Relationship between impairments and gait performance after stroke: a summary of rele- vant research R W Bohannon EdD, PT, NCS’, A Williams Andrews MS, PT2 ‘School of Allied Health, University of Connecticut, Storrs, and Hartford Hospital, Hartford, Connecticut; *University of North Carolina Hospitals, Chapel Hill, North Carolina, USA Summary This review was undertaken to describe the relationship between clinical measures of impairment and gait performance in patients with stroke. Twenty-four studies were analysed. Four major gait variables and five major impairments were identified. Strength of the paretic lower extremity and standing balance were found consistently to be correlated with gait performance. These variables therefore represent appropriate targets for evalua- tion and treatment in patients with stroke. Key words: Stroke, rehabilitation, muscle performance, gait, balance Gait & Posture 1995; Vol. 3: 236-240, December Introduction For professionals involved in rehabilitation, a major emphasis of patient care is the reduction of disabil- ity, that is, the functional consequences of disease. Although disability can be, and often is, treated directly, it can also be approached by way of its under- lying impairments. The basis of such an approach is the presumed relationship between impairment and disabil- ity. In spite of published statements of scepticismlJ, considerable research evidence exists that supports the relationship between impairment and disability, at least in patients with stroke. Much of the research involving patients with stroke focuses on gait, which is the func- tional activity most valued by them3.4. The purpose of this review is to summarize research in which the rela- tionships between impairments and gait performance after stroke have been described. Studies focusing on concurrent relationships are emphasized. Received: 20 April 1994 Accepted: 8 February 1995 Correspondence and reprint requests to: Richard W. Bohannon EdD PT NCS, School of Allied Health, U-101, University of Connecticut, Storrs, CT 06269, USA 0966-6362/95 $9.50 0 Elsevier Science B.V. All rights reserved Review process This review involved the examination of over 300 arti- cles relevant to stroke and gait that were published over the past 25 years. The examined articles were identified through a search of personal files and six databases (Medline, Science Citation Index, Excerpta Medica, Physiotherapy Index, Cumulative Index of Nursing and Allied Health Literature, and Gaitbib). Selected from among the articles were all clinical research studies in which: (1) impairments and gait performance were quantified in an identifiable manner, and (2) concurrent relationships between impairments and gait perfor- mance were either described or could be derived with inferential statistics. The statistics and their significance (P < 0.05) were documented. Review findings Twenty-four studies published between 1970 and 1993 were found that met the inclusion criteria5-*s. All but five enrolled less than 50 subjects. Four major clinical measurements were used in the studies to describe gait: speed (comfortable and maximum)5m19, cadence (comfortable and maximum)5.9,1*,*5,1*, endurance (distance and steps)9JOJ, and independence5~9.‘4~17X?2-28. The inde- pendent variables whose relationships with gait per-