Age and Ageing Advance Access originally published online on November 21, 2008
Age and Ageing 2009 38(1):100-103; doi:10.1093/ageing/afn232
Temporal relationship between influenza infections and subsequent first-ever stroke incidence
SIR—an excess of deaths from cardiovascular disease in winter months that might be related to previous acute respiratory infections has been reported [1–3]. Only few studies of reasonable size and length have reported a peak of stroke incidence in the respective winter months in the Northern (December to March) and Southern (July to August) hemisphere [4, 5]. Furthermore, only one large-scale study has reported a significant association between respiratory infection and subsequent stroke [6]. However, this study did not differentiate between stroke subtypes despite different underlying pathomechanisms between ischaemic and haemorrhagic strokes.A number of studies have failed to demonstrate a protective effect of antibiotics on cardiovascular risk underlining the importance of viruses for respiratory infections, of which influenza represents a major contributor [7, 8]. Additionally, influenza infections are routinely monitored by national surveillance systems in England and Wales. Therefore, we took advantage of an established active national surveillance system monitoring influenza infections in England and Wales (http://www.hpa.org.uk) and compared these infection rates with subtype-specific first-ever stroke data from the South London Stroke Register (SLSR) in order to assess the seasonality of stroke incidence in the SLSR and the temporal relationship between influenza and ischaemic and/or haemorrhagic strokes.
Ascertainment of incident strokes
Stroke patients were identified within the SLSR. The SLSR is a prospective ongoing population-based stroke register set up in 1995, recording all first-ever strokes in patients of all age groups for an inner area of South London based on 22 electoral wards in Lambeth and Southwark. Data collected between 1995 and 2004 were used in this analysis. The total source population of the SLSR area was 271,817 at the Census 2001.
Stroke was defined according to WHO criteria [9]. Special trained study nurses and field workers collected all data prospectively. A study clinician verified the diagnosis of stroke. Strokes were classified as first ever if there was no history of a previous stroke and as recurrent if a stroke was recorded prior to the current event. Classification of pathological stroke subtype [ischaemic stroke (IS), primary intracerebral haemorrhage (PICH) and subarachnoid haemorrhage (SAH)] was based on results from at least one of the following: brain imaging (computerised tomography or magnetic resonance imaging), cerebrospinal fluid analysis (in all living cases of SAH where brain imaging was not diagnostic) or necropsy examination. Details of the SLSR are reported elsewhere [10].
Influenza data
Weekly number of influenza patients in England and Wales by time period was obtained from the Health Protection Agency (HPA) for the period from 1995 to 2004 (http://www.hpa.org.uk). The HPA centre for infectious diseases (Cfl) monitors and records influenza cases in England and Wales. Clinical data were obtained from GP's surgeries which reported the consultations for influenza-like illness and other acute respiratory illness. The clinical cases that were confirmed by NHS and HPA laboratories were reported to the Cfl. The virological data that were used in this study were collected by two sentinel schemes: the Respiratory Diseases Department, CfI, collaborative study and the Virus Reference Department & Royal College of General Practitioners, collaborative study (http://www.hpa.org.uk).
Statistical analysis
The seasonal distributions of stroke and influenza cases were separately analysed with the R package surveillance [11]. The algorithm by Farrington et al. was used to identify outbreaks of the two time series with a random error level of 
= 0.0001 [12]. Zero counts of weekly influenza or stroke data were replaced by 0.1 to ensure the convergence of the algorithm. We estimated seasonal trends based on LOESS as suggested by Cleveland and colleagues [13].
We performed a multivariate time series analysis to assess an association between influenza infections and strokes using the established ‘3h-algorithm’. This algorithm is based on a multivariate branching process with immigration and yields robust and valid estimations [14]. We considered different time lags between preceding influenza and stroke from 1 week up to 5 weeks to identify the length of increased hazard after an influenza infection. We used likelihood ratio tests estimating the goodness of fit of different models.
All calculations were carried out with the statistical software package R version 2.6.1 [15].
Ethics
Patients and/or their relatives gave written informed consent to participate in stroke-related studies within the SLSR. The design was approved by the ethics committees of Guy's and St Thomas’ Hospital Trust, King's College Hospital, Queens Square, and Westminster Hospital (London); no additional approval was required for this study.
Between 1995 and 2004 a total of 2,874 patients with first-ever stroke were registered in the SLSR. Overall 2,092 (73%) ischaemic and 566 (20%) haemorrhagic strokes were registered throughout the study period. The seasonal trend of first-ever ischaemic strokes showed three peaks, a peak from end December to end January, a peak at the end of April and a peak from mid-September to mid-October (Figure 1). The distribution of haemorrhagic strokes showed only one peak from mid-February to beginning of May (data not shown).
|
In the respective time period, 21,473 influenza cases were reported to the HTA. Most cases occurred between the end of December and March. Three outbreaks could be identified for influenza by the Farrington algorithm, in 1998, 1999 and 2003. Although slightly more influenza infections could be observed than expected for certain weeks in 2001, 2002 and 2004, these only occurred in few single, not linked weeks (Figure 2). In contrast, no outbreak in ischaemic, haemorrhagic or overall strokes was identified by the Farrington algorithm during the entire observation period (data not shown).
|
Multivariate analysis of the influenza and ischaemic stroke time series yielded a significant association between influenza and first-ever strokes with a time lag of 1 week and 2 weeks (both P < 0.001). For a time lag of more than 2 weeks, no significant association could be observed (e.g. P = 0.315 for a time lag of 3 weeks).
We also observed an association between influenza and haemorrhagic strokes for a time lag of 1, 2, 3 and 4 weeks with P-values <0.001, while a time lag of 5 or more weeks did not show any association (e.g. P = 0.401 for time lag of 5 weeks). Both ischaemic and haemorrhagic strokes showed the best fit for the shortest time lag after influenza infection (1 week) and a decreasing goodness of fit for larger time lags in terms of likelihood ratios.
A temporal relationship between influenza infections and strokes was observed. The strength of the relationship was inversely related to the length of the time lag between influenza and stroke pointing to a contemporary increased stroke risk after influenza. This relationship could partly explain the increased number of ischaemic strokes within 2 weeks and the increased number of haemorrhagic strokes within 4 weeks after the seasonal influenza peak. However, the distribution of ischaemic strokes had also an early autumn peak outside the influenza season indicating that preceding influenza might only be a trigger besides various other risk factors contributing to stroke aetiology and possibly responsible for the other peaks outside the influenza season. Additionally, the relationship between influenza and stroke was not specific with regard to different subtypes except with respect to the time element.
We observed a statistical link between preceding influenza infections and strokes that is similar to the only study reporting respiratory infections related to stroke incidence [6]. However, this study looked at respiratory infections in general, pointing to the nonspecific role of influenza. Additionally, evidence from the literature shows that stroke might be related to other acute infections like urinary tract infections [16]. A review also reported a strong evidence of an association between inflammation and ischaemic stroke, possibly due to atherogenic mechanisms and/or possible plaque ruptures [17].
Additionally, levels of inflammatory markers such as fibrinogen and platelets are known to exhibit a biphasic distribution among the elderly with increasing levels in late winter and a second peak in late summer/early autumn [18, 19], while protein C and antiplasmin were lower in winter suggesting an increased risk of atherothrombotic disease [20]. A small subsequent study by the same authors failed to show an association between winter infections and variations in C-reactive protein and fibrinogen levels, however, possibly due to the lack of sample power [21]. Nevertheless, the two fibrinogen peaks might partially coincide with two of the ischaemic stroke peaks in the SLSR occurring in early spring and autumn.
The observed link between influenza and haemorrhagic strokes suggests that other reasons than the possible inflammatory mechanisms for ischaemic strokes may be involved in the aetiology of chronic vascular diseases after acute infections. Reduced endothelial reactivity that has been observed after infections might play a role in the increased haemorrhagic stroke risk after influenza [22]. The longer lasting increased hazard from the time of infection to haemorrhagic stroke with 4 weeks in total compared to ischaemic stroke with 2 weeks in total after influenza further indicates different underlying mechanisms for both diseases. However, our data do not allow the study of potential pathological mechanisms.
While influenza data were obtained for England and Wales, stroke cases were only registered in the SLSR area. Unfortunately, we lacked information on weekly influenza numbers for the SLSR area only. A bias due to varying influenza seasons in the SLSR area compared to the whole area of England and Wales cannot be ruled out, but seems to be unlikely due to (1) the rapid spread of influenza infections and (2) London and Southern England contributing to more than one-third of all registered influenza cases.
Recurrent strokes might also be triggered by influenza. However, this possible association was not assessed since only first-ever strokes were considered due to appropriate sample power.
An unadjusted statistical link between influenza infections and first-ever strokes was observed. However, a peak of ischaemic strokes could also be observed in early autumn that is unlikely to be related to influenza. We unfortunately lacked information on additional seasonal factors such as temperature, humidity, weather patterns or RSV infections that should also be considered in future studies. Additionally, other acute infections were also reported to be related to first-ever strokes; general rather than influenza-specific inflammatory responses might be instrumental. Since both ischaemic and haemorrhagic stroke were associated with preceding influenza, exact mechanisms of acute infections and subsequent vascular disease have to be further disentangled.
- Excess of deaths from cardiovascular disease in winter months have been described, but only one study reported a relationship between respiratory infections and total stroke incidence without considering the subtype.
- A seasonal trend for ischaemic strokes with three peaks in winter, early spring and early autumn could be observed, while there was only one long-lasting late winter/early spring peak for haemorrhagic strokes.
- There was a significant increase of ischaemic strokes within 2 weeks after influenza infection and within 4 weeks after infection for haemorrhagic strokes.
The authors declare no conflict of interest.
The study was funded by the Northern& Yorkshire NHS R&D Programme in Cardiovascular Disease and Stroke, Guy's and St Thomas’ Hospital Charity, Stanley Thomas Johnson Foundation, Department of Health, UK. The authors acknowledge financial support from the Department of Health via the National Institute for Health Research (NIHR) Biomedical Research Centre award to Guy's & St Thomas’ NHS Foundation Trust in partnership with King's College London.
Acknowledgements
We wish to thank all the patients and their families and the health care professionals involved. Particular thanks to all the field workers and the whole team who have collected data since 1995 for the South London Stroke Register.
Division of Health and Social Care, Department of Public Health Sciences, King's College London, 42 Weston St, 7th Floor, Capital House, London SE1 3QD, UK
* To whom correspondence should be addressed Email: michael.toschke{at}kcl.ac.uk
References
- Curwen M. Excess winter mortality: a British phenomenon? Health Trends (1990) 22:169–75.
- Kloner RA, Poole WK, Perritt RL. When throughout the year is coronary death most likely to occur? A 12-year population-based analysis of more than 220 000 cases. Circulation (1999) 100:1630–4.
[Abstract/Free Full Text] - Meier CR, Jick SS, Derby LE, Vasilakis C, Jick H. Acute respiratory-tract infections and risk of first-time acute myocardial infarction. Lancet (1998) 351:1467–71.[CrossRef][Web of Science][Medline]
- Nakaji S, Parodi S, Fontana V, et al. Seasonal changes in mortality rates from main causes of death in Japan (1970–1999). Eur J Epidemiol (2004) 19:905–13.[CrossRef][Web of Science][Medline]
- Wang Y, Levi CR, Attia JR, D'Este CA, Spratt N, Fisher J. Seasonal variation in stroke in the Hunter Region, Australia: a 5-year hospital-based study, 1995–2000. Stroke (2003) 34:1144–50.
[Abstract/Free Full Text] - Clayton TC, Thompson M, Meade TW. Recent respiratory infection and risk of cardiovascular disease: case-control study through a general practice database. Eur Heart J (2007).
- Danesh J. Antibiotics in the prevention of heart attacks. Lancet (2005) 365:365–7.[Web of Science][Medline]
- Taylor-Robinson D, Boman J. The failure of antibiotics to prevent heart attacks. BMJ (2005) 331:361–2.
[Free Full Text] - Hatano S. Experience from a multicentre stroke register: a preliminary report. Bull World Health Organ (1976) 54:541–53.[Web of Science][Medline]
- Wolfe CD, Rudd AG, Howard R, et al. Incidence and case fatality rates of stroke subtypes in a multiethnic population: the South London Stroke Register. J Neurol Neurosurg Psychiatry (2002) 72:211–6.
[Abstract/Free Full Text] - Höhle M. An R-package for the surveillance of infectious diseases. In: Proceedings in Computational Statistics, 17th Symposium—Rizzi A, Vichi M, eds. 28 August–1 September 2006, Rome, Italy. Heidelburg: Physica-Verlag, 2006; 1327– 34.
- Farrington CP, Andrews NJ, Beale AD, Catchpole MA. A statistical algorithm for the early detection of outbreaks of infectious disease. J R Stat Soc A (1996) 159:547–63.
- Cleveland RB, Cleveland WS, McRae JE, Terpenning I. STL: a seasonal-trend decomposition procedure based on loess. J Off Stat (1990) 6:3–73.
- Held L, Hohle M, Hofmann M. A statistical framework for the analysis of multivariate infectious disease surveillance counts. Stat Model (2005) 5:187–99.
- R Development Core Team. R: A Language and Environment for Statistical Computing (2006) Vienna, Austria: R Foundation for Statistical Computing.
- Smeeth L, Thomas SL, Hall AJ, Hubbard R, Farrington P, Vallance P. Risk of myocardial infarction and stroke after acute infection or vaccination. In: N Engl J Med (2004) 351:2611–8.
[Abstract/Free Full Text] - Lindsberg PJ, Grau AJ. Inflammation and infections as risk factors for ischemic stroke. Stroke (2003) 34:2518–32.
[Abstract/Free Full Text] - Crawford VL, McNerlan SE, Stout RW. Seasonal changes in platelets, fibrinogen and factor VII in elderly people. Age Ageing (2003) 32:661–5.
[Abstract/Free Full Text] - Stout RW, Crawford V. Seasonal variations in fibrinogen concentrations among elderly people. Lancet (1991) 338:9–13.[CrossRef][Web of Science][Medline]
- Stout RW, Crawford VL, McDermott MJ, Rocks MJ, Morris TC. Seasonal changes in haemostatic factors in young and elderly subjects. Age Ageing (1996) 25:256–8.
[Abstract/Free Full Text] - Crawford VL, Sweeney O, Coyle PV, Halliday IM, Stout RW. The relationship between elevated fibrinogen and markers of infection: a comparison of seasonal cycles. Q J Med (2000) 93:745–50.
- Elkind MS, Sciacca RR, Boden-Albala B, et al. Leukocyte count is associated with reduced endothelial reactivity. Atherosclerosis (2005) 181:329–38.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||