Task-Specific Locomotor / Gait Training

The single strongest message of the ANPT Locomotor clinical practice guideline (CPG) is about intensity and specificity: to improve walking speed and distance, get the patient walking, and make that walking hard enough. The CPG states that clinicians should use moderate- to high-intensity walking training to improve walking speed and distance in individuals more than 6 months after an acute-onset central nervous system injury (chronic stroke, incomplete spinal cord injury, or traumatic brain injury), compared with alternative interventions (Source: Hornby et al., 2020). For stroke, this is a strong recommendation (evidence quality I–II) (Source: Hornby et al., 2020).

"Moderate-to-high intensity" has a number attached. The CPG defines the effective aerobic intensity as 60%–80% of heart-rate reserve, or up to 85% of maximum heart rate (Source: Hornby et al., 2020). (Heart-rate reserve = maximum heart rate − resting heart rate; target zones are calculated from it with the Karvonen method.) This is meaningfully higher than the comfortable pace most patients self-select, which is exactly the point — low-intensity walking is the comparator that higher-intensity training outperforms.

The CPG also tells you what not to rely on when the goal is faster, farther walking. As the means to improve walking speed and distance in chronic stroke, clinicians should not use seated/standing non-walking balance training, body-weight–supported treadmill training (BWSTT), or robotic-assisted gait training — each a strong recommendation against (Source: Hornby et al., 2020). Note the framing: these are judged against the specific outcome of walking speed/distance, not banned for all purposes. Virtual reality is the opposite case — the CPG recommends using it coupled with walking practice (strong) (Source: Hornby et al., 2020). Strength training, cycling/recumbent stepping, and circuit training are weaker "may consider" options for stroke (Source: Hornby et al., 2020).

Where the guidelines differ (worth flagging). The older AHA/ASA stroke rehabilitation guideline lists body-weight–supported and robot-assisted walking as options that "may be considered," particularly for patients who are nonambulatory or have low ambulatory ability early after stroke (Source: Winstein et al., 2016). The more recent Locomotor CPG recommends against these specifically for improving walking speed and distance in chronic stroke (Source: Hornby et al., 2020). These are not flatly contradictory — they address different populations (early/nonambulatory vs. chronic ambulatory) and a different goal — but when you meet the conflict, default to the more recent, population-matched guidance and be clear about which population you're treating.

Safety matters at higher intensities: there may be increased cardiovascular risk without monitoring, so the CPG advises cardiovascular monitoring, consultation with the patient's physician before higher-intensity training, and excluding individuals with significant cardiovascular history for whom their physician does not recommend it (Source: Hornby et al., 2020).

Jargon: BWSTT = walking on a treadmill while a harness offloads part of body weight. HR reserve = max HR − resting HR.

---