Incidence, Types, Risk Factors, and Outcome of Stroke in a
Developing Country
The Trivandrum Stroke Registry
Sapna E. Sridharan, MD; J.P. Unnikrishnan, MPhil; Sajith Sukumaran, MD; P.N. Sylaja, MD;
S. Dinesh Nayak, MD; P. Sankara Sarma, PhD; Kurupath Radhakrishnan, MD
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
Background and Purpose—Despite increasing burden of stroke in developing countries, population-based data are rare.
Through the Trivandrum Stroke Registry, we intend to assess incidence, types, risk factors, and outcome of stroke
among urban and rural dwellers of a South Indian community.
Methods—We ascertained all first-ever strokes occurring among 741 000 urban and 185 000 rural inhabitants of Trivandrum,
Kerala. In addition to Steps 1 and 2 of World Health Organization STEPS Stroke Manual, we used multiple supplementary
methods to maximize ascertainment of nonfatal and nonhospitalized fatal stroke events in the community.
Results—During a 6-month period, 541 strokes were registered, 431 in the urban and 110 in the rural communities. Stroke
occurred at a median age of 67 years; only 3.8% of patients were aged ⱕ40 years. Adjusted annual incidence rates
per 100 000 were 135 (95% confidence interval 123 to 146) for total, 135 (122–148) for urban, and 138 (112–164) for
rural populations, and 74.8 (66.3 to 83.2), 10.1 (7.0 to 13.2), and 4.2 (2.2 to 6.1) for ischemic stroke, intracerebral
hemorrhage, and subarachnoid hemorrhage, respectively. There was more stroke of undetermined type in the rural
community. One or more modifiable risk factors were identified in 90% patients. More rural male patients smoked
tobacco. The 28th day case fatality rate was 24.5% for urban and 37.1% for rural populations (P⫽0.011).
Conclusions—There are more similarities than differences between developing and developed countries in the
epidemiology of stroke. Compared to urban stroke patients, rural ones are less likely to be optimally investigated and
treated. (Stroke. 2009;40:1212-1218.)
Key Words: epidemiology 䡲 incidence 䡲 outcome 䡲 risk factors 䡲 stroke
S
troke is a global health problem and is a leading cause of
adult disability.1 Of 35 million deaths attributable to
chronic noncommunicable diseases that occurred worldwide
in 2005, stroke was responsible for 5.7 million (16.6%)
deaths, and 87% of these deaths occurred in low-income and
middle-income counties.2 Driven by increasing size and aging
of populations, and escalating prevalence of risk factors such
as hypertension, tobacco use, unhealthy diet, physical inactivity, and obesity, stroke is becoming a major cause of
premature death and disability in developing countries.3 This
prompted the World Health Organization (WHO) to launch
the Global Stroke Initiative aimed to generate populationbased data on burden of stroke and to use such data to evolve
strategies for prevention and management.4 The feasibility of
WHO STEPS Stroke Surveillance System (STEPS Stroke),5
which provides a standardized approach to stroke data collection, analysis, interpretation, and dissemination, has been
successfully tested in several developing countries.6
India, with more than 1 billion inhabitants, is undergoing
remarkable economic and demographic changes in recent
years resulting in a transition from poverty-related infectious
and nutritional deficiency diseases toward lifestyle-related
cardiovascular and cerebrovascular diseases.7,8 Despite rapid
economic boom, a large segment of the Indian population still
lives in poverty. Given the anticipated increase in burden of
stroke in coming years and limited availability of organized
stroke care services to the majority of people in India, it
would be logical to place greater emphasis on populationbased stroke prevention strategies. However, very little reliable information is currently available regarding epidemiology of stroke in India. The reported frequency, pattern, risk
factors, and outcome of stroke from India are largely derived
from hospital-based observations.8,9 Very few epidemiological studies on stroke that are available from India10 –12 do not
fulfill the criteria for an ideal study of stroke incidence.13
The Trivandrum Stroke Registry intends to overcome
this deficiency. Using WHO STEPS Stroke protocol, we
intensively monitored first-ever-in-a-lifetime (first-ever)
stroke occurrences among urban and rural dwellers of a
Received July 16, 2008; final revision received August 30, 2008; accepted October 1, 2008.
From the Department of Neurology and Achutha Menon Center for Health Science Studies, Sree Chitra Tirunal Institute for Medical Sciences and
Technology, Trivandrum, Kerala, India.
Correspondence to Kurupath Radhakrishnan, MD, Professor and Head, Department of Neurology, Sree Chitra Tirunal Institute for Medical Sciences
and Technology, Trivandrum – 695 011, Kerala, India. E-mail [email protected]
© 2009 American Heart Association, Inc.
Stroke is available at http://stroke.ahajournals.org
DOI: 10.1161/STROKEAHA.108.531293
1212
Sridharan et al
Epidemiology of Stroke in Kerala, India
1213
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
Figure. Map of India, the state of Kerala, and the areas surveyed (shaded) in Trivandrum (Thiruvananathapuram). SCTIMST indicates
Sree Chitra Tirunal Institute for Medical Sciences and Technology.
South Indian community to ascertain information on incidence, types, risk factors, morbidity, and mortality of
stroke.
Subjects and Methods
Study Area
The southern state of Kerala covers an area of 39 000 Km2 spread
over 14 districts along the western coast of Indian peninsula (Figure)
and is inhabited by more than 32 million people, who are distin-
guished from the rest of India by the high level of literacy (91%)
and health awareness.14 The life expectancy of Kerala population
is 71 years for males and 76 years for females.15 The per capita
domestic product of Kerala state at current prices is Indian
Rupees 39 315 (U.S. dollar 983).16 This study was conducted in
Trivandrum (Thiruvananathapuram in the regional language
Malayalam), the southernmost district of Kerala (Figure), under
the supervision of Sree Chitra Tirunal Institute for Medical
Sciences and Technology, a medical institution of national
importance for cardiovascular and neurological disorders situated
in Trivandrum city.
1214
Stroke
April 2009
The Trivandrum Stroke Registry
The Trivandrum Stroke Registry is based on the WHO STEPS
Stroke Manual.5 STEPS Stroke consists of gathering information
using standardized questionnaire about hospitalized stroke events
(Step 1), fatal stroke events in the community (Step 2), and nonfatal
stroke events in the community (Step 3).5 In addition to Steps 1 and
2, we used multiple overlapping supplementary methods as described below to maximize ascertainment nonfatal and nonhospitalized fatal stroke events in the community.
Preparatory Phase
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
During January and February 2005, we defined the geographical area
of the study and its population, and all medical care facilities existing
in the study area. We choose an urban area of Trivandrum Corporation covering 140 km2 with a population of 741 307 (363 704
males and 377 603 females) and a rural area covering 60 km2
situated 20 km away from the Trivandrum City comprising a
population of 184 560 (92 067 males and 92 493 females; Figure).
The medical care facilities were personally visited by the principal
investigator (K.R.) along with the other 2 neurologists involved with
the study (P.N.S., S.D.N.) and a medical social worker (J.P.U.) to
ensure cooperation of local physicians and to decide on the frequency of hospital visits to be undertaken for case ascertainment. We
then classified the medical facilities in the study area according to the
number of stroke cases encountered by them into those which needed
to be visited daily (ⱖ6 stroke admissions per week); those which
needed to be visited twice a week (3 to 5 stroke admissions per
week), and those which could be visited once a week or on call (ⱕ2
stroke admissions per week). Two research assistants and a medical
social worker once a week contacted each general practitioner in the
urban and rural study areas and the hospitals bordering them to
gather information about possible cases of stroke. Three brain
imaging centers in the urban study area were accessed weekly as a
supplementary search strategy. In addition, patients attending the
physiotherapy and rehabilitation facilities at the Trivandrum Medical
and Ayurveda Colleges were also monitored to ascertain incident
stroke cases. One of the neurologists (K.R., P.N.S., or S.D.N.)
accompanied the research assistants during their visits periodically to
supervise data collection and to verify accuracy of the data collected.
The study had the approval of the institutional review boards of all
participating hospitals.
To ascertain fatal stroke events, we scrutinized the death registry
maintained by the Trivandrum Corporation Office to which more
than 95% of all deaths occurring in the community are registered. On
scanning through the death registrations, we noticed that nearly 90%
of the deaths reported were from the hospitals and 10% were
domiciliary deaths. Whereas for deaths reported from the hospital,
the cause of death was evident, for deaths occurring at homes, the
cause was often obscure. We authorized the medical social worker
(J.P.U.) to visit the houses of all patients whose cause of death was
uncertain in the death certificate and to conduct a verbal autopsy
using a structured questionnaire to identify the cause of death, after
obtaining informed consent from a senior member of the household.
Case Ascertainment Phase
After test running the procedures during March 2005, we ascertained
incident cases of first-ever stroke among the study population during
the 6-month period from April 1, 2005 to September 30, 2005. We
used the WHO definition of stroke as “a focal (or at times global)
neurological impairment of sudden onset, and lasting ⱖ24 hours (or
leading to death), and of presumed vascular origin.”5 This definition
excluded transient ischemic attack (TIA) and recurrent stroke.
Diagnosis of stroke was based on information obtained from the
patient, caregiver, or hospital records. Based on the noncontrast
computed tomography (CT) scan findings, types of stroke was
categorized as ischemic stroke (IS), intracerebral hemorrhage (ICH),
subarachnoid hemorrhage (SAH), or undetermined (no brain imaging).13 We used the following definitions for major stroke risk
factors17: hypertension, systolic blood pressure (BP) ⱖ140 mm Hg,
diastolic BP ⱖ90 mm Hg, or any treatment for high BP; diabetes
mellitus, fasting plasma glucose ⱖ126 mg/dL or single value ⱖ200
mg/dL, previous diagnosis and treatment; smoking, current use of
inhaled tobacco; and hypercholesterolemia, previous diagnosis, receiving lipid lowering drugs, or total cholesterol ⱖ240 mg/dL or low
density lipoprotein ⱖ160 mg/dL. In this part of India, tobacco
smoking among females is nonexistent. Past history of TIA and
ischemic heart disease could not be reliably obtained. Disability
status at 28th day was measured using Modified Rankin Scale18
obtained either by telephonic interview, or hospital or home visits by
the medical social worker.
Verification Phase
We verified the residential status of each case, and ⱖ1 year
residency in the study area was considered as eligibility for inclusion. At least 2 of the 3 study neurologists (K.R., P.N.S., S.D.N.)
reviewed each case data to verify the diagnosis of stroke, and
completeness and accuracy of the information collected, and to avoid
duplication of cases collected through multiple sources. Problem CT
scans were discussed with neuroradiologists. The death certificates
were carefully scrutinized, and supporting evidence for a stroke
death was obtained from hospital records. The medical social worker
(J.P.U.) visited the houses of all unattended domiciliary deaths
during the study period (excluding those attributable to accidents and
suicides) to conduct verbal autopsies. The verification process
continued until August 2006, when the data were considered suitable
for analyses.
Statistical Methods
The 2001 Indian official census provided the population data.14 The
incidence rates were adjusted by direct method using WHO World
Standard Population, which has been constructed based on the
projected average world population age-structure for the period 2000
to 2025.19 We calculated 95% confidence intervals (CI) of crude and
adjusted incidence rates assuming a Poisson distribution for observed
rates. To compare data from urban and rural populations, we used
Student t test, Fisher exact test, and Pearson ␹2 test, as appropriate. A
probability value of ⱕ0.05 was considered significant.
Results
Incidence Data
Of the 541 validated first-ever stroke events ascertained
during the study period, 431 occurred in the urban community
and 110 occurred in the rural community. Of them, 36 (8.4%)
urban patients and 24 (21.8%) rural patients were ascertained
by verbal autopsy. From the rural community, whereas 52
(47.3%) patients were ascertained from hospitals outside the
survey area, 21 (19.1%) patients had seen only the general
practitioners and were not hospitalized. Table 1 shows crude
and age- and sex-specific and adjusted stroke incidence rates.
The crude annual incidence per 100 000 for all strokes was
117 for the total population (115 for males and 119 for
females), 116 for the urban population, and 119 for the rural
population. The age-adjusted incidence rates per 100 000 per
year were 135 (95% CI 123 to 146) for total, 143 (125–160)
for male and 128 (113–143) for female, and 135 (122–148)
for urban and 138 (112–164) for rural populations.
Age Distribution
The median age of stroke patients was 67 years, which did not
differ between urban and rural populations. Seventy-seven
percent of patients were aged ⱖ60 years. Stroke in the young,
defined as first stroke occurring at ⱕ40 years, comprised only
18 (3.8%) patients in this population-based study. The incidence rates increased steadily with age for both genders in
urban and rural communities (Table 1).
Sridharan et al
Table 1.
Epidemiology of Stroke in Kerala, India
1215
Incidence Rates of Stroke per 100 000 Inhabitants per Year in Trivandrum, Kerala, India
Urban Community
Males
Females
Population
Cases
Population
Cases
Rate
89 306
0
0
0
0
25 618
0
0
23 388
0
0
0
0
15–24
67 205
1
3.2
68 970
0
0
1
1.6
17 572
0
0
17 315
0
0
0
0
25–34
60 552
1
3.5
66 496
5
15.9
6
10.1
14 603
0
0
15 821
1
14.1
1
7.4
35–44
52 015
7
28.8
54 880
8
30.9
15
29.9
12 403
1
18.2
12 868
2
34.6
3
26.6
45–54
41 996
22
106.9
40 985
17
82.8
39
94.9
9764
9
185.1
9361
4
95.2
12
141.0
55–64
24 737
45
363.1
26 930
45
330.6
90
346.0
6028
15
487.2
6395
10
313.5
22
397.3
65–74
15 430
76
983.9
19 694
76
775.0
152
865.7
4046
24
1173
4741
18
751.6
37
943.4
7001
48
1371.3
10 341
78
1,508.5
126
1462.3
2041
12
1176
363 704
201
110.5
377 603
230
121.6
431
116.4
92 076
61
141.7 (122.1–161.3)
Cases
Rate
Cases
130.1 (113.3–146.9)
Rate
135.0 (122.0–148.0)
Rate
Both Genders
0
Age adjusted rates
(95% CI)
Population
Females
0
All ages
Rate
Males
94 769
ⱖ75
Cases
Both Genders
⬍15
Age-Group (yr)
Population
Rural Community
131.6
Cases
Rate
2603
13
998.9
23
1116.2
92 493
49
105.2
110
119.4
163.4 (122.4–204.4)
115.3 (83.0–147.6)
138.0 (112.0–164.0)
CI indicates confidence interval.
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
Stroke Types
CT brain was available in 344 (63.6%) and MRI in 33
patients (6.1%). We could categorize strokes into IS, ICH,
and SAH in 372 (68.8%) patients, which were diagnosed in
83.6%, 11.6%, and 4.8%, respectively. The distribution of
these stroke types did not differ between the urban and rural
Table 2. Distribution of Patients According to Stroke
Characteristics Compared Between Urban and Rural
Communities in Trivandrum, Kerala, India
populations (Table 2). Stroke undetermined type was more in
the rural population (rural 43.6% versus urban 28.1%,
P⫽0.003) related to less frequent availability of brain imaging (rural 56.4% versus urban 71.9%, P⫽0.018). The incidence rates according to stroke types are shown in Table 3.
Age adjusted incidence rates for the total (urban plus rural)
population per 100 000 per year for IS, ICH, and SAH were
74.8, 10.1, and 4.2, respectively.
Risk Factor Distribution
Urban (n⫽431)
n (%)
Rural (n⫽110)
n (%)
Significance
P
Available
310 (71.9)
62 (56.4)
0.003
Not available
121 (28.1)
48 (43.6)
IS
260 (83.9)
51 (82.3)
NS
ICH
33 (10.6)
10 (16.1)
NS
SAH
17 (5.5)
1 (1.6)
NS
The distribution of conventional stroke risk factors was
remarkably similar among the urban and rural communities
(Table 2). Hypertension was the most frequent and occurred
in 450 (83.2%) patients. Seventy of 261 male patients
(26.8%) smoked tobacco, whereas none of the female patients
smoked. Compared to urban males, more rural males smoked
tobacco (22.8% versus 39.3%, P⫽0.013). Half of the patients
had diabetes mellitus and 26% had dyslipidemia. None of the
risk factors was seen in only 28 (5%) patients. One risk factor
was present 208 (38.4%) patients, two in 227 (42.0%), and
three or more in 78 (14.4%) patients.
Hypertension
352 (81.6)
98 (89.0)
NS
Patient Outcomes
Diabetes mellitus
210 (48.7)
61 (55.7)
NS
Smoking (males)
46 (22.8)
24 (39.3)
0.013
109 (25.3)
29 (26.7)
NS
37 (8.7)
5 (4.9)
NS
Mortality
One hundred forty-seven patients did not survive beyond the
28th poststroke day - fatality rate 27.2% (24.5% for urban and
37.1% for rural population, P⫽0.011; Table 2). One hundred
six (72.1%) of these deaths occurred within 10 days of stroke
onset.
Characteristics
Brain imaging
Stroke types, by brain
imaging
Risk factors
Dyslipidemia
Atrial fibrillation
No. of risk factors
None present
22 (5.1)
6 (5.9)
NS
One only
173 (40.1)
35 (31.7)
NS
Two
181 (42.1)
46 (41.6)
NS
Three or more
55 (12.7)
23 (20.8)
0.034
106 (24.5)
41 (37.1)
0.011
Mild
126 (42.3)
19 (43.2)
NS
Moderate
126 (42.3)
21 (47.7)
NS
46 (15.4)
4 (9.1)
NS
28th day case fatality
28th day disability (n 342)
Severe
IS indicates ischemic stroke; ICH, intracerebral hemorrhage; SAH, subarachnoid hemorrhage; NS, not significant, n, No. of patients.
Disability
For the 394 patients, who survived beyond the 28th day of
stroke onset, functional outcome was available in 342
(86.8%) patients. Mild disability (Rankin score ⱕ2) was
observed in 145 (42.4%) of them. Whereas 50 (14.6%)
patients were bedridden (Rankin score 5), 147 (43%) were
moderately disabled (Rankin score 3 or 4; Table 2). There
was no significant difference in the functional outcome
between males and females (P⫽0.179), between urban and
rural patients (P⫽0.515), or across different age groups
(P⫽0.526).
1216
Stroke
April 2009
Table 3.
Incidence Rates per 100 000 per Year According to Stroke Type in Trivandrum, Kerala, India
IS
Age-Group (yr)
Cases
SAH
Rate
95% CI
Cases
Rate
95% CI
Cases
Rate
95% CI
3.7
2.6 –5.3
3
0.9
0.4 –1.7
2
0.6
0.2–1.3
65.4
51.3–82.3
7
14.3
8.6–22.7
2
4.1
1.7–9.0
219.3
185.7–257.4
4
12.7
6.5–22.9
5
15.9
8.7–27.2
521.5
458.4–591.0
11
49.9
33.1–72.8
4
18.1
9.3–32.7
72
627.9
533.5–734.5
15
130.8
91.8–181.6
4
34.9
17.9–62.9
301
65.0
59.8–70.2
40
8.6
6.8–10.5
17
3.7
2.4–4.9
74.8
66.3–83.2
10.1
7.0–13.2
4.2
2.2–6.1
ⱕ44
13
45–54
32
55–64
69
65–74
115
ⱖ75
Total
Age–adjusted
ICH
IS indicates ischemic stroke; ICH, intracerebral hemorrhage; SAH, subarachnoid hemorrhage; CI, confidence interval.
Discussion
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
This WHO-sponsored 6-month survey was intended to explore the feasibility of establishing a stroke registry in a
developing region. Implementing STEPS Stroke beyond Step
1 (hospital-based case ascertainment) is considered to be a
challenge in developing countries because of poor accessibility to health care facilities to majority of population, underreporting of deaths in the community, and unreliability of
death certificate diagnosis.4,20 We believe that the Trivandrum Stroke Registry to a great extent overcame these
hurdles by the following means. First, we ascertained incident
strokes from both urban and rural populations in a geographical region with well-developed primary and secondary medical care facilities. Second, in addition to intensive surveillance of all hospitals where stroke patients are likely to get
admitted, we contacted medical practitioners, and brain imaging centers and physiotherapy facilities in the study area to
ascertain nonhospitalized stroke events. It has been shown
that these supplementary methods can lead to near-complete
case ascertainment.21 Compared to other developing regions,
chance of a patient from the highly literature and health
conscious Kerala population, with stroke related symptoms
lasting beyond 24 hours, not to seek medical consultation is
very unlikely. Third, in the geographical region where nearly
all deaths in the community get registered, through verbal
autopsy, we could verify stroke related deaths in instances
where death certificate diagnosis was uncertain. Fourth, at
least 2 experienced neurologists reviewed the clinical and
imaging data to achieve diagnostic accuracy of stroke and its
type. All 3 incidence studies on stroke from India were
undertaken in relatively small populations from northeastern
part of the country.10 –12 The methodology of 6-month houseto-house surveys used in 2 of these studies was more suited
for prevalence estimation of stroke rather than incidence, and
data were not presented in mid-decade age bands.11,12 Furthermore, only 10% of death certificates were available for
investigators’ review in the most recently reported survey
from the city of Kolkata, West Bengal.12
The wide geographical variation in the epidemiological
indices of stroke around the world can be real (related to
differences in genetic susceptibility and exposure to risk
factors) or apparent (attributable to differences in availability
and accessibility to medical care and imaging facilities).
Furthermore, there are marked differences between stroke
epidemiological studies in definition of stroke, size of population surveyed, methods of case ascertainment, and quality
of data ascertained and reported. Feigin et al13 reviewed 15
population-based incidence studies on stroke (hereafter called
Feigin review) from 13 countries that more or less met the
requirements of ideal epidemiological studies such as welldefined population denominator, diagnostic accuracy, and
nearly complete case ascertainment. In Feigin review, agestandardized annual incidence of all types of stroke per 1000
population aged ⱖ55 years ranged from 4.2 to 6.5, and
age-specific incidence of stroke increased progressively with
each decade of life approaching rates of 12.0 to 20.0 for those
aged 75 to 84 years.13 In Trivandrum, stroke occurred at rate
of 7.1 per 1000 per year in people ⱖ55 years, and the rate
escalated to 13.3 for the ⱖ75 years age-group. The mean age
at stroke onset at Trivandrum of 65 years males and 67 years
females is within the ranges of 60.8 to 75.3 years for males
and 66.6 to 78.0 years for females cited in the Feigin
review.13 Similarly, the frequency of IS, ICH, and SAH in
Trivandrum of 83.6%, 11.6%, and 4.8%, respectively, are
more or less within the ranges of 67.3% to 80.5%, 6.5% to
19.6%, and 0.8% to 7.0%, respectively, reported in Feigin
review.13 Stroke undetermined type (no brain imaging) was
31.2% in Trivandrum study compared to 2.0% to 14.5% in
Feigin review. The case fatality of total strokes in Trivandrum
of 27.2% did not vary beyond the range of 17% to 33% noted
in Feigin review.13 Despite differences in methods of data
collection, the 3 previous stroke incidence studies from
eastern India have shown results more or less in agreement
with ours. Nearly 60% of our patients were significantly
disabled (Rankin score ⱖ3) 1 month poststroke, which is very
similar to the outcome of patients from the Polish National
Stroke Registry.22 Our inability to find more disability among
older patients may be related to relatively small number of
very old patients.
It is believed that the average age of patients with stroke in
developing countries is 15 years younger than that in developed countries.4,23 In India, nearly one-fifth of patients with
first ever strokes admitted to hospitals are aged ⱕ40 years.9
In a recent hospital-based study of stroke from sub-Saharan
Africa, the mean age of patients was 58 years.24 High
prevalence of stroke risk factors leading onto premature
atherosclerosis, unreliability of hospital statistics in estimating the age distribution of stroke occurrences in the commu-
Sridharan et al
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
nity, and population structure with smaller proportion of
people ⱖ60 years have been implicated as reasons for the
increased occurrence of stroke in the young in developing
countries.23,25 In the Trivandrum Stroke Registry, only 3.8%
of incident strokes occurred in people aged ⱕ40 years, 9.5%
aged ⱕ50 years, and 18.1% aged ⱕ55 years; these data are
very similar to that in another community-based study from
northeastern India12 and those from developed countries.13
The reported occurrence of stroke in the young in developed
countries, therefore, appears to be largely an artifact related to
hospital-based case ascertainment.
It has been estimated that hypertension causes 54% of
stroke mortality in low-income and middle-income countries,
followed by high cholesterol (15%) and tobacco smoking
(12%).2 Among our community-based stroke patients, nearly
85% had hypertension, half had diabetes mellitus, a quarter
had dyslipidemia, and one-fifth of males smoked tobacco.
Because we inquired only about current smoking, we could
have underestimated the real contribution of this risk factor.
In a case– control study from the same geographical region,
compared to community-based controls, young stroke subjects were 8 times more likely to have had smoked tobacco.17
Only 5% of Trivandrum Stroke Registry patients did not have
any modifiable risk factors, whereas more than half of them
had cooccurrence of ⱖ2 risk factors. However, in the absence
of reliable data on prevalence of above risk factors in the
Trivandrum population, we are neither able to calculate the
attributable risks nor able to comment on the causal role of
these risk factors in our patient population.
Very few studies on stroke have focused on the urban-rural
dichotomy.26,27 It is uncertain as to whether the urban-rural
differences in the epidemiological indices of stroke observed
in these studies are real or could be explained by the
dissimilarities in the quality of the data, which are more
difficult to obtain reliably in a rural setting compared to an
urban setting. Studies from India on cardiovascular risk
factors have shown a 2 to 3 times high prevalence of
hypertension, hyperlipidemia, obesity, diabetes mellitus, and
smoking (in males) in urban compared to rural communities.28,29 Because of the demographic transition occurring in
majority of developing countries in recent years, the urbanrural difference in the risk factor distribution is fast disappearing. In Trivandrum, more rural males with stroke smoked
tobacco. There were more strokes undetermined type in the
rural community because of a lack of availability of imaging
information. The stroke incidence did not differ between
urban and rural communities. Significantly more of our rural
patients compared to urban died within 1 month, probably
reflecting inadequate facilities in the rural area to manage
them during the early poststroke period.
We acknowledge the following limitations of our study.
First, with the research fund allocated for this feasibility
study, intensive surveillance was possible only for a 6-month
period. We could not account for any seasonal variation in
incidence of stroke, as our study did not run for a complete
1-year period. Second, patients with trivial strokes, particularly in the rural area, might not have been ascertained
through any of the facilities we utilized. Third, to ensure a
uniform data collection, we ascertained only selected consis-
Epidemiology of Stroke in Kerala, India
1217
tently ascertainable risk factors and could not obtain reliable
information on some of the important ones like ischemic and
valvular heart diseases. Fourth, our incidence rates for the
stroke types are likely to be underestimates because of the
fact that one-third of patients had no brain imaging. Moreover, our inability to show any difference in the incidence
rates of ICH and SAH between the urban and rural communities may be attributable to the small number of patients with
these stroke types. Fifth, the questionnaire we used for verbal
autopsy remains to be validated. Sixth, we used the 2001
census data to calculate stroke incidence rates in 2005.
Because of very successful implementation of family planning measures, Kerala has achieved a near zero population
growth in recent years.30 Therefore, it is unlikely that trivial
change in the Kerala population demographics during the
brief period of 2001 to 2005 would have affected our results.
Last, as India is a vast country with marked socioeconomic
diversity, our results cannot be extrapolated to other parts of
the country, maybe not even to other southern regions.
However, despite these limitations, as elaborated in the
second paragraph of discussion, our incidence and mortality
indices are in agreement with those from ideal stroke epidemiological studies from developed countries analyzed in
Feigin review.13 The multi-center longitudinal populationbased national stroke registry, which the Indian Council of
Medical Research is planning to establish, may address the
above listed deficiencies of the present study.
Conclusions
This epidemiological study illustrates that there are more
similarities than differences between developing and developed countries in incidence, age distribution, types, mortality,
and morbidity of stroke. We could not substantiate the usually
held notion that stroke occurs more in the younger age group
in low-income counties. Compared to urban stroke patients,
rural ones were less likely to be optimally investigated and
treated. Nearly 90% of stroke victims had one or more
modifiable risk factors. The necessity for educating the
people and health care professionals about implementing
effective stroke preventive strategies cannot be overemphasized. The information that nearly two-thirds of stroke survivors were significantly disabled at 1 month after stroke
demands better care during acute stage to minimize disability
and long-term strategies to rehabilitate the disabled.
Acknowledgments
We thank the people, physicians, and hospital administrators in the
study area for their help in gathering the data.
Sources of Funding
The study was supported by a grant from the South East Asia
Regional Office of the World Health Organization, New Delhi,
India.
Disclosures
None.
References
1. Donnan GA, Fisher M, Macleod M, Davis SM. Stroke Lancet. 2008;371:
1612–1623.
1218
Stroke
April 2009
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
2. Strong K, Mathers C, Bonita R. Preventing stroke: saving lives around the
world. Lancet Neurol. 2007;6:182–187.
3. World Health Organization. The World Health report 2004: changing
history. Geneva: WHO, 2004.
4. Bonita R, Mendis S, Truelsen T, Bogousslavsky J, Toole J, Yatsu F. The
global stroke initiative. Lancet Neurol. 2004;3:391–393.
5. WHO STEPS Stroke Manual. http://www.who.int/chp/steps/Stroke/
en.pdf, last accessed on July 2, 2008.
6. Truelsen T, Heuchmann PU, Bonita R, Arjundas G, Dalal P, Damasceno
A, Nagaraja D, Ogunniyi A, Oveisgharan S, Radhakrishnan K,
Skvortsoya VI, Stakhovakaya V. A standard tool for developing stroke
registers in low and middle income countries: experience from a feasibility study of a stepwise approach to stroke surveillance (STEPS Stroke).
Lancet Neurol. 2007;6:134 –139.
7. Reddy KS. Cardiovascular diseases in non-western countries. N Engl
J Med. 2004;350:2438 –2440.
8. Dalal PM. Burden of stroke: Indian perspective. Int J Stroke. 2006;1:
164 –166.
9. Pandian JD, Padma V, Vijaya P, Sylaja PN, Murthy JMK. Stroke and
thrombolysis in developing countries. Int J Stroke. 2007;2:17–26.
10. Banerjee TK, Mukherjee CS, Sarkhel A. Stroke in the urban population
of Calcutta – an epidemiological study. Neuroepidemiology. 2001;20:
201–203.
11. Bhattacharya S, Saha SP, Basu A, Das SK. A 5 years prospective
incidence, mortality and morbidity profile of stroke in a rural community
in eastern India. J Indian Med Assoc. 2005;103:655– 659.
12. Das SK, Banerjee TK, Biswas A, Roy T, Raut DK, Mukherjee CS,
Chaudhuri A, Hazra A, Roy J. A prospective community-based study of
stroke in Kolkata, India. Stroke. 2007;38:906 –910.
13. Feigin VL, Lawes CMM, Bennett DA, Anderson CS. Stroke epidemiology: a review of population-based studies of incidence, prevalence, and
case-fatality in the late 20th century. Lancet Neurol. 2003;2:43–53.
14. Census of India 2001. http://www.censusindia.gov.in/Census_
Data_2001/ States_at_glance/ State_Links/32_ker.pdf, last accessed on
July 2, 2008.
15. Government of India. Family welfare statistics in India – 2006. New
Delhi: Ministry of Health and Family Welfare, 2007;A-21.
16. Kerala State Planning Board. Economic Review 2007. http://www.
Keralaplanning board.org/Economic_Review_2007, last accessed July 2,
2008.
17. Lipska K, Sylaja PN, Sarma PS, Thankappan KR, Kutty VR, Vasan RS,
Radhakrishnan K. Risk factors for acute ischaemic stroke in young adults
in South India. J Neurol Neurosurg Psychiatry. 2007;78:959 –963.
18. de Hann R, Limburg M, Bossuyt P, van der Meulen J, Aaronson N. The
clinical meaning of Rankin “handicap” grade after stroke. Stroke. 1995;
26:2027–2030.
19. Ahmad OB, Boschi-Pinto C, Lopez AD, Murray CJL, Lozano R, Inoue
M. Age standardization of rates: a new WHO standard. GPE Discussion
Paper Series: No. 31. EIP/GPE/EBD, World Health Organization. http://
www.who.int/infobase/publicfiles/who_standardpopulation_discussion31.
pdf, last accessed July 3, 2008.
20. Truelsen T, Bonita R, Jamrozik K. Surveillance of stroke: a global
perspective. Int J Epidemiol. 2001;30:S11–S16.
21. Coull AJ, Silver LE, Bull LM, Giles MF, Rothwell PM, on behalf of the
Oxford Vascular (OXVASC) Study. Direct assessment of completeness
of ascertainment in a stroke incidence study. Stroke. 2004;35:2041–2047.
22. Niewada M, Skowronska M, Ryglewicz D, Kaminski B, Czlonkowska A,
on behalf of the Polish National Stroke Prevention and Treatment Collaborative Group. Stroke. 2006;37:1837–1843.
23. Poungvarin N. Stroke in the developing world. Stroke. 1998;352:19 –22.
24. Walker RW, Rolfe M, Kelly PJ, George MO, James OFW. Mortality and
recovery after stroke in the Gambia. Stroke. 2003;34:1604 –1609.
25. Radhakrishnan K, Ashok PP, Sridharan R, Mousa ME. Stroke in the
young: incidence and pattern in Benghazi, Libya. Acta Neurol Scand.
1986;73:434 – 438.
26. Correia M, Silva MR, Matos I, Magalhaes R, Lopes JC, Ferro JM, Silva
C. Prospective community-based study of stroke in northern Portugal.
Incidence and case fatality in rural and urban populations. Stroke. 2004;
35:2048 –2053.
27. Zhang X-H, Guan T, Mao J, Liu L. Disparity and its time trends in stroke
mortality between urban and rural populations in China 1987 to 2001.
Changing patterns and their implications for public health policy. Stroke.
2007;38:3139 –3144.
28. Singh RB, Sharma JP, Rastogi V, Raghuvanshi RS, Moshiri M, Verma
SP, Janus ED. Prevalence of coronary artery disease and coronary risk
factors in rural and urban populations of north India. Eu Heart J. 1997;
18:1728 –1735.
29. Goyal A, Yusuf S. The burden of cardiovascular disease in the Indian
subcontinent. Indian J Med Res. 2006;124:235–244.
30. Family Welfare Statistics in India-2006. Government of India, Ministry
of Health and Family Welfare, New Delhi, 2007.
Incidence, Types, Risk Factors, and Outcome of Stroke in a Developing Country: The
Trivandrum Stroke Registry
Sapna E. Sridharan, J.P. Unnikrishnan, Sajith Sukumaran, P.N. Sylaja, S. Dinesh Nayak, P.
Sankara Sarma and Kurupath Radhakrishnan
Downloaded from http://stroke.ahajournals.org/ by guest on June 15, 2017
Stroke. 2009;40:1212-1218; originally published online February 19, 2009;
doi: 10.1161/STROKEAHA.108.531293
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2009 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/40/4/1212
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Stroke is online at:
http://stroke.ahajournals.org//subscriptions/