Gait recovery is a major goal in rehabilitation programs for people with stroke and is very important in their ability to regain independence in activities of daily life. Recent research suggests that arm swing and walking performance are closely related and affect each other. Applying additional weights to the arms could be a possible intervention to facilitate hemiparetic gait. The purposes of this dissertation were to investigate the effects of adding weights to the arms on the arm swing amplitude, gait performance, and muscle activity in healthy people (study 1) and people with hemiparesis following stroke (study 2).

In the first study, twenty-two healthy subjects (10 male, 12 female;​ M±SD age:​ 23.5±2.1 years, height:​ 1.71±0.07 m, mass:​ 67.8±10.9 kg, BMI:​ 23.0±3.1) walked at their preferred speed across a level 6 m walkway under different weight carriage conditions (no weight, C1;​ weight (0.45 kg) on unilateral arm, C2;​ weights on bilateral arms, C3;​ waist weights, C4). When additional weights were added, subjects significantly increased gait speed (p=0.002, C1:​ 1.21±0.08 (M±SD);​ C2:​ 1.25±0.11;​ C3:​ 1.24±0.11;​ C4:​ 1.25±0.11 m/s) with an increase in cadence (p<​0.001, C1:​ 106.14±5.05;​ C2:​ 108.92±5.48;​ C3:​ 108.14±5.26;​ C4:​ 108.00±5.66 steps/min) and electromyographic (EMG) activity in the tibialis anterior (p=0.007, C1:​ 21.05±4.59;​ C2:​ 25.10±6.10;​ C3:​ 23.93±4.75;​ C4:​ 24.33±6.32 μV) muscle.

In the second study, nine hemiparetic stroke (8 male, 1 female;​ M±SD age:​ 58.0±6.8 years, height:​ 1.74±0.07 m, mass:​ 78.9±12.0 kg, BMI:​ 25.9±3.8, stroke duration:​ 91.4±67.9 months) and nine healthy (8 male, 1 female;​ M±SD age:​ 57.0±6.2 years, height:​ 1.74±0.08 m, mass:​ 74.2±8.0 kg, BMI:​ 24.4±1.3) subjects participated in the study. All subjects walked at their preferred speed across a level 6 m walkway under different weight carriage conditions (stroke group:​ no weight, C1;​ weight on uninvolved arm, C2;​ weight on involved arm, C3;​ bilateral arm weights, C4;​ healthy group:​ no weight, C1;​ weight on dominant arm, C2;​ weight on non-dominant arm, C3;​ bilateral arm weights, C4). The effects of arm weights on gait performance in stroke patients and healthy subjects were statistically inconclusive (αadjusted<​p<​0.05) due to low statistical power caused by the small sample size and a large variability between subjects. In stroke patients, gait speed (p=0.048, C1:​ 0.64±0.26 (M±SD);​ C2:​ 0.66±0.26;​ C3:​ 0.70±.25;​ C4:​ 0.69±0.27 m/s) and integrated EMG (iEMG) values from the tibialis anterior on the involved side (p=0.018, C1:​ 49.59±13.3;​ C2:​ 45.00±12.7;​ C3:​ 43.29±14.0;​ C4:​ 44.88±13.9 μV) exhibited changes with the addition of arm weights that were not statistically detectable using the more stringent Hochberg correction. In healthy subjects, iEMG values from the posterior deltoid variable on the non-dominant side was also statistically inconclusive (p=0.022, C1:​ 24.98±30.0;​ C2:​ 25.37±28.5;​ C3:​ 30.13±31.7;​ C4:​ 28.88±33.3 μV). Although no statistically conclusive findings were observed, several variables exhibited large and potentially clinically meaningful changes. When individual subject response patterns were explored using descriptive statistics (mean difference between weight carriage conditions) and effect sizes (Cohen’s d), 7 of 9 stroke subjects increased gait speed in the changes from C1 to C2 (M±SD increase:​ 0.060±0.029 m/s;​ 4 subjects exhibiting large d effect sizes, where 0.8 is considered a large d effect size), 7 of 9 subjects increased gait speed from C1 to C3 (M±SD increase:​ 0.091±0.053 m/s;​ 6 subjects exhibiting large d effect sizes), and 8 of 9 subjects increased gait speed from C1 to C4 (M±SD increase:​ 0.056±0.052 m/s;​ 5 subjects exhibiting large d effect sizes). Five of 9 healthy subjects increased gait speed in the changes from C1 to C2 (M±SD increase:​ 0.078±0.034 m/s;​ 5 subjects exhibiting large d effect sizes), 7 subjects increased gait speed from C1 to C3 (M±SD increase:​ 0.076±0.072 m/s;​ 4 subjects exhibiting large d effect sizes), and 6 subjects increased gait speed from C1 to C4 (M±SD increase:​ 0.054±0.030 m/s;​ 5 subjects exhibiting large d effect sizes).

These results indicate that additional weights can improve gait performance in some individuals. The observed increases in gait speed in the current study demonstrated encouraging results for higher functioning stroke patients who exhibit gait impairment and asymmetry. The addition of weights to the arms merits further investigation as a potential rehabilitation in people with stroke-related gait disturbances