Sex Differences in Cardiac Rehabilitation Outcomes

Abstract

Cardiovascular disease (CVD) is a leading cause of morbidity and mortality in males and females in the United States and globally. Cardiac rehabilitation (CR) is recommended by the American Heart Association/American College of Cardiology for secondary prevention for patients with CVD. CR participation is associated with improved CVD risk factor management, quality of life, and exercise capacity as well as reductions in hospital admissions and mortality. Despite these advantageous clinical outcomes, significant sex disparities exist in outpatient phase II CR programming. This article reviews sex differences that are present in the spectrum of care provided by outpatient phase II CR programming (i.e., from referral to clinical management). We first review CR participation by detailing the sex disparities in the rates of CR referral, enrollment, and completion. In doing so, we discuss patient, health care provider, and social/environmental level barriers to CR participation with a particular emphasis on those barriers that majorly impact females. We also evaluate sex differences in the core components incorporated into CR programming (e.g., patient assessment, exercise training, hypertension management). Next, we review strategies to mitigate these sex differences in CR participation with a focus on automatic CR referral, female-only CR programming, and hybrid CR. Finally, we outline knowledge gaps and areas of future research to minimize and prevent sex differences in CR programming.

1. Introduction

Cardiovascular disease (CVD) is a leading cause of morbidity and mortality in males and females in the United States and globally¹. Sex differences exist in the development and diagnosis of CVD (e.g., presentation) that portends sex-specific disparities in clinical management and outcomes. For example, females are often older, have more comorbidities, and present with different symptomology when diagnosed with coronary artery disease^{2, 3}. It is now recognized that female-specific risk factors heighten CVD risk for females (e.g., preeclampsia and early menopause)^3–5. Despite this, a large proportion of primary care physicians and cardiologists generally lack awareness/knowledge relating to CVD risk and management specifically of females⁶. Further, there is evidence that females with coronary artery disease exhibit worse short-term outcomes compared to males^{2, 7–9}. As such, optimal secondary prevention strategies are crucial to improve patient care and outcomes for females with CVD.

The American Heart Association/American College of Cardiology (AHA/ACC) guidelines recommend cardiac rehabilitation (CR) for secondary prevention for patients with CVD (Class I recommendation)¹⁰. The Centers for Medicare and Medicaid Services currently provide coverage for CR for patients following acute myocardial infarction, coronary artery bypass surgery, heart valve repair or replacement, percutaneous transluminal coronary angioplasty or coronary stenting, and heart or heart-lung transplant as well as those with stable angina pectoris and heart failure with reduced ejection fraction. CR is a comprehensive, multidisciplinary chronic disease management program that incorporates core components including patient assessment, nutritional counseling, cardiovascular risk factor management, psychological management and counseling, exercise training and physical activity counseling^{11, 12}. As a result, substantial evidence indicates CR participation is effective in improving CVD risk factor management, quality of life, psychological management, and exercise capacity (discussed below). Importantly, CR participation is also associated with reductions in hospital readmissions and mortality^13–23. Despite these advantageous clinical outcomes, sex disparities are present in outpatient phase II CR programming (i.e., from referral to clinical management). In this review, we have focused on sex differences in CR programming (and subsequent outcomes) where we discuss potential contributing biological, environmental, and social factors. To date, the role that gender plays in impacting CR programming has been minimally investigated and, as such an important gap that needs to be addressed (see ‘Knowledge gaps and future directions’). In this review, we will first discuss sex differences across the spectrum of care provided by the CR program. In doing so, we will discuss sex differences in CR referral, enrollment, and completion as well as each of the core components of CR. These sex differences in CR participation and the core component outcomes are highlighted in Figure 1. Next, we will evaluate strategies to mitigate these sex differences in CR participation. Lastly, we will outline knowledge gaps and propose future directions to progress the field forward and identify mechanisms for these sex differences in CR programming as well as interventions to mitigate sex-specific disparities in clinical management and outcomes.

Figure 1.

This summary figure illustrates the occurrence of sex differences in CR participation as well as the core components of CR, based on the studies described herein. The down arrow represents a poorer outcome for females compared to males. The bidirectional arrow represents either a similar outcome between the sexes or that inadequate data is available on sex-specific differences for that outcome.

2. Sex differences in CR participation

CR participation rates are low in the United States and around the world. In 2016, ~24% of eligible Medicare beneficiaries participated in CR with ~57% and 27% of those beneficiaries completing >24 and 36 CR sessions, respectively²⁴. Importantly, sex-specific disparities occur in CR referral, enrollment, and completion. Li and colleagues using the American Heart Association Get with the Guidelines Coronary Artery Disease registry (n=48,993), found that females were 12% less likely to be referred to CR compared to males²⁵. Further, it has been reported that females are less likely than males to 1) enroll into CR, 2) attend CR sessions, and 3) complete the CR program^{16, 24, 26–29}. For example, Ritchey and colleagues found that females were 9% and 13% less likely to enroll and complete CR compared to males among Medicare beneficiaries in 2016²⁴. These sex differences were found after adjusting for age, which is an important consideration as females tend to have CVD events at a later age and age disparities are highly prevalent in CR participation^{24, 30}. These sex-specific disparities in CR participation also exist despite female-specific guideline recommendations in 2004 promoting CR³¹. Importantly, a dose response relationship exists between the number of CR sessions attended and the associated reduction in mortality and major adverse cardiac and cerebrovascular event rates^{16, 17, 20, 21, 23, 32}. Despite the low CR participation rates in females, CR has been shown to elicit similar or greater mortality benefit in females compared to males^{13, 14, 16, 19, 22, 33}.

Sex differences in the temporal trends in CR participation are unclear. In the United States, CR referral rate was reported to increase from 2007 to 2012 in males and females following acute myocardial infarction to the same extent (~7%)³⁴. Similarly, the proportion of females referred to CR in a single center in Toronto, Canada increased from 1996 to 2010 (23% to 32%)³⁵. In contrast, temporal improvements in CR participation have not been reported in females. In the United States, previous studies have found that CR participation in Olmsted County, MN for females following acute myocardial infarction remained unchanged from 1982 to 1998 as well as 1987 to 2010^{13, 36}. Among Medicare beneficiaries with HFrEF, CR participation has increased for males from 2014 to 2016, but not females²⁹. It is important to stress the lack of CR coverage by the Centers for Medicare and Medicaid Services for HF with preserved ejection fraction (HFpEF) despite the improvements in exercise capacity and quality of life following exercise training, as a greater proportion of HFpEF patients are females (versus HFrEF)^{37, 38}. Since 2012, the Million Hearts Cardiac Rehabilitation Collaborative (organized by the Centers for Disease Control and Prevention and Centers for Medicare and Medicaid Services) has been focused on implementing specific evidence-based interventions and benchmarks, which include increasing CR referral and participation to 70% of eligible patients by 2022³⁹. Further, there has recently been a greater emphasis on CVD management for females with the identification of female-specific risk factors coupled with the development of female-specific guidelines for CVD as well as multiple CV health awareness campaigns and initiatives directed towards females (e.g., AHA Go Red for Women)^{4, 40–42}. Future studies are warranted to determine if CR referral and participation rates have subsequently increased as a result of these initiatives as well as if the sex-specific disparities in CR participation have been minimized. Taken together, these findings provide an important foundation for identifying barriers to CR participation as well as potential mitigation strategies that ultimately improve CR participation in females.

Identifying barriers to CR participation has been a major topic of interest to develop and implement strategies to improve CR participation for all patients. Barriers to CR participation can occur at patient, health care provider, and social/environmental levels and are comprised of both modifiable and non-modifiable factors. The low CR referral rates (for males and females) is often considered the most important (and potentially the easiest barrier to overcome) as CR referral is required for CR enrollment. Although the barriers to CR referral are multifactorial, referral failure plays a significant role⁴³. To this point, females following revascularization procedures were less likely to receive a referral from their health care provider for CR compared to males⁴⁴. As elaborated below (see ‘Potential strategies to mitigate sex differences’), the implementation of an automatic electronic medical record-based CR referral system is a viable strategy to mitigate these sex differences in CR referral⁴⁵.

Among the numerous patient, health care provider, and social/environmental factors that impact CR utilization, there are many that particularly impact females^{46, 47}. For example, females are more likely to report transportation problems, family responsibilities, having multiple medical issues, and the perception of exercise as tiring or painful as factors contributing to not enrolling into CR⁴⁸. Strong endorsement from health care providers as well as spousal and family support to attend CR are directed toward females less frequently than males^{44, 49}. Further, older age (≥70 years) and lower education levels are negatively associated with CR enrollment for females^{50, 51}. Previous studies have also been conducted identifying sex-specific barriers for CR adherence/completion. Marzolini and colleagues found that the most common reason to withdraw from CR was lack of interest for both males and females⁵². Females, compared to males, were more likely to withdrawal from CR due to musculoskeletal issues, multiple medical issues, transportation problems, and family obligations⁵². Taken together, individualized CR programming that incorporates sex-specific factors will likely have a significant impact on CR participation as barriers can differ between sexes. Important, however, these interventions should be further tailored beyond biological sex as not all males or females exhibit the same barriers to CR participation. These findings indicate that tailored strategies need to be developed to improve CR referral and participation rates for both males and females.

3. Sex differences in the core components of CR

Patient assessment:

Patient assessment is the first core component to occur and provides the framework for optimal CR programming for each patient (e.g., individualized exercise prescription)¹¹. Often occurring prior to and/or at the beginning of CR, patient assessment includes medical history, physical examination, and testing (e.g., cardiopulmonary exercise testing). Despite its importance, there is paucity of studies investigating patient assessment strategies in CR as well as potential sex differences. De Feo and colleagues reported information related to patient assessments (via survey) from 1,677 males and 604 females enrolled in CR programs in Italy⁵³. Females were less likely to perform cardiopulmonary exercise tests at CR entry and completion compared to males, but females were more likely to perform submaximal exercise tests (i.e., 6-minute walk test). As discussed below, these sex differences have important implications for individualization of the exercise prescription (i.e., exercise intensity) used during the CR sessions. A comprehensive investigation on factors (e.g., sex, age, obesity) associated with patient assessments should be prioritized as it may reveal additional sex disparities which impact individualized treatment plans that are developed and used during the course of CR.

Exercise training:

Patients enrolled in CR are prescribed an individualized exercise prescription that generally begins with 20 minutes and progresses to 30 to 45 minutes of supervised aerobic exercise with an additional 10 to 15 minutes of strength training. To date, studies investigating sex differences in exercise prescription in the CR setting have focused on aerobic exercise training. It is important to note that strength training with and without aerobic exercise is associated with reductions in mortality with females exhibiting a greater mortality benefit in the general population^54–56. Future studies translating these findings to the CR setting as well as investigating sex differences would be highly impactful. Herein, we focus on sex differences in exercise capacity specific to aerobic exercise training in the CR setting. Previous studies have shown that exercise training performed in the outpatient phase II CR setting results in a 13–35% increase in peak aerobic capacity (i.e., peak oxygen uptake (VO₂peak)) with the majority of patients exhibiting improvements (~14–24% of CR patients do not exhibit an improvement in exercise capacity following CR)^57–61. These improvements have important clinical implications as increasing VO₂peak (and post-CR VO₂peak⁶²) is associated with lower readmission and mortality rates^{59, 60, 63}. In the general population, sex differences exist in VO₂peak that become more apparent with increasing age⁶⁴. Consistent with this, females enrolled in CR generally exhibit a lower VO₂peak prior to CR initiation compared to males^{57, 58, 65}. Importantly, numerous studies have reported sex differences in VO₂peak following CR with females having less of an improvement^{57–59, 66}. For example, Rengo and colleagues recently investigated if sex differences are present in the VO₂peak improvement following CR programming in 1,789 patients⁵⁷. Females exhibited less of an improvement in VO₂peak following CR than males (13 vs. 17%) despite performing a similar number of CR sessions suggesting that females garner less improvement in VO₂peak from each CR session. Further, a higher percentage of females failed to increase VO₂peak from pre to post-CR compared to males (24% vs. 16%)⁵⁷. These findings support the notion that established CR exercise prescription practices (based primarily on data from males) may result in females receiving suboptimal exercise ‘dosage’⁶⁷. It is likely that females (as well as males not exhibiting VO₂peak improvement) require a greater duration of CR (i.e., more CR sessions), higher exercise intensity, and/or increased frequency of these CR sessions to achieve VO₂peak improvement^{58, 61, 68}; however, future studies are clearly needed to confirm this hypothesis.

Many factors including exercise intensity, musculoskeletal issues, and differences in underlying physiology contribute to these sex differences in VO₂peak following CR. For example, higher exercise intensity (prescribed as a % of VO₂peak or peak heart rate) is one of the strongest predictors of increasing VO₂peak following CR^{58, 61}. However, it has been hypothesized that females are prescribed more conservative exercise intensities during CR because they are often older and have more comorbidities, which may contribute to their attenuated VO₂peak improvement following CR⁵⁷. Further, females often report higher ratings of perceived exertion for a given relative exercise intensity compared to males in the CR setting⁵⁸. Future studies investigating optimal exercise prescription strategies for females are necessary to maximize the VO₂peak improvement following CR. To this point, a recent study in patients with CVD investigated if sex differences existed in the improvement of estimated VO₂peak following high-intensity interval training performed 2 days per week for 10 weeks and comprised of 4 minutes of high intensity interspersed with 3 minutes of low intensity for 25 minutes⁶⁹. Females exhibited similar improvements in estimated VO₂peak compared to males following 10 weeks of high-intensity interval training performed in the CR setting (8.8 vs. 9.0%)⁶⁹. However, future studies that directly measure VO₂peak are necessary to confirm this finding because sex differences in VO₂peak improvement following CR can be masked by indirect measures of VO₂peak⁵⁷. Musculoskeletal issues are common in patients participating in CR with ~50% of these patients reporting musculoskeletal issues/complaints (e.g., lower back and knee pain primarily due to arthritis/joint problems)⁷⁰. Further, these musculoskeletal issues/complaints significantly contribute to disability and functional impairment⁷¹. Importantly, females with CVD are more likely to exhibit musculoskeletal issues than males with knee, hip, and back pain as being the most commonly reported concerns^{57, 70, 72–74}. As noted above, females are more likely to withdraw from a CR program compared to males due to musculoskeletal issues⁵². Further, these issues will negatively impact the patient’s ability to perform specific exercise modalities and intensities and as a result, these patients may have lesser increases in VO₂peak following CR⁷³. Taken together, individualized exercise prescription tailored to circumvent the most commonly reported musculoskeletal issues/complaints (of which the majority occur in females) may be an effective strategy to optimize CR adherence. Lastly, it is important to note that sex differences exist in multiple organ systems and the organ systems response to training (e.g., cardiac function and skeletal musculature)^{75, 76}. The extent to which these sex differences in exercise training mediated physiological adaptations translate to patients with CVD undergoing standard CR programming remains unclear.

Physical activity counseling:

The AHA/ACC Secondary Prevention goal for moderate to vigorous physical activity is ≥ 150 minutes per week⁷⁷. Similar to the general population, the vast majority (~80–85%) of patients enrolled in CR do not meet this goal^{78, 79}. A recent meta-analysis including randomized controlled trials found that CR was not associated with increased moderate to vigorous physical activity (three studies were included), but evidence of increased physical activity was seen with more steps per day following CR (five studies were included)⁸⁰. Large retrospective studies have found that moderate to vigorous physical activity (via self-report) increased following CR; however, ~40% of the patients did not meet the target goal of ≥ 150 minutes per week^{78, 79}. With respect to sex differences, females were less likely to reach the moderate to vigorous physical activity goal than males following CR⁷⁹. Investigating and quantifying time spent performing sedentary behaviors (e.g., sitting, watching television, and lying down) is also crucial as prolonged sedentary time is associated with greater risk of CVD development and mortality⁸¹. This is especially important in the CR population as these patients have been reported to have higher levels of sedentary time compared to healthy adults⁸². In 2,584 patients who had previously participated in CR, Bakker and colleagues found that male sex was associated with higher levels of sedentary time (measured via survey)⁸². These authors also investigated if physical activity behaviors (i.e., sedentary time and moderate to vigorous physical activity) changed with CR participation⁸². In 106 CR patients, sedentary time decreased following CR, while no changes were found in moderate to vigorous physical activity (both objectively measured via accelerometry). As low moderate to vigorous physical activity combined with high sedentary time elicit the most detrimental health outcomes⁸³, innovative strategies are needed to positively modify physical activity behaviors (i.e., increase moderate to vigorous physical activity and decrease prolonged sedentary time) in males, and particularly in females, following CR in both the short and long-term (e.g., PATHway I Trial⁸⁴). We direct the reader to a position statement on assessing physical activity by the American Association of Cardiovascular and Pulmonary Rehabilitation for a comprehensive explanation on the different physical activity domains, physical activity assessments, and practical considerations for measuring physical activity in CR programs⁸⁵.

Nutritional counseling:

As optimal dietary patterns (e.g., Mediterranean diet) are associated with lower risk of coronary heart disease and mortality^{86, 87}, nutritional counseling and subsequent optimization of dietary practices are a critical component of outpatient phase II CR programming. Nutritional counseling is comprised of dietary habit assessment followed by individualized prescribed dietary modification (by a dietician) with the goal of long-term adherence¹¹. In the general population, females are more likely to consume fruits and vegetables compared to males and less likely to adhere to a Western dietary pattern suggesting that females in CR may need different nutritional counseling^{88, 89}. Males and females referred to cardiologists similarly perceive diet as important for treating and preventing heart disease⁹⁰. However, there is minimal data available regarding sex differences in nutritional counseling and diet modification with CR programming. For example, Timin and colleagues investigated a nutritional education intervention consisting of two nutritional education classes and an individual diet counseling session during CR and reported that biological sex was not associated with the likelihood of improving diet or self-efficacy for diet behaviors⁹¹. However, males enrolled in CR perceive a higher degree of difficulty in following nutritional advice, poorer diet choices due to work, and lower satisfaction of low-fat foods compared to females⁹². These data highlight the need for novel studies investigating sex differences in diet modification and behaviors in response to CR on short and long-term physiological adaptations as well as patient and clinical outcomes⁹³.

Weight management:

Patients who are overweight and obese (defined as a body mass index (BMI) of 25–29.9 and ≥30 kg/m², respectively) are highly prevalent in the CR setting. For example, Bader and colleagues investigated the prevalence of obesity in 449 patients enrolled in their CR program and found that obese and overweight patients comprised 88% of their cohort⁹⁴. A follow-up study investigated how clinical characteristics of patients enrolled in CR have changed from 1996 to 2015 and found an increase in the prevalence of obesity during this time period⁶⁵. However, no sex differences were present in BMI or frequency of obesity (as indicated by BMI) in patients at CR entry which is consistent with previous studies^{58, 65, 94–96}. The AHA/ACC Secondary Prevention goal for weight management is a BMI of 18.5 to 24.9 kg/m²; however, this goal is one of the least likely to be achieved at CR completion^{77, 78}. It is recommended that overweight and obese patients enrolled in CR reduce body weight by >5–10% over a 6-month time period¹¹. However, previous studies have reported smaller reductions in body weight following CR (~0.5 to 3%)^{57, 58, 94–97}. Larger cohort studies investigating sex differences in CR have reported conflicting findings regarding weight management. Specifically, studies by Terada and colleagues as well as Lavie and Milani found that females, compared to males had smaller reductions in body weight following CR^{95, 96}. In contrast, Rengo and colleagues reported no sex differences in body weight improvement following CR (males and females both exhibited ~0.5 to 1% reductions in weight)⁵⁷. It is important to note that although BMI is often used to define obesity status, it does not characterize body composition (i.e., it does not differentiate between lean and fat mass). There is the potential that using BMI as a surrogate for ‘body composition’ may conceal reductions in adiposity through improvements in lean mass particularly in females⁹⁸. For example, Lavie and colleagues have shown that females exhibited significant reductions in body fat percentage (measured via skin fold), while no differences were present in BMI following 12 weeks for CR^{96, 99}. As a sexual dimorphism exists regarding the impact of lean and fat mass composition on cardiovascular mortality in the general population¹⁰⁰, future studies investigating sex differences in lean and fat mass composition as well as fat mass distribution (subcutaneous versus visceral) using gold standard measures such as dual energy X-ray absorptiometry following CR as well as subsequent patient-centric outcomes would provide valuable insights.

Lipid management:

Over the last 10–15 years, AHA/ACC guidelines on lipid management for secondary prevention have evolved with current guidelines recommending the use of high-intensity statin therapy for patients ≤75 years of age with atherosclerotic CVD who are not at very high risk with the goal of reducing low-density lipoprotein cholesterol (LDL-C) by ≥50%¹⁰¹. For those patients with atherosclerotic CVD at very high risk, high-intensity or maximal statin therapy is recommended (as well as non-statin therapy when needed)¹⁰¹. Importantly, a recent study including 88,256 patients found that high-intensity statin use following hospitalization for myocardial infarction for secondary prevention was lower for females compared to males (RR: 0.91 95% CI: 0.90 to 0.92)¹⁰². Further, there is evidence that high-intensity statin adherence is lower in females compared to males at 6 and 24 months following discharge for myocardial infarction among Medicare beneficiaries 66 to 75 years of age¹⁰³. As attending CR programming is associated with greater medication adherence/compliance (with a CR session dose response relationship present)^{104, 105}, CR attendance may minimize these sex differences in high-intensity statin use. Recently, Lin and colleagues investigated high-intensity statin use in 375 patients enrolled in CR from 2014 to 2015 who were ≤75 years of age. These authors found that the frequency of high-intensity statin use increased in only ~5% of the patients from CR entry to CR completion (CR entry: 211 vs. CR completion: 229). It is important to note that the increase in frequency of high-intensity statin use occurred in males only. To date, studies investigating sex differences in lipid management during CR have included patients primarily when the AHA/ACC lipid management guidelines recommended an LDL-C target goal of <100 mg/dL using statin therapy (<70 mg/dL for high-risk patients))^77–79. Approximately 50 to 65% of patients met this LDL-C target goal at CR entry, with large cohort studies (n=8,929 – 117,913) reporting that LDL-C decreased by 14–35 mg/dL (reductions of ~15–36%) following CR^{78, 79, 106} (with smaller cohort studies reporting smaller or no changes in LDL-C^{95–97, 99, 107, 108}). At CR entry, females were less likely to be at the LDL-C target goal compared to males but are reported to have a lower or similar frequency of meeting the LDL-C target goal at CR completion than males^{79, 106}. It is important to note that biological sex and menopause appear to modulate the lipid profile^109–111. Lavie and colleagues have proposed that the actuality/degree of improvement in lipid management following CR may be more apparent in subgroups of patients (i.e., with abnormal baseline levels) enrolling in CR¹¹². Taken together, the preliminary finding that sex differences exist in high-intensity statin use during CR coupled with the paucity of data investigating lipid management during CR following the 2013 and 2018 AHA/ACC lipid management guidelines signify an important gap in our clinical knowledge that must be addressed.

Diabetes management:

The prevalence of diabetes in patients enrolled in CR is ~22% and has been estimated to increase 0.3% each year⁶⁵. Although the prevalence of diabetes appears to be similar between males and females enrolled in CR^{65, 79, 113}, females with diabetes are less likely to attend CR compared to both females without diabetes and males (with and without diabetes)¹¹³. The AHA/ACC Secondary Prevention goal for CR patients with diabetes is to achieve a HbA1c of <7%⁷⁷. In a study including 2,522 patients with diabetes participating in CR, Turk-Adawi and colleagues found that males and females reduced HbA1c to the same extent following CR. However, females were less likely to achieve the AHA/ACC Secondary Prevention goal of a HbA1c <7.0% compared to males (adjusted OR: 0.50, 95% confidence interval: 0.32–0.78) at CR completion⁷⁹. These findings clearly indicate that innovative strategies to optimize diabetes management with CR programming are needed for females. Further, future studies providing a better understanding of the sex-specific pathophysiological mechanisms responsible for altered diabetes management in CR patients (e.g., differential pattern of adiposity storage) would be of great importance.

Hypertension management:

The prevalence of hypertension in patients enrolled in CR is estimated to be 60 to 70% and has increased over the past 20 years⁶⁵. The AHA/ACC Secondary Prevention goals for hypertension management is a systolic and diastolic blood pressure of <140 and 90, respectively (<130 and 80 for patients with diabetes, heart failure or chronic kidney disease¹¹)⁷⁷. At CR enrollment, females generally have a higher prevalence of hypertension diagnosis than males with menopause status likely playing an important role^{65, 114}. As such, it is not surprising that females enrolled in CR are less likely to meet the AHA/ACC Secondary Prevention Goals for systolic blood pressure at CR entry⁷⁹. Sex differences in blood pressure management during the CR program have been minimally investigated compared to other core components^{79, 95, 96, 115, 116}. In the study by Turk-Adawi and colleagues, significant (albeit small) reductions in systolic blood pressure in females (−1.6 mmHg), but not males were found following CR, while both groups exhibited similar reductions in diastolic blood pressure (0.5 to 1 mmHg)⁷⁹. The minimal reductions in blood pressure with CR may be reflective of a large proportion of patients already meeting the blood pressure AHA/ACC Secondary Prevention goals prior to CR entry with no sex differences present in the frequency of patients meeting these goals following CR⁷⁹ (>90% for males and females for systolic and diastolic blood pressure at CR completion^{78, 79}).

Smoking cessation:

The AHA/ACC Secondary Prevention goal for tobacco smoking is complete cessation⁷⁷. The prevalence of tobacco smoking in patients enrolled in CR is 5 to 16%^{65, 78, 79}. Although females have been reported to have a lower prevalence of tobacco smoking at CR entry compared to males^{79, 115}, this is not a consistent finding as other studies have found no sex differences^{65, 95, 96}. Strategies implemented during the CR program for smoking cessation are critical as smoking cessation is associated with a ~36% reduction in mortality in patients with CVD¹¹⁷. Despite the mortality benefit and cost-effectiveness of smoking cessation, smoking cessation is challenging due to the physical and psychological impact of nicotine addiction as well as social and behavioral factors. Importantly, CR attendance is associated with smoking cessation following CR with similar benefits reported for males and females⁷⁹. However, challenges exist regarding smoking cessation in the CR setting. First, smokers with an indication for CR are less likely to enroll and complete CR¹¹⁸. Second, ~50% of the patients enrolled in CR who smoke tobacco will continue following CR^{79, 119}. Third, only a small proportion of CR programs offer the behavioral and pharmacological interventions necessary for smoking cessation^{11, 118}. These data highlight the importance of developing multidisciplinary strategies and interventions in CR to maximize opportunities for complete smoking cessation.

Psychosocial management:

It is well recognized that psychosocial distress is associated with CVD development and mortality^{120, 121}. Psychosocial distress is multifactorial, but includes depression, anxiety, and hostility. Although inconsistent, females have been reported to have a higher prevalence of psychosocial distress compared to males at CR entry^{95, 115, 122–125}. Importantly, exercise training performed in the CR setting is associated with substantial reductions in the prevalence of psychosocial distress^{107, 108, 120, 125, 126}. To date, sex differences in psychosocial distress in response to CR have been controversial with studies reporting blunted, similar, and greater improvements for females compared to males^{69, 95, 122, 123, 127}. Numerous factors are responsible for these discrepant findings; however, two contributing factors include the specific component of psychosocial distress that was assessed and the indication for CR. Specifically, Hazelton and colleagues assessed levels of depression, anxiety, panic, anger, and relationship satisfaction in males (n=258) and females (n=122) prior to and following CR¹²⁵. Males and females exhibited similar improvements for depression, anxiety, and anger following CR, while females exhibited worse panic scores. These authors further investigated if sex differences were present in the CR mediated improvement in psychosocial distress in those patients exhibiting clinical levels of symptoms¹²⁵. The frequency of patients reporting clinically relevant symptoms of depression and anxiety decreased significantly and to a similar extent in males and females following CR. However, the frequency of patients reporting clinically relevant symptoms of anger decreased in males, but not females. Further, the frequency of relationship dissatisfaction was ~2 times more frequent in females than males prior to CR with neither sex exhibiting improvements following CR. Regarding the indication for CR, a recent study by Terada and colleagues investigated if sex differences existed in the improvement of anxiety and depression following CR for patients following coronary artery bypass graft (CABG) surgery and percutaneous coronary intervention (PCI)¹²³. In contrast to males, females following PCI exhibited worse anxiety levels following CR, while males and females following CABG exhibited similar improvements in anxiety¹²³. These findings emphasize that the manner in which the components of psychosocial distress are reported (change in score vs. change in frequency of those patients with clinically relevant symptoms) and the indication for CR can influence the identification of sex differences and may have important implications for the development of individualized therapeutic strategies.

Interdependence of core components:

There is preliminary evidence suggestive of interdependence of core components. Specifically, previous studies have reported that improvements in exercise capacity following CR are linked to improvements in psychosocial distress^{128, 129} and, importantly, those patients who do not exhibit improvements in both exercise capacity and psychosocial distress have a greater risk of mortality^{107, 126}. Consistent with this, recent evidence suggests that comprehensive CR delivery is most beneficial for patient outcomes (i.e., revascularization) as the core components can interact synergistically^{130, 131}. Future studies are necessary to elucidate interdependence of other core components as well as potential sex differences considering the numerous sex disparities in CR core components detailed above.

4. Potential strategies to mitigate sex differences in CR

One of the most opportunistic strategies to mitigate sex differences in CR is minimizing referral failure and subsequently improving CR referral. One recommended strategy to help alleviate this barrier is implementation of an automatic electronic medical record-based CR referral system^{39, 42, 132}. For example, Grace and colleagues surveyed coronary artery disease patients from 11 Ontario hospitals and found that automatic CR referral resulted in a higher CR referral rate compared with usual routes of referral (70% vs. 32%) with the addition of a staff member or liaison to help coordinate CR referral further increasing CR referral rates (85%)⁴⁵. Importantly, they found that automatic CR referral alone and with the liaison eliminated sex differences in CR referral rate and resulted in females being 6.5 to 10 times more likely to be referred to CR¹³³. It has been put forth that implementation of automatic CR referral with a liaison in all hospitals in the United States has the potential to triple the current CR referral rates³⁹, which has patient-oriented implications as CR referral is associated with reduced rehospitalization rates, greater survival rates, and CR enrollment^{25, 45, 133}.

Another ideal target to minimize sex differences in CR outcomes is female-specific modifications to outpatient phase II CR programming. Over the past decade, there has been the needed emergence of prospective studies investigating female-only CR programming^134–140. For example, Beckie and Beckstead investigated if a tailored CR program for females was more efficacious in improving many of the core components of CR outlined above compared to traditional CR programming^134–137. The tailored CR programming consisted of a behavioral change intervention (via motivational interviewing guided by the transtheoretical model) and females exclusively exercising with only females in the CR facility. Compared to traditional CR, females enrolled in the tailored CR program attended more CR sessions (i.e., higher CR adherence), had greater reductions in depressive symptoms, and greater increases in quality of life and perceptions of health^134–136. Further, the tailored CR program elicited similar improvements in resting systolic blood pressure, lipid profile, and exercise capacity in these females compared to the traditional CR program¹³⁷. These provide the foundation for future innovative, individualized CR programming opportunities for females. However, there are important challenges that need to be overcome associated with female-only CR programming that include logistics (e.g., this previous trial provided only one timeslot per day for female-only CR), cost, and flexibility (i.e., not all females prefer female-only CR¹³⁸)¹⁴¹. Taken together, the continued development of strategies that tailor CR programming for females is an essential approach to maximize CR outcomes for these patients.

Recently, there has been the accelerated implementation of innovative CR delivery strategies that have the potential to mitigate the sex differences elaborated above. The principal example is hybrid CR where, in addition to conventional center-based CR, the patient has the opportunity to engage in CR programming outside the confines of the CR center (e.g., at home)¹⁴². When performed at a location other than the conventional CR center, hybrid CR can be delivered in a synchronous or asynchronous manner (audiovisual care during exercise versus communication (e.g., telephone calls, texting) before or following exercise, both via telecommunication). Hybrid CR has the potential to address many of the CR participation barriers that are highly prevalent for females including transportation and family responsibilities (limiting the ability to attend center-based CR during the weekday). To this latter point, there is evidence that home-based CR may be a viable option to improve CR adherence in females¹³⁹. As recently detailed, synchronous and asynchronous hybrid CR is associated with specific disadvantageous and barriers that need to be addressed and future research will be critical to optimize this CR delivery strategy¹⁴³. Nevertheless, hybrid CR coupled with widely available telehealth technologies has the potential to improve CR participation particularly in women.

5. Knowledge gaps and future directions

There are numerous future lines of research that need to be addressed to better understand these sex differences in CR programming and clinical outcomes as well as strategies to mitigate these sex differences (see Table 1). First and foremost, there is a dire need for the inclusion of more females in prospective trials investigating CR programming and subsequent outcomes as females remain underrepresented in most scientific research and CR studies in particular²⁸. To this point, meta-analyses on the effect of CR on mortality and morbidity in females have not been able to be conducted due to the lack of available CR data in females^{144, 145}. Recommendations have recently been outlined by Norris and colleagues that range from investing in research training that is focused on sex differences in biomedical science that spans the clinical to translational spectrum to the development and implementation of policies that mandate reporting data by sex and eliminate barriers for the inclusion of females in research studies and clinical trials⁴.

Table 1.

Knowledge gaps and future directions

Comprehensive investigation of sex differences in the temporal trends of CR referral, enrollment, and completion by CR indication Determination of factors (e.g., biological sex) associated with specific patient assessments performed in the CR setting Identification of the optimal exercise prescription (e.g., non-traditional modality, greater duration of CR, and/or increased exercise intensity) to maximize VO2peak improvement following CR for females Mechanistic elucidation of potential sex differences in exercise training mediated physiological adaptations in patients with CVD Investigating the impact of CR attendance (and biological sex) on the frequency of those patients meeting guideline recommended lipid management in the CR setting Development of innovative, female-specific strategies to optimize diabetes management with CR programming Novel study designs investigating sex differences in diet modification and behaviors in response to CR Identifying the impact of gender verus biological sex on CR participation and outcomes Understanding the impact of female-specific risk enhancing factors on the spectrum of care provided by the CR program

CR, cardiac rehabilitation; CVD, cardiovascular disease; VO₂peak, peak oxygen uptake

Second, there is emerging evidence that suggests that gender, defined here as a psychosocial construct incorporating social norms, expectations, roles, and identities, plays an important role in clinical management and outcomes for patients with CVD^{146, 147}. In regards to gender and CR, Marzolini investigated the predictors of withdrawing from CR in males and females⁵². These authors identified younger age, marital status (i.e., being single), comorbidities (i.e., obesity and diabetes), lack of commonly prescribed cardiac medication (i.e., lipid lowering therapy and beta-blocker medication), and lower pre-CR VO₂peak as independent predictors of CR withdrawal, while biological sex was not a significant predictor of CR withdrawal. Future studies are necessary to directly determine the impact of gender versus biological sex on CR participation and barriers thereof, core components of CR, and subsequent outcomes.

Third, it is clear that exercise prescription in the outpatient phase II CR setting is not optimal for all patients and females specifically, as evidenced by the lower CR adherence and blunted improvements in VO₂peak in females following CR. As females more often perceive exercise as tiring or painful than males⁴⁸, future studies are needed investigating innovative exercise prescription strategies utilizing alternative approaches for the modality, intensity, and frequency. Recently, Vidal-Almela and colleagues identified numerous alternative modalities often employed in the community setting such as dancing, group walking, and Zumba that may be viable aerobic activities for females to incorporate in the CR program¹⁴⁸. To this point, Belardinelli and colleagues compared waltz dancing to cycling exercise when performed 3 times per week for 8 weeks in heart failure patients¹⁴⁹. These authors found that waltz dancing elicited similar increases in VO₂peak compared to cycling exercise (19 vs. 16%). Further, a meta-analysis by Oosenbrug and colleagues found that attending more sessions per week (i.e., 3 or more) eliminated the sex differences in CR adherence²⁸. There is a clear need for additional prospective, randomized clinical trials investigating innovative CR programming strategies for females (e.g., different exercise prescription strategies and female-specific tailored CR programming based on theoretical frameworks of behavioral change as discussed above).

Fourth, it is now recognized that sex-specific risk enhancing factors exist including premature menopause and preeclampsia with other female-specific factors also elevating CVD risk such as adverse pregnancy outcomes (e.g., gestational hypertension and diabetes), ovarian failure, polycystic ovarian syndrome, certain hormone-based contraceptive methods, and inflammatory disorders⁵. The extent as to which these female-specific factors influence the spectrum of care provided by the CR program (i.e., from referral to clinical management) needs to be addressed.

Fifth, the core components for CR were discussed in the framework of the AHA/ACC guidelines herein, while we have referenced and discussed findings from authors and CR centers in the United States, Canada, Italy, and others. It is important to note that several societies across the world have their own secondary prevention guidelines. Although outside the scope of this review, it is important to investigate if sex differences in CR programming and outcomes vary across countries as significant differences exist in CR availability (e.g., insurance), how CR is incorporated into the health system, core components of CR programming, and CR delivery models.

Nonstandard Abbreviations and Acronyms:

ACC

American College of Cardiology

AHA

American Heart Association

BMI

body mass index

CABG

coronary artery bypass graft

CR

cardiac rehabilitation

CVD

cardiovascular disease

HFpEF

heart failure with preserved ejection fraction

HFrEF

heart failure with reduced ejection fraction

LDL-C

low-density lipoprotein cholesterol

PCI

percutaneous coronary intervention

VO₂peak

peak oxygen uptake