Barriers and facilitators of domain-specific physical activity: a systematic review of reviews

The health benefits of physical activity are well established [1]. However, global progress to increasing physical activity has been slow [2]. Worldwide, 27.5% of adults [3] and 81% of adolescents [4] do not meet the recommended levels of physical activity. Physical activity can be undertaken in different domains, named leisure, travel, work or education, and household [1]. These domains reflect when, where, and how physical activity is performed according to the routine of daily living, opportunities, duties, and culture. For instance, a significant fraction of the volume of moderate-to-vigorous physical activity comes from occupational and household activities (less volitional domains) [5]. In contrast, the recreational domain (most volitional domain) usually contributes the least to the total physical activity volume [5].

There is some evidence that physical activity performed in different domains may have different effects on health [6, 7]. For instance, harmful health outcomes have been associated with high levels of occupational physical activity, with an 18% increase in the risk of premature mortality compared to those in less physically demanding jobs [7]. The effects of physical activity on mental health may also vary according to domain, with one meta-analysis indicating that, compared with other domains, recreational physical activity could be more effective in preventing ill mental health than other domains [8].

A number of intrapersonal, interpersonal, environmental, cultural, socio-economic, and political factors can influence individual and population patterns of physical activity [9, 10]. However, the direction and magnitude of the relationship between these factors and physical activity level may depend on the physical activity domain (leisure, travel, work or education, and household) of interest. Mitigating barriers and strengthening facilitators in the different domains is key to enable the adoption and maintenance of a more active lifestyle for all. Therefore, evidence on domain-specific barriers and facilitators becomes central to the development of more effective physical activity promotion actions. Thus, our objective was to conduct a systematic review of reviews on barriers and facilitators of domain-specific physical activity.

We conducted a systematic review of reviews following the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines [11]. The study protocol was registered and approved in the International Prospective Register of Systematic Reviews (PROSPERO) under the code CRD42020209710.

After screening and quality assessment, 44 systematic reviews [14‐57] investigating the association between barriers and facilitators and domain-specific physical activity were included in our analysis. Table 3 presents study designs, age groups, domains of physical activity, and barriers and facilitators included by each systematic review, alongside the methodological quality rating. Thirty-three reviews included original studies with longitudinal designs (e.g., experiments, prospective cohorts). However, evidence from cross-sectional studies predominated in 26 out of 34 reviews that included this type of study design. Eleven reviews included qualitative studies and four considered mixed-methods approaches. Regarding the age groups investigated, 15 reviews included children, 22 included adolescents, 25 included adults, and 22 older adults. Leisure was the most investigated physical activity domain (n = 32), followed by travel (n = 22), work or education (n = 6), and household (n = 1).

Our systematic review of reviews provides the most comprehensive overview up to this date of the current evidence base on barriers and facilitators of domain-specific physical activity behaviour. Our findings show that the evidence base is largest for leisure-time, followed by travel-related physical activity, whereas a very limited number of reviews were dedicated to physical activity in work, educational and domestic settings. Across all physical activity domains, factors related to the built environment were more abundant in the reviews than intra and interpersonal factors, and almost no reviews investigated programmatic factors. Very consistent associations could be observed between a range of intrapersonal factors and leisure-time physical activity. Almost no reviews synthesized the association between intrapersonal factors and physical activity in the other domains. Results for social and built environmental factors were moderately consistent at the best, across all domains.

Our study has some limitations. For some barriers or facilitators, the most recent systematic review might have been conducted years ago, so our findings might not incorporate the results of the most recent original studies. Since the COVID-19 pandemic, a number of studies have been conducted to investigate changes in physical activity behaviour that occurred in this new scenario, and the factors associated with these behavioural changes. However, a rapid search and screening conducted in September 2022 found only two additional systematic reviews (out of 5153 entries) published since September 2020 that focused on barriers and facilitators of domain-specific physical activity behaviour [58, 59], whose results were largely in line with the findings of this study. Second, we used a very broad inclusion criteria to capture as much of the evidence available as possible. Consequently, the reviews included in our study vary in the inclusion criteria they applied, for instance, in terms of population groups (e.g., only men or women, specific age groups, only socially or economic disadvantaged people), locations (e.g., urban or rural, specific regions of the globe), and methodological design of the original studies. It is possible that certain barriers and facilitators are more consistently associated to domain-specific physical activity in some groups of the population or locations than others, and that combining the results of all reviews might mask these patterns. Third, the quality and interpretation of our synthesis are affected by the methodological quality of the reviews, and the original studies they included. Of the 44 reviews included in our study, only three were rated as having high methodological quality according to AMSTAR-2, with other 31 rated as having moderate quality. Moreover, we excluded 27 reviews from our synthesis because their critically low methodological quality according to AMSTAR-2. Of the 44 reviews included, thirty-four included cross-sectional studies, which was the predominant study design in 26 of these reviews. Fourth, the reviews and the original studies they included might differ in their operational definition of a same barrier or facilitator, and in how they measured domain-specific physical activity.

As expected, the evidence base is larger for the leisure domain, in all groups of factors investigated. We observed consistent evidence of association for a range of intrapersonal factors, including better skills, higher motivation and goal setting, positive beliefs about the physical activity consequences, lack of time and easy access to concurrent behaviours, negative emotions related to physical activity practice, and worse health conditions. We also found evidence of moderate consistency that general social support and support from family members are positively associated with leisure-time physical activity. Results were largely mixed for built environment factors, with results indicating either positive or no evidence of association. Because of the volitional nature of leisure activities, personal predisposition is a necessary condition to engage in recreational physical activity. Even though social and built environment factors can contribute to behavioural adoption and maintenance, their effects might not be observed if personal predisposition to engage in physical activity does not reach a certain threshold, which can explain the more mixed results observed for these two sets of factors.

Attention to the factors that enable and prevent travel-related physical activity has been increasing, but very much related to the built environment. Evidence of moderate consistency indicated a positive relationship with walkability and existence of facilities that support active travel. However, travel behaviour, including active travel, is shaped by influences at the macro, meso, and micro level [60]. The disproportional emphasis in the macro level (built environment) with limited understanding of the micro- and meso-level factors that affect travel decisions and behaviours very likely will be insufficient to design effective active travel promotion strategies. For instance, travel decisions are affected by circumstantial factors (e.g., journey purpose) that make one mode (e.g., bicycle) more or less appealing for a given journey. Also, Mattioli et al. suggest that some people might be consciously dependent of a mode (e.g., car) regardless of other circumstances (e.g., availability of other modes) [60]. Therefore, more work is needed to understand what factors, at all levels, affect active travel behaviour.

The evidence gap is even more salient for facilitators and barriers of physical activity in work, educational and domestic settings. Even though different domains of physical activity may impact health in different ways [6, 61, 62] public health messaging encourages that physical activity be incorporated throughout the day (i.e., in different domains) as part of an active lifestyle [1]. However, reductions in work and domestic physical activity have been observed and forecasted across the globe as a result of economic and social transitions [63], with likely larger impacts in low- and middle-income countries, where these domains respond for the larger fraction of physical activity volume in adults [5]. Hence, understanding what facilitates and prevents physical activity in work, educational and domestic settings is as important as in the leisure and transport domains for a successful day-long approach to physical activity promotion.

It is important to acknowledge that most of the original studies in the reviews we included were conducted in high-income Western settings, and that what facilitates or prevents domain-specific physical activity might be different in other locations, due to cultural, socio-economic, and environmental differences, for instance. Also, even though the evidence base was summarized for each factor in isolation, these factors are likely interdependent, with different combinations of factors affecting differently the capability, opportunity, and motivation [64] to be more active within and across physical activity domains. Hence, it is important that future studies on facilitators and barriers consider the broader context and underlying conditions in which specific factors seem to be more or less likely to affect physical activity behaviour.

Looking at the findings by categories of barriers and facilitators, built environment factors accounted for 447 synthesis units, equivalent to 60% of all units investigated in this study. This is more than three times the number of synthesis units in the intrapersonal (n = 142) and social environment and interpersonal (n = 136) categories. Only 25 synthesis units have been reported about programmatic factors. This imbalance in the evidence base is even more evident when we consider the domains. For instance, almost 70% of all synthesis units in the travel domain are related to built environment, more than six times the number of units for intrapersonal factors (11%). Even though there is a consensus in the field that physical activity is a multi-dimensional, multi-factorial behaviour, it is evident that there has been a disproportional emphasis in the investigation of built environment barriers and facilitators that needs to be addressed in the future.

Our study indicates a number of research opportunities, across both domains and categories of barriers and facilitators, that need to be addressed if we want to have a better understanding of the factors that affect domain-specific physical activity behaviour. First, there is a clear deficit of evidence of barriers and facilitators in work and domestic settings, which correspond to the larger portion of the moderate-to-vigorous physical activity volume [5]. Second, it seems that most of the evidence generated so far is about built environment factors, even though evidence still uncertain about a number of potential intrapersonal and interpersonal barriers and facilitators. Third, it is important to acknowledge that these barriers and facilitators, and perhaps the domains, very likely interact between themselves over the day, with different combinations of factors creating sufficient conditions to have more physically (in)active lifestyle patterns. These knowledge gaps limit a holistic understanding of the conditions that affect domain-specific physical activity behaviour and, consequently, the design of promotion strategies that can effectively incentivize and sustain more active lifestyles.

Our study provides a picture of the research conducted so far on intrapersonal, interpersonal, and built environment factors that can prevent or facilitate physical activity behaviour across domains. Even though it is accepted that knowledge of the barriers and facilitators of physical activity across domains is necessary to design promotion strategies that support more active lifestyles over the day, the evidence base is still limited and biased towards the leisure domain and built environment factors. Efforts and resources are required to diversify and strength the evidence base required to create the conditions for more physically active societies.