Association between Helicobacter pylori Infection and Stroke: A Meta-analysis of Prospective Observational Studies Min Yu, MS,* Yangbo Zhang, MS,† Zhen Yang, MD,* Jiangwu Ding, PhD,‡ Chuan Xie, MD,* and Nonghua Lu, PhD*
Background: Some studies have suggested an association between Helicobacter pylori infection and the risk of stroke, but the relationship remains controversial. The aim of this study was to obtain a more comprehensive estimate of H. pylori on the risk of stroke by performing a meta-analysis. Methods: A computerized search of PubMed, EMBASE, and the Cochrane library (including CENTRAL) up to February 2014 was performed to identify eligible studies. Prospective studies reported that a multivariate-adjusted estimate for the association between H. pylori and stroke were included. A random-effects model was used to calculate the overall combined risk. Results: Ten prospective observational studies (6 cohort studies, 4 nested case– control, or case–cohort studies within cohort studies) were included in the meta-analysis. The overall combined odds ratio for H. infection and stroke was .96 (95% confidence interval, .78-1.14). Similar results were yielded in patients with cytotoxin-associated gene-A seropositive strains. The combined estimates were robust across sensitivity analyses and had no observed publication bias. Conclusions: In conclusion, our formal meta-analysis indicated no strong association between H. pylori infection and stroke, neither in those with cytotoxin-associated gene-A–positive infection. We believe that future epidemiologic studies of H. pylori and stroke are unlikely to be fruitful. Key Words: Helicobacter pylori—stroke— prospective study—meta-analysis. Ó 2014 by National Stroke Association
Introduction From the *Department of Gastroenterology, The First Affiliated Hospital of Nanchang University, Nanchang; †Department of Neurology, The Second Affiliated Hospital of Nanchang University, Guangzhou; and ‡Department of Oncology, The Second Affiliated Hospital of Nanchang University, Nanchang, People’s Republic of China. Received March 12, 2014; accepted April 3, 2014. Conflicts of interest: None declared. N.-H.L. and M.Y. conceived and designed the experiments. M.Y., Z.Y., and C.X. analyzed the data. M.Y. and N.-H.L. wrote the article. M.Y., Y.-B.Z., and Z.Y. performed the literature search and the data extraction. All authors saw and approved the final version of the manuscript. Address correspondence to Nonghua Lu, Department of Gastroenterology, The First Affiliated Hospital of Nanchang University, 17 Yongwaizheng St, Nanchang, Jiangxi 330006, People’s Republic of China. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.04.020
Stroke is the second leading cause of death and serious disability worldwide.1 It causes a significant public health burden. According to estimates, about 795,000 Americans experience a new or recurrent stroke annually—on average, 1 stroke for every 40 individuals.2 The conventional risk factors, which manifest as hypertension, diabetes mellitus, smoking, obesity, serum lipids, and family history, do not fully account for all the cases, especially young subjects, and often do not have any of these factors.3 It is, therefore, important to identify additional, treatable causes of stroke that could lead to more effective prevention. There has been increasing evidence that, in addition to established risk factors, markers of inflammation and chronic infectious diseases, including Helicobacter pylori infection, may be linked to stroke and other ischemic disease.4-6 Helicobacter pylori is a gram-negative bacterium infecting 50% of the world’s population.7 However, in some
Journal of Stroke and Cerebrovascular Diseases, Vol. 23, No. 9 (October), 2014: pp 2233-2239
2233
M. YU ET AL.
2234
regions and countries of the world, more than 80% of the populations are infected.8 Helicobacter pylori infection results in chronic gastritis, ulcers, stomach cancer, and many extra-gastrointestinal conditions. It has been suggested that the production of excessive amounts of pro-inflammatory factors and cross-mimicry between H. pylori and host antigens may contribute to the development of gastric mucosal damage and extra-digestive manifestations associated with H. pylori infection.9-11 During the past decades, many epidemiologic studies have focused on the association between H. pylori infection and the risk of stroke.12-14 However, the role of H. pylori in stroke is still controversial.15-17 Previous meta-analyses have shown a significantly increased risk for stroke in H. pylori–infected individuals, especially in those infected with cytotoxin-associated gene A (CagA)–positive strains.18,19 But the data from these included by these meta-analyses were limited to case–control studies, in which selective bias and insufficient cofounding from risk factors remained an alternative explanation for significant association between H. pylori and stroke. Additionally, several recent population-based cohort studies yielded contrasting results.20-22 To obtain a more comprehensive estimate of the putative influence of the H. pylori on stroke, we conducted a meta-analysis of prospective studies to determine the association between H. pylori and stroke.
Materials and Methods Literature Search A systematic search of PubMed, EMBASE, and the Cochrane library (including CENTRAL) was performed to identify potentially relevant publications from the date of database origination through February 2014. The following key words were used in our search strategies: ‘‘helicobacter pylori,’’ ‘‘H. pylori,’’ ‘‘helicobacter,’’ or ‘‘hp’’ and ‘‘stroke,’’ ‘‘brain ischemic,’’ ‘‘transient brain ischemia,’’ ‘‘cerebral arterial accident,’’ ‘‘non-ischemic stroke,’’ ‘‘ischemic stroke,’’ ‘‘cerebrovascular accident,’’ ‘‘intracranial artery disease,’’ or ‘‘cardiovascular disease.’’ The search was restricted to human studies. No language restrictions were set. In addition, the reference lists of all retrieved articles, and in reviews and abstracts from recent conferences, were manually searched. When the same or similar patient cohort was included in several publications, only the most recent or informative report was selected for analysis.
Study Selection Studies were initially selected based on their titles and abstracts. Two reviewers (M.Y. and Y.-B.Z) independently screened all abstracts to determine whether the studies met the inclusion criteria. Differences were resolved by a third investigator (Z.Y.). Studies were considered
eligible if they met the following criteria: (1) the design was a prospective study; (2) the study population was patients with H. pylori infection; (3) reported quantitative estimates of the multivariate-adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for fatal or nonfatal stroke associated with H. pylori; hazard ratio (HR)/relative risk were considered equivalent to OR. The prospective nested case–control studies or case–cohort studies within cohort studies were also included for the meta-analysis because of their advantages comparable with cohort studies in verifying a causal relationship. Studies were excluded if (1) the study design was retrospective, (2) unadjusted estimates were reported, and (3) studied the association in patients with diabetes, nephropathy, etc.
Data Abstraction and Quality Assessment To ensure homogeneity in data gathering and entry, the data extraction was conducted by 2 experienced investigators working independently (M.Y. and Y.-B.Z). A third investigator (Z.Y.) was called on to resolve any differences, and complete consensus was reached for all the main variables to be assessed in the analysis. Data were recorded as follows: the first author’s last name (year of publication), country, study design, setting, size of the cohort, mean age, follow-up time, assessment of H. pylori and stroke, number of cases, cofounding covariates, and results. The Newcastle–Ottawa Scale was used to assess the methodologic quality of studies.23 In this scale, observational studies were scored across 3 categories: selection (4 questions) and comparability (2 questions) of study groups and ascertainment of the outcome of interest (3 questions). An item scored 1 if the study met the demands; otherwise, a score of 0 was given for that item. The final score ranged from 0 to 9 for each study (recorded in Table 1), with higher scores indicating better methodology. We defined studies of high or low quality based on the median overall score among all studies.
Statistical Analyses OR with 95% CI were used as a common measure of the association between H. pylori and risk of stroke across studies. Taking both within-study and between-study variabilities into account, we used a random-effects model to combine outcome data (expressed as ORs and 95% CIs).24 Heterogeneity was assessed by the I2 statistic, a quantitative measure of inconsistency across studies.25 To explore possible explanations for heterogeneity, we conducted stratified analyses by CagA situation, study design, setting, pathogen species studied, and study quality. We also assessed the influence of individual studies on the combined risk estimate by sequentially excluding 1 study in each turn to test the robustness of the main results. Potential publication bias was examined by Begg test and Egger test.26,27 A trim and fill analysis was performed to identify and correct asymmetry of the
Table 1. Main characteristics of included studies Study design
Jefferson et al31
United States
Cohort
Community based
1642
Schottker et al21
German
Cohort
Population based
9953
5.1
Chen et al28
United States
Cohort
Population based
9895
6
Smieja et al29
Canada
Cohort
Population based
3168
4.5
Elkind et al22
United States
Cohort
Community based
1625
8
Benjamin et al30
South Africa
Cohort
Clinic based
205
Coles et al15
Australia
Case– Population cohort based
Ikeda et al20
Japan
Whincup et al16
Britain
Heuschmann et al32 German
Setting
Study size
Followup (y) Helicobacter 10
Stroke
Serological Any type of stroke based on self-report and stroke death based on ICD-10 codes Serological Any type of stroke based on self-reported and physician-diagnosed and stroke death based on ICD-10 codes Serological Any type of stroke death based on ICD-9 or ICD-10 codes Serological Not reported
Stroke cases
Study quality
Age, sex, education, DM, AF, smoking, BMI, CHD/PAD, HTN, and hyperlipidemia Age, sex, education, BMI, smoking, alcohol, SBP, physical inactivity, CVD, DM, and TC
8
153
Age, sex, education, BMI, race, and smoking
6
107
Age, sex, smoking, ramipril, DM, HTN, and hypercholesterolemia Age, sex, race, education, SBP, HDL, LDL, BG, alcohol, smoking, WC, physical activity, and CAD Sex, age, smoking, alcohol intake, BMI, WC, SBP, DBP, FPG, and leukocytes Age, sex, BMI, SBP, TC, DM, Hb, triglycerides, treatment for HTN, and smoking Age, sex, BMI, date of blood sampling, time since last meal, and study location, alcohol, treatment for HTN, SBP, smoking status, and log-hsCRP Age, social class, smoking, SBP, BG, and heightstandardized FEV
5
164
241
Serological Any type of stroke based on physician diagnosed
67
10
Serological Not reported
36
1612
17
Case– Population control based
29,876
12
Serological Any type of fatal or nonfatal stroke based on ICD-10 codes Serological Any type of stroke based on ICD codes confirmed by medical records
Case– Population control based
7735
12
Case– Populationbased 100,330 control
1
Serological Any type of stroke based on physician diagnosed and stroke death based on ICD codes Serological Ischemic stroke based on World Health Organization criteria
Cofoundings
119
600
137
145
Age, sex, DM, HTN, smoking, residence, TC, CI, education, OR for CP
9
7
7
ASSOCIATION BETWEEN H. PYLORI INFECTION AND STROKE
Country
Reference
7
7
6
5
2235
Abbreviations: AF, atrial fibrillation; BG, blood glucose; BMI, body mass index; CAD, coronary artery disease; CVD, cardiovascular disease; CHD, coronary heart disease; CI, chronic inflammation; CP, Chlamydia pneumoniae; CRP, c-reactive protein; DM, diabetes mellitus; FEV, forced expiratory volume; FPG, fasting plasma glucose; DBP, diastolic blood pressure; Hb, hemoglobin; HR, heart rate; HDL, high-density lipoprotein; HTN, hypertension; ICD, International Classification of Disease; LDL, low-density lipoprotein; NA, not available; PAD, peripheral arterial disease; SBP, systolic blood pressure; TC, total cholesterol; TG, triglycerides; WC, waist circumference; OR, odds ratio.
M. YU ET AL.
2236
funnel figure caused by publication bias. All calculations were conducted by Stata, version 12.0 (Stata Corporation, College Station, TX). All P values were 2 sided, and the significance level was .05.
Results Identification of Eligible Studies Based on our search criteria, we identified 461 articles from Pubmed, EMBASE, and Cochrane library. Of these articles, most were excluded after reviewing titles and abstracts, mainly because they were reviews, letters, comments, retrospective studies, or not relevant to our analysis, leaving 49 that appeared to meet our selection criteria. Of those, 39 were excluded articles according to the inclusion of the present meta-analysis, among which 15 were excluded for irrelevant studies, 11 for retrospective case–control studies, 8 for unavailable data for analysis, 4 for diabetes or hematodialysis at study entry, and 1 for no adjusted estimate. Finally, a total of 10 prospective articles with 1769 cases of stroke and 166,041 participants were included in the meta-analysis.15,16,20-22,28-32 A detailed flowchart of the selection process was shown in Figure 1.
Study Characteristics The main characteristics of the 10 included studies were presented in Table 1. These studies were published between 1996 and 2013. Six were cohort studies,15,21,22,28-31 and 4 were case–control16,20,32 or case–cohort15 studies nested in prospective cohort studies. Among the 10 studies, 4 were conducted in American countries,22,28,29,31 3 from European 16,21,32 countries, and the remaining 3 from Asian, Africa, and Oceania countries, respectively.20,30,15 The number of participants ranged from 205 in the clinic-based study by Longo-Mbenza et al30 to 100,330 in the Erlangen Stroke Project study by Heuschmann et al.32 The follow-up duration ranged from 1 to 17 years. All studies assessed H. pylori infection by their use of seropositivity, whereas the definition of stroke varied each study. All the included studies adjusted important confounding factors: age, gender, diabetes, hypertension, smoking, social class, BMI, and hyperlipidemia. Besides, 5 of these studies were intended to examine only H. pylori agent and others more than 1. All studies provided adjusted risk estimates, and the overall quality of studies was good (range, 5-9).
Main Results A total of 13 comparisons (3 studies did have CagAspecific data) investigated the association between H. pylori infection and the risk of stroke. Pooling all 13 comparisons, H. pylori was not significantly associated with an increased risk of fatal and nonfatal stroke (OR, .96; 95% CI, .78-1.14; P 5 .000; Fig 2). Between-study heteroge-
Figure 1.
Identification of eligible studies from different databases.
neity was observed (I2 5 48%, P 5 .027), which did not alter much in the subgroup analysis. Figure 3 showed the pooled OR for stroke stratified by CagA situation, study design, setting, pathogen species studied, and methodologic quality. The OR for clinicbased study was 3.6 (95% CI, .2–7.0; P 5 .000). Increases in stroke events were also found in community-based, high-quality, more pathogen, and non–CagA-positive groups; however, the associations were not statistically significant.
Sensitivity Analysis and Publication Bias Heterogeneity disappeared when the study performed by Chen et al28 was deleted, indicating that the study was the main origin of heterogeneity. However, the results of the meta-analysis remained nonsignificant, indicating that our study was stable (data not shown). There was no statistical evidence of publication bias among studies by using both Begg test and Egger test (P 5 .85 and .68, respectively).
Discussion To our knowledge, this is the so far first meta-analysis that estimated the association of H. pylori infection with stroke through a combination of prospective studies. Contrast with previous meta-analyses based on case– control studies, we did not observe a positive association between H. pylori infection and the risk of stroke, neither in those with CagA-positive infection. The combined estimates were robust across sensitivity analyses and subgroup analyses and had no observed publication bias.
ASSOCIATION BETWEEN H. PYLORI INFECTION AND STROKE
Figure 2.
2237
Association between Helicobacter pylori and risk of stroke. Abbreviations: CI, confidence interval; OR, odds ratio.
Overwhelming evidence indicates that H. pylori infection is associated with the development of stroke. However, with respect to epidemiologic studies, most studies to date in supportive of H. pylori role in stroke were case– control or cross-sectional.12-14,33 These studies were hampered by their small size and consequent possibility of selection bias and insufficient adjustment for possible confounder variables as mentioned earlier. Furthermore, most were intended to examine only H. pylori agent; it was, therefore, likely that these studies may have been overpowered to be able to detect a possible significant difference between H. pylori infection and the risk of stroke, and the positive associations might be because of chance. The same limitations also applied to the meta-analyses based on these studies, so the association between stroke and H. pylori may be spurious, and the question of whether H. pylori infection is in fact associated with the development of stroke remains unclear. With respect to prospective studies, the literature on the association between H. pylori and stroke is not consistent, most with reports of positive or null associations. Interestingly, a more recent population-based German cohort of 9953 older participants showed a significant inverse association between CagA positivity and cardiovascular mor-
tality; and the inverse association in stroke was supported by a following prospective cohort analyses in a nationally representative sample of 9895 participants enrolled in the National Health and Nutrition Examination Survey III to evaluate H. pylori influence on total and category-specific mortality.21,28 Therefore, it is highly necessary to conduct a quantitative and systematic summary of the evidence using rigorous methods. In the present meta-analysis, in accordance with recently published meta-analyses on coronary artery disease, we found no association between either H. pylori or CagA and the risk of stroke. The role of chronic inflammation in atherogenesis has been studied extensively over the past decades. It has been repeatedly hypothesized that inflammation, immune-mediated vascular damage, direct bacterial invasion of atherosclerosis plaques, and hyperhomocysteinemia might be among the mechanisms contributing to stroke risk attributed to H. pylori.34-36 Therefore, drawing a positive link between H. pylori and stroke seems more plausible theoretically. In contrast, prospective studies have shown negative or less extreme results as confirmed by our meta-analysis. Several plausible mechanisms may explain the discrepancy in our meta-analysis. Maybe H. pylori can cause
2238
Figure 3. Analyses of subgroups relating Helicobacter pylori to stroke. Abbreviations: CI, confidence interval; OR, odds ratio.
a low-grade chronic inflammatory response and promote the development of atherosclerosis, but the clinical effect itself is not strong enough to lead to the final stroke incidence or death. Additionally, most of the postulated pathways by which infection with H. pylori may increase the risk of stroke are hypothesis after all and need to be confirmed or refuted by future research. Another plausible mechanism is that the possible link between H. pylori and stroke was limited to ischemic stroke caused by an atherosclerotic mechanism, whereas most studies covered patients with hemorrhagic stroke or cardioembolic stroke, which might underestimate or distort the association to some extent. However, information on subtypes of stroke was not available to further investigate the association in ischemic stroke. Therefore, future studies could be scheduled to assess the effect of H. pylori on ischemic stroke subtype only, but the association is likely to be modest at best because of the fact that ischemic stroke accounts for more than 70% of all strokes. Heterogeneity did not alter much in the subgroup analysis; however, it disappeared when the study performed by Chen et al28 was deleted, indicating that the study was the main source of heterogeneity. Stroke mortality did not parallel to its incidence. This study was intended to evaluate the influence of H. pylori on total and categoryspecific mortality from a global perspective; so, it was entirely possible to obtain a completely new conclusion. It found an inverse association of H. pylori and stroke mortality (HR .69, 95% CI, .44-1.18), and the inverse association was stronger for CagA-positive strains (HR .45, 95% CI, .27-.76). Although this inverse association might be counterintuitive on first view, one potential explanation might be a constant stimulation of regulatory T cells in subjects with H. pylori colonization, which may
M. YU ET AL.
be protective against stroke risk. In the long run, some protective effects of H. pylori infection are not impossible. Compared with the previous meta-analysis, our study has several more important strengths. A major one is its inclusion of prospective studies—the assessment of stroke risk factors and other potential confounders was carried out before the stroke events defining the cases and was considerably more detailed—they have more strength in adjusting the appropriate cofounding factors and providing stronger evidence of a causal relation. In addition, to our knowledge, this is the largest metaanalysis on H. pylori and the risk of stroke; with the accumulating evidence and large sample size of included studies, we have great statistical power to provide precise and reliable risk estimates. Furthermore, we assessed the quality of individual studies using the Newcastle–Ottawa Scale, all of them were of high quality, and no publication was detected. Therefore, the results should be more reliable, and we believe that further epidemiologic studies of H. pylori and stroke are unlikely to be fruitful. Despite these advantages, several limitations should be considered while interpreting the results of this study. First of all, given the observational nature of all included studies, one of the greatest limitations is that these studies could not establish or refute a causal relationship; thus, there might be still a need for large interventional randomized trials. Another limitation is that all our studies used seropositivity as an indicator of H. pylori infection; they could not accurately assess the actual condition of pathogen exposure. The serological positivity can persist despite the bacterial eradication or the fact that it could disappear with the presence of gastric atrophy, so patients who might have a lifelong seropositivity may be negative if studied by a direct method. Besides, eradication of the infection after baseline might also cause a decrease in the observed association between H. pylori and stroke; therefore, the results would be more reliable if the authors also showed H. pylori status at the onset of stroke. Third, it would be interesting to determine whether the association differed by stroke subtypes as mentioned earlier, but few data were available for a stratified analysis. Finally, the possible limitation is because of language bias. We attempted to minimize this bias by searching 3 major electronic databases with no language restriction. However, some articles published in Chinese or other non-English languages might not appear in international journal databases and could be missed by our searches. In conclusion, this is the so far first meta-analysis that estimated the association of H. pylori infection with stroke through a combination of prospective studies. Our formal meta-analysis indicated no strong association between H. pylori infection and stroke. We believe that future epidemiologic studies of H. pylori and stroke are unlikely to be fruitful.
ASSOCIATION BETWEEN H. PYLORI INFECTION AND STROKE
References 1. Donnan GA, Fisher M, Macleod M, et al. Stroke. Lancet 2008;371:1612-1623. 2. Roger VL, Go AS, Lloyd-Jones DM, et al. Heart disease and stroke statistics—2011 update: a report from the American Heart Association. Circulation 2011;123:e18-e209. 3. Helfand M, Buckley DI, Freeman M, et al. Emerging risk factors for coronary heart disease: a summary of systematic reviews conducted for the U.S. Preventive Services Task Force. Ann Intern Med 2009;151:496-507. 4. Manolakis A, Kapsoritakis AN, Potamianos SP. A review of the postulated mechanisms concerning the association of Helicobacter pylori with ischemic heart disease. Helicobacter 2007;12:287-297. 5. Kriszbacher I, Koppan M, Bodis J. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med 2005; 353:429-430. 6. McColl BW, Allan SM, Rothwell NJ. Systemic infection, inflammation and acute ischemic stroke. Neuroscience 2009;158:1049-1061. 7. Bodger K, Crabtree JE. Helicobacter pylori and gastric inflammation. Br Med Bull 1998;54:139-150. 8. Porras C, Nodora J, Sexton R, et al. Epidemiology of Helicobacter pylori infection in six Latin American countries (SWOG Trial S0701). Cancer Causes Control 2013; 24:209-215. 9. Roubaud BC, Franceschi F, Salles N, et al. Extragastric diseases and Helicobacter pylori. Helicobacter 2013; 18(Suppl 1):44-51. 10. Genta RM. Helicobacter pylori, inflammation, mucosal damage, and apoptosis: pathogenesis and definition of gastric atrophy. Gastroenterology 1997;113(6 Suppl):S51-S55. 11. Liao W. Endotoxin: possible roles in initiation and development of atherosclerosis. J Lab Clin Med 1996;128:452-460. 12. De Bastiani R, Gabrielli M, Ubaldi E, et al. High prevalence of Cag-A positive H. pylori strains in ischemic stroke: a primary care multicenter study. Helicobacter 2008;13:274-277. 13. Sawayama Y, Ariyama I, Hamada M, et al. Association between chronic Helicobacter pylori infection and acute ischemic stroke: Fukuoka Harasanshin Atherosclerosis Trial (FHAT). Atherosclerosis 2005;178:303-309. 14. Pietroiusti A, Diomedi M, Silvestrini M, et al. Cytotoxinassociated gene-A–positive Helicobacter pylori strains are associated with atherosclerotic stroke. Circulation 2002;106:580-584. 15. Coles KA, Knuiman MW, Plant AJ, et al. A prospective study of infection and cardiovascular diseases: the Busselton Health Study. Eur J Cardiovasc Prev Rehabil 2003;10:278-282. 16. Whincup PH, Mendall MA, Perry IJ, et al. Prospective relations between Helicobacter pylori infection, coronary heart disease, and stroke in middle aged men. Heart 1996;75:568-572. 17. Tan HJ, Goh KL. Extragastrointestinal manifestations of Helicobacter pylori infection: facts or myth? A critical review. J Dig Dis 2012;13:342-349. 18. Zhang S, Guo Y, Ma Y, et al. Cytotoxin-associated gene-Aseropositive virulent strains of Helicobacter pylori and atherosclerotic diseases: a systematic review. Chin Med J (Engl) 2008;121:946-951.
2239
19. Wang ZW, Li Y, Huang LY, et al. Helicobacter pylori infection contributes to high risk of ischemic stroke: evidence from a meta-analysis. J Neurol 2012;259:2527-2537. 20. Ikeda A, Iso H, Sasazuki S, et al. The combination of Helicobacter pylori- and cytotoxin-associated gene-A seropositivity in relation to the risk of myocardial infarction in middle-aged Japanese: The Japan Public Health Center-based study. Atherosclerosis 2013;230:67-72. 21. Schottker B, Adamu MA, Weck MN, et al. Helicobacter pylori infection, chronic atrophic gastritis and major cardiovascular events: a population-based cohort study. Atherosclerosis 2012;220:569-574. 22. Elkind MS, Ramakrishnan P, Moon YP, et al. Infectious burden and risk of stroke: the northern Manhattan study. Arch Neurol 2010;67:33-38. 23. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010;25:603-605. 24. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177-188. 25. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ 2003;327:557-560. 26. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50:1088-1101. 27. Egger M, Davey SG, Schneider M, et al. Bias in metaanalysis detected by a simple, graphical test. BMJ 1997; 315:629-634. 28. Chen Y, Segers S, Blaser MJ. Association between Helicobacter pylori and mortality in the NHANES III study. Gut 2013;62:1262-1269. 29. Smieja M, Gnarpe J, Lonn E, et al. Multiple infections and subsequent cardiovascular events in the Heart Outcomes Prevention Evaluation (HOPE) Study. Circulation 2003; 107:251-257. 30. Longo-Mbenza B, Nsenga JN, Mokondjimobe E, et al. Helicobacter pylori infection is identified as a cardiovascular risk factor in Central Africans. Vasc Health Risk Manag 2012;6:455-461. 31. Sealy-Jefferson S, Gillespie BW, Aiello AE, et al. Antibody levels to persistent pathogens and incident stroke in Mexican Americans. PLoS One 2013;8:e65959. 32. Heuschmann PU, Neureiter D, Gesslein M, et al. Association between infection with Helicobacter pylori and Chlamydia pneumoniae and risk of ischemic stroke subtypes: results from a population-based case-control study. Stroke 2001;32:2253-2258. 33. Gabrielli M, Santoliquido A, Cremonini F, et al. CagApositive cytotoxic H. pylori strains as a link between plaque instability and atherosclerotic stroke. Eur Heart J 2004;25:64-68. 34. Nilsson HO, Pietroiusti A, Gabrielli M, et al. Helicobacter pylori and extragastric diseases—other Helicobacters. Helicobacter 2005;10(Suppl 1):54-65. 35. Laurila A, Bloigu A, Nayha S, et al. Association of Helicobacter pylori infection with elevated serum lipids. Atherosclerosis 1999;142:207-210. 36. Hoffmeister A, Rothenbacher D, Bode G, et al. Current infection with Helicobacter pylori, but not seropositivity to Chlamydia pneumoniae or cytomegalovirus, is associated with an atherogenic, modified lipid profile. Arterioscler Thromb Vasc Biol 2001;21:427-432.
Background: Some studies have suggested an association between Helicobacter pylori infection and the risk of stroke, but the relationship remains controversial. The aim of this study was to obtain a more comprehensive estimate of H. pylori on the risk of stroke by performing a meta-analysis. Methods: A computerized search of PubMed, EMBASE, and the Cochrane library (including CENTRAL) up to February 2014 was performed to identify eligible studies. Prospective studies reported that a multivariate-adjusted estimate for the association between H. pylori and stroke were included. A random-effects model was used to calculate the overall combined risk. Results: Ten prospective observational studies (6 cohort studies, 4 nested case– control, or case–cohort studies within cohort studies) were included in the meta-analysis. The overall combined odds ratio for H. infection and stroke was .96 (95% confidence interval, .78-1.14). Similar results were yielded in patients with cytotoxin-associated gene-A seropositive strains. The combined estimates were robust across sensitivity analyses and had no observed publication bias. Conclusions: In conclusion, our formal meta-analysis indicated no strong association between H. pylori infection and stroke, neither in those with cytotoxin-associated gene-A–positive infection. We believe that future epidemiologic studies of H. pylori and stroke are unlikely to be fruitful. Key Words: Helicobacter pylori—stroke— prospective study—meta-analysis. Ó 2014 by National Stroke Association
Introduction From the *Department of Gastroenterology, The First Affiliated Hospital of Nanchang University, Nanchang; †Department of Neurology, The Second Affiliated Hospital of Nanchang University, Guangzhou; and ‡Department of Oncology, The Second Affiliated Hospital of Nanchang University, Nanchang, People’s Republic of China. Received March 12, 2014; accepted April 3, 2014. Conflicts of interest: None declared. N.-H.L. and M.Y. conceived and designed the experiments. M.Y., Z.Y., and C.X. analyzed the data. M.Y. and N.-H.L. wrote the article. M.Y., Y.-B.Z., and Z.Y. performed the literature search and the data extraction. All authors saw and approved the final version of the manuscript. Address correspondence to Nonghua Lu, Department of Gastroenterology, The First Affiliated Hospital of Nanchang University, 17 Yongwaizheng St, Nanchang, Jiangxi 330006, People’s Republic of China. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.04.020
Stroke is the second leading cause of death and serious disability worldwide.1 It causes a significant public health burden. According to estimates, about 795,000 Americans experience a new or recurrent stroke annually—on average, 1 stroke for every 40 individuals.2 The conventional risk factors, which manifest as hypertension, diabetes mellitus, smoking, obesity, serum lipids, and family history, do not fully account for all the cases, especially young subjects, and often do not have any of these factors.3 It is, therefore, important to identify additional, treatable causes of stroke that could lead to more effective prevention. There has been increasing evidence that, in addition to established risk factors, markers of inflammation and chronic infectious diseases, including Helicobacter pylori infection, may be linked to stroke and other ischemic disease.4-6 Helicobacter pylori is a gram-negative bacterium infecting 50% of the world’s population.7 However, in some
Journal of Stroke and Cerebrovascular Diseases, Vol. 23, No. 9 (October), 2014: pp 2233-2239
2233
M. YU ET AL.
2234
regions and countries of the world, more than 80% of the populations are infected.8 Helicobacter pylori infection results in chronic gastritis, ulcers, stomach cancer, and many extra-gastrointestinal conditions. It has been suggested that the production of excessive amounts of pro-inflammatory factors and cross-mimicry between H. pylori and host antigens may contribute to the development of gastric mucosal damage and extra-digestive manifestations associated with H. pylori infection.9-11 During the past decades, many epidemiologic studies have focused on the association between H. pylori infection and the risk of stroke.12-14 However, the role of H. pylori in stroke is still controversial.15-17 Previous meta-analyses have shown a significantly increased risk for stroke in H. pylori–infected individuals, especially in those infected with cytotoxin-associated gene A (CagA)–positive strains.18,19 But the data from these included by these meta-analyses were limited to case–control studies, in which selective bias and insufficient cofounding from risk factors remained an alternative explanation for significant association between H. pylori and stroke. Additionally, several recent population-based cohort studies yielded contrasting results.20-22 To obtain a more comprehensive estimate of the putative influence of the H. pylori on stroke, we conducted a meta-analysis of prospective studies to determine the association between H. pylori and stroke.
Materials and Methods Literature Search A systematic search of PubMed, EMBASE, and the Cochrane library (including CENTRAL) was performed to identify potentially relevant publications from the date of database origination through February 2014. The following key words were used in our search strategies: ‘‘helicobacter pylori,’’ ‘‘H. pylori,’’ ‘‘helicobacter,’’ or ‘‘hp’’ and ‘‘stroke,’’ ‘‘brain ischemic,’’ ‘‘transient brain ischemia,’’ ‘‘cerebral arterial accident,’’ ‘‘non-ischemic stroke,’’ ‘‘ischemic stroke,’’ ‘‘cerebrovascular accident,’’ ‘‘intracranial artery disease,’’ or ‘‘cardiovascular disease.’’ The search was restricted to human studies. No language restrictions were set. In addition, the reference lists of all retrieved articles, and in reviews and abstracts from recent conferences, were manually searched. When the same or similar patient cohort was included in several publications, only the most recent or informative report was selected for analysis.
Study Selection Studies were initially selected based on their titles and abstracts. Two reviewers (M.Y. and Y.-B.Z) independently screened all abstracts to determine whether the studies met the inclusion criteria. Differences were resolved by a third investigator (Z.Y.). Studies were considered
eligible if they met the following criteria: (1) the design was a prospective study; (2) the study population was patients with H. pylori infection; (3) reported quantitative estimates of the multivariate-adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for fatal or nonfatal stroke associated with H. pylori; hazard ratio (HR)/relative risk were considered equivalent to OR. The prospective nested case–control studies or case–cohort studies within cohort studies were also included for the meta-analysis because of their advantages comparable with cohort studies in verifying a causal relationship. Studies were excluded if (1) the study design was retrospective, (2) unadjusted estimates were reported, and (3) studied the association in patients with diabetes, nephropathy, etc.
Data Abstraction and Quality Assessment To ensure homogeneity in data gathering and entry, the data extraction was conducted by 2 experienced investigators working independently (M.Y. and Y.-B.Z). A third investigator (Z.Y.) was called on to resolve any differences, and complete consensus was reached for all the main variables to be assessed in the analysis. Data were recorded as follows: the first author’s last name (year of publication), country, study design, setting, size of the cohort, mean age, follow-up time, assessment of H. pylori and stroke, number of cases, cofounding covariates, and results. The Newcastle–Ottawa Scale was used to assess the methodologic quality of studies.23 In this scale, observational studies were scored across 3 categories: selection (4 questions) and comparability (2 questions) of study groups and ascertainment of the outcome of interest (3 questions). An item scored 1 if the study met the demands; otherwise, a score of 0 was given for that item. The final score ranged from 0 to 9 for each study (recorded in Table 1), with higher scores indicating better methodology. We defined studies of high or low quality based on the median overall score among all studies.
Statistical Analyses OR with 95% CI were used as a common measure of the association between H. pylori and risk of stroke across studies. Taking both within-study and between-study variabilities into account, we used a random-effects model to combine outcome data (expressed as ORs and 95% CIs).24 Heterogeneity was assessed by the I2 statistic, a quantitative measure of inconsistency across studies.25 To explore possible explanations for heterogeneity, we conducted stratified analyses by CagA situation, study design, setting, pathogen species studied, and study quality. We also assessed the influence of individual studies on the combined risk estimate by sequentially excluding 1 study in each turn to test the robustness of the main results. Potential publication bias was examined by Begg test and Egger test.26,27 A trim and fill analysis was performed to identify and correct asymmetry of the
Table 1. Main characteristics of included studies Study design
Jefferson et al31
United States
Cohort
Community based
1642
Schottker et al21
German
Cohort
Population based
9953
5.1
Chen et al28
United States
Cohort
Population based
9895
6
Smieja et al29
Canada
Cohort
Population based
3168
4.5
Elkind et al22
United States
Cohort
Community based
1625
8
Benjamin et al30
South Africa
Cohort
Clinic based
205
Coles et al15
Australia
Case– Population cohort based
Ikeda et al20
Japan
Whincup et al16
Britain
Heuschmann et al32 German
Setting
Study size
Followup (y) Helicobacter 10
Stroke
Serological Any type of stroke based on self-report and stroke death based on ICD-10 codes Serological Any type of stroke based on self-reported and physician-diagnosed and stroke death based on ICD-10 codes Serological Any type of stroke death based on ICD-9 or ICD-10 codes Serological Not reported
Stroke cases
Study quality
Age, sex, education, DM, AF, smoking, BMI, CHD/PAD, HTN, and hyperlipidemia Age, sex, education, BMI, smoking, alcohol, SBP, physical inactivity, CVD, DM, and TC
8
153
Age, sex, education, BMI, race, and smoking
6
107
Age, sex, smoking, ramipril, DM, HTN, and hypercholesterolemia Age, sex, race, education, SBP, HDL, LDL, BG, alcohol, smoking, WC, physical activity, and CAD Sex, age, smoking, alcohol intake, BMI, WC, SBP, DBP, FPG, and leukocytes Age, sex, BMI, SBP, TC, DM, Hb, triglycerides, treatment for HTN, and smoking Age, sex, BMI, date of blood sampling, time since last meal, and study location, alcohol, treatment for HTN, SBP, smoking status, and log-hsCRP Age, social class, smoking, SBP, BG, and heightstandardized FEV
5
164
241
Serological Any type of stroke based on physician diagnosed
67
10
Serological Not reported
36
1612
17
Case– Population control based
29,876
12
Serological Any type of fatal or nonfatal stroke based on ICD-10 codes Serological Any type of stroke based on ICD codes confirmed by medical records
Case– Population control based
7735
12
Case– Populationbased 100,330 control
1
Serological Any type of stroke based on physician diagnosed and stroke death based on ICD codes Serological Ischemic stroke based on World Health Organization criteria
Cofoundings
119
600
137
145
Age, sex, DM, HTN, smoking, residence, TC, CI, education, OR for CP
9
7
7
ASSOCIATION BETWEEN H. PYLORI INFECTION AND STROKE
Country
Reference
7
7
6
5
2235
Abbreviations: AF, atrial fibrillation; BG, blood glucose; BMI, body mass index; CAD, coronary artery disease; CVD, cardiovascular disease; CHD, coronary heart disease; CI, chronic inflammation; CP, Chlamydia pneumoniae; CRP, c-reactive protein; DM, diabetes mellitus; FEV, forced expiratory volume; FPG, fasting plasma glucose; DBP, diastolic blood pressure; Hb, hemoglobin; HR, heart rate; HDL, high-density lipoprotein; HTN, hypertension; ICD, International Classification of Disease; LDL, low-density lipoprotein; NA, not available; PAD, peripheral arterial disease; SBP, systolic blood pressure; TC, total cholesterol; TG, triglycerides; WC, waist circumference; OR, odds ratio.
M. YU ET AL.
2236
funnel figure caused by publication bias. All calculations were conducted by Stata, version 12.0 (Stata Corporation, College Station, TX). All P values were 2 sided, and the significance level was .05.
Results Identification of Eligible Studies Based on our search criteria, we identified 461 articles from Pubmed, EMBASE, and Cochrane library. Of these articles, most were excluded after reviewing titles and abstracts, mainly because they were reviews, letters, comments, retrospective studies, or not relevant to our analysis, leaving 49 that appeared to meet our selection criteria. Of those, 39 were excluded articles according to the inclusion of the present meta-analysis, among which 15 were excluded for irrelevant studies, 11 for retrospective case–control studies, 8 for unavailable data for analysis, 4 for diabetes or hematodialysis at study entry, and 1 for no adjusted estimate. Finally, a total of 10 prospective articles with 1769 cases of stroke and 166,041 participants were included in the meta-analysis.15,16,20-22,28-32 A detailed flowchart of the selection process was shown in Figure 1.
Study Characteristics The main characteristics of the 10 included studies were presented in Table 1. These studies were published between 1996 and 2013. Six were cohort studies,15,21,22,28-31 and 4 were case–control16,20,32 or case–cohort15 studies nested in prospective cohort studies. Among the 10 studies, 4 were conducted in American countries,22,28,29,31 3 from European 16,21,32 countries, and the remaining 3 from Asian, Africa, and Oceania countries, respectively.20,30,15 The number of participants ranged from 205 in the clinic-based study by Longo-Mbenza et al30 to 100,330 in the Erlangen Stroke Project study by Heuschmann et al.32 The follow-up duration ranged from 1 to 17 years. All studies assessed H. pylori infection by their use of seropositivity, whereas the definition of stroke varied each study. All the included studies adjusted important confounding factors: age, gender, diabetes, hypertension, smoking, social class, BMI, and hyperlipidemia. Besides, 5 of these studies were intended to examine only H. pylori agent and others more than 1. All studies provided adjusted risk estimates, and the overall quality of studies was good (range, 5-9).
Main Results A total of 13 comparisons (3 studies did have CagAspecific data) investigated the association between H. pylori infection and the risk of stroke. Pooling all 13 comparisons, H. pylori was not significantly associated with an increased risk of fatal and nonfatal stroke (OR, .96; 95% CI, .78-1.14; P 5 .000; Fig 2). Between-study heteroge-
Figure 1.
Identification of eligible studies from different databases.
neity was observed (I2 5 48%, P 5 .027), which did not alter much in the subgroup analysis. Figure 3 showed the pooled OR for stroke stratified by CagA situation, study design, setting, pathogen species studied, and methodologic quality. The OR for clinicbased study was 3.6 (95% CI, .2–7.0; P 5 .000). Increases in stroke events were also found in community-based, high-quality, more pathogen, and non–CagA-positive groups; however, the associations were not statistically significant.
Sensitivity Analysis and Publication Bias Heterogeneity disappeared when the study performed by Chen et al28 was deleted, indicating that the study was the main origin of heterogeneity. However, the results of the meta-analysis remained nonsignificant, indicating that our study was stable (data not shown). There was no statistical evidence of publication bias among studies by using both Begg test and Egger test (P 5 .85 and .68, respectively).
Discussion To our knowledge, this is the so far first meta-analysis that estimated the association of H. pylori infection with stroke through a combination of prospective studies. Contrast with previous meta-analyses based on case– control studies, we did not observe a positive association between H. pylori infection and the risk of stroke, neither in those with CagA-positive infection. The combined estimates were robust across sensitivity analyses and subgroup analyses and had no observed publication bias.
ASSOCIATION BETWEEN H. PYLORI INFECTION AND STROKE
Figure 2.
2237
Association between Helicobacter pylori and risk of stroke. Abbreviations: CI, confidence interval; OR, odds ratio.
Overwhelming evidence indicates that H. pylori infection is associated with the development of stroke. However, with respect to epidemiologic studies, most studies to date in supportive of H. pylori role in stroke were case– control or cross-sectional.12-14,33 These studies were hampered by their small size and consequent possibility of selection bias and insufficient adjustment for possible confounder variables as mentioned earlier. Furthermore, most were intended to examine only H. pylori agent; it was, therefore, likely that these studies may have been overpowered to be able to detect a possible significant difference between H. pylori infection and the risk of stroke, and the positive associations might be because of chance. The same limitations also applied to the meta-analyses based on these studies, so the association between stroke and H. pylori may be spurious, and the question of whether H. pylori infection is in fact associated with the development of stroke remains unclear. With respect to prospective studies, the literature on the association between H. pylori and stroke is not consistent, most with reports of positive or null associations. Interestingly, a more recent population-based German cohort of 9953 older participants showed a significant inverse association between CagA positivity and cardiovascular mor-
tality; and the inverse association in stroke was supported by a following prospective cohort analyses in a nationally representative sample of 9895 participants enrolled in the National Health and Nutrition Examination Survey III to evaluate H. pylori influence on total and category-specific mortality.21,28 Therefore, it is highly necessary to conduct a quantitative and systematic summary of the evidence using rigorous methods. In the present meta-analysis, in accordance with recently published meta-analyses on coronary artery disease, we found no association between either H. pylori or CagA and the risk of stroke. The role of chronic inflammation in atherogenesis has been studied extensively over the past decades. It has been repeatedly hypothesized that inflammation, immune-mediated vascular damage, direct bacterial invasion of atherosclerosis plaques, and hyperhomocysteinemia might be among the mechanisms contributing to stroke risk attributed to H. pylori.34-36 Therefore, drawing a positive link between H. pylori and stroke seems more plausible theoretically. In contrast, prospective studies have shown negative or less extreme results as confirmed by our meta-analysis. Several plausible mechanisms may explain the discrepancy in our meta-analysis. Maybe H. pylori can cause
2238
Figure 3. Analyses of subgroups relating Helicobacter pylori to stroke. Abbreviations: CI, confidence interval; OR, odds ratio.
a low-grade chronic inflammatory response and promote the development of atherosclerosis, but the clinical effect itself is not strong enough to lead to the final stroke incidence or death. Additionally, most of the postulated pathways by which infection with H. pylori may increase the risk of stroke are hypothesis after all and need to be confirmed or refuted by future research. Another plausible mechanism is that the possible link between H. pylori and stroke was limited to ischemic stroke caused by an atherosclerotic mechanism, whereas most studies covered patients with hemorrhagic stroke or cardioembolic stroke, which might underestimate or distort the association to some extent. However, information on subtypes of stroke was not available to further investigate the association in ischemic stroke. Therefore, future studies could be scheduled to assess the effect of H. pylori on ischemic stroke subtype only, but the association is likely to be modest at best because of the fact that ischemic stroke accounts for more than 70% of all strokes. Heterogeneity did not alter much in the subgroup analysis; however, it disappeared when the study performed by Chen et al28 was deleted, indicating that the study was the main source of heterogeneity. Stroke mortality did not parallel to its incidence. This study was intended to evaluate the influence of H. pylori on total and categoryspecific mortality from a global perspective; so, it was entirely possible to obtain a completely new conclusion. It found an inverse association of H. pylori and stroke mortality (HR .69, 95% CI, .44-1.18), and the inverse association was stronger for CagA-positive strains (HR .45, 95% CI, .27-.76). Although this inverse association might be counterintuitive on first view, one potential explanation might be a constant stimulation of regulatory T cells in subjects with H. pylori colonization, which may
M. YU ET AL.
be protective against stroke risk. In the long run, some protective effects of H. pylori infection are not impossible. Compared with the previous meta-analysis, our study has several more important strengths. A major one is its inclusion of prospective studies—the assessment of stroke risk factors and other potential confounders was carried out before the stroke events defining the cases and was considerably more detailed—they have more strength in adjusting the appropriate cofounding factors and providing stronger evidence of a causal relation. In addition, to our knowledge, this is the largest metaanalysis on H. pylori and the risk of stroke; with the accumulating evidence and large sample size of included studies, we have great statistical power to provide precise and reliable risk estimates. Furthermore, we assessed the quality of individual studies using the Newcastle–Ottawa Scale, all of them were of high quality, and no publication was detected. Therefore, the results should be more reliable, and we believe that further epidemiologic studies of H. pylori and stroke are unlikely to be fruitful. Despite these advantages, several limitations should be considered while interpreting the results of this study. First of all, given the observational nature of all included studies, one of the greatest limitations is that these studies could not establish or refute a causal relationship; thus, there might be still a need for large interventional randomized trials. Another limitation is that all our studies used seropositivity as an indicator of H. pylori infection; they could not accurately assess the actual condition of pathogen exposure. The serological positivity can persist despite the bacterial eradication or the fact that it could disappear with the presence of gastric atrophy, so patients who might have a lifelong seropositivity may be negative if studied by a direct method. Besides, eradication of the infection after baseline might also cause a decrease in the observed association between H. pylori and stroke; therefore, the results would be more reliable if the authors also showed H. pylori status at the onset of stroke. Third, it would be interesting to determine whether the association differed by stroke subtypes as mentioned earlier, but few data were available for a stratified analysis. Finally, the possible limitation is because of language bias. We attempted to minimize this bias by searching 3 major electronic databases with no language restriction. However, some articles published in Chinese or other non-English languages might not appear in international journal databases and could be missed by our searches. In conclusion, this is the so far first meta-analysis that estimated the association of H. pylori infection with stroke through a combination of prospective studies. Our formal meta-analysis indicated no strong association between H. pylori infection and stroke. We believe that future epidemiologic studies of H. pylori and stroke are unlikely to be fruitful.
ASSOCIATION BETWEEN H. PYLORI INFECTION AND STROKE
References 1. Donnan GA, Fisher M, Macleod M, et al. Stroke. Lancet 2008;371:1612-1623. 2. Roger VL, Go AS, Lloyd-Jones DM, et al. Heart disease and stroke statistics—2011 update: a report from the American Heart Association. Circulation 2011;123:e18-e209. 3. Helfand M, Buckley DI, Freeman M, et al. Emerging risk factors for coronary heart disease: a summary of systematic reviews conducted for the U.S. Preventive Services Task Force. Ann Intern Med 2009;151:496-507. 4. Manolakis A, Kapsoritakis AN, Potamianos SP. A review of the postulated mechanisms concerning the association of Helicobacter pylori with ischemic heart disease. Helicobacter 2007;12:287-297. 5. Kriszbacher I, Koppan M, Bodis J. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med 2005; 353:429-430. 6. McColl BW, Allan SM, Rothwell NJ. Systemic infection, inflammation and acute ischemic stroke. Neuroscience 2009;158:1049-1061. 7. Bodger K, Crabtree JE. Helicobacter pylori and gastric inflammation. Br Med Bull 1998;54:139-150. 8. Porras C, Nodora J, Sexton R, et al. Epidemiology of Helicobacter pylori infection in six Latin American countries (SWOG Trial S0701). Cancer Causes Control 2013; 24:209-215. 9. Roubaud BC, Franceschi F, Salles N, et al. Extragastric diseases and Helicobacter pylori. Helicobacter 2013; 18(Suppl 1):44-51. 10. Genta RM. Helicobacter pylori, inflammation, mucosal damage, and apoptosis: pathogenesis and definition of gastric atrophy. Gastroenterology 1997;113(6 Suppl):S51-S55. 11. Liao W. Endotoxin: possible roles in initiation and development of atherosclerosis. J Lab Clin Med 1996;128:452-460. 12. De Bastiani R, Gabrielli M, Ubaldi E, et al. High prevalence of Cag-A positive H. pylori strains in ischemic stroke: a primary care multicenter study. Helicobacter 2008;13:274-277. 13. Sawayama Y, Ariyama I, Hamada M, et al. Association between chronic Helicobacter pylori infection and acute ischemic stroke: Fukuoka Harasanshin Atherosclerosis Trial (FHAT). Atherosclerosis 2005;178:303-309. 14. Pietroiusti A, Diomedi M, Silvestrini M, et al. Cytotoxinassociated gene-A–positive Helicobacter pylori strains are associated with atherosclerotic stroke. Circulation 2002;106:580-584. 15. Coles KA, Knuiman MW, Plant AJ, et al. A prospective study of infection and cardiovascular diseases: the Busselton Health Study. Eur J Cardiovasc Prev Rehabil 2003;10:278-282. 16. Whincup PH, Mendall MA, Perry IJ, et al. Prospective relations between Helicobacter pylori infection, coronary heart disease, and stroke in middle aged men. Heart 1996;75:568-572. 17. Tan HJ, Goh KL. Extragastrointestinal manifestations of Helicobacter pylori infection: facts or myth? A critical review. J Dig Dis 2012;13:342-349. 18. Zhang S, Guo Y, Ma Y, et al. Cytotoxin-associated gene-Aseropositive virulent strains of Helicobacter pylori and atherosclerotic diseases: a systematic review. Chin Med J (Engl) 2008;121:946-951.
2239
19. Wang ZW, Li Y, Huang LY, et al. Helicobacter pylori infection contributes to high risk of ischemic stroke: evidence from a meta-analysis. J Neurol 2012;259:2527-2537. 20. Ikeda A, Iso H, Sasazuki S, et al. The combination of Helicobacter pylori- and cytotoxin-associated gene-A seropositivity in relation to the risk of myocardial infarction in middle-aged Japanese: The Japan Public Health Center-based study. Atherosclerosis 2013;230:67-72. 21. Schottker B, Adamu MA, Weck MN, et al. Helicobacter pylori infection, chronic atrophic gastritis and major cardiovascular events: a population-based cohort study. Atherosclerosis 2012;220:569-574. 22. Elkind MS, Ramakrishnan P, Moon YP, et al. Infectious burden and risk of stroke: the northern Manhattan study. Arch Neurol 2010;67:33-38. 23. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010;25:603-605. 24. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177-188. 25. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ 2003;327:557-560. 26. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50:1088-1101. 27. Egger M, Davey SG, Schneider M, et al. Bias in metaanalysis detected by a simple, graphical test. BMJ 1997; 315:629-634. 28. Chen Y, Segers S, Blaser MJ. Association between Helicobacter pylori and mortality in the NHANES III study. Gut 2013;62:1262-1269. 29. Smieja M, Gnarpe J, Lonn E, et al. Multiple infections and subsequent cardiovascular events in the Heart Outcomes Prevention Evaluation (HOPE) Study. Circulation 2003; 107:251-257. 30. Longo-Mbenza B, Nsenga JN, Mokondjimobe E, et al. Helicobacter pylori infection is identified as a cardiovascular risk factor in Central Africans. Vasc Health Risk Manag 2012;6:455-461. 31. Sealy-Jefferson S, Gillespie BW, Aiello AE, et al. Antibody levels to persistent pathogens and incident stroke in Mexican Americans. PLoS One 2013;8:e65959. 32. Heuschmann PU, Neureiter D, Gesslein M, et al. Association between infection with Helicobacter pylori and Chlamydia pneumoniae and risk of ischemic stroke subtypes: results from a population-based case-control study. Stroke 2001;32:2253-2258. 33. Gabrielli M, Santoliquido A, Cremonini F, et al. CagApositive cytotoxic H. pylori strains as a link between plaque instability and atherosclerotic stroke. Eur Heart J 2004;25:64-68. 34. Nilsson HO, Pietroiusti A, Gabrielli M, et al. Helicobacter pylori and extragastric diseases—other Helicobacters. Helicobacter 2005;10(Suppl 1):54-65. 35. Laurila A, Bloigu A, Nayha S, et al. Association of Helicobacter pylori infection with elevated serum lipids. Atherosclerosis 1999;142:207-210. 36. Hoffmeister A, Rothenbacher D, Bode G, et al. Current infection with Helicobacter pylori, but not seropositivity to Chlamydia pneumoniae or cytomegalovirus, is associated with an atherogenic, modified lipid profile. Arterioscler Thromb Vasc Biol 2001;21:427-432.
