In normal subjects, the level and variability of blood pressure decrea
se during non-rapid eye movement (non-REM) sleep. In contrast, sleep a
pnea is associated with large swings in nocturnal pressure. In this st
udy, we evaluated a computer-derived index of all-night blood pressure
variability in normotensive snorers with or without sleep apnea. We a
lso examined this index in snorers receiving medical treatment for coe
xistent ischemic heart disease. Beat-to-beat blood pressure was record
ed with a photoplethysmographic device (Finapres) throughout polysomno
graphy. Subjects were categorized into four groups: those without card
iovascular disease without or with sleep apnea (greater than or equal
to 15 apnea plus hypopnea per hour of sleep), and those with ischemic
heart disease without or with sleep apnea. A frequency distribution hi
stogram of all increases and decreases of blood pressure according to
their amplitudes was drawn and the SD of the distribution used as an e
stimation of variability. Mean systolic and diastolic pressures during
the total sleep time were not different among the four groups. In con
trast, the SD of the distribution of systolic and diastolic pres sure
variations that were higher in the apneic than in the nonapneic groups
(P < .05) correlated with apnea plus hypopnea (P < .0001) and transie
nt electroencephalographic arousal number per hour of sleep (P < .0001
). In both apneic and nonapneic subjects, blood pressure variability a
s assessed by SD decreased during stages 3 and 4 of non-REM sleep comp
ared with stages 1 and 2 and REM sleep (P < .001). Blood pressure vari
ability was similarly increased in apneic subjects with or without isc
hemic heart disease. We speculate that in apneic individuals with coex
istent ischemic heart disease, pressure variability that is increased
despite treatment with P-blockers or calcium antagonists may be a risk
factor for acute coronary events.