Definition of Stroke

Epidemiological studies are dealing with population-level assessments of disease occurrence and determinants of health and disease. It is important to define the disease in a standardized way and using methods that are suitable for large population examinations. Therefore, the definition of a disease may differ to some extent from that used in individuals in clinical medicine where it is possible to apply more detailed and expensive diagnostic tools. It has been agreed that for epidemiological purposes stroke is primarily defined by clinical findings and symptoms [1]: rapidly developed signs of focal (or global) disturbance of cerebral function lasting more than 24 hours (unless interrupted by surgery or death), with no apparent cause other than a vascular origin. This approach may be supplemented with neuroimaging, but even with advanced imaging techniques the diagnosis is based on clinical signs. Therefore, precise definitions of clinical signs are needed. WHO definitions are [1]:

Definite focal signs:

• 
  

  unilateral or bilateral motor impairment (including dyscoordination)

  
  
• 
  

  unilateral or bilateral sensory impairment

  
  
• 
  

  aphasis/dysphasis (non-fluent speech)

  
  
• 
  

  hemianopia (half-sided impairment of visual fields)

  
  
• 
  

  diplopia

  
  
• 
  

  forced gaze (conjugate deviation)

  
  
• 
  

  dysphagia of acute onset

  
  
• 
  

  apraxia of acute onset

  
  
• 
  

  ataxia of acute onset

  
  
• 
  

  perception deficit of acute onset.

Not acceptable as sole evidence of focal dysfunction, although strokes can present in this way, these signs are not specific and cannot therefore be accepted as definite evidence of stroke:

• 
  

  dizziness, vertigo

  
  
• 
  

  localized headache

  
  
• 
  

  blurred vision of both eyes

  
  
• 
  

  dysarthria (slurred speech)

  
  
• 
  

  impaired cognitive function (including confusion)

  
  
• 
  

  impaired consciousness

  
  
• 
  

  seizures.

Neuroimaging studies are needed for classification of stroke by subtypes: subarachnoid hemorrhage, intracerebral hemorrhage, and brain infarction (necrosis). Although there may be large variations in stroke subtype distributions between populations, thrombotic and embolic strokes are responsible for about 80–85% of all strokes in the Indo-European populations, and as low as 65% in some Asian populations. Subarachnoid hemorrhage represents 5–10% of all strokes, and may often occur in middle-aged people [2], while both intracerebral and especially thrombotic and embolic stroke events increase markedly with age.

This classic WHO definition is mainly clinical and does not account for recent advances in science and diagnostic technology. Therefore, the Stroke Council of the American Heart Association/American Stroke Association developed an expert consensus document for an updated definition of stroke in 2013 [3] that attempted to take into account advances in basic science, neuropathology, and neuroimaging of the central nervous system (Box 6.1). Central nervous system infarction was defined as brain, spinal cord, or retinal cell death attributable to ischemia, based on neuropathological, neuroimaging, and/or clinical evidence of permanent injury. Central nervous system infarction occurs over a clinical spectrum: ischemic stroke specifically refers to central nervous system infarction accompanied by overt symptoms, while silent infarction by definition causes no known symptoms. Stroke also broadly includes intracerebral hemorrhage and subarachnoid hemorrhage. The aim of this updated definition of stroke was to incorporate clinical and tissue criteria that may be used in clinical practice, research, and public health aspects.

For ischemic stroke, further subtypes have been proposed, since prognosis, recurrence rate, and management of acute phase vary by subtype. In the Oxfordshire Community Stroke Project Subtype Classification (OCPS) four clinically identifiable subgroups of cerebral infarction were defined [4]. Of the patients 17% had large anterior circulation infarcts with both cortical and subcortical involvement (total anterior circulation infarcts, TACI), 34% had more restricted and predominantly cortical infarcts (partial anterior circulation infarcts, PACI), 24% had infarcts clearly associated with the vertebrobasilar arterial territory (posterior circulation infarcts, POCI), and 25% had infarcts confined to the territory of the deep perforating arteries (lacunar infarcts, LACI). Other criteria called TOAST (Trial of ORG 10172 in Acute Stroke Treatment) propose five subtypes: large artery atherosclerosis, cardioembolism, small artery occlusion, stroke of other determined cause, and stroke of undetermined cause [5].

The Scope of the Problem

Stroke is the second leading cause of death worldwide in the adult population, the first being coronary heart disease [6]. Stroke is an increasing problem in low- and middle-income countries where over 80% of all stroke deaths occur [2, 6–11]. Stroke is the second leading cause of disease burden (as measured in disability-adjusted life years [DALYs]) after coronary heart disease in 2016 [12]. It caused about 4.4 million deaths worldwide in 1990, 5.4 million in 1999, 5.7 million in 2004, and 5.5 million in 2016, with two-thirds of these deaths occurring in less-developed countries [6, 8–10, 13]. In 2010, globally the highest burden of stroke was found among people living in low- and middle-income countries. DALYs due to stroke were 62.67 per million person-years in high-income countries, corresponding to 4.5% of the total DALYs, the estimate for low-and-middle-income countries was 9.35 per million person-years corresponding to 6.3% of the total DALYs in the human population [10]. The burden of stroke is particularly high in Eastern Europe, North Asia, Central Africa, and the South Pacific with a 10-fold difference in stroke mortality and morbidity rates between the most affected and least affected countries [10, 11].

The Stroke Experts Panel of the Global Burden of Disease (GBD) group has also reported estimates for ischemic and hemorrhagic stroke in all ages for 188 countries during 1990 to 2013 [14]. In 2013, there were globally almost 25.7 million stroke survivors (71% with IS), 6.5 million deaths from stroke (51% died from IS), 113 million DALYs due to stroke (58% due to IS), and 10.3 million new strokes (67% IS). During 1990 to 2013, there were statistically significant reductions in the incidence, mortality, and DALY rates of ischemic stroke. For hemorrhagic stroke there were statistically non-significant increases in the incidence and prevalence, and decreases in the mortality and DALY rates (Figure 6.1). There was a significant increase in the absolute number of DALYs due to ischemic stroke, and of deaths from ischemic and hemorrhagic strokes, survivors, and incident events for both ischemic and hemorrhagic stroke (Table 6.1). The preponderance of the burden of stroke continued to reside in developing countries, comprising 75.2% of deaths from stroke and 81.0% of stroke-related DALYs. The recent estimate indicated that in the USA the cost of stroke (direct and indirect costs together) was $73.7 billion [3] in 2010.

Figure 6.1 Age-adjusted DALYs, mortality, incidence, and prevalence rates of ischemic (IS, blue) and hemorrhagic (HS, red) strokes per 100 000 people (with 95% uncertainty intervals) in 1990, 2005, and 2013.

(From Feigin et al. 2015 [14] with permission from S Karger AG, Basel.)

The rapid increase in life expectancy in most parts of the world has a major effect on the burden of stroke, since stroke is a disease of older people. Even though the stroke risk in the middle-aged and young-old population has significantly decreased in many countries, the onset of the disease might have moved to older ages. Therefore, the overall burden measured by incidence, prevalence, and DALYs related to stroke globally has not changed much during the last decade [14].

Occurrence of Stroke: Incidence, Prevalence, Mortality, and Case Fatality

Although stroke is considered as a major public health problem, data are still limited to epidemiology of stroke besides mortality statistics. There are several measures to describe the occurrence of stroke from an epidemiological (and to some extent also clinical) perspective. These are: the incidence (or event rate), prevalence, mortality, and case fatality (reverse of survival). The incidence means the number of first stroke events (either non-fatal or fatal) per population during a certain time period, for instance events/100 000 population/year. While it is useful to know the incidence (occurrence of first stroke events), in most populations data may be available on mortality from stroke only, but not on non-fatal events. The case fatality at the stroke event is usually determined as the proportion of deaths occurring during the first 4 weeks after the onset of a stroke event, but can be presented for any other time period as well. It gives information about the severity of stroke and may also reflect the efficacy of early management of acute stroke. The relative frequency of different subtypes of stroke varies among ethnic groups and populations. This variation is mostly due to genetic differences, but also due to differences in lifestyle and risk factor profiles.

Mortality

Since cause-specific mortality data are routinely collected in all high-income countries and also in many low- and middle-income countries nationwide, stroke mortality data are available in a large proportion of the world’s population. However, the coverage and accuracy of stroke mortality varies among countries. In some countries validation studies have been carried out using standard methodology [1]. Data from many countries show that stroke mortality rates have declined over recent decades, most notably in Japan, Australia, North America, and Western Europe [15]. Mortality from stroke was highest in the world in Finland in the 1970s, together with Japan [15–19], but reduced dramatically in these countries over the past decades.

The geographic differences in stroke mortality among countries are huge. The highest rate was found in Russia and several other countries from the previous Soviet Union [11]. Also China, Mongolia, countries from the Middle East and North Africa, Brazil, the Caribbean region, and the Pacific islands have high incidence of ischemic stroke. The lowest rate was seen in the Seychelles and Switzerland. Most of the low mortality countries were found in Western Europe, North and Central America, Japan, Australia, and New Zealand (Figure 6.2). Mortality rates were 2-fold higher in developing countries than in developed countries. In 2013, there were approximately 5.6 million deaths from stroke globally; ischemic and hemorrhagic strokes had an equal share of these deaths [14]. The highest ischemic stroke mortality rates (124–174 per 100 000 person-years) were observed in Russia and Kazakhstan, with the lowest (at or below 25 per 100 000 person-years) seen in Western Europe and North and Central America, a 7-fold difference [14].

Figure 6.2 Age-standardized annual mortality rates (per 100 000) of ischemic stroke (A) and hemorrhagic stroke (B) in 2013. There is large geographical variation in stroke burden, with the highest stroke mortality rates in developing countries.

(From Feigin et al. 2015 [20], with permission from S Karger AG, Basel.)

Incidence

The incidence gives the rate of new cases of stroke within a time period (usually per year) in a specified population. The incidence depends on the effect of risk factors that are causing the disease. There are few incidence studies with validated data from stroke registers or other sources. The highest rate was found in Russia and also several other countries from the previous Soviet Union are among the top 20. Also China, Mongolia, countries from the Middle East and North Africa, Brazil, the Caribbean region, and the Pacific islands have a high incidence of ischemic stroke. The GBD Stroke Panel Experts Group has presented data on incidence and mortality of stroke for all countries to assess the burden of ischemic and hemorrhagic stroke between regions and over time [14]. The analysis was based on 119 studies, 58 in high-income and 61 in low-income countries. They used a GBD analytic technique (DisMod-MR) [21] to calculate the country-specific incidence for 1990 and 2010. The highest incidence rate for ischemic stroke was found in Russia and also several other countries from the previous Soviet Union and Balkans were among the top. Also China, Mongolia, countries from the Middle East and North Africa, Brazil, the Caribbean region, and the Pacific islands have high incidence of ischemic stroke.

The incidence of stroke has declined sharply in Finland during the last decades [17], and in 1998 it was 241/100 000, not far from other Western industrialized countries, after a steady fall of about 3% per year throughout the years studied. Other countries that already had comparatively low stroke incidence rates in the 1980s, for example New Zealand [22], the USA [23], or Denmark [24], have reported no further fall in stroke incidence, while an increase in the incidence of stroke has been observed in Eastern Europe and Russia [2, 9, 10, 25–27]. In Shanghai, China, almost no decline in incidence of stroke but a clear decline in stroke mortality was reported [28]. Central and Eastern European countries have the highest incidence and mortality rates through Europe. The improvements in stroke prevention and treatment in Central and Eastern European countries did not completely reach the quality parameters present in Western European countries [29]. Incidence studies from Greece [30] and Bulgaria [31] confirm the findings from mortality statistics showing the populations in the Balkan region have a very high risk of stroke. Geographic comparisons of stroke incidence are useful for identifying populations at high risk and developing hypotheses for prevention of stroke. The differences observed between countries in mortality rates, and even more in incidence rates, are, however, difficult to interpret, as they largely depend on the study design, the accuracy of the data collection, and the time point when the measurements were made.

Prevalence

Prevalence means the proportion of people in the population who have the disease at a particular time point. In the case of stroke, the prevalence shows the number of stroke survivors in the population. It therefore means the number of incident cases minus people deceased in stroke attack. The prevalence is more difficult to estimate than the incidence or mortality, since the number of stroke survivors living in the target population needs to be known, and this cannot be derived from routine statistics of hospital admissions (a usual source for incidence) or death register (the source for mortality). Both the incidence and mortality vary over time and geographically depending on many factors such as age, socioeconomic status, lifestyles, comorbidities, and health services. Thus, the prevalence is a result from many medical and non-medical factors. Nevertheless, the GBD Stroke Panel Experts Group estimated prevalence of stroke during 1990 to 2013 [32]. Among adults aged 20–64 years, the global prevalence of hemorrhagic stroke in 2013 was 3.7 million and the prevalence of ischemic stroke was 7.3 million (Table 6.2). Globally, the number of people with prevalent hemorrhagic stroke increased from 2.0 million in 1990 to 3.7 million in 2013, and the number of those with prevalent ischemic stroke from 3.8 million to 7.3 million. In 2013, of people living with hemorrhagic stroke 2.7 million lived in developing countries and 1.1 million in developed countries; and of people living with ischemic stroke 3.2 million lived in developing countries and 4.0 million in developed countries. The prevalence per 100 000 population for hemorrhagic stroke increased from 75 to 130 in developed countries and from 72 to 81 in developing countries during 1990 to 2013. The increases for ischemic stroke prevalence per 100 000 population were from 292 to 497 in developed and from 86 to 97 in developing countries, respectively.

Case Fatality

The case fatality indicates the proportion of people who died in acute stroke attack. It reflects the severity of the underlying disease or efficacy of the management of acute stroke, or both. Before the era of thrombolysis or thrombectomy no proven life-saving clinical treatment for acute stroke was available; therefore, the case fatality mainly reflected the disease severity. It can be calculated in various ways: within a defined time period, e.g. 1-day, 1-week, 4-week, 1-year period, etc., or in relation to the management of stroke, e.g. out-of hospital, in-hospital, before thrombolysis, after thrombolysis, etc. The overall case fatality (the proportion of deaths among all strokes) varies markedly among populations, even within a country, and is roughly 20% within the first month, and subsequently increases by around 5% each year after the acute stroke event. The large variation in case fatality of stroke was demonstrated well in the WHO Monitoring of Trends and Determinants in Cardiovascular Disease (MONICA) Stroke Study: among men the case fatality of stroke ranged from 12% in northern Sweden to 53% in Moscow, Russia [25]. Overall, the case fatality was high in all Eastern European countries. In women, the difference in case fatality of stroke between populations was larger than in men, ranging from 16% in Kuopio, Finland to 57% in Moscow. The INTERSTROKE study has shown that stroke case fatality and functional outcome is dependent on the case mix presenting in different global regions on the one hand, but also on the access to stroke management services, such as CT scanning, stroke unit treatment, and anti-platelet therapy on the other [33]. Good functional outcome (modified Rankin Scale 0–3) is achievable in 90% of stroke victims in high-income countries compared to 78% in low-income or middle-income countries [33].

Trends in Stroke Event Rates, Case Fatality, and Mortality of Stroke

In the WHO MONICA Stroke Study stroke event rates, case fatality, and mortality (obtained from both the register and in routine mortality statistics) [34] of stroke event rates declined in 9 of 14 populations in men and 8 of 14 populations in women. In men, the case fatality of stroke declined in seven populations, increased in eight, and fluctuated only slightly in two. Among women, a decline in case fatality was seen in eight populations, no obvious change was seen in three, and an increase was observed in three. The trends in case fatality were statistically significant among men in only two populations with declining trends and in two with increasing trends. Among women, there was a significant downward trend in four populations. Of the 14 populations, stroke mortality declined in 8 populations among men and 10 populations among women. Thus, stroke trends have not been uniform among countries. Stroke mortality increased in all the Eastern European populations except in Warsaw, Poland. In Beijing, China and in the nine Western European populations, stroke mortality declined.

Effects of changes in incidence and improved survival on the downward trend in stroke mortality are not easy to quantify and compare, due to the difficulty of measuring accurately the incidence of stroke. In the WHO MONICA Stroke Study populations, changes in stroke mortality, whether declining or increasing, were principally attributable to changes in case fatality rather than changes in event rates (i.e. incidence of stroke) [34]. Since only limited advances in acute stroke care took place during that time, it is likely that the natural history of stroke events has changed and they have become less severe. Effects of risk factor changes on stroke mortality have obviously been important, such as reduced blood pressure levels, smoking, and dietary changes (especially lower salt intake), but actual data on these issues are limited.

The analysis by the GBD Stroke Panel Experts Group showed that globally there were significant increases in absolute numbers and prevalence of both hemorrhagic and ischemic stroke for people aged 20–64 years [32]. There was a 20% decline in the number of total stroke deaths in this age group in developed countries, but a 37% increase in developing countries (Figure 6.3). Death rates for all strokes declined significantly from 47 to 39 per 100 000 population per year from 1990 to 2013 in developing countries and from 33 to 24 per 100 000 population per year in developed countries. Regarding stroke subtypes, for hemorrhagic stroke mortality there was a significant decrease for adults aged 20–64 only in developed countries from 20 to 14 per 100 000 population per year between 1990 and 2013, but no significant change was detected in ischemic stroke mortality. Globally, there was a 24% increase in total DALYs with a 20% and 37% increase in hemorrhagic stroke and ischemic stroke numbers, respectively.

Figure 6.3 Percentage change from 1990 to 2013 in total stroke mortality by Global Burden of Disease region in the age group 60–64. Significant declines on mortality were seen in the high-income regions of Asia Pacific and other developed regions of the world, whereas increases in mortality were seen in African regions and Oceania.

(From Krishnamurthi et al. 2015 [32] with permission from S Karger AG, Basel.)

After the WHO MONICA Stroke Study no formal multinational comparison of stroke incidence has been organized. There are, however, data from many countries. Feigin at al. carried out a systematic review of published stroke incidence studies from 1970 to 2008 [2]. They found adequate data from 47 centers in 28 countries. Over the four decades, age-adjusted stroke incidence rates in high-income countries decreased by 42% (from 163 per 100 000 person-years in 1970–1979 to 94 per 100 000 person-years in 2000–2008; p = 0.0004), whereas in low- to middle-income countries the stroke incidence rates more than doubled (52 per 100 000 and 117 per 100 000 person-years, respectively; p <0.0001). During 2000–2008, for the first time stroke incidence rates in low- to middle-income countries exceeded the rate observed in high-income countries. Early stroke case fatality was decreasing in both high-income and low- to middle-income countries, but, overall, early stroke case fatality in low- to middle-income countries in the past decade was 25% higher than early stroke case fatality in high-income countries.

The Swedish National Stroke Register, Riks-Stroke, has demonstrated that it is possible to develop nationwide data collection system for acute stroke events [35]. Riks-Stroke is the world’s longest-running national stroke quality register (established in 1994), includes all 76 hospitals in Sweden admitting acute stroke patients, and covers approximately 85% of all stroke events. In Canada, the Canadian Institute for Health Information’s Hospital Morbidity Database that includes ICD codes has been used to assess trends in hospital admissions and in-hospital case fatality for stroke [36]. The age- and sex-standardized rate of hospital admissions decreased by 28% for stroke, but case fatality decreased only by 9% during 1994–2004. The average annual rate of decline in stroke mortality was about 3%. In Finland the national Hospital Discharge Register has been used to evaluate the outcome of stroke patients [18]. During 1999 to 2007 stroke outcome in Finland has improved as length-of-stay in hospital decreased for ischemic stroke patients. Active treatment during the acute phase of stroke has become more specialized, which was shown to be associated with improved patient outcome. Nevertheless, a significant proportion of Finnish stroke patients still did not receive optimal care. The situation is likely to be similar in other countries.

Recently, a national stroke clinical registry was established in Australia. This protocol was published with pilot data that showed that younger stroke patients exhibited distinct differences from their older counterparts with regard to demographic and clinical characteristics, prescription of anti-hypertensive medications, and residual health status [37]. In other publications from this registry treatment and outcome data of stroke events were presented [38].

In the USA stroke mortality fell by 33.5% from 1996 to 2006, with the total number of stroke deaths declining by 18.4% [3]. The previously set goal of a 25% reduction was exceeded in 2008. A study from Dublin, Ireland, in 2005–2006 found the crude incidence of 165 per 100 000 person-years for first-ever stroke and 28 for recurrent stroke, and 45 first-ever transient ischemic attack [39]. Age-adjusted stroke rates in Dublin were higher than those reported earlier in nine other recent population-based samples from high-income countries. Data from Oxfordshire [40] and London, UK [41], and from Beijing, Shanghai, and Changsha, China [42], suggest that the implementation of preventive treatments and decreases in risk factors at the population level have contributed to the significant fall in stroke incidence. On the other hand, the data from Belarus during 2001–2003 confirmed that the stroke incidence in Eastern Europe is higher than in Western Europe [42]. Incidence studies from Greece [30] and Bulgaria [31] confirm the findings from mortality statistics showing the populations in the Balkan region have a very high risk of stroke.

In Dijon, France, attack rates increased significantly over time irrespective of the stroke subtype [43]. In contrast, overall stroke mortality rate declined with decreasing rates for ischemic stroke, but no change occurred for intracerebral hemorrhage and subarachnoid hemorrhage. Between the first (1987–1991) and the last (2007–2012) study periods, the annual number of stroke patients who survived beyond 30 days rose by 55%. Thus, increasing attack rates and decreasing mortality have led to a rise in the number of stroke survivors in the population, thus indicating a growing need for services for stroke victims. Therefore, stroke mortality rates and their changes cannot provide satisfactory information for planning of health services for stroke.

Risk Factors for Stroke

Stroke has a multifactorial origin and a plethora of putative and confirmed risk factors have been listed and tested in various types of studies. Knowledge about risk factors is essential for prediction and prevention of the disease. The assessment of the global epidemiology is severely hindered by the lack of any kind of data on stroke occurrence and risk factors in most populations in the world. Although over 65% of all deaths due to stroke occur in developing countries, studies of stroke epidemiology in these populations hardly exist.

Risk factors for stroke are discussed in more detail in Chapter 7. The American Heart Association Stroke Council’s Scientific Statement Oversight Committee guideline has provided an overview of the evidence on various established and potential stroke risk factors and proposed recommendations for the reduction of stroke risk in 2006 [44] with the latest extensive update in 2017 [3]. This paper represents probably the most thorough assessment of the prediction and potential for the prevention of stroke. The committee used systematic literature reviews published since 2001, reference to previously published guidelines, personal files, and expert opinions to summarize existing evidence on standard criteria. Risk factors or risk markers for a first stroke were classified according to their potential for modification (non-modifiable, modifiable, or potentially modifiable) and strength of evidence (well documented or less well documented). It is important to note that risk factors can be biological, behavioral, environmental, and social. For instance, in a multivariable analysis, the association of stroke mortality rates with gross national income among countries remained significant even after adjustment for national indicators of cardiovascular disease risk, with a 4% reduction in stroke mortality for every additional US$1 000 in gross national income per capita.

Non-modifiable risk factors include age, sex, low birth weight, race/ethnicity, and genetic factors. Well-documented and modifiable risk factors include arterial hypertension, exposure to cigarette smoke, air pollution, low socioeconomic status, diabetes mellitus, atrial fibrillation and certain other cardiac conditions, dyslipidemia, carotid artery stenosis, excessive alcohol drinking, sickle-cell disease, post-menopausal hormone therapy, poor diet, physical inactivity, and obesity and central body fat distribution. Less well-documented or potentially modifiable risk factors include the metabolic syndrome, drug abuse, oral contraceptive use, sleep-disordered breathing, migraine headache, elevated gamma-glutamyl transferase, hyperhomocysteinemia, elevated lipoprotein(a), elevated lipoprotein-associated phospholipase, hypercoagulability, low-grade inflammation, and chronic infection.

In the INTERSTROKE study comparing 3 000 patients and controls from 22 countries the following significant risk factors for stroke were found: history of hypertension, current smoking, waist-to-hip ratio, low healthy diet risk score, low physical activity, diabetes mellitus, regular alcohol intake, psychosocial stress, depression, cardiac causes, and ratio of apolipoproteins B to A1. Taken together, these risk factors explain up to 90% of the PAR (population attributable risk) of stroke [45].

Whether and to what extent risk factors for the main types of stroke differ has not been commonly agreed. A recent study comprising 712 433 participants in the UK Million Women Study without prior stroke, heart disease, or cancer reported behavioral and related factors at baseline (1999–2007) and were followed up by record linkage to national hospital admission and death databases. After 12.9 (SD 2.6) years of follow-up, 8 128 women had an incident ischemic stroke, 2 032 had intracerebral hemorrhage, and 1 536 had subarachnoid hemorrhage [46]. In women with diabetes, the risk of ischemic stroke was substantially increased (RR 2.01), risk of intracerebral hemorrhage was increased slightly (RR 1.31), but risk of subarachnoid hemorrhage was reduced (RR 0.43) (heterogeneity by stroke type, p <0.0001). Current smokers were at an increased risk of all three stroke types (although greater for subarachnoid hemorrhage [≥15 cigarettes/d vs. never smoker, RR 4.75) than for intracerebral hemorrhage (RR 2.30) or ischemic stroke (RR 2.50). Also hypertension increased the risk of all stroke types: subarachnoid hemorrhage (RR 1.06), intracerebral hemorrhage (RR 1.37), and ischemic stroke (RR 1.72). Obesity was associated with an increased risk of ischemic stroke, but a decreased risk of hemorrhagic stroke.

Although non-modifiable risk factors cannot be a target to interventions to reduce them, they are very important for the risk assessment, and cannot be overlooked. Such factors can interact with modifiable risk factors and/or their treatment. Non-modifiable risk factors also have a significant effect on recovery and prognosis, and they must be considered in clinical management of stroke patients. Age is probably the most important determinant of stroke; the risk of stroke doubles for each successive decade after age 55 years [47, 48] (Figure 6.4).

Figure 6.4 Incidence per 100 000 of ischemic and hemorrhagic stroke in females and males by 5-year age bands in 1990–2013 (from [49] with permission from S Karger AG, Basel). Stroke risk increases after the age of 55; age-related patterns are relatively similar for men and women across time and stroke types.

Racial or ethnic-specific stroke risk is difficult to interpret. While within a country such as the USA clear ethnic group differences exist (African Americans [50–52] and some Hispanic Americans [53, 54] have higher stroke incidence and mortality rates when compared with European Americans), globally stroke mortality does not follow any specific ethnic patterns [55]. Nevertheless, it is well known that intracerebral bleeding is more common in oriental than other populations, and subarachnoid hemorrhage is most common in Finland and Sweden [15, 56].

Both paternal and maternal history of stroke are associated with an increased stroke risk [3, 57–60]. It is not necessarily “stroke genes” that are behind this familial aggregation, but one or more of the mechanisms that may contribute to it, such as (1) familial occurrence of risk factors for stroke, (2) genetic susceptibility to these risk factors, (3) familial sharing of environmental/lifestyle factors associated with stroke, and (4) the interaction between genetic and environmental effects [58–60]. Currently, rapid advances in genetic research are taking place and have resulted in the identification of genes associated with stroke and its subtypes. Low birth weight is another risk factor for stroke [61, 62], as it is for cardiovascular disease in general. Although these risk factors themselves cannot be modified, it does not mean that the stroke risk in such individuals could not be modified. In people with non-modifiable risk factors for stroke, it is particularly important to pay attention to the control of modifiable risk factors in order to prevent acute stroke events. Also, screening for family history of stroke can provide an opportunity for earlier detection and management of modifiable risk factors [63].

There are several well-documented medical conditions and diseases that have importance as risk factors for stroke. These are described in the next chapter (Chapter 7) by Brainin et al. In this chapter, some general observations are made on lifestyle factors and their relative importance for stroke incidence or recurrence is reported.

In the WHO MONICA Project, repeated population surveys of cardiovascular risk factors and continuous monitoring of stroke events was conducted in 35–64-year-old people over a 7–13-year period in 15 populations in nine countries. Stroke trends were compared with trends in individual risk factors and their combinations [64]. A 3–4-year time lag between changes in risk factors and change in stroke rates was considered. Population-level trends in systolic blood pressure showed a strong association with stroke event trends in women, but there was no association in men. In women, 38% of the variation in stroke event trends was explained by changes in systolic blood pressure. Combining trends in daily cigarette smoking, serum cholesterol, and body-mass index with systolic blood pressure into a risk score explained only a small additional fraction of the variation in stroke event trends.

Prediction of Stroke in Patients with TIA

Ischemic stroke is often preceded by early symptoms, i.e. a transient ischemic attack (TIA) [65]. Whereas early definitions of TIA were based on the duration of clinical symptoms (less than 24 hours) recent studies using modern brain imaging showed that the time period of underperfusion leading to permanent injury is highly variable, and that 30–50% of patients with symptoms lasting less than 24 hours had relevant ischemic lesions. TIA and ischemic stroke can be seen as a continuum and TIA should thus be seen as a warning sign for acute stroke and as an opportunity for prevention [3, 66]. The risk of an acute stroke event after a TIA attack has been previously underestimated due to issues in study designs [67, 68]. Hospital- and population-based cohort studies have reported 7-day risks of stroke of up to 10% [69–74], and the increased risk for stroke and cardiovascular events was found to be sustained for at least 5 years after TIA occurred [75]. However, a substantial international variation exists as to how patients with suspected TIA are managed in the acute phase. Therefore, models with predictors for long-term risk of stroke after TIA or minor stroke have been developed [69–72, 76–78]. Simple risk scores to assess high versus low stroke risk in TIA patients are clinically useful for the decision regarding which patients need urgent treatment.

The most important clinical symptoms associated with TIA were motor and speech symptoms and long duration of TIA symptoms. Because the diagnosis of TIA is largely clinical and is dependent on patients’ reliability in reporting the event, these symptoms might also be those symptoms reported most accurately.

Early scoring systems such as the California score [69], the ABCD score [79], and the ABCD2 [80] score were based on stroke risk factors and clinical symptoms. One of the first scores, the ABCD, has been developed and validated by Rothwell et al. to predict stroke during the first 7 days after a TIA attack [79]. A six-point score derived (Age [>60 years = 1], Blood pressure [systolic ≥140 mmHg and/or diastolic ≥90 mmHg = 1], Clinical features [unilateral weakness = 2, speech disturbance without weakness = 1, other = 0], and Duration of symptoms in minutes [≥60 = 2, 10–59 = 1, <10 = 0]) was highly predictive of 7-day risk of stroke in patients with probable or definite TIA (p <0.0001), in the Oxford Vascular Study population-based cohort of all referrals with suspected TIA (p <0.0001), and in the hospital-based weekly TIA clinic-referred cohort (p = 0.006). Subsequently the ABCD score and the California score were unified resulting in the ABCD2 score which added diabetes to the ABCD score [80]. While the ABCD2 score has been recommended by many guidelines for risk stratification of TIA/minor stroke, it may have some limitations. The ABCD2 has low specificity and seems not to identify patients with carotid stenosis or atrial fibrillation needing urgent intervention. A meta-analysis of 29 studies (13766 TIA patients) showed a sensitivity of 86.7%, but only a specificity of 35.4% for recurrent stroke within 7 days for patients with an ABCD2 score ≥4 (medium to high risk). Furthermore, 20% of patients with an ABCD2 <4 (low risk) had a >50% carotid stenosis or atrial fibrillation [81].

Therefore, new scores (e.g. CIP-model, ASPIRE approach, ABCD2-I, ABCD3, ABCD3-I, ABCDE+) were developed adding imaging or etiology to the ABCD2 score. For example, the ABCD2-I added brain imaging only, whereas the ABCD3-I includes diffusion-weighted imaging (DWI), carotid stenosis >50%, and a TIA occurring within 1 week of the first. A pooled analysis of individual patient data comparing the validity of the ABCD2, the ABCD2-I, and the ABCD3-I in 2 176 patients from 16 cohort studies concluded that both the ABCD2-I and the ABCD3-I had better predictive ability than the ABCD2 score, and that the ABCD3-I was superior to the ABCD2-I [82]. The ABCDE+ score which includes point-weighting for etiology was shown to be superior to the ABCD2 score [83].

The changes in the definition of TIA from time-based (symptom duration below/above 24 hours) to tissue-based (without/with corresponding brain lesion on CT or MRI) may question the validity of the ABCD3-I risk score. However, a retrospective data analysis showed that the ABCD3-I score performed equally in TIA patients in tissue- as well as time-based definition [84].

A disadvantage of these new approaches is that they require new brain imaging techniques and diagnostic equipment which are not available in all countries and all clinics where patients with TIA are treated.