(Stroke. 1997;28:1913-1918.)
© 1997 American Heart Association, Inc.
Articles |
From the Department of Neurology, Austin and Repatriation Medical Centre, University of Melbourne (R.X.Y., H.M. O'M., G.A.D.); Department of Epidemiology and Preventive Medicine, Monash University (J.J.M., A.G.T.); and Department of Neurology, Royal Melbourne Hospital, University of Melbourne (S.M.D.) (Australia). This multicenter study was coordinated by the Department of Neurology, Austin and Repatriation Medical Centre.
| Abstract |
|---|
|
|
|---|
Methods A total of 201 consecutive patients with first-onset stroke due to cerebral infarction aged 15 to 55 years (mean, 45.5 years) were accrued from four teaching hospitals during 1985 to 1992 and compared with their age- and sex-matched neighborhood controls. Information concerning potential risk factor exposure status was collected by structured questionnaire at interview. Stroke risks were estimated by calculating the odds ratios with multivariate logistic regression.
Results Significantly increased risk of stroke was found
among those with diabetes (odds ratio, 11.6 [95% confidence
intervals, 1.2 to 115.2]), hypertension (6.8 [3.3 to 13.9]), heart
disease (2.7 [1.1 to 6.4]), current cigarette smoking (2.5 [1.3,
5.0]), and long-term heavy alcohol consumption (
60 g/d) (15.3 [1.0
to 232.0]). However, heavy alcohol ingestion (
60 g) within 24 hours
preceding stroke onset was not a risk factor (0.9 [0.3 to 3.4]).
Conclusions Diabetes, hypertension, heart disease, current smoking, and long-term heavy alcohol consumption are major risk factors for stroke in young adults. Given that the majority of these factors are either correctable or modifiable, prevention strategies may have the potential to reduce the impact of stroke in this age group.
Key Words: cerebral infarction risk factors young adults
| Introduction |
|---|
|
|
|---|
| Subjects and Methods |
|---|
|
|
|---|
Control Subjects
Community-based control subjects individually matched with case
subjects by the same sex and age (within 5 years) were identified by
using "neighborhood" selection techniques, as described
before.16 17 Here a standardized method of identifying a
matched control subject was used by starting in the house on the
immediate right of that of the subject in the same street. Sequential
houses to the right were then visited until a matched control subject
was found. To eliminate bias due to omitting people at work or absent
for other reasons, all people in each house were identified and, if the
house was empty, at least three return visits were made outside working
hours. People with a history of documented stroke were excluded.
Information Ascertainment
Information about potential risk factor exposure status of
eligible case and control subjects was collected by a constructed
questionnaire at interview, as previously
described.16 17
Definitions
Definitions were as follows: Stroke due to cerebral
infarction: acute onset of a focal neurological deficit lasting
longer than 24 hours, in which CT scan excludes causes other than
cerebral infarction. Those with deficits lasting less than 24 hours but
CT or MRI evidence of infarction were not included. Lacunar
syndrome: acute onset of one of the five recognized lacunar
syndromes (pure motor hemiplegia, ataxic hemiparesis,
dysarthriaclumsy hand syndrome, sensorimotor stroke, and pure sensory
stroke) in which CT had excluded underlying cerebral
hemorrhage. In many cases the site of infarction was identified
on CT scan, but this was not an absolute requirement for classification
as a lacunar syndrome. Thromboembolic infarction: acute
onset of focal neurological deficit with documentation of the site of
infarction on CT scan in either cerebral hemisphere or hind brain, in
which the mechanism of infarction was attributed to large-vessel
extracranial or intracranial vascular disease. Cardiac embolic
cerebral infarction: acute onset of a focal neurological deficit
in which the site of infarction had been documented on CT scan in the
presence of atrial fibrillation, myocardial infarction within the
preceding 3 weeks, or cardiomyopathy. In some cases
cerebral angiography or noninvasive studies of the extracranial
circulation were done to help exclude carotid occlusive disease as a
casual mechanism, but this was not an absolute requirement.
Cerebral infarct site or mechanism uncertain: acute onset of
a focal neurological deficit in which the site of infarction or the
mechanism of its genesis was unclear, but nonvascular causes were
excluded by the CT scan. Current smoker: a person who smoked
at least one cigarette per day for the preceding 3 months or more.
Ex-smoker: a person who smoked at least one cigarette per
day for 3 months or more at some period during his/her life but has not
smoked for the preceding 3 months or more. Never smoker: a
person who does not meet the criteria for a current smoker or
ex-smoker. Ever-drinker/never drinker: a subject who has
ever/never drunk alcohol in his/her lifetime. Hypertension, Heart
disease, Diabetes mellitus, and High cholesterol:
history of such advised by a medical practitioner.
Use of oral contraceptives: past and/or current use of birth
control pills either for contraception or other indications.
Physical exercise: current engagement in active physical
exercise that caused perspiration and breathlessness, such as brisk
walking, running, swimming, cycling, squash, and vigorous team sports.
According to the exercise frequency, this variable was reduced to
three subcategories of "never/rarely" (reference category), "1
to 2 times per week," and "
3 times per week." Mechanisms of
stroke were as defined previously.16
Statistical Analysis
A conditional logistic regression model (EGRET19
statistical package) was used to analyze the case-control
pair-matched data set. Odds ratios of stroke for risk factors under
study were estimated with adjustments for potentially confounding
variables by multivariate analysis of the
regression model. The significance level was set at P<.05.
The two-sample t test was used to compare mean ages between
the two groups.
| Results |
|---|
|
|
|---|
|
The frequency distribution of presumed stroke mechanisms is shown in
Table 2
. While the majority (52%) of the
cases were presumably thromboembolic, the mechanism and site of
cerebral infarction in 14% of cases remained uncertain.
|
Comparison of the prevalence rates of and odds ratios for the main risk factors in case subjects recruited during 1985 to 1988 and 1988 to 1992 did not show statistically significant differences (current smoking, 63%, 54%; hypertension, 45%, 50%; heart disease, 20%, 23%, respectively; all P>.2 for ratio of their two odds ratios). This provides evidence to suggest that the case subjects recruited in the two periods in the same region were more likely to be homogeneous in terms of risk factor exposure and can therefore be considered as a single group for analysis.
The estimates of stroke risk for the examined risk factors are show in
Table 3
. A significantly increased risk
of stroke was associated with diabetes mellitus (odds ratio, 11.6
[95% confidence interval, 1.2 to 115.2]); hypertension (6.8 [3.3 to
13.9]); heart disease (2.7 [1.1 to 6.4]); current smoking (2.5 [1.3
to 5.0]); and long-term heavy drinking (
60 g alcohol per day) (15.3
[1.0 to 232.0]). However, recent heavy drinking (
60 g alcohol
within the preceding 24 hours) was not a significant risk factor (0.9
[0.3 to 3.4]). The remaining factors also did not reach significance.
These odds ratios and confidence intervals largely varied by no more
than 10% when heart disease was excluded from the
multivariate analysis. The exceptions to this
were current smoking, diabetes, and ever use of the oral contraceptive
pill, in which cases the multivariate odds ratios and
95% confidence intervals without adjustment for heart disease were 2.9
(1.4 to 5.6), 13.6 (1.4 to 128.2), and 0.7 (0.2 to 2.0), respectively.
No effects of interactions between the examined risk factors on the
risk of stroke were found.
|
| Discussion |
|---|
|
|
|---|
Hypertension is widely considered as a major contributor to stroke in
the general population. However, hypertension-related risk of stroke in
the young has been infrequently assessed with adjustment for
confounding variables,20 although the prevalence of
hypertension in stroke patients at younger ages has been occasionally
reported.4 14 21 22 23 24 25 Although the age-specific prevalence
rates of hypertension (22%, 26%, and 33% in the age groups of 15 to
25, 15 to 35, and 15 to 45 years, respectively) in the case subjects of
the present study were close to or slightly higher than those in
other hospital-based case-control studies,4 14 23 24 25 26 27 28 29 the
overall prevalence rates of hypertension were 49% in the case subjects
and 16% in the control subjects, indicating a 6.8-fold increased risk
of stroke for young hypertensives (Table 4
). In the interpretation of this
finding, the possibility of bias, confounding effects, and chance in
producing the result must be considered. First, we designed this study
to try to minimize potential biases in each stage of subject selection,
data collection, and analysis, although the possibility that
the hospital-based nature of the study may have selectively attracted
hypertensive subjects cannot be excluded. Second, the potentially
confounding effects of sex, age, and socioeconomic and regional factors
have been controlled for by selecting sex- and age-matched neighbors as
control subjects. Other confounding variables including smoking,
diabetes, and heart disease have been adjusted for by
multivariate regression. Finally, since the Australian
community prevalence rate of hypertension is approximately 14% among
people aged 20 to 59 years according to a national survey performed in
1989 by Australian Heart Foundation (AHF),30 the rate of
19% among the control subjects aged 15 to 60 years was unlikely to be
unusually low to result in an overestimate of hypertension-related risk
of stroke in this case-control study. Therefore, it appears that the
observed high risk of cerebral infarction associated with hypertension
in the young group most likely reflects a true association between
hypertension and stroke due to cerebral infarction in young adults, at
least in this study population. Clearly, this association needs to be
examined further.
|
Similarly, the role of diabetes mellitus in the development of stroke has been well documented in general populations, in particular the elderly,31 32 33 34 35 36 37 38 39 but not yet in young adults. Most studies have only described a prevalence rate of diabetes mellitus as a possible cause or risk factor without estimating the relative risk for diabetes per se.4 14 23 24 25 While some authors reported that the prevalence of diabetes in patients with ischemic stroke under 50 years was not significantly different from that of patients with infectious diseases at similar age,40 others estimated the relative risk of stroke for diabetes.31 41 42 In the present study 15 case subjects (7%) and only one (0.5%) control subject had a history of diabetes, presenting a more than 10-fold increased risk of stroke related to diabetes. However, the interpretation of this high risk estimate with a wide 95% confidence interval is difficult because the frequency of diabetes among control subjects was relatively low, probably by chance, compared with those community prevalence rates observed in the AHF survey (1.2% in the group aged 20 to 59 years),30 a Swedish study (1.5% for those aged 50 years),43 and the Honolulu Heart Program (9.2% among those aged 45 to 59 years).34
The estimate of a 2.7-fold increased risk of stroke due to cerebral infarction associated with heart disease in this study was in agreement with those found in general populations of previous studies.37 38 44 45 Twenty-one percent of cases in the current study were considered to have a mechanism of embolism from the heart, a finding similar to that of Adams et al,14 who found a cardiac source of embolism in 23% of cases. Since the risk of stroke may vary with different types of heart diseases,20 45 a more refined definition is critically important in further study of cardiac risk factors.
The health effect of cigarette smoking in young people is an important public issue because of the increasing prevalence of smoking in this age group.46 47 48 The AHF national survey found current smokers in one of four Australians (24.5%) aged 20 to 59 years.30 Compared with the community prevalence, a slightly higher rate in the control subjects (33%) and a much higher rate in the case subjects (56%) were observed in this study and indicated a 2.6-fold increased risk of cerebral infarction for the young current smokers. If this estimate is correct, the attributable risk would be considerable because of the high prevalence of young smokers in the community. At least from the perspective of stroke risk, education concerning the adverse effects of cigarette smoking is appropriate for the young and is particularly important because of the reported increase in smoking prevalence in the young in some societies.49
Alcohol consumption as a risk factor for stroke in young adults has
been observed in some studies.50 51 52 53 54 55 56 57 58 59 We found that alcohol
consumption was a risk factor when long-term heavy ingestion of 60
g or more alcohol per day was considered. However, recent intake of
large amounts (
60 g alcohol within 24 hours preceding stroke onset,
which also included no habitual drinkers) was not a risk factor. This
pattern is somewhat different from that found by Hillbom et
al,57 who showed that ingestion of alcohol more than
40 g within 24 hours preceding disease onset was an independent
risk factor in their population. When we considered alcohol drinking as
a dichotomous variable, no increase in stroke risk was found.
Our study has many of the limitations associated with retrospective case-control methodology. As mentioned earlier, many of the potential biases of case selection and influence of confounding variables have been minimized by study design, particularly by using the neighborhood control technique. Here past access to medical care of case subjects compared with control subjects is likely to be similar, so that definitions concerning past illnesses are more likely to be standardized. Even so, the measurement of past risk factor exposure is always particularly difficult and undoubtedly results in an underestimation of the prevalence in both case and control subjects. The use of oral contraceptive use is a case in point in which past use was probably underestimated and current usage was low. Here, the combined "ever used" grouping was used, which is less clinically useful. In view of recent findings that low-dose oral contraceptives probably do not increase stroke risk,28 29 it would be of interest to develop an expanded sample size to test this hypothesis in our population.
There are four directly comparable case-control studies of stroke in the young. Hillbom et al52 studied 75 consecutive subjects aged 16 to 40 years with first-ever ischemic brain infarction and compared these with 133 hospital-based control subjects. They found that recent alcohol ingestion, arterial hypertension, cardiac disease, diabetes, and migraine were significant risk factors, whereas smoking was not. The lack of association with smoking in their study may be in part due to use of hospital-based control subjects, in which a bias toward smoking in the control group may exist.60 Haapaniemi et al59 focused on weekend and holiday timing of stroke onset as a risk factor for ischemic stroke in young women. Among all cases (male and female) 141 subjects were studied, but a control group was not used. In our group, the timing of onset of stroke was not analyzed. Marini et al26 studied 333 subjects with stroke aged 15 to 44 years and used both hospital- and community-based (railway workers) controls. Like us, they found the vascular risk factors of hypertension, diabetes, and smoking to be associated with stroke, as did Rohr et al,27 who found these risk factors to be particularly important among blacks.
In summary, in the population studied here, the risk factors for ischemic stroke in those younger than 55 years were hypertension, diabetes, heart disease, and current smoking, all factors that are likely to promote premature atherosclerosis. Screening for atherosclerosis in this high-risk group may therefore be reasonable. Although this group of young patients with stroke represents only approximately 10% of the total, long-term disability in this group is of greater consequence than in older subjects. Quantitation of risk factors is therefore particularly important, and those identified are either modifiable or preventable. These findings need to be further validated in community-based studies.
| Acknowledgments |
|---|
| Footnotes |
|---|
Received January 22, 1997; revision received July 15, 1997; accepted July 15, 1997.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
V. M. Bhat, J. W. Cole, J. D. Sorkin, M. A. Wozniak, A. M. Malarcher, W. H. Giles, B. J. Stern, and S. J. Kittner Dose-Response Relationship Between Cigarette Smoking and Risk of Ischemic Stroke in Young Women Stroke, September 1, 2008; 39(9): 2439 - 2443. [Abstract] [Full Text] [PDF] |
||||
![]() |
K Lipska, P N Sylaja, P S Sarma, K R Thankappan, V R Kutty, R S Vasan, and K Radhakrishnan Risk factors for acute ischaemic stroke in young adults in South India J. Neurol. Neurosurg. Psychiatry, September 1, 2007; 78(9): 959 - 963. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Wendel-Vos, A. Schuit, E. Feskens, H. Boshuizen, W. Verschuren, W. Saris, and D Kromhout Physical activity and stroke. A meta-analysis of observational data Int. J. Epidemiol., August 1, 2004; 33(4): 787 - 798. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. L. Nightingale and R. D.T. Farmer Ischemic Stroke in Young Women: A Nested Case-Control Study Using the UK General Practice Research Database Stroke, July 1, 2004; 35(7): 1574 - 1578. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. D. Lee, A. R. Folsom, and S. N. Blair Physical Activity and Stroke Risk: A Meta-Analysis Stroke, October 1, 2003; 34(10): 2475 - 2481. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. Novak, A. Chowdhary, B. Farrar, H. Nagaraja, J. Braun, R. Kanard, P. Novak, and A. Slivka Altered cerebral vasoregulation in hypertension and stroke Neurology, May 27, 2003; 60(10): 1657 - 1663. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. A. Beckman, M. A. Creager, and P. Libby Diabetes and Atherosclerosis: Epidemiology, Pathophysiology, and Management JAMA, May 15, 2002; 287(19): 2570 - 2581. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Hajat, R. Dundas, J. A. Stewart, E. Lawrence, A. G. Rudd, R. Howard, and C. D. A. Wolfe Cerebrovascular Risk Factors and Stroke Subtypes : Differences Between Ethnic Groups Stroke, January 1, 2001; 32(1): 37 - 42. [Abstract] [Full Text] [PDF] |
||||
![]() |
J.A. Papadakis, E.S. Ganotakis, and D.P. Mikhailidis Beneficial effect of moderate alcohol consumption on vascular disease: myth or reality? The Journal of the Royal Society for the Promotion of Health, March 1, 2000; 120(1): 11 - 15. [Abstract] [PDF] |
||||
![]() |
M. Hillbom, H. Numminen, and S. Juvela Recent Heavy Drinking of Alcohol and Embolic Stroke Stroke, November 1, 1999; 30 (11): 2307 - 2312. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. Guerrero-Romero and M. Rodriguez-Moran Proteinuria Is an Independent Risk Factor for Ischemic Stroke in Non–Insulin-Dependent Diabetes Mellitus Stroke, September 1, 1999; 30(9): 1787 - 1791. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. R. Evenson, W. D. Rosamond, J. Cai, J. F. Toole, R. G. Hutchinson, E. Shahar, and A. R. Folsom Physical Activity and Ischemic Stroke Risk : The Atherosclerosis Risk in Communities Study Stroke, July 1, 1999; 30(7): 1333 - 1339. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Seppa and P. Sillanaukee Binge Drinking and Ambulatory Blood Pressure Hypertension, January 1, 1999; 33(1): 79 - 82. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1997 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |