New paradigms to address long-term exercise adherence, an Achilles heel of lifestyle interventions

Regular physical activity is widely recommended for the primary and secondary prevention of cardiovascular disease (CVD), and formal cardiac rehabilitation (CR) programs are indicated for a range of cardiovascular and post-surgical conditions. Heart failure (HF) is one of the conditions for which multi-component CR and even exercise training (ET) alone improve both exercise capacity and clinical outcomes^1,2 As a result, CR (including ET, medication education, dietary recommendations, and psychosocial support) and ET alone are codified in US and European guidelines for management of patients with HF regardless of ejection fraction, as class 1 to 2a recommendations. Of note, the efficacy of exercise intervention is significantly impacted by adherence.² Short-term adherence rates of up to ~80% can be achieved in rigorous clinical trials of exercise interventions,² but longer-term sustainability remains either poor or untested globally. This is particularly true once individuals transition out of in-person supervised sessions to the home environment. Several clinical factors are associated with adherence, but these account for very little of the variance in individual adherence.² This leaves an enormous gap in the knowledge base necessary to sustain exercise therapy in this high-risk population. Traditional efforts to improve adherence (as detailed below) have demonstrated important, albeit modest benefits.³ For HF in particular, there are relatively few recent or ongoing clinical trials investigating adherence to exercise. Increased efforts through alternative paradigms may be necessary to reduce the knowledge gap around effective methods to enhance adherence (including in diverse populations). From this perspective, we briefly review traditional methods to optimize adherence in exercise for CVD, with HF as a case study, and discuss innovative strategies to overcome barriers to long-term adherence.

ADHERENCE AND TRADITIONAL METHODS

The definition of exercise adherence varies across studies. It is typically recorded as percentage of visits attended and/or minutes per week of prescribed exercise completed.^1–3 The greatest adherence rates have historically been achieved through supervised in-person visits under clinical trial conditions. Transitioning to local facility or home-based exercise regimens has posed challenges for maintenance of adherence.¹ Traditional methods to target long-term engagement have included specific exercise prescriptions, education programs, activity logs, remote monitoring, motivational interviewing/coaching, problem-solving support, logistical/transportation assistance, home environment assessments, and resource provision (e.g., home supplies and local parks/fitness facilities).^2,3 These interventions have supported a degree of intervention sustainability; but results have been variable and relatively modest if successful in improving adherence rates in HF.³ For instance, the HEART Camp multicomponent intervention (81% HFrEF, 19% HFpEF) is one of the few studies to have demonstrated long-term efficacy associated with better adherence at 12 months and 18 months, but adherence rates remained modest overall (42% versus 28% in usual care at 12 months, and 35% versus 19% at 18 months).⁴ Similarly, HF-ACTION was globally the largest, extended trial of exercise training in HFrEF with sites in the US, Canada, and France. It included several features to target adherence, including measuring session attendance, providing activity logs, performing telephone and clinic follow-up, and monitoring heart rate data during the home exercise training phase. Nevertheless, adherence (defined as exercise minutes per week at or above exercise prescription, as measured by participant reporting) attenuated to 40% with median exercise time more than 15%–20% below target over the first 12 months in this closely followed cohort.¹ Similarly, the recent OptimEx-Clin program, studying exercise intensities in HFpEF across 5 European sites, demonstrated ~48–59% adherence (defined as ≥ 70% of scheduled exercise sessions) in the 9-month home-based phase of the program.⁵

NEW STRATEGIES FOR EXERCISE ADHERENCE

With this background, new strategies may be required to drive longer-term sustainability. For instance, one might consider loading and maintenance dose theory in exercise for HF states (Figure). This principle – long-term pharmacologic and device therapy with adjustments over time – is well-established across multiple fields including oncology and HF. On the other hand, exercise interventions, often due to resource and financial constraints, are typically utilized as initial intensive therapy and transitioned ‘back’ to patient-directed home/local exercise habits; but the long-term ‘legacy’ effects of exercise interventions in HF are unproven. Utilization of intermittent, in-person ‘maintenance doses’ of exercise training could improve both efficacy and long-term adherence. These intermittent doses could be deployed as scheduled, planned regimens, or triggered by recurrences, which, once stabilized, could be treated with re-loading of ET dose (as in chemotherapy regimens). Overcoming reimbursement barriers (varying across countries) and leveraging technology such as mobile health platforms (particularly in rural communities where direct access to exercise locations may be more difficult) may prove critical to success of this paradigm. Still, the investment may be justified, particularly in a condition such as HFpEF, where there are few effective therapies and for which physical inactivity likely plays a driving role in the underlying pathophysiology and disease progression.

Figure:

Innovative paradigms for adherence and sustainability in exercise interventions for heart failure

Another strategy would be to consider risk stratification specific to adherence/sustainability risk. For instance, barriers to adherence could be assessed at initiation of intervention with patients then stratified depending on risk assessment for poor adherence into lower-touch, lower cost regimens versus closer monitoring, higher cost regimens. This could help address the issue of limited in-person, facility resources and also help to drive towards greater equity by increasing services in at-risk, underserved populations. This is a paradigm for future consideration, given the current inability to predict the variance in long-term adherence with traditional models; a combination of advanced modeling and/or direct participant interviews and surveys may be necessary to close this gap in understanding and enabling this approach.

Third, the use of behavioral economics strategies in direct patient care is an area of growing interest in CVD; however, this approach remains relatively underutilized, particularly in exercise interventions for patients with HF. Utilizing tools such as insurance incentives, commitment contracts, and social networks have the potential to promote greater longevity and sustainability in achieving exercise targets; these strategies could be phenotype-directed to address issues of patient heterogeneity, prominent in HFpEF for instance. Small, short-term studies have efficacy in increasing physical activity in cardiovascular patients, but further research is necessary in the HF field.

Each of these strategies might be reasonably employed in industrialized countries or developing nations, although facility access and transportation may present challenges for intermittent dosing of exercise therapy and behavior economics strategies would vary across regional settings. Across all of these strategies, trials focusing primarily on the outcome of intermediate-term and long-term adherence may be most beneficial for the field; this could come through studies across the early phases of the NIH Stage Model for Behavioral Intervention Development including basic science (Stage 0), intervention development and pilot testing (Stage 1), and traditional efficacy testing (Stage II). For instance, Stage I/II studies might be pursued in the form of a platform trial, assessing adherence over time across intervention strategies and pivoting through interim analyses. Sample sizes could be more modest for these initial trials, particularly in HFpEF for instance, followed by larger trials with broader eligibility (e.g. variety of settings - primary/tertiary, rural/urban facilities) including clinical outcomes and adherence outcomes. Several lines of research would be useful including: 1) developing improved data on barriers to greater adherence or reasons for withdrawal from intervention programs; 2) determination of minimal necessary exercise ‘dosing’ in specific disease states such as HF; and 3) a greater understanding of the complex cardiovascular-peripheral-neurological pathways driving motivation for and response to exercise. Technology-enabled platforms for telemonitoring, prompts, and therapeutic training may help support testing of these paradigms. Moreover, the lexicon utilized in these efforts should be chosen with care. The reframing of pharmacologic ‘non-compliance’ to ‘barriers to adherence’ for medication use represents one patient-centered approach. This reframing may prove to be as, or more, important in exercise interventions, which are intrinsically more challenging for patients, particularly those with exertion-limiting conditions like HF.

CONCLUSIONS

Traditional models have produced a strong body of evidence regarding the mechanisms and clinical benefits of exercise and rehabilitation interventions in CVD. Yet, long-term adherence to exercise once in-person interventions cease in CVD disease states such as HF remains poor or unproven. Innovative paradigms, such as intermittent dosing, risk stratification by barriers to adherence, and behavioral economics tools may be required to overcome this Achilles heel of exercise therapy.