Hemorrhagic stroke incidence is declining faster than ischemic stroke in Joinville, Brazil: A population-based study over 18 years and systematic review

Norberto L Cabral, Pedro Telles Cougo-Pinto, Pedro SC Magalhaes, Alexandre L Longo, Claudio H Amaral, Gerson Costa, Felipe I Reis, Anderson RR Gonçalves, Octavio M Pontes-Neto.

Joinville Stroke Registry, Joinville, Brazil.


Background Although the absolute number of ischemic stroke (IS) cases is twice that of hemorrhagic stroke (HS) cases worldwide, the overall burden of HS is higher. This study evaluated the incidence trends of all stroke types in Joinville, Brazil and compared them to other studies done in Latin America and the Caribbean.

Methods Using multiple overlapping sources, we ascertained the incidence of all first-ever strokes from 1995–2013. Premorbid cerebrovascular risk factors and one year functional outcome were registered from 2005–2013. In the same period, we did a systematic review on population- or hospital-based studies in Latin America and the Caribbean.

Results We registered 320, 759, 859, and 922 first-ever strokes in 1995, 2005–2006, 2010–2011, and 2012–2013, respectively. From 1995–2013, the age-adjusted incidence of all strokes decreased 37% (95% CI 32–42%). From 2005–2013, the HS incidence decreased 60% (95% CI 13–86), IS decreased 15% (95% CI 1–28), and subarachnoid hemorrhage incidence remained stable. The proportions of IS and HS patients with regularly treated hypertension increased 60% (p = 0.01) and 33% (p = 0.01), respectively. The proportions of IS and HS patients that quit smoking increased 8% (p = 0.03) and 17% (p = 0.03), respectively. A similar incidence trend was observed in all stroke in other Latin America and Caribbean countries, however data was scarce.

Conclusions Stroke incidence has been decreasing in Joinville over the last 18 years, more so for HS than IS. Better control of hypertension and tobacco use might explain these findings.





Regardless of the country, stroke remains a current and huge problem of health agenda worldwide. Data from Global Stroke Burden Disease Study reports a significant 22% increase in hemorrhagic stroke (HS) incidence and a non-significant 6% increase in ischemic stroke (IS) incidence in 61 low- and middle-income countries from 1990–2010 but no simultaneous changes in IS or HS. The same study  also shown that in Brazil, from 1990 to 2013, prevalence of ischemic stroke (IS) increased 90 % and hemorrhagic stroke ((HS) increased 96%.1  In recent years, three main actions directly aiming to mitigate stroke burden have been implemented in Brazil: (i) the National Stroke Policy Act, which is a national program dedicated to stroke launched in 2012; (ii) free access to the main antihypertensive and anti-diabetic medications, and statins; and (iii) more restrictive antismoking laws. However, the actual impacts of these policies on stroke incidence remain unclear. We aim to know the incidence trends of all types of stroke and their subtypes in Joinville, Brazil from 1995–2013; the trends of IS, HS, and subarachnoid hemorrhage (SAH) , 30-day and 1-year functional status from 2005–2013; We further addressed whether similar trends has been reported in other countries from Latin America and the Caribbean.  

The Joinville Stroke Registry is a population-based registry that has been on-going since 2005. Data on all cases of first-ever strokes occurring in Joinville residents in 1995 and 2005–2006 have been published previously.2 In brief, using multiple overlapping sources, we identified all inpatients and outpatients with stroke as described by Sudlow and Warlow as well as the three Stroke-Steps modular program proposed by the World Health Organization. We included all cases of patients of any age diagnosed with any type of first-ever or recurrent IS, HS, or SAH who were residents of Joinville. The study nurses recorded the premorbid use of medication (regular or not) and cardiovascular risk factors in face-to-face meetings with patients or their relatives. All of these data were extracted from patients with first-ever or recurrent strokes in 2005–2006 and 2012–2013. We determined the modified Rankin scale (mRS) scores by telephone interviews at 30 days and one year after discharge to evaluate the degree of disability and dependence of the patients.


Systematic review

We systematically reviewed population- or hospital-based studies assessing the incidence of stroke and stroke pathological types in Latin America and the Caribbean from 1995–2013 in order to assess whether there is any apparent trend. We included all relevant studies identified in a previous rigorous systematic review of stroke incidence studies published by Lavados et al. in 2007 that included studies published with no date limit until September 2006.3 We supplemented this with a comprehensive electronic search of Medline, Embase and Scielo for studies published in any language between Oct 2006 to February 2015 using a combination of MeSH terms and the following terms: (1) ‘Latin America’, ‘South America’, or ‘Caribbean’; (2) ‘stroke’, ‘cerebrovascular disease’, ‘ischemic stroke’, ‘hemorrhagic stroke’, ‘cerebral infarction’, ‘intracerebral hemorrhage’, ‘parenchymal hemorrhage’ or ‘subarachnoid hemorrhage’; (3) ‘epidemiology’, ‘incidence’, ‘first-ever stroke’, ‘population-based studies’. We also perused the reference lists of all relevant identified studies and review articles for additional potentially relevant studies. The world age-adjusted incidence rates of stroke and stroke type when available were extracted and compared by time period.

We registered 320, 759, 859, and 922 first-ever strokes in Joinville in 1995, 2005–2005–2006, 2010–2011, and 2012–2013 years, respectively. The age-standardized incidence of first-ever stroke stratified by sex and age also decreased significantly over time (Table 1). The reduction was 11% greater in men [42% (95% CI 35–49%)], than women [31% (95% CI 23–39%)] and 16% greater in young people [≤44 years: 54% (95% CI 41–66%)] than older people [>44 years: 38% (95% CI 33–43%)].

We stratified the incidences among IS, HS, and SAH from 2005–2013. The weight of the decrease in age-adjusted stroke incidence was proportionally higher of HS than IS, whereas that of SAH remained stable (Table 2). In the last eight years, the incidences of IS and HS showed significant absolute decreases of 15% (95% CI 1–28%) and 60% (95% CI 13–86%), respectively. Meanwhile, the incidence of SAH showed a 29% non-significant absolute decrease (95% CI −15–92%). We also compared the functional statuses (i.e. mRS scores) of IS and HS 30 days and one year after the first-ever stroke between 2005–2006 and 2012–2013. Worryingly, after 1 year, we found a non significant increase of 32% of functional dependence among IS patients and a non significant increase of  124 % among HS patients.  The case fatality rates of IS and HS (i.e. the proportion of patients with an mRS score of 6) tended to decrease over the last eight years, although not significantly.

In the systematic review, we extracted 512 abstracts. From those, we identified five studies of stroke incidence in the region (not including our prior reports from Joinville, Brazil); and three of these—from Iquique, Chile, and Martinique (in the Caribbean)—were population-based (Figure). Two of these studies involved populations from the Caribbean (Martinique4 and Barbados5), one from Central America (Durango, Mexico6), and two from South America (Iquique, Chile7 and Matão, Brazil8). Only one study, from Martinique, presents incidence results at different time points, showing a declining incidence of stroke that was only significant among IS events (Figure).

The premorbid cardiovascular risk factors of IS and HS patients with first-ever events were compared between 2005–2006 and 2012–2013 (Table 3). The proportions of IS and HS patients with regularly treated hypertension increased significantly by 60% (p = 0.0001) and 33% (p = 0.01), respectively. Among IS patients, the mean cholesterol level and proportion of patients with levels ≥6.0 mmol/L decreased significantly by 8% and 13%, respectively (both p = 0.0001). However, no changes were observed in HS patients. The proportions of IS and HS patients who quit smoking increased significantly by 8% (p = 0.03) and 17% (p = 0.03), respectively. The prevalence of obesity (i.e. BMI > 30) in both IS and HS patients increased significantly by 8% (p = 0.006) and 13% (p = 0.03), respectively.


Thus, this study, which extracted new data from our registry until 2013, revealed that the stroke incidence has been decreasing in Joinville over the last 18 years. To our knowledge, this is the first study showing a long-term decreasing trend in stroke incidence in a population-based setting in a middle-income country. Indeed, the available data regarding trends in stroke incidence in Latin America and the Caribbean is extremely scant.

There are some potential explanations for these results. 9 From 1991 to 2010, Brazil’s HDI almost doubled from 0.493 (very low) to 0.727 (high). Meanwhile, the HDI of Joinville, an industrial city in Southern Brazil, increased from 0.585 (66th in among Brazilian cities) in 1991 to 0.711 (79th) in 2000 and to 0.809 in 2010 (21st).9

Our findings showed the proportional decrease in the incidence of HS (60%; 95% CI 13–86%) was greater than that of IS (15%; 95% CI 1–28%). In fact, a similar trend was reported in the Global Burden of Disease Study: the incidence of HS has been decreasing significantly in high-income regions of North America and Western Europe as well as countries in tropical regions and Southern Latin America.1 However, it remains unclear whether the rate of cerebral haemorrhage is declining among stroke patients in South America as addressed by Del Brutto and Del Brutto in 2014.10 Similarly, in Santiago, Chile, the proportion of HS decreased from 38% in 1997 to 16% in another hospital-based study in 2006.11

In order to understand why stroke rates have been falling over the last 18 years in Joinville, we compared the prevalence of premorbid cardiovascular risk factors. The hypertension population-attributable risk for IS and HS are 45% and 74%,12 respectively; whereas the smoking population-attributable risk is higher in IS (21%) than HS (10%).12 Both the IS and HS cohorts showed significant increases in the proportion of patients under regular treatment for hypertension and the number of people who quit smoking. We hypothesise the stroke incidence decrease is more pronounced for HS than IS because of various factors interacting at local and national levels of socioeconomic development. At the local level, two accomplishments should be stressed. First, Joinville was the first city in Brazil to establish a stroke unit.13 The stroke units’ teams coordinates the World Stroke Day since 2006. In February 2011, the Federal Government initiated a new program to control hypertension and diabetes, increasing the distribution of medications for these conditions to the population at no cost to more than 20,000 popular pharmacies.14 After one year, the number of people who received free medication increased 264% from 853,000 to 3.2 million per month, including free access to new drugs such as amlodipine, enalapril, and losartan.14 A meta-analysis published in 2012, reported that hypertension prevalence in Brazil decreased 6% in the last three decades.15 Thus, increased access to an expanded number of antihypertensive medications in the context of differing hypertension population-attributable risks (i.e. higher in HS than IS) might be another explanation for our findings.

The present results show the proportions of both IS and HS patients who have quit smoking have increased significantly over the last eight years. A phone survey conducted in all Brazilian capitals from 2008 to 2013 shown a reduction in the prevalence of smoking in the population with and without health insurance of 0.72% and 0.69% per year respectively.16 Moreover, a large cohort study in Denmark with 14 years of follow-up examined the combined effect and interactions among socioeconomic status, smoking, and hypertension on IS and HS incidence; the results show that reducing smoking in those with low socioeconomic status and/or hypertension may reduce the social inequality of stroke incidence.17

In conclusion, the incidences of major stroke subtypes in Joinville, Brazil have decreased consistently over the last 18 years. Although causality cannot be proven, the reduction of smoking habit and free access to new and effective drugs for hypertension might have contributed to the current decreasing trend of stroke incidence in our setting. Nevertheless, more prospective population-based studies are necessary in Brazil as well as in other low- and middle-income countries in order to clarify the current stroke burden.



  1. Krishnamurthi RV, Feigin VL,Forouzanfar MH,Mensah GA,Connor M, Bennett DA, Moran AE, Sacco RS, Anderson LM, Truelsen T, O’Donnell M, Venketasubramanian N, Barker-Collo S, Lawes CMM, Wang W, Shinohara Y, Witt E, Ezzati M, Naghavi M, Murray C, on behalf of the Global Burden of Diseases,Injuries, Risk Factors Study 2010 (GBD 2010) and the GBD Stroke Experts Group. Global and regional burden of first-ever ischaemic and haemorrhagic stroke during 1990–2010: findings from the Global Burden of Disease Study 2010. Lancet Glob Health 2013; 1 (5): e259–81.
  2. Cabral NL, Gonçalves ARR, Longo AL, Moro CHC, Costa G, Amaral CH, Souza MV, Eluf-Neto J, Fonseca LAM.Trends in stroke incidence, mortality and case fatality rates in Joinville, Brazil: 1995-2006. J Neurol Neurosurg Psychiatry 2009; 80: 749–54.
  3. Lavados PM, Hennis AJ, Fernandes JG, Medina MT,Legetic B,Hoppe A, Sacks C,Jadue L, Salinas R, MDi Stroke epidemiology, prevention, and management strategies at a regional level: Latin America and the Caribbean. Lancet Neurol 2007; 6: 362–72.
  4. Olindo S, Chausson N, Mejdoubi M, Severine Jeannin S, Rosillette K,Saint-Vil M, Signate A, Edimonana-Kaptue M, Larraillet V,Cabre P,Smadja D, Joux J. Trends in incidence and early outcomes in a Black Afro-Caribbean population from 1999 to 2012: Etude Réalisée en Martinique et Centrée sur l’Incidence des Accidents Vasculaires Cérébraux II Study. Stroke 2014; 45: 3367–73.
  5. Corbin DOC, Poddar V, Hennis A, Gaskin A, Rambarat C,Wilks R, Wolfe CDA, Fraser HS, Incidence and case fatality rates of first-ever stroke in a Black Caribbean population: The Barbados Register of Strokes. Stroke 2004; 35: 1254–8.
  6. Cantu-Brito C, Majersik JJ, Sánchez BN, Ruano A, Quiñones G, Arzola J,Morgenstern LB. Hospitalized stroke surveillance in the community of Durango, Mexico: the brain attack surveillance in Durango study. Stroke 2010; 41: 878–84.
  7. Lavados PM, Sacks C, Prina L, Escobar A, Tossi C, Araya F, Feuerhake W, Galvez M, Salinas R, Alvarez G.Incidence, 30-day case-fatality rate, and prognosis of stroke in Iquique, Chile: a 2-year community-based prospective study (PISCIS project). Lancet 2005; 365: 2206–15.
  8. Minelli C, Fen LF, Minelli DPC. Stroke incidence, prognosis, 30-day, and 1-year case fatality rates in Matão, Brazil: a population-based prospective study. Stroke 2007; 38: 2906–11.
  9. Albright MK. United Nations. Foreign Policy 2003; 138: 16–24.
  10. Del Brutto Oh, Del Brutto VJ. Is the rate of cerebral hemorrhages declining among stroke patients in South America? Int J Stroke 2014; 9 (2): 207–9.
  11. Diaz T V, Illanes DS, Reccius MA, Manterola VJL, Cerda CP, Recabarren LC, González VR. Evaluation of a stroke unit at a university hospital in Chile. Rev Med Chil 2006; 134: 1402–8.
  12. O’Donnell MJ, Denis X, Liu L, Liu L, Zhang H, China SL,Rao-Melacini P, Rangarajan S, Islam S, Pais P,McQueen MJ, Mondo C, Damasceno A, Lopez-Jaramillo P, Hankey GJ, Dans AL, Yusoff K,Truelsen T, Diener HC, Sacco R, Ryglewicz D, Czlonkowska A. Risk factors for ischaemic and intracerebral haemorrhagic stroke in 22 countries (the INTERSTROKE study): A case-control study. Lancet 2010; 376: 112–23.
  13. Stroke Unit Trialists’ Collaboration. Organised inpatient (stroke unit) care for stroke. Cochrane Database Syst Rev 2007 Oct 17; 4: CD000197.
  14. Martins SCO, Pontes-Neto OM, Alves CV, Freitas GR, Oliveira-Filho J, Tosta ED, Cabral NL and on behalf of Brazilian Stroke Network Past, present, and future of stroke in middle-income countries: the Brazilian experience. Int J Stroke 2013; 8 Suppl A1: 106–11.
  15. Picon RV, Fuchs FD, Moreira L B, Riegel G, Fuchs SC. Trends in prevalence of hypertension in Brazil: a systematic review with meta-analysis. Plos One 2012; 7(10): e48255.
  16. Malta DC, Bernal RTI, Oliveira, M. (2015). Trends in risk factors chronic diseases, according of health insurance, Brazil, 2008-2013.Ciência & Saúde Coletiva, 2015; 20 (4): 1005-1016.
  17. Nordahl H, Osler M, Frederiksen BL, Andersen I, Prescott E, Overvad K, Diderichsen F, Rod NH. Combined effects of socioeconomic position, smoking, and hypertension on risk of ischaemic and hemorrhagic stroke. Stroke 2014; 45(9): 2582–7.


Acknowledgments We thank our colleagues at all state-run units and private services of Joinville as well as the Joinville Neurologic Clinic staff. The Joinville Secretary of Health has funded the Joinvile Stroke Registry since 2005.


Funding Sources Joinville City Health Department. Brazilian National Council for Scientific and Technological Development (CNPq) – Grants: 402396/2013-8 and 402388/2013-5.