Health Behavior

Abstract

Health behavior is a type of behavior: the actions a person takes that bear on their health, whether to protect it or to place it at risk. The construct was given its modern shape by Kasl and Cobb, who distinguished the conduct of a well person acting to stay healthy from that of a person who feels ill, and it now anchors a large body of social-cognitive theory explaining why people so often fail to act in their own long-term interest. This article develops it across levels: how it is defined and distinguished from illness and sick-role behavior, the subtypes the Medical Subject Headings vocabulary files beneath it, the expectancy-value and self-regulatory theories that predict it, the gap between intention and action, and the framework-based approach to changing it.

Keywords: health behavior, health belief model, theory of planned behavior, intention–behavior gap, behavior change

A person resolves each January to exercise, knows the evidence for it, intends it sincerely, and by February has stopped. Another keeps smoking for decades against full knowledge of the harm, then quits abruptly after a single conversation with a physician. The actions that most shape how long and how well a person lives—what they eat, whether they move, whether they smoke, whether they attend a screening—are not reliably governed by what they know or even by what they intend. Health behavior is the term the medical and psychological literature uses for this class of action, and the puzzle of why it resists both knowledge and good intention is one of the organizing problems of health psychology (Kasl & Cobb, 1966).

The stakes are large and increasingly behavioral. As infectious disease gave way to chronic disease as the main cause of death in industrialized countries, the principal determinants of mortality shifted from pathogens to patterns of conduct—what Matarazzo called behavioral pathogens, the habits that cause disease, set against the behavioral immunogens that protect against it (Matarazzo, 1980). This article treats health behavior as a problem in the psychology of action: how it is defined, what forms it takes, what theories best predict it, why prediction so often outruns control, and what the evidence says about changing it.

What Health Behavior Is

Health behavior is any activity a person undertakes that affects their physical health, whether or not that is its purpose, ranging from deliberate protective acts such as exercising, eating well, and attending a screening to risk-taking acts such as smoking or excessive drinking. The foundational distinction is due to Kasl and Cobb, who separated three kinds of health-related action: health behavior, the activity of a person who believes themselves well, undertaken to prevent disease or detect it while still asymptomatic; illness behavior, the activity of a person who feels ill, directed at discovering what is wrong and finding relief; and sick-role behavior, the activity of a person already defined as ill, directed at getting well (Kasl & Cobb, 1966). In the Medical Subject Headings vocabulary, Health Behavior is the descriptor that names the first of these, the conduct of the ostensibly well person as it bears on health.

The distinction matters because the three kinds of action answer to different psychologies. Health behavior proper is undertaken in the absence of symptoms, which is precisely what makes it hard: its costs are immediate and certain while its benefits are distant and probabilistic, so it depends on belief about a future that has not yet arrived rather than on the pressing feedback of pain. This is why a person can know that exercise prevents disease and still not exercise, and why the field that studies health behavior is a branch of the psychology of motivation and self-regulation rather than of medicine as such (Matarazzo, 1980). The modern construct also widens Kasl and Cobb's focus from discrete acts to the habitual patterns—diet, activity, substance use—that accumulate over a lifetime into the behavioral determinants of chronic disease.

Figure 1

Kasl and Cobb's Three Kinds of Health-Related Behavior

Types of Health Behavior

The clinical description above cuts the construct by the person's state. A second cut comes from the classification itself: the Medical Subject Headings vocabulary places Health Behavior at tree position F01.145.488, beneath the broader descriptor Behavior, and hangs seven narrower descriptors below it. These subtypes are not mutually exclusive—a person who enters a smoking-cessation program is engaged in treatment adherence, smoking reduction, and smoking cessation at once, and the categories overlap freely—so the list is a taxonomy of indexing, not a partition of conduct. It is worth stressing that MeSH is an indexing classification built to retrieve the literature, not a theory of how health-relevant action carves into natural kinds; the subtypes are how the field files its papers, not a claim about joints in nature. Table 1 lists the direct children of the descriptor; several are treated at length in their own articles.

The classification shows at a glance that health behavior spans both the protective and the risky: self-examination and sleep hygiene are things a well person does to stay well, while the smoking cluster names the single most consequential behavioral pathogen and the long effort to undo it. The same theoretical machinery—belief about threat, expectation of outcome, confidence in one's own capability—is brought to bear on all of them.

Theories of Health Behavior

The intellectual core of the field is a family of expectancy-value theories, so called because each predicts action from the value a person places on an outcome multiplied by their expectation of attaining it. The earliest was the health belief model, developed by Irwin Rosenstock and colleagues in the 1950s to explain why people failed to take up free preventive services, which holds that a person acts when they believe they are susceptible to a condition, that the condition is severe, that action would bring benefits, and that the benefits outweigh the barriers (Rosenstock, 1974). Rogers extended the threat side of this picture into protection motivation theory, which separates a person's appraisal of the threat from their appraisal of the coping response, and showed that fear appeals change behavior only when the person also believes an effective response is available to them (Rogers, 1975).

The most influential model of the deliberative route to action is Ajzen's theory of planned behavior, which predicts behavior from the intention to perform it, and predicts intention in turn from three determinants: the person's attitude toward the behavior, the subjective norm they perceive from others, and their perceived behavioral control—how easy or hard they expect the action to be (Ajzen, 1991). Perceived behavioral control is close kin to the construct that runs through Bandura's social cognitive theory, self-efficacy: the belief that one can organize and execute the action required. Bandura argued that self-efficacy governs not only whether a person forms a health goal but how much effort they invest and how they respond to setbacks, making it a determinant of both the adoption and the maintenance of health behavior (Bandura, 1998). Across all of these models the recurring lesson is that information alone does not move behavior; belief about personal risk, about the value of acting, and above all about one's own capability does.

The Intention–Behavior Gap

The deliberative models share a common pivot: intention is treated as the proximal cause of behavior, the point at which belief becomes action. Yet the single most robust finding in the field is that intention predicts behavior only imperfectly. A conceptual and empirical review by Sheeran found that intentions account for a substantial but far from complete share of the variance in later behavior, and that the shortfall is systematic rather than random (Sheeran, 2002). The asymmetry is revealing: people who do not intend to act almost never do, but a large fraction of those who sincerely intend to act nonetheless fail to—the inclined abstainers. This pattern is what the field calls the intention–behavior gap (Sheeran & Webb, 2016).

The gap matters because it relocates the problem of health behavior. If intention translated reliably into action, the task would be simply to persuade—to change attitudes and beliefs until intention formed. Because it does not, a second, volitional phase of self-regulation becomes the real battleground: forming concrete plans, managing competing goals, coping with cues and cravings, and recovering from lapses. Interventions that target this phase, such as prompting a person to specify exactly when, where, and how they will act, close part of the gap precisely because they operate after intention is already in place (Sheeran & Webb, 2016). The gap also explains why maintenance is a distinct problem from initiation: a systematic review of behavior-change theories found that keeping a new behavior going draws on partly different mechanisms—habit, self-regulatory capacity, and a supportive environment—from those that start it (Kwasnicka et al., 2016).

Changing Health Behavior

The accumulation of competing theories created its own problem: an intervention designer faced dozens of overlapping models with no principled way to choose among them. Two responses organized the field. The first was the transtheoretical model of Prochaska and DiClemente, which proposed that change proceeds through identifiable stages—precontemplation, contemplation, preparation, action, and maintenance—and that an intervention should be matched to the person's current stage rather than applied uniformly (Prochaska & DiClemente, 1983). The second was an insistence that interventions be built from explicit theory in the first place, since a review of public-health practice found that theory-based interventions consistently outperformed those designed by intuition (Glanz & Bishop, 2010).

The most systematic synthesis is the COM-B model and its Behaviour Change Wheel, developed by Susan Michie and colleagues, which holds that any behavior is the product of three necessary conditions—Capability (the physical and psychological capacity to act), Opportunity (the physical and social affordances that permit it), and Motivation (the reflective and automatic processes that energize it) (Michie et al., 2011). The model's practical force is diagnostic: before choosing an intervention, the designer asks which of the three is the binding constraint, because an intervention aimed at motivation will fail if the real obstacle is a lack of capability or opportunity. This reframing absorbs the older theories—the belief models populate motivation, self-efficacy sits in psychological capability—and explains why single-target interventions so often disappoint (Rhodes et al., 2019). It also makes room for the determinants that individual psychology leaves out: a large share of the variation in health behavior tracks social and economic position, so that changing behavior at the population level is as much a matter of altering opportunity as of changing minds (Short & Mollborn, 2015; Kelly & Barker, 2016).

Worked Example

The theory of planned behavior makes its prediction concrete by combining its three determinants into an intention score. In its measurement form, intention is modeled as a weighted sum of attitude, subjective norm, and perceived behavioral control, with the weights estimated from data and reflecting how much each determinant matters for the particular behavior and population (Ajzen, 1991). Each determinant is rated on a scale—here 1 to 7—and the predicted intention falls on the same scale.

Consider a person contemplating regular exercise, with the three determinants measured as attitude A = 6 (they view exercise very favorably), subjective norm SN = 4 (they feel only moderate social pressure), and perceived behavioral control PBC = 5 (they are fairly confident they can do it). Suppose the weights estimated for this behavior are β_A = 0.45, β_SN = 0.20, and β_PBC = 0.30. The predicted intention is the weighted sum:

Intention = (0.45 × 6) + (0.20 × 4) + (0.30 × 5) = 2.70 + 0.80 + 1.50 = 5.00 on the 1–7 scale.

An intention of 5.00 is well above the midpoint, so the model predicts this person will form a reasonably strong intention to exercise. The computation also shows where the leverage lies: because attitude carries the largest weight, the favorable attitude is doing most of the work, while the moderate subjective norm contributes least. Yet the intention–behavior gap enters here too. Empirically, intention correlates with later behavior at roughly r = 0.47, which means intention explains about r² ≈ 0.22, or 22%, of the variance in whether the person actually exercises (Sheeran, 2002). A strong intention of 5.00 therefore tilts the odds toward action without guaranteeing it—which is exactly why the volitional strategies of the previous section are needed to convert the remaining intenders.

Discussion

The study of health behavior has moved, over half a century, from cataloguing acts to modeling the psychology that produces them, and from a confident expectancy-value optimism to a more sober appreciation of its limits. The expectancy-value models were a genuine advance: they showed that health behavior is lawful, predictable from a person's beliefs about threat and capability rather than mysterious, and they gave self-efficacy its central and durable place. But their common reliance on intention as the final cause proved to be their weak point. The discovery that intention explains only a fifth or so of the variance in behavior was not a refinement of the models but a challenge to their architecture, forcing the field to take seriously the volitional processes that operate after a goal is set (Sheeran, 2002; Sheeran & Webb, 2016).

The second major correction was the recognition that individual cognition is only part of the story. The COM-B synthesis made explicit what the belief models had backgrounded: that capability and opportunity are preconditions for action no amount of motivation can replace, and that opportunity in particular is distributed by social and economic structure rather than chosen (Michie et al., 2011; Short & Mollborn, 2015). This is why the most knowledge-dependent interventions—telling people what is good for them—are also the least effective, and why the behaviors most resistant to change are concentrated where opportunity is scarcest (Kelly & Barker, 2016). The field's current center of gravity is therefore a layered one: individual self-regulation nested within, and constrained by, the affordances of the environment.

Current Directions

Contemporary work has largely abandoned the search for a single best theory in favor of two complementary programs. The first is integrative and diagnostic: rather than pitting the health belief model against the theory of planned behavior, researchers treat their constructs as a shared pool and ask which combination predicts a given behavior in a given population, an approach exemplified by the COM-B framework and by syntheses of the physical-activity literature that map the overlapping theories onto a common structure (Rhodes et al., 2019). The second program focuses on maintenance as a problem in its own right, recognizing that the mechanisms that start a behavior differ from those that sustain it; a systematic review identified distinct maintenance theories built around habit, self-regulatory resources, identity, and environmental support, and argued that interventions should be designed for the long run rather than assuming that initiation will persist (Kwasnicka et al., 2016). Running through both programs is a growing attention to the automatic, non-reflective determinants of behavior—cues, habits, and context—and to the social gradient that places much of the variation in health behavior outside the reach of persuasion (Short & Mollborn, 2015).

Common Misconceptions

People behave unhealthily because they lack information.

Knowledge is necessary but rarely sufficient: smokers know the risks, and most people who fail to exercise can recite its benefits. The determinants that actually move behavior are beliefs about personal susceptibility and, above all, about one's own capability to act, so information campaigns alone change little (Rosenstock, 1974; Bandura, 1998).

Forming a firm intention is the hard part, after which action follows.

The intention–behavior gap shows the opposite pattern: intention is relatively easy to produce and predicts behavior only weakly, so a large fraction of sincere intenders never act. The difficult, under-appreciated work lies in the volitional phase of translating intention into sustained behavior (Sheeran & Webb, 2016).

Health behavior is purely a matter of individual choice.

Much of the variation in whether people smoke, exercise, or eat well tracks social and economic position, because opportunity—the time, money, safety, and social support an action requires—is unequally distributed. Treating health behavior as choice alone misattributes a structural problem to personal failing (Short & Mollborn, 2015; Kelly & Barker, 2016).

Glossary

Attitude (toward a behavior).

In the theory of planned behavior, a person's overall favorable or unfavorable evaluation of performing the behavior; one of the three determinants of intention.

Behavior.

The broad MeSH parent category of observable activity under which health behavior is classified.

Behavioral pathogen.

Matarazzo's term for a personal habit that causes or promotes disease, such as smoking or physical inactivity; its protective opposite is a behavioral immunogen.

COM-B model.

The account that behavior arises from the interaction of Capability, Opportunity, and Motivation; the hub of the Behaviour Change Wheel.

Expectancy-value theory.

A family of models that predict action from the value a person places on an outcome multiplied by their expectation of attaining it.

Health behavior.

Action undertaken by a person who believes themselves well, directed at preventing disease or detecting it while asymptomatic; more broadly, any conduct affecting health.

Health belief model.

Rosenstock's expectancy-value model predicting health action from perceived susceptibility, severity, benefits, and barriers.

Illness behavior.

Kasl and Cobb's category for the activity of a person who feels ill, directed at discovering what is wrong and obtaining relief.

Intention.

A person's self-instruction to perform a behavior; the proximal predictor of action in the deliberative models and the quantity measured in the theory of planned behavior.

Intention–behavior gap.

The robust finding that intention predicts behavior only imperfectly, so that many sincere intenders fail to act (the inclined abstainers).

Perceived behavioral control.

In the theory of planned behavior, a person's belief about how easy or difficult performing the behavior would be; closely related to self-efficacy.

Protection motivation theory.

Rogers's model separating appraisal of a health threat from appraisal of the coping response, explaining when fear appeals change behavior.

Self-efficacy.

Bandura's construct for a person's belief in their capability to organize and execute an action; among the most consistent predictors of health behavior.

Sick-role behavior.

Kasl and Cobb's category for the activity of a person already defined as ill, directed at getting well, including treatment adherence.

Subjective norm.

In the theory of planned behavior, the perceived social pressure to perform or not perform the behavior; one of the three determinants of intention.

Transtheoretical model.

Prochaska and DiClemente's stage model of change (precontemplation, contemplation, preparation, action, maintenance), matching interventions to a person's readiness.

Key Researchers

Icek Ajzen

(University of Massachusetts Amherst, Professor Emeritus of Psychology). Social psychologist who, with Martin Fishbein, developed the theory of reasoned action and then the theory of planned behavior, the most widely applied model of deliberative health behavior. ORCID

Albert Bandura

(Stanford University; 1925–2021). Originator of social cognitive theory and the construct of self-efficacy, which became the single most consistent psychological predictor of whether people adopt and maintain health behaviors. Wikipedia

Susan Michie

(University College London, Professor of Health Psychology and Director of the Centre for Behaviour Change). Developed the COM-B model and the Behaviour Change Wheel, which reorganized a fragmented field of theories into a single diagnostic framework for designing interventions. ORCID

James O. Prochaska

(University of Rhode Island; 1942–2023). With Carlo DiClemente, created the transtheoretical model and its stages of change, shifting intervention design toward matching strategy to a person's readiness to change. ORCID

Irwin M. Rosenstock

(University of Michigan; 1925–1998). Co-developer of the health belief model, the first systematic psychological theory of why people do or do not take preventive health action. Obituary

Paschal Sheeran

(University of North Carolina at Chapel Hill, Professor of Psychology and Neuroscience). Established the empirical scale of the intention–behavior gap and the volitional strategies, such as implementation intentions, that help close it. ORCID

Frequently Asked Questions

What is health behavior?

It is any action a person takes that affects their physical health—protective acts such as exercise, healthy eating, and screening, and risk acts such as smoking. In its original and narrower sense it is the activity of a person who believes themselves well, undertaken to prevent disease or detect it while still asymptomatic (Kasl & Cobb, 1966).

How does health behavior differ from illness behavior?

Kasl and Cobb distinguished three kinds by the person's state: health behavior is what a well person does to stay well, illness behavior is what a person who feels ill does to find out what is wrong, and sick-role behavior is what someone already diagnosed does to get well. The same act can fall into any of the three depending on why it is done (Kasl & Cobb, 1966).

What are the main theories of health behavior?

The dominant ones are expectancy-value models: the health belief model, protection motivation theory, the theory of planned behavior, and social cognitive theory. Each predicts action from beliefs about threat, the value of acting, and—crucially—the person's confidence in their own capability (Rosenstock, 1974; Ajzen, 1991).

Why do people fail to act on health advice they believe?

Because intention predicts behavior only weakly. The intention–behavior gap is the finding that a large share of people who sincerely intend to act never do, so the hard work lies in the volitional phase of translating intention into sustained action rather than in forming the intention (Sheeran & Webb, 2016).

What is self-efficacy and why does it matter?

Self-efficacy is a person's belief that they can perform the action in question. Bandura showed it governs whether a person sets a health goal, how much effort they invest, and how they respond to setbacks, which makes it one of the most consistent predictors of both intention and behavior (Bandura, 1998).

What is the COM-B model?

It is the account that behavior requires Capability, Opportunity, and Motivation together, developed by Susan Michie and colleagues as the hub of the Behaviour Change Wheel. Its value is diagnostic: it directs an intervention at whichever of the three is the real obstacle rather than assuming the problem is motivation (Michie et al., 2011).

Does knowing the risks change behavior?

Rarely on its own. Information is necessary but seldom sufficient—smokers know the risks and most sedentary people know the benefits of exercise. What moves behavior is belief about personal susceptibility and capability, together with the opportunity to act (Rosenstock, 1974; Kelly & Barker, 2016).

Is health behavior just a matter of individual choice?

No. Much of the variation in whether people smoke, exercise, or eat well follows social and economic position, because the time, money, safety, and support an action requires are unequally distributed. Changing behavior at the population level is as much about altering opportunity as about changing minds (Short & Mollborn, 2015).

References