Diagnosis and Severity. Acad Pediatr 2017;4: 349–355. [Google Scholar]
- 18.Luthar SS, Cicchetti D, Becker B. The construct of resilience: a critical evaluation and guidelines for future work. Child Dev 2000;71(3):543–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Masten A, Garmezy N. Risk, vulnerability, and protective factors in developmental psychopathology. In: Lahey B, Kazdin A, editors. Advances in clinical child psychology. Vol. 8. New York, NY. Plenum Press; 1985: 1–52. [Google Scholar]
- 20.Garner A, Yogman M. Preventing childhood toxic stress: partnering with families and communities to promote relational health. Pediatrics. 2021;148(2). [Google Scholar]
- 21.National Scientific Council on the Developing Child. Supportive Relationships and Active Skill-Building Strengthen the Foundations of Resilience: Working Paper No. 13 Cambridge, MA: National Scientific Council on the Developing Child; 2015. https://developingchild.harvard.edu/resources/supportive-relationships-and-active-skill-building-strengthen-the-foundations-of-resilience. Accessed April 28, 2025. [Google Scholar]
- 22.Schneider M, VanOrmer J, Zlomke K. Adverse Childhood Experiences and Family Resilience Among Children with Autism Spectrum Disorder and Attention-Deficit/Hyperactivity Disorder. J Dev Behav Pediatr 2019;40(8): 573–580. [DOI] [PubMed] [Google Scholar]
- 23.Malhi GS, Das P, Bell E, et al. Modelling resilience in adolescence and adversity: A novel framework to inform research and practice. Translational Psychiatry. 2019;9(1): 1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Cowen EL, Wyman PA, Work WC, et al. Follow-up study of young stress-affected & stress-resilient urban children. Development and Psychopathology. 1997;9: 564–577. [Google Scholar]
- 25.Julian MM, Lumeng JC. The Biopsychosocial Model: Understanding Multiple, Interactive Influences on Child Development and Behavior. In: Feldman HM, Elias ER, Blum NJ et al. , eds. Developmental-Behavioral Pediatrics, 5th edition. Elsevier; 2023:11–18. [Google Scholar]
- 26.Bethell C, Jones J, Gombojav N, Linkenbach J, Sege R. Positive Childhood Experiences and Adult Mental and Relational Health in a Statewide Sample: Associations Across Adverse Childhood Experiences Levels. JAMA Pediatr 2019. Nov 1;173(11):1110. [Google Scholar]
- 27.Okwori G Role of Individual, Family, and Community Resilience in Moderating Effects of Adverse Childhood Experiences on Mental Health Among Children. J Dev Behav Pediatr 2022. Sep 1;43(7): e452–e462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Uddin J, Alharbi N, Uddin H, et al. Parenting stress and family resilience affect the association of adverse childhood experiences with children’s mental health and attention-deficit/hyperactivity disorder. J Affect Disord 2020;272; 104–109. [DOI] [PubMed] [Google Scholar]
- 29.Keane K, Evans RR. Exploring the relationship between modifiable protective factors and mental health issues among children experiencing Adverse Childhood Experiences Using a Resilience Framework. J Child Adolesc Trauma. 2022;14;15(4): 987–998. [Google Scholar]
- 30.Child and Adolescent Health Measurement Initiative (CAHMI) 2021. 2019, National Survey of Children’s Health: Child and Family Health Measures and Subgroups, SPSS Codebook, Version 1.0, Data Resource Center for Child and Adolescent Health supported by Cooperative Agreement from the U.S. Department of Health and Human Services, Health Resources and Services Administration (HRSA), Maternal and Child Health Bureau (MCHB). www.childhealthdata.org. Accessed May 3, 2024. [Google Scholar]