Predicting Dosage of Balance Training Interventions

Introduction: Balance training is an evidence-based fall prevention intervention, and optimal dosage parameters (frequency, intensity, type, time) are essential for effectiveness in adults with balance impairments and fall risk. Evidence is limited on the dosage parameters clinicians use in practice. This study’s purpose was to determine the time spent in active exercise and balance-specific activities during routine physical therapy and examine if any patient characteristics predict time spent exercising. 

Methods: Trained researchers conducted real-time naturalistic observations of 135 outpatient sessions from 98 adults with balance impairments, categorizing activities via a standardized Activity Categorization Form and Balance Intensity Scale. Pearson and Spearman correlations examined relationships between active exercise time and patient variables; multiple linear regression evaluated prediction of active exercise time and proportion of balance activity time.

Results: Mean active exercise time was 49.7% ± 19.1% and mean time spent on balance activities 28.1% ± 18.6%). Patients were a mean age of 70.4 years ± 12, 69.6% female, with mean chronicity of 9 years. Common diagnoses included musculoskeletal and neuromuscular conditions. Patients had mean 2.38 ± 4.2 one-year fall history and most required no-to-minimal assistance for mobility. Significant associations existed between active time and assistance level for transfers (ρ=0.226, p=0.009), standing (ρ=0.173, p=0.046), walking (ρ=0.185, p=0.033), and stairs (ρ=0.235, p=0.007). Age, sex, fall history, chronicity, pain, and assistive device use were not associated with active time. Only assistance-level for transfers (ρ=0.291, p=0.0006) showed a statistically significant correlation with balance activity time. A regression model predicted active exercise time from transfer assistance (β=0.096, p=0.001) and age (β=−0.002, p=0.038, R²=0.103). Regression model predicted the proportion of balance activities time from transfer assistance (β=0.105, p=0.0001; R²=0.110).

Conclusions: Patients requiring greater assistance with functional mobility engage in more, and older patients in slightly less, of the active exercise necessary for change in fall risk.