Adult Mental Health, Substance Use Disorders, and Functional Outcomes of Children Resilient to Early Adversity

Of the 1,266 participants followed into adulthood, only 25.6% (N=325) had not met criteria for either a full psychiatric diagnosis or a subthreshold disorder (i.e., some symptoms but no psychiatric diagnosis) by age 16. Participants without childhood psychiatric problems had lower levels of all types of childhood adversities (Table 1). Nevertheless, many individuals in this group had experienced at least one form of adversity during childhood.

Cumulative childhood exposure to adversity was estimated by counting the number of categories of adversity experienced (of the five that were assessed). Figure 1 shows the relative proportions of participants with and without psychiatric problems, stratified by the number of exposures to childhood adversity. With each additional domain of childhood adversity experienced (i.e., treating number of adversities as a continuous variable), the risk for having a childhood disorder increased significantly (risk ratio=1.7, 95% CI=1.5–1.9, p<0.001). Risk was highest among individuals who experienced multiple forms of childhood adversity: only 12.2% (N=63 of 650) of individuals experiencing adversity in two or more domains were free of childhood psychiatric problems, compared with 52.4% (N=262 of 616) of participants experiencing adversity in one or fewer domains (p<0.001). Thus, most children without a psychiatric problem had exposure to fewer types of childhood adversity. The approximately 12% of children who did not have mental health problems despite a level of childhood adversity that is commonly associated with psychiatric problems would be considered resilient based on much of the literature (i.e., a high level of risk coupled with no psychiatric problems). Children in this resilient group did not significantly differ by sex (53.1% male, 46.9% female; p=0.74). Subsequent analyses compared this resilient group with two other groups: children without psychiatric problems and a low level of adversity exposure (low risk/no disorder) and children with psychiatric problems (childhood disorder). A comparison of these three groups in terms of childhood adversities is provided in Table S3 in the online supplement.

The next question was how these groups, defined by their childhood adversity coupled with their childhood psychiatric history, fared in adulthood. First, we looked at adult psychiatric status at ages 25 and 30 (Table 2 and Figure 2). Compared with the low-risk/no-disorder group, the resilient group was at greater risk of having an anxiety disorder (risk ratio=2.9, 95% CI=1.0–9.1, p=0.04) or a depressive disorder (risk ratio=4.5, 95% CI=1.1–16.7, p=0.03) in adulthood. There were no differences in risk of diagnosis of adult psychiatric disorders between resilient individuals and those with a childhood disorder. Findings differed for substance use disorders. The resilient group was not at increased risk for adult substance use disorders compared with the low-risk/no-disorder group (p values ranged from 0.96 to 0.34). The resilient group generally had lower levels of substance use diagnoses than the childhood-disorder group, although these differences did not meet conventional statistical thresholds. Thus, compared with the low-risk/no disorder group, the resilient group was at increased risk for adult emotional disorders but not substance use disorders.

We also compared the groups in terms of four domains of adult functioning: physical health, criminality, financial or educational, and social functioning (Table 2). Scale scores were standardized to facilitate comparisons, with higher scores indicative of better function (Figure 3). The low-risk/no-disorder group had the highest level of adult functioning, with the exception of criminal behavior. Compared with the low-risk/no-disorder group, the resilient group had worse physical health (means ratio=0.7, 95% CI=0.5–0.9, p=0.008) and financial or educational functioning (means ratio=0.6, 95% CI=0.5–0.7, p<0.001); however, the groups were similar in the domains of criminality and social functioning. The resilient group had better functioning compared with the group of participants with childhood psychiatric problems in the domains of health (means ratio=1.4, 95% CI=1.2–1.8, p=0.001), criminality (means ratio=2.5, 95% CI=1.8–3.5, p<0.001), and social functioning (means ratio=1.5, 95% CI=1.2–1.9, p<0.001), but not financial or educational functioning. Thus, the well-being of adults who had been characterized as resilient children was better than that of the adults with childhood psychiatric problems but sometimes worse than that of adults exposed to less childhood adversity.

To test whether the results were robust to alternative group definitions, models were retested with a resilient group that was defined based solely on exposure to maltreatment, rather than a broader range of adversity variables (e.g., low socioeconomic status and family dysfunction). Thus, individuals in the resilient group had experienced childhood maltreatment (i.e., physical abuse, sexual abuse, and neglect), but they had not met criteria for either a full psychiatric diagnosis or a subthreshold disorder. Results from this reanalysis are provided in Table S4 in the online supplement. All statistically significant associations in primary analyses were also observed in this sensitivity analysis.

Finally, one possible explanation for the observed pattern of results is that children in the resilient group continued to experience adversity at a high level in adulthood. To test this possibility, we compared adult exposure to traumatic events across the three groups. Exposure to trauma in adulthood was reported by 23.3% of participants in the resilient group (N=9 of 63), compared with 2.9% of those in the low-risk/no-disorder group (N=14 of 262) and 8.3% in the childhood-disorder group (N=81 of 941). The level of adult trauma exposure in the resilient group was higher than that in both the low-risk/no-disorder group and the childhood-disorder group (p values, 0.002 and 0.03, respectively). Inclusion of adult trauma in the primary analyses partially attenuated the association between childhood resilience and adult anxiety and depressive disorders.

Although psychiatric problems are a common experience in childhood (6, 7, 10), about one in four children in the Great Smoky Mountains Study failed to experience any such problems, even at a subthreshold level. The vast majority of children without psychiatric problems had limited exposure to the types of early childhood adversity that typically evoke such problems. Only about one in eight of the children exposed to multiple forms of adversity did not display subthreshold or fully diagnostic psychiatric problems in childhood, thus meeting common definitions of resilience. In other words, when comprehensively measuring exposure to adversity and psychological problems over longer periods of time within childhood, resilience to childhood adversity is uncommon.

But how does this small group fare in terms of their well-being in the long run? Children characterized as resilient had a greater risk of depression and anxiety in adulthood, but not substance use disorders, compared with children exposed to fewer forms of adversity and remained free of childhood psychiatric disorders. In terms of their everyday functioning, children characterized as resilient were doing much better than participants with childhood psychiatric problems, but these resilient children were doing more poorly in the domains of physical health and financial or educational functioning than participants who had been exposed to fewer types of adversity and did not experience childhood psychiatric problems. Thus, exposure to childhood adversity appears to have a significant long-term effect that is delayed among individuals who initially appeared to be resilient.

Persistent mental health was strongly associated with minimal exposure to adversity. These individuals are fortunate. The importance of being fortunate when it comes to experiencing few adversities has been reported frequently (18, 20, 30). In this study, nine out of 10 children who never displayed psychiatric problems were exposed to adversity in only one domain or were free of exposure to childhood adversity. Indeed, childhood resilience was rare among individuals exposed to multiple forms of childhood adversity. There are a number of reasons for this finding. First, our analysis included children with subthreshold psychiatric problems, who are sometimes categorized as resilient in other studies. Second, to avoid misclassification, our analyses included only participants who had multiple assessments during childhood. Third, we used a cumulative measure across domains and time to define adversity status. In other words, adversity status in our study was based on many aspects of the childhood environment, including home and school settings, across several years (17, 31). Failure to measure these adversities or focusing on a single form of adversity might lead to the misclassification of the child’s overall history of adversity and thus overestimate the number of children who do not develop psychological problems in the face of adversity.

Good mental health in the face of significant adversity (i.e., resilience), however, did not always persist into adulthood. This finding was anticipated by a previous study that discussed the potential for early adversity to have either “sensitizing” or “steeling” effects on individuals in response to challenges in their future (32). Some authors have argued that resilience to childhood adversity could be advantageous: successfully overcoming significant adversity early in life could prepare individuals for facing adversity after reaching maturity. Conversely, lack of exposure to all forms of adversity could deprive children of opportunities to cultivate the coping skills necessary to navigate the vicissitudes of life. However, neither of these reasonable hypotheses was fully supported by our findings.

Exposure to multiple forms of adversity had long-lasting effects in some domains, even if the effect was not immediately apparent (i.e., sleeper effect [33]), and these effects were not pervasive across all areas of life. Resilient individuals generally did better than individuals who had childhood psychiatric problems, particularly when it came to adult substance use disorders. This finding is particularly notable given the strong association between childhood adversity and subsequent substance use problems. This study suggests that some of the risk associated with childhood adversity may be mediated by childhood mental health problems. Potential “costs” of resilience have also been reported in a follow-up to the Kauai Longitudinal Study, with some of the children who were characterized as resilient experiencing stress-related health problems as adults (12). More recently, physiological costs of psychosocial resilience have been identified by Brody and colleagues in their sample of at-risk children from the southern United States (34). Although resilient individuals may escape their childhood relatively unscathed, the stress of maintaining psychological health despite adversity may catch up with them later in development. Additionally, adversity continues to occur throughout the lifespan. In our study, resilient children who had experienced multiple forms of adversity in childhood were at the greatest risk for exposure to trauma in adulthood.

This study included a sample of participants that was representative of the southeastern United States but not representative of the U.S. population. Compared with the U.S. population, African American and Latino people were underrepresented, and the American Indian population was overrepresented. However, the rates of childhood psychiatric disorders in this study are similar to those in other representative samples that included higher levels of Hispanic and African American youths (7). Although our measure of early adversity included risk factors with strong evidence of association with multiple common childhood psychiatric disorders, the degree of adversity is often subjective, and other experiences with the potential to influence emotional and behavioral functioning may have been omitted. Whereas our cumulative childhood adversity scale provides a broad picture of environmental risk, a similar measure of genetic risk was not tested because of the lack of replicated variants of risk for common childhood diagnostic disorders (e.g., 35). Finally, the group of interest (i.e., resilient individuals) was small, and these analyses need to be tested in other studies that have carefully assessed psychosocial functioning longitudinally from childhood to adulthood.

In this sample, resilience was uncommon, and for many, it did not last. Exposure to adversity had an influence, even if the effects were delayed. In the unlikely scenario that it were possible to identify all the factors that promote resilience, identify all the children at significant risk, and teach these children to implement resilience factors effectively, the public health effect would likely be modest at best, and the children might nevertheless develop subsequent problems. Alternatively, reducing childhood exposure to adversity could strongly promote long-term health and exert a more substantial and long-lasting impact on public health. For example, reducing childhood poverty—which is often viewed as a “risk of risk,” because it co-occurs with and spawns so many other forms of adversity—could be a promising approach, which might be implemented through, for example, the expansion of the child tax credit or a universal basic income (36). This is not to suggest that resilience-based interventions do not have a role within comprehensive efforts to reduce the burden of mental illness. But such efforts should come only in the context of widespread public health efforts to minimize exposure to childhood adversity. Finally, an unintended consequence of the notion of resilience for individuals who eventually develop psychological problems is stigma. On this point, our findings are quite clear: the development of some level of mental health problems is the normative response to significant adversity. Individual resilience followed by persistent mental health is rare and may not be a reasonable goal. Public health efforts should prioritize reducing risk and treating individuals who are ill.