Published online before print December 4, 2003, doi: 10.1161/01.STR.0000106914.60740.78
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2004;35:7.)
© 2004 American Heart Association, Inc.
Original Contributions |
Risk Factors for First-Ever Stroke in Older People in the North East of England
A Population-Based Study
From the School of Population and Health Sciences (Epidemiology & Public Health) (H.R., J.G., T.D., R.W., N.S., R.T.) and School of Clinical Medical Sciences (H.R., T.D.), University of Newcastle upon Tyne, and Northumbria Healthcare NHS Trust (H.R., T.D.), North Shields, Tyne and Wear, UK.
Correspondence to Dr Helen Rodgers, Reader in Stroke Medicine, School of Population and Health Sciences (Epidemiology & Public Health), The Medical School, Newcastle upon Tyne, NE2 4HH, UK. E-mail helen.rodgers{at}newcastle.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Background and Purpose— Although older people potentially have most to gain from prevention, they have been excluded from or underrepresented in many stroke incidence studies. We sought to determine the risk factors for stroke in older people.
Methods— A 5-year follow-up study of a population-based cohort of 4440 subjects aged >65 years in northern England. Subjects were recruited from 27 general practices between 1995 and 1997.
Results— A total of 329 out of 4351 subjects with follow-up data suffered a first-ever stroke. On multivariate analysis, risk factors for stroke in older people included atrial fibrillation (hazard ratio [HR], 2.03 [95%CI, 1.31 to 3.16]); previous transient ischemic attack (1.87 [95% CI, 1.27 to 2.76]); smoking (1.72 [95% CI ,1.28 to 2.32]); and cardiovascular disease (1.55 [95% CI, 1.19 to 2.03]). The HR per 10-mm Hg increase in systolic blood pressure was 1.15 (95% CI, 1.06 to 1.24). Age was associated with a HR of 1.74 (95% CI, 1.42 to 2.12) per 10-year increase.
Conclusions— "Classic risk factors" increase the risk of stroke in older people. Stroke is not an inevitable consequence of aging, so by identifying and modifying risk factors in older people there are opportunities to reduce the incidence and mortality of this devastating condition.
Key Words: elderly • epidemiology • risk factors • stroke
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The incidence of stroke rises markedly with age, with 75% occurring over the age of 65. Few studies have specifically addressed stroke risk factors in older people, although they potentially have the most to gain from prevention.1–5 Intervention studies of stroke prevention, eg, cholesterol lowering and treatment of hypertension, have shown that older people benefit from risk factor modification.6,7 Indeed, as they have a higher absolute risk of stroke compared with younger people, they have more to gain in terms of absolute risk reduction.8,9 Given the aging population, there is a need for prospective population-based cohort studies to enable assessment of the predictive value of risk factors on subsequent stroke mortality and morbidity in older people. We report the relative contribution of risk factors for first-ever stroke in a population-based cohort of older people.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We undertook a population-based survey of older people to determine the prevalence of atrial fibrillation (AF) in the United Kingdom.10 A random age- and sex-stratified community sample of 4773 residents of southeast Northumberland, England, aged >65 was recruited from 27 contiguous general practices between 1995 and 1997. The study area covered both rural and urban areas and the population is 99% white. We reviewed the primary care records (paper and electronic) of 4440 of these patients to supplement data obtained from 3678 (77%) who attended an initial screening visit during which they completed a questionnaire about sociodemographic details, health and lifestyle, and health status. Blood pressure and a limb-lead ECG were recorded. Table 1 shows risk factor definitions and data sources.
|
Primary and secondary care records were flagged to enable notification of cases of first-ever stroke, transient ischemic attack (TIA), and deaths in the 5 years subsequent to recruitment. Medical records were reviewed to seek further information about these events. Details of TIA were obtained to ensure that symptoms had not persisted beyond 24 hours and hence were due to stroke. Stroke was defined according to World Health Organization (WHO) criteria.11 Results of brain imaging and postmortem examinations were obtained. Stroke subtypes were recorded. At the end of the study period, all primary care notes were reviewed to ensure completeness of case ascertainment. When subjects had moved out of the area, a postal questionnaire was sent to their new general practitioner. The study was approved by Northumberland Ethics Committee and data stored in accordance with the UK Data Protection Act.
Statistical Analysis
The strength of the association between the various risk factors and time to first-ever stroke were assessed using Cox regression procedures. Initially each risk factor was included separately as an explanatory variable in a Cox proportional hazards model. In a second step, each risk factor was included in a model that also contained age and sex as covariates. A third step used stepwise modeling procedures to identify a set of risk factors such that each was significantly associated with risk of stroke even when estimates of hazard ratios were adjusted for all other risk factors included in the model. The first stage of this final step considered the various highly confounded indicators of heart disease separately to identify the one that best predicted risk of stroke by using both forward and backward elimination procedures. The final stage was to examine the relative importance of this variable alongside all the risk factors, again using both forward and backward elimination. A significance level of 5% was used as a guide as to whether a variable should be added to and retained in the model. Analysis was undertaken with SPSS (version 10) software.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The cohort has been described elsewhere10; 4440 had no stroke prior to entry. Eighty-nine patients had no follow-up data beyond baseline. Baseline and follow-up data were obtained from 4351 (98%) (the Figure).
|
During the 5-year follow-up period, 329 of 4351 (7.6%) had a first-ever stroke: 179 of 329 (54%) were due to cerebral infarction; 27 (8%), primary intracerebral hemorrhage; 2 (1%), subarachnoid hemorrhage; and 121 (37%) did not have brain imaging. Seventy-seven (23%) had a total anterior circulation stroke; 107 (33%), partial anterior circulation stroke; 69 (21%), lacunar stroke; 51 (16%), posterior circulation stroke; and 25 (8%) were unclassifiable. The 30-day case fatality was 95 of 329 (29%).
The crude prevalence of risk factors and hazard ratios adjusted for age and sex are given in Table 2. The following were associated with an increased risk of stroke: history of TIA (HR 2.18), atrial fibrillation (2.12), peripheral vascular disease (1.86), smoking (1.79), hypertension (1.68), heart failure (1.42), left ventricular hypertrophy (1.38), angina (1.34), ischemic heart disease (1.32), and male sex (1.28). Both age and blood pressure were associated with increased risk of stroke (an increase of 10 years in age and 10 mm Hg in blood pressure yield hazard ratios of 2.14 and 1.15, respectively). Those with poorer scores on the physical function and mental health subscales of the SF-36 were also at increased risk. Cardiovascular disease (CVD)—myocardial infarction (MI), angina, peripheral vascular disease (PVD), or heart failure—a variable used in the Framingham Study, had a HR of 1.57.12 No statistically significant relationship was found between first-ever stroke and diabetes, alcohol, MI, body mass index (BMI), and deprivation score.
|
Both forward and backward elimination procedures indicated that the most predictive of the cardiovascular indicators was CVD. Once this variable was included, none of the other cardiovascular variables were significant. The final stepwise selection of variables resulted in a final model that included age, atrial fibrillation, history of TIA, smoking, CVD, history of hypertension, and systolic blood pressure (Table 3).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Despite being predominantly a disease of older people, the majority of previous population-based cohort studies have excluded or underrepresented older people.13–19 Studies of older people have been undertaken in the United States,1–3 Australia,4 and Finland5 but we believe that this is the first population-based study to specifically address stroke risk factors for older people in the United Kingdom.
Our main effects model identified atrial fibrillation, history of hypertension, TIA, CVD, current smoking, systolic blood pressure, and increasing age as risk factors, confirming the importance of these "classic risk factors" in older people. Male sex, left ventricular hypertrophy (LVH), angina, ischemic heart disease, PVD, heart failure, and global health status were significant on univariate analysis but did not form part of the main effects model. We did not confirm diabetes, alcohol, MI, BMI, or socioeconomic status as risk factors in older people.
In the United Kingdom, the Health Authority Register is a reliable method for identifying representative samples of older people.20 Studies from Dubbo, Australia, and Turku, Finland also identified representative populations with high participation rates.4,5 The American Cardiovascular Health Study was a selected population, with only 57% of eligible subjects participating.2
We specifically addressed risk factors for first-ever stroke while other studies have included both first-ever and recurrent events4,5 or TIA and first-ever stroke.1 Ideally, suspected cases of stroke should be examined as soon as possible after the event to confirm the diagnosis. However, as the accuracy of the clinical diagnosis of stroke is reasonably good,21 we used a notification system and expert notes review to identify and confirm cases. We did not contact study participants after the initial assessment and therefore could not identify cases who did not seek medical attention. Participants in the Framingham and Cardiovascular Health Studies attended regular study clinics,1,2 and in addition to seeking data from hospital records and national death registers, the Dubbo study and the Turku study sent questionnaires to study participants.4,5
As with previous studies, our study has considered stroke as a single disease. The risk factors for different etiologies of stroke may differ in the strength and even direction of association, so ideally stroke type and subtype should be identified and analyzed separately. However, we were able to distinguish cerebral infarct from hemorrhage for 73% of strokes. Our initial study sought to determine the prevalence of atrial fibrillation; we did not take blood from participants so we lack measures of cholesterol or glucose.
The incidence of stroke doubles each decade over the age of 55 and it has been suggested that age is a proxy marker for duration of risk factor exposure.8 Our results were similar to those of the Dubbo study, which reported a relative risk of 1.06 (95% CI, 1.04 to 1.09) per year increase in age,10 and the Rochester study, which found a relative risk of 1.9 per 10 years.3 Age-adjusted stroke incidence rates are 30% higher in men than in women22; we have also found men to have an increased risk of stroke. Low socioeconomic status has been reported as a risk factor for stroke,23 but using the Townsend Deprivation Score,24 we did not find any relationship between socioeconomic status and stroke in older people. The British Regional Heart Study found no differences between social classes in stroke risk following adjustment for other risk factors.13 There are problems with assigning socioeconomic status to people beyond retirement age, which may explain our findings.
Hypertension is the most important modifiable risk factor for stroke, and in our cohort a history of hypertension was an independent risk factor. The relative risk of stroke also increased by 1.15 (95% CI, 1.08 to 1.22) per 10-mm Hg rise in systolic blood pressure. The relative risk of stroke has previously been shown be 2.3 per 10-mm Hg rise in diastolic blood pressure.25 The strength of association between blood pressure and stroke declines with increasing age26 because of the increased prevalence of hypertension in older people, which may explain our findings.8
LVH has previously been shown to be an important stroke risk factor even after adjustment for coexistent hypertension.27 Despite using the same ECG criteria as the Framingham study, LVH was not an independent risk factor for stroke. Others have also not found an independent association.4,28
CVD (MI, angina, PVD, or heart failure) was an independent risk factor, with a relative risk of 1.55. This is similar to that found in subjects aged 55 to 84 in the Framingham study, where a relative risk of 1.68 was found for men and 1.54 for women.29 As with the Turku study, TIA was an independent risk factor for stroke in older people.5
The prevalence of atrial fibrillation rises markedly with age, affecting 4.7% of those age >65 years.10 Our study confirms that atrial fibrillation is independently associated with an increased risk of stroke in older people. In the Framingham study, the relative risk for stroke for patients with atrial fibrillation rose markedly with age (RR 2.6 aged 60 to 69 years and RR 4.5 >80 years).29 Anticoagulation significantly reduces the risk of stroke, and at the time of the screening visit, 23% of our study were taking warfarin10; this may in part explain our lower relative risk. Neither the Dubbo nor Rochester study found an independent association with atrial fibrillation, perhaps due to the low prevalence of atrial fibrillation (2%) in Dubbo and the younger age group in Rochester.3,4
The Framingham study found obesity to be an independent risk factor for stroke.30 As with other studies, we did not find that obesity measured by BMI was a risk factor for stroke in older people.4 Previous studies have found that central obesity may be a more important risk factor than BMI. As prospective cohort studies have confirmed diabetes as a risk factor for stroke,31 we were surprised that diabetes was not an independent risk factor in our cohort.
The relative risk for smoking and stroke was similar to that found in the Framingham study, which found a 2-fold increase.29 A meta-analysis found 1.4 to 1.6 increase in relative risk for smokers.32 Using alcohol as a dichotomous variable, we did not find alcohol consumption to be a risk factor for stroke. There was a large proportion of nondrinkers and very few heavy drinkers in our cohort, so a more detailed analysis was not feasible. It has been suggested that there is a J-shaped relationship between alcohol consumption and ischemic stroke and that high alcohol consumption is being associated with primary intracerebral hemorrhage.33
The SF-36 was used to measure global health status, and both the physical and mental health subscales were associated with increased risk of stroke on univariate analysis but did not appear in the main-effects model. Physical activity has been shown to be associated with reduced risk of stroke.34,35 House et al reported an increased risk of stroke with life events,36 and in the Dubbo study a higher depression score was also associated with increased risk.4 Both physical and mental health subscores may be confounders as markers for other causes of ill health.
As the incidence of stroke rises markedly with age, given the aging population, stroke will be a major public health problem for the foreseeable future.37 However, stroke is not an inevitable consequence of aging, so by identifying and modifying risk factors for stroke in people of all ages, there are opportunities to reduce the incidence and mortality of this devastating condition.
|
Acknowledgments |
|---|
We are grateful to the Stroke Association and Northumbria Healthcare NHS Trust, who funded the follow-up phase of this study. The original cohort study was funded by the Stroke Association, Medical Research Council (UK), Northern Regional Health Authority, and Northumberland Health Authority. We would like to thank all general practitioners who participated. Particular thanks are due to Dawn Winpenny, Judy Murdy, Emma Hutchinson, Salma Ayis, and Mark Sudlow.
Received July 10, 2003; revision received September 2, 2003; accepted September 23, 2003.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
1. Wolf PA, D’Agostino RB, O’Neal MA, Sytkowski P, Kase CS, Belanger AJ, et al. Secular trends in stroke incidence and mortality: the Framingham Study. Stroke. 1992; 23: 1551–1555.
2. Fried LP, Kronmal RA, Newman A, Bild DE, Mittelmark MB, Polak JF, et al. Risk factors for 5-year mortality in older adults: the Cardiovascular Health Study. JAMA. 1998; 297: 585–592.
3. Davis PH, Dambrosia JM, Schoenberg BS, Schoenberg DG, Pritchard DA, Lilienfeld AM, et al. Risk factors for ischemic stroke: a prospective study in Rochester, Minnesota. Ann Neurol. 1987; 22: 319–327.[CrossRef][Medline] [Order article via Infotrieve]
4. Simons LA, McCallum J, Friedlander Y, Simons J. Risk factors for ischemic stroke: Dubbo study of the elderly. Stroke. 1998; 29: 1341–1346.
5. Kaarisalo MM, Immonen-Raiha P, Marttila RJ, Lehtonen A, Torppa J, Tuomilehto J. Long-term predictors of stroke in a cohort of people aged 70 years. Arch Gerontol Geriatr. 2000; 31: 43–53.[CrossRef][Medline] [Order article via Infotrieve]
6. Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial. Lancet. 2002; 360: 7–22.[CrossRef][Medline] [Order article via Infotrieve]
7. SHEP Co-operative Research Group. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension: final results of the Systolic Hypertension in the Elderly Programme (SHEP). JAMA. 1991; 265: 3255–3264.
8. Psaty BM, Koepsell TD, Manolio TA, Longstreth WT, Wagner EH, Wahl PW, et al. Risk ratios and risk differences in estimating the effect of risk factors for cardiovascular disease in the elderly. J Clin Epidemiol. 1990; 43: 961–970.[CrossRef][Medline] [Order article via Infotrieve]
9. Thomson R. How can epidemiological studies help us to prevent stroke? The example of atrial fibrillation. Age Ageing. 2002; 31: 9–16.
10. Sudlow M, Thomson R, Thwaites B, Rodgers H, Kenny RA. Prevalence of atrial fibrillation and eligibility for anticoagulants in the community. Lancet. 1998; 352: 1167–1671.[CrossRef][Medline] [Order article via Infotrieve]
11. Aho K, Harmsen P, Hatano S, Marquardsen J, Smirnov V, Strasser T. Cerebrovascular disease in the community: results of a WHO Collaborative Study. Bull World Health Organ. 1980; 58: 113–130.[Medline] [Order article via Infotrieve]
12. D’Agostino RB, Wolf PA, Belanger AJ, Kannel WB. Stroke risk profile: adjustment for antihypertensive medication: the Framingham Study. Stroke. 1994; 25: 40–43.[Abstract]
13. Shaper AG, Phillips AN, Pocock SJ, Walker M, Macfarlane PW. Risk factors for stroke in middle aged British men. BMJ. 1991; 302: 1111–1115.
14. Lindenstrom E, Boysen G, Nyboe J. Risk factors for stroke in Copenhagen, Denmark, I. basic demographic and social factors. Neuroepidemiology. 1993; 12: 37–42.[Medline] [Order article via Infotrieve]
15. Hart CL, Hole DJ, Smith GD. Risk factors and 20-year stroke mortality in men and women in the Renfrew/Paisley study in Scotland. Stroke. 1999; 30: 1999–2007.
16. Abbott RD, Behrens GR, Sharp DS, Rodriguez BL, Burchfiel CM, Ross GW, et al. Body mass index and thromboembolic stroke in non-smoking men in older middle age. Stroke. 1994; 25: 2370–2376.[Abstract]
17. Tanne D, Yaari S, Goldbourt U. Risk profile and prediction of long-term ischemic stroke mortality: a 21-year follow-up in the Israeli Ischemic Heart Disease (IIHD) Project. Circulation. 1998; 98: 1365–1371.
18. Haheim LL, Holme I, Hjermann I, Leren P. Smoking habits and risk of fatal stroke: 18 years follow up of the Oslo Study. J Epidemiol Commun Health. 1996; 50: 621–624.
19. Greenwood R, McCarron P, Elwood P, Shlomo YB, Bayer A, Baker I, et al. The incidence and aetiology of stroke in the Caerphilly and Speedwell Collaborative Studies I: methods and incidence of events. Public Health. 2001; 115: 4–11.[Medline] [Order article via Infotrieve]
20. O’Mahony PG, Thomson RG, Rodgers H, Dobson R, James OFW. Accuracy of the Family Health Service Authority Register as a sampling frame for population studies. J Epidemiol Commun Health. 1997; 51: 206–207.
21. Sandercock P, Molyneux A, Warlow C. Value of computed tomography in patients with stroke. BMJ. 1985; 290: 193–197.
22. Sacco RL, Benjamin EJ, Broderick JP, Dyken M, Easton JD, Feinberg WM, et al. American Heart Association Prevention Conference IV: prevention and rehabilitation of stroke risk factors. Stroke. 1997; 28: 1507–1517.
23. Acheson RM, Sanderson C. Stroke: social class and geography. Popul Trends. 1978; 12: 13–17.
24. Morris R, Carstairs V. Which deprivation index? A comparison of selected deprivation indexes. J Public Health Med. 1991; 13: 318–326.
25. Lewington S, Clarke R, Qizilbash N, Peto R, Collins R. Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet. 2002; 360: 1903–1913.[CrossRef][Medline] [Order article via Infotrieve]
26. Prospective Studies Collaboration. Cholesterol, diastolic blood pressure and stroke: 13,000 strokes in 450,000 people in 45 prospective cohorts. Lancet. 1995; 346: 1647–1653.[CrossRef][Medline] [Order article via Infotrieve]
27. Bikkina M, Levy D, Evans JC, Larson MG, Benjamin EJ, Wolf PA, et al. Left ventricular mass and risk of stroke in an elderly cohort: the Framingham Heart Study. JAMA. 1994; 272: 33–36.
28. Welin L, Svardsudd K, Wilhelmsen L, Larsson B, Tibblin G. Analysis of risk factors for stroke in a cohort of men born in 1913. N Engl J Med. 1987; 317: 521–526.[Abstract]
29. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke. 1991; 22: 983–988.
30. Wolf PA, Kannel WB. Reduction of stroke through risk factor modification. Semin Neurol. 1986; 6: 243–253.[Medline] [Order article via Infotrieve]
31. Sacco RL. Risk factors, outcomes, and stroke subtypes for ischemic stroke. Neurology. 1997; 49: S39–S44.
32. Shinton R, Beevers G. Meta-analysis of relation between cigarette smoking and stroke. BMJ. 1989; 298: 789–794.
33. Camargo CA Jr. Moderate alcohol consumption and stroke: the epidemiologic evidence. Stroke. 1989; 20: 1611–1626.
34. Abbott RD, Rodriguez BL, Burchfiel CM, Curb JD. Physical activity in older middle-aged men and reduced risk of stroke: the Honolulu Heart Program. Am J Epidemiol. 1994; 139: 881–893.
35. Wannamethee G, Shaper AG. Physical activity and stroke in British middle-aged men. BMJ. 1992; 304: 597–601.
36. House A, Dennis M, Mogridge L, Hawton K, Warlow C. Life events and difficulties preceding stroke. J Neurol Neurosurg Psychiatry. 1990; 53: 1024–1028.
37. Abbott R, Curb J, Rodriguez B, Masaki K, Popper J, Ross G, et al. Age-related changes in risk factor effects on the incidence of thromboembolic and haemorrhagic stroke. J Clin Epidemiol. 2003; 56: 479–486.[CrossRef][Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  D. J Stott, P. Welsh, A. Rumley, M. Robertson, I. Ford, N. Sattar, R. G J Westendorp, J W. Jukema, S. M Cobbe, and G. D O Lowe Adipocytokines and risk of stroke in older people: a nested case-control study Int. J. Epidemiol., February 1, 2009; 38(1): 253 - 261. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Y. Zubkov, A. A. Rabinstein, A. Dispenzieri, and E. F.M. Wijdicks Primary systemic amyloidosis with ischemic stroke as a presenting complication Neurology, September 11, 2007; 69(11): 1136 - 1141. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. J. Hankey Potential New Risk Factors for Ischemic Stroke: What Is Their Potential? Stroke, August 1, 2006; 37(8): 2181 - 2188. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R Thomson, M Murtagh, and F-M Khaw Tensions in public health policy: patient engagement, evidence-based public health and health inequalities Qual. Saf. Health Care, December 1, 2005; 14(6): 398 - 400. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. D. Syme, A. W. Byrne, R. Chen, R. Devenny, and J. F. Forbes Community-Based Stroke Incidence in a Scottish Population: The Scottish Borders Stroke Study Stroke, September 1, 2005; 36(9): 1837 - 1843. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||