Facilitators, barriers, and interventions for exercise adherence in patients with prediabetes: a mixed-methods systematic review

Abstract

Background:

As a critical preclinical stage of type 2 diabetes (T2D), prediabetes necessitates urgent lifestyle interventions to halt disease progression. Despite robust evidence supporting the efficacy of exercise in improving glycemic control, exercise adherence remains suboptimal among individuals with prediabetes.

Objectives:

This review aimed to identify barriers and facilitators to exercise adherence in individuals with prediabetes while also extracting evidence-based behavior change techniques (BCTs) that may enhance physical activity maintenance.

Design:

Mixed-methods systematic review

Data sources and methods:

A comprehensive literature search was performed in 10 databases from inception to November 2024. Guided by the JBI Manual for Evidence Synthesis, a convergent integrated approach was used to map barriers and facilitators to the Behavior Change Wheel (BCW) and Theoretical Domains Framework (TDF). TDF domains were ranked by frequency across studies, while adherence-promoting interventions were mapped to BCW intervention functions and BCTs. Barriers were systematically linked to BCW intervention functions and BCTs to guide implementation design.

Results:

A total of 36 studies were included, consisting of 9 qualitative studies, 24 quantitative studies (15 randomized controlled trials, 5 cross-sectional studies, and 4 quasi-experimental studies), and 3 mixed-methods studies. The synthesis identified “lack of time,” “lack of appropriate exercise equipment or venues,” and “health-related physical limitations” as predominant barriers. Conversely, “perceived health benefits,” “social support,” and “tools or methods to help with exercise adherence” emerged as key facilitators. These factors were mapped to five primary TDF domains: reinforcement, environmental context and resources, behavioral regulation, social influences, and skills. Furthermore, seven intervention functions derived from the BCW were linked to 21 BCTs.

Conclusion:

This review highlights multifaceted determinants of exercise adherence in individuals with prediabetes and identifies theory-informed strategies to address barriers to adherence. Determinants supported more consistently across studies may provide a stronger basis for intervention development, whereas less frequently reported findings should be interpreted more cautiously.

Trial registration:

PROSPERO CRD42024605588.

Plain language summary

What helps or stops people with prediabetes from exercising, and how can we support them?

Prediabetes is a condition where blood sugar levels are higher than normal but not yet high enough to be called type 2 diabetes. Exercising is one of the best ways to stop prediabetes from getting worse. However, many people find it hard to stick to an exercise routine. In this study, we looked at existing research to find out what stops people with prediabetes from exercising (barriers) and what helps them keep going (facilitators). We found that the biggest hurdles are a lack of time, not having the right equipment or places to exercise, and physical health problems. On the other hand, understanding the health benefits, getting support from friends and family, and using helpful tools make people more likely to exercise. To help people overcome these hurdles, we suggest creating personalized exercise plans. These plans should include specific techniques like setting clear goals, making action plans, and building social support. By using these simple strategies, doctors and nurses can better help people with prediabetes stay active, which may help reduce their risk of developing type 2 diabetes.

Introduction

Prediabetes, defined as abnormal glucose homeostasis (plasma glucose levels between normal and diabetic thresholds), includes impaired fasting glucose (IFG), impaired glucose tolerance (IGT), or their combination. This condition poses a global public health challenge: >70% of affected individuals may develop diabetes without intervention. ¹ In 2021, ~762 million adults (20–79 years) had prediabetes, with projected global prevalences of IGT and IFG rising to 10.0% (638 million) and 6.5% (414 million), respectively. ² A 2020 meta-analysis (129 studies, 10 million individuals) showed prediabetes was associated with 13% higher all-cause mortality, 15% higher cardiovascular disease, 16% higher coronary heart disease, and 14% higher stroke risk versus normal glycemia (median follow-up: 9.8 years). These risks were more pronounced in individuals with preexisting atherosclerotic cardiovascular disease ³ : during 3.2 years of follow-up, prediabetes was associated with 36% higher all-cause mortality and 37% higher cardiovascular disease risk. Lifestyle modifications and pharmacological interventions delay progression to diabetes,^4,5 but medications like metformin and acarbose are inferior to lifestyle interventions. ⁶ Pharmacological approaches also carry higher risk and cost burdens, making lifestyle intervention the preferred clinical strategy. ¹

Regular physical activity is a core component of lifestyle interventions for prediabetes, improving glycemic control, lipid profiles, insulin resistance, and inflammatory markers. ⁷ Aerobic exercise and reduced sedentary time significantly improve cardiometabolic outcomes, while combined aerobic-resistance training enhances pancreatic β-cell function. ⁸ A systematic review of 19 studies showed structured exercise programs delay disease progression, reduce type 2 diabetes (T2D) risk, and improve health parameters in prediabetic individuals. ⁹ Despite established metabolic benefits, exercise adherence remains suboptimal in this population. Compounding this, limited research has systematically investigated multifactorial adherence determinants—particularly the long-term efficacy of adherence-promoting interventions. ¹⁰ To address this gap, it is imperative to systematically identify the barriers to and facilitators of exercise adherence in individuals with prediabetes and to develop targeted interventions to overcome these barriers.

Current research on exercise adherence in individuals with prediabetes includes both quantitative and qualitative approaches. Quantitative studies primarily examine the effects of different exercise modalities on adherence-related outcomes,^11,12 whereas qualitative studies explore patients’ subjective experiences, as well as their perceived barriers and facilitators to exercise. ¹³ However, although a substantial number of primary studies have investigated barriers to and facilitators of exercise adherence and proposed corresponding interventions, no systematic review has yet integrated and analyzed these findings comprehensively. This lack of synthesis hinders a comprehensive and objective understanding of the relevant influencing factors and limits the development of evidence-based intervention strategies. Therefore, this study aimed to conduct a mixed-methods systematic review to synthesize this evidence.

With the advancement of healthcare knowledge and evidence-based medicine, increasing attention has been paid to the use of theoretical frameworks to guide behavioral interventions, rather than relying solely on experience or assumptions. Behavior change frameworks help researchers understand the determinants of health-related behaviors and identify ways to support behavioral modification. These frameworks have been widely applied to promote changes in various health-related behaviors, such as medication adherence, smoking cessation, and physical activity.^14–16 Theory-informed approaches to behavior change have also been shown to be more effective than interventions developed without an explicit theoretical foundation. ¹⁷

Among the most widely applied frameworks in this domain are the Behavior Change Wheel (BCW), the Theoretical Domains Framework (TDF), and the behavior change techniques (BCTs). The BCW outlines three core components for comprehensively analyzing facilitators of and barriers to individual behavior change: capability, opportunity, and motivation (collectively known as the COM-B model). ¹⁸ Surrounding the COM-B model, the BCW identifies nine intervention functions that can be used to address barriers to behavior change. To enable a more precise diagnosis of the barriers and facilitators identified within the COM-B model of the BCW, the TDF was applied. The TDF further elaborates on the three components of the COM-B model, enabling a detailed classification of determinants influencing capability, opportunity, and motivation.^19,20 Meanwhile, the BCTs are used to translate intervention functions of the BCW into more specific and actionable techniques. ²¹ BCTs represent the “active ingredients” required to implement the intervention functions, ²² providing a standardized terminology for specifying concrete, replicable intervention components. These integrated frameworks have demonstrated utility in synthesizing behavioral determinants across diverse healthcare contexts, ranging from physical activity promotion to chronic disease management.^23–25 Therefore, this study applies the BCW, TDF, and BCTs to systematically analyze barriers and facilitators of exercise adherence in individuals with prediabetes and to identify targeted intervention strategies that address these determinants.

In summary, while numerous studies have investigated factors influencing exercise adherence in prediabetes and proposed corresponding interventions, no review has yet systematically integrated this evidence using behavioral theoretical frameworks. This gap limits understanding of how these factors interact and hinders the development of targeted strategies. This review therefore aims to synthesize quantitative, qualitative, and mixed-methods evidence through the BCW, TDF, and BCTs to identify barriers and facilitators and to specify applicable intervention strategies for exercise adherence in individuals with prediabetes.

Methods

This mixed-methods systematic review was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020. ²⁶ The protocol was registered with PROSPERO in November 2024 (registration number CRD42024605588).

Search strategy

A comprehensive literature search was performed across 10 databases, including PubMed, Embase, Cochrane Library, Web of Science, PEDro, CINAHL, China National Knowledge Infrastructure (CNKI), Chinese Biomedical (CBM), Wanfang Data, and VIP Database for Chinese Technical Periodicals (VIP), to identify barriers and facilitators of exercise adherence in individuals with prediabetes. Reference lists of retrieved articles were manually screened for additional studies, and gray literature was searched through Google Scholar and OpenGrey. Because this review employed a convergent integrated mixed-methods approach and did not perform meta-analysis, formal statistical assessment of publication bias (e.g., funnel plots) was not appropriate. The search span covered database inception to November 2024. Combinations of MeSH terms, Emtree synonyms, and free words were used in the literature search. The search terms included “prediabetic state/prediabetes/state, prediabetic/IGT” AND “exercise/ motor activity/physical activity/sedentary behavior/walking” AND “patient compliance/barrier*/enabler*” were used without date restrictions. The full database-specific search strategies are provided in Supplemental Material 1.

Inclusion and exclusion criteria

The inclusion criteria were as follows: (a) participants with prediabetes aged ⩾18 years; patients diagnosed with prediabetes according to the American Diabetes Association (ADA) and World Health Organization (WHO) criteria: FBG: 100–125 mg/dL or HbA1c 5.7% to 6.4%. (b) Quantitative, qualitative, or mixed-methods study, quantitative studies included the factors affecting exercise adherence, and qualitative studies included the perspectives of prediabetic patients, family caregivers, health professionals participating in exercise interventions. The exclusion criteria were as follows: (a) Studies reporting on factors influencing adherence to multi-component interventions. (b) Studies published in languages other than English or Chinese. This restriction was applied to ensure the feasibility of data extraction and the accuracy of the synthesis, given the linguistic capabilities of the review team, while still capturing the majority of relevant international and region-specific evidence. (c) Studies with incomplete data or data that could not be analyzed. (d) Literature with redundant publication.

Selection process

Two researchers (Y.Y.J. and G.H.T.) independently screened literature and extracted data after importing documents into EndNote X9. The screening process included: removing duplicates; screening titles/abstracts to exclude irrelevant studies; and reviewing full texts for eligibility. Discrepancies were resolved by discussion, with a third researcher (L.X.Y.) consulted for unresolved issues.

Quality assessment

The Mixed Methods Appraisal Tool (MMAT) 2018 version was used to assess study quality, as it offers a unified instrument with strong content validity for appraising diverse study designs. ²⁷ Three researchers (Y.Y.J., G.H.T., and L.X.Y.) independently evaluated methodological rigor using MMAT, which assesses five study types: qualitative, RCTs, non-randomized, quantitative descriptive, and mixed-methods. As overall scoring was not recommended, each criterion was rated “Yes,” “No,” or “Can’t tell” as applicable. ²⁸ Discrepancies were resolved by consensus, with a senior researcher (C.H.) consulted for unresolved issues.

Data extraction

Data were independently extracted by two researchers (Y.Y.J. and G.H.T.) using Excel spreadsheets, capturing variables including author(s), publication year, country, study design, sample size, participant characteristics, data collection/analysis methods, identified barriers/facilitators, and intervention components. Qualitative studies were coded for themes, interpretations, and quotations, while quantitative data were abstracted as numerical findings and interpretations. Extracted factors were categorized as barriers, facilitators, or mixed factors (dual/ambiguous roles) in tabular format. Discrepancies in extraction were resolved by consensus or consultation with a third researcher (C.H. or C.X.Y.).

Data analysis

Data analysis was conducted using the JBI convergent integrated approach for mixed-methods systematic reviews. ²⁹ In accordance with this approach, quantitative findings were transformed into qualitative form (“qualitizing”) and then integrated with qualitative evidence for combined synthesis.

Data transformation

Following data extraction, quantitative data from quantitative studies (including quantitative components of mixed-methods studies) were transformed into qualitative form through “qualitizing”. ³⁰ This process involved converting statistical findings about factors influencing exercise adherence into narrative statements describing their directional influence. For instance, the finding that “a negative association was observed between the initial mean steps per day and the intention to do more exercise within the next month, but only in the older age group (p < 0.05)” was transformed into the qualitative finding: “Lack of self-efficacy and behavioral inertia” serves as a barrier to exercise adherence in older patients. To minimize interpretive bias specific to this transformation step, all qualitization coding was performed independently by the same two reviewers who conducted the data extraction. Any discrepancies in interpretation were resolved through consensus discussion with the same three senior reviewers as above. This procedure is consistent with JBI guidance for mixed-methods systematic reviews.

Data synthesis and integration

We employed a convergent integrated approach for data synthesis and integration. This involved merging qualitized quantitative data with original qualitative data. First, barriers and facilitators were identified through deductive thematic analysis and coded into the COM-B model of the BCW and the TDF domains, followed by inductive thematic analysis. During this process, overlapping constructs were merged and refined to eliminate redundancies and ensure conceptual distinctiveness. Concurrently, adherence-promoting interventions were extracted and deductively thematic analysis coded into the BCW’s intervention functions and corresponding BCTs. Finally, the identified barriers were systematically linked with the most relevant intervention functions and BCTs. The most prevalent barriers and corresponding intervention strategies were prioritized to ensure practical applicability. A formal CERQual assessment was not undertaken for the integrated findings. This review followed the JBI convergent integrated approach for mixed-methods systematic reviews, ²⁹ which focuses on data extraction, transformation, and integration of qualitative evidence with qualitized quantitative findings; however, current JBI guidance does not yet provide an established method for applying CERQual to these integrated mixed-methods findings. Therefore, we interpreted the synthesized findings cautiously, with particular attention to the number of contributing studies and their methodological quality as assessed by the MMAT. Given the substantial methodological and clinical heterogeneity across the included studies, and following a JBI convergent integrated approach for mixed-methods systematic reviews, a quantitative meta-analysis was not feasible.

Results

Study selection

The search identified 4312 relevant studies from 10 databases. 2768 studies were excluded due to duplication, and 1278 studies were omitted based on titles and abstracts. Of these, 266 studies were selected for full-text screening. After reviewing the full texts, 36 articles met the eligibility criteria. The reasons for exclusion and the process details are given in Figure 1.

Figure 1.

Flowchart for search and inclusion of studies.

Study characteristics

Supplemental Material 2 describes included studies (36 total: 24 quantitative, 9 qualitative, 3 mixed-methods). Qualitative studies used interviews/focus groups to explore participants’ attitudes/motivations, while quantitative research employed standardized tools to measure variables (e.g., physical activity, glucose indices). Mixed-methods studies integrated both approaches. Four studies reported adverse events (muscle soreness and exercise-related injuries, 2 cases each), with high-intensity training linked to higher injury risk. Adherence rates ranged from 50% to 96.3%, mostly between 70% and 90%.

Risk of bias in studies

Supplemental Material 3 summarizes the overall quality of included studies. Of the 36 studies, 18 met all or 80% of quality criteria, 11 met 60%, and 7 met ⩽40%. While qualitative studies generally demonstrated high adherence to reporting standards, significant methodological variations were observed among quantitative studies. The lower ratings were primarily attributed to critical methodological flaws, including substantial attrition bias, insufficient reporting of randomization and allocation concealment mechanisms, and the absence of blinding for outcome assessors. The MMAT discouraged the exclusion of studies with low methodological quality and therefore included the full range of studies for analysis.

Results of syntheses

Results of barriers and facilitators

A total of 47 distinct influencing factors (16 facilitators, 20 barriers, and 11 mixed factors) were extracted and coded into relevant TDF domains following a rigorous refinement process to eliminate redundancies and merge overlapping constructs. Table 1 presents barrier/facilitator themes/subthemes via inductive analysis of TDF-coded data, with details in Supplemental Material 4. TDF domains were ranked by study frequency (Table 2), with the top five being reinforcement (n = 24), environmental context and resources (n = 23), behavioral regulation (n = 18), social influences (n = 16), and skills (n = 15). The Optimism domain was unrepresented across all 36 included studies.

Table 1.

Barrier and facilitator themes and sub-themes derived from the COM-B model and the TDF.

Barriers	Facilitators	Themes	TDF domains	COM-B model of BCW
• Lack of knowledge about how to exercise • Lack of knowledge about the importance of exercise • Lack of awareness of prediabetes (symptoms, risks, and trends)	• Knowledge about how to exercise — —	Knowledge about exercise Awareness of prediabetes (symptomatic/risks/trends)	Knowledge	Capability
• Cognitive decline	—	Cognitive capability	Memory, attention, and decision processes
• Difficulty in decision-making	—	Decision-making
—	• Healthcare professionals’ assistance with patients’ memory skills	Healthcare professionals’ assistance with patients’ memory skills
• Tools do not help improve exercise adherence	• Tools help improve exercise adherence	Tools or methods to help with exercise adherence (e.g., exercise diary)	Behavioral regulation
• Methods do not help improve exercise adherence	• Methods help improve exercise adherence
—	• Self-management skills	Self-management skills
—	• Integrating exercise into daily life	Integrating exercise into daily life
• Physical function	—	Physical function	skills
• Patients may not have the ability to access professional guidance or resources	• Patients may have the ability to access professional guidance or resources	Patient’s ability to access specialized guidance or resources
—	• Tailored/flexible exercise programs	Tailored/flexible exercise programs	Environmental context and resources	Opportunity
• Inclement weather • Unsafe exercise environments • Lack of appropriate exercise equipment or venues	— — • Appropriate exercise equipment or venues	Physical environment factors
• Work-related limitations • Family responsibilities • Other hobbies/commuting takes up time for exercise • Lack of time for which no reason was mentioned	— — — —	Lack of time for exercise
• Financial costs	• Financial costs	Financial costs
• Common culture	—	Common culture
— — • Lack of support from outside the exercise group (e.g., family, friends)	• Support from exercise partners • The exercise instruction and emotional support from professionals • Support from outside the exercise group (e.g., family, friends)	Support from others	Social influences
• Lack of trust in professionals	• Trust in professionals	(Lack of) Trust in professionals
• Gender or sociocultural and religion	—	Gender or sociocultural and religion
• Being gazed at while exercising	—	Being gazed at while exercising
—	• Rewarded by others for consistent exercise	Rewarded by others for consistent exercise	Reinforcement	Motivation
—	• Health benefits of exercise	Health benefits of exercise
—	• Repeat to become habitual	Repeat to become habitual
• Adverse effects of exercise	—	Adverse effects of exercise
—	• Enjoying exercise	Enjoying exercise	Emotion
• Feel boring	—	Feel boring
• Emotional problems (fear/depression)	—	Emotional problems (fear/depression)
• Tiredness	—	Tiredness
• Social professional role and identity	—	Social professional role and identity	Social or professional role and identity
• Lack of self-efficacy and behavioral inertia	• Self-efficacy and behavioral inertia	Self-efficacy and behavioral inertia	Belief about capabilities
—	• Control beliefs and behavioral autonomy	Control beliefs and behavioral autonomy
— — —	• Contemplation Stage • Preparation stage • Action stage	Stages of change model	Intentions
• Lack of intention to adhere to exercise	• Intention to adhere to exercise	(lack of) Intention to adhere to exercise
—	• The goal/motivation of promoting health	The goal/motivation of promoting health	Motivation and goals
• Setting unrealistic goals	—	Setting unrealistic goals
• Fear of exercise injury or discomfort	—	Fear of exercise injury or discomfort	Belief about consequences
• Lack of expectations regarding exercise benefits	• Expectations of exercise benefits	Expectations of exercise benefits

TDF, theoretical domains framework.

Table 2.

Frequencies of the number of studies identified by each TDF domain presented in rank order, and frequencies of barriers, facilitators, and mixed (facilitators/barriers).

TDF domains	Number of studies identified	Barriers	Facilitators	Mixed
Reinforcement	24	5	22	/
Environmental context and resources	23	18	5	9
Behavioral regulation	18	/	4	15
Social influences	16	3	9	12
Skills	15	9	/	8
Emotion	11	5	9	/
Belief about consequences	10	2	/	10
Knowledge	9	5	/	5
Beliefs about capabilities	9	/	1	9
Intentions	8	/	4	5
Motivation and goals	8	1	7	/
Memory, attention, and decision processes	3	2	2	/
Social professional role and identity	1	/	1	/
Optimism	/	/	/	/

TDF, theoretical domains framework.

Results of interventions

Based on intervention factors extracted from the literature that facilitate adherence, we mapped the intervention content to the relevant intervention functions of the BCW ¹⁸ and identified seven major intervention functions and 21 BCTs capable of achieving these functions. ²¹ Table 3 summarizes the BCW intervention functions and the BCTs identified in the included studies. The three most frequently reported intervention functions were enablement (n = 5), training (n = 4), and persuasion (n = 3). The top three BCTs were instruction on how to perform the behavior (n = 10), social support (unspecified; n = 10), and action planning (n = 9). Supplemental Material 5 provides a detailed list of all BCTs derived from the studies, along with their corresponding citations. Critically, this study established direct correspondence between identified barrier factors, intervention functions, and BCTs (Table 3), proposing aligned intervention functions and BCTs for each barrier.

Table 3.

BCTs targeting key barriers to exercise adherence.

Barriers	Themes	TDF domains	COM-B model of BCW	Intervention functions of BCW	BCTs
Lack of knowledge about how to exercise	Knowledge about exercise	Knowledge	Capability (psychological)	Education	4.2 Information about antecedents
Lack of knowledge about the importance of exercise					4.3 Re-attribution
Lack of awareness of prediabetes (symptoms, risks, and trends)	Awareness of prediabetes (symptomatic/risks/trends)
• Cognitive decline	Cognitive capability	Memory, attention and decision processes		Enablement	3.2 Social support (practical)
				Environmental restructuring	7.1 Prompts/cues
• Difficulty in decision-making	Decision-making			Training	4.1 Instruction on how to perform the behavior
• Tools do not help improve exercise adherence	Tools or methods to help with exercise adherence (e.g., exercise diary)	Behavioral regulation		Training	2.4 Self-monitoring of outcome(s) of behavior
• Methods do not help improve exercise adherence
• Patients may not have the ability to access professional guidance or resources	Physical function	Skills	Capability (physical)	Training	2.3 Self-monitoring of behavior
• Physical function					4.1 Instruction on how to perform the behavior
• Work-related limitations	Lack of time for exercise	Environmental context and resources	Opportunity (physical)	Enablement	1.4 Action planning
• Family responsibilities
• Other hobbies/commuting takes up time for exercise
• Lack of time for which no reason was mentioned				Training	2.1 Monitoring of behavior by others
• Inclement weather	Physical environment factors			Environmental restructuring	12.1 Restructuring the physical environment
• Unsafe exercise environments
• Lack of appropriate exercise equipment or venues
• Financial costs	Financial costs				12.2 Restructuring the social environment
• Common culture	Common culture				12.5 Adding objects to the environment
• Lack of support from outside the exercise group (e.g., family, friends)	Support from others	Social influences	Opportunity (social)	Enablement	3.1 Social support (unspecified)
• Lack of trust in professionals	(Lack of) Trust in professionals
• Gender or sociocultural and religion	Gender or sociocultural and religion
• Being gazed at while exercising	Being gazed at while exercising
• Adverse effects of exercise	Adverse effects of exercise	Reinforcement	Motivation (automatic)	Education	5.2. Salience of consequences
• Feel boring	Feel boring	Emotion		Incentivization	10.2 Material reward (behavior); 10.3 Non-specific reward
• Emotional problems (fear\depression)	Emotional problems (fear/depression)
• Tiredness	Tiredness
• Lack of self-efficacy and behavioral inertia	Self-efficacy and behavioral inertia	Beliefs about capabilities	Motivation (reflective)	Enablement	1.2 Problem-solving
				Persuasion	15.1 Verbal persuasion about capability
• Fear of exercise injury or discomfort	Fear of exercise injury or discomfort	Beliefs about consequences		Education	5.1 Information about health consequences
• Lack of expectations regarding exercise benefits	Expectations of exercise benefits
• lack of intention to adhere to exercise	(Lack of) Intention to adhere to exercise	Intentions		Incentivization	2.4 Self-monitoring of outcome(s) of behavior
				Persuasion	2.2 Feedback on behavior
• Setting unrealistic goals	Setting unrealistic goals	Motivation and goals		Enablement	1.1 Goal setting (behavior)

Discussion

This review identified multiple determinants of exercise adherence in individuals with prediabetes and linked these determinants to BCW intervention functions and specific BCTs. Importantly, the strength of evidence was not uniform across findings. Determinants such as lack of time, limited access to exercise facilities or resources, perceived health benefits, and social support were more consistently supported across the included studies, suggesting that they may be of greater practical relevance for intervention design. By contrast, findings related to less frequently represented domains, particularly memory, attention and decision processes, may reflect a narrower evidence base and should therefore be interpreted with greater caution. To our knowledge, this review is among the first mixed-methods systematic reviews to examine exercise adherence in people with prediabetes using the TDF and the BCW. By linking identified determinants with potential intervention functions and BCTs, this review provides a theory-informed basis for developing more targeted and feasible exercise interventions. Nevertheless, the confidence placed in specific findings should be considered in light of both the number and methodological quality of the contributing studies.

Capability in COM-B: Barriers and corresponding BCTs

Capability is defined as the individual’s psychological and physical capacity to engage in the activity concerned. ¹⁸ This study identified key determinants within this domain, encompassing both physical capacity (physical skills) and psychological capacity (including knowledge, decision-making processes, and behavior regulation).

In the dimension of physical capability, 15 studies highlighted reduced physical activity engagement due to health-related physical limitations such as aging or chronic conditions,^31,32 findings that are relatively well supported and converge with prior research on barriers to general physical activity. ³³ Physical discomfort, pain, and comorbidities were the most frequently cited barriers to physical activity engagement,^31,34,35 suggesting that these factors represent relatively robust targets for intervention. For patients with chronic pain or other arthritis-related functional limitations, non-weight-bearing exercises are recommended as alternatives to walking or jogging. ³⁶ Exercise guidance should be tailored to each patient’s physical condition. Post-exercise and healthcare professionals analyze data from mobile health (mHealth) devices to evaluate exercise patterns and physiological indicators, enabling timely resolution of activity-related challenges.^37–39 The subtheme based on the skill of patients obtaining professional guidance or resources emerged from eight studies, with patients’ inability to access corresponding medical resources and exercise methods contributing to poorer adherence.^40,41 This study highlights the significant role of healthcare professionals’ educational effectiveness in primary care for addressing this barrier factor. Training and guidance in exercise expertise, effective communication, and meaningful encouragement were found to influence exercise adherence. ⁴² Multidisciplinary clinical teams should actively guide patients in optimizing community resources. ³⁴ National health promotion campaigns should emphasize the adoption of non-weight-bearing exercises in prediabetes management.^34,43

Within the psychological capacity dimension, 15 studies identified tools such as exercise diaries, self-monitoring technological devices, and mobile device reminders as critical components of behavioral regulation in exercise interventions,^44,45 suggesting that behavior regulation is one of the more consistently supported determinants in this review. Achieving a permanent shift in physical activity patterns is a long-term process. Evidence suggests that interventions to promote physical activity should initiate with high-frequency contact, followed by a gradual reduction in intervention frequency, thereby facilitating the integration of physical activity into daily routines and the formation of sustainable habits. ⁴⁶ Concurrently, cultivating awareness of self-management is considered a crucial component of successful behavior maintenance. ⁴⁷ However, four studies noted that some participants failed to consistently utilize monitoring tools and frequently prioritized non-exercise activities when confronted with competing demands. ⁴⁸ As this finding was derived from a relatively small subset of studies, its relative importance should be interpreted with some caution. This highlights the need for social support. Studies have established support systems through organizing outdoor activities, medical expert lectures, and encouraging family members to engage in exercise together.^13,38,49 Simultaneously, enhancing education and training for grassroots community healthcare providers strengthens the multi-stakeholder support network. Additionally, reminders delivered via text messages or mobile devices, along with repetitive information dissemination through text, images, and videos, serve to facilitate behavior implementation. ³⁸ The phenomenon where patients prioritize discontinuing exercise reflects insufficient awareness among these participants regarding the importance of exercise, along with deficits in both exercise knowledge and prediabetes-related knowledge.^13,35 Several studies included in this review synthesize educational strategies for individuals with prediabetes across domains such as health education, counseling, and information dissemination. Within the realm of education and counseling, 75% of participants expressed a demand for exercise-related consultations. ⁵⁰ To address this, multifaceted educational initiatives have been implemented. These include structured Baduanjin training programs, group-based health education workshops, ³⁹ and tailored prediabetes education in rural healthcare clinics. The latter involves instructional manuals detailing risk mitigation strategies, exercise duration guidelines, and safety protocols. ⁵¹ Diabetes specialist nurses deliver comprehensive education encompassing prediabetes pathophysiology, nutritional management, and blood glucose monitoring. Concurrently, physicians should clearly communicate the health risks associated with prediabetes and the metabolic benefits of evidence-based exercise regimens.^39,52 Furthermore, after fully explaining the intervention tasks and mechanisms on-site, key messages are reinforced through new media platforms to amplify program effectiveness. ³⁸ Based on these findings, it is recommended to prioritize exercise intervention programs in regions with high prediabetes prevalence. Public health campaigns should emphasize the physiological benefits of regular physical activity while strengthening health literacy initiatives to improve disease prevention capabilities in at-risk populations. ⁵³

Opportunity in COM-B: Barriers and corresponding BCTs

Opportunity is defined as all the factors external to the individual that enable or prompt the behavior. ¹⁸ In this study, the core factors within this dimension include physical opportunities (environmental context and resources) and social opportunities (social influence).

Within the physical opportunity dimension, perceived lack of time emerged as the most prevalent barrier to exercise engagement, consistent with prior research findings.^54,55 Because this barrier was repeatedly identified across a relatively large number of included studies, it may represent one of the most robust and clinically relevant determinants for intervention prioritization. Time constraints were further compounded by prolonged commuting durations, familial obligations, intensive extracurricular commitments, or competing leisure pursuits.^56,57 Targeted strategies addressing common barriers such as occupational time constraints, for example scheduling weekend exercise sessions, emphasize the importance of maintaining exercise routines despite competing priorities. ⁵⁷ Studies further demonstrate that reducing response costs enables intervention recipients to eliminate factors undermining exercise adherence and strengthen management of interfering variables. For instance, for patients intending to walk to work but averse to sweating, mobilizing family members to prepare towels, water, and change of clothes has been effective. ⁴⁹ Minimizing disruptive factors significantly enhances individual motivation and adherence to physical activity. Additionally, digital tools such as WeChat pedometers, smart bracelets, and mobile apps can be used to quantify physical activity (steps, duration, calorie expenditure) and support self-check-in, heart rate monitoring, and YOGA-DP journaling for self-monitoring.^41,44,58 Meanwhile, sharing steps or check-ins on social media feeds promotes peer interaction and fosters social support through collective encouragement.^45,59 These approaches facilitate the overcoming of practical obstacles through adaptive exercise planning, ultimately optimizing engagement experiences and fostering long-term physical activity sustainability. A recurring theme among individuals with prediabetes was the expressed need for easily accessible exercise options, facilities, and equipment. Physical activity participation was frequently deemed impractical when fitness facilities or activity-friendly public spaces (e.g., parks) were geographically distant from residences or workplaces.^34,40 Additional barriers included inadequate access to safe exercise environments and adverse weather conditions, which collectively diminished exercise initiation and maintenance. Financial costs emerged as a significant barrier to exercise engagement. The expenses associated with accessing exercise facilities and enrolling in exercise programs consistently deterred participation. ⁶⁰ Given the socioeconomic constraints and living conditions of certain populations, healthcare professionals can help mitigate environmental barriers by designing home-based physical activity programs. ⁴³ Concurrently, workplace and community settings should also play important roles in promoting exercise engagement through proactive improvements to exercise environments and the provision of accessible public fitness infrastructure. ³⁴ A notable finding was the identification of linguistic and cultural disparities as obstacles to exercise education. Because this theme was less frequently reported than major environmental barriers such as time and access, it should be interpreted more cautiously, although it may still be highly relevant for specific populations. Some women reported feeling comfortable in informal settings where their native language is spoken, and they perceived communication barriers with healthcare teams as impediments to exercise adherence. This issue has rarely been addressed in prior studies, though this finding aligns with previous research. ⁶¹ Variations in perceptions and preferences toward exercise modification across demographic groups, influenced by age, gender, and income levels, ⁶² underscore the importance for healthcare providers to develop personalized exercise prescriptions aligned with individuals’ physiological profiles and sociocultural contexts. Tailored interventions enhance feasibility, engagement willingness, and long-term adherence. ⁵³

Within the social opportunity dimension, support from others emerged as the most frequently cited facilitator. Peer, familial, and social support were emphasized as critical enablers,^42,63 corroborating prior evidence that professional guidance and supervision from healthcare providers significantly improve adherence. ⁶⁴ Given that these facilitators were consistently identified across multiple studies, they may be regarded as relatively strong targets for intervention development. However, three studies further highlighted that lack of trust in health professionals was identified as a significant barrier. ³⁸ As this finding was supported by a comparatively smaller number of studies, it should be interpreted with some caution, although it still suggests that patient–provider trust may be an important issue in certain contexts. This barrier can be addressed through the provision of exercise-related knowledge training, the enhancement of effective communication, and the delivery of motivational encouragement to improve patient trust. ⁴² Family-centered interventions, including educational workshops and collaborative exercise participation, have been shown to increase patient motivation. ⁴⁹ Additionally, exercise engagement is further promoted by family-mediated reminders and monitoring, which help individuals maintain consistent participation. Several studies documented gender-and age-related sociocultural norms that restrict the types of physical activities deemed appropriate for women, particularly in male-dominated exercise environments. ⁵⁶ Cultural and religious barriers were also identified, with Islamic women reporting constraints due to modesty requirements such as hijab-wearing, which limited their ability to use outdoor fitness equipment in uncovered parks. ⁴⁰ Additionally, some women avoided exercising in public spaces when men were present. These findings align with prior research highlighting similar sociocultural restrictions.^23,65 Although these findings were not among the most frequently reported determinants in this review, they may still be particularly relevant when designing culturally sensitive interventions for specific subgroups. Therefore, providing women with more safe and easily accessible recreational activities can contribute to enhancing their overall activity levels, thereby addressing gender disparities in health outcomes.

Motivation in COM-B: Barriers and corresponding BCTs

Motivation is defined as all those brain processes that energize and direct behavior, not just goals and conscious decision-making. ¹⁸ This study identified factors related to reflective motivation (social/professional role and identity, beliefs about capabilities/consequences, intentions, and motivation and goals) and automatic motivation (reinforcement, emotions).

Reflective motivation, being a conscious process of planning and evaluation grounded in beliefs regarding good and bad, constitutes a key dimension within this domain. The expertise of professionals in exercise guidance emerged as the sole identified facilitator in this domain. Professional integrity among certified yoga instructors, coupled with clients’ trust in their trainers’ competence, ⁴¹ demonstrated significant potential to enhance exercise adherence through role-specific behavioral reinforcement. Multiple studies identified self-efficacy as a major determinant of physical activity behaviors. ⁵¹ Self-efficacy is defined as an individual’s belief in their ability to successfully complete a task or manage specific situations. ⁶⁶ It reflects confidence in one’s capacity to execute behaviors necessary to achieve desired outcomes. Within self-determination theory, motivational sustainability arises when perceived achievement aligns with competence needs, fostering autonomous regulation that perpetuates behavior maintenance. Higher motivational levels correspond to stronger exercise adherence intentions, and this effect is potentiated when combined with setting health achievement goals, exerting a more positive influence on physical activity engagement.^46,67 Tailored coaching and verbal encouragement have been shown to be effective in enhancing self-efficacy.^36,68 Conversely, laziness and unrealistic goals can act as factors impeding exercise.^31,32 Research indicates that breaking large goals into smaller subgoals improves feasibility, ⁶⁷ and healthcare professionals should assist in developing detailed exercise plans specifying frequency, intensity, and duration parameters.^45,51 Tailoring such targets to individual baseline fitness levels and clinical profiles may improve exercise adherence.^46,59 To address motivational barriers like laziness, studies suggest strategies such as personal comparisons of weight loss outcomes before and after intervention or group-based initiatives for mutual encouragement and feedback.^37,52,57 These approaches amplify intrinsic motivation and create accountability, working in conjunction with structured goal-setting frameworks to enhance exercise engagement. Within the domain of belief about consequences, positive outcome expectations regarding exercise benefits are associated with enhanced physical activity adherence, ⁶⁹ while fear of exercise-induced injury or discomfort emerges as a prevalent barrier. ⁵⁶ Although these findings were supported across several studies, the strength of support appeared to vary by subtheme; therefore, interpretation should consider some variation in the breadth of the underlying evidence. One approach to addressing this barrier is for physicians to inform patients that prediabetes elevates the risk of developing diabetes and cardiovascular disease, while emphasizing the benefits of exercise in reducing diabetes incidence. ⁵²

Autonomous motivation encompasses innate emotional responses, impulses, and desires that operate independently of deliberate cognitive processing. Five studies identified habitual integration of exercise into daily routines as a key driver of sustained physical activity, ⁴⁴ suggesting that habit formation may be an important facilitator of long-term adherence within the current evidence base. However, two studies highlighted that exercise-induced musculoskeletal discomfort (e.g., muscle soreness, joint pain) acted as barriers to continued participation.^34,39 As this barrier was identified in only a small number of studies, it should be interpreted cautiously, although it may still be clinically important for specific subgroups. The studies included in this review have found that such barriers can be addressed through health education focusing on exercise precautions. Innovatively, a “negative case” behavioral intervention model has been designed in prior research, aiming to educate participants via real-life scenarios of exercise-related risks. ³⁸ This approach can enhance self-protection awareness and alleviate fears of exercise-induced discomfort, thereby bridging the gap between intention and action in physical activity engagement. Intrinsic facilitators such as exercise enjoyment, perceived health benefits, and positive affective states were strongly associated with adherence,^6,58 whereas anxiety, disinterest, or exercise-induced discomfort and pain diminished compliance.^42,47 Compared with the major barriers and facilitators more consistently identified across studies, these emotional determinants should be interpreted with somewhat greater caution because the supporting evidence was narrower. In this study, negative emotions can be alleviated through verbal encouragement or gift rewards for exercise adherence, aiming to stimulate participants’ enthusiasm and promote exercise compliance.^45,67

Implications for practice

This review examines the barriers and facilitators influencing exercise adherence in individuals with prediabetes, along with theory-informed strategies to enhance compliance. Exercise adherence, as a pivotal indicator of behavioral change, is shaped by multifactorial influences encompassing individual capacity (e.g., physical fitness), social influences (e.g., familial support), and motivational drivers (e.g., self-efficacy). Incorporating these determinants into tailored interventions may help improve long-term feasibility and engagement. Nevertheless, balancing heterogeneous determinants and implementing integrated strategies remains a challenge in both research and clinical practice. These findings may help healthcare professionals and intervention designers prioritize determinants and strategies that appear most relevant within the current evidence base. However, their practical application should be considered in context, including local resources, service settings, and individual patient needs, and should not be interpreted as prescriptive recommendations.

Recent advancements in artificial intelligence (AI) and digital health technologies have introduced novel possibilities for exercise interventions. By integrating mHealth technologies and wearable devices, researchers can monitor physical activity behaviors in real-time while delivering personalized feedback and guidance. ⁷⁰ Additionally, virtual reality (VR) and augmented reality (AR) technologies have been utilized to create immersive exercise experiences that enhance patient engagement and motivation. ⁷¹

The studies included in this review predominantly originate from developed nations and rely on conventional intervention approaches, with limited incorporation of eHealth strategies. Future research should prioritize investigations into exercise adherence among prediabetic populations in low-and middle-income countries (LMICs) and evaluate the applicability and effectiveness of AI-driven interventions across diverse cultural and socioeconomic contexts. Furthermore, the determinants and intervention strategies identified in this review require validation through high-quality RCTs. Subsequent studies should elucidate the causal mechanisms underlying behavioral change and explicitly document the technical and methodological frameworks employed, thereby establishing a robust scientific foundation for evidence-based exercise interventions.

Strengths and limitations

The strength of this review lies in its systematic methodology for identifying relevant studies across published literature, encompassing all study designs (quantitative, qualitative, and mixed-methods) and integrating both deductive and inductive analytical approaches. The applied theoretical frameworks not only facilitated the analysis of barriers and facilitators but also enabled the linkage of the TDF to evidence-based BCTs via the BCW. This process identified actionable intervention functions for improving exercise adherence in prediabetic populations, thereby bridging theoretical determinants to practical implementation strategies.

This study has several limitations. First, the inclusion of only English and Chinese publications may have omitted relevant evidence published in other languages, which may have introduced selection bias and limited the generalizability of the findings. In addition, although gray literature was searched, the possibility of publication bias and incomplete retrieval of relevant studies cannot be excluded. Studies reporting null, negative, or less prominent findings may have been underrepresented, which may have affected the completeness of the evidence base and the relative prominence of some synthesized determinants and intervention strategies. Second, as the majority of included studies were conducted in developed countries, the generalizability of our findings to developing regions requires further validation. Third, the application of the TDF presented significant challenges in the precise categorization of key determinants due to interdisciplinary conceptual overlaps, which may introduce classification biases. Prior research has explicitly acknowledged these taxonomic difficulties, ⁷² emphasizing the need to refine the dimensional design and logical architecture of the framework to systematically enhance its explanatory power and practical utility in complex behavioral contexts. Finally, although dual independent coding and consensus procedures were used to enhance the reliability of the data transformation process, qualitizing quantitative findings into qualitative constructs necessarily involved interpretive judgments that may extend beyond the original statistical results. This represents an inherent epistemological limitation of the convergent integrated approach and should be considered when interpreting the synthesized findings.

Conclusion

This review highlights the key determinants of exercise adherence in individuals with prediabetes. However, these determinants were supported by evidence of varying breadth and methodological quality. The synthesis identifies “lack of time,” “lack of appropriate exercise equipment or venues,” and “health-related physical limitations” as predominant barriers. Conversely, “perceived health benefits,” “social support,” and “tools or methods to help with exercise adherence” emerged as key facilitators. Together, these more consistently identified barriers and facilitators may provide a more reliable basis for intervention development. In contrast, less frequently reported determinants should be interpreted more cautiously and warrant further empirical validation. Future interventions should consider prioritizing theory-informed strategies that integrate specific BCTs, including goal setting, action planning, and social support. Further rigorous empirical validation, particularly through randomized controlled trials, is needed to evaluate the efficacy of these targeted strategies and to inform technology-enabled personalized management.