Influenza Vaccination Is Associated With a Reduced Risk of Stroke
Background and Purpose— Vascular death rates and hospitalizations for stroke are increased during influenza epidemics. Influenza vaccination may prevent stroke. We investigated whether influenza vaccination is associated independently with reduced odds of stroke and whether effects are confined to stroke subgroups and winter seasons and shared by other vaccinations.
Methods— During 18 months, we performed standardized interviews assessing vaccination status, risk factors, health-related behavior, and socioeconomic factors in 370 consecutive patients with ischemic or hemorrhagic stroke or transient ischemic attack (TIA) and 370 age- and sex-matched control subjects selected randomly from the population.
Results— Influenza vaccination during the last vaccination campaign was less common in patients (19.2%) than control subjects (31.4%; P<0.0001). After adjustment for vascular risk factors, education, health-related behavior and other factors, influenza vaccination (odds ratio [OR], 0.46; 95% CI, 0.28 to 0.77) but not other combined recent vaccinations (OR, 0.80; 95% CI, 0.42 to 1.43) were associated with reduced odds of stroke/TIA. Significant effects were found in men, older subjects (>65 years), subjects with previous vascular diseases, and regarding ischemic stroke; nonsignificant trends existed in women, younger subjects, and regarding hemorrhagic stroke. In etiologic subgroups of cerebral ischemia, similar effects were found. No protective effects were found during summer months; however, results also varied considerably between both winter seasons examined.
Conclusions— These results support the hypothesis that influenza vaccination may be associated with reduced stroke risk. However, residual confounding cannot be excluded, and interventional studies are required to evaluate the role of influenza vaccination in stroke prevention.
The generally accepted risk factors for stroke are not sufficient to explain the total burden of stroke in populations. Recent studies have indicated that acute and chronic infections may also contribute to stroke risk.1 Each year, ≈5% of adults develop symptomatic influenza, a worldwide distributed contagion. Influenza causes a broad range of illness, from symptomless infection to various respiratory syndromes and disorders affecting the heart, brain, and other organs, to fulminant primary viral and secondary bacterial pneumonia.2 During influenza epidemics, hospitalizations for stroke and cardiac diseases increase, and more than half of the excess mortality during such epidemics was attributed to causes other than influenza, including cardiovascular diseases and stroke.2–4⇓⇓
Influenza vaccination is an effective measure to reduce hospitalization and death from all causes in the elderly and work absenteeism in adults of working age.2 Two studies showed that influenza vaccination is associated with reduced cardiac morbidity,5,6⇓ whereas another study did not find such benefit.7 A study investigating 90 ischemic stroke patients and 180 community controls aged ≥60 years found that influenza vaccination was associated with reduced odds of stroke after adjustment for potential confounders (odds ratio [OR], 0.50; 95% CI, 0.26 to 0.94).8 In a large observational study with >140 000 elderly subjects, influenza vaccination was associated with a significant reduction of hospitalizations for cerebrovascular disease (−16% 1998 to 1999; −23% 1999 to 2000) and all-cause mortality (−48% 1998 to 1999; −50% 1999 to 2000) during influenza seasons.9
We performed a case-control study to investigate whether influenza vaccination is associated independently with reduced odds of stroke and whether any effect is confined to older subjects, stroke subgroups and winter seasons, and shared by other vaccinations.
Subjects and Methods
We performed a case-control study with consecutive patients admitted for ischemic or hemorrhagic stroke or transient ischemic attack (TIA) between November 2001 and April 2003. Control subjects were selected randomly from the general population and matched to patients 1-to-1 for age, sex, and area of residence. Exclusion criteria were inability to give informed consent or to follow the interview in native language because of aphasia, reduced consciousness, or other reasons. Among 385 eligible patients, 370 (96.1%) agreed to study participation. All patients received a computed tomography or MRI scan. Ischemic stroke was defined as a focal neurological deficit of acute onset lasting >24 hours or a new ischemic lesion on neuroimaging and cerebral hemorrhage ruled out by neuroimaging. TIA was diagnosed if a sudden deficit lasted <24 hours and neuroimaging showed no new vascular lesion. Hemorrhagic stroke was diagnosed by neuroimaging showing cerebral bleeding. Etiologic subgroups of cerebral ischemia were based on Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria.10 Control subjects were selected from a random sample of the inhabitants of 55 communities representing the majority of the catchment area of the hospital. The random sample contained name, sex, age, and address of 2% of all inhabitants. The first potential control subject who was identified in this sample and matched to the respective patient for age, sex, and area of residence was selected and contacted by mail and later by phone. Among 430 subjects contacted by phone, 370 (86.0%) agreed to study participation.
Subjects were interviewed in a standardized way by trained interviewers using a structured questionnaire. The interview was performed face-to-face for patients and over the phone for control subjects. The interview contained questions regarding vascular risk factors and diseases, other severe and mainly disabling diseases, chronic infections (eg chronic bronchitis), smoking, alcohol consumption, medication, education, current profession and marital status, family history, physical activity, symptoms of recent infection, lifestyle factors, health-related behavior, and influenza vaccination during the last influenza vaccination campaign and during the last 5 years, as well as other previous vaccinations. The period of the influenza campaign was defined as September and October and events after November 1 were related to the directly preceding influenza campaign (eg, patients with stroke in October 2002 and their respective control subjects were asked for vaccination in the 2001 campaign; patients with stroke in November 2002 and their respective control subjects were asked for vaccination in the 2002 campaign). Vaccinations during winter months after November 1 were also acknowledged if performed before events. Self-reported risk factors and diseases were acknowledged if subjects affirmed that a physician had made the respective diagnosis previously. To assess health-related lifestyle, we asked for information on medical and dental routine check-ups, use of antibiotics, and usual behavior in acute febrile infection (Table 1⇓). The ethics committee approved the study protocol. All subjects gave informed consent.
We compared the frequency of influenza vaccination and other variables using the McNemar test. Conditional logistic regression analysis was used to adjust for other variables and included all generally accepted stroke risk factors (hypertension, diabetes mellitus, smoking, previous stroke/TIA, and hyperlipidemia) plus all factors being significant (P<0.05) in univariate analysis. Subgroup analyses were performed for patients with ischemic stroke, TIA, and hemorrhagic stroke, and regarding season (November to April versus May to October), age, sex, previous vascular disease, diabetes mellitus, and whether influenza vaccination was recommended according to current national recommendations (age >60, previous vascular disease, diabetes mellitus, chronic airway disease, immunodeficiency, and other chronic diseases). Because of lower numbers of subjects, a reduced model was used for subgroup analyses only containing those generally accepted risk factors that were significant in multivariate analyses with all subjects. Data were analyzed using the software package SAS.
Patients had received an influenza vaccination less often than control subjects either during the last vaccination campaign (OR, 0.51; 95% CI, 0.36 to 0.72) or during any campaign within the last 5 years (OR, 0.48; 95% CI, 0.35 to 0.66). Other vaccinations were also less often reported by patients than control subjects; however, the difference was of borderline significance. Conventional risk factors were all more common in patients than control subjects. In addition, patients more often reported a history of chronic bronchitis, >2 episodes of flu-like illnesses per year, shorter school education, the readiness to continue work despite febrile infection, and less often leisure time physical activity than control subjects (Table 1⇑). In multivariate analyses, influenza vaccination within the previous campaign was associated with a reduced risk of stroke/TIA (OR, 0.46; 95% CI, 0.28 to 0.77), as was any influenza vaccination within 5 years (OR, 0.46; 95% CI, 0.28 to 0.74), whereas combined other vaccinations were not correlated with stroke risk (Table 2). Recent tetanus vaccination was correlated with reduced odds of stroke/TIA in univariate (OR, 0.54; 95% CI, 0.31 to 0.93) but not multivariate analysis (OR, 0.90; 95% CI, 0.42 to 1.95).
Influenza vaccination was associated with a reduced risk of stroke/TIA in men, older subjects, subjects with an indication for influenza vaccination according to current guidelines, and subjects with previous vascular disease, and it tended to be correlated with a reduced risk in women, younger subjects, and subjects without present indication. The odds of ischemic stroke were reduced significantly, and the odds of hemorrhagic stroke but not TIA tended to be reduced by recent influenza vaccination. In all etiologic subgroups of ischemic stroke/TIA, there was a trend toward reduced risk associated with recent influenza vaccination (Table 3). Vaccination rates were 11.6% in patients and 37.0% in control subjects presenting during the first winter season (November to April; OR [multivariate model], 0.11; 95% CI, 0.03 to 0.41), 25.9% in patients, and 31.5% in control subjects investigated during the summer season (OR, 0.99; 95% CI, 0.28 to 3.51), and 21.8% in patients and 25.0% in control subjects presenting during the second winter season (OR, 0.90; 95% CI, 0.37 to 2.19). Among patients, 135 (36.5%) reported fever or at least one symptom indicative of respiratory infection (cough, hoarseness, rhinitis) within 1 week before stroke/TIA. The rate of recent infection was 36.5% (109 of 299) in patients with and 36.6% (26 of 71) in those without influenza vaccination.
In line with recent studies,8,9⇓ we found that influenza vaccination is associated with reduced risk of stroke. In addition to previous studies, we differentiated between stroke etiologies and found that influenza vaccination was associated with protection from ischemic stroke and with a trend toward protection from hemorrhagic stroke but not from TIA. Effects in etiologic subgroups of cerebral ischemia were of similar size; however, caution is warranted because of small numbers in several subgroups.
Vaccination behavior may reflect general attitudes toward health issues, and therefore, socioeconomic factors and health consciousness may have an important impact on influenza vaccination. Thus, we analyzed a large number of potential confounders. Interestingly, patients reported frequent flu-like illnesses more often than control subjects, a finding that was independent from other covariates and may indicate a greater risk of infection-associated morbidity. More patients than control subjects reported to continue work despite acute febrile disease, indicating a less careful attitude toward health-related risks or greater social pressure in times of high unemployment. However, on adjustment for these and several other potential confounders, recent influenza vaccination still was associated strongly with odds of stroke/TIA. Although this could indicate a specific preventive effect, confounders not assessed in our study (such as nursing home residence and functional status indicators) could still have contributed to the effect.
Given the lack of medical record review, results may also have been influenced by a recall bias, with patients remembering vaccinations less often than controls, although it is unlikely that this would account for most of the effect. We also tried to elucidate whether more antecedent vaccinations may also provide protection, but subjects were unsure about annual vaccinations, and because of missing values, we could not analyze this issue in greater detail. However, subjects recalled whether they had ≥1 influenza vaccination within 5 years, and here, results were similar to findings regarding the last campaign. Influenza vaccination is recommended before the influenza period; however, subjects can also be vaccinated later. Our study design includes the possibility of a higher chance of controls subjects to have received the vaccine before the interview than patients before stroke/TIA if subjects were vaccinated after November 1. However, effects during the last months of the influenza season (February to April), when vaccination is hardly performed anymore, were similar (univariate OR, 0.41; 95% CI, 0.20 to 0.82; multivariate OR, 0.52; 95% CI, 0.19 to 1.44) than for the whole period, arguing against any bias by time of vaccination.
In contrast to previous studies,8,9⇓ we also included younger persons because beneficial effects may not be restricted to the elderly, and national vaccination guidelines also include younger subjects with several previous and chronic diseases. Preventive effects were smaller in younger patients and did not reach statistical significance after adjustment for several covariates. Even in the elderly, vaccination rates were lower in our study compared with recent reports from other countries,8,9⇓ but the vaccination frequency among our control subjects was comparable to rates reported for our country. The patient population at our university center is younger than in population-based studies and skewed to a preponderance of atherothrombotic and uncommon etiologies that are more common in young and middle-aged patients.
Several case-control studies and a recent large study based on within-person comparisons showed that acute infection is a trigger factor for stroke.12–17⇓⇓⇓⇓⇓ Therefore, influenza vaccination may protect from stroke by preventing infections with influenza viruses and superimposed bacterial infections. Procoagulant mechanisms, fever, dehydration, and proteolytic lesions to the vessel wall are mechanisms that could increase the risk of ischemic and hemorrhagic stroke after infections. The rate of self-reported fever or symptoms of respiratory infection was not different between vaccinated and nonvaccinated patients, a result that does not exclude that the rate of specific infection, especially influenza infection, and of subclinical infection was lower in vaccinated subjects. It is a limitation of this and previous studies that acute influenza infection was not investigated. Future studies may use surveillance data to define seasons of active influenza as the primary outcome period.
The efficacy of vaccination to prevent influenza is estimated to be ≈80%.2 About 62% of all strokes would have to be associated with influenza in order for vaccination to reduce stroke by ≈50%. This rate appears implausibly high, and it is much higher than those found in studies assessing the etiology of infections before stroke.13,16⇓ Influenza vaccination may also be effective by avoiding subclinical infection and by other not yet defined mechanisms. However, it is also possible that our results are in part attributable to bias by residual confounding. To study the specificity of influenza vaccination, we also assessed the role of other vaccinations. These vaccinations combined, and tetanus vaccination as the most prevalent vaccination, were more common in control subjects but not associated with stroke protection after adjustment for other variables, rather suggesting that not vaccinating, per se, is protective. However, the power to detect any effect was limited because of the low prevalence of other vaccinations.
Studying effects during different seasons could help to elucidate how influenza vaccination may contribute to stroke protection. Influenza and other respiratory infections are more common during winter than summer months, and this could contribute to the higher incidence of stroke during the colder season found in most, although not all, respective studies.1 To test the hypothesis that influenza vaccination is associated with lower odds of stroke only during winter and not during summer months, we investigated effects during the whole year.8,9⇓ In fact, no protective effect could be found during summer months; however, even both winter seasons assessed yielded widely differing results. Most likely, this result reflects the play of chance and not different vaccination behavior in both years in the populations investigated. Given this difference between 2 winter seasons, our results regarding the effect during summer months may not definitely exclude any effect during warmer seasons, and further, preferably larger, studies are required to study this issue.
In a first small randomized study with acute myocardial infarction or planned angioplasty or stenting, influenza vaccination was associated with a reduced rate of cardiovascular death and a reduced rate of death, reinfarction, and rehospitalization for ischemia.18 Potentially, influenza vaccination may prevent strokes; however, because of the retrospective design of our study, definite conclusions cannot be drawn. To test whether influenza vaccination can be added to the recommendations for stroke prevention, interventional studies in stroke patients are required.
- Received January 25, 2005.
- Revision received April 18, 2005.
- Accepted April 26, 2005.
Lindsberg P, Grau AJ. Inflammation and infections as risk factors for ischemic stroke. Stroke. 2003; 34: 2518–2532.
Nicholson KG, Wood JM, Zambon M. Influenza. N Engl J Med. 2003; 362: 1733–1745.
Housworth J, Langmuir AD. Excess mortality from epidemic influenza, 1957–1966. Am J Epidemiol. 1974; 100: 40–48.
Alling DW, Blackwelder WC, Stuart-Harris CH. A study of excess mortality during influenza epidemics in the United States 1968–1976. Am J Epidemiol. 1981; 113: 30–43.
Siscovick DS, Raghunathan TE, Lin D, Weinmann S, Arbogast P, Lemaitre RN, Psaty BM, Alexander R, Cobb LA. Influenza vaccination and the risk of primary cardiac arrest. Am J Epidemiol. 2000; 152: 674–677.
Naghavi M, Barlas Z, Siadaty S, Naguib S, Madjid M, Casscells W. Association of influenza vaccination and reduced risk of recurrent myocardial infarction. Circulation. 2000; 102: 3039–3045.
Jackson LA, Yu O, Heckbert SR, Psaty BM, Malais D, Barlow WE, Thompson WW. Influenza vaccination is not associated with a reduction in the risk of recurrent coronary events. Am J Epidemiol. 2002; 156: 634–640.
Lavallée P, Perchaud V, Gautier-Bertrand M, Grabli D, Amarenco P. Association between influenza vaccination and reduced risk of brain infarction. Stroke. 2002; 33: 513–518.
Adams HP Jr, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, Marsh EE III. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. Trial of Org 10172 in Acute Stroke Treatment. Stroke. 1993; 24: 35–41.
Syrjänen J, Valtonen VV, Iivanainen M, Kaste M, Huttunen JK. Preceding infection as an important risk factor for ischaemic brain infarction in young and middle aged patients. BMJ. 1998; 296: 1156–1160.
Grau AJ, Buggle F, Heindl S, Steichen-Wiehn C, Banerjee T, Maiwald M, Rohlfs M, Suhr H, Fiehn W, Becher H, Hacke W. Recent infection as a risk factor for cerebrovascular ischemia. Stroke. 1995; 26: 373–379.
Macko RF, Ameriso SF, Barndt R, Clough W, Weiner JM, Fisher M. Precipitants of brain infarction. Roles of preceding infection/inflammation and recent psychological stress. Stroke. 1996; 27: 1999–2004.
Bova IY, Bornstein NM, Korczyn AD. Acute infection as a risk factor ischemic stroke. Stroke. 1996; 27: 2204–2206.
Grau AJ, Buggle F, Becher H, Zimmermann E, Spiel M, Fent T, Maiwald M, Werle E, Zorn M, Hengel H, Hacke W. Recent bacterial and viral infection is a risk factor for cerebrovascular ischemia. Clinical and biochemical analysis. Neurology. 1998; 50: 196–203.
Gurfinkel EP, de la Fuente RL, Mendiz O, Mautner B; for the FLUVACS Study Group. Influenza vaccine pilot study in acute coronary syndromes and planned percutaneous coronary interventions. Circulation. 2002; 105: 2143–2147.