Makoto Suzuki, Yoshitsugu Omori, Seiichiro Sugimura, Masaaki Miyamoto, Yuko Sugimura, Hikari Kirimoto, Sumio Yamada
DOI: 10.2340/16501977-0877

Objective: To investigate the recovery pattern of bilateral upper extremity muscle strength and to predict the recovery of strength early after stroke using a logarithmic regression model.
Design: Longitudinal study.
Subjects: Twenty-one inpatients with post-stroke hemiparesis were enrolled. The mean time after stroke event was 7. 1 days (standard deviation (SD) 3. 5 days).
Methods: Bilateral elbow flexion and extension strengths were assessed separately with a hand-held dynamometer. Grip strength was also assessed with a Jamar dynamometer. These measurements were carried out on 4 occasions: baseline assessment within 2 weeks following stroke onset, and at weeks 1, 2 and 3 following baseline evaluation. Recovery of flexion and extension strengths was predicted using a logarithmic model using scores at the initial 2 evaluations.
Results: The time course of recovery for bilateral upper extremities resembled a logarithmic function. Moreover, on the basis of a logarithmic regression model, baseline measures of bilateral strengths sampled from 2 time-points during recovery could be applied to predict the pattern of recovery accurately during the subacute stroke phase (R2=0. 74–0. 95, p < 0. 0001).
Conclusion: Upper extremity muscle strength improved significantly in a similar pattern on the sides contralateral and ipsilateral to the brain lesion. Moreover, a logarithmic regression model accurately predicted both measures.