Commentary: The Intention–Behavior Gap in Physical Activity

When considering motivation toward physical activity, there may be nothing more fundamental than the mismatch between good intentions and the struggle to achieve desired levels of physical activity. The article by Rhodes and Brennan (1) is a timely and important contribution to a better understanding of what has become known as the “intention–behavior gap” in physical activity. Indeed, psychological and other behavioral scientists have grappled with this issue for decades, probably stemming from the inclusion of intention as a fundamental construct in many theories. As stated by Rhodes and Brennan (1), there are at least nine theories evident in the physical activity literature that include some notion of intention. Although the most well-known is probably the theory of planned behavior (because of its key link between intention and behavior) (2), others include measures of intention within the broader social cognitive, stage-based, dual-process, and action control approaches. Intention is thought to be the central proximal predictor of behavior, although it can be reflected as a decision or assessed for its strength of commitment (3). The proximal nature of the variable means that many other factors, such as attitudes and perceptions of control and confidence, will affect behavior through their influence on intentions. But, as shown in the Rhodes and Brennan article (1), it is not as simple as this. In earlier work, it was identified that having stable intentions, being conscientious, and having feelings of “anticipated regret” (at the possibility of not being physically active) will all strengthen the link between intentions and behavior (4). But, as shown by Rhodes and Brennan (1), this list of potential moderators could be expanded. Importantly, they have categorized moderators into eight possible clusters: demographic, personality, physical capability, environmental opportunity, social opportunity, and variables labeled “regulatory,” “reflective,” and “reflexive.” Figure 3 in the Rhodes and Brennan article (1)—an excellent visual summary in its own right—shows that many of these categories, although logically linked to the intention–behavior gap, currently lack good evidence. The regulatory, reflective, and reflexive variables are particularly interesting and worthy of further discussion. They also mirror some important trends in research on determinants of physical activity—a longstanding field of study (5,6). Researchers have been stating for over 30 years that only about 35% of the variance in behavior was explained by commonly studied social cognitive variables, including intentions: “Social cognitive models operate mainly in terms of predisposition and for this reason cannot explain more than a part of the variance in exercise behavior” (7). This suggests further approaches are needed without abandoning longstanding variables. Regulatory processes, as described by Rhodes and Brennan (1), “represent the behavioral, cognitive, and affective tactics used to enact intentions…They are generally considered the hallmark of most theories that have focused on the translation of intention into behavior.” Interestingly, what have been thought to be practical strategies to close the intention–behavior gap, such as implementing intention action plans (e.g., “if-then” plans), have only mixed evidence for effectiveness and hence are labeled as “indeterminate” in Rhodes and Brennan’s Figure 3 (1). There is also a longer list of variables for which there is no association (e.g., mindfulness, fatigue, and ego depletion). The category of reflective variables represents “consciously deliberated expectations of behavioral enactment” (1). These tend to reflect typical social cognitive variables that have been studied for many years in exercise and health psychology, including goals, attitudes, motivation, and confidence. Rhodes and Brennan (1) suggest that evidence for such variables in closing the intention–behavior gap is strongest for intention stability and commitment, affective attitude, and perceived behavioral control/self-efficacy. The reason that affective attitude is included here and not in the reflexive category is that it involves expectations of affect, such as pleasure, thus involving more deliberative (cognitive) processes. It is concluded that anticipated enjoyment and pleasure are likely to play key roles in the intention–behavior gap. These might need greater emphasis in future interventions. Moreover, we need greater clarity on the nature and measurement of enjoyment. Currently, it is a construct used rather loosely in the extant literature (6). Regulatory and reflective processes tend to use relatively slow cognitive processing, such as instrumental judgments and weighing up of pros and cons of behavioral decision making. These have been the cornerstone of theoretical work in exercise and health psychology for decades. However, although they still have utility, it is questionable how much people are willing and able to undergo such cognitive effort to change their behavior, especially for effortful behavior such as physical activity. This has led researchers to increase their focus on less consciously driven processes, such as affect, and this is included in Rhodes and Brennan’s analysis (1) as “reflexive processes.” Reflexive processes include affective reactions, habit, and identity. This is partly akin to the automatic motivation alluded to in the popular COM-B (Capability, Opportunity, and Motivation Behavior model) element of the Behaviour Change Wheel framework (8). As stated by Rhodes and Brennan (1), “reflexive motivational processes (i.e., learned associations, cue-driven behavior, and motivation with low awareness) had been traditionally overlooked in understanding [physical activity]…yet are often considered critical drivers of [physical activity] maintenance.” To date, evidence suggests that the intention–behavior gap in physical activity is associated with identity/schema, but not in-task or post-task affect. Evidence on habit appears to be indeterminate. It is recommended that greater attention be paid to reflexive processes. Recent automatic and dual-process models, such as the affective-reflective theory (9), theory of effort minimisation (10), and multi-process action control framework (11), all require greater scrutiny and application to intention–behavior linkages across health behaviors (12). The closing of the intention–behavior gap in physical activity research remains a critical endeavor. I concur with Rhodes and Brennan (1) that the intention–physical activity gap “remains a central challenge...the importance of identifying what predicts the translation of intention into [physical activity] is paramount for the advancement of both theory and practice.” Identifying important categories of variables and processes is an excellent step forward, but many unanswered questions remain. ACKNOWLEDGMENTS The publication of this article does not constitute endorsement by the American College of Sports Medicine. CONFLICTS OF INTEREST AND SOURCE OF FUNDING S.J.H.B has current research funding from the Australian Research Council, but has no other conflicts of interest or funding to report related to this article.