Limitations of gait speed as an independent predictor of mortality and morbidity in cardiac patients.

Afilalo et al. (1) report that an impaired gait speed (a simple measure of frailty) can be used to identify elderly patients at high risk of major in-hospital events after cardiac surgery. They defined the primary predictor slow gait speed as the time taken to walk 5 m in more than 6 s. Gait speed is already an established marker of exercise capacity in the elderly (2), and thus the results of the study are not surprising. The overwhelming amount of previous studies including statements of consensus definitions for sarcopenia and cachexia in elderly and in chronically ill patients defines slow gait speed as a walking speed 0.8 m/s on the 4-m walk test (3,4). We thus ere surprised to see that Afilalo et al. (1) used a different efinition of slow gait speed and suggest using a 5-m walk test. For he 4-m gait speed test, a very large body of population-based data nd normal values are available (from investigations in many housands of subjects [4]). To reinvent the wheel may sometimes e a good idea, but it seems that this is not such an occasion, articularly not if we want our studies to be accepted outside of ardiology by general medicine and geriatrics; after all many of our atients are elderly. Afilalo et al. (1) suggest that a time of 6 s to alk 5 m is “normal” (i.e., 0.833 m/s), but where is the evidence f this? This small study with 131 patients cannot establish normality.” Very large population-based studies found a cutoff of .8 m/s (3). We in cardiology should use these cutoffs as well, at east until we have proved that they are not useful in our patients which seems doubtful). We suggest focusing on the use of the revious and established definition for the 4-m walk test in future rials. This would allow better comparability between previous, ngoing, and future studies in the field of frailty, sarcopenia, nd/or cachexia in patients with heart disease as well as other hronic illnesses. Regarding the survival analysis, we only want to state that ccording to their Table 1, it appears that many important arameters known to affect prognosis of such patients (including nemia, estimated glomerular filtration rate, body temperature, nd plasma levels of natriuretic peptides) were not considered. ence, we find the statement that gait speed (regardless of how it s measured) is an “incremental predictor of mortality and major orbidity” in elderly patients undergoing cardiac surgery too broad nd likely wrong in selected patients. We believe that the value of sing slow gait speed as a reliable marker for surgical or other utcomes in patients with cardiac illness still needs to be better efined.