A series of stepwise forward multiple regressions was com-puted to explain the variance of each SF-36 scale by the inde-pendent variables NYHA functional class, peak oxygen uptake,
six minute walk test, left ventricular ejection fraction,
duration of disease, and age. Variability in short form scale
physical functioning was explained by NYHA functional class
(40%) and to a much lower degree by the six minute walk test
(11%). In all other SF-36 scales, only between 10% and 28% of
the variance could be explained by NYHA functional class,
while none of the other variables contributed significantly
(table 3).