Implementation intentions for physical activity in supportive and unsupportive environmental conditions: An experimental examination of intention–behavior consistency

Abstract

The primary objective of this investigation was to examine the joint influences of intention strength, executive control resources (ECRs) and implementation intentions on physical activity performance in differentially supportive environmental conditions. In two studies, participants were randomly assigned to implementation intentions or control conditions. Individual differences in ECRs, intention strength and physical activity behavior were measured at baseline in the laboratory; follow-up physical activity was measured via an online questionnaire 7 days after the baseline laboratory session. Study 1 was conducted under activity-supportive environmental conditions (i.e., warm weather, little precipitation), whereas Study 2 was conducted under non-supportive environmental conditions (i.e., cold weather, moderate precipitation). In Study 1, a two-way interaction between ECR and intention strength was observed, such that those with relatively stronger ECR demonstrated stronger intention–behavior relationships than those with relatively weaker ECR; this effect was invariant across experimental conditions. In Study 2, a significant 3-way interaction was observed, such that those with lower ECRs showed greater intention–behavior correspondence in the experimental group compared to the control. Together these findings suggest that the beneficial effects of implementation intentions on intention–behavior correspondence for physical activity may be more potent under challenging environmental conditions, and that they may be of special benefit for those with initially low ECRs.

Introduction

Behavioral intention is among the most well-studied predictors of health behavior (Godin & Kok, 1996). Despite this, there have been misgivings about its role as an exclusive proximal determinant. For instance, Webb and Sheeran (2006) conducted a meta-analysis of experimental studies of intention–behavior relationships, and found that moderate-to-large changes in intention resulted in only small-to-moderate changes in subsequent behavior. Moreover, the potency of the intention effect was subject to modification based on a number of other variables, including controllability of the behavior and habit strength. Specifically, increases in behavioral intention were more likely to cause increases in subsequent behavior when the behavior had weak behavioral precedent (i.e., low habit strength) or when the behavior was under the control of the individual (Webb & Sheeran, 2006).

Although some facets of control exist on the level of perception, other facets have a neurobiological basis and may exert their effects on behavior at a level that is not always consciously appreciated by the individual. Executive control resources (ECRs), for example, are thought to arise from operation of the prefrontal cortext and associated neural systems (Miller & Cohen, 2001). Though there are several conceptualizations of the units that underlie ECRs, most would agree that they are minimally composed of inhibitory, working memory and attentional set shifting abilities (Miyake, Friedman, Emerson, Witzki, & Howerter, 2000).

To date, studies have implicated low ECRs (or suboptimal function of the brain regions that support them) in medication non-adherence (Hinkin et al., 2002), poor dietary choice (Allan et al., 2010, Hall, in press), physical inactivity/non-adherence (Hall et al., 2008, McAuley, 2011) and poor sleep hygiene (Hall, Elias, & Crossley, 2006). They have also found a place in recent theoretical models of health behavior (Hall and Fong, 2007, Hofmann et al., 2008). Although ECRs can be adversely affected by incidental factors (e.g., sleep deprivation, alcohol ingestion), individual differences in trait ECRs among cognitively intact individuals are subject to high genetic heritability (Friedman et al., 2008). As such, compensatory strategies that enhance intention–behavior consistency might be useful for those with low ECRs.

Implementation intentions are simple “if–then” plans that specify where and when an action is to be performed (Gollwitzer, 1999, Sheeran et al., 2007). These plans have been shown to be effective for facilitating a variety of behaviors, and typically outperform simply articulating the nature of the goal itself. Implementation intentions are thought to work by increasing the saliency of action cues, and therefore making people more likely to act when relevant opportunities present themselves (Webb & Sheeran, 2006).

Given their demonstrated facilitating effect for health protective behaviors (e.g., Armitage, 2004, Sheeran and Orbell, 2000), implementation intentions constitute a strategy that might serve to bolster self-regulatory success among those with low ECRs, particularly when behaviors are difficult to implement. Physical activity is one such behavior, and indeed implementation intentions have been shown to be effective interventions for facilitating physical activity among both healthy (Prestwich, Perugini, & Hurling, 2010) and clinical (Latimer, Ginis, & Arbour, 2006) populations, although there have been a few exceptions (e.g., Buddin & Sagarin, 2007). A recent meta-analysis concluded that the overall effect size of implementation intentions for increasing physical activity is .31, a small-to-medium sized effect, with greater effects documented in young adult samples and when barrier management is included in the implementation intention strategy (Blanger-Gravel, Godin, & Amireault, 2011).

Because implementation intentions are designed to enhance the individual's ability to respond to opportunities to act, and develop plans for overcoming obstacles, it is possible that one moderator of implementation intention efficacy is the ease of implementation of the behavior in question (Brandstatter et al., 2001, Dewitte et al., 2003, Gollwitzer and Brandstatter, 1997). Put differently, when behaviors must be implemented in environmental contexts that are unsupportive, implementation intentions should have stronger effects. With respect to physical activity, one very important cyclical factor that can affect ease of implementation is seasonal weather patterns, and this is particularly true in parts of the world that experience cold temperatures and snow during part of the year. This seasonal variation introduces a number of new barriers (e.g., locating indoor venues, use of specialized equipment, etc.) and intensifies the amount of planning required to enact an intention to be physically active over the course of a given week.

In the current set of studies, we randomly assigned individuals to receive implementation intentions or control conditions under activity-supportive (summer season; Study 1) and activity non-supportive (winter season; Study 2) environmental conditions, and examined intention–behavior consistency in each group as a function of ECR level. We hypothesized that low ECR participants in the physical activity implementation intention conditions would demonstrate stronger intention–behavior relationships than the no treatment or control conditions, but that the effects would be more pronounced under non-supportive environmental conditions (i.e., winter season) than supportive conditions (i.e., summer season).