The maximum force a muscle can produce depends on its cross-sectional area(CSA). However, the exact interpretation of this relationship has been a matter of controversy. Recently, the controversy has centered on whether the measurements are best correlated using regression analysis or ratio standards. Applying regression analysis to this problem implies that all the experimental error is in the measurement of force. Thus, confusion may arise by failure to take account of errors in the measurement of CSA. Using a statistical model, we show how regression analysis can be misleading as error is introduced into the measurement of CSA as well as that of force. Because neither the errors in force nor CSA can be quantified in the experimental situation, we conclude that ratio standards are less likely to mislead although the accuracy of the result depends on the degree of correlation between force and CSA in the muscle measured.
Institute of Human Performance, University College London, Royal National Orthopedic Hospital, Stanmore, Middlesex HA7 4LP, UK and Department of Medicine for the Elderly, Conquest Hospital, St. Leonards-on-Sea, East Sussex TN37 7RD, UK
Submitted for publication April 1996.
Accepted for publication November 1996.
The work was supported by South-East and North-West Thames Regional Health Authorities, Research into Aging and Action Research.
Address for correspondence: Dr Stuart Bruce, Institute of Human Performance, University College London, Royal National Orthopedic Hospital, Brockley Hill, Stanmore, Middlesex HA7 4LP, UK.