From the Department of Preventive Medicine, Rush Medical College of Rush
University, and Rush–Presbyterian–St Luke's Medical Center,
Chicago, Ill.
Correspondence to William J. Elliott, MD, PhD, Department of Preventive Medicine, Rush–Presbyterian–St Luke's Medical Center, 1725 West Harrison St, Suite 117, Chicago, IL 60612.
Methods—A meta-analysis of 31 publications reporting the
circadian timing of 11 816 strokes was performed, subdividing (when
possible) by the type of stroke, according to the time of onset of
symptoms. When precise timing was not given, strokes were distributed
evenly (that is, biasing toward the null hypothesis of lack of
circadian variation).
Results—All subtypes of strokes displayed a significant
(P<0.001) circadian variation in time of onset, whether
divided into 3-, 4-, or 6-hour time periods. There was a 49% increase
(95% confidence interval, 44% to 55%) in stroke of all types between
6 AM and noon (compared with expectations if no circadian
variation was present), which is a 79% (95% confidence interval,
72% to 87%) increase over the normalized risk of the other 18 hours
of the day. There were 29% fewer strokes between midnight and 6
AM, a 35% decrease compared with the other 18 hours of the
day. All three subtypes of stroke had a significantly higher risk
between 6 AM and noon (55% for 8250 ischemic
strokes; 34% for 1801 hemorrhagic strokes, and 50% for 405 transient
ischemic attacks).
Conclusions—These data support the presence of a circadian
pattern in the onset of stroke, with a significantly higher risk in the
morning.
Early studies of the timing of acute stroke, however, indicated that
many afflicted patients reported awakening with new neurologic
deficits, and several reports indicated that acute strokes tended to
occur either during the evening hours or during
sleep.6 7 8 This led to the conclusion that
especially because acute therapies for stroke-in-evolution were not
particularly effective, there was little reason to consider acute
stroke as a medical emergency because the onset of symptoms was thought
to occur during sleep, when most patients would not recognize them.
More recently, data have been reported from many different countries
regarding both timing of the onset of acute stroke and the subtype of
stroke. A systematic review of these data was therefore undertaken to
consider whether there was a period during the 24 hours of the day when
stroke onset was more likely, to estimate the level of excess risk, and
to determine whether this period of increased risk was different for
various subtypes of stroke (ischemic, hemorrhagic, or transient
ischemic attack).
Because most publications did not report an hourly breakdown of the
onset of acute stroke but instead 2- to 20-hour time blocks, it was
necessary to divide the 24-hour day into standardized blocks of time,
and reallocate the observed numbers of strokes accordingly. When
precise timing was not given, or when the original report did not
include data from each of the 24 hourly divisions, the number of
strokes were distributed evenly throughout the appropriate hours. This
technique systematically biases the meta-analysis toward the
null hypothesis of a lack of circadian variation. In addition, a
"worst-case scenario" was considered, for which all of the
"untimed" strokes cited in the original reports were arbitrarily
assigned to the time period having the fewest observed strokes.
Statistical testing for the meta-analysis was performed with
traditional methodology for homogeneity and significance;
Across all original reports, there were 1222 strokes for which the time
of onset was unrecorded or uncertain. If one places all these
"untimed" strokes in the time period of lowest risk (a
"worst-case scenario"), there is still a significant circadian
variation in the risk of stroke (
It is also possible to rearrange the strokes for which the time of
onset was reported into other periods of interest, slicing the 24 hours
of the day into different time blocks. The most analogous to the
primary result given above is the 3 AM to 9 AM
to 3 PM to 9 PM quarters of the day. Allocating
the strokes reported into these time periods results in a 17% (95%
CI, 12% to 22%) increase in relative risk for all stroke between 3:01
and 9 AM and a 23% (95% CI, 18% to 29%) increase in
risk between 9:01 and 3 PM (compared with numbers expected
if there had been no circadian variation in stroke onset
(
If one reanalyzes the reported strokes according to their time
of onset during six 4-hour periods of the day, a highly significant
circadian variation is seen in the meta-analysis
(
Because there are some reports from Japan,19
especially regarding hemorrhagic stroke,20 21
which suggest that there may be differences in circadian variation of
stroke timing according to the subtype of stroke of interest,
meta-analyses of ischemic and hemorrhagic stroke
(including subarachnoid and intracerebral
bleeds), and transient ischemic attack were also carried out.
The results (when distributed into the 6-hour time periods beginning
immediately after midnight) are shown in Figure 1
As with the results of all
meta-analyses,22 these conclusions should
be interpreted cautiously. The reports that provide the raw data for
the meta-analysis were seldom population based, may have been
subject to "publication bias," and typically rely on recall of
information from patients or witnesses as to what time the stroke
symptoms actually began. The raw data are not "wake-time adjusted"
and do not identify strokes occurring among individuals who work night
or evening shifts, who have a higher blood pressure on arising but not
in the typical 6 to 8 AM time frame. There are few reports
that divide the numbers of strokes according to each of the 24 hours of
the day, and it is possible that the categorization of stroke
frequencies into the arbitrary time periods for the
meta-analysis is incorrect. It would be expected, however, that
any such inaccuracies would bias the results toward the null
hypothesis, because distributing the number of strokes evenly
throughout the period reported should increase the chances that no
significant circadian variation in stroke frequency would be noted.
Last, the results of the smaller time periods (3- or 4-hour
"slices") may underestimate the true circadian variation in stroke
onset because statistical power of the meta-analysis diminishes
as the number of time periods increases.
The finding that the early morning hours (and not the nighttime hours)
have the highest risk of the onset of stroke symptoms has two broad
implications. The first is that patients should no longer be told that
stroke symptoms are not a medical emergency. Although this may have
been sound public policy when acute treatments for stroke were not
available, there is now some evidence that acute emergent treatments
for cerebral ischemia can be delivered in a timely fashion and
result in improved long-term outcomes.23 24 The
results of this meta-analysis contradict older conclusions that
"strokes are more likely to occur during
sleep."19 This meta-analysis indicates
that irrespective of the type of stroke, most patients will be awake
when the onset of stroke symptoms occurs. The recognition of new
neurologic deficits should prompt afflicted patients and their families
to consider these as a medical emergency (or "brain attack"). The
second implication of these conclusions has to do with some modalities
useful in stroke prevention. Blood pressure is often considered one of
the most powerful risk factors for stroke and has a circadian variation
that essentially parallels the circadian variation in stroke onset.
Antihypertensive agents administered in the morning ought to have a
long duration of action to still have an effect on the early morning
rise in blood pressure. It is tempting to speculate that
antihypertensive agents that specifically target the early morning rise
in blood pressure and heart rate, without reducing blood pressure
severely during the night, might be more advantageous in controlling
the 20% rise in blood pressure during the hours around awakening. This
appears also to be the time of day associated with an increased risk of
stroke, myocardial infarction, and sudden cardiac death. At least one
long-term clinical trial is currently enrolling patients to examine
this question.25
This meta-analysis of 11 816 strokes provides strong evidence
that the onset of stroke symptoms has a circadian variation, with a
higher risk in the early morning hours (6 AM to noon), and
lower risk during the nighttime period (midnight to 6 AM).
Approximately 1 of every 8 strokes (1 of 7 ischemic strokes, 1
of 10 hemorrhagic strokes, and 1 of 8 transient ischemic
attacks) is attributable to the morning excess.
Received January 26, 1998;
revision received March 6, 1998;
accepted March 6, 1998.
© 1998 American Heart Association, Inc.
Original Contributions
Circadian Variation in the Timing of Stroke Onset
A Meta-analysis
Abstract
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Background and Purpose—Acute
myocardial infarction and sudden death display a circadian rhythm, with
a higher risk between 6 AM and noon. Some reports suggest
that stroke does not follow such a circadian variation and that
hemorrhagic stroke occurs more often during the evening.
Key Words: circadian variation • meta-analysis • stroke onset
Introduction
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Several types of
cardiovascular events, including acute myocardial
infarction and sudden cardiac death, display significant circadian
variation in the timing of onset of symptoms. A recent
meta-analysis of the 30 reports from across the world, which
included 66 635 acute myocardial infarctions, has demonstrated a 40%
excess risk between 6 AM and noon compared with the rest of
the day.1 A similar meta-analysis of 19
studies involving 19 390 sudden cardiac deaths indicated a 29%
increase in risk for this 6 AM to noon time
period.1 Some believe this morning excess of
cardiovascular risk parallels the usual circadian
pattern of physical activity, blood pressure, plasma
catecholamines, and/or plasma
cortisol.2 3 4 5
Subjects and Methods
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
A MEDLINE search of publications in all languages from 1966 to
December 1997 was performed with both text word searching and the
appropriate MESH headings of "circadian variation," and "stroke,
cerebrovascular accident, transient ischemic attack, or brain
attack." The bibliographies of each of the retrieved publications
that contained primary data about the timing of the onset of stroke
were reviewed, and relevant citations in these listings were also
evaluated for inclusion in the meta-analysis. Reported data
about timing of stroke, the hourly time periods involved, and the
subtype of stroke were abstracted and entered into a computerized
spreadsheet/database.9 When the same population
was reported more than once,10 11 12 only the most
descriptive report was included. Instead of reporting the hourly
incidence of stroke onset, several publications instead provided data
only in the form of mathematical models (for example, cosinor
analysis)13–15; this necessitated
estimating hourly stroke onset rates from the parameters
given in the reports, according to standard
procedures.14 15 16 Some
reports17 also presented the primary data
as figures, in which case, estimates of hourly stroke onset were
derived from photocopied enlargements of the published figures and
appropriate interpolation. When reports identified the onset of stroke
symptoms during sleep, data that distributed the risk across the known
hours of sleep were retained in the meta-analysis. Within each
report, data from patients whose stroke onset times were unknown (and
could not be identified as occurring during sleep) were excluded from
the meta-analysis (when it was possible to distinguish them in
the original publication), because reports about strokes without times
of onset were also excluded from the meta-analysis. 
2 techniques (for goodness of fit to the null
model of equal distribution of strokes) were used, along with 95%
confidence intervals, to evaluate the circadian pattern of stroke
onset.18 Two strategies were used to estimate the
relative risk of strokes occurring at specific time periods: One
assumed that all strokes would be evenly distributed in onset among the
24 hours of the day, and therefore compared the observed proportion
relative to the proportion expected, based on the total number
reported. The second method is based on a comparison of the observed
number of strokes compared with the average for the other hours of the
day, normalized for hours under consideration.1
In this analysis, the expected number of events that were
reported to have occurred, for instance, between 6 AM and
noon, were compared with the number of events occurring in the
remaining 18 hours of the day (divided by 3, to normalize for the
number of hours in the time period under consideration), and the
relative risk was the number of strokes actually reported, divided by
the number expected.
Results
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Thirty-one
publications6 7 8 13 14 15 17 19 20 21 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 were
retrieved that contained primary data regarding strokes with known
times of onset; these included 11 816 patients and are summarized in
the Table
. Despite some rather large
differences across studies in reported sample size (59 to 1075),
outcomes (fatal versus nonfatal), and types of stroke studied
(ischemic versus hemorrhagic versus other), most of the studies
showed a similar diurnal pattern of stroke incidence. When all strokes
are categorized according to the midnight to 6 AM to noon
to 6 PM divisions of the day, as in Cohen et
al,1 a "morning excess" of all stroke is seen
between 6 AM to noon (Figure 1, bars A), which corresponds to a 49%
increased relative risk (95% confidence interval [CI]: 44% to 55%)
compared with the number expected if there had been no circadian
variation in stroke onset (2=1133, 3
df, P<0.001), or a 79% increased relative risk
(95% CI, 72% to 87%) compared with the normalized rate for the
remaining 18 hours of the day. The 6-hour time period with the lowest
risk for all stroke is between midnight and 6 AM, which has
a 29% (95% CI, 25% to 32%) lower relative risk (compared with
expected number of strokes had there been no circadian variation) or
35% (95% CI, 32% to 38%) lower relative risk than expected if
compared with the other 18 hours of the day.
View this table:
[in a new window]
Table 1. Strokes of All Types, Divided into 6-Hour Time Blocks,
Beginning at Midnight
View larger version (50K):
[in a new window]
Figure 1. Circadian patterns of onset of symptoms of stroke,
subdivided by subtype of stroke, according to four 6-hour time periods.
Dark bars (A) show the distribution of all types of stroke (n=11 816);
diagonally-striped bars (B) correspond to ischemic stroke
(n=8250); gray bars (C) represent hemorrhagic stroke (n=1801);
and light bars (D) show the data for transient ischemic attacks
(n=405). 2=725, 3
df, P<0.001). The 6 AM to noon time
period then has a 35% (95% CI, 30% to 41%) increased risk compared
with the expected number of strokes had there been no circadian
variation, or 54% (95% CI, 48% to 60%) increased risk if compared
with the normalized rate for the other 18 hours of the day. 2=599, 3 df, P<0.001).
If the normalized rate for the remaining 18 hours of the day is used as
the comparator, there is a 24% (95% CI, 19% to 30%) increased
relative risk between 3:01 and 9 AM and a 34% (95% CI,
28% to 40%) increased relative risk of stroke in the next 6-hour
period. The time period between 9:01 PM and 3
AM has the lowest number of strokes (33% [95% CI, 30%
to 37%] reduced relative risk compared with expectations if there had
been no circadian variation and 40% [95% CI, 38% to 43%] reduced
relative risk compared with the other 18 hours of the day). 2=879, 5 df, P<0.001).
The highest risk is found between 8:01 AM and noon (a 45%
[95% CI, 38% to 52%] increase compared with what would have been
expected if there were no circadian variation in stroke onset and a
59% [95% CI, 51% to 68%] increase compared with the normalized
rate for the remaining 20 hours of the day); the lowest is found
between midnight and 4 AM (35% [95% CI, 31% to 40%]
or 40% [95% CI, 35% to 45%] reduced relative risk, respectively).
Finally, if the reported strokes are redistributed according to their
times of onset during the eight 3-hour periods of the day, there is
still a significant result in the meta-analysis, rejecting the
null hypothesis of no circadian variation in time of stroke onset
(2=1202, 7 df,
P<0.001). The time period of highest risk is found between
6:01 and 9 AM (58% [95% CI, 48% to 67%] increase
compared with the expected value if all strokes had been evenly
distributed and a 72% [95% CI, 62% to 83%] increase compared with
the value expected for the other 21 hours in the day), with the time
between 9:01 and noon following close behind (42% and 51% [95% CI,
41% to 61%] increase in relative risk, respectively). The time
period with the lowest number of events is from midnight to 3
AM (36% [95% CI, 31 to 40%] and 44% [95% CI, 40%
to 48%] decrease in relative risk, respectively). 
. These data suggest
that for each subtype of stroke studied, there is an increase in risk
during the early morning hours. There were 21 studies (including 8250
patients) of ischemic stroke, which was 55% (95% CI, 48% to
62%) more likely between 6 AM and noon. The 13 studies of
1801 patients with hemorrhagic stroke also showed a significant
circadian variation (Figure 1B); the risk was 34% (95% CI, 21% to
49%) greater between 6 AM and noon. The "morning
excess" is even statistically significant for transient
ischemic attack (TIA), in which only two studies form the basis
for a meta-analysis of 405 patients, but a
2 value of 95 (3 df,
P<0.001) is obtained, indicating a 50% (95% CI, 20% to
85%) increased risk of TIA between 6:01 AM and noon and an
even more impressive 76% (95% CI, 64% to 84%) decreased risk of TIA
between midnight and 6 AM (all compared with the random
distribution of TIAs across the 24 hours of the day). The
pertinent percentages for these time periods (with the 18 remaining
hours of the day as comparator) are an 80% (95% CI, 43% to 126%)
increase, or an 81% (95% CI, 72% to 87%) decrease, in risk for 6
AM to noon or midnight to 6 AM, respectively. A
statistically significant circadian pattern of timing of stroke onset
was seen for all subtypes of strokes, even when the 24 hours of the day
were divided into 4-hour (Figure 2) or
3-hour periods (data not shown).
View larger version (57K):
[in a new window]
Figure 2. Circadian patterns of onset of symptoms of stroke,
subdivided by subtype of stroke, according to six 4-hour time periods.
Dark bars (A) show the distribution of all types of stroke (n=11 816);
diagonally-striped bars (B) correspond to ischemic stroke
(n=8250); gray bars (C) represent hemorrhagic stroke (n=1801);
and light bars (D) show the data for transient ischemic attacks
(n=405).
Discussion
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
These data from a meta-analysis of the world's 31
published studies on the circadian timing of stroke onset indicate that
like acute myocardial infarction and sudden cardiac death, there is an
increased risk of the onset of acute stroke during the early morning
hours. The data are remarkably consistent across the various
subtypes of stroke, and indicate, for ischemic stroke,
hemorrhagic stroke, and even transient ischemic attacks, that
the excess risk during the 6 AM to noon time period is
significantly higher than would be expected by chance: 89%, 52%, and
80% (95% CI, 89% to 99%, 36% to 69%, and 43% to 126%,
respectively, compared with the normalized risk for the other 18 hours
of the day). Similarly, there is a significantly lower risk of stroke
during the nighttime hours (midnight to 6 AM) for each
stroke subtype: 30%, 54%, and 81% (95% CI, 26% to 33%, 48% to
60%, 72% to 87%, respectively, compared with the normalized risk for
the other 18 hours of the day). Even when all the strokes in the
original publications for which a time of onset could not be accurately
determined are incorporated into the meta-analyses in a
"worst-case scenario" in each division of the hours of the day,
there is a significant circadian variation in the timing of stroke
onset.
Footnotes
Presented in part at the 12th Annual Meeting of the American Society of Hypertension, San Francisco, Calif, May 30, 1997.
References
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
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]
|
|
|
|
 |
|
 R. Manfredini, B. Boari, and R. Salmi Higher Stroke Mortality on Weekends: Are All Strokes the Same? Stroke, October 1, 2007; 38(10): e112 - e112. [Full Text] [PDF]
|
|
|
|
 |
|
 T. A. Martino, N. Tata, G. A. Bjarnason, M. Straume, and M. J. Sole Diurnal protein expression in blood revealed by high throughput mass spectrometry proteomics and implications for translational medicine and body time of day Am J Physiol Regulatory Integrative Comp Physiol, September 1, 2007; 293(3): R1430 - R1437. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Authors/Task Force Members:, G. Mancia, G. De Backer, A. Dominiczak, R. Cifkova, R. Fagard, G. Germano, G. Grassi, A. M. Heagerty, S. E. Kjeldsen, et al. 2007 Guidelines for the Management of Arterial Hypertension: The Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC) Eur. Heart J., June 11, 2007; (2007) ehm236v1. [Full Text] [PDF]
|
|
|
|
 |
|
 T. A. Martino, N. Tata, D. D. Belsham, J. Chalmers, M. Straume, P. Lee, H. Pribiag, N. Khaper, P. P. Liu, F. Dawood, et al. Disturbed Diurnal Rhythm Alters Gene Expression and Exacerbates Cardiovascular Disease With Rescue by Resynchronization Hypertension, May 1, 2007; 49(5): 1104 - 1113. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J-C Chen Death stroked by dusty air: more mysteries to be solved Occup. Environ. Med., January 1, 2007; 64(1): 3 - 4. [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]
|
|
|
|
 |
|
 K Madin and P Iqbal Twenty four hour ambulatory blood pressure monitoring: a new tool for determining cardiovascular prognosis. Postgrad. Med. J., September 1, 2006; 82(971): 548 - 551. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. B. Klerman Clinical Aspects of Human Circadian Rhythms J Biol Rhythms, August 1, 2005; 20(4): 375 - 386. [Abstract] [PDF]
|
|
|
|
|
|
N U Weir and A M Buchan A study of the workload and effectiveness of a comprehensive acute stroke service J. Neurol. Neurosurg. Psychiatry, June 1, 2005; 76(6): 863 - 865. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Sista, J. C.-K. Lai, O. Eradiri, and K. S. Albert Pharmacokinetics of a Novel Diltiazem HCl Extended-Release Tablet Formulation for Evening Administration J. Clin. Pharmacol., October 1, 2003; 43(10): 1149 - 1157. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. R. Black, W. J. Elliott, G. Grandits, P. Grambsch, T. Lucente, W. B. White, J. D. Neaton, R. H. Grimm Jr, L. Hansson, Y. Lacourciere, et al. Principal Results of the Controlled Onset Verapamil Investigation of Cardiovascular End Points (CONVINCE) Trial JAMA, April 23, 2003; 289(16): 2073 - 2082. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Bernatova, M. P. Key, J. B. Lucot, and M. Morris Circadian Differences in Stress-Induced Pressor Reactivity in Mice Hypertension, November 1, 2002; 40(5): 768 - 773. [Abstract] [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]
|
|
|
|
|
|
G. Y.H. Lip, E. K.H. Tan, C. K.Y. Lau, and S. Kamath Diurnal Variation in Stroke Onset in Atrial Fibrillation Stroke, June 1, 2001; 32(6): 1443 - 1448. [Full Text] [PDF]
|
|
|
|
|
|
V. L. Feigin, C. S. Anderson, N. E. Anderson, J. B. Broad, M. J. Pledger, and R. Bonita Is There a Temporal Pattern in the Occurrence of Subarachnoid Hemorrhage in the Southern Hemisphere? : Pooled Data From 3 Large, Population-Based Incidence Studies in Australasia, 1981 to 1997 Stroke, March 1, 2001; 32(3): 613 - 619. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P.A. Nyquist, R.D. Brown Jr., D.O. Wiebers, C.S. Crowson, and W.M. O'Fallon Circadian and seasonal occurrence of subarachnoid and intracerebral hemorrhage Neurology, January 23, 2001; 56(2): 190 - 193. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Mancia and G. Parati Ambulatory Blood Pressure Monitoring and Organ Damage Hypertension, November 1, 2000; 36(5): 894 - 900. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Anonymous Epidemiology of Aneurysmal Subarachnoid Hemorrhage in Australia and New Zealand : Incidence and Case Fatality From the Australasian Cooperative Research on Subarachnoid Hemorrhage Study (ACROSS) Stroke, August 1, 2000; 31(8): 1843 - 1850. [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]
|
|
|
|
|
|
H.-R Arntz, S.N Willich, C Schreiber, T Bruggemann, R Stern, and H.-P Schultheiss Diurnal, weekly and seasonal variation of sudden death. Population-based analysis of 24061 consecutive cases Eur. Heart J., February 2, 2000; 21(4): 315 - 320. [Abstract] [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]
|
|
|
|
|
|
S. Chaturvedi, H. P. Adams Jr, and R. F. Woolson Circadian Variation in Ischemic Stroke Subtypes Stroke, September 1, 1999; 30(9): 1792 - 1795. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||