Over the past years, the focus of research on attention-deficit-hyperactivity disorder (ADHD) has shifted towards its developmental course rather than its childhood characteristics. While the significance of ADHD beyond childhood has become clearer [1
], insight into the factors that explain individual differences over the course of ADHD is rather limited. One important factor in the variable course of ADHD symptoms may be the exposure to chronic stress. Grant and colleagues [3
] reviewed the association between stressors and child- and adolescent psychopathology and showed a bidirectional, continuing, cycle of stressors leading to enhanced symptoms, in turn leading to more stressors, and so forth. Although their paper was not focused on ADHD, previous studies have reported that exposure to stress and ADHD are associated (e.g., family conflict, poor home conditions, death of a close family member or friend [4
]. These studies suggest the existence of a stress–psychopathology–stress cycle underlying ADHD, in which (families of) individuals with ADHD are not only exposed to “bad luck” stress but also generate stress exposure themselves by creating, for example, peer conflict or financial problems. It may be hypothesized that differences in stress exposure are linked to differences in the course of ADHD, with a reduced stress–ADHD–stress cycle in those who gradually remit and an enhanced cycle in those who persist. Importantly, stress exposure and management of stressful circumstances are in part modifiable, offering the potential for clinical intervention. The literature on stress exposure and (persistence of) ADHD is scant. In the present paper, we study whether individual differences in psychosocial stress exposure between childhood and adulthood are associated with the course of ADHD. Further, we study whether such differences are associated with the core symptoms of ADHD, or with co-occurring problems, including reduced effortful control, emotion dysregulation, and internalizing symptoms.
Traditionally, stress–psychopathology relationships are studied in the context of internalizing problems, in particular depression, anxiety disorders, and medically unexplained somatic complaints [6
] rather than ADHD. Therefore, when studying stress-exposure and the course of ADHD, not only the core symptoms of ADHD, but also the classic stress-related anxiety, depression, and somatic complaints need to be considered as possible comorbid outcomes. It may be hypothesized that individuals with ADHD who are exposed to stress are characterized both by a more persistent form of ADHD and by the onset of comorbid internalizing problems alongside this persistent ADHD trajectory [9
]. In addition, the effects of stress exposure in children with ADHD may also differ from the classic stress related internalizing problems. Recent accounts of ADHD symptomatology have proposed that emotional regulation problems (e.g., low frustration tolerance, explosive anger) are an important aspect of ADHD [10
]. Although not studied as such, it is plausible that children with ADHD who are in stressful circumstances may express emotional regulation problems rather than internalizing problems. For example, laboratory experiments indicate that children with ADHD are more emotionally intense and less proficient in anger management during frustrating, stress-inducing tasks relative to children without ADHD [12
In the present study, we investigate the manifestation of differential developmental trajectories, in which individual differences in stress exposure, core ADHD symptoms, internalizing problems, and emotional regulation problems are clustered. We hypothesize that high exposure to stressful life events is related to a more persistent and complex form of ADHD (i.e., multi-problem, including internalizing and emotional regulation problems), while low-stress exposure relates to remitting ADHD over the course of adolescence. We focus on four waves of longitudinal data collected between age 11–19 from a large number of individuals derived from the TRAILS population and clinical cohorts [14
]. This developmental period (early adolescence to young adulthood) is characterized by significant normative developmental changes as well as individual differences over the course of ADHD, stress exposure, internalizing, and emotional regulation problems. We apply multivariate latent class growth analysis on repeated measurements of stress exposure, ADHD, internalizing, and emotional regulation symptoms. This approach makes full use of the repeated simultaneous measurement of all variables and identifies subgroups of individuals with trajectories that shift jointly across multiple domains [15
]. This allows identification of subgroups of individuals with ADHD who have distinct stress exposure patterns combined with a different course of ADHD and co-occurring problems.
Our study examined the hypothesis that high-stress exposure is associated with a poor ADHD course from pre-adolescence to young adulthood. We first showed that stress exposure was consistently higher in the four ADHD subgroups as compared to the non-ADHD subgroups, in line with existing literature [4
]. Second, the two non-remitting ADHD subgroups were associated with very high, curvilinear stress exposure (peak in mid-adolescence), while the two remitting ADHD subgroups were related to lower as well as declining stress exposure, therefore, supporting our hypothesis. Third, the findings indicated that the two persisting ADHD combined types with very high-stress exposure were associated with either co-occurring severe and persistent emotion dysregulation (irritability, and especially extreme reactivity and frustration) or elevated irritability and elevated and increasing anxiety and depression symptoms. In terms of demographic characteristics, comparatively low SES and low IQ characterized both trajectories, while stimulant use was more likely in the former group, and females were more often present in the latter group. Findings suggest that the stress–psychopathology–stress cycle may play an important facilitating or sustaining role in the continuation of symptoms of ADHD from childhood to adolescence. Stimulant use does not seem to be protective herein.
The two persisting combined type trajectories of ADHD were continuously exposed to higher levels of stress compared to the two less severe remitting ADHD trajectories (i.e., moderate combined type and inattentive type). This can be explained by both heightened sensitivity to and reduced regulation of stress, as these two vulnerabilities hold true particularly for adolescents with combined type ADHD. Children with ADHD enter adolescence with—on average—a less matured prefrontal cortex [27
]. Adolescence, in particular, is a period of important changes in brain connectivity within the prefrontal cortex and with other brain regions, supporting the maturation of cognitive control, such as emotion regulation abilities and impulse control abilities [28
]. During adolescence, the developing prefrontal cortex is especially vulnerable to the effects of stress [31
]. Our findings support the idea that prefrontal cortex immaturity at the start of adolescence and slower prefrontal maturation during adolescence makes children with ADHD particularly vulnerable in terms of both self-generation of stress and the potentially harmful effects of stress on prefrontal functions of the brain, setting the stage for a persistent and comorbid form of ADHD. Poor socio-economic circumstances and lower IQ may play an additional role herein.
Of strong interest is the finding that high-stress exposure was associated with two persistent ADHD trajectories with diverging comorbid trajectories: one characterized by persistently elevated irritability, as well as elevated and increasing internalizing problems but without the more extreme reactivity and frustration; the other characterized by severe ongoing emotional regulation problems (irritability, extreme reactivity, frustration) but without the internalizing problems. This finding supports earlier work showing etiological overlap (in cognitive control) between internalizing and externalizing symptoms [32
]. At the same time, our findings indicate that a subset of children with the most severe persistent ADHD and exposed to high stress are not vulnerable for internalizing problems. Trait impulsivity in young childhood is thought to be one of the driving liabilities to combined-type ADHD among children and is a core predisposing vulnerability to all (adult) externalizing psychopathology [34
]. The diverging trajectories of the two most severe ADHD groups in our sample may be explained by the differential presence of early trait anxiety, which is thought to attenuate impulsivity to some extent while representing a vulnerability for progression into more severe internalizing psychiatric outcomes [34
]. Indeed, our findings show differences in anxiety/depression, impulsivity, extreme reactivity, and frustration that were apparent at age 11, suggesting that the divergence of both pathways likely started before the onset of adolescence. Trait impulsivity and anxiety are thought to fully manifest as psychiatric symptoms upon interaction with the environment [34
]. Chronic stress exposure may be one such potent transactional factor along the ADHD course from young childhood into adulthood.
Intriguingly, we did not identify subgroups for whom high-stress levels lead to trajectories of internalizing problems or severe emotion dysregulation problems in the absence of ADHD. Only a “mild internalizing” subgroup was found with enhanced but much lower stress levels. This finding may be taken to indicate that stress-exposure and ADHD are even more intertwined than currently acknowledged in the literature which is focused on the role of stress in depression and anxiety. Importantly, the finding should not be interpreted to show the irrelevance of the classic stress–stress cycle. Rather, since we oversampled male participants, this cycle is more likely in our non-sampled screen-negative female participants. It is likely that our identified “mild internalizing” subgroup will develop more severe internalizing problems when they reach their early twenties when depression and several anxiety disorders have their peak incidence. Our findings are suggestive of a different underlying etiology for anxiety-depression with and without ADHD, respectively. We tentatively propose a neurodevelopmental subtype of “internalizing ADHD,” which is different from the adolescent onset of internalizing problems without ADHD. This distinction between “neurodevelopmental” and “traditional” internalizing problems is strongly consistent with a series of recent papers on the role of childhood irritability in depression [35
]. These authors similarly propose the existence of “two depressions.” Here we confirm the neurodevelopmental route to depression and show that (similar to “typical” i.e., adolescent and young adulthood-onset depression) this route is strongly intertwined with stress exposure.
Our study has strong assets. We applied multivariate latent class growth modeling that captures this heterogeneity in both stress exposure and behavioral symptoms. This approach makes full use of the repeated simultaneous measurement of all variables by parsing subgroups of adolescents with specific forms of continuity and change [15
]. In addition, our same age cohort study design is critical for studying the heterogeneity of the ADHD course. Often, ADHD participants of varying ages are combined in a sample, confounding the heterogeneity that exists among same-age individuals with heterogeneity caused by differential development over time. That is, by combining various ages, it is, for example, ignored that hyperactivity in childhood, typical for ADHD at that age, has a different meaning in adolescence, when it is less commonly reported [37
]. Our study design of same-age individuals followed over time is very strong in disentangling momentaneous heterogeneity from developmental heterogeneity.
Our study has limitations as well. The current study of trajectories does not allow inferences of causality; rather our analyses show the strong intertwinement of high-stress exposure with two qualitatively different trajectories of persistent ADHD, both different from remitting ADHD. We regard our study, therefore, as the first necessary step in showing the relevance of stress exposure for the course of ADHD, which has been ignored so far in the literature. Future observational and treatment studies may extend our work by focusing on the causal relations among stress exposure and the symptoms of ADHD over time. As the early stages of a progressive course of ADHD were already apparent in pre-adolescents, we recommend to not only study the potential causal role of stress during adolescence but also pre-adolescence and even pre-onset to ADHD. We propose to study the consequences of stress exposure on ADHD symptoms both at the microlevel, capturing day-to-day causal relations, and at the macrolevel for causal changes over longer periods of time.
Our findings suggest that taking into account stressful environmental factors should be an inherent part of the diagnosis, prognosis, and treatment of ADHD. The clinical profile and level of stress exposure at the start of adolescence are strong markers for what comes next: preadolescents living in chronic stressful conditions who have combined type ADHD symptoms and enhanced anxiety and depression are likely to have persistent ADHD and future exacerbation of anxiety and depression. Conversely, under similar stressful conditions, high parent-rated compared to self-rated emotion regulation problems mark the highly severe and persistent course of both ADHD and emotion dysregulation symptoms. High-quality normative data for instruments that adequately tap these behaviors, derived from pre-adolescents with ADHD may aid clinical thinking about the prognosis. Importantly, the extent to which individuals expose themselves to stressful conditions and how they manage stress are potentially modifiable, thereby providing a therapeutic angle for prevention or interruption of the negative cycle of persistent ADHD and (development of) comorbid conditions.
In sum, we conclude that the stress–psychopathology–stress cycle likely plays an important facilitating and sustaining role in the persistence of ADHD and (development of) comorbid problems from childhood to adolescence. We showed that ADHD combined type with high-stress exposure can either take the form of co-occurring severe and persistent emotion dysregulation (irritability, frustration, extreme reactivity) or persistently high irritability and increased anxiety and depression symptoms. The former trajectory fits well with recent literature that emphasizes that emotion dysregulation is present in a substantial subset of children with ADHD and we show that the course of this group into adulthood is unfavorable. The latter trajectory fits well with a recent emphasis on a neurodevelopmental route to depression. Conversely, low and declining stress exposure and remitting or inattentive ADHD may accumulate towards desistence from both internalizing and externalizing problems. Clinicians may, therefore, include the presence of stressful conditions in the diagnostic process for prognosis and potential prevention or interruption of an adverse ADHD course.