Has Stroke Shifted from Hemorrhagic to Ischemic in China?

Gelin Xu,Xinfeng Liu

DOI: https://doi.org/10.1161/01.str.0000210186.07530.69

IF: 10.17

2006-01-01

Stroke

Abstract:HomeStrokeVol. 37, No. 4Has Stroke Shifted From Hemorrhagic to Ischemic in China? Free AccessLetterPDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessLetterPDF/EPUBHas Stroke Shifted From Hemorrhagic to Ischemic in China? Gelin Xu, MD and Xinfeng Liu, MD Gelin XuGelin Xu Department of Neurology, Jinling Hospital, Nanjing University School of Medicine, Nanjing, Jiangsu Province, China and Xinfeng LiuXinfeng Liu Department of Neurology, Jinling Hospital, Nanjing University School of Medicine, Nanjing, Jiangsu Province, China Originally published1 Apr 2006https://doi.org/10.1161/01.STR.0000210186.07530.69Stroke. 2006;37:941–942To the Editor:In China, which has the largest population and greatest number of stroke victims in the world, data concerning stroke epidemiology are very limited. Guidelines of stroke prevention and management designed for Chinese were tailored, to a considerable degree, according to the epidemiological, etiological and clinical profiles mirrored from western populations. The higher morbidity and mortality of stroke in Chinese further bulks the need of nationwide epidemiological data for establishing population specialized strategies for stoke prevention and management.We read with great interest the study by Jiang et al concerning the stroke incidence and trends in 3 large cities in China.1 It was shown that incidence of intracranial hemorrhage (ICH) decreased significantly at an annual rate of 12.0% in Beijing, 4.4% in Shanghai, and 7.7% in Changsha; at the time, incidence of ischemic stroke (IS) increased in Beijing (5.0% per year) and Shanghai (7.7%) during the 1990s. As reasoning this opposite trends of ischemic and hemorrhagic stroke, the authors wrote:These (lifestyle) changes may have resulted in increased prevalence of obesity, hypertension, diabetes, and hypercholesterolemia in Chinese populations and may also help explain the increasing trend of IS as observed in this study. Uncontrolled hypertension and high prevalence of smoking might partially explain the hemorrhagic stroke predominance among Chinese in the past, and the increasing awareness of hypertension control and decrease in cigarette smoking may have contributed to its decrease in the past decade.These explanations are ridiculous because it means hypertension and other risk factors were increasing when interpreting the trends of IS, while these risk factors were controlled and decreasing when interpreting the trends of ICH. Furthermore, well-designed large epidemiological studies indicated prevalence of cigarette smoking was increasing in the 1980s and 1990s,2 and percentage of smokers contemplating quitting was low in China.3 The incidence of hypertension and diabetes mellitus were also increased throughout the 1990s in China.4,5 Under these circumstances, it is rather arbitrary to interpret the 12% annual decrease of ICH incidence detected in this study as a result of some underlying changes of risk factors in Chinese population.On the other hand, the results that IS increased and ICH decreased during 1990s can be more reasonably explained by the misclassification of stroke patients, especially during the early days of this study when CT and MRI scans were less available. In the absence of neuroimaging results, IS may be misclassified as ICH and vice versa. This study did not provide substantive classification criteria applied at time of stroke onset. The Atherosclerosis Risk in Communities (ARIC) criteria for stroke diagnosis and classification used in this study were published many years after the initiation of the study.6 In the ARIC study, the rate of cases with CT scan is higher (84%), and stroke patients were possibly classified as definite cerebral infarction or definite ICH only when autopsy results or CT/MRI scans were available. For stroke patients without neuroimaging or autopsy results, a probable classification was made. When a stroke case was equivocal, it was classified following the hierarchy of ICH, subarachnoid hemorrhage, embolic brain infarction and thrombotic brain infarction. This indicates that equivocal cases were more probably classified as ICH than as IS. This arbitrary allocation will undoubtedly increase the possibility of IS being misclassified as ICH especially in patients without CT or MRI scans. Many patients who had been diagnosed in the early days of this study (eg, >50% of the patients diagnosed in 1991 in Shanghai) did not have a CT or MRI scan. For these patients ICH may be overdiagnosed.In a large multicenter collaborative study which included Beijing and Shanghai as this study did, Zhang and colleagues showed proportion of clinically diagnosed7 ICH in total stroke was increased gradually during the early years of the 1990s (which are contrary to the results of this study), a time when clinical application of CT and MRI was booming in mainland China. But when ratio of patients with CT scan reached a satisfying 85% after 1996, proportion of ICH kept considerably equivalent over years. These results indicate that inconsistency in stroke classification related to increasing use of CT/MRI scan may be a prominent obstruction in evaluating the temporal trends of stokes components.This article provides valuable information of stroke incidence in urban China. However, because of significant design flaw and lack of valid criteria for stroke classification, the conclusion that the decreased ICH and increased IS was caused by underlying changes of risk factors in Chinese populations is misleading.1 Jiang B, Wang WZ, Chen H, Hong Z, Yang QD, Wu SP, Du XL, Bao QJ. Incidence and trends of stroke and its subtypes in China: results from three large cities. Stroke. 2006; 37: 63–68.LinkGoogle Scholar2 Yang G, Fan L, Tan J, Qi G, Zhang Y, Samet JM, Taylor CE, Becker K, Xu J. Smoking in China: findings of the 1996 National Prevalence Survey. JAMA. 1999; 282: 1247–1253.CrossrefMedlineGoogle Scholar3 Yang G, Ma J, Chen A, Zhang Y, Samet JM, Taylor CE, Becker K. Smoking cessation in China: findings from the 1996 national prevalence survey. Tob Control. 2001; 10: 170–174.CrossrefMedlineGoogle Scholar4 Ueshima H, Zhang XH, Choudhury SR. Epidemiology of hypertension in China and Japan. J Hum Hypertens. 2000; 14: 765–769.CrossrefMedlineGoogle Scholar5 Chan JC, Ng MC, Critchley JA, Lee SC, Cockram CS. Diabetes mellitus–a special medical challenge from a Chinese perspective. Diabetes Res Clin Pract. 2001; 54: S19–S27.CrossrefMedlineGoogle Scholar6 Rosamond WD, Folsom AR, Chambless LE, Wang CH, McGovern PG, Howard G, Copper LS, Shahar E. Stroke incidence and survival among middle-aged adults: 9-year follow-up of the Atherosclerosis Risk in Communities (ARIC) cohort. Stroke. 1999; 30: 736–743.CrossrefMedlineGoogle Scholar7 Zhang LF, Yang J, Hong Z, Yuan GG, Zhou BF, Zhao LC, Huang YN, Chen J, Wu YF; Collaborative Group of China Multicenter Study of Cardiovascular Epidemiology. Proportion of different subtypes of stroke in China. Stroke. 2003; 34: 2091–2096.LinkGoogle ScholarstrokeahaStrokeStrokeStroke0039-24991524-4628Lippincott Williams & WilkinsJiang Bin, , MD, MPH and Wang Wen-zhi, , MD01042006Response:We thank Drs Xu and Liu for their interest in our recent publication on the stroke trends in 3 cities of China,1 and we reply to clarify their questions about our stroke diagnostic criteria and classification in our study and provide some further information regarding our study that we could not include in the original report because of the space limitation. However, we believe what we observed in our study was real, and our inference of the shift from hemorrhage to ischemic stroke pattern in China was correctly drawn.The criteria for stroke diagnosis and classification were similar to those used in the Atherosclerosis Risk in Communities (ARIC) Study, both of which were adapted from published criteria from the National Survey of Stroke.2 Minor modifications were made to facilitate the field data collection in China. The major differences between the criteria used in these 2 studies,1,3 other than languages, included: (1) "unilateral numbness of 2 or more body parts" was regarded as a major symptom in our study but a minor symptom in the ARIC study; (2) in the ARIC Study, a hierarchical case definition was used for the rare situation where the cases met criteria for 2 different diagnostic categories. In our study, we defined such cases as "undetermined", even for cases with CT/MRI information. Conversely, cases only with neuroimaging data without clinical symptoms and signs were excluded. As in many other epidemiological studies, misclassifications were expected, but diagnostic errors should be few.We shared Dr Xu et al's concern that >10% of the stroke subtypes were classified based on clinical examinations. This was true for the first several years of follow-up. CT/MRI technologies have become readily available in Beijing and Shanghai (>90%) since mid-1990s. Further, the high hospitalization rate and high quality medical service in these large cities should have to some extent alleviated such a concern. As noted by Dr Xu et al, ICH in total stroke gradually increased in the study by Zhang et al in early 1990s.4 However, CT use was slightly lower in that study compared to ours, and the increasing trend of ICH in total stroke stopped since 1996 when CT was used in 85% of their case diagnoses.1,4We believe that the trend we observed in our study reflects the changes of stroke incidence and composition in our study populations. Further, our observation was consistent with the findings from the MONICA study in Beijing during 1984 to 1999.5 However, our potential explanations are speculative and we have made it clear in our discussion. More studies are needed to confirm our findings in both urban and rural Chinese populations and to find out the reasons underlying these changes. Previous Back to top Next FiguresReferencesRelatedDetails April 2006Vol 37, Issue 4 Advertisement Article InformationMetrics https://doi.org/10.1161/01.STR.0000210186.07530.69PMID: 16567662 Originally publishedApril 1, 2006 PDF download Advertisement