From the Department of Neurology, Hospital Universitari La Fe (A.L.,
J.T., L.L., C.V., M.B.), Valencia, and the Department of Neurology, Hospital
General (D.G.), Castellón, Spain.
Correspondence to Aida Lago, Department of Neurology, Hospital Universitari la Fe, Avda Campanar 21, 46.009 Valencia, Spain. E-mail alagom{at}meditex.es
Methods—A consecutive series of 1223 patients with
ischemic stroke was admitted at 2 reference hospitals; the time
of onset of symptoms was obtained, differentiating between onset while
asleep and awake. We compared circadian rhythm between stroke types and
between first-ever and recurrent stroke.
Results—The onset time was known in 914 patients; 25.6%
experienced onset on awakening [higher incidence in thrombotic and
lacunar stroke (28.9% and 28.4%, respectively) than in embolic stroke
(18.8%)]. For all stroke subtypes, there was a significant diurnal
variation, with a morning peak between 6 AM and noon; after
redistributing the hour of onset of patients awakening with stroke, the
morning peak was minimal in all types of stroke. There were no
differences in circadian rhythm between patients with first-ever and
recurrent stroke.
Conclusions—Only hospitalized patients were studied. There is a
circadian rhythm in all types of stroke, with higher frequency during
the day and lower frequency in the last hours in the evening. The
highest incidence in the early hours of the morning can be
overestimated, due to patients who awaken with stroke. There is no
difference in circadian rhythm between first-ever stroke and recurrent
stroke.
Some authors have noticed circadian alterations in blood pressure in
patients with stroke, with reduction of nocturnal blood pressure
decline.15 16 17 18 19 To investigate the possibility
that strokes occur more often at certain times of the day and that
different types of stroke may have different circadian variation, we
examined data from a hospital-based stroke register. If patients with
previous stroke have modified circadian patterns of blood pressure,
presumably the onset of stroke presentation would be
different; therefore, we analyzed the possible differences
between patients with first-ever stroke and patients with recurrent
stroke.
Ischemic stroke was classified as lacunar stroke, nonlacunar
stroke, and undetermined stroke; nonlacunar stroke was subdivided into
stroke with a cardiac source of embolism [presumed embolic stroke
(PES)] and stroke without a cardiac source of embolism [presumed
thrombotic stroke (PTS)]. We distinguished between patients with
first-ever stroke and patients with recurrent stroke.
The time of onset of symptoms was obtained from the patients whenever
possible and otherwise was elicited from relatives. For patients who
first noticed their stroke symptoms on awakening, it was presumed that
stroke had occurred during sleep. For patients who first noted symptoms
on awakening from sleep, the time of awakening was used initially as
the time of onset. Frequency of onset was analyzed for twelve
2-hour periods and four 6-hour intervals in a day. A second
analysis was done in which strokes with onset during sleep were
assumed to have occurred at some time during the preceding time frame.
The
In 1049 patients (85.77%), it was known if the onset was during daily
activity or sleep. Patients had stroke onset on awakening (25.6%),
lacunar stroke (28.4%), PTS (28.9%), PES (18.8%), and undetermined
stroke (17.9%). The difference was significant between PTS and PES
(P=0.002) and between lacunar and PES
(P=0.015).
The exact time of stroke onset (or awakening with stroke) was known in
914 patients (74.7%). There were little differences (Table 1
Frequency (2-hour intervals) of onset of ischemic stroke is
seen in Table 2
The stroke cases by 6-hour time period are seen in Table 3
We analyzed age, sex, hypertension, diabetes mellitus, and
ischemic cardiopathy according to the stroke onset time and the
stroke subtype; there were no significant differences in
univariate and multivariate
analyses (data not shown). Of 914 cases with accurate time of
onset, 719 had a first-ever stroke (mean age, 69.6 years; range, 15 to
97 years), and 155 had recurrent stroke (mean age, 71.7 years; range,
39 to 88 years). Forty patients in whom a history of previous stroke
was unknown were excluded. Stroke occurred on awakening in 25.9% with
first-ever stroke versus 20.2% in patients with recurrent
stroke (P=0.133, NS). The distribution by stroke subtypes
was similar: lacunar first-ever stroke (22%); lacunar recurrent stroke
(19.6%); first-ever PTS (50%), recurrent PTS (42.9%); first-ever PES
(23%); recurrent PES (30.8%); first-ever undetermined stroke (5%);
and undetermined recurrent stroke (30.8%).
There was no difference in the time of presentation between
the 2 groups (first-ever and recurrent stroke) (Table 4
Conclusions
Ischemic stroke has a circadian rhythm, with a higher frequency
during the day and a lower frequency, obvious in all the studied groups
(lacunar, thrombotic, embolic stroke), between 6:01 PM and
12:00 AM. The morning peak in stroke frequency in the 6:01
AM to 12:00 PM period could be overestimated.
In acute myocardial ischemia, the time of onset is easily
determined,20 21 corresponding with the onset of
thoracic pain and does not occur when the patient with ischemic
stroke is discovered on awakening; the onset may have occurred at any
time during sleep. Unfortunately, at present there is no marker
that indicates the time of stroke onset.
To minimize the problem, different authors have redistributed patients
with stroke on awakening in the 6 or 8 previous
hours,4 6 8 which results in a marked increase in
the interval from 6 AM to 12 noon. We can also redistribute
the patients, as in this work, in the previous time frame, ie, from
12:00 AM to 6:00 AM; in this case, there is a
considerable reduction in the morning peak stroke time.
Stroke is classified in a heterogeneous group of vascular
diseases, unlike myocardial ischemia, with different
ethiopathogenic mechanisms. Arterial pressure, as
other factors, may play an important role, favoring an increase in
morning stroke onset. But these same factors, such as lower
arterial pressure and heart rate during the night may
contribute, through a hemodynamic mechanism, to a
stroke onset during the sleeping hours, particularly in thrombotic
stroke. Our results show that thrombotic and lacunar strokes have a
higher onset during sleeping hours when compared with embolic stroke,
in agreement with previous authors.2 22
It has been suggested that circadian rhythm of arterial
blood pressure could be disturbed in patients who have suffered from
stroke; however, we do not find differences in the circadian pattern
between first-ever stroke and recurrent stroke.
Received February 18, 1998;
revision received June 25, 1998;
accepted June 25, 1998.
© 1998 American Heart Association, Inc.
Original Contributions
Circadian Variation in Acute Ischemic Stroke
A Hospital-Based Study
Abstract
Top
Abstract
Introduction
Subjects and Methods
Results
References
Background and Purpose—We
investigated circadian rhythm in ischemic stroke onset and its
subtypes, differentiating between first-ever stroke and recurrent
stroke.
Key Words: circadian rhythm • stroke onset • ischemia
Introduction
Top
Abstract
Introduction
Subjects and Methods
Results
References
The existence of a circadian rhythm in ischemic
stroke has been established, with a higher frequency in the early hours
of the morning.1 2 3 4 5 6 7 8 9 10 The cause of this diurnal
variation is uncertain. It has been related to the circadian rhythm of
fibrinolysis,11 platelet
aggregability,10 and, mainly,
arterial blood pressure, with its minimum value during
sleep and maximum value in the early hours in the morning, in both
normotensive and hypertensive patients.12
However, it has been suggested that an increase in morning stroke onset
could be due to patients awakening with neurological deficits as a
result of a stroke that could have occurred during the
night.13 14
Subjects and Methods
Top
Abstract
Introduction
Subjects and Methods
Results
References
We studied a consecutive series of 1223 patients with
ischemic stroke admitted at 2 reference hospitals and included
in the same stroke register from March 1994 to May 1996. All patients
had at least 1 CT scan performed in the first 48 hours. 
2 test was applied to the number of
observed versus expected strokes during each 6-hour interval.
Differences in demographic data and risk factors for stroke (diabetes,
hypertension, and ischemic cardiopathy) among the four 6-hour
intervals in a day were evaluated by 1-way analyses of variance
and contingency tables; multiple regression analysis was used
to assess the relation between the stroke onset time (dependent
variable) and demographic data and stroke risk factors.
P<0.05 was considered statistically significant.
Results
Top
Abstract
Introduction
Subjects and Methods
Results
References
We studied 1223 patients with ischemic stroke (696 men and
527 women) ranging in age from 15 to 97 years (median age, 72 years).
Patients had diabetes (30.5%), hypertension (58.2%), atrial
fibrillation (24.4%), and ischemic
cardiomyopathy (23.6%). There were 261 lacunar
stroke (21.3%), 897 nonlacunar stroke (73.3%), and 65 undetermined
(5.3%). In nonlacunar stroke, 303 (33.8%) were PES, and 594 (66.2%)
were PTS. 
) in age, sex, and stroke subtype
between patients with known (914 patients) and unknown (309 patients)
onset time.
View this table:
[in a new window]
Table 1. Characteristics of Patients With Known and Unknown
Stroke Onset Time . There are significant
differences according to the stroke onset time in all groups
analyzed.
View this table:
[in a new window]
Table 2. Stroke Cases by 2-h
Period , with the minimum peak at night and the
maximum peak between 6:01 AM and noon in all the groups.
When we redistributed patients with presenting stroke on awakening
in the previous time frame (patients who awake with stroke between 6.01
AM and noon are included in the previous 6 hours), there is
a significant variation in daytime, but when we compared the
first 2 intervals, the difference is nonsignificant except for embolic
stroke (P=0.049).
View this table:
[in a new window]
Table 3. Stroke Cases by 6-h Period,
% ), nor when the diagnostic
categories were compared. Table 4 shows only global data.
View this table:
[in a new window]
Table 4. First-ever and Recurrent Stroke by 6-h
Period
This study concerns only hospitalized patients, not including
those who never reach the hospital or who are not admitted; however, a
community-based study in Italy7 analyzing
circadian variation in stroke offers similar results to hospital-based
studies.
References
Top
Abstract
Introduction
Subjects and Methods
Results
References
This article has been cited by other articles:
 |
|
 |
|
  C. Foerch, H.-W. Korf, H. Steinmetz, M. Sitzer, and for the Arbeitsgruppe Schlaganfall Hessen (ASH) Abrupt Shift of the Pattern of Diurnal Variation in Stroke Onset With Daylight Saving Time Transitions Circulation, July 15, 2008; 118(3): 284 - 290. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Omama, Y Yoshida, A Ogawa, T Onoda, and A Okayama Differences in circadian variation of cerebral infarction, intracerebral haemorrhage and subarachnoid haemorrhage by situation at onset J. Neurol. Neurosurg. Psychiatry, December 1, 2006; 77(12): 1345 - 1349. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Kriszbacher, M. Koppan, J. Bodis, H.-K. Yip, S.-S. Chen, and M.-C. Chen Aspirin for Stroke Prevention Taken in the Evening? * Response: Stroke, December 1, 2004; 35(12): 2760 - 2762. [Full Text] [PDF]
|
|
|
|
 |
|
 I. Casetta, E. Granieri, E. Fallica, O. la Cecilia, E. Paolino, and R. Manfredini Patient Demographic and Clinical Features and Circadian Variation in Onset of Ischemic Stroke Arch Neurol, January 1, 2002; 59(1): 48 - 53. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Passero, F. Reale, G. Ciacci, and E. Zei Differing Temporal Patterns of Onset in Subgroups of Patients With Intracerebral Hemorrhage Stroke, July 1, 2000; 31(7): 1538 - 1544. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Bassetti and M. Aldrich Night time versus daytime transient ischaemic attack and ischaemic stroke: a prospective study of 110 patients J. Neurol. Neurosurg. Psychiatry, October 1, 1999; 67(4): 463 - 467. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||