Background

Currently, physical inactivity (PIA) and sedentary behavior (SB) are considered global health challenges hampering the achievement of the United Nations' (UN) third sustainable development goal (SDG). PIA and SB are responsible for 1.6 million deaths per year (27% due to diabetes and 20% due to cardiovascular disease [CVD]) [1]. More than 31% of premature deaths attributed to non-communicable diseases (NCDs) occur in physically inactive populations and are responsible for US $54 billion per year of direct care costs and US $14 billion per year of indirect costs (i.e., a loss of productivity) [1].

It is important to differentiate between three unique concepts: physical activity (PA), PIA, and SB. The WHO defines PA as “any bodily movement produced by skeletal muscles that requires energy expenditure.” The WHO defines PIA as any activity below the threshold of 150 min per week of moderate or vigorous PA. SB is defined as any waking behavior that leads a person to consume 1.5 metabolic equivalents or less (e.g., sitting, reclining, or lying down) [2].

A recent meta-analysis revealed that prolonged SB is associated with an elevated risk of morbidity and mortality from NCDs. This risk can be reduced or even eliminated by engaging in PA. However, if SB is very high (SB time exceeding 7 h) the risk of mortality and morbidity from NCDs is independent of the level of PA [3]. Both PIA and SB carry a high risk of developing an NCD. PIA is a major risk factor for CVD [4], type 2 diabetes [5], high blood pressure [6], cancer [7] and drug use [8]. However, SB is associated with a 30% increase in CVD [9] as well as a 55% increase in the risk of endometrial cancer [10] and elevated blood pressure [11]. These risks are exacerbated when combined with insufficient PA [12]. Thus, interventions aimed at reducing PIA and SB are estimated to reduce the risk of CVD, type 2 diabetes, depression, and cancer by 35%, 40%, and 35%, respectively [1].

In recent years, increased attention has been given to designing combined interventions, targeting both PIA and SB, to appropriately prevent and contribute to the management of NCDs for better health and well-being outcomes [13]. These interventions need to involve behavioral changes and to be informed by behavioral theories according to the WHO and other global health institutions, communities of researchers, practitioners, and policymakers [14,15,16,17].

Behavioral theories and Behavior Change Techniques (BCT)

Behavioral theories explain why, when, and how an individual behavior does (or does not) occur. They highlight that the mechanism of change at play, if targeted, will alter the behavior at the individual, interpersonal, or community level. These mechanisms are central to the design of theory-informed behavior change interventions (BCI) [19], which are complex social adaptive systems (e.g., multiple health behavioral change interventions (BCIs) targeting simultaneously or sequentially two or more health behaviors, that comprise interacting components and sensitivity to context, with emergent intended and unintended effects at different levels: the individual, interpersonal, community (organizational, environmental, national, and global) levels [20,21,22,23].

According to Hayden [24], behavioral theories can be classified into three categories based on their levels of action: 1) Intrapersonal or individual-level theories focus on personal determinants that influence behavior (e.g., knowledge, attitudes, beliefs, and motivation). Examples include the health belief model (HBM) (Hoch, Baum 1958; [25], the theory of planned behavior (TPB) [26], and self-determination theory (SDT) [27]. 2) Interpersonal level theories highlight the influence of others in shaping one’s behavior; social cognitive theory (SCT) [28] is the most commonly used interpersonal-level theory. 3) Community-level theories aim to affect or modify the social systems within which actors interact. These social systems include organizations institutions, and public policies, among others. Examples of community-level theories include diffusion of innovation theory (Valente & Rogers, 1995) [29] and the social ecological model (SEM) [30].

In practice, behavioral theories are translated into BCIs; these are implemented through the use of BCTs, which are interactive, reproducible elements of an intervention that facilitate the alteration of the mechanism of change or the causal pathway toward the intended behavioral outcome [31, 32].

Recent research has urged scholars to place more emphasis on understanding how and in which context a BCI addressing PIA or SB will lead to desired or unexpected outcomes and impacts [33]. However, the answer remains elusive. To close this gap, we aimed to map out the different types of BCIs geared toward PIA and SB and their underlying theories and techniques. We focused on mapping out different interventions to reduce PIA and SB and identified the underlying behavioral theories and BCTs used. We also aimed to assess the extent of behavioral theories use in the design of BCIs. Our review will provide decision-makers and behavioral designers with a unique systematic and comprehensive mapping of BCI targeting PA and SB using behavioral change theories, tools, and techniques.

Methods

We adopted the scoping review methodology as defined by Arksey and O’Malley [34] and refined by the Joanna Briggs Institute (JBI) [35].

Specifying the review question

During different research team meetings, we iteratively refined our review question as follows: What are the different behavioral theories and BCTs used in theory-informed interventions focused on PIA and SB? To construct a suitable search strategy, we employed the health behavior, health context, exclusion, models, and theories (BeHEMoth) framework [36, 37] (see Table 1), which is especially relevant for identifying interventions based on behavioral theories. We then followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) reporting guidelines to report the results of our scoping review [38].

Table 1 The BeHEMoth framework
Full size table

We included only interventions addressing PIA and SB or both. We excluded interventions adressing other health behaviors such as,nutrition, smoking, and sleep (see Table 2).

Table 2 Inclusion criteria for the population, concept, and context models
Full size table

Search strategy

We searched four databases (Scopus, Web of Science [WoS], PubMed, and Google Scholar) (see Supplementary file 1 and Table 2). We manually searched for gray literature on institutional sites and used reference tracking to identify additional papers. We combined search terms for theories (“Logic model” OR “Theory of change” OR “Outcome of change” OR “Program* theory” OR "Program*logic" OR “Logical framework” AND “Behavioral change intervention”) with search terms addressing BCIs: “Behavioral change interventions” AND keywords for “physical activity” OR “sedentary” OR “physical inactivity” OR “exercise” OR “fitness.”

Study selection

The study selection was carried out by two researchers, HK and ZB. We included only empirical studies of interventions addressing SB, PIA, or PA that explicitly used behavioral theories in the context of healthcare. Table 2 guided the definition of our inclusion criteria using the PCC (population, concept, context) framework (JBI) [35]. We included papers published in French, English, and Arabic between January 2010 and November 2023. All study designs were included. We excluded reviews, study protocols, feasibility studies, books, book chapters, commentaries, and letters to editors (See supplementary file 2).

Data charting

Data extraction was guided by, and adapted from the Cochrane Handbook of Systematic Review of Interventions for describing the characteristics of interventions [39] (see Table 3). We first extracted data about the general characteristics of the included studies (author, year, country, type of article, study population). Then, we extracted data about the following characteristics of behavioral theories -informed interventions: 1) theories, models or conceptual frameworks; 2) types of interventions; 3) behavioral theories; 4) BCTs; 5) targeted behavior (SB, PIA, PA, or both); and 6) level of intervention (individual, interpersonal, and environmental) (see supplementary file 3).

Table 3 Data charting form adapted from Higgins et al. (2019)
Full size table

Data analysis, coding and synthesis

BCI

As we aimed to identify the underlying behavioral theories and BCTs used to inform the design of BCIs, we employed a BCI upper-level ontology [40] that coded different forms of BCIs. This taxonomy provides a helpful model for systematically and uniformly describing the upper-level components of BCIs; this enabled us to describe BCIs based on theory and to create a map of the different contexts, BCI content, mode of delivery, and BCI outcomes (see Table 4 and supplementary file 4).

Table 4 Thematic analysis, coding using different taxonomies of BCT interventions, theories, and techniques
Full size table

Mode of delivery

We coded the different modes of delivery using the taxonomy developed by [41].

Behavioral theories

To comprehensively describe the theories used to inform the design of interventions, we used the taxonomy of behavioral theories developed by Michie [19] and we refined it based on Hayden [24]. This taxonomy outlines key behavioral theory constructs (definitions, interest, use, the context of theory development).

We further assessed the intensity and degree of theory use in BCIs (an analysis of how interventions have actually been implemented according to the stated theory) as developed by Michie, 2010 [42] and refined by Bluethmann, 2017 [43] to fit the context of PA. This taxonomy included the following criteria: 1) a theory was mentioned, 2) relevant constructs were targeted, 3) each intervention technique was explicitly linked to at least one theoretical construct, 4) participants were selected or screened based on prespecified criteria (e.g., a construct or predictor), 5) interventions were tailored to different subgroups, 6) at least one construct or theory mentioned in relation to the intervention was measured post-intervention, 7) all measures of theory were presented with some evidence of their reliability, and 8) the results were discussed in relation to the theory.

The most prevalent theories are the transtheoretical model (TTM) of change [44], the TPB [26], SCT [28], information motivation behavior (IMB) [45], the HBM [46], SDT [27], and the health action process approach (HAPA) [19, 47].

Behavioral change techniques (BCTs)

We finally coded the BCTs using the V1 taxonomy [31]. The taxonomy of BCTs synthesizes 93 BCTs classified into 16 domains: 1) goals and planning, 2) feedback and monitoring, 3) social support, 4) shaping knowledge, 5) natural consequences, 6) comparison of behavior, 7) associations, 8) repetition and substitution, 9) comparison of outcomes, 10) rewards and threats, 11) regulation, 12) antecedents, 13) identity, 14) scheduled consequences, 15) self-belief, and 16) covert learning.

Results

Search results

As indicated in Fig. 1, we identified a total of 1,173 studies during systematic searches in four electronic databases. After removing duplicates (n = 165), we screened 1,027 articles for eligibility. We excluded 945 studies during the title and abstract screening. We extracted and analyzed 82 full-text studies for eligibility and excluded 53 (see the reasons for exclusion in Fig. 1 and Supplementary File 2). We screened the reference lists of the included studies for additional relevant articles (n = 19). We finally included a total of 29 articles.

Fig. 1
figure 1

PRISMA flowchart

Full size image

In the following paragraphs, we describe the general characteristics of the included studies, the features of theory-informed BCIs (the intervention model, behavioral theories, and BCTs), and the extent of theory use in the included studies.

General characteristics of the included studies

Most of the included studies were carried out in high-income countries (n = 23): the US (n = 5) [48,49,50,51,52], the UK (n = 5) [53,54,55,56], Australia (n = 3) [57,58,59], Belgium (n = 3) [60,61,62], the Netherlands (n = 2) [63, 64], Canada (n = 2) [65, 66], Jordan (n = 2) [67, 68], Iran (n = 2) [69, 70], Italy (n = 1) [71], Qatar (n = 1) [72], Portugal (n = 1) [73], Spain (n = 1) [74], and Germany (n = 1) [75].

Intervention duration

The duration of the BCIs varied from six weeks to three years. Most interventions were carried out in a short period, ranging from one to four months (n = 14); others lasted five to six months (n = 7). Only six interventions lasted over twelve months (n = 6) (see Table 5).

Table 5 Generals characteristics of included studies
Full size table

Study design, context, and participants

All studies used experimental designs, including randomized controlled trials (n = 13), cluster randomized trials (n = 6), and multisite RCTs (n = 2) and quasi-experimental studies (n = 8). These studies took place in diverse settings and targeted various populations (see Table 5).

Nine studies were conducted in the workplace [50, 56,57,58, 61, 63, 70, 72, 75]. Seven studies reported interventions for people with chronic illnesses, diabetes (n = 3) [65, 69, 71], obesity (n = 1) [48], cardiovascular disease (n = 1) [68], and Parkinson’s disease (n = 1) [64] as well as for survivors of breast cancer (n = 1) [59]. Other studies included different groups such as older adults (n = 4) [52, 55, 66, 76], healthy adults (n = 2) (53. 62), university students (n = 2) [67, 74], and preschool children (n = 2) [51, 54] (see Table 5).

Description of theory based BCI

In our scoping review, we identified 29 articles describing interventions informed by behavioral theories targeting SB and PIA. Among these, fifteen articles aimed to address PIA to meet guideline recommendations, while eleven focused on reducing SB. Three articles combined interventions to reduce SB and increase PA (see Table 6).

Table 6 Description of interventions according to the behavioral change intervention ontology
Full size table

In the following, we will describe the content of BCIs, levels of interventions, mode of delivery and reported outcomes (see Table 6).

Content of BCIs

Most BCI interventions adopted educational methods (n = 20) aimed at raising awareness of the importance of meeting PA recommendations and breaking the vicious cycle of SB [18, 48, 51, 53,54,55,56, 59, 62, 64,65,66,67, 69,70,71, 73, 75, 76, 79]. These interventions also included communication strategies (n = 14): motivational interviews (n = 4) [68, 75, 76, 80], and coaching (n = 10) (face-to-face consultations or phone calls) [18, 53, 57, 59, 62, 64, 65, 67, 71, 73]. Social support to implement interventions was used nine times [51, 52, 56, 57, 62, 66, 69, 76, 79], and physical exercise training was used 8 times [52, 54, 64, 66, 70, 71, 73,74,75,76, 80]. Finally, digital interventions (devices, desktops, m-health) were used in most interventions (n = 16)0.2

Levels of interventions

The majority of interventions involved individual-level BCIs (n = 15). Few studies combined the individual level of the interpersonal level (e.g., peer support) (n = 6) [52, 56, 62, 66, 69, 76], and six studies combined the individual level with organizational-level interventions (n = 6) [50, 51, 54, 63, 71, 72]. Only two studies can be described as systemic BCIs addressing the individual, interpersonal, and organizational levels (n = 2) [57, 79] (see Table 6).

Heterogeneity of modes of delivery

The modes of delivery of BCIs were often mixed. BCIs included face-to-face delivery in most cases (n = 24) with single individuals (n = 6) [57, 59, 64, 65, 68, 73] or with groups of people (n = 10) [18, 48, 51, 54, 56, 69,70,71, 74, 76] or a combination of both modes of delivery (n = 8) [52, 53, 55, 58, 62, 66, 67, 75]. The electronic mode of delivery was often employed (n = 15), including messaging (n = 3) [67, 68, 70], computer-based delivery (n = 6) [48, 61, 63, 72, 74, 77], and digital devices (wearable or mobile devices) (n = 13) [18, 48, 50, 53, 58, 59, 61, 63, 65,66,67, 71, 74]. The printing mode of delivery was also utilized less frequently (n = 10).

Reported outcomes

Twenty-five of the 29 interventions mentioned a decrease in PIA and SB, while four studies [18, 54, 65, 74] found no changes in SB or PIA. These four interventions specifically targeted preschool children, school-age students, and adults at risk of diabetes. Four studies reported mixed results and inconclusive evidence. One study showed a significant decline in SB without any change in the level of PA [72] (see Table 6).

Behavioral theories

Our scoping review showed that the authors of the included studies referred to 15 behavioral theories (n = 15) (see Table 7 and Supplementary file 5). Most of the included studies used at least one of the four following theories: SCT (n = 14), SDT (n = 6), the TTM (n = 6), the TPB (n = 6), the SEM (n = 5), and the HBM (n = 5). Most interventions used either a single theory (n = 13) or a combination of two BCTs (n = 12). Only two interventions did not explicitly define the theoretical constructs guiding the development of the BCIs.

Table 7 Behavioral change theories and key constructs used in the design of BC interventions
Full size table

The SCT was the most commonly used theory. Five interventions used SCT as a single theory (n = 5) [48, 50, 51, 62, 69], whereas eight employed a combination of other behavioral theories: SDT [65], TPB [6, 68, 65], TTM [64, 65], HBM, SEM [57, 64,65,66, 79], behavioral choice theory [18], and protection motivation theory (PMT) [65]. Interventions rooted in SCT addressed specific psychological and social constructs ranging from one to four constructs per intervention. The most frequently used constructs were self-efficacy, self-regulation, observational learning, and positive reinforcement (see Table 5). SCT was used almost equally to reduce SB and PIA.

PA interventions mostly involved individual behavioral theories (SDT, SRT, TPB, TTM, HAPA), with a focus on reducing the intention-to-action gap. Conversely, the theories employed to reduce SB are primarily interpersonal (SCT, SET, SiS) and environmental (SEM). They seek to make behavior more socially acceptable, encouraging and influencing the behavior of others. Additionally, restructuring the environment is a central component of interventions aimed at reducing SB in the workplace.

Our scoping review showed that most interventions targetted the following individual-level constructs: self-efficacy (n = 16), motivation (n = 10), self-regulation [9], and the interpersonal level illustrated by using subjective norms (n = 5) and basic psychological needs (n = 4). Few studies have addressed environmental factors (e.g., institutional, community, society) (n = 7). The SB interventions used essentially socioecological constructs (n = 4) and enhanced self-efficacy (n = 6), self-regulation (n = 5), and modeling (n = 4). PIA BCI interventions were more centered on individual-level constructs such as motivation (n = 10), intention (n = 5), and controlled volition (n = 6) (see Table 8 and Supplementary file number 2).

Table 8 Classification of included studies using the BCTs taxonomy
Full size table

Our scoping review revealed some discrepancies in the characteristics of PIA interventions compared with those of SB interventions. The latter were considered systemic interventions based on SCT and SEM. They combined multilayered actions at the macro-level (environmental restructuring), the meso-level (social and peer pressure) and the micro-level (by activating intrapersonal and interpersonal mechanisms of change). In contrast, BCI targeting PIA were mostly focused on the individual level of change by using individual intrapersonal theories (SDT, TTM, TPB, HAPA, and PMT).

Behavior change techniques

All interventions were designed as multicomponent interventions integrating various behavior change techniques (see Table 7).

Our scoping review revealed that the scholars of the included studies used a set of 25 BCTs. On average, six to nine BCTs were used in an intervention (a minimum of 5 and a maximum of 12).

Social support, which is unspecified, was the most commonly used type of BCTs and involved targeting the interpersonal level (social influence) (n = 28), followed by goal setting, targeting the individual level (goal and intention) (n = 24); solving problems and identifying barriers at the individual level (belief capability) (n = 18); instruction on how to perform behavior at the individual level; self-monitoring of behavior (n = 17); feedback on the outcome of behavior at the individual level (n = 14); information about health consequences at the individual level (n = 14); social rewards targeting the interpersonal level (reinforcement and social influence) (n = 9); restructuring the physical environment targeting the environmental level (n = 6); and materiel rewards, targeting the interpersonal level (reinforcement) (n = 4). In our scoping review, most BCTs targeted the interpersonal level and the individual level followed by the environmental level.

Common characteristics of BCI with no modifications to PIA or SB

These interventions were based on educational, self-monitoring and the use of a coaching strategy involving distinct connected devices that targeted adults at risk of metabolic diseases or diabetes type 2) [18, 65] or preschool children, students, and adults at risk of metabolic diseases [54, 74], or a single individual level of behavioral change. They used face-to-face training sessions. Key contextual conditions that prevent the effectiveness of theory-informed interventions include the absence of parental involvement in BCTs targeting children [54], a lack of peer support in interventions involving students [74], and the absence of illness in interventions targeting adults [18, 65].

Description of studies reporting positive changes in PIA and SB

The included studies, mostly carried out in the workplace (n = 9), used a combination of education, training, and communication strategies (motivational interviews or coaching), along with social support and environmental restructuring. The included studies emphasized the importance of systemic-level interventions combining actions at the individual (face-to-face and digital interventions using wearable devices, desktops, and apps) and interpersonal (social support and group interventions) levels with macro-level environmental restructuring. Environmental restructuring encompasses interventions such as installing pedals and workstations, sending email reminders, and even using digital health apps [50, 57, 58, 63, 72]; it also focuses on reinforcing the knowledge and skills of actors and providing social support through group interventions. In contrast, other studies reported that BCIs targeting individuals with chronic diseases (e.g., CVD [68], diabetes [65, 69, 71], Parkinson’s disease [64], obesity [48], and cancer survivors [59] are essentially individually focused and underwent substantive changes in PIA and SB. These studies suggest that patients with NCDs are more committed to education and that coaching interventions intrinsically motivate people to follow PA recommendations [59, 64, 68, 71].

Intensity of theory use

We found heterogeneous use of theory in the implemented interventions. Fifteen interventions involved an intensive degree of theory use (level 3). Eleven interventions entailed moderate levels of theory (Level 2), and three interventions utilized a low level of theory (Level 1) (see Table 9).

Table 9 Intensity/degree of use theory in based theory-intervention
Full size table

Discussion

In sum, our scoping review showed that most interventions used a combination of similar modes of delivery, design, and components (education, training/coaching, regulation, and the use of connected devices), and BCIs were mostly individually focused and based, in most cases, on education and self-monitoring.

Most interventions were focused on individual levels of behavior changes and involved a multitude of intrapersonal behavioral theories and wearable devices for monitoring, using diverse BCTs with a focus on social support and goal setting. Only two studies can be considered systemic level theory informed BCIs addressing both individual intrapersonal drivers (e.g., motivation, attitude, perceived norms, self-efficacy, etc.) combined with interpersonal interventions (group and social support interventions) and macro-level interventions, such as environmental restructuring in the workplace.

Our scoping review indicated that single digital technology-based web apps informed by intrapersonal theories, such as the TPB, self-regulation, and SDT, had no significant effects. Hence, there is a need to combine intrapersonal theories with interpersonal and environmental interventions for better adherence to interventions and the adoption of a desired behavior [81, 82]. Indeed, interventions informed by the HBM, aimed at addressing an individual’s perceptions of PA and increasing one’s level of PA, have shown no significant effect [83].

The relevance of systemic theory informed BCIs stems from the complexity of causal processes underlying SB and PIA, which are considered a consequence of intricate interactions between intertwined levels of structure and agency [16, 18]. PIA and SB are influenced by individual, interpersonal, and organizational and broader contextual factors [84] (Heath et al., 2012). At the individual level, behavior is defined by people’s awareness, cognition, beliefs, and skills. At the interpersonal level, behavior is impacted by the extent to which social support is received from family and friends. At the organizational level, behaviors are constrained by cultural norms and practices in the workplace. At the broader level, behaviors are constrained by contextual factors at the national and global levels, such as legal frameworks, environmental restructuring, political and socioecological factors shaping individuals’ architecture of choice, and their day-to-day decision-making [16, 18].

This suggests the importance of considering the notion of “reciprocal determinism,” which refers to the dynamic interaction between personal social, and environment factors and behavior [24]. The environment plays a significant role in the acquisition of PA behaviors and, consequently, in behavioral change [85]; it can encompass the immediate environment around the individual (one’s parents, workplace, neighbors, and community) as well as the interpersonal environment of the community. As such, PA is conditioned by the individual’s motivation (which can be intrinsic or extrinsic) [86], physical ability, social support, the availability of wearable device pedometers or accelerometers [87]  and the existence of an enabling living environment (sport fields, space, resources), and regulatory enabling policies (breaks/leave from work, health insurance) [16].

At the national and global levels, individual behaviors are often constrained or facilitated by national legal contexts and restructuring policies of the built environment, including public transit, green spaces, parks, and recreational facilities [88]. Thus, environmental restructuring can be a good example of the complementarity and synergies of interventions, as shown by Dugdill, who highlighted the relevance of macro-level interventions to alter the workplace, where people spend a great deal of time. Systemic interventions, in line with those used by [89, 90], that combine multiple levels of interventions (individual, interpersonal, and environmental) may have synergistic effects on behavioral changes compared with individually focused interventions (face-to-face and digital interventions).

Our review underscores the importance of environmental restructuring as a complementary intervention to individually focused BCIs. In the workplace, this can include promotion of managers’ leadership such that they serve as role models for employees, as suggested by [91,92,93]. As a consequence, employees may perceive strong social influence and peer pressure, which may increase their self-efficacy and self-regulated behaviors [94]. These interventions seem to foster social identification, social comparison, and socialization mechanisms by increasing individuals’ adherence to BCIs in the workplace [91,92,93].

In addition, at the organizational level, employees’ behaviors are often influenced by organizational policies promoting PA in the workplace [87]. Moreover, the broader context plays a role in shaping the individuals’ behaviors. For instance, Davis [21] reported that behavioral modeling is only effective if individuals see other active people in their social context. Other scholars have shown that a lack of perceived security (crimes, sexual harassment, incivility) may reduce people’s willingness to carry out outdoor PA [95].

Our scoping review indicates that in the context of school BCIs, in line with other findings [96, 97], children may also benefit from systemic interventions by reducing their screen time usage through school policies and receiving individual training sessions to enable them to reduce their SB while also engaging with their parents (interpersonal and social influence) through role modeling. However, more attention is needed to develop systemic BCIs based on multiple-level interventions, such as individual coaching, mentoring, interpersonal social support, and altering the physical and cultural environment [98].

Our scoping review, in line with [82, 94, 99] and [100], has shown the usefulness of SCT in explaining how the training and empowerment of individuals enhance their self-efficacy, self-regulation, their perceived benefits, and risk and control volition, which may prove appropriate in the context of PA and SB interventions.

Our scoping review demonstrated, in line with previous systematic reviews [101], that using a combination of multiple behavioral change techniques is associated with an increased overall effect of the intervention and the adoption of desired behavioral outcomes. Techniques include, for instance, social support, goal setting, and self-monitoring, in line with other studies [102, 103].

Figure 2 shows a tentative integrative framework that incorporates three levels of interventions (environmental, interpersonal, and intrapersonal) and may be useful for helping program designers to build theory informed BCIs on the basis of a multilayered theoretical model. For instance, at the intrapersonal level, one might use the HBM combined with the TTM and SDT. However, at the interpersonal level, program designers might use SCT and behavioral choice theory. At the environmental level, one can use environmental theories such as social influence strategies ( see Fig. 2).

Fig. 2
figure 2

Integrative framework of theories and constructs for effective BCT interventions

Full size image

These constructs serve as mechanisms of action at the individual and interpersonal levels. This finding aligns with the results regarding the contribution of SCT and its constructs in predicting and adopting active behavior.

Study limitations and research gaps

In our review, we identified a lack of comprehensive reporting by scholars of key theoretical constructs underlying the design of BCI. We may have missed other relevant literature, as we had to make some trade-offs between comprehensiveness, depth of analysis and feasibility (Arksey, 2005). However, we performed a systematic, comprehensive search of four databases, including Google Scholar, to identify contextually rich gray literature. In addition, two reviewers screened the titles and abstracts, and extracted the data. Our findings also suggest that many theory-informed interventions do not use theoretical constructs appropriately; however, a call for improving the reporting and quality of intervention fidelity is needed while promoting the use of standardized tools such as Michie’s taxonomy of BCIs [40] and BCTs [104].

Our scoping review included only experimental studies that lacked sufficient descriptions of the role of context in shaping the characteristics of interventions and their mechanisms of action. Thus, more attention should be paid to promoting evaluation using context-sensitive methods and approaching theory-based evaluation, realistic evaluation [105], qualitative comparative analysis [106], and contribution analysis [107]. Further research is needed to unpack the black box of behavioral theories -informed interventions by unraveling what works for whom and in what context.

Further studies are also needed to examine the role of individual and digital interventions, which we insufficiently explored in our review. More rigorous systematic and meta-analyses are needed to complement the results of this descriptive, explorative scoping review and to provide evidence of the effectiveness of Theory -informed BCI [85].

Conclusion

Our review offers an innovative approach to systematically categorize behavioral theories interventions using a set of appropriate behavioral theories taxonomies, tools, and techniques, and provides working examples of how these taxonomies can be applied to assess the theory use and the described characteristics of BCT theory-informed interventions. Our study suggests an integrative framework to help program designers develop interventions while implying that specific behavioral theories and BCTs can be used at every level of intervention (the individual, interpersonal and environmental, policy and global levels). In sum, the congruence between behavioral theories, the implementation settings, and the characteristics of the targeted subpopulations needs to be considered when designing behavioral theories interventions to reduce PIA and SB. One size does not fit all. We also recommend, in line with (Noar et al., 2008), that behavioral change practitioners select theories and techniques based on their congruence with participants’characteristics and the nature of the context.