These epidemiological findings converge with evidence from neurobiology about numerous effects of childhood stress on brain and physical systems (Glaser 2000
). Extreme, traumatic or repetitive childhood stressors such as abuse, witnessing or being the victim of domestic violence, and related types of ACES are common, tend to be kept secret, and go unrecognized by the outside world. Likewise, the fight-or-flight response among children exposed to these types of stressors, and the attendant release of endogenous catecholamines and adrenal corticosteroids are both uncontrollable and invisible (Perry 1998
; Teicher 2002
; De Bellis 1994
; Scaer 2001
). Furthermore, the detrimental effects of traumatic stress on developing neural networks and on the neuroendocrine systems that regulate them have until recently remained hidden even to the eyes of most neuroscientists. However, the information and data that we present herein suggest that this veiled cascade of events represents a common pathway to a variety of important long-term behavioral, health, and social problems ().
Summary of the convergence between neurobiological effects of childhood maltreatment with ACE study epidemiological findings
The convergence of evidence from neurobiology and epidemiology calls for an integrated perspective on the origins of health and social problems throughout the lifespan. This constellation of effects from childhood stressors calls to mind the wisdom of Occam’s razor, a celebrated dictum in medicine, which holds that if a single unifying explanation can be found for multiple symptoms and problems, then it is likely that the correct explanation lies in the simplest account (Lo Re and Bellini 2002
). In the context of what we present herein, the application of this dictum has the potential to unify and improve our understanding of many seemingly unrelated, but often co-morbid health and social problems that have historically been seen and treated as categorically independent in Western culture.
Certain neurobiological findings are especially congruent with the data from the ACE Study reported herein (). Magnetic resonance imaging (MRI) has revealed reductions in hippocampus (Bremner 1997
; Stein 1997
), and amygdala (Driessen 2000
; Schmahl 2003
) volumes as well as deficits in verbal declarative memory measured with neuropsychological testing (Teicher 2000
; Heim and Nemeroff 2001
) among women who were sexually abused as children. The hippocampus plays a role in memory storage and retrieval; we found that impaired memory of childhood increases as the ACE score increases. Neurobiological evidence supports the hypothesis of dysfunction in hippocampus, amygdala, medial prefrontal cortex, and other limbic structures believed to mediate anxiety and mood dysregulation following early abuse (Teicher 2002
). We, in turn, demonstrated a graded relationship of the ACE score to affective symptoms and unexplained periods of panic among our study participants. We found that a history of hallucinations increases as the ACE score increases; these symptoms may be related to alterations in hippocampal and/or prefrontal cortical function. The amygdala plays a critical role in fear responses and probably sexual and aggressive behaviors (Pinchus and Tucker 1978
) and in the current study we show strong relationships of the ACE score to sexual behaviors, poor anger control, and the risk for perpetrating intimate partner violence.
The current study adds support for numerous effects of childhood adverse experiences on physical health. Stress is known from animal studies to be associated with a broad range of effects on physical health, including cardiovascular disease, hypertension, hyperlipidemia, asthma, metabolic abnormalities, obesity, infection, and other physical disorders (Musselman 1998
; Kaplan 1982
; Rozanski, McEwen and Stellar 1993; Anda 1993
). Findings of increased obesity as the ACE score increases in this study and reported elsewhere (Williamson 2002
) are consistent with animal studies showing that stress, acting through the effects of glucocorticoids on the glucocorticoid receptor on intra-abdominal adipocytes, leads to increased intra-abdominal fat which carries its own independent mortality risk.
We found a strong relationship between early adverse experience and substance use and abuse (illicit drugs, alcohol, and nicotine) later in life. Studies in animals show that early stressors lead to increased activity of the locus coeruleus with resultant increased release of nor-epinephrine in the brain (Abercrombie and Jacobs 1987
). Substances such as heroin and alcohol decrease firing of the locus coeruleus, while substance withdrawal has the opposite effect (Bermner 1996). Consistent with this, the onset of substance abuse corresponds to the time of traumatization in PTSD patients, and these patients report that heroin and alcohol decrease symptoms of PTSD (Bremner 1996b
). Stress also results in altered release of dopamine in the nucleus accumbens (striatum), the primary reward system within the brain (Deutch and Roth 1990
). Smoking causes release of dopamine in this area, which is felt to underlie the addictive properties of nicotine (Volkow 2003
). Early adverse experiences may disrupt this dopamine circuit, leading to increased risk of smoking, with its attendant negative health consequences. In summary, findings from animal studies provide a physiological rationale for how early stress can be associated with substance abuse and smoking in later life.
Another interesting finding is the relationship between ACE score and sexuality (early intercourse, promiscuity, sexual dissatisfaction) in adulthood. Animal studies show that early stressors result in long-term changes in peptides such as oxytocin that regulate pair bonding and social attachment (Insel and Winslow 1998
; Francis 2002
). Early adverse experiences may disrupt the ability to form long-term attachments in adulthood. The unsuccessful search for attachment may lead to sexual relations with multiple partners, with resultant promiscuity and other issues related to sexuality.
The monoamine neurotransmitter systems (norepinephrine, dopamine, serotonin) (Valentsein 1998) act within a primary regulatory system of large neural networks; these monoamine systems help to orchestrate complex neural functions. Their ubiquitous patterns of connectivity originate in the lower regions of the brain and send projections throughout the brain; in addition, they receive input from the autonomic nervous system and peripheral sensory apparatus (Foote 1983
). In young animals, experimental manipulation of these systems can create behaviors similar to those seen in abuse victims, including aggression, eating problems, alcohol use, stress-response dysfunction, hyper-reactivity, anergy, and many other behavioral problems. A similar situation exists in humans in whom monoamine dysfunction has been hypothesized in a host of neuropsychiatric syndromes, including aggressive and violent behavior, suicidality, alcoholism, substance abuse and dependence, depression, anxiety disorders, and social/relational problems. We know from several studies that the functioning of these monoamine systems in adults is influenced by childhood experiences (De Bellis 1999b
; Whitfield 2003b
). In addition, a recent study of a polymorphism for the promoter region of the serotonin transporter (5-HTT) gene found that childhood maltreatment increased the risk of depression in early adulthood for persons with the common “short” allele compared to persons with the long allele; the short allele is associated with lower transcriptional efficiency of the promoter (Caspi 2003
). Not surprisingly, many currently prescribed psychoactive drugs act by altering the dynamics of these monoamine systems. In some circumstances, the effects of these drugs may have caused an 182 oversight of the important distinction between understanding intermediary mechanisms (alterations in monoamine neurotransmitter systems) and recognizing the underlying causes of these alterations (childhood traumatic stress).
Numerous studies have shown that early abuse survivors have multiple overlapping psychiatric disorders (Kessler 1995
) which have been described as “comorbidity”. The term comorbidity, however, can imply that these represent unique disorders with distinct etiologies (Lillienfeld 2003). An alternative explanation is that several disorders (e. g., depression, PTSD, dissociative disorders, substance abuse, borderline personality disorder) have to varying degrees a common etiology and are modulated by genetics (Caspi 2003
) and repeated exposure to stress such as childhood maltreatment. Indeed, the term “trauma spectrum disorders” has been used to describe these overlapping conditions (Bremner 2003b
). In addition, the artificial distinction between psychiatric and physical disorders has represented an impediment to the effective treatment of the numerous problems among survivors of childhood maltreatment. Epidemiological findings are consistent with a need to develop more broad based approaches to addressing the wide spectrum of effects of childhood maltreatment ().
There are several potential limitations with retrospective reporting of childhood experiences and self-reporting of the outcome measures. For example, respondents may have had difficulty recalling certain childhood events (Edwards 2001
) or may choose not to disclose certain experiences or personal behaviors. Longitudinal follow-up of adults whose childhood abuse was documented has shown that their retrospective reports of childhood abuse are likely to underestimate
actual occurrence (DellaFemina 1990
; Williams 1995
). Interestingly, evidence of the effects of traumatic stress in childhood on the hippocampus provides a neurophysiologic explanation for this phenomenon. Difficulty recalling childhood events likely results in misclassification (classifying persons truly exposed to ACEs as unexposed) that would bias our results toward the null (Rothman and Greenland 1998
). Thus, this potential weakness probably resulted in underestimates of the true strength of the relationships between ACEs and the 18 outcomes we examined.
The historical mind-body dichotomy that persists in traditional Western medical training points medical researchers and clinicians away from risk factors that may be judged psychosocial. Thus, the original traumatic pathophysiological insults may be “silent” until much later in life (Brown 2001
; Putnam 1998
), when they are likely to be overlooked by investigators and clinicians who are understandably prone to focus on proximate determinants of human well-being. This leads to treatment of symptoms
without a full understanding of their potential origins in the disruptive effects of ACEs on childhood neurodevelopment.
The argument for a causal relationship between ACEs and a variety of outcomes is strengthened by the combined evidence from neurobiology and epidemiology. This argument is important because evidence of causation affects decisions about prognosis, diagnosis, and treatment and can enhance understanding of the role of the childhood stressors on the developing brain in producing changes in affect, behavior, and cognition (Putnam 1998
We summarize the application of Sir Bradford Hill’s 9 criteria for establishing an argument for causation (van Reekum 2001
) in the context of this converging evidence:
● Demonstration of a strong association between the causative agent and the outcome. The strength of the relationship between ACEs and numerous outcomes is consistently strong as reported herein.
● Consistency of findings across research sites and methods. Numerous studies in different study populations and measures of abuse, neglect, and related experiences have shown relationships of ACEs to a variety of symptoms and behaviors.
● Specificity. In the context of the converging evidence from epidemiology and neurobiology, specificity is lacking, but this in no way detracts from the argument of causation. The ACE score is a combined score representing cumulative stress and was not designed to provide evidence of specificity. Moreover, ACEs would be expected to be associated with multiple outcomes because of their effects on a variety of brain structures and functions.
● Temporal sequence. Most of the outcomes presented herein occurred during adulthood; the exposures (childhood experiences) clearly antedate the outcomes in these cases.
● Biological gradient. The “dose-response” relationship between the number of ACEs and each of the outcomes (as well as the number of comorbid outcomes) is strong and graded. This is consistent with cumulative effects of childhood stress on the developing brain.
● Biological plausibility. The strength of the convergence between epidemiology and neurobiology is most evident here. Recent studies from the neurosciences show that childhood stress can affect numerous brain structures and functions providing convincing biologic plausibility for the epidemiologic findings.
● Coherence.“The term coherence implies that a cause and effect interpretation for an association does not conflict with what is known about the natural history and biology of the disease (Rothman 1998).” In fact, recent research shows that childhood maltreatment interacts with a common functional polymorphism in the promoter region of the serotonin transporter 5-HTT, resulting in higher risk of depression and suicidality (Caspi 2003), both of which are associated with the ACE score. This information is consistent with an effect of early maltreatment on monoamine pathways known to be involved in depressive disorders.
● Experimental evidence. This is the most persuasive evidence, but for ethical reasons randomized experiments depend on animal studies. Evidence from studies in rodents and primates show that stressful exposures induce neuroanatomical and neurophysiologic differences as well as aggression and drug seeking behaviors.
● Analogous evidence. A widely acknowledged analogy for an exposure causing a multitude of outcomes (as seen with ACEs, including a dose-response relationship) is the causal relationship of cigarette smoking to cardiovascular diseases, neoplasms, lung disease, and other health problems (CDC, 2002).
In conclusion, there is a striking convergence of recent findings from the neurosciences with those from a large epidemiologic study of the long-term effects of ACEs which has the potential to open multidisciplinary approaches to studying and improving human well-being. Current practices of medicine and public health are fragmented by categorical funding, organizational boundaries, and a symptom-based system of medical care. Prevention and remediation of our nation’s leading health and social problems is likely to benefit from understanding that many of these problems tend to be co-morbid and may have common origins in the enduring neurodevelopmental consequences of abuse and related adverse experiences during childhood.