Initial interventions for the affected arm are aimed at eliciting small voluntary or reflexive involuntary movements. Training the arm for functional activities can start when the arm and hand begin to overcome gravity. If no hand dexterity is apparent by six weeks after stroke,3 further rehabilitation should emphasize maintaining a comfortably mobile arm.
Protocols for constraint-induced–movement therapy are based on the theory that nonuse of an arm may result from the great effort required to perform tasks with the affected hand.29 This therapeutic approach emphasizes intensive task-specific practice with the affected hand, for three to six hours a day over a period of from two to six weeks. As in other task-oriented interventions, the therapy progresses by gradual approximations from minimal movements to fuller actions of reaching for and grasping and pinching items that a patient would ordinarily use (Table 1). Some protocols restrain the unaffected hand to force greater use of the affected hand. The key requirement, and the primary limitation, of this approach is that patients must have at least 10 degrees of finger and wrist extension, which equates to rather good motor control, in order to benefit from the therapy. As a result, perhaps only 10 percent of patients may benefit from this strategy.30 Clinical trials indicate that patients who practice at this intensity increase the amount and efficiency of use of the affected hand by 20 to 50 percent, whether therapy is begun while they are inpatients or a year after the stroke.31,32
Many technology-assisted approaches to improve the functional use of a hand have been tried. In small trials, electromyographic biofeedback, which is used to help patients focus on increasing the force of contraction necessary to move a paretic wrist or to improve pinching and grasping, has increased movement but has less often improved the functional use of the hand.33,34 Larger randomized trials of the use of acupuncture to improve functional use of the hand have shown little or no benefit when sham puncture (i.e., insertion of the needle outside a traditional zone or without stimulation) served as a control.35
Spasticity is often blamed for poor hand function in patients with minimal wrist and finger extension but some preservation of flexion. Contracture and change in the morphologic features of muscle36 contribute to a flexed posture, but poor motor control with weakness, not hypertonia, is the primary cause of disability.37 In a randomized trial, the injection of botulinum toxin into the muscles of the arm reduced excessive flexion and the associated pain, spasms, or postures that interfered with patients' self-care,38 but the induced muscle weakness usually does not improve the functional use of the hand. The injection must be followed by stretching exercises and treatment of pain that may exacerbate the hypertonicity.
Walking
Independent walking is a primary goal and a reasonable expectation for most patients. Inpatients who develop at least partial movement against gravity for hip flexion and knee extension will progress from single steps taken between parallel bars to at least slow walking for short distances with the use of a walker or cane and hands-on assistance from another person for balance. A molded plastic orthosis for the ankle and foot may stabilize a weak ankle and knee to improve foot clearance and prevent the knee from buckling39 without impeding the subsequent recovery of dorsiflexion in the ankle. The recovery of the ability to walk even short distances requires considerable practice. Even in a dedicated rehabilitation unit, however, patients may spend little more than 15 minutes a day engaged in mobility tasks.40
Randomized trials show that patients who walk slowly (i.e., less than 0.4 m per second) can improve their speed and endurance significantly even when treatment begins from 3 months to more than 12 months after a stroke. Successful interventions involve one or more cycles of from 12 to 20 hours of task-oriented exercise over a period of four weeks, such as practicing walking outdoors or on a treadmill or obstacle course and exercising to improve leg strength and balance.10,25 Training on a treadmill with the use of partial body-weight support (the attachment of an overhead lift to a harness on the patient's chest) allows patients to take more steps at faster speeds than may otherwise be feasible. Randomized trials initiated during inpatient rehabilitation, however, have shown clinically insignificant increases in speed and distance of walking, as compared with conventional training. More data are needed to make it possible to assess whether this approach can improve walking among patients who persistently walk poorly at three to six months after a stroke, when gains have reached an apparent plateau after conventional therapy.41
Exercise and Strengthening
In patients with hemiparesis, the rates of production of muscle force, power, speed of sequential movements, and resistance to fatigue are impaired. Although neural factors that affect motor control account for much of the impairment, changes in muscle fibers and atrophy induced by inactivity may contribute. In randomized trials, progressive resistance exercises performed three to four times weekly for a period of from 6 to 12 weeks by patients with adequate motor control improved strength and functional activities.42 Fitness often declines in disabled persons, but trials show gains with progressive aerobic exercises, such as walking on a treadmill three days a week, that are tailored to each patient's tolerance, even when the exercise is initiated years after a stroke.43
Other Factors Affecting Rehabilitation
Attention to modifiable conditions that may interfere with the effectiveness of efforts toward rehabilitation is routinely warranted. For example, preexisting conditions such as painful osteoarthritis or cardiopulmonary disease may limit exercise tolerance. Other modifiable factors that limit participation in exercise include sleep disorders, pain, adverse effects of medications (such as orthostatic hypotension and impaired concentration), mood disorders, and incontinence or urinary retention. Depression is especially prevalent, affecting 25 to 40 percent of patients within the first year after a stroke.44 Randomized, placebo-controlled trials of citalopram, fluoxetine, methylphenidate, and nortriptyline suggest that mid-range doses of these medications (e.g., a dose of 50 to 100 mg when a low dose is <50 mg and a high dose is >100 mg) increase patients' participation in rehabilitation activities45 and may lessen cognitive deficits in some depressed patients.46 Other issues that are not medical, such as the costs of a caregiver and of remodeling the home to eliminate physical barriers, also affect patients' efforts and goals for rehabilitation.
Areas of Uncertainty
Formal therapy is often stopped when patients show no qualitative gains after a few weeks of treatment. A plateau in recovery, however, does not necessarily imply a diminished capacity for further gains in physical speed or precision or in learning a new task. At present, the opportunity to achieve maximal improvement is probably constrained by a lack of adequate data to define the optimal intensity (performance time, pace, and duration) of training strategies for specific disabilities. Functional neuroimaging studies may, in the future, help to guide decisions about the type and duration of treatment by providing insight into the maximal cortical reorganization that can be achieved with a particular therapy over time47,48; however, this possibility requires much more research.
Several potential interventions need further study. Small trials have shown modest clinical improvement in disabilities after stroke with the use of the following techniques: electrical stimulation over the surface of muscles to contract them for simple movements, such as grasping, or to assist ankle dorsiflexion while walking49; intense practice with electromechanical devices that assist in reaching or stepping50; noninvasive stimulation of the peripheral nerve of the arm51 or direct stimulation of the motor cortex over the hand representation52 to augment cortical plasticity and learning during arm therapies; pharmacotherapy with agonists of dopamine, acetylcholine, and serotonin, which may modulate neurotransmission and learning53; and the use of mental imagery of an action,54 which may enhance training because it activates many of the same cortical neurons that are involved in performing the action. Phase 1 trials are beginning in order to assess the safety of the injection of drugs into the cerebrospinal fluid or of cells into brain tissue to replace neurons and promote dendrite sprouting and axon regeneration,55 with the goal of possible neural repair.