Insights and Advances in Physical Activity Research on Youth with Type 1 Diabetes

Abstract

Purpose of Review:

Most youth with type 1 diabetes (T1D) do not meet the guidelines for physical activity engagement, thereby diminishing potential benefits to physical and mental health. This review synthesizes the recent literature on physical activity among youth with T1D and offers recommendations for future research.

Recent Findings:

Studies highlight challenges related to the use of inconsistent measurement tools, which prevent definitive conclusions about the mechanistic factors underlying low physical activity in youth. There has been limited research in young children and youth newly diagnosed with T1D. Additionally, most interventions to promote physical activity in youth with T1D have involved structured and supervised exercise sessions, leaving a gap in knowledge regarding the potential impact of unstructured and unsupervised exercise interventions. To address these gaps, rigorous studies employing validated measures of physical activity in youth are needed. Interventions should incorporate developmentally appropriate behavioral science theories and emerging technologies in their design. Additional priorities include integrating diabetes technologies into clinical care, more real-world data to improve the accuracy of machine learning models for predicting dysglycemia, and advancing personalized mHealth interventions to promote physical activity in youth.

Summary:

While physical activity is an important area of pediatric diabetes research, gaps remain in our knowledge and intervention development. Physical activity consultations should be a part of routine diabetes care for youth. Research can inform these consultations by providing strategies to promote physical activity uptake and maintenance and by exploring ways to leverage new technologies to help youth with T1D exercise safely.

Introduction

Global estimates indicate that more than 1,200,000 individuals under 20 years of age are currently living with type 1 diabetes (T1D) [1]. In the United States, T1D affects approximately 185,000 youth and if current trends continue, experts predict that over 300,000 will be living with T1D by 2060 [2]. Youth living with T1D are at risk for both physical and psychosocial challenges, mental health disorders [3], poor peer relationships [4], and delayed achievement of key developmental tasks (e.g., forming a positive self-concept) [5]. Yet findings indicate that moderate to vigorous physical activity (MVPA) can play a crucial role in ameliorating the physical [6,7] and psychosocial [6–9] impact of T1D on youth. For instance, there is evidence that associates physical activity with improved cardiovascular function [10], body composition [7], fitness [7], bone mineral density [11], and extended partial remission time [12] among youth with T1D. Additionally, because increased energy demands during exercise trigger glucose transport, glucose uptake, and increased insulin sensitivity [13], physical activity can lead to lower blood glucose [14–16] and HbA1c levels [6,10,12,17–19] for youth with T1D.

Through a psychosocial lens, the data suggest forms of physical activity (e.g., recreational activities, team sports) offer invaluable opportunities to build social capital, form peer relationships, and practice leadership and communication skills [20]. As a result, physical activity not only has numerous physiological benefits, but it can serve as a key resource to help youth manage many psychosocial challenges associated with T1D [6–9]. Thus, a common failure to meet MVPA guidelines, particularly during early development, stands to undermine children and adolescents’ self-management and physical health as well as disrupt their personal growth.

The purpose of this current review is to synthesize the recent physical activity literature on youth with T1D since Tully and colleagues’ (2016) review [21], with an eye toward exploring the methodological challenges of studying physical activity and the specific contextual factors impacting physical activity in youth living with T1D. Additionally, this review provides an update on available interventions targeting physical activity in youth with T1D, and concludes with guidance regarding future research inquiries. Though physical activity is also an important treatment component for youth living with type 2 diabetes, the contextual factors impacting physical activity for youth with type 2 diabetes are different and therefore beyond the scope of the current article.

Physical Activity Measurement

One inherent challenge in studying physical activity is measurement. Self-report has been the dominant assessment method for youth living with T1D and youth without T1D [22,23]. This method has the benefit of being both low cost and highly feasible. However, it is also vulnerable to response or recall bias, which can reduce accuracy [23]. For instance, research suggests there is no correlation between objective measures of physical activity and self-report among young children, and even in older children, the correlation between objective measures and self-report is very low [24]. Accelerometer-based physical activity assessment can offer an objective measure and balance rigorous assessment and feasibility. Accelerometers measure movement data over time and can provide reliable estimates of sedentary, light, moderate, and vigorous activity in children based on validated algorithms [25]. Furthermore, because accelerometers can be used to measure real-world activity, they offer a critical layer of ecological validity for behavioral science research. At the other end of the spectrum, there are laboratory-based physical activity measures [26] such as the Bruce protocol, which measures VO₂ max through a progressively demanding treadmill test [27,28]. These methods deliver gold-standard accuracy, but they are expensive and burdensome, and may lack ecological validity.

Accelerometers strike a balance between cost, feasibility, and accuracy in measuring physical activity. While they yield the most accurate activity data and can capture whole-body movement when worn at the hip (near the iliac crest), hip placement can be cumbersome[25]. Participants often remove waist-worn devices at night and forget to put them back on—resulting in reduced wear time [29]. In contrast, wearing wrist-worn accelerometers on the non-dominant wrist greatly improves compliance [30] but often produces inflated activity readings compared to hip placement [31]. Therefore, balancing acceptability with data accuracy is a crucial consideration when designing projects that utilize accelerometers.

Specific to projects using accelerometers in youth with T1D, it may be important to consider the burden of adding another device. There is evidence suggesting many youth with T1D report wear-related issues with their diabetes devices, and more than a third report that devices are a general burden [32]. Additionally, youth with T1D might find it particularly challenging to wear an accelerometer at their hip, as this space may compete with placement of their glucose sensor or pump infusion site. Finally, it is not clear whether existing algorithms that classify physical activity for youth have been validated in youth with T1D. We know that immediate glycemic responses to real-world physical activity vary based on factors such as activity type, duration, and glucose availability among youth with T1D [14]. As such, it may be necessary to establish tailored scoring criteria to identify sedentary, light, moderate, and vigorous activity in youth with T1D.

As another method to assess real-world physical activity, researchers have begun to use consumer-grade wearable activity trackers (e.g., Fitbit) [33]. These devices have the benefit of low participant burden and immediate feedback, which may enhance engagement [34], making them attractive for studies—especially when participants are allowed to keep the device as an incentive [35]. Unfortunately, consumer-grade activity wearables often tend to overestimate physical activity relative to research-grade accelerometers [33], though it may be possible to minimize this bias by choosing wearables that allow for raw data access [36] or paying for platforms like Fitabase to retrieve and process raw data from consumer devices [37].

Physical Activity Guidelines and Associated Outcomes

Current guidelines for children and adolescents living with T1D recommend 60 minutes of daily MVPA, including muscle- and bone-strengthening exercises at least three days a week [38]. These recommendations align with physical activity guidelines set by the World Health Organization [39] and the United States Department of Health and Human Services [40]. While most youth do not meet MVPA recommendations, data suggest youth with T1D are more sedentary and engage in significantly less MVPA than their counterparts without T1D [41,42]. Specifically, results from a recent meta-analysis suggest youth with T1D engage in nearly 13 fewer minutes of MVPA and 63 more minutes of sedentary behavior per day than youth without T1D [41].

Though youth with T1D may not engage in physical activity at rates commensurate with clinical care guidelines or comparable to their peers without T1D, findings regarding the health benefits of regular physical activity in youth with T1D are robust. For instance, youth with T1D who engage in more physical activity consistently demonstrate lower glucose [14–16] and HbA1c [6,10,12,17–19] levels than their less active peers. One study examining glycemic effects during physical activity found that greater reductions in glucose is associated with lower baseline HbA1c, participation in noncompetitive activities, lower BMI, and reduced fear of hypoglycemia (FH) [14]. Extending these insights to examine the post-exercise period, researchers found that glucose levels were lowest eight to sixteen hours post-exercise [15] and that these post-exercise declines were most pronounced in youth who were recently diagnosed, had lower baseline HbA1c, experienced a greater drop in glucose during exercise, and engaged in longer exercise sessions [15]. Other research has identified additional moderators of post-exercise glucose levels, including glucagon levels [43] and diet [44].

With respect to exercise effects on HbA1c, data show lower HbA1c levels following high intensity exercise [10,18], longer exercise programs [10], and combination aerobic and resistance training routines [10] in youth. There are also important benefits for physically active youth with new-onset T1D who may experience higher rates of partial remission for at least two years compared to less active youth [12]. Finally, there is evidence that physical activity enhances psychosocial outcomes for youth with T1D. For example, adolescents who engage in more physical activity have demonstrated improved family functioning [6], quality of life [45], and lower levels of diabetes-related distress [8].

Nevertheless, engaging in physical activity is not without risk for youth with T1D. Regulatory hormones stimulated when energy demands are high trigger glucose production and a rise in blood glucose levels [46]. After vigorous physical activity (VPA), circulating insulin can cause excessive glucose uptake that depletes the body of its blood glucose without restoring what has been used, resulting in delayed onset hypoglycemia [46]. It is also possible for glucose uptake to occur during muscle contraction without requiring insulin, which can increase the risk of later hypoglycemia [46]. Simply put, while brief bouts of exercise may lead to temporary increases in blood glucose levels and possible hyperglycemia [46], sustained VPA may result in post-exercise hypoglycemia. Though the same problem does not appear related to MVPA [47], youth living with T1D must remain vigilant and know how to effectively manage both high and low glucose levels to safely engage in physical activity.

Influences on Physical Activity Engagement

Fear of Hypoglycemia

Unlike their peers without T1D, youth with T1D encounter unique challenges that may perpetuate a cycle of insufficient MVPA. Fear of hypoglycemia (FH), a condition characterized by excessive fear and an over reliance on hypoglycemia avoidance behaviors, is one barrier that has been explored in youth. Unfortunately, while multiple scales have been validated to measure FH in youth [48,49] and parents [49], no scale has been designed to specifically assess fear related to exercise-induced hypoglycemia in adults, youth, or parents. In the absence of a specific measure to assess fear of exercise-induced hypoglycemia, three studies have employed measures of general FH with mixed results. Roberts et al. (2020) used the child and parent versions of the Hypoglycemia Fear Survey in a large national sample of youth with T1D and parents [50]. Their results found no association between parent-reported FH and youth-reported physical activity, though youth-reported FH was positively associated with youth-reported minutes of VPA, challenging the idea that FH is a barrier to physical activity [50]. In contrast, a recent study measuring youth physical activity via accelerometer found that youth with higher FH engaged in fewer high intensity physical activity sessions than youth with lower FH [51]. Likewise, in a study examining the association between parent-reported FH and minutes of physical activity among preschoolers with T1D, researchers found parents’ report of hypoglycemia worry negatively correlated with children’s daily minutes of MVPA, suggesting parent FH could be a barrier to physical activity in very young children [52]. There is one other study which employed a single item to measure FH in youth [53]. The results of this study furthers the controversy as the researchers found that youth who reported no FH engaged in less total and light physical activity than youth who reported some FH [53]. Thus, the current state of research examining youth and parent FH as potential barriers to physical activity provides no definitive conclusions. However, it is possible a developmentally tailored measure of exercise-induced FH could help clarify any association.

Risk of Hypoglycemia

To obtain a broader assessment of potential barriers to physical activity in youth, some studies have used an adapted version of the Barriers to Physical Activity in Diabetes-Type 1 (BAPAD1) [18,54–58]. The BAPAD1 has 11 items that represent potential T1D-specific barriers and general barriers to physical activity. Among the barriers specific to T1D is an item that evaluates perceived risk of hypoglycemia. Here the results appear more uniform with all studies suggesting that youth perceive risk of hypoglycemia as very to slightly unlikely to be a barrier to physical activity [18,56–58]. It is notable that one study found a weak positive correlation between the risk of hypoglycemia item on the BAPAD1 and youth-reported FH scores [51]. Thus, while related, it appears that they represent separate constructs or barriers to physical activity and therefore should not be used interchangeably in research.

Physical Fitness

There are mixed results when examining physical fitness level as barrier to physical activity as well. In one study, recruiting youth and young adults with T1D between 5 and 20 years, low fitness level was identified as the second leading barrier to physical activity [58]. However, this contrasts with other studies which have found that low fitness was among the least impactful barriers to physical activity, especially among children [55, 58]. It is possible age could be a moderator in the equation of physical activity engagement among youth with T1D [59]. This hypothesis would align with research on adults with T1D which has found low fitness to be a key barrier to physical activity [59].

Social Support

There is robust evidence that social support from family members, peers, and knowledgeable experts (e.g., coaches) is a key mediator of physical activity among youth with T1D. These persons act as important models and organizers, providing encouragement and logistical aid to help youth navigate physical activity-related barriers [60]. Among these relationships, the parent-child dyad plays a central role. For instance, parents can help youth overcome practical barriers to physical activity, such as access to exercise equipment, transportation, and identifying alternative activities during bad weather [56,58–61]. Parents also serve as role models for physical activity engagement; research indicates that children of active parents demonstrate higher levels of MVPA and less screen time [62]. However, parental anxiety about their child’s ability to prevent hypoglycemic episodes can lead to some reluctance in encouraging physical activity engagement [60], underscoring the potential impact caregivers can have both in facilitating or impeding youth physical activity. Furthermore, teachers and healthcare professionals can shape physical activity engagement in youth through their personal beliefs about exercise and their confidence in providing guidance on safe physical activity [58,60,63].

Though coaches can play an important role in promoting physical activity among youth, research examining their impact on youth with T1D is limited [64]. It is safe for most youth with T1D to participate in sports [65], but there is evidence that coaches who lack confidence in adolescents’ ability to self-manage their T1D during sports can reduce adolescents’ satisfaction and increase their frustration, especially among athletes [66]. Existing research on children and adolescents in the general population indicates that coaching behaviors—specifically support for players’ autonomy—is crucial for athlete satisfaction and is indirectly related to engagement [67]. Qualitative reports indicate that youth with T1D and their parents believe coaches lack important knowledge related to T1D, while coaches themselves vary in their sense of responsibility and perceived capacity to assist in T1D management [64]. Coaches do, however, express a desire for more information on T1D, and—given the time constraints and additional responsibility of overseeing multiple players—would prefer brief informational guides [64]. Thus, efficient educational interventions for coaches may enhance mutual trust and overall physical activity safety and engagement.

As youth mature, peer relationships increasingly influence health behaviors. While encouragement and mutual participation in recreational activities may enhance physical activity engagement among youth with T1D [58], peer perceptions may also have a negative impact. For example, diabetes-related embarrassment has been shown to mediate the relationship between peer victimization and reduced physical activity [68]. In addition, the desire for social acceptance and the pressure to fit in can lead youth to neglect glucose monitoring or delay treatments [69,70], increasing the risk of hypoglycemia during periods of elevated physical activity. Given the frequency of peer interactions in school and extracurricular settings, it is possible identifying modifiable factors and teaching behavioral strategies to cope with stigma may increase physical activity among youth with T1D.

Interventions

In a recent meta-analysis of 14 physical activity promoting interventions, researchers demonstrated that youth with T1D who received exercise training had significantly reduced HbA1c levels compared to those who did not (Hedges’ g = −0.38, 95% CI [−0.66, −0.11]) [10]. However, heterogeneity across study designs complicates drawing conclusions about overall efficacy.

Most studies tested an aerobic-only [71–76] or combined aerobic-resistance program [77–82], while one study examined resistance-only training [83] and another examined a Pilates program [84]. Intervention durations ranged widely from 8 to 32 weeks, with significant variation in frequency and session length [10]. With the exception of Salem and colleagues’ (2010) [81], studies all had samples of fewer than 50 participants and only four studies included children under the age of nine. For only half the studies, researchers measured adherence to the training programs [71,72,77–81], while only one study tested for maintenance effects [80] or changes in youths’ typical physical activity [73].

It is notable that all the interventions—including three studies that were not included in the meta-analysis [85–87]—involved physical activity programs supervised by experts (e.g., nutritionists, sports scientists, and medical doctors). To date, few unsupervised interventions have been tested [88–90]. Moreover, among available studies, the results do not seem promising. The trials either failed to increase physical activity levels [88], demonstrated an increase in MVPA but no group difference between intervention and control groups [89], or failed to collect baseline measures of PA [90]. Only two trials have aimed to leverage social cognitive theory to increase PA engagement [89,90]. Though their results were generally negligible, it is a promising next step to see diabetes researchers design and test physical activity interventions based on established theories of behavior change.

Gaps in the Research

While interventions to promote physical activity among youth with T1D have made significant strides, few projects have distinguished between young children, school-age children, adolescents, and young adults. Therefore, to advance the field, interventions must adopt a developmentally tailored approach. It is also important to design and test interventions that incorporate proven behavior change theories. A review of intervention studies on preteens with T1D (N=24) identified significant weaknesses concerning the integration of theories, including 1) a common failure to explicitly state and justify the choice of theoretical models used in an intervention, 2) a rare use of age-appropriate theoretical frameworks, and 3) a reliance on individual-level behavior change theories which may overlook sociocultural dynamics that could be better captured by integrating models from the social and behavioral sciences [91].

Furthermore, combining theoretical frameworks from the social and behavioral sciences offers an opportunity to craft more personalized interventions. Advocates for study designs rooted in behavioral economics highlight how this interdisciplinary approach could be adapted to serve younger populations due to its flexible reward systems and focus on natural human motives such as Present Bias, Optimism Bias, Loss Aversion, Default Bias, and Social Norms [92]. Related, targeting tendencies such as delay discounting (i.e., preference for immediate reward over future gains) may help reshape ineffective behavioral patterns that reduce physical activity. Indeed, increased levels of delay discounting have been linked to less treatment adherence [93] and higher glucose levels [94] among youth with T1D. Finally, interventions that leverage behavioral science theories should aim to increase self-regulation abilities and better align the incentives with physical activity engagement. Because youth and adults face different challenges managing T1D [95], interventions must target the developmentally unique motives and barriers to engage youth. Grounding these efforts in social and behavioral science principles will be crucial for developing effective strategies to promote physical activity.

Innovations in Technology

Technologies such as continuous glucose monitors (CGM), hybrid closed-loop insulin delivery systems, mobile apps, and personalized text messaging prompts have the potential to reduce barriers to physical activity [96–99]. One such way is through real-time CGM readings which provide insights before, during, and after exercise that can optimize physical performance and safety [46]. Moreover, devices that support data sharing provide an extra safeguard by allowing parents immediate access to important health information about their child’s glucose levels [46]. These technologies have already been shown to increase VPA and reduce FH in youth [100]. Yet caution should still be exercised. While CGM users may report less FH and engage in more VPA, both effects may make youth more vulnerable to hypoglycemia as a consequence of fewer hypoglycemia avoidance behaviors [100]. Thus, it is important to provide education on safe glucose management during exercise even among youth using CGM.

Closed-loop technologies, which automatically titrate basal insulin based on CGM readings, offer a promising solution to help with glucose management before, during, and after physical activity and should therefore continue to be the focus of research. While confounding variables, including personal fitness, hormone functioning, diet, and exercise type, duration, and intensity, remain a significant factor [46], it is possible more data will help overcome this hurdle and increase these systems self-operating capabilities. Currently, predictive models of dysglycemia have been limited by the poor precision of CGMs and suboptimal performance of existing algorithms during physical activity [101]. In time, though, it is possible precision will improve along with our algorithms. One promising area of exploration is in the use of machine learning to predict dysglycemia risk in youth with T1D during and after exercise. Generally, these models have shown high predictive precision [102], suggesting that with further model training, they could have a dramatic impact on youth with T1D by providing them with advance notice for when to initiate preventative measures for dysglycemia during and after exercise.

Though the Dexcom G-series, Freestyle Libre-series, and Omnipod Dash all have the capacity to integrate with third-party apps on Android and Apple products, another gap in research is the effect of these systems on physical activity in youth with T1D. It is possible that merging data from physical activity trackers (i.e., Apple Watches and Fitbits), CGMs, insulin pumps, and self-reported diet and exercise, may help to inform more personalized behavior change recommendations to promote physical activity in youth with T1D. Currently, one study in individuals with type 2 diabetes offers promising results [103] and could possibly be a model for future studies in youth with T1D.

Finally, in one recent paper, researchers described their co-design of an mHealth app to empower youth with T1D to exercise safely. Key features that youth identified for optimizing the exercise intervention included CGM integration for easy data access, individualized feedback on personal health trends, diet tracking, and synchronization across apps (e.g., music, FitBit) [96]. These data suggest tailored mHealth tools may have the potential to create more engaging, personalized, and sustainable exercise routines in youth with T1D. Thus, another important area for future research in youth with T1D may be in the design of T1D-specific and exercise-focused mHealth interventions.

Conclusion

Physical activity offers youth with T1D an array of physical and psychosocial benefits, including improved metabolic control, cardiovascular health, and psychological well-being. Despite these advantages, a significant proportion of youth with T1D fail to meet recommended physical activity guidelines. To advance our understanding of factors related to physical activity in youth with T1D, we need more rigorous studies using validated and T1D-specific measures. To enhance physical activity engagement among youth with T1D, research should integrate approaches from the behavioral and social sciences into intervention design, leverage emerging diabetes technologies for safe glucose regulation during and after exercise, and incorporate strategies that address both individual and systemic barriers to physical activity. Within specific populations, there is a gap in research among young children living with T1D. There is also more to learn in both measurement and intervention development (e.g. integration of technology, social networks, behavior change theories) in the pursuit of lower HbA1c levels among newly diagnosed youth [104,105]. In general, with the release of new medical innovations and technologies to support T1D self-management, researchers should prioritize studies examining how they may also work to promote physical activity in youth.

Lastly, from the point of diagnosis onward, clinicians should engage in routine physical activity consultations with youth. These consultations should emphasize the potential physiological and psychological benefits of regular physical activity, provide instruction in safe glucose management strategies during exercise, offer personalized recommendations for physical activity, and help youth to problem solve potential barriers to exercise. Additionally, it may be helpful to involve caregivers, teachers, coaches, and peers in physical activity programs to further promote youth engagement. It is possible application of a multifaceted approach to promoting physical activity can help to increase the proportion of youth with T1D who meet their daily physical activity target.