Muscles are usually stimulated by shocks delivered at some constant rate. H
owever, human thenar motor units generate optimum force per pulse when exci
ted by impulse trains that begin with one or two short interpulse intervals
("doublets"), followed by longer intervals. Our aim was to determine wheth
er the rate of force and force-time integral reduction during fatigue of th
enar muscles is influenced by an initial doublet, and/or the number of puls
es per train. We first matched thenar force-time integral using two differe
nt pulse patterns, one of which began with a doublet. Fatigue induced by tr
ains that contained a doublet resulted in slower rates of force and force-t
ime integral reduction and smaller increases in half-relaxation time than t
hat evoked by bursts of 40-Hz stimulation. When the force was measured in e
ach protocol after equal numbers of pulses had been delivered, the force lo
ss was still significantly less for pulse trains containing a doublet, Thes
e results have useful implications when designing stimulation to strengthen
weak muscles or to drive paralyzed muscles. (C) 2000 John Wiley & Sons, In
c.