Patient characteristics associated with participation in cardiorespiratory exercise during stroke rehabilitation: a multisite observational cohort study

This multisite observational cohort study found that only 40.1% of stroke rehabilitation patients participated in cardiorespiratory exercise, with participation significantly hindered by cardiac and mental health conditions or severe mobility limitations, and even among those who exercised, merely 1.5% met recommended intensity and duration guidelines.

The Gist

Imagine your body after a stroke is like a car that has just been in a major accident. The engine (your brain) has taken a hit, and the wheels (your limbs) might not turn as smoothly as before. To get this car back on the road, you need to do more than just fix the dents; you need to rebuild the engine's stamina. That "stamina" is called cardiorespiratory fitness—your heart and lungs' ability to keep you going when you're walking, climbing stairs, or just getting dressed.

This study is like a massive inspection of four different auto repair shops (hospitals) in Ontario, Canada. The researchers wanted to answer three big questions:

1. How many patients actually got their engines tuned up with cardio exercise?
2. What kind of patients got the "tune-up," and who got left out?
3. Did the mechanics (therapists) follow the manufacturer's manual (medical guidelines) when they did the work?

Here is the breakdown of what they found, using some everyday metaphors.

1. The "Participation Rate": Only 4 out of 10 Got the Tune-Up

Out of nearly 1,500 patients who came in for rehab, only about 40% actually did any cardio exercise.

• The Analogy: Imagine a gym where 100 people sign up for a membership, but only 40 of them actually step on a treadmill.
• The Twist: Even among those who exercised, only about 26% had it officially written into their "repair plan" (prescribed by a doctor). The rest just happened to do it "incidentally" (maybe they hopped on a bike because it was there, not because it was a formal order).

2. The "Gatekeepers": Who Got Left Out?

The researchers looked for patterns to see who was more likely to get the exercise and who wasn't. Think of these as traffic lights controlling who gets to drive.

The Red Lights (Barriers to Exercise):

• Heart Trouble: If a patient already had heart disease or a weak heart pump, they were much less likely to exercise. It's like a mechanic saying, "We can't rev this engine; it might blow a gasket."
• Mental Health Struggles: Patients dealing with depression or anxiety were less likely to participate. It's like trying to drive a car when you're too scared to turn the key.
• The Wheelchair Trap: Patients who arrived in a wheelchair (unable to walk at all) were significantly less likely to get cardio exercise. The study suggests that because they needed a wheelchair, the therapists might have thought, "They can't walk, so we can't do cardio."
• Low Walking Ability: Even if they could walk a little, if they needed a lot of help, they were less likely to get the exercise.

The Green Lights (Factors that Helped):

• The Walker/Rollator: Patients who could walk with a walker or rollator were more likely to get exercise. It seems having some independence was the key to getting the green light.
• Having a Goal: If the patient said, "I want to get my endurance back," they were twice as likely to get the exercise.
• Time: Patients who stayed in rehab longer had more chances to get the exercise.

3. The "Manual Check": Did They Follow the Rules?

This is the most surprising part. Medical guidelines say cardio exercise should be done at a moderate pace for at least 20 minutes to actually rebuild that engine stamina.

• The Reality Check: The study found that only 1.5% of patients (9 out of 601) actually met this standard.
• The Analogy: Imagine a nutritionist telling you to eat a full, healthy meal for 20 minutes to get strong. But 98% of the people only ate a single cracker for 10 minutes and called it a meal.
• Why? The average session was only about 13 minutes. Therapists often didn't have the right equipment, the time, or the confidence to push patients to that "moderate" level. They were mostly using a specific machine called a "recumbent stepper" (a seated stepping machine), which is great, but they didn't keep the patients on it long enough to make a real difference.

The Big Picture

The study concludes that while stroke rehab is great for learning to walk again, it's missing a huge opportunity to rebuild heart and lung health.

• The Problem: We are treating the "wheels" (legs) but often ignoring the "engine" (heart/lungs), especially for people who are sicker, have mental health struggles, or use wheelchairs.
• The Solution: We need to stop treating heart disease or mental health as a reason to stop exercise. Instead, we need to find safe, adapted ways to get everyone moving, and we need to make sure they move long enough (20+ minutes) to actually get the benefits.

In short: The repair shops are open, and the tools are there, but they aren't using the full repair kit on enough cars, and they aren't driving them long enough to really fix the engine.

Technical Summary

1. Problem Statement

Stroke survivors frequently suffer from low cardiorespiratory fitness (CRF), which impairs their ability to perform activities of daily living and increases long-term disability. While clinical guidelines recommend moderate-intensity cardiorespiratory exercise (≥20 minutes/session) for stroke rehabilitation to improve CRF and neuroplasticity, implementation in routine care is inconsistent.

• The Gap: Previous research suggests participation rates vary widely (6–42%) and are often lower than prescription rates. It remains unclear which specific patient characteristics (demographic, clinical, functional) predict participation in routine care across multiple sites, and whether prescribed exercises adhere to evidence-based dosage guidelines.
• Context: The study also sought to evaluate if the COVID-19 pandemic altered participation trends.

2. Methodology

• Study Design: Multisite observational cohort study.
• Setting: Four urban academic stroke rehabilitation centers in Ontario, Canada (two inpatient/outpatient, one outpatient-only, one inpatient-only).
• Participants:
  • Cohort: All patients admitted to stroke rehabilitation during specific windows: March/October 2019 (pre-pandemic) and a 12-month period starting in 2021 (post-pandemic).
  • Sample Size: 1,497 patients included in the final analysis (after excluding 60 due to opt-outs, missing charts, or readmissions).
• Data Collection: Retrospective chart review.
  • Variables: Demographics, socio-economic status, time post-stroke, length of stay (LOS), comorbidities (cardiac, mental health, etc.), and functional assessments (FIM, FAC, CMSA, MOCA, NIH-SS).
  • Exercise Classification:
    • Prescribed: Exercise included in the treatment plan.
    • Incidental: Exercise documented in logs but not in the formal plan.
    • Adherence: Defined as moderate intensity (RPE ≥3/10 or HR ≥55% max) for ≥20 minutes per session.
• Statistical Analysis:
  • Descriptive statistics and hypothesis testing (t-tests, Chi-square, Mann-Whitney U) to compare participants vs. non-participants.
  • Multivariable Logistic Regression: Stepwise regression to identify predictors of participation.
  • Model Validation: 500 iterations of repeated sample splitting (90% training, 10% testing) to assess Area Under the Receiver Operating Characteristic (AUROC), sensitivity, and specificity.
  • Imputation: Missing data handled via the missForest package (random forest imputation).

3. Key Contributions

• Multisite Generalizability: Unlike previous single-site studies, this research aggregates data from four distinct centers, providing a robust view of routine care practices in Ontario.
• Differentiation of "Prescribed" vs. "Incidental": The study distinguishes between formally prescribed exercise and incidental participation, revealing that a significant portion of exercise occurs outside formal treatment plans.
• Quantification of Guideline Adherence: It provides concrete data on the gap between clinical guidelines (20 mins moderate intensity) and actual clinical practice.
• Predictive Modeling: The study develops and validates predictive models for participation, identifying specific barriers (e.g., cardiac disease, wheelchair use) and facilitators (e.g., walker use, specific goals).

4. Key Results

• Participation Rates:
  • Overall: 40.1% (601/1,497) of patients participated in cardiorespiratory exercise.
  • Prescription vs. Participation: Only 26.4% had exercise formally prescribed; 13.8% participated incidentally.
  • Pandemic Impact: Participation increased slightly post-pandemic (41.4%) compared to pre-pandemic (34.7%), suggesting the pandemic was not the primary driver of low participation rates.
• Predictors of Participation (Odds Ratios - OR):
  • Negative Predictors (Barriers):
    • Diagnosed cardiac disease (OR = 0.74).
    • Poor left ventricular function (OR = 0.09).
    • History of mental health conditions (OR = 0.69).
    • Lower Functional Ambulation Category (FAC) scores (OR = 0.74).
    • Wheelchair use at admission (OR = 0.46 for indoor/outdoor use).
  • Positive Predictors (Facilitators):
    • Use of a walker or rollator at admission (OR = 3.22 for indoor/outdoor use).
    • Having a specific cardiorespiratory fitness goal in the treatment plan (OR = 2.13).
    • Longer length of stay (OR = 1.01 per day).
• Model Performance:
  • Full sample model AUROC: 0.732 (Moderate discrimination).
  • Inpatient-only model AUROC: 0.778 (Moderate to good discrimination).
• Guideline Adherence:
  • Critical Finding: Only 1.5% (9/601) of participants met the evidence-based guidelines (moderate intensity for ≥20 minutes).
  • Duration: The average session duration was 12.8 minutes. Over 80% of sessions were <20 minutes.
  • Intensity: Most sessions relied on equipment resistance/cadence rather than heart rate or RPE monitoring.
  • Testing: Only 2.2% underwent standardized Cardiopulmonary Exercise Testing (CPET); 58.7% used the 2-minute walk test.

5. Significance and Implications

• Clinical Practice Gap: There is a substantial disconnect between clinical guidelines and actual practice. While 40% of patients engage in some cardiorespiratory exercise, the vast majority do not receive the therapeutic dose required to elicit physiological adaptations.
• Barriers to Care: The study identifies that clinicians may be overly cautious regarding patients with cardiac comorbidities or severe mobility deficits (wheelchair users), potentially denying them beneficial exercise. Conversely, patients with moderate mobility deficits (using walkers) are more likely to participate.
• Mental Health: The association between mental health conditions and lower participation highlights the need for integrated mental health support to facilitate exercise engagement.
• Future Directions:
  • Rehabilitation programs require strategies to safely prescribe exercise for patients with cardiac and mobility comorbidities (e.g., adapted equipment, gradual progression).
  • Clinician training is needed to move beyond "incidental" exercise to structured, goal-oriented programs that meet dosage guidelines.
  • Implementation of standardized exercise testing (CPET or sub-maximal tests) is critical to safely prescribe and monitor intensity.

Conclusion: The study concludes that improving cardiorespiratory exercise participation in stroke rehabilitation requires addressing specific barriers (cardiac, mental health, mobility) and shifting clinical focus toward ensuring prescribed exercises meet evidence-based intensity and duration thresholds, rather than just increasing participation rates.