Associations between real-life upper and lower limb activity and clinical assessments, in individuals with therapy-dependent spasticity after stroke

To investigate the relationships between real-life limb activity (measured by accelerometers) and clinical assessments in individuals with therapy-dependent spasticity in the chronic post-stroke phase.

Inclusion criteria: adults with confirmed stroke at least 3 months prior, referred to the spasticity team, able to walk at least 10 m with/without assistance. Four ActiGraph accelerometers (worn on wrists and ankles) measured real-life activity performance for 2 to 4 days while participants maintained their daily routines. Data was expressed as Vector Magnitude (VM) Counts per Minute for each accelerometer, and the activity ratio (more-affected/less-affected extremity) assessed asymmetry. Standardized clinical scales assessed impairment and activity capacity, including motor function (Fugl-Meyer Assessment, FMA), spasticity (Modified Ashworth Scale, MAS), upper extremity activity capacity (ARAT), and walking speed. Perceived activity performance was assessed using ABILHAND and ABILOCO. Data analyses required at least 20 hours data. Spearman correlation coefficients was used to assess the relationships between real-life activity and clinical assessments.

35 individuals (mean 56.8 years, 54% female), 57% ischemic strokes were included. Correlation strength was categorized as: 0.00–0.25 (very low), 0.26–0.49 (low), 0.50–0.69 (moderate), 0.70–0.89 (high), 0.90–1.00 (very high). Correlation between the real-life activity performance of the more-affected arm (accelerometry) and self-perceived performance in manual tasks (ABILHAND, 0.74) and locomotion (ABILOCO, 0.81) was strong. For the more-affected leg, accelerometry data showed strong correlations with self-perceived performance (ABILHAND, 0.70; ABILOCO, 0.82) and observed walking speed (0.76). Real-life arm ratio also correlated well with clinical assessments: arm motor function (FMA, 0.80), leg motor function (0.73), upper extremity activity capacity (ARAT, 0.77), walking speed (0.80), and self-perceived performance (ABILHAND, 0.73; ABILOCO, 0.82). In contrast, correlations for the less-affected limbs were generally lower (0.0018–0.49). All accelerometer-derived activity measures exhibited very low to low correlations (0.080–0.49) with spasticity. 

Using accelerometers to measure real-life limb activity is feasible for individuals with therapy-dependent stroke-related spasticity. It is recommended to focus on measuring real-life activity performance in the more-affected limbs or across both sides of the body given the typical  post-stroke asymmetry in limb function. The strong correlation between self-perceived activity performance and accelerometer-based activity of the more-affected limbs suggests these tools align with the objective accelerometer measures.

To improve daily functioning, objective outcome tools should be integrated into clinical and home settings. Understanding real-life activity can guide personalized rehabilitation, inform decisions, and support self-management in stroke recovery.