We tested the hypothesis that the cardiac-related rhythm in sympathetic ner
ve discharge (SND) results from the forcing of a central oscillator to the
frequency of the heart beat by pulse-synchronous baroreceptor afferent nerv
e activity. For this purpose, time series analysis was used to examine the
phase relations between the brachial arterial pulse (AP) and cardiac-relate
d activity recorded from the postganglionic inferior cardiac sympathetic ne
rve (CN) in urethan-anesthetized cats. Specifically, we made cycle-by-cycle
measurements of peak systolic blood pressure, heart period, CN burst ampli
tude, and the phase angle (and corresponding interval) between peak systole
and the next peak of CN activity. As the steady-state level of systolic bl
ood pressure was raised by increasing the rate of a constant intravenous in
fusion of phenylephrine, we observed transitions from no phase-locking of C
N activity to the AP to either phase-locking of variable strength or phase
walk through part of the cardiac-cycle on the time scale of respiration. Ph
ase walk is defined as a progressive and systematic change in the phase lag
of cardiac-related CN activity relative to peak systole. Raising blood pre
ssure strengthened phase-locking and either increased or decreased the mean
interval between peak systole and the next peak of CN activity even when t
he change in heart period was small. CN burst amplitude and the interval be
tween peak systole and the next peak of CN activity were inversely related,
but the strength of the relationship varied considerably with experimental
conditions. The relationship was strongest during phase walk. Step-wise in
creases in blood pressure induced by abdominal aortic obstruction led to an
abrupt increase in the phase lag of CN activity relative to peak systole e
ven when heart rate was not changed. We refer to such changes as sharp phas
e transitions that are a general property of dynamical nonlinear systems. T
he results support the view that the cardiac-related rhythm in SND is a for
ced nonlinear oscillation rather than the consequence of periodic inhibitio
n of randomly generated activity.