Adverse childhood experiences (ACEs), defined as preventable, potentially traumatic events that occur in childhood (0–17 years) (CDC, 2019), are associated with a wide range of lifelong adverse behavioral, health, and social outcomes. These experiences can have significant impacts on adolescents’ mental and physical health (Grant et al., 2003; Shonkoff, 2012). In the short-term, ACEs such as childhood physical abuse, assault, and sexual abuse, can cause acute stress reactions, post-traumatic stress disorder, and increase risk for suicidality (Ali et al., 2024; Low et al., 2012; Serafini et al., 2015; Thompson et al., 2023). Longer term, ACEs are associated with a host of negative physical, mental, and behavioral health outcomes, such as cardiovascular disease, cancer, and depression (Bellis et al., 2019; Daníelsdóttir et al., 2024; Felitti et al., 1998; March-Llanes et al., 2017; Petruccelli et al., 2019; Serafini et al., 2015).

Assessing ACEs among children and adolescents often relies—at least in part—on parent report (Loveday et al., 2022; Popp et al., 2020; Szilagyi et al., 2016). Parent report is a common and generally accepted approach to collecting data on children’s health and experiences, including ACEs. Examples of national parent-report surveys on health and ACEs in the United States include the National Survey of Children’s Health (NSCH) (National Survey of Children’s Health (NSCH) data query, 2022) and National Health Interview Survey (NHIS) (National Health Interview Survey, 2022 survey description, 2023).

Extant research suggests poor agreement between caregivers’ and children’s reports of ACEs. Most studies of concordance on violence-related ACEs suggest poor agreement between caregivers’ reports and children’s self-reports, with caregivers consistently underestimating the violence exposure reported by their child (Richters & Martinez, 1993; Thomson et al., 2002). One study concluded that for children of all ages, caregivers underestimated their child’s violence victimization 58 % of the time; and for children 12 years or older, caregivers underestimated this 75 % of the time (Howard et al., 1999). A 2015 study of children, mothers, and fathers completing the Life Incidence of Traumatic Events Questionnaire found that agreement was low across many potentially traumatic events, including non-interpersonal events (Tingskull et al., 2015). A recent web panel study using Cohen’s Kappa estimates to examine parent-child concordance on children’s ACEs found moderate to substantial agreement between reporters for household challenge-related ACEs (e.g., parental separation/divorce, household substance use), but fair or less agreement between reporters for ACEs related to violence and neglect (e.g., emotional abuse, contact sexual violence, physical neglect) (Licitis et al., 2024).

Lack of concordance between parents and children on reported ACEs may be, in part, related to children experiencing events outside of the supervision of their caregivers (Thomson et al., 2002). Additional potential contributors include poor relationship or communication between parent and child; fear of disclosure; fear of child welfare involvement; unwillingness to self- or proxy-report on certain behaviors or outcomes; stigma; and informants’ unique perspectives, all of which may apply differently to parents versus children (De Los Reyes & Kazdin, 2005; Dirks et al., 2012; Knight et al., 2000).

Little is known about how sociodemographic factors such as race and ethnicity, household income, and parental education influence parent-child concordance on ACEs. A 2022 familial violence study found several sociodemographic factors associated with discordant reports of violence (compared to concordant reports of violence), including Hispanic ethnicity, married parents, and higher household income (Hogan et al., 2023). Licitis et al. (2024) found parent-child pairs with lesbian, gay, bisexual, or another sexual identity; non-Hispanic Black; Hispanic; or non-Hispanic multiracial or another race teens were more likely to be discordant in reporting ACEs compared to parent-child pairs with heterosexual or non-Hispanic White teens. They also found some differences in concordance by household income for some individual ACEs (e.g., witnessing community violence, household substance misuse, emotional neglect).

Quantifying the extent the prevalence of childhood ACEs differs based on type of reporter (parent versus child) and type of ACE is critical for understanding the strengths and limitations of each reporter for research and public health surveillance efforts. Additional studies are also needed to understand sociodemographic factors that may be associated with concordance on reporting of ACEs. The major objectives of the present study are to examine parent-teen concordance of ACEs in a national sample of teens, defined as the level of agreement between parent and teen responses, and to explore sociodemographic predictors of concordance.

Table 1 presents the weighted percent distribution of parent-reported sociodemographic characteristics of the study sample. Overall, about half of the teens were male (51.4 %), aged 12–14 years (51.4 %), non-Hispanic White (46.1 %) and had parents with a college degree or more (52.6 %). A large majority of teens lived in two-parent families (82.7 %).

Overall, 27.7 % of parents reported their child had at least one ACE compared to 43.8 % of teens who reported at least one ACE (Fig. 1; Table 2). A lower percentage of parents reported each ACE examined, when compared to their teens (victim of/witnessed violence in the neighborhood: 6.5 % vs. 15.5 %; parental incarceration or detention: 6.2 % vs. 9.2 %; household mental illness: 11.3 % vs. 21.1 %; household substance use: 9.8 % vs. 16.7 %; experienced emotional abuse: 4.6 % vs. 16.1 %; experienced unmet basic needs: 4.1 % vs. 7.1 %; experienced racism: 6.9 % vs. 18.5 %). Z-tests indicated these differences were significant for every ACE except parental incarceration or detention and experienced unmet basic needs.

Fig. 2 presents percent agreement and disagreement between parents and teens for ACEs. In general, percent agreement between parents and their teens was high, percentages ranging from 83.6 % to 94.2 %. Where disagreement existed, it was more common that parents did not report an ACE when the teen reported it (e.g. household mental illness: parent no/teen yes, 12.7 %).

Weighted Cohen’s kappa and PABAK statistics are presented in Table 2. Cohen’s kappa showed fair agreement between parent-teen pairs for any ACE (k = 0.39), experienced racism (k = 0.26), being a victim of or witnessing violence in the neighborhood (k = 0.27), experienced emotional abuse (k = 0.28), and unmet basic needs (k = 0.36). Moderate agreement was shown for household mental illness (k = 0.41), parental incarceration or detention (k = 0.53) and household substance use (k = 0.50).

PABAK values showed higher agreement compared to Cohen’s kappa, as anticipated given the low prevalence of several ACEs. Agreement for any ACE was moderate (k = 0.41) and substantial for household mental illness (k = 0.67) and experienced racism (k = 0.69). Agreement was almost perfect (k’s: 0.60–0.80) for all other ACEs (k’s ranging from 0.73 to 0.88).

Prevalence and bias indices are also presented in Table 2 for each ACE. In general, the prevalence index showed that the probability of concordance of a parent and teen both reporting “no” was higher than that of both the parent and teen both reporting “yes”. Bias ranged from 0.02 for experienced unmet basic needs and parental incarceration or detention, to 0.17 for any ACE.

In this large, nationally representative survey of teens, teens were more likely to report ACEs than their parents. Nearly 44 % of adolescents reported any ACE versus 28 % of parents reporting on their child’s behalf. These findings are consistent with previous studies, which have found discrepancies between parent and child reports of ACEs, with parents consistently less likely to report ACEs compared to children (Schreier et al., 2005; Stover et al., 2010), particularly for violence- and neglect-related ACEs (Licitis et al., 2024; Tang et al., 2024). Despite this difference in reporting, nationally representative self-reported data from adolescents are limited for most ACEs (Anderson et al., 2022). Parent reported prevalence of individual ACEs in this NHIS-Teen sample was comparable to the subset of parallel items that appear on the NSCH among those aged 12–17 years (e.g., prevalence household substance use was 9.8 % on NHIS, and 11.9 % on NSCH).

The largest relative differences between teen- and parent-reported ACEs were for experiencing any ACE, experiencing emotional abuse, experiencing racism, and household mental illness. The greater level of discordance for these ACEs may reflect parental unawareness for things that may occur outside the home (e.g. racism), or a difference in perception between parents and teens on what may constitutes emotional abuse or mental illness.

Concordance between adolescents and parents ranged from fair–moderate by weighted Cohen’s Kappa statistics to substantial–almost perfect when using PABAK. Given the strengths and limitations of different concordance measures, as discussed previously, our study uses both weighted Cohen’s Kappa and PABAK to provide a more comprehensive view of concordance. For all ACEs, PABAK values were higher than Cohen’s Kappa statistics, suggesting that in part, a portion of discordance observed was due to low prevalence of certain ACEs. These findings are consistent with prior studies (Tang et al., 2024; Licitis et al., 2024) who found differences between parent-reported and child-reported victimization when using Cohen’s Kappa, but strong agreement when using PABAK. However, given that some PABAK values still demonstrated lower levels of agreement (e.g. Any ACE, PABAK = 0.41; household mental illness, PABAK = 0.67), these results also suggest that not all discordance observed was a result of low ACE prevalence (Feinstein et al., 1990; Tingskull et al., 2015).

Our study presents new insight into the likelihood of concordance between adolescents and their parents based on sociodemographic characteristics and the type of ACE. While not significant for all ACEs, concordance tended to be higher among parent-teen pairs when the teen was male, aged 12–14 years, non-Hispanic White, and where the parents’ highest level of education was a college degree or higher. This echoes findings from Tang et al. (2024), which found that several community-level and demographic variables were associated with lower odds of concordance in parent-report and child-report of certain social victimization experiences, including parental perception of neighborhood safety, Black/African American race, or Hispanic/Latinx ethnicity. Concordance in reporting experiencing racism was less likely between parents and teens belonging to all other race and Hispanic origin groups compared to non-Hispanic White teens. The reasons for this finding are likely complex and multifactorial, but may include cultural norms (Elliott & Urquiza, 2006) and parents’ own experiences with racial discrimination (Holloway & Varner, 2021).

This study is among the first to provide comparisons of parent- and teen-reported information using nationally representative data, which few, if any, prior studies have had the ability to do. Our study adds to the literature highlighting the importance of assessing ACEs via self-report when developmentally appropriate (as with teens) (Anderson et al., 2022; Tingskull et al., 2015). When direct reports from children are not possible, our study helps clarify interpretation of parent-reported findings on ACEs. In addition, because NHIS-Teen collects a comprehensive range of ACEs, some of which are not measured by other nationally representative teen-reported surveys (e.g. racial discrimination), these data are helpful to quantify associations between ACEs and health outcomes.

ACEs are prevalent among teens, and teen self-report of experiences are different from parent-report. These findings provide important context for interpreting ACEs data collected from parents. The prevalence of discordance between parent and teen for ACEs, particularly where teens report an ACE and parents did not, highlights the importance of collecting data directly from teens, and may help inform interventions focused on individuals with ACEs.

Supplementary data to this article can be found online at https://doi.org/10.1016/j.chiabu.2025.107339.