Clinimetrics and Allostatic Load

Research on childhood stress and allostasis is surprisingly recent. It was less than 40 years ago, in 1985, that Vincent Felitti, who directed the Southern California Kaiser Permanente Obesity Clinic, met a 28-year-old woman. She asked Felitti for his help. She weighed 408 pounds, and she announced to Felitti that she was there to become thin [1].

Felitti had developed the weight loss program at Kaiser Permanente and over the years had considerable success in helping patients lose substantial amounts of weight. In just a little over a year, the woman lost 276 pounds, reaching her previously stated goal of 132 pounds. But like so many of the women in Felitti’s program, this patient experienced the same weight gain that he saw in others. In just a matter of weeks after reaching her target weight, she gained nearly 40 pounds.

Felitti was determined to find out why. He reviewed the medical records of his patients and carried out statistical analyses of the cohort, looking for common determinants of recurrent weight gain. Nothing in the records or in the statistical models explained why the women who had been so successful at losing weight could not keep it off. Then Felitti did something no one had thought to do; he carried out an intimate, personal medical history of the patient’s biography. When Felitti asked the woman why she thought she had gained back so much weight in such a short time, she explained that a co-worker had started to give her new attention and even sexually propositioned her. Shortly afterward, she began binge eating.

The woman reported that gaining weight was a way to make herself less attractive to men, an attraction that filled her with anxiety. When Felitti probed further, the woman confessed to having been sexually abused by her grandfather starting at age 10. Felitti went on to conduct similar interviews with many other patients in his clinic. Out of 286 people in the weight loss program he interviewed, Felitti discovered that 55% of the obese women reported sexual abuse as a child. [1] Later, Robert Anda, a CDC epidemiologist, heard Felitti present his research at a conference in 1990, and Anda proposed a large-scale survey of 17,000 individuals to answer questions about adverse childhood experiences (ACEs) and develop a predictive model. The adverse experiences of childhood were organized into 10 categories: physical, sexual, or psychological abuse; emotional or physical neglect; divorce; mental illness; substance abuse; domestic violence; criminality among household members [2].

The survey reported that 64% of participants had at least one of 10 adverse events, 20% had at least 3, and 12% had four or more. On average, for people with at least 4 adverse events, the risk of suicide, alcoholism, and smoking all increased substantially. Those with at least 6 ACEs lived 20 years less, on average, than other people in the survey. Over the ensuing years, the ACEs have become embedded in medicine and public health as a tool for detecting toxic stress in children and adolescents that is associated with later life mental and physical illness.

Yet, it is hard to ignore the irony in the use of a statistical score developed on a population to substitute for the in-depth personal interview that was the hallmark of Felitti’s method. He did not tally up all the childhood stresses his obese patient experienced. He found in her biography the single core experience that created a fundamental source of distress and appreciated its consequences in her later adult life. She may well have scored zero on all the other ACE components and appeared to be at low statistical risk for later life problems, and it would not have mattered. Her one powerful trauma discovered through a deeply personal patient-physician interview was the only risk that mattered.

More recently, the understanding of individual-level stress has been enabled by the development of a major conceptual advance in stress-related research. Allostasis, a new model of physiological regulation that complements homeostasis, was proposed by Peter Sterling and Joseph Eyer in 1988 [3]. Unlike the rigid constancy of homeostasis, allostasis suggests that physiological stability is achieved through variation and is distinguished by predictive regulation orchestrated by the brain. An excellent contemporary view of these ideas is given in Peter Sterling’s recent book, “What is Health? Allostasis and the Evolution of Human Design (2020)” [4]. When allostasis cannot be maintained, a new notion, introduced in 1993 by McEwen and Stellar and which they refer to as allostatic load, becomes salient. Conceptually, allostatic load refers to the wear and tear on the body that accumulates as the individual is exposed to repeated or chronic stress. When there is an inability to recover from such persistent challenge, allostatic load moves into a state of allostatic overload, and clinical intervention may be necessary [5].

The focus on the individual’s level of stress and response is a feature that separates allostasis and allostatic load from the average population-level score that is characteristic of ACEs. While ACEs may have value in defining risk in public health, they are less useful in guiding the clinical decision-making that is fundamental to clinical medicine. In this issue of the journal, Lucente and Guidi provide new insight into these distinctions between population stress measures and clinical measures. Their paper, “Allostatic Load in Children and Adolescents: Systematic Review,” is notable for several reasons [6].

Most research on allostatic load and overload relies on biological markers that, in aggregate, reflect disordered functioning of several physiological systems, including neuroendocrine and immune systems, the hypothalamic-pituitary axis, cardiovascular and endocrine systems, as well as the regulation of sleep and circadian processes [7]. However, this literature is largely static in character and does not develop the idea of cumulation of allostatic load as a dynamic process. Indeed, there is, at the present time, no full operationalization of the concept of allostatic load for the whole person. However, a close approximation to it was first proposed by Koob and Moal [8] where they consider a dynamic formulation of drug addiction framed entirely in an allostasis and allostatic load context. Of particular importance in this work is the integration of social and stressful challenges together with biology. The drug addiction formulation has, more recently, been further developed by Koob and Schulkin [9].

In stark contrast to these versions of allostatic load, Lucente and Guidi provide insights beyond biomarkers by introducing a clinimetric formulation and assessment of the concept of allostatic load. These measures were developed recently and are included in the Diagnostic Criteria for Psychosomatic Research under a specific diagnostic category [10‒12]. The clinimetric criteria require not simply the presence of a stressor that exceeds the individual’s coping abilities but also clinical manifestations of distress felt by the person.

The clinimetric tools are essential for making allostatic load, and by inference, ACEs, relevant to actual clinical practice by introducing biomarkers in the context of a patient’s biography, or life experience [13‒18]. In doing so, it also enables the physician to personalize interventions that help mitigate toxic stress and its impact on health.

It is not surprising that Lucente and Guidi found that most of the studies in their systematic review relied very substantially on static measures of biomarkers, thereby reflecting the fact that very little progress has been made toward the development of a fully dynamic formulation of allostatic load since the initial attempt to operationalize the idea in 1997 [7]. The clinimetric formulation, in addition to the above-mentioned drug addiction studies, is one of the very few examples of real progress in this area, and its direct connection to clinical practice is laudable. Their review confirmed that greater allostatic load – the static version – is associated with poorer health outcomes for children and adolescents. However, they emphasize the important opportunity to apply clinimetric criteria, thereby personalizing interventions that would improve the health and well-being of children now and in the future.

The associations between poverty and higher levels of the current version of allostatic load that they report are not unexpected. An interesting finding that deserves greater attention is the role played by accessibility to green spaces. Many poorer neighborhoods lack the amenities that green spaces provide to create a greater sense of calm and stability. Lucente and Guidi noted that the greater the distance from school to the nearest green space, the higher the allostatic load in children. The consistency of this association argues for urban planners to pay close attention to creating more green space in neighborhood development.

Allostatic load in children and adolescents was found to be associated with worse physical and mental health later in life. Abnormal nutritional status, both malnutrition and overweight, stood out as prominent disorders whose long-term consequences are profound and yet preventable. The authors rightly stress that psychotherapeutic strategies aimed at improving coping with stress, such as mindfulness-based approaches and cognitive-behavioral strategies, may be of substantial value to children’s health. Over the past 20 years, speculation about this point has been drastically reduced by the introduction of Well-Being Therapy and its use in clinical practice [18] and in positive mental health promotion in Italian schools. Lucente and Guidi reinforce this point by making the important suggestion that promoting healthy lifestyle modifications among children and adolescents through the pursuit of euthymia may provide enduring benefits.

As a final point, the emphasis on allostatic load and ACEs fosters a focus on negative experience. In childhood and adolescence, it would be particularly helpful to bring in a preventive emphasis with a focus on interventions designed to promote positive psychological functioning and the absence of cumulating allostatic load. As indicated above, such a program has been in progress in a limited set of Italian schools with early signs of successful promotion of positive functioning [18]. Much more in this direction deserves attention with the objective being to promote allostasis and keep allostatic load, however it is measured, as low as possible.

Realizing the hopefulness that characterized the conclusions of the review will require moving away from an approach to stress in children and adolescents measured by ACEs to an approach that focuses on the person by an informed and expert clinician. The lesson of Felitti should not be a blind adherence to a statistical score that measures the average risk of a population of children. Rather, we should recommit to further development of the clinimetric indexes first proposed by Fava et al. and a renewed emphasis on unpacking the specific features in any person that are most impactful for their later health. Until we do, neither allostatic load nor ACEs will suitably point the way to improved child health.