Home Medicine The role of exercise modality on psychological, behavioral, and fitness outcomes among individuals at risk of type 2 diabetes: preliminary evidence from the CHOICE pragmatic randomized trial
Article Open Access

The role of exercise modality on psychological, behavioral, and fitness outcomes among individuals at risk of type 2 diabetes: preliminary evidence from the CHOICE pragmatic randomized trial

  • Alexandre Santos , Kaja Falkenhain , Jonathan P. Little , Nikhil R. Patel , Joel Singer , Frank Halperin , Kevin Pistawka and Mary E. Jung EMAIL logo
Published/Copyright: December 1, 2025
Download Article (PDF)
Translational Exercise Biomedicine
From the journal Translational Exercise Biomedicine Volume 2 Issue 4

Abstract

Objectives

To determine whether providing a choice between high-intensity interval training (HIIT) and moderate-intensity continuous training (MICT) within a 4-week diabetes prevention program may lead to greater perceived autonomy support, motivation regulation, free-living physical activity, and cardiorespiratory fitness 6 months post-intervention when compared to imposed exercise.

Methods

In a preliminary pragmatic randomized trial, 77 individuals at risk of type 2 diabetes (M age=61.5 (±9.8) years; N females=58) were randomized to one of three exercise conditions: HIIT, MICT, or the choice thereof (CHOICE). Perceived autonomy support was assessed post-intervention. Changes in motivation, physical activity, and cardiorespiratory fitness were assessed 6 months post-intervention. Linear mixed models and Bonferroni-adjusted pairwise comparisons on estimated marginal means were used to derive effect estimates after adjusting for stratified allocation factors.

Results

Perceived autonomy support was not different among conditions [F (2, 47)=0.068, p=0.934]. No effects were detected for motivation regulation, physical activity, or cardiorespiratory fitness (ps>0.05). Participants in the CHOICE condition self-reported significantly more physical activity 6 months post-intervention compared to pre-intervention [t (31)=2.922, p=0.019]. Improvements in cardiorespiratory fitness were seen in CHOICE [t (65)=2.509, p=0.044] and MICT [t (65)=3.492, p=0.003].

Conclusions

Providing choice between HIIT and MICT did not significantly affect individuals’ perceived autonomy support or motivation regulation compared to imposed exercise. However, physical activity and cardiorespiratory fitness improved over time for the CHOICE condition. Providing choice between HIIT and MICT may be a feasible exercise strategy among this population.

Keywords: behavior change; high-intensity interval training; moderate-intensity continuous training; pragmatic randomized trial; prediabetes; self-determination theory

Introduction

The reduction of type 2 diabetes (T2D) incidence and its associated complications by increasing physical activity behavior among individuals at risk of T2D has been well documented [1]. Increases in weekly moderate-to-vigorous physical activity (MVPA) in this population can lead to decreased body mass, lower blood glucose, and improved cardiorespiratory fitness [1]. Diabetes prevention programs (DPPs) are a commonly used strategy for fostering changes in physical activity behavior, yet such programs have traditionally been costly and resource-intensive, thus impeding the translation of findings for use in “real-world” community settings [2]. Coupled with relatively low physical activity levels among individuals at risk of T2D compared to the general population [3], there is a clear need for pragmatic trials on DPPs to discern potential strategies that increase physical activity in real-world community settings. Indeed, the call for more pragmatic trials has been documented for decades [4], with outcomes being directly relevant to participants, communities, healthcare practitioners, and decision-making processes relevant to development of new guidelines and interventions [5].

Physical activity behavior change is complex. Numerous theories exist that aim to explain and predict changes in physical activity behavior, with considerable overlap among them [6]. Michie and colleagues [7] propose that among these theories, behavior is driven by three overarching constructs: one’s capabilities, opportunities, and motivations to perform such behavior. This model, known as COM-B, can act as a framework for the development of pragmatic health behavior interventions. COM-B is also useful in its direct recommendation of which specific behavior change techniques (BCTs) should be targeted within an intervention to drive changes in behavior such as physical activity [7], 8]. Targeting specific BCTs within DPPs for the promotion of free-living physical activity behavior may therefore be a tangible strategy in real-world settings.

Of relevance to this study, BCTs that target individuals’ motivations to engage in physical activity may hold promise in changing behavior. Self-determination theory (SDT) is a theory that explains the relationship between motivation and behavior [9]. More autonomous forms of motivation regulation are consistently positively related to healthy behaviors, including physical activity [10]. A sub-theory of SDT, Basic Psychological Needs (BPN) postulates that fostering three basic psychological needs of autonomy, competence, and relatedness leads to better autonomous motivation regulation, more positive affect towards physical activity, and improved long-term physical activity engagement [10], [11], [12], [13]. Taken together, the promotion of the three BPNs by using BCTs related to motivation in DPPs may prove to be a useful pragmatic strategy to increase individuals’ autonomous motivation regulation, which in turn may lead to better physical activity in free-living conditions and improvements in cardiorespiratory fitness.

A recent article by Teixeira and colleagues [14] classified potential BCTs that may influence each of the three BPNs. In their process model adapted from the original model by Ryan and colleagues [12], these motivation-based behavior change techniques (MBCTs) are thought to specifically support the needs of autonomy, competence, and relatedness, which are hypothesized to subsequently increase autonomous motivation regulation, the degree of participation in health behaviors (i.e., physical activity), and ultimately cardiorespiratory fitness. Of interest, the provision of choice is listed as an MBCT that has the potential to foster one’s sense of autonomy, thereby leading to increases in perceived autonomy support [14]. In applying this to efforts in increasing physical activity within DPPs, providing a choice between different exercise modalities and intensities could be more conducive to fostering perceived autonomy, autonomous motivation regulation, physical activity, and cardiorespiratory fitness than the more traditional prescriptive methods commonly used in DPPs.

High-intensity interval training (HIIT) is an exercise modality that has recently gained considerable momentum as a potential option in addition to traditional moderate-intensity continuous training (MICT) exercise protocols. Despite cumulative research demonstrating comparable psychological and physiological responses to both exercise modalities [15], with further evidence suggesting additional benefits for HIIT protocols compared to MICT (i.e., waist circumference, percent fat mass, maximal volume of oxygen consumption [V̇O2max]) [16], a debate in the literature exists as to whether HIIT is an appropriate option for populations that are insufficiently active [17]. On one side, a recent narrative opinion by Ekkekakis and Biddle [18] argues that individuals are unable to maintain adherence to vigorous-intensity exercises for a prolonged period (>12 months) and that HIIT protocols do not induce higher free-living physical activity rates when compared to MICT. Therefore, the use of HIIT as a public health strategy should be called into question. On the other hand, a systematic review and meta-analysis by Santos and colleagues [19] including 188 studies found no differences in free-living physical activity adherence between HIIT and MICT protocols among individuals who were insufficiently active or living with a medical condition, with adherence rates for both conditions being above 60 % of prescribed sessions, suggesting that both exercise modalities are equally viable options. However, the need for more research was also noted by the authors over concerns of robustness, heterogeneity, and quality of the evidence.

Instead of pitting one exercise modality against the other, a tactical solution to enhance autonomy within pragmatic interventions such as DPPs may be to provide the choice between HIIT or MICT. Irrespective of the side of the debate, both perspectives acknowledge the potential importance of providing choice, with Ekkekakis and Biddle [18] stating “Notwithstanding the importance of allowing people choice of exercise modalities and intensities…” and Santos and colleagues [19] stating that “… perhaps providing a choice between [HIIT or MICT] may prove optimal when developing physical activity recommendations and designing exercise interventions.” Furthermore, the need for the integration of behavioral techniques (i.e., BCTs) within pragmatic interventions for the promotion of free-living physical activity has been repeatedly expressed from both sides.

The purposes of this study were to determine whether the provision of choice between HIIT and MICT within a pragmatic DPP would lead to increases in individuals’ perceived autonomy support post-intervention, more autonomous forms of motivation regulation, and improvements in 6-month free-living physical activity and cardiorespiratory fitness when compared to the prescription of either HIIT or MICT only. Based on the theoretical underpinnings of SDT [9], 11], 12] and the adapted process model by Teixeira and colleagues [14], we hypothesized that, compared to being prescribed a particular type of exercise, individuals at increased risk of developing T2D who were given the choice in exercise type would report greater perceived autonomy support after a DPP, as well as increases in motivation regulation, free-living physical activity, and cardiorespiratory fitness 6 months post-intervention. The summary of this article is presented in Figure 1.

Figure 1: Graphical representation of this study. Key points: (1) this pragmatic randomized trial among adults at risk of type 2 diabetes (n=77) explored whether providing a choice between two exercise modalities (HIIT and MICT) within a community diabetes prevention program would yield favorable outcomes. (2) Providing a choice in exercise modality significantly increased cardiorespiratory fitness and self-reported physical activity behavior. Perceived autonomy support and motivation scores were high for all exercise modalities. (3) This study translates research evidence into pragmatic settings by addressing the ongoing uncertainty of optimal exercise prescription for individuals at risk of type 2 diabetes. It may also support future physical activity guidelines and intervention development. Figure created with BioRender.
Figure 1:

Graphical representation of this study. Key points: (1) this pragmatic randomized trial among adults at risk of type 2 diabetes (n=77) explored whether providing a choice between two exercise modalities (HIIT and MICT) within a community diabetes prevention program would yield favorable outcomes. (2) Providing a choice in exercise modality significantly increased cardiorespiratory fitness and self-reported physical activity behavior. Perceived autonomy support and motivation scores were high for all exercise modalities. (3) This study translates research evidence into pragmatic settings by addressing the ongoing uncertainty of optimal exercise prescription for individuals at risk of type 2 diabetes. It may also support future physical activity guidelines and intervention development. Figure created with BioRender.

Materials and methods

The CHOICE pragmatic trial was a prospective 3-arm, multi-center, parallel-condition unblinded randomized trial conducted in a western city of Canada. Reporting of this trial followed the CONSORT 2010 guidelines for reporting randomized controlled trials and its extension for reporting pragmatic trials [5]. Ethical approval was granted by a university’s Clinical Research Ethics Board (H16-02028-A021). The datasets produced and analyzed during the current study are available from the corresponding author upon reasonable request.

Participants

Potentially eligible participants for this trial were referred by their family physician to partake in a community-based DPP (Small Steps for Big Changes) if they had a hemoglobin A1c (HbA1c) test within the American Diabetes Association’s (ADA) prediabetes range of 5.7–6.4 % in the last 6 months [20]. Potential participants who did not have a recent HbA1c test result but scored a 5 or higher on the ADA diabetes risk assessment were also invited to participate [20]. Advertisements for recruitment included pamphlets in healthcare offices, website adverts, social media posts, and word of mouth. Potential participants were screened by a member of the research team via phone to ascertain if they met the following inclusion criteria: between 18 and 75 years of age, engaged in two or fewer bouts of purposeful physical activity per week in the last 6 months, had a recent HbA1c test result between 5.7 and 6.4 % or an ADA risk assessment score≥5, and were cleared to engage in physical activity as determined by the Get Active Questionnaire [21]. Participants were excluded from this trial if they had a previous diagnosis of T2D, were taking glucose-lowering medications or beta-blockers, had a history of cardiovascular disease, were diagnosed with uncontrolled hypertension (resting blood pressure>160/90), or had explicit contraindications to exercise. All participants provided written informed consent prior to enrollment.

Estimation of cardiorespiratory fitness

Participants who met the eligibility criteria and provided consent subsequently completed a predicted maximal exercise stress test at a local cardiac rehabilitation clinic supervised by a certified exercise physiologist prior to beginning the intervention. Depending on each participant’s baseline physical ability and/or comfort level with treadmills, the exercise physiologist employed a Bruce or Modified Bruce protocol for the stress test [22], 23]. Protocol stages were performed on a treadmill and incremented sequentially every 3 min until participants reached volitional exhaustion (i.e., either by reaching their predicted maximum heart rate [HRmax] based on 211 – 0.64·age [24] or a perceived rating of exertion [RPE]≥18) [25], or the test was stopped by the exercise physiologist according to safety cut-offs (blood pressure>250/120, oxygen saturation<82 %, physical concerns). Participants’ heart rate, blood pressure, oxygen saturation, and RPE were monitored at each stage of the protocol. Stress test results were reviewed by a cardiologist to determine whether exercise was safe for the individual or whether any underlying cardiac issues were present that needed further attention. Additionally, results from the stress test were used to obtain each participant’s resting heart rate and HRmax to calculate individualized exercise intensity zones for the intervention. This test also served to estimate cardiorespiratory fitness based on the total test duration (details found in outcomes section below). The stress test was repeated with each participant 6 months post-intervention to assess changes in cardiorespiratory fitness.

Small Steps for Big Changes

Participants that were cleared to exercise by a cardiologist were enrolled in Small Steps for Big Changes (SSBC), a 4-week community-based DPP. The full details of SSBC, including its efficacy and effectiveness in laboratory and community settings have been published previously [26], [27], [28], [29]. Briefly, SSBC consists of six one-on-one sessions with a trained community coach (i.e., fitness professionals working at a recreation facility) over a 4-week period, delivered either virtually or in-person at one of three community recreation facilities. Each session entails a brief behavioral counselling component during which trained coaches use motivational interviewing-informed techniques to foster participants’ self-efficacy and determination for changing and maintaining healthy nutritional and physical activity behaviors. Information pertaining to specific content delivered during sessions and fidelity of program delivery by community coaches has been previously published [30], [31], [32], [33]. Community coaches subsequently follow up with participants 1, 6, and 12 months after the 4-week supervised intervention.

In addition to behavioral counselling, SSBC participants also complete supervised progressive exercise during each of the six sessions. For this trial, SSBC participants provided additional consent for a sub-study for which they were randomly allocated to one of three conditions: HIIT was introduced and performed for all 6 sessions (HIIT-only); MICT was introduced and performed for all 6 sessions (MICT-only); and HIIT and MICT were introduced during sessions 1 and 2 in a randomized counterbalanced order, and participants were provided the choice of which exercise to perform during sessions 3–6 (CHOICE). In all three conditions, participants were able to choose whether to exercise on a treadmill, exercise bicycle, or elliptical. Exercise duration for each session was designed to elicit matching work volumes based on prescribed intensities for both HIIT and MICT during the progressive nature of the intervention. For example, a participant’s total work done for session one of HIIT at the prescribed intensity theoretically equated to that participant’s total work done for session one of MICT at the prescribed intensity [26], 27]. The use of progressive exercise was intended to help familiarize participants with exercise modalities they may not have been familiar with and/or ease the introduction of exercise to those who were previously insufficiently active. Exercise was individualized for each participant by prescribing exercise intensities based on their measured resting heart rate and HRmax from the stress test using the Karvonen formula [34]. Exercise intensity compliance was measured during each exercise session via heart rate monitors [Polar H10, Polar Electro] and participants’ RPE [25]. Exercise intensity was recorded by the supervising community coach at the beginning and at 25 , 50, and 100 % of exercise completion, and coaches encouraged participants to adjust speed, resistance, or incline if target heart rate was not achieved. Following the 4-week supervised intervention, participants in all three conditions were encouraged to continue performing MVPA on their own according to Canada’s physical activity recommendations for adults [35] but were not told to exclusively perform either HIIT or MICT due to the pragmatic nature of the SSBC program and this trial. Participants’ physical activity in free-living conditions was tracked using two objective measures and one self-report measure. Exercise intensity was guided by one of the objective measures via heart rate. Apart from condition allocation, all other aspects of the DPP were identical for all three conditions.

High-intensity interval training

Participants who were randomized to the HIIT-only condition began each exercise session with a 3-min warmup at a comfortable self-selected pace. During sessions 1 and 2, participants completed five 30-s intervals>80 % HRmax interspersed by four 60-s periods of light recovery at a self-selected pace. For sessions 3 and 4, duration of high-intensity intervals increased from 30 to 45 s, and for sessions 5 and 6, duration was increased again to 60 s, so that by the end of the supervised intervention, participants were completing five 60-s intervals interspersed by four 60-s recovery periods. All sessions concluded with a 2-min cooldown. Including warmup and cooldown, the total duration of sessions ranged between 11.5 and 14 min for sessions 1 to 6, respectively. The HIIT protocol employed was based on previous findings [26], 27] and has been shown to elicit desired intensity targets (>80 % HRmax) despite the low volume of 60-s intervals [28]. Progression was modified in this study from 60 to 30 s to ease previously insufficiently active participants into high-intensity training. Despite this, intensity checks throughout each session and subsequent adjustments to the speed and/or incline ensured target intensity was reached by the second interval and remained elevated for the remaining intervals.

Moderate-intensity continuous training

Participants who were randomized to the MICT-only condition exercised continuously for 20 min at an intensity between 60 and 80 % HRmax during sessions 1 and 2. Duration of exercise was increased to 25 min for sessions 3 and 4, and to 30 min for sessions 5 and 6. There were no warmup or cooldown periods, resulting in total session durations ranging from 20 to 30 min for sessions 1 to 6, respectively.

Choice

Participants who were randomized to the CHOICE condition were introduced to either HIIT or MICT during session 1 and introduced to the other exercise modality during session 2. The introduction order of HIIT or MICT was randomized in a counterbalanced fashion to diminish potential anchoring bias. HIIT and MICT protocols were the same as in the other two conditions, and progression of exercise followed the same convention.

Outcomes

Exercise modality frequency

To determine the modality of exercise that participants engaged in during the 6-month follow-up period, a 2-item self-report questionnaire was created and administered to a sub-set of individuals in each group. The Exercise Modality Frequency Questionnaire (EMFQ) prompted participants with the passage, “In the past 4 weeks, on average, how many sessions of [HIIT or MICT] did you do each week?” A definition of each exercise modality was provided and examples of each type of exercise were included (e.g., HIIT: walking/running up and down hills; MICT: walking on flat ground at a moderate pace, etc.). For participants who indicated engaging in HIIT or MICT at least once a week, an open text box was provided to describe the nature of their exercise (e.g., walking around the neighborhood with the dog). The EMFQ was administered at the 6-month follow-up appointment as a manipulation check.

Perceived autonomy support

Individual participants’ rating of perceived autonomy support was collected at the end of the 4-week intervention phase using the Learning Climate Questionnaire (LCQ) adapted by Williams and Deci [36]. The LCQ is a 15-item self-report questionnaire that measures perceived autonomy support and has been previously validated and shown to have strong internal reliability [36], 37]. The questionnaire uses a 7-point Likert scale ranging from 1 “strongly disagree” to 7 “strongly agree”. For this study, the anchoring questions were modified to replace the term “instructor” with “coach”. Scoring of individual results was done by averaging individual responses after reverse-coding item #13. Participants answered the LCQ by completing the form online at the end of their last supervised session (post-intervention) with their community coach.

Motivation regulation

Participants’ motivation regulation was measured using the Behavioral Regulation in Exercise Questionnaire-2 (BREQ-2) [38]. The BREQ-2 is a 19-item self-report questionnaire that measures one’s motivation to exercise according to the stages of motivation presented in SDT’s continuum of the locus of causality [9], including amotivation, external, introjected, identified, and intrinsic forms of regulation. The questionnaire uses a 5-point Likert scale ranging from 0 “not true for me” to 4 “very true for me”. BREQ-2 has been previously validated in a variety of settings including physical fitness [39]. For this study, the relative autonomy index (RAI) was used to provide an index of the degree to which participants felt self-determined ranging from −24 to +20, with higher positive scores indicating more autonomous forms of motivation regulation. The RAI was calculated by first averaging each subscale, then multiplying each average by a subscale weighting, and subsequently summing the weighted subscale scores. Weighting for each subscale were as follows: amotivation (−3), external regulation (−2), introjected regulation (−1), identified regulation (+2), intrinsic regulation (+3). The BREQ-2 was administered through an online form at three timepoints: at the beginning of the first session with the community coach (pre-intervention), at the end of the last session (post-intervention), and 6 months post-intervention.

Free-living physical activity

Free-living physical activity was assessed using three different methods to assist with reported variability in physical activity behavior depending on how it is measured [40]. The Godin Leisure-Time Exercise Questionnaire (GLTEQ) was used to assess self-reported physical activity adherence [41]. The GLTEQ requires individuals to report the frequency of bouts of at least 15 min of strenuous, moderate, and mild exercise during a typical week. Scoring of the GLTEQ consists of first multiplying reported frequencies by their corresponding metabolic equivalent (MET) values (strenuous: 9 METs, moderate: 5 METs, mild: 3 METs), and then summing the products for a total score. Interpretation of the scores follows the categories proposed by Godin, where a score of 24 units or higher indicates an individual is active, a score between 14 and 23 indicates an individual is moderately active, and a score below 14 indicates an individual is insufficiently active [41]. For this study, self-reported frequencies of moderate and strenuous activity were also used separately to estimate weekly engagement in free-living physical activity at each intensity. The GLTEQ has been previously validated and shown to be comparable to accelerometry data for measuring MVPA [42]. The GLTEQ was administered to each SSBC participant at the beginning of the first intervention session and 6 months post-intervention via an online form.

Free-living physical activity was also measured using 7-day triaxial accelerometry data. Validity and accuracy of accelerometry data in free-living conditions has been previously reported [40]. Participants were asked to wear an Actigraph GT3X-BT for seven consecutive days at the top of their right hip at a sampling rate of 100 Hz. Participants were instructed to wear the device throughout the waking hours of the day and remove the device during sleeping hours. Accelerometers were also removed during water-based activities. Accelerometry data was considered valid if the participant wore the accelerometer for a minimum of 10 h/day and at least 4 of the 7 days of the week, consistent with parameters previously recommended [43]. Non-wear time was defined as any period of 60 min or longer with no recorded activity counts. Physical activity was measured by summing activity counts in 60-s epochs, and cut-off activity count thresholds for defining intensity followed Troiano and colleagues’ convention [44]: sedentary (0–99), light (100–2,019), moderate (2,020–5,998), and vigorous (5,999 and above). Free-living physical activity was operationalized as average daily minutes of MVPA and was calculated by summing both moderate and vigorous activity count minutes and subsequently dividing by the number of valid wear days. Daily minutes of MVPA was used instead of total weekly MVPA minutes to account for individual differences in valid wear days. Wear-time analysis and scoring of data were performed using Actilife v6.13.4 (Actigraph LLC, Pensacola, FL, USA). Accelerometry data was collected for two 7-day periods (the week prior to the intervention and the week following the 6-month follow-up appointment).

A wearable activity tracker (Fitbit Luxe™) was also provided to each participant to track physical activity behavior in free-living conditions. The validity and utility of using different wearable activity trackers, including certain models of Fitbits, for tracking physical activity has been previously summarized [40]. However, we are not aware of any study validating the accuracy of the Luxe™ model since it is relatively new. Minute-by-minute physical activity data was captured by the devices and imported into the Fitabase software (Small Steps Labs LLC, San Diego, CA, USA) whenever participants synced their devices to the Fitbit mobile application. In congruence with previous research [45] and similar to accelerometry-based parameters, we defined a valid wear day as any day with a minimum of 10 h of wear time determined by continuous minute-by-minute heart rate recordings. Non-wear time was defined as a period of 60 continuous minutes with no heart rate recordings. A minimum of 4 out of 7 consecutive days of the week were required for the week to be considered valid. Wear-time analysis was conducted using an R package created by a member of the research team. Free-living physical activity was operationalized as average daily minutes of MVPA and was calculated by summing the daily minutes spent in moderate and intense activities per day and subsequently dividing by the number of valid wear days. Activity intensity levels were determined by using the Fitbit-derived algorithms, and number of minutes per day at each level were imported into Fitabase. Although the algorithms are proprietary, Fitbit provides crude MET equivalents [46] for each activity intensity (sedentary: <1.5 METs, light: between 1.5 and 3 METs, moderate: between 3 and 6 METs, intense: >6 METs). For consistency in reporting and comparing, Fitbit data were collected during the same 7-day periods as accelerometry data for each individual pre-intervention and 6 months post-intervention.

Cardiorespiratory fitness

Participants’ cardiorespiratory fitness was estimated at baseline and 6 months post-intervention based on the time participants were able to sustain in the Bruce Treadmill exercise stress test performed at the cardiac rehabilitation clinic [22]. Calculation of cardiorespiratory fitness was conducted according to Bruce and colleagues’ equation for estimated V̇O2max in mL kg−1 min−1 and has been shown to account for 85.5 % of explained variation compared to observed V̇O2max [22]. For those who performed a Modified Bruce protocol [23], time spent in the modified stages was not included in the calculation of V̇O2max. Calculations were stratified by biological sex, where w is the weighting factor for sex (1=male; 2=female) and t is the duration of the protocol in seconds:

V ˙ O 2 max = 6.70 − 2.82 w + 0.056 t

Changes consequent to COVID-19

Recruitment for this trial commenced in February 2019 using a monthly rolling recruitment strategy (∼6–10 participants/month). Originally, follow-up appointments were planned to discern changes in the outcome variables 12 months post-intervention. However, at the beginning of the 12-month follow-up phase (February 2020), the COVID-19 pandemic resulted in the curtailment of all research activities at the host institution. A total of 75 participants were not able to complete follow-up assessments, and data were therefore lost for this first cohort. Recruitment for this trial could not be resumed until October 2021. To complete data collection in a timely manner given these unanticipated resource constraints, follow-up assessments were conducted 6 months post-intervention for this second cohort as opposed to the original 12-month timepoint. The complete loss of data on the first 75 participants to enroll in the trial and subsequent curtailment of research for more than 1.5 years therefore significantly limited the final study sample size.

Sample size

Limited studies have been conducted that assess perceived autonomy support after adults are provided a choice between different types of exercise in an intervention. A study by Lonsdale and colleagues [47] assessed the change in perceived autonomy support among adolescents enrolled in physical education programs. Comparisons between the control group (treatment as usual) and a group that was provided choice in what exercises were performed during class demonstrated a small-to-medium effect favoring the choice group (d=0.39). Based on these results, we anticipated a crude small-to-medium effect size in favor of the choice condition in the present study (d=∼0.40). Using a fixed effects one-way ANOVA at an alpha level of 0.05 and 80 % power, we estimated that 60 participants per condition (n=180) were required to detect significant between-condition differences in perceived autonomy support post-intervention. Based on pilot findings by Locke and colleagues [27], this sample size was then increased by 20 % to account for participant attrition, resulting in a total sample size of 216. The severe COVID-19 pandemic limitations impacted the ability to achieve the estimated sample size and data are thus reported for n=77.

Randomization

SSBC participants were randomly allocated into one of the three arms of this study in a 1:1:1 allocation ratio by using a computer random-number generator that produced variable permuted block sizes. Conditions were stratified for biological sex (male and female) and age (18–45 and 46–75 years) based on the potential effect of age on cardiorespiratory fitness [16]. Allocation sequencing was performed by a member of the research team not involved in any other aspect of this trial and allocation was concealed from participants until they began the intervention.

Blinding

Given the pragmatic nature of this trial and the fact that participants and community coaches needed to know which exercise modality to engage in during the exercise sessions, blinding of participants and those delivering the intervention was not possible. Data collected for each participant were blinded during data analysis by replacing the name of the allocated condition with a numerical digit known only to a member of the research team not involved in the data analysis phase so that data analysts would not be biased by condition allocation.

Statistical methods

Anthropometric and demographic information were summarized as means (standard deviations) for continuous data and n (%) for categorical data, unless otherwise stated. Medians and ranges are reported for 6-month exercise modality frequency to represent the central tendency and dispersion more accurately. Descriptive information was aggregated using SPSS (version 28).

Blinded data analysis was completed on an intention-to-treat basis as all randomized participants were included in the analyses irrespective of their program compliance. No missing data were imputed as per contemporary guidelines [48]. Using R (version 4.3.0) and associated packages (sjPlot, emmeans, effectsize, lmerTest, lme4), a linear model with restricted maximum likelihood estimation was used to detect a between-condition difference in perceived autonomy support post-intervention. The model included treatment condition (i.e., HIIT, MICT, or CHOICE) as a fixed factor, and stratified allocation factors (i.e., age and biological sex) as covariates. Similarly, linear mixed models were used to analyze changes in motivation regulation, physical activity behavior, and cardiorespiratory fitness between conditions. Linear mixed models included fixed effects for timepoint (pre-intervention, post-intervention, 6 months post-intervention), treatment condition, as well as the interaction thereof, stratified allocation factors, and a random effect for participants to address non-independence of measurements arising due to the repeated-measures design of this study. The intercepts and slopes of each individual were allowed to vary for each linear model. F-statistics were used as the omnibus tests to detect main and interaction effects in each linear model, and Bonferroni-adjusted preplanned pairwise comparisons of estimated marginal means were conducted to derive effect estimates and accompanying 95 % confidence intervals (CIs) after adjusting for included covariates. A two-sided alpha of 0.05 was used for all statistical analyses. Model assumptions were assessed visually using normal probability plots and residuals vs. fitted values plots (i.e., homoscedasticity). When assumptions were violated, a non-parametric equivalent test (i.e., Kruskal-Wallis test) was also performed to determine whether assumption violations had a significant impact on parametric test results.

Results

Recruitment for this trial began in October 2021, with the last 6-month post-intervention appointment occurring in July 2023. A flow of participants throughout this study is presented in Figure 2. Of the 77 participants randomized to one of the three intervention conditions, 74 completed the 4-week intervention. At the 6-month timepoint, two participants dropped out from the HIIT condition: one due to back surgery and another due to a cancer diagnosis, neither of which were related to this study. One participant dropped out from the CHOICE condition due to moving countries. One participant from each condition (n=3) was lost to follow-up. Including those who did not start the intervention (n=3), overall compliance to the 4-week supervised intervention was 96 % (18 %). Compliance to the supervised intervention for HIIT, MICT, and CHOICE conditions were 96 % (19 %), 96 % (20 %), and 97 % (17 %), respectively. Medians and ranges for each condition’s reported engagement in HIIT and MICT during the 6-month free-living period are summarized in Table 1. Baseline demographic and anthropometric characteristics of participants are summarized in supplementary file 1. No adverse events due to participation in this study were reported.

Figure 2: CONSORT flow diagram describing the process of participant enrollment, random assignment, attrition, and data analysis.
Figure 2:

CONSORT flow diagram describing the process of participant enrollment, random assignment, attrition, and data analysis.

Table 1:

Self-reported weekly frequency of HIIT and MICT engagement by condition 6-months post-intervention.

HIIT (n=14) MICT (n=9) CHOICE (n=14)
HIIT frequency per week 2.5 (0–4) 0 (0–4) 3 (1–4)
MICT frequency per week 3 (0–8) 5 (3–8) 3.5 (0–12)

  1. Values provided are medians (ranges). HIIT, high-intensity interval training; MICT, moderate-intensity continuous training.

Perceived autonomy support

Twenty-five participants did not have data for perceived autonomy support post-intervention. Based on the remaining 52 participants, assumptions of normality and residual homoscedasticity were violated. Given the robustness of a linear model design, coupled with the ability to include stratified allocation factors as covariates in the model, the linear model was still used as the main analysis approach. The linear model showed no main effect of condition after adjusting for stratified biological sex and age [F (2, 47)=0.068, p=0.934, η p 2 =0.003]. Estimated marginal means and accompanying 95 % CIs of perceived autonomy support for each condition are visualized in Figure 3. Pre-specified Bonferroni-adjusted between-condition pairwise comparisons based on effect estimates and accompanying 95 % CIs for perceived autonomy support and other outcomes are summarized in Table 2. A sensitivity analysis using the non-parametric Kruskal-Wallis test was conducted to determine whether any differences in results may occur due to assumption violations. Kruskal-Wallis test showed similar results to the linear model [χ 2 (2)=0.734, p=0.69].

Figure 3: Estimated marginal means of perceived autonomy support (n=52) measured post-intervention via the learning climate questionnaire [36]. Error bars represent 95 % confidence intervals. HIIT: high-intensity interval training; MICT: moderate-intensity continuous training; a.u.: arbitrary units.
Figure 3:

Estimated marginal means of perceived autonomy support (n=52) measured post-intervention via the learning climate questionnaire [36]. Error bars represent 95 % confidence intervals. HIIT: high-intensity interval training; MICT: moderate-intensity continuous training; a.u.: arbitrary units.

Table 2:

Effect estimates for within- and between-condition comparisons in primary outcomes.

Outcome HIITa (n=24) MICTa (n=24) CHOICEa (n=21) CHOICE-HIITb CHOICE-MICTc HIIT-MICTd
Perceived autonomy support, a.u. – – – −0.08 (−0.58–0.42) −0.01 (−0.51–0.50) 0.07 (−0.42–0.57)
Motivation regulation, a.u.
 Post-interventione 1.29 (−1.23–3.81) 1.70 (−0.76–4.15) 1.90 (−0.54–4.33) 0.60 (−2,90–4.11) 0.20 (−3.26–3.66) −0.40 (−3.92–3.12)
 6-Month follow-upe 0.57 (−2.14–3.27) 1.05 (−2.13–4.22) 1.73 (−0.76–4.23) 1.16 (−2.52–4.84) 0.69 (−3.35–4.73) −0.48 (−4.65–3.70)
MVPA (accelerometry), min· day−1 −7.59 (−17.1–1.87) −2.40 (−11.7–6.93) −4.34 (−14.4–5.69) 3.25 (−10.5–17.0) −1.94 (−15.6–11.8) −5.19 (−18.5–8.10)
MVPA (fitbit), min· day−1 −8.59 (−24.2–6.99) −7.66 (−22.5–7.19) −11.80 (−24.8–1.18) −3.21 (−23.5–17.1) −4.14 (−23.9–15.6) −0.93 (−22.4–20.6)
GLTEQ, a.u. 7.54 (−7.69–22.8) 15.94 (−2.19–34.1) 16.84 (2.25 – 31.4)f 9.30 (−11.7–30.4) 0.90 (−22.3–24.1) −8.41 (−32.0–15.2)
Cardiorespiratory fitness (V̇O2max), mL kg −1  min −1 1.23 (−0.70–3.16) 2.69 (0.80 – 4.58)f 2.06 (0.04 – 4.08)f 0.83 (−1.96–3.62) −0.63 (−3.39–2.14) −1.46 (−4.16–1.25)

  1. MVPA, moderate-to-vigorous physical activity; GLTEQ, godin leisure time exercise questionnaire; V̇O2max, relative maximal volume of oxygen consumption per minute; HIIT, high-intensity interval training; MICT, moderate-intensity continuous training; a.u.: arbitrary units. All data are presented as effect estimates (95 % CI) based on intention-to-treat analyses. aDenotes within-condition effect estimates compared to pre-intervention. bBonferroni-adjusted pairwise comparison between CHOICE, and HIIT, conditions. cBonferroni-adjusted pairwise comparison between CHOICE, and MICT, conditions. dBonferroni-adjusted pairwise comparison between HIIT, and MICT, conditions. eCompared to baseline. fDenotes a significant difference (p<0.05) at a two-sided alpha of 0.05.

Motivation regulation

Assumptions for the linear mixed model were met, and a main effect of time was detected (F (2, 77)=4.041, p=.021, η p 2 =0.10) after adjusting for stratified allocation factors. No main effects of condition or interaction were detected (ps>.05). Within-condition significance dissipated when pairwise comparisons were Bonferroni-corrected. Estimated marginal means and accompanying 95 % CIs at each timepoint are depicted in Figure 4.

Figure 4: Estimated marginal means of motivation regulation (n=68) measured pre-, post-, and 6-months post-intervention via the behavioral regulation in exercise Questionnaire-2 [38]. Error bars represent 95 % confidence intervals. HIIT: high-intensity interval training; MICT: moderate-intensity continuous training; a.u.: arbitrary units.
Figure 4:

Estimated marginal means of motivation regulation (n=68) measured pre-, post-, and 6-months post-intervention via the behavioral regulation in exercise Questionnaire-2 [38]. Error bars represent 95 % confidence intervals. HIIT: high-intensity interval training; MICT: moderate-intensity continuous training; a.u.: arbitrary units.

Free-living physical activity

Linear mixed model assumptions were met for all measures of physical activity. Estimated marginal means and accompanying 95 % CIs for each method of measurement are depicted in Figure 5. For accelerometry-derived MVPA, a main effect of time was detected [F (1, 53)=4.532, p=0.038, η p 2 =0.08] but no condition or interaction effect (ps>0.05). Similarly, a main effect of time was found for Fitbit-derived MVPA [F (1, 45)=7.692, p=0.008, η p 2 =0.15], with no condition or interaction effects (ps>0.05). Significance in both instances dissipated when within-condition pairwise comparisons were Bonferroni-corrected. A main effect of time was detected for self-reported physical activity adherence [F (1, 36)=13.128, p=0.0009, η p 2 =0.27], but no condition or interaction effects were found (ps>0.05). Based on Bonferroni-adjusted pairwise comparisons, only participants in the CHOICE condition self-reported being significantly more physically active 6 months post-intervention when compared to baseline [t (31)=2.922, p=0.019].

Figure 5: Estimated marginal means of free-living physical activity measured via (A) accelerometer-derived daily minutes of moderate-to-vigorous physical activity (MVPA; n=68), (B) fitbit-derived daily minutes of MVPA (n=65), and (C) self-reported weekly exercise (n=55). Error bars represent 95 % confidence intervals. GLTEQ: Godin leisure time exercise questionnaire [41]; HIIT: high-intensity interval training; MICT: moderate-intensity continuous training; a.u.: arbitrary units. All measurements were collected pre- and 6-months post-intervention. *Denotes a statistically significant difference at an alpha of 0.05.
Figure 5:

Estimated marginal means of free-living physical activity measured via (A) accelerometer-derived daily minutes of moderate-to-vigorous physical activity (MVPA; n=68), (B) fitbit-derived daily minutes of MVPA (n=65), and (C) self-reported weekly exercise (n=55). Error bars represent 95 % confidence intervals. GLTEQ: Godin leisure time exercise questionnaire [41]; HIIT: high-intensity interval training; MICT: moderate-intensity continuous training; a.u.: arbitrary units. All measurements were collected pre- and 6-months post-intervention. *Denotes a statistically significant difference at an alpha of 0.05.

Cardiorespiratory fitness

The linear mixed model for cardiorespiratory fitness met assumptions of normality and homoscedasticity. Estimated marginal means and accompanying 95 % CIs for each treatment condition are shown in Figure 6. A main effect of time was detected [F (1, 65)=18.979, p<0.0001, η p 2 =0.23]. Main effects of condition and condition-by-timepoint interaction were not statistically significant (ps>0.05). Statistically significant improvements in cardiorespiratory fitness 6 months post-intervention when compared to baseline values were seen among participants in both the CHOICE condition [t (65)=2.509, p=0.044] and MICT condition [t (65)=3.492, p=0.003].

Figure 6: Estimated marginal means of cardiorespiratory fitness (n=68) measured via Bruce predicted maximal exercise testing [22]. Error bars represent 95 % confidence intervals. HIIT: high-intensity interval training; MICT: moderate-intensity continuous training. Exercise testing was completed pre- and 6-months post-intervention. *Denotes a statistically significant difference at an alpha of 0.05.
Figure 6:

Estimated marginal means of cardiorespiratory fitness (n=68) measured via Bruce predicted maximal exercise testing [22]. Error bars represent 95 % confidence intervals. HIIT: high-intensity interval training; MICT: moderate-intensity continuous training. Exercise testing was completed pre- and 6-months post-intervention. *Denotes a statistically significant difference at an alpha of 0.05.

Discussion

This pragmatic randomized trial sought to determine whether the provision of choice between HIIT and MICT within a 4-week DPP would lead to greater perceived autonomy support post-intervention when compared to those that were prescribed HIIT or MICT only. Contrary to our hypotheses, we did not find evidence that providing a choice between the two exercise modalities augmented participants’ perceived autonomy support. Such results could be due to several reasons. First, our power to detect between-condition differences in this outcome was severely reduced by the smaller-than-intended sample size due to research interruptions by the COVID-19 pandemic. In addition, a large loss of data (n=25) resulting from human error further compounded this sample size issue, increasing the probability of a type 2 error for this analysis. Second, based on the data that was included, all three randomized conditions averaged near the learning climate questionnaire’s maximum score of 7 with minimal variability (Figure 3), which may be indicative of a ceiling effect in this outcome measure. It may be the case that participants in all three conditions already felt autonomously supported throughout SSBC irrespective of the exercise modality they engaged in due to the abundance of other intervention components within this DPP aimed at fostering autonomous motivation. Indeed, a recent study by MacPherson and colleagues [30] reported the use of 43 different BCTs and 20 motivational interviewing techniques within SSBC’s intervention phase. The effect of changing only one MBCT by providing choice between exercise modalities may not have been strong enough to significantly alter perceived autonomy support when it is already high in this cohort. Despite the small sample size, the results add some evidence that SSBC participants demonstrate high perceived autonomy support after partaking in this DPP. As SSBC is combining multiple motivational strategies rather than solely relying on one strategy, it may serve as a useful example to other health interventions that aim to promote autonomy in their respective populations, though we are unable to ascertain statistical significance.

We also sought to discern any differences in motivation regulation at the conclusion of the intervention phase, as well as 6 months post-intervention. No statistically significant within-condition changes or between-condition differences in motivation regulation were detected. As visually seen in Figure 4, all three conditions increased their autonomous motivation regulation from pre-intervention to post-intervention, with the MICT and CHOICE conditions maintaining such increases 6 months post-intervention. Despite the non-significant findings, the current results suggest that motivation regulation is largely unaffected by either a prescription or provision of choice between HIIT or MICT within SSBC. Similar to perceived autonomy support, the lack of between-condition effects indicates that the use of HIIT, MICT, or a combination of both seem to be feasible exercise strategies within a DPP to foster and/or maintain more autonomous motivation regulation. Contrary to the existing debate on whether HIIT is more feasible than MICT in free-living conditions [17], [18], [19], it may be that the modality/intensity of exercise provided is less important than auxiliary strategies within an intervention such as the incorporation of BCTs that play a more influential role on motivational constructs related to physical activity behavior.

The notion that type of exercise is not the primary determinant of physical activity behavior is further supported when examining the outcome measures of free-living physical activity. Both accelerometry- and Fitbit-derived data showed a significant main negative effect of time in MVPA per day from pre-intervention to 6 months post-intervention, with no between-condition differences (Figure 5). However, this effect dissipated when comparisons were Bonferroni-corrected. Interestingly, there was a main positive effect of time for self-reported physical activity levels from pre-intervention to 6 months post-intervention, with only the CHOICE condition reaching statistical significance after Bonferroni corrections, although the other two conditions would likely show similar significance levels with a larger sample size due to similar effect sizes. These results highlight the existing variability and inaccuracies of currently used methods of physical activity measurement in free-living conditions [40]. There could be various causes to such inconsistencies between objective and subjective measures of physical activity, some of which may include measurement error from worn devices, incorrect use by participants, an overinflation in self-reported results, and recall bias. Technological and self-report advances for measuring free-living physical activity with the capacity to capture a diversity of exercise modalities, including continuous and interval-type exercise, are needed to truly understand behavior in unsupervised settings. In a preliminary attempt to quantify the free-living frequencies of HIIT and MICT engagement, the self-reported exercise modality results highlight the potential patterns arising from the prescription of exercise and provision of choice, with those in the CHOICE and HIIT conditions engaging in both types of exercise and those in the MICT condition engaging in moderate intensity exercise almost exclusively (Table 1). Validation of tools such as this with bigger samples is warranted and may be a potential next step in the measurement of free-living exercise behavior.

Beyond the limitations of physical activity measurement, the results from all three methods of measurement show that there were no between-condition differences in 6-month changes of physical activity behavior from baseline. This further supports the idea that HIIT, MICT, or a combination thereof may all be feasible options to implement in DPPs. Consistent with a previous systematic review on this topic [19], efforts at improving long-term physical activity engagement (whether moderate- or high-intensity) among this population through other strategies may prove to be more beneficial than exercise modality itself.

Participants’ cardiorespiratory fitness increased from pre-intervention to 6 months post-intervention, with both the CHOICE condition and MICT condition showing a statistically significant increase over time in the pre-planned Bonferroni-adjusted within condition comparisons (Figure 6) Of note, cardiorespiratory fitness did not statistically increase for the HIIT condition. This result may be in part due to the HIIT regimen used in this study, where a progressive protocol comprised of only five intervals of relatively low duration was employed. It may be the case that the training stimulus was insufficient to elicit significant cardiorespiratory adaptations, especially if participants continued to engage in the noted HIIT protocol in free-living conditions. Perhaps increases in the duration and/or number of intervals during HIIT sessions (e.g., traditional 4x4min intervals [49], 10x1 min intervals [50]) are warranted and would elicit similar results observed in the CHOICE and MICT conditions. Despite no time effect in the HIIT condition, between-condition differences in the change of cardiorespiratory fitness were not detected, suggesting that both HIIT and MICT are similarly viable exercise modalities to potentially improve fitness parameters among individuals at risk of developing T2D. This is in line with previous research on this population [16].

Interestingly, a decrease in physical activity behavior measured via accelerometry and Fitbit data did not translate to decreases in cardiorespiratory fitness. Instead, cardiorespiratory fitness increased over time, supporting the findings of self-reported physical activity levels. Though this phenomenon may be due to potential measurement errors of wearable devices or self-report data, it may also point to limitations in the time-restricted nature of wearable technology data (i.e., 7-day timeframe), and the need for novel techniques to assess physical activity behavior continuously in free-living conditions. It would also be of value to explore what other factors may be influencing this increase in cardiorespiratory fitness among SSBC participants beyond free-living physical activity behavior.

Strengths and limitations

To our knowledge, this is the first pragmatic randomized trial that has focused on the utility of providing a choice between HIIT and MICT within a DPP on individuals’ psychological, behavioral, and fitness outcomes. The pragmatic nature of this study allows for high ecological validity, making results relevant to real-world applications and potentially generalizable to other populations in need of health interventions. This trial was embedded within a community-based DPP, and the information gathered can be helpful for the undergoing refinement of the intervention. The rigorous methodology, randomization process, and statistical analytical plan account for some possible confounding factors and missing data on some of the outcomes explored. However, the main limitation of this study is a relatively small sample size consequential to the COVID-19 pandemic, particularly the loss of 75 participants’ data from this study’s initial wave. The small sample size compared to the a-priori power analysis significantly raises the probability of a type 2 error and diminishes the utility of the proposed statistical plan to accurately estimate population parameters of measured outcomes. Additionally, data loss in our primary outcome variable further diminished our power to detect between-condition differences and may have had an influential role in the results observed. In free-living conditions, we were unable to isolate just exercise type and control for all other confounders due to this study being conducted pragmatically within a community DPP. Furthermore, the use of an estimation for cardiorespiratory fitness with inherent measurement error as opposed to a traditional metabolic measure limits the accuracy of the point estimates of this outcome and should be interpreted with caution.

Future directions

Although this study provides preliminary evidence on the utility of providing choice between different types of exercise within SSBC, further research is needed to determine the optimal combination of techniques within DPPs for the greatest likelihood of reducing the incidence of T2D via increases in free-living physical activity and resulting cardiorespiratory fitness. It would also be interesting to explore additional factors beyond just physical activity behavior (i.e., weight loss, dietary patterns) that may play a role in determining cardiorespiratory fitness among individuals at increased risk of developing T2D. Future research on measuring free-living physical activity would benefit from novel and validated methods of measurement that are able to accurately capture and differentiate information across diverse exercise modalities. Lastly, replication of this study with a bigger sample and across multiple other populations of interest may be warranted to improve confidence in results and increase generalizability of findings.

Conclusions

The provision of choice between HIIT and MICT within a brief 4-week DPP did not significantly influence individuals’ perceived autonomy support or motivation regulation; however, self-reported free-living physical activity and cardiorespiratory fitness 6 months post-intervention significantly increased for those given a choice. Irrespective of the exercise engaged in, post-intervention perceived autonomy support and motivation regulation were relatively high among SSBC participants and cardiorespiratory fitness improved 6 months post-intervention, indicating that a combination of techniques used within this DPP may be effective at fostering autonomous motivation and translates to positive fitness adaptations. Results also suggest that such combination of auxiliary techniques may be more predictive of physical activity behavior and cardiorespiratory fitness than comparing HIIT vs. MICT; therefore HIIT, MICT, or other exercise modalities may be feasible options among individuals at increased risk of developing T2D. Further advancements in free-living physical activity measurement techniques are needed to address limitations found in current methods of measurement.

Corresponding author: Dr. Mary E. Jung, Professor, Faculty of Health and Social Development, University of British Columbia – Okanagan Campus, 1238 Discovery Avenue, V1V-1V9, Kelowna, BC, Canada, E-mail:

Funding source: Heart and Stroke Foundation of Canada

Award Identifier / Grant number: G-18-0022225

Acknowledgment

The authors would like to thank Jordelle Dupre, Jacqueline Gabelhouse, and Michelle Ungaro for their assistance during stress tests. We would also like to acknowledge all participants in this study. This project was conducted on the unceded, ancestral lands of the Syilx Nation.

  1. Research ethics: Ethical approval was granted on March 4, 2019 by the University of British Columbia’s Clinical Research Ethics Board (H16-02028-A021). This study was conducted in accordance with the Declaration of Helsinki (2013).

  2. Informed consent: Written informed consent was obtained from all individuals included in this study prior to enrollment.

  3. Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission. A.S. conceived the research questions and methods, completed data collection, analyses, and interpretations, and drafted the manuscript. K.F. and J.S. provided support in data analyses and assisted with the critical revision of this manuscript. J.L. assisted with the conception of this study and assisted with the critical revision of this manuscript. F.H., K.P., and N.P. provided support in data collection and assisted with the critical revision of this manuscript. M.J. assisted with the conception of this study, provided supervision of the project, and assisted with the critical revision of this manuscript.

  4. Use of Large Language Models, AI and Machine Learning Tools: None declared.

  5. Conflict of interest: The authors declare no conflict of interest. Jonathan P. Little serves as a Deputy Editor-in-Chief for Translational Exercise Biomedicine, but was not involved in the handling, editorial review, or decision-making process for this manuscript. Mary E. Jung serves as an editorial board member for Translational Exercise Biomedicine, but was not involved in the handling, editorial review, or decision-making process for this manuscript.

  6. Research funding: This trial was funded by a Grant-in-Aid program provided to the corresponding author by the Heart & Stroke Foundation of Canada (G-18-0022225).

  7. Data availability: The data that support the findings of this study are available from the corresponding author, M.E.J., upon reasonable request.

  8. Clinical Trial Registration: This trial was registered with the clinicaltrials.gov registry and given the identifier (NCT03576924).

References

1. Jadhav, RA, Hazari, A, Monterio, A, Kumar, S, Maiya, AG. Effect of physical activity intervention in prediabetes: a systematic review with meta-analysis. J Phys Act Health 2017;14:745–55. https://doi.org/10.1123/jpah.2016-0632.Search in Google Scholar PubMed

2. Diabetes Prevention Program Research Group. The diabetes prevention program (DPP): description of lifestyle intervention. Diabetes Care 2002;25:2165–71. https://doi.org/10.2337/diacare.25.12.2165.Search in Google Scholar PubMed PubMed Central

3. Farni, K, Shoham, DA, Cao, G, Luke, AH, Layden, J, Cooper, RS, et al.. Physical activity and pre-diabetes – an unacknowledged mid-life crisis: findings from the NHANES 2003-2006. PeerJ 2014;2:e499. https://doi.org/10.7717/peerj.499.Search in Google Scholar PubMed PubMed Central

4. Tunis, SR, Stryer, DB, Clancy, CM. Practical clinical trials: increasing the value of clinical research for decision making in clinical and health policy. JAMA 2003;290:1624–32. https://doi.org/10.1001/jama.290.12.1624.Search in Google Scholar PubMed

5. Zwarenstein, M, Treweek, S, Gagnier, JJ, Altman, DG, Tunis, S, Haynes, B, et al.. Improving the reporting of pragmatic trials: an extension of the CONSORT statement. BMJ 2008;337:a2390. https://doi.org/10.1136/bmj.a2390.Search in Google Scholar PubMed PubMed Central

6. Michie, S, West, R, Campbell, R, Brown, J, Gainforth, H. ABC of behaviour change theories. Sutton. UK: Silverback Publishing; 2014.Search in Google Scholar

7. Michie, S, Atkins, L, West, R. The behaviour change wheel: a guide to designing interventions. Sutton, UK: Silverback Publishing; 2014.Search in Google Scholar

8. Carraça, E, Encantado, J, Battista, F, Beaulieu, K, Blundell, J, Busetto, L, et al.. Effective behavior change techniques to promote physical activity in adults with overweight or obesity: a systematic review and meta-analysis. Obes Rev 2021;22:e13258. https://doi.org/10.1111/obr.13258.Search in Google Scholar PubMed PubMed Central

9. Deci, EL, Ryan, RM. Intrinsic motivation and self-determination in human behavior. New York, NY: Plenum; 1985.10.1007/978-1-4899-2271-7Search in Google Scholar

10. Teixeira, PJ, Carraca, EV, Markland, D, Silva, MN, Ryan, RM. Exercise, physical activity, and self-determination theory: a systematic review. Int J Behav Nutr Phys Act 2012;9:78. https://doi.org/10.1186/1479-5868-9-78.Search in Google Scholar PubMed PubMed Central

11. Ryan, RM, Deci, EL. Self-determination theory: basic psychological needs in motivation, development, and wellness. New York, NY: Guilford Publications; 2017.10.1521/978.14625/28806Search in Google Scholar

12. Ryan, RM, Patrick, H, Deci, EL, Williams, GC. Facilitating health behaviour change and its maintenance: interventions based on self-determination theory. Eur Health Psychol 2008;10:2–5.Search in Google Scholar

13. Teixeira, DS, Marques, MM, Palmeira, AL. Associations between affect, basic psychological needs and motivation in physical activity contexts: systematic review and meta-analysis. Rev Iberoam Psicol Ejerc Deporte 2018;13:225.Search in Google Scholar

14. Teixeira, PJ, Marques, MM, Silva, MN, Brunet, J, Duda, J, Haerens, L, et al.. A classification of motivation and behavior change techniques used in self-determination theory-based interventions in health contexts. Mot Sci 2020;6:438–55. https://doi.org/10.1037/mot0000172.Search in Google Scholar

15. De Nardi, AT, Tolves, T, Lenzi, TL, Signori, LU, Silva, AMVD. High-intensity interval training versus continuous training on physiological and metabolic variables in prediabetes and type 2 diabetes: a meta-analysis. Diabetes Res Clin Pract 2018;137:149–59. https://doi.org/10.1016/j.diabres.2017.12.017.Search in Google Scholar PubMed

16. Guo, Z, Li, M, Cai, J, Gong, W, Liu, Y, Liu, Z, et al.. Effect of high-intensity interval training vs. moderate-intensity continuous training on fat loss and cardiorespiratory fitness in the young and middle aged a systematic review and meta-analysis. Int J Environ Res Public Health 2023;20:4741. https://doi.org/10.3390/ijerph20064741.Search in Google Scholar PubMed PubMed Central

17. Biddle, SJH, Batterham, AM. High-intensity interval exercise training for public health: a big HIT or shall we HIT it on the head? Int J Behav Nutr Phys Act 2015;12:95. https://doi.org/10.1186/s12966-015-0254-9.Search in Google Scholar PubMed PubMed Central

18. Ekkekakis, P, Biddle, SJH. Extraordinary claims in the literature on high-intensity interval training (HIIT): IV. Is HIIT associated with higher long-term exercise adherence? Psychol Sport Exerc 2023;64:102295. https://doi.org/10.1016/j.psychsport.2022.102295.Search in Google Scholar PubMed

19. Santos, A, Braaten, K, MacPherson, M, Vasconcellos, D, Vis-Dunbar, M, Lonsdale, C, et al.. Rates of compliance and adherence to high-intensity interval training: a systematic review and meta-analyses. Int J Behav Nutr Phys Act 2023;20:134. https://doi.org/10.1186/s12966-023-01535-w.Search in Google Scholar PubMed PubMed Central

20. American Diabetes Association. Classification and diagnosis of diabetes: standards of medical care in diabetes – 2019. Diabetes Care 2019;42:s13–28. https://doi.org/10.2337/dc19-S002.Search in Google Scholar PubMed

21. Canadian Society for Exercise Physiology [Internet]. CSEP get active questionnaire 2023[cited 2023 Jul 20]. Available from: https://csep.ca/2021/01/20/pre-screening-for-physical-activity/.Search in Google Scholar

22. Bruce, RA. Exercise testing of patients with coronary heart disease. Principles and normal standards for evaluation. Ann Clin Res 1971;3:323–32.Search in Google Scholar

23. Bruce, RA, Kusumi, F, Hosmer, D. Maximal oxygen intake and nomographic assessment of functional aerobic impairment in cardiovascular disease. Am Heart J 1973;85:546–62. https://doi.org/10.1016/0002-8703(73)90502-4.Search in Google Scholar PubMed

24. Nes, BM, Jansky, I, Wisloff, U, Stoylen, A, Karlsen, T. Age-predicted maximal heart rate in healthy subjects: the HUNT fitness study. Scand J Med Sci Sports 2013;23:697–704. https://doi.org/10.1111/j.1600-0838.2012.01445.x.Search in Google Scholar PubMed

25. Borg, GA. Psychophysical bases of perceived exertion. Med Sci Sports Exerc 1982;14:377–81. https://doi.org/10.1249/00005768-198205000-00012.Search in Google Scholar

26. Jung, ME, Bourne, JE, Beauchamp, MR, Robinson, E, Little, JP. High-intensity interval training as an efficacious alternative to moderate-intensity continuous training for adults with prediabetes. J Diabetes Res 2015;2015:191595. https://doi.org/10.1155/2015/191595.Search in Google Scholar PubMed PubMed Central

27. Locke, SR, Bourne, JE, Beauchamp, MR, Little, JP, Barry, J, Singer, J, et al.. High-intensity interval or continuous moderate exercise: a 24-week pilot trial. Med Sci Sports Exerc 2018;50:2067–75. https://doi.org/10.1249/MSS.0000000000001668.Search in Google Scholar PubMed

28. Jung, ME, Locke, SR, Bourne, JE, Beauchamp, MR, Lee, T, Singer, J, et al.. Cardiorespiratory fitness and accelerometer-determined physical activity following one year of free-living high-intensity interval training and moderate-intensity continuous training: a randomized trial. Int J Behav Nutr Phys Act 2020;17:25. https://doi.org/10.1186/s12966-020-00933-8.Search in Google Scholar PubMed PubMed Central

29. Bean, C, Dineen, T, Locke, S, Bouvier, B, Jung, ME. An evaluation of the reach and effectiveness of a diabetes prevention behaviour change program situated in a community site. Can J Diabetes 2021;45:360–8. https://doi.org/10.1016/j.jcjd.2020.10.006.Search in Google Scholar PubMed

30. MacPherson, MM, Dineen, TE, Cranston, KD, Jung, ME. Identifying behaviour change techniques and motivational interviewing techniques in small steps for big changes: a community-based program for adults at risk of type 2 diabetes. Can J Diabetes 2020;44:719–26. https://doi.org/10.1016/j.jcjd.2020.09.011.Search in Google Scholar PubMed

31. Dineen, TE, Bean, C, Jung, ME. Implementation of a diabetes prevention program within two community sites: a qualitative assessment. Implement Sci Commun 2022;3:11. https://doi.org/10.1186/s43058-022-00258-6.Search in Google Scholar PubMed PubMed Central

32. Dineen, TE, Banser, T, Bean, C, Jung, ME. Fitness facility staff demonstrate high fidelity when implementing an evidence-based diabetes prevention program. Transl Behav Med 2021;11:1814–22. https://doi.org/10.1093/tbm/ibab039.Search in Google Scholar PubMed PubMed Central

33. Cranston, KD, Ivanova, E, Davis, C, Jung, ME. Fidelity and maintenance of motivational interviewing skills in diabetes prevention program coaches: a pilot study. Int J Health Promot Educ 2022:1–13. https://doi.org/10.1080/14635240.2022.2136976.Search in Google Scholar

34. Karvonen, MJ, Kentala, E, Mustala, O. The effects of training on heart rate; a longitudinal study. Ann Med Exp Biol Fenn 1957;35:3017–315.Search in Google Scholar

35. Ross, R, Chaput, J-P, Giangregorio, LM, Janssen, I, Saunders, TJ, Kho, ME, et al.. Canadian 24-hour movement guidelines for adults aged 18-64 years and adults aged 65 years or older: an integration of physical activity, sedentary behaviour, and sleep. Appl Physiol Nutr Metab 2020;45:S57–102. https://doi.org/10.1139/apnm-2020-0467.Search in Google Scholar PubMed

36. Williams, GC, Deci, EL. Internalization of biopsychosocial values by medical students: a test of self-determination theory. J Pers Soc Psychol 1996;70:767–79. https://doi.org/10.1037//0022-3514.70.4.767.Search in Google Scholar PubMed

37. Levesque-Bristol, C, Knapp, TD, Fisher, BJ. The effectiveness of service-learning: it’s not always what you think. J Exp Educ 2010;33:208–24. https://doi.org/10.1177/105382590113300302.Search in Google Scholar

38. Markland, D, Tobin, V. A modification to the behavioral regulation in exercise questionnaire to include an assessment of amotivation. J Sport Exerc Psychol 2004;26:191–6. https://doi.org/10.1123/jsep.26.2.191.Search in Google Scholar

39. Murcia, JAM, Gimeno, EC, Camacho, AM. Measuring self-determination motivation in a physical fitness setting: validation of the behavioral regulation in exercise Questionnaire-2 (BREQ-2) in a Spanish sample. J Sports Med Phys Fitness 2007;47:366–74.Search in Google Scholar

40. Dowd, KP, Szelicki, R, Minetto, MA, Murphy, MH, Polito, A, Chigo, E, et al.. A systematic literature review of reviews on techniques for physical activity measurement in adults: a DEDIPAC study. Int J Behav Nutr Phys Act 2018;15:15. https://doi.org/10.1186/s12966-017-0636-2.Search in Google Scholar PubMed PubMed Central

41. Godin, G. The godin-shephard leisure-time physical activity questionnaire. Health Fitness J Can 2011;4:18–22. https://doi.org/10.14288/hfjc.v4i1.82.Search in Google Scholar

42. Motl, RW, Bollaert, RE, Sandroff, BM. Validation of the godin leisure-time exercise questionnaire classification coding system using accelerometry in multiple sclerosis. Rehab Psychol 2018;63:77–82. https://doi.org/10.1037/rep0000162.Search in Google Scholar PubMed

43. Migueles, JH, Cadenas-Sanchez, C, Ekelund, U, Nystrom, CD, Mora-Gonzalez, J, Lof, M, et al.. Accelerometer data collection and processing criteria to assess physical activity and other outcomes: a systematic review and practical considerations. Sports Med 2017;47:1821–45. https://doi.org/10.1007/s40279-017-0716-0.Search in Google Scholar PubMed PubMed Central

44. Troiano, RP, Berrigan, D, Dodd, KW, Mâsse, LC, Tilert, T, McDowell, M, et al.. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc 2008;40:181–8. https://doi.org/10.1249/mss.0b013e31815a51b3.Search in Google Scholar PubMed

45. Orstad, SL, Gerchow, L, Patel, NR, Reddy, M, Hernandez, C, Wilson, DK, et al.. Defining valid activity monitor data: a multimethod analysis of weight-loss intervention participants’ barriers to wear and first 100 days of physical activity. Informatics 2021;8:39. https://doi.org/10.3390/informatics8020039.Search in Google Scholar PubMed PubMed Central

46. Fitbit Health Solutions [Internet]. Fitbit web API data dictionary – glossary 2022 [cited 2023 Jul 20]. Available from: https://healthsolutions.fitbit.com/fitbit-web-api-data-dictionary/#glossary.Search in Google Scholar

47. Lonsdale, C, Rosenkranz, RR, Sanders, T, Peralta, LR, Bennie, A, Jackson, B, et al.. A cluster randomized controlled trial of strategies to increase adolescents’ physical activity and motivation in physical education: results of the motivating active learning in physical education (MALP) trial. Prev Med 2013;57:696–702. https://doi.org/10.1016/j.ypmed.2013.09.003.Search in Google Scholar PubMed

48. Chakraborty, H, Gu, H. A mixed model approach for intent-to-treat analysis in longitudinal clinical trials with missing values. Research Triangle Park, NC: RTI Press; 2009.10.3768/rtipress.2009.mr.0009.0903Search in Google Scholar PubMed

49. Moholdt, TT, Amundsen, BH, Rustad, LA, Wahba, A, Lovo, KT, Gulliksta, LR, et al.. Aerobic interval training versus continuous moderate exercise after coronary artery bypass surgery: a randomized study of cardiovascular effects and quality of life. Am Heart J 2009;158:1031–7. https://doi.org/10.1016/j.ahj.2009.10.003.Search in Google Scholar PubMed

50. Little, JP, Gillen, JB, Percival, ME, Safdar, A, Tarnopolsky, MA, Punthakee, Z, et al.. Low-volume high-intensity interval training reduces hyperglycemia and increases muscle mitochondrial capacity in patients with type 2 diabetes. J Appl Physiol 2011;111:1554–60. https://doi.org/10.1152/japplphysiol.00921.2011.Search in Google Scholar PubMed

Supplementary Material

This article contains supplementary material (https://doi.org/10.1515/teb-2025-0026).

Received: 2025-08-08
Accepted: 2025-11-19
Published Online: 2025-12-01

© 2025 the author(s), published by De Gruyter on behalf of Shangai Jiao Tong University and Guangzhou Sport University

This work is licensed under the Creative Commons Attribution 4.0 International License.

Articles in the same Issue

  1. Frontmatter
  2. Section: Integrated exercise physiology, biology, and pathophysiology in health and disease
  3. The impact of exercise on the role of lipid droplets in maintaining health: a narrative review
  4. Effects of different interventions on microgravity-induced bone loss
  5. Section: Personalized and advanced exercise prescription for health and chronic diseases
  6. A therapeutic exercise program for adolescents engaged in gender diversity services: study protocol for a non-randomised clinical trial
  7. The role of exercise modality on psychological, behavioral, and fitness outcomes among individuals at risk of type 2 diabetes: preliminary evidence from the CHOICE pragmatic randomized trial
  8. Section: Physical activity/inactivity and health across the lifespan
  9. Serum metabolic alterations after two weeks of step-reduction and following four weeks of exercise rehabilitation in older adults: a secondary analysis of the ENDURE randomised controlled trial
  10. Calibration and evaluation of MET models for estimating energy expenditure using thigh and ankle-worn move 4 accelerometer
  11. Section: Sports medicine and movement science
  12. Muscular activity of hamstrings under attentional focus instructions: an electromyography based study
  13. Associations between anthropometric outcomes and fat percentage with physical performance in professional soccer players: a cutoff points approach
https://doi.org/10.1515/teb-2025-0026
Keywords for this article
behavior change; high-intensity interval training; moderate-intensity continuous training; pragmatic randomized trial; prediabetes; self-determination theory
Creative Commons
BY 4.0

Articles in the same Issue

  1. Frontmatter
  2. Section: Integrated exercise physiology, biology, and pathophysiology in health and disease
  3. The impact of exercise on the role of lipid droplets in maintaining health: a narrative review
  4. Effects of different interventions on microgravity-induced bone loss
  5. Section: Personalized and advanced exercise prescription for health and chronic diseases
  6. A therapeutic exercise program for adolescents engaged in gender diversity services: study protocol for a non-randomised clinical trial
  7. The role of exercise modality on psychological, behavioral, and fitness outcomes among individuals at risk of type 2 diabetes: preliminary evidence from the CHOICE pragmatic randomized trial
  8. Section: Physical activity/inactivity and health across the lifespan
  9. Serum metabolic alterations after two weeks of step-reduction and following four weeks of exercise rehabilitation in older adults: a secondary analysis of the ENDURE randomised controlled trial
  10. Calibration and evaluation of MET models for estimating energy expenditure using thigh and ankle-worn move 4 accelerometer
  11. Section: Sports medicine and movement science
  12. Muscular activity of hamstrings under attentional focus instructions: an electromyography based study
  13. Associations between anthropometric outcomes and fat percentage with physical performance in professional soccer players: a cutoff points approach
Sign up now to receive a 20% welcome discount
Institutional Access
How does access work?
Have an idea on how to improve our website?
Please write us.
© 2026 De Gruyter Brill
Downloaded on 26.1.2026 from https://www.degruyterbrill.com/document/doi/10.1515/teb-2025-0026/html
Scroll to top button