Dal Bello‐Haas 2007.

Methods	6 month parallel group, randomized trial.
Participants	27 people with definite or probable, probable with laboratory‐supported ALS (El Escorial criteria), aged 41 to 80 years. Early stage ALS (Stage 1 or 2, Sinaki & Mulder staging). FVC > 90% predicted and ALSFRS > 30. Able to ambulate with or without assistive device. Able to breathe without any form of mechanical ventilation. Able to understand and comply with instructions.
Interventions	Exercise group: thrice weekly individualized, progressive, moderate intensity, moderate load resistance exercises and daily stretching exercises performed at home. Control group: daily stretching exercises performed at home.
Outcomes	Outcomes measured monthly for 6 months. Function, using ALSFRS*. Quality of life, using SF‐36*. Fatigue using FSS*. Muscle strength, using MVIC*.
Notes	Multi‐site trial. Treatment group drop‐out rates: 15.4.6% at 3 months, 38.4% at 6 months. Control group drop‐out rates: 0% at 3 months, 28.6% at 6 months. Continued high drop‐out rates continued after 6 months.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"twenty‐seven eligible, consecutive subjects were randomly assigned" (p. 2004) Comment: probably done
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding (performance bias and detection bias) All outcomes	Low risk	"all monthly evaluations were performed by a second physical therapist ...who was blinded to group assignment" (p. 2004) Comment: probably done "participants were prescribed an exercise program by a research physical therapist who was unblinded to group assignment" (p. 2004) Participants were not blinded to group assignment Comment: blinding of participants and prescribing physical therapist not feasible
Incomplete outcome data (attrition bias) All outcomes	Low risk	Enrolled subjects with FVC ≥90% of predicted and ALSFRS ≥ 30; high drop‐out rates: 28.6% usual care group and 38.4% resistance exercise group at 6 months; trend toward worse Physical Role SF‐36 subscale scores in the usual care group at baseline "dropout rates and reasons for dropout did not differ significantly between groups" (p. 2004) "we completed a stringent intention‐to‐treat analysis and imputed usual care group means at month 6 for missing data points for both the resistance exercise and the usual care group subjects" (p. 2004) Comment: probably done
Selective reporting (reporting bias)	Low risk	All data published
Other bias	Low risk	None noted