Effects of Strength Training in Amyotrophic Lateral Sclerosis: How Much Do We Know?

The impact of various forms of exercise on function in patients with amyotrophic lateral sclerosis (ALS) has been a matter of debate and study for many years. The common mantra for most of the 20th century was that exercise was contraindicated in ALS for fear that driving sick motor units would accelerate disease. This hypothesis has been tested both in animal models and in human studies. In animals, studies have predominantly but not completely supported a view of exercise being beneficial. Kaspar et al. evaluated exercise duration from 2 to 12 hours in mice bearing the transgene for human mutant superoxide disumutase 1 (SOD1), and found a dose effect suggesting that 6 hours of exercise improved survival more than either 2 or 12 hours, but no dose was harmful. Additional studies have found modest survival benefits with a variety of exercise regimens as well. However, other studies have suggested increased survival and preservation of motor neurons at moderate exercise levels, but worsened survival at more intense levels. In addition, there may be an exercise specific effect; Deforges et al. noted that a high-intensity running regimen had no effect on either survival or spinal motor neuron counts, whereas a similar duration swimming regiment dramatically improved both of these. Human studies have evaluated both aerobic exercises and resistance training, and have either shown no effect or a benefit depending on the intervention and on the study. A limitation of all studies is that they are small, and the outcome measures assessed vary significantly. Bello-Haas et al. performed a small, unblinded study using personalized resistance exercises for 6 months, and found both an improvement in strength of exercised muscles as well as improvement in function, as measured by the ALS Functional Rating Scale—Revised (ALSFRS-R). A similar study using resistance exercises showed a small positive functional effect and improved strength at 3 months but not at 6 months. More recently, a multicenter, randomized trial of 60 patients compared standard care to an aerobic conditioning group and a resistance training group over a 6-month period. Aerobic conditioning was better tolerated than resistance training, but there were no changes either in the ALSFRS-R or in strength measures as compared with standard of care. Conversely, a 5-week, partially blinded trial combining aerobic and isometric resistance exercises showed improvements in muscle power, oxygen consumption, and fatigue in the exercising group. Overall, human studies evaluating exercise in ALS with respect to global function are difficult to perform and to compare, in contrast to animal models, which have a much more uniform phenotype. Because ALS patients at study entry vary tremendously in phenotype, exercise regimens have to be individualized, and it is likely that, if exercise is effective, very weak muscles may behave differently than strong muscles. In addition, the variation in outcomes measured make comparison of one study to another very difficult. One way to approach phenotypic variability is to focus on specific muscle groups or functional activities. A study looking at tongue force training in SOD1 transgenic mice compared a variety of tongue movements as well as bulbar motor neuron counts in transgenic animals with and without force training. The force training sessions were brief (6 minutes or less per day); however, animals undergoing this training actually performed worse than those not trained, without changes in motor neuron counts in the hypoglossal nucleus. Focused human studies have primarily evaluated respiratory and bulbar function. Pinto et al. evaluated the effect of inspiratory muscle training over 4 months in a small, randomized trial. Overall, there were no significant effects either on function as measured by the ALSFRS-R or on specific respiratory muscle measurements, including sniff nasal inspiratory pressure (SNIP), which directly assesses the muscles exercised. However, the small size of the study (13 subjects per group) likely precluded the ability to observe potentially important effects. Other studies were even smaller; taken together, no strong evidence for an effect on any respiratory measure See article on pages 40–46 in this issue.