Developing a Theory of Health Behavior: An Interview with Howard Leventhal, PhD

Martin S. Hagger, PhD^✉; Theories and Techniques of Behavior Change Interventions SIG Co-Chair

Howard Leventhal, PhD

 

How did you first become interested in developing a theory of health behavior?

I never thought of myself as focused on “developing a theory of health behavior”. Our focus was to describe what individuals, see, feel, think and expect when acting to control a health threat or acting to enhance health; that is to understand the mechanisms generating these actions in real time. It didn’t matter whether the “action” was latent, vaguely conceived and unexpressed or clearly expressed in thought or behavior. This focus on specific, underlying processes, was conditioned in part by the state of the discipline in the 1950s and 60s; “big” theories, for example, Hull and classical reinforcement positions, were suspect. It no longer seemed reasonable to explain all behavior with a single principle, and individuals espousing one-size fits all theoretical models were not warmly received.

On a more personal side, I was interested in how things worked and taking things apart and building stuff was what I did for as far back as I can remember. I can see myself, not yet 4, trying to construct a dirigible using my older cousin’s erector set. It soon became clear that it would be extremely difficult if not impossible to construct an elliptical form given the shape and rigidity of the available components. The lesson learned was simple and clear; what you can build is constrained by the materials and tools at hand. Understanding how things work and building things and models, may require searching for and creating new materials and tools.
 

What sort of research is needed to make health behavior theories more effective and more useful?

Creating interdisciplinary teams to develop methods for describing and understanding the mechanisms generating perceptions, thoughts and actions taken to enhance and control threats to health in real time. And recognizing that understanding underlying mechanisms is not the same as predicting specific behaviors and outcomes. Inaccurate, indeed nonsensical models can make extremely accurate predictions.

The focus on mechanisms has important consequences. First, it requires teams of investigators from multiple disciplines who look at behavior from different, yet related perspectives. Our teams included social and clinical psychologists, psychometricians, statisticians, physicians and nurses from varying medical specialties, as well as patients who were full team members!

Second, our group was open to all available methods for collecting and examining evidence; qualitative interviews and focus groups, analysis of cross sectional and longitudinal, descriptive data, randomized trials and the arbiter of arbiters, the multifactorial experiment. Each provided insight into the mechanisms generating behavior.

Third, we collected data from multi-factorial experiments in “real world” settings, not just the laboratory. Observations were made and experiments conducted with patients in hospital. Bringing reality into the laboratory was central to our program, as it was to that of Stanley Milgram, my “next door” office mate and close friend.
 

What is your perspective on the current direction of research on the predominant theories in health behavior?

Different types of theory are essential in behavioral health. Epidemiological data, numbers from large numbers of respondents differing in ethnicity, gender, age, and context, (morbidities, insured vs. uninsured, etc.) provide population based estimates critical for decisions involved in funding and providing facilities and health care services. Findings from carefully designed epidemiological studies can also provide valuable clues regarding the processes underlying individual action.

Most behavioral health investigators are focused on how individuals obtain, evaluate and respond to information urging them to act to treat or prevent signs of somatic distress and dysfunction, or a specific illness. The emphasis on participant “beliefs” in the various iterations of the “Theory of Planned Behavior” were central concepts in the psychology of William James and John Dewey. For example, Planned Behavior investigators have implied they had invented the concept of “action plans” but failed to reference studies that examined the interdependence of “action plans” with health warnings published over the past 55 years. The model is similar to other decision frameworks, e.g., “Health Belief Theory” that represent “beliefs” about illnesses and actions, paying little attention to the processes that generated these mental contents and their correlated responses.

In contrast, we attempted to capture the operation of the mechanisms underlying health related actions. Our early studies were published under the “self-regulation” banner, and relabeled four decades ago, as suggested by Marie Johnston, as the “Common-Sense Model (CSM)”. The CSM focuses on the somatic and functional changes that lead an individual to notice and question whether they are experiencing something new, a minor nuisance, or a major threat. Somatic changes are experienced against the background of the everyday self: “How its sensations and functions vary in real time.” The meaning/response to these variations likely depends on the organ system or functional arena, varying across systems. How much do we know about this multi-component, multi-level “entity”, our “everyday self”. I suspect not as much as we think we do. Why are some somatic functions attended to ad-nauseum in some but not all people? Does the time between the onset and response to a functional change differ by its intensity, or by area of function, e.g., change in gait, hand strength, vision, hearing? Are these differences in the occurrence and/or awareness of signals due to innate factors, and how does their action change over the lifespan?

Questions respecting individual function must be connected to contextual factors. Do the social and physical environments (climate; food availability and variety) affect these outcomes, enhancing the instability or “noise” generated by particular systems? Are these features unique to the self or are they shared and is the individual aware of the sharing? And if aware, does the individual perceive them to be shared by birth, i.e., genetically, by participation in common environments, occupations, religious affiliation, food histories, etc., or are they seen as unique? The CSM treats these factors as components of an active mechanism operating in real time recognizing how little we know about how changes in function of different systems will be experienced, understood and responded to and the antecedent and contextual physical and social factors affecting these processes.

The above questions are starting points for the CSM investigator. Each raises questions about the consistency among an individual respondent’s “priors”; are the expectations associated with a deviation in somatic experience or function similar or different across perceptual and abstract components. Discrepancies due to the multi-component nature of events, experienced and conceptualized, are readily seen with respect to the resources used for control. Our preferred “common sense” remedies for cuts and bruises, GI upset, visual disturbances, often fail to perform as expected. How do these internal inconsistencies get resolved? Do their resolutions affect how we view other treatments and their proponents, doctors, nurses, nutritionists, etc.? It may seem odd to separate the various sources of our expectancies given the rapidity with which we cycle between them. How to describe and examine these internal processes in real time and real social contexts, poses a major challenge to our skill in experimental design and modelling.