From the Department of Biometry and Epidemiology (D.T.L.; D.L.Barker) and
the Department of Neurology (D.L.Bachman, T.D.C., S.T., H.K.), Medical
University of South Carolina, Charleston, SC.
Correspondence to Daniel T. Lackland, DrPH, Department of Biometry and Epidemiology, Medical University of South Carolina, Charleston, SC 29425-2203. E-mail lackland{at}musc.edu
Methods—Strokes were identified in the Anderson and Pee Dee areas
of South Carolina. All hospitalized and out-of-hospital deaths
occurring during 1990 among the residents of these 2 areas were
included. Strokes were classified by an independent panel of
neurologists using a standard protocol that included specific criteria
for stroke and subtypes.
Results—The overall age-adjusted stroke incidence rates (per
100 000 population) were significantly higher in the Pee Dee
population (293.1) compared with Anderson (211.2). The geographic
differences were more dramatic in the younger age groups of 35 to 64
years. Likewise, incidence rates for blacks were nearly twice
the rates for whites. The rates in the Pee Dee were higher than the
rates from other studies in the United States and other parts of the
world. Although the stroke subtypes did not vary between the 2 regions,
race-sex differences were identified.
Conclusions—High stroke incidence and disease rates persist for
all 4 race-sex groups in the Southeast and reflect similar risks as
mortality rates. However, geographic variability in stroke rates
suggests that the pattern of disease in the region is not so much a
"belt" of increased stroke in contiguous areas but rather more a
"necklace" of different levels of risk. These results should be
useful in the identification of factors associated with this geographic
enigma.
Stroke risks for this geographic region of the country have primarily
been based on mortality rates as identified by death certificates.
Although this approach is convenient because of the availability of
mortality data, the results of such analyses are most useable
to generate hypotheses. The limitations of death certificates severely
reduce their validity in drawing conclusions about the true estimates
of cerebrovascular disease incidence.13 14 In
addition, stroke mortality rates are affected by case-fatality
rates.15 16 The risk assessment of geographic
variation for cerebrovascular disease is further complicated by the
lack of stroke incidence data for the 2 primary ethnic populations of
the Southeast. Incidence rates are generally considered a more accurate
reflection of disease level than mortality rates, which are affected by
case fatality and survival.17 18 The methodology
for determining population stroke incidence is a critical consideration
involving problems of case ascertainment and epidemiological review of
multiple data sources.19 20 21 22 23 24 This study is one of
the first major efforts to measure variation in stroke incidence and
cerebrovascular disease in a southeastern stroke belt population.
Case Ascertainment
Hospital Cases
Death Cases
Stroke Classification
A review panel of 3 neurologists independently reviewed each case.
Cases were randomly assigned by the lead neurologist (D.
Bachman). The study neurologists were not familiar with any of the
cases. Likewise, the panel was blinded to the community of origin for
the cases.
Cases were designated as probable stroke, possible stroke, and
uncertain cases (see Appendix
Statistical Analysis
Interrater agreement was assessed with a 10% random sample of the
cases rated by the lead neurologist (D. Bachman), after he was blinded
to the initial rating. The
The majority (51%) of the 674 cases were classified as uncertain
(Table 2
"First-ever" strokes represented 76% of cases
ascertained. Similar rates of first-ever strokes were detected for the
4 race-sex groups (white males, 77%; white females, 77%; black males,
74%; and black females, 76%). Nearly all of the 35- to 44-year-old
case subjects were first-ever strokes. The other age groups were
similar, with nearly three fourths of the cases identified as
first-ever strokes. First-ever strokes were similar for the 2 sites
(Anderson, 78%; Pee Dee, 74%).
All-criteria incidence rates for Anderson and Pee Dee are
presented in Table 3
Atheroembolic stroke was the most prevalent subtype, accounting for
>50% of the events in both Anderson and the Pee Dee area (see the
Figure
The fatality rates for all criteria cases were not significantly
different between the 2 sites (Anderson, 14.7%; Pee Dee area, 12.8%).
Likewise, the case-fatality rates were similar for the 4 race-sex
groups by Anderson and Pee Dee area (white males, 14.0% and 9.6%;
white females, 14.0% and 12.7%; black males, 27.3% and 19.2%; and
black females, 12.5% and 20.0%, respectively).
Classification of stroke cases, particularly from death certificates
and cases with insufficient information in the medical record, is a
critical consideration in the assessment of cerebrovascular disease
patterns. This study provides an opportunity to identify a range of
stroke incidence. Disease rates, as determined by probable/possible
criteria, reflect the minimum incidence rates for the population.
Likewise, incidence rates based on all-criteria cases would
represent the maximum rates of stroke and probably an
overestimate of the true incidence. Nonetheless, the complete range of
stroke rates determined by this study is considerably greater than the
levels reported from other incidence studies from
Framingham,38 Rochester,39
Missouri,40 Connecticut,41
and Norway.42 For example, the stroke incidence
rates for 45- to 54-year-old white males in the Pee Dee area ranged
from 2.4/1000 population per year (probable/possible) to 4.0 (all
criteria). This range is clearly higher than the rates from these other
population studies: Framingham,38 2.2;
Rochester,39 0.6;
Missouri,40 0.3;
Connecticut,41 1.2; and
Norway,42 0.6. Similar patterns for 55- to
64-year-old white males were found, with the range in the Pee Dee area
from 5.7 to 10.6 compared with Framingham38 at
4.6, Rochester39 3.6,
Missouri40 3.9,
Connecticut41 4.6, and
Norway42 2.2. For white males 65 to 74 years old,
the range in the Pee Dee area was 6.2 to 16.1/1000 population compared
with Framingham38 at 9.8,
Rochester39 8.9, Missouri40
7.4, Connecticut41 11.9, and
Norway42 7.3. Similar patterns are seen from
other population studies.43 44 45 The conservative
probable/possible category of stroke classification was higher in the
Pee Dee area than in nearly all of the comparison
populations.38 39 40 41 42 The differences are even
greater when all criteria incidence rates are compared. The higher
incidence rates in South Carolina are most dramatic when compared with
the other population studies performed in younger subjects (45 to 64
years). Comparison rates for blacks are not available, but the higher
incidence rates for blacks are much higher than the rates reported for
these other population studies.
The geographic variation in stroke and higher rates in the
Southeast46 47 have been reported. However, those
previously reported populations were separated by considerable distance
in contrast to the significantly different incidence rates of the
relatively proximal areas of Pee Dee and Anderson. Although blacks have
substantially higher rates of stroke overall, the geographic variation
between these 2 study sites is more evident for white males and
females. In fact, the incidence rates for blacks did not significantly
differ from site to site.
The reasons there are differences in stroke incidence rates between
these 2 communities are not known. Available medical care facilities
and manpower appear to be similar between the 2 areas. The profiles of
stroke subtypes, which are associated with different risk
factors,48 49 50 are similar between the areas. The
case-fatality rates were also similar for the Anderson and Pee Dee
populations. Although socioeconomic status is associated with
stroke,51 the education and economic profiles of
these 2 communities are also similar. Likewise, there is no difference
in the percentage of Medicaid admissions for stroke between the
Anderson and Pee Dee areas. Behavioral risk factors (including
hypertension, smoking, and obesity) are strongly associated with
stroke.52 53 54 55 56 However, these factors are similar
for these 2 communities.26 57 Perhaps an obscure
interaction of socioeconomic status, access to primary care, and the
prevalence of risk factors may yet explain the differences in stroke
rates. Or, some as-yet unidentified environmental risk factor, such as
magnesium in drinking water58 59 or an early life
event affecting birth weight, may play an important role in the
geographic variation of stroke incidence.60
Geographic variation in disease rates for heart disease was also seen
between these 2 areas, with the incidence of acute myocardial
infarction and congestive heart failure higher for white males in the
Pee Dee area.25 27 61 These differences in heart
disease and stroke rates suggest a variation in overall risk for
atherosclerotic disease.
The present study demonstrates that high stroke rates persist
in the southeastern United States and confirms the findings of other
mortality and morbidity reports for the
region.4 5 6 7 8 9 10 62 63 64 65 66 Although declines and changes
in stroke mortality have been reported from other
investigations,67 68 69 70 71 the high incidence rates (a
measure of disease72) reported here indicate that
the populations residing in the Southeast have an excess
cerebrovascular disease risk. The geographic variability in stroke
incidence rates suggests that the region is not so much a "belt" of
areas of contiguous increased stroke risks as a "necklace" of areas
of varying risks. Significant regional variability of stroke rates
within a defined geographical area may help to identify additional
important risk factors for stroke that need to be further
investigated.
Conclusions
Received March 26, 1998;
revision received July 2, 1998;
accepted July 2, 1998.
2.
Ostfeld SM. A review of stroke
epidemiology. Epidemiol Rev. 1980;2:136–152.
3.
Anderson RN, Kochanek KD, Murphy SL. Report of final
mortality statistics, 1995. Monthly Vital Statistical
Reports. 1997;45(suppl 2)1–80.
4.
Lanska DJ. Geographic distribution of stroke mortality
in the United States, 1936–1941 to 1979–1981. Neurology. 1993;43:1839–1851.
5.
Borhani N. Changes and distribution of mortality in
cerebrovascular disease. Am J Public Health. 1965;35:673–681.
6.
Lackland DT, Moore MA. Hypertension-related mortality
and morbidity in the Southeast. South Med J. 1997;90:191–198.[Medline]
[Order article via Infotrieve]
7.
Howard G, Evans GW, Pearce K, Howard VJ, Bell RA,
Mayer ES, Burke GL. Is the stroke belt disappearing? An
analysis of racial, temporal and age effects.
Stroke. 1995;26:1153–1158.
8.
Lanska DJ, Kuller LH. The geography of stroke
mortality in the United States and the concept of a stroke belt.
Stroke. 1995;26:1145–1149.
9.
Howard G, Howard VJ. The end of the stroke belt? It
may be too early to declare victory. Stroke. 1995;26:1150–1152.
10.
Hall WD, Farrario CM, Moore MA, Hall JE, Flack JM,
Cooper W, Simmons JD, Egan BM, Lackland DT, Perry M, Roccella E.
Hypertension-related morbidity and mortality in the Southeastern United
States. Am J Med Sci. 1997;313:1–15.[Medline]
[Order article via Infotrieve]
11.
Kuller LH, Bolker A, Saslaw MS, Paegel BL, Sisk C,
Borhani N, Wray JA, Anderson H, Peterson D, Winkelstein W, Cassel J,
Spiers D, Robinson AL, Curry A, Lilienfield AM, Seltser R. Nationwide
cerebrovascular disease mortality study, III: accuracy of the clinical
diagnosis of cerebrovascular disease. J Epidemiol. 1969;90:556–566.
12.
Lanska DJ, Peterson PM. Geographic variation on
reporting of stroke deaths to underlying or contributing causes in the
United States. Stroke. 1995;26:1999–2003.
13.
Florey CDV, Sater MG, Acheson RM. A study of the
validity of the diagnosis of stroke in mortality data. Am J
Epidemiol. 1969;89:15–24.
14.
Iso H, Jacobs DR, Goldman L. Accuracy of death
certificate diagnosis of intracranial hemorrhage and
nonhemorrhagic stroke. Am J Epidemiol. 1990;132:993–998.
15.
Howard G. Decline in stroke mortality in North
Carolina–description, predictions and a possible underlying cause.
Ann Epidemiol. 1993;3:488–492.[Medline]
[Order article via Infotrieve]
16.
Howard G, Craven TE, Sanders L, Evans GW. Relationship
of hospitalized stroke rate and in-hospital mortality to the decline in
US stroke mortality.
Neuroepidemiology. 1991;10:251–259.[Medline]
[Order article via Infotrieve]
17.
Kuller LH, Bolker A, Saslaw MS. Nationwide
cerebrovascular disease mortality study, II: comparison of clinical
records and death certificates. Am J Epidemiol. 1969;90:545–555.
18.
Gillum RF. Community surveillance for
cardiovascular disease: disease, methods, problems,
applications: a review. J Chronic Dis. 1978;31:87–94.[Medline]
[Order article via Infotrieve]
19.
Allen CMC. Differential diagnosis of acute stroke: a
review. J Roy Soc Med. 1984;77:878–881.[Medline]
[Order article via Infotrieve]
20.
Oxfordshire Community Stroke Project. Incidence of
stroke in Oxfordshire: first year's experience of a community stroke
registry. BMJ. 1983;287:713–717.
21.
Allen CMC. Clinical diagnosis of the acute stroke
syndrome. Q J Med. 1983;208:515–523.
22.
Foulkes MA, Wolf PA, Price TR, Mohr JP, Hier DM. The
stroke data bank: design, methods and baseline characteristics.
Stroke. 1988;19:547–554.
23.
Wolf PA, D'Agostino RB, Belanger AJ, Kannel WB.
Probability of stroke: A risk profile from the Framingham Study.
Stroke. 1991;22:312–318.
24.
Sacco RL. Ischemic stroke. In: Gorelick PB,
Alter M, eds. Handbook of
Neuroepidemiology. New York, NY: Marcel
Dekker; 1994:77–121.
25.
Keil JE, Gazes PE, Litaker MS, Saunders DE, Weinrich
MC, Baroody NB, Lackland DT, Hudson MB. Changing patterns of acute
myocardial infarction: decline in period prevalence and delay in onset.
Am Heart J. 1989;117:1022–1029.[Medline]
[Order article via Infotrieve]
26.
Lackland DT, Orchard TJ, Keil JE, Saunders DE, Wheeler
FC, Adams-Campbell LL, McDonald RH, Knapp RG. Are race differences in
the prevalence of hypertension explained by body mass and fat
distribution? A survey in a biracial population. Int J
Epidemiol. 1992;21:236–246.
27.
Lackland D, Berg P, Keil J, Jones E, Baroody N,
Saunders D, Gazes P. AMI incidence trends in the U.S. strokebelt.
Can J Cardiol. 1997;13(suppl B):112B. Abstract.
28.
WHO MONICA Project, Principal Investigator. The
World Health Organization MONICA (Monitoring Trends and Determining in
Cardiovascular Disease): a major international
collaboration. J Clin Epidemiol. 1988;41:105–114.[Medline]
[Order article via Infotrieve]
29.
Sandercock P, Molyneaux A, Warlow C. Value of
computerized tomography in patients with stroke: the Oxfordshire
Community Stroke Project. BMJ. 1985;290:193–197.
30.
Robins M, Baum HM. The national survey of stroke
incidence. Stroke. 1981;2(suppl):45–58.
31.
Kannel WB, Dawber TR, Cohen MS, McNamara PM. Vascular
disease of the brain—epidemiologic aspects: the Framingham Study.
Am J Public Health. 1965;55:1355–1366.
32.
Matsumoto N, Whisnant JP, Kurland LT, Okazaki H.
National history of stroke in Rochester, Minnesota, 1955 through 1969:
an extension of previous study, 1945 through 1954. Stroke. 1973;4:20–29.
33.
Rosner B. Fundamental Biostatistics. 3rd ed.
Boston, Mass: PWS–Kent Publishing Company; 1996.
34.
Bonita R. Epidemiology of
stroke. Lancet. 1992;339:342–344.[Medline]
[Order article via Infotrieve]
35.
Malmgren R, Waslow C, Bamford J, Sandercock P.
Geographical and secular trends in stroke incidence. Lancet. 1987;331:1196–1200.
36.
Hong Y, Bots ML, Pan X, Wang A, Jing H, Hofman H, Chen
H. Stroke incidence and mortality in rural and urban Shanghai from 1984
through 1991. Stroke. 1994;25:1165–1169.[Abstract]
37.
WHO MONICA Project. Stroke incidence and mortality
unrelated to stroke risk factors in the WHO MONICA Project: an
ecological study of 18 populations. Stroke. 1997;28:1367–1379.
38.
Wolf PA, Cobb JL, D'Agostino.
Epidemiology of stroke. In: Barnett HJM, Stein
BM, Mohr JP, Yatsu FM, eds. Stroke: Pathology, Diagnosis, and
Management. New York, NY: Churchill-Livingstone; 1992.
39.
Brown RD, Whisnant SJ, Sicks JD, O'Fallon WM, Wiebers
DO. Stroke incidence, prevalence and survival secular trends in
Rochester, Minnesota, 1986 through 1989. Stroke. 1996;27:373–380.
40.
Eckstrom PT, Brand FR, Edlavitch SA, Parrish HM.
Epidemiology of stroke in a rural area.
Public Health Rep. 1969;84:878–882.[Medline]
[Order article via Infotrieve]
41.
Eisenberg H, Morrison JT, Sullivan P, Foote FM.
Cerebrovascular accidents, influence and survival rates in a defined
population, Middlesex County, Connecticut. JAMA. 1964;198:883–888.
42.
Ellekjaer H, Holmen J, Indredovik B, Lerent A.
Epidemiology of stroke in Innherred, Norway,
1994 to 1996: incidence and 30-day care fatality rate.
Stroke. 1997;28:2180–2184.
43.
Herman B, Leyten ACM, Van Luiyk JH, Frenken LW,
Opdecoul AA, Schulte BP. Epidemiology of stroke
in Tolberg, the Netherlands. Stroke. 1982;13:829–835.
44.
Truelson T, Prescott E, Gronbuek M, Schnohr P, Boysen
G. Trends in stroke incidence: the Copenhagen City Heart Study.
Stroke. 1997;28:1903–1907.
45.
McGovern PG, Burke GL, Sprafka JM, Xue S, Folsom AR,
Blackburn H. Trends in mortality, morbidity, and risk factor levels for
stroke from 1966 through 1990: the Minnesota Heart Survey.
JAMA. 1992;268:753–759.
46.
Gillum RF, Ingram DD. Relation between residence in the
southeast region of the United States and stroke. Am J
Epidemiol. 1996;144:665–673.
47.
Stolley PD, Kuller LH, Nefzger MD, Fonasia S,
Lillienfield AM, Miller GD, Diamond EL. Three-area epidemiological
study of geographic differences in stroke mortality, II: results.
Stroke. 1977;8:551–557.
48.
Caplan LR. Intracerebral
haemorrhage. Lancet. 1992;339:656–658.[Medline]
[Order article via Infotrieve]
49.
Sacco RL, Kargman DE, Gu Q, Zamanilla MC.
Race-ethnicity and determinants of intracranial atherosclerotic
cerebral infarction: the northern Manhattan stroke study.
Stroke. 1995;26:14–28.
50.
Wolf PA. An overview of the
epidemiology of stroke. Stroke.
1990;21(suppl II):II-4–II-6.
51.
Modan B, Wagener DK. Some epidemiological aspects of
stroke: mortality/morbidity trends, age, sex, race, socioeconomic
status. Stroke. 1992;23:1230–1236.
52.
Kannel WB, Wolf PA. Inference from secular trend
analysis of hypertension control. Am J Public
Health. 1992;82:1593–1595. Editorial comment.
53.
Wolf PA, Kannel WB, McNamara PM, Gordan T. The role of
impaired cardiac function in atheroembolic brain infarctions: the
Framingham Study. Am J Public Health. 1973;63:52–58.
54.
Howard G, Manolio TA, Burke GL, Wolfson SK, O'Leary
DH. Does the association of risk factors and
atherosclerosis change with age? An analysis of
the combined ARIC and CHS cohorts. Stroke. 1997;28:1693–1701.
55.
Bonita R, Beaglehole R. The enigma of the decline in
stroke deaths in the United States: the search for an explanation.
Stroke. 1996;27:370–372.
56.
Kannel WB, Wolf PA, Venter J, McNamara PM.
Epidemiologic assessment of the role of blood pressure in stroke: the
Framingham Study. JAMA. 1996;276:1269–1278.
57.
Wheeler FC, Lackland DT, Mace ML, Reddick A, Hogelin G,
Remington PL. Evaluating South Carolina's community
cardiovascular disease prevention project.
Public Health Rep. 1991;106:536–544.[Medline]
[Order article via Infotrieve]
58.
Yamori Y, Nara Y, Mizushima S, Sawamura M, Horie
R. Nutritional factors for stroke and major
cardiovascular diseases: international epidemiological
comparison of dietary prevention. Health Rep. 1994;6:22–27.[Medline]
[Order article via Infotrieve]
59.
Alturu BM, Altura BT. Magnesium and
cardiovascular biology: an important link between
cardiovascular risk factors and atherogenesis.
Cell Mol Biol Res. 1995;41:347–359.[Medline]
[Order article via Infotrieve]
60.
Martyn CN, Barker DJ, Osmond C. Mothers' pelvic size,
fetal growth, and death from stroke and coronary heart disease
in men in the U.K. Lancet. 1996;348:1464–1468.[Medline]
[Order article via Infotrieve]
61.
Lackland DT, Berg PH, Keil JE, Baroody NB, Gazes PC,
Jones FG, Saunders DE Jr. Racial differences in congestive heart
failure (CHF) associated with acute myocardial infarction (AMI).
J Heart Failure. 1995;2:249. Abstract.
62.
Pickle LW, Mungiole M, Gillum RF. Geographic variation
in stroke mortality in blacks and whites in the United States.
Stroke. 1997;28:1639–1647.
63.
Wing S, Casper M, Davis WB, Pellon A, Riggan W, Tyroler
HA. Stroke mortality maps: United States whites aged 35–74 years,
1962–1982. Stroke. 1988;19:1507–1513.
64.
Mason TJ, Fruumeni JF Jr, Hoover R, Blot WJ. An atlas
of mortality from selected diseases. NIH Publication No. 81–2397.
Washington, D. C. U. S. Government Printing Office;
1981.
65.
Heyman A, Tyroler HA, Caussel JC, O'Fallon WM, Davis
L, Muhlbaier L. Geographic differences in mortality from stroke in
North Carolina, I: analysis of death certificates.
Stroke. 1976;7:41–45.
66.
Sauer HI. The enigma of the Southeast: trace elements
in environmental health. In: Hemphill DD, ed. Proceedings of
University of Missouri 9th Annual Conference on Trace Substances in
Environmental Health. Columbia, Missouri, 1975.
67.
Wolf PA, D'Agostino RB, O'Neal MA, Sytkowski P, Kare
CS, Belanger AJ, Kannel WB. Secular trends in stroke incidence and
mortality: the Framingham Study. Stroke. 1993;23:1551–1555.
68.
Wolf PA, D'Agostino RB. Secular trends in stroke in
the Framingham Study. Ann Epidemiol. 1993;3:471–475.[Medline]
[Order article via Infotrieve]
69.
Casper ML, Wing S, Anda RF, Escobedo LG, Taylor HA. The
shifting stroke belt: changes in the geographic pattern of stroke
mortality in the United States, 1962 to 1988. Stroke. 1995;26:755–760.
70.
Lanska DJ, Peterson PM. Effects of interstate migration
on the geographic distribution of stroke mortality in the United
States. Stroke. 1995;26:554–561.
71.
Lanska DJ, Peterson PM. Geographic variation in the
decline of stroke mortality in the United States. Stroke. 1995;26:1159–1165.
72.
Shahar E, McGovern PG, Pankow JS, Doliszny KM, Smith
MA, Blackburn H, Luepken RV. Stroke rates during the 1980s: the
Minnesota Stroke Survey. Stroke. 1997;28:275–279.
© 1998 American Heart Association, Inc.
Original Contributions
The Geographic Variation in Stroke Incidence in Two Areas of the Southeastern Stroke Belt
The Anderson and Pee Dee Stroke Study
Abstract
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
Appendix 1
References
Background and Purpose—South
Carolina and the southeastern United States have maintained the highest
stroke mortality in the country. The Anderson and Pee Dee Stroke Study
is an assessment of cerebrovascular disease incidence in 2
geographically defined communities in the stroke belt.
Key Words: blacks • cerebrovascular disorders • epidemiology • geography • incidence
Introduction
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
Appendix 1
References
Although cerebrovascular disease mortality has been
declining for the past 2 decades, stroke remains the third leading
cause of death in the United States.1 2 3 However,
significant geographic variation in stroke mortality within the United
States has been detected.4 The southeastern
region of the United States, dubbed the "stroke belt," has been
recognized as an area of excess cerebrovascular disease mortality for 4
decades.5 Although some recent investigations
have suggested changes in the contiguous geographic pattern of
mortality, areas of excessive stroke death rates continue to be
reported in the Southeast.6 7 8 9 The reasons for
these areas of high mortality risk remain an
enigma10,11; however, death certificate reporting
practices do not explain the geographic variation in stroke
mortality.12
Subjects and Methods
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
Appendix 1
References
Study Population Areas
The 2 South Carolina population areas used in this study were
Anderson county and the Pee Dee area, which includes Darlington and
Florence counties. The 2 areas are separated by approximately 200
miles. In 1990, both areas had total populations of slightly less than
100 000. The median education level of both areas was approximately 12
years, with 16.1% of the Anderson adults having <9 years of education
compared with 16.3% from the Pee Dee area. Unemployment levels were
similar at 7.4%. Also, 9.7% of the Anderson population in 1990 was
below the poverty level compared with 11.3% of the Pee Dee residents.
Although socioeconomic status indicators are similar for the 2 areas,
some racial variation is seen, with a higher black proportion in the
Pee Dee area (33%) than in Anderson (18%). The 2 areas are considered
predominantly rural, with the city of Anderson, in Anderson county, and
Florence, in the Pee Dee area, as urban centers. The Pee Dee area has 7
hospitals, including 1 regional medical center, whereas Anderson is
served by 1 hospital, a large regional medical center. The 2
populations have participated in cardiovascular disease
studies in the past, including a surveillance study of acute myocardial
infarction and population risk factors.25 26 27
This prior experience in assessing the incidence of acute myocardial
infarction in these communities identified very little out-migration
for acute health care. Therefore, almost all cases requiring
hospitalization can be captured by reviewing the discharge records
from the index hospitals. Hospital discharge records were reviewed
from the entire state system to determine that all cases for residents
had been included from out-of-area hospitals.
Stroke cases were identified from hospital records,
emergency room records, death certificates, coroners' records,
physicians' reports, and informant interviews. All case subjects 35 to
74 years of age were included in the study. Case subjects 75 years of
age and older were excluded because of the difficulty in complete
accounting as a result of different treatment and diagnosis for elderly
patients, a higher rate of old events, and higher rates of
out-of-hospital case management such as nursing homes. Multiple events
during the year were considered as 1 case.
Each hospital identified all hospitalizations and emergency room
presentations with discharges of cerebrovascular disease
(International Classification of Diseases, Revision 9, codes
430 to 438) occurring during the calendar year 1990. These selection
criteria are broad to maximize the selection of all acute stroke cases.
Hospitalizations for residents of Anderson, Darlington, and Florence
counties occurring outside of the 3-county area were also identified
through the South Carolina hospital discharge system. Records were
manually abstracted by trained abstractors using standard methodology.
The abstract forms and protocol were adapted from similar population
studies of stroke incidence, and included modules of initial
symptomatology, physical examination reports, comorbid conditions,
associated discharges, diagnostic procedures including
CT/MRI scan reports, hospital treatments, and supporting information
from the medical records.28 29 30 31 32
Death certificates were obtained for all mortality in stroke
cases occurring in residents of the 3-county area. Certificate
information was merged with abstract information from hospitalized
events. Out-of-hospital events with no prior hospital discharges in
1990 were considered new cases. In addition to death certificate
information, coroners' reports, autopsy reports, physicians' reports,
and informant interviews were included in the assessment. Case fatality
rates were determined for deaths occurring for hospitalized cases of
stroke.
A stroke was defined as a neurological deficit with an onset of
72 hours or less, and a duration of 24 hours or greater due to
ischemic infarction or intraparenchymal hemorrhage.
Subarachnoid hemorrhage and intraparenchymal
hemorrhage due to berry aneurysm, trauma, neoplasm, or
infection were not considered stroke for this study. 
). Cases that met the criteria for 1 of
the 5 stroke subtypes (atheroembolic stroke, cardioembolic stroke,
lacunar stroke, intraparenchymal hemorrhage, and indeterminate)
with a CT scan and/or MRI scan that demonstrated an acute lesion
corresponding to the clinical findings were designated as "probable
stroke." Cases meeting the clinical criteria for stroke in which a CT
or MRI scan did not demonstrate an acute lesion, or no scan was
performed, were designated as "possible strokes." The remaining
cases were designated "uncertain strokes." "All-criteria
strokes" include those classified as probable, possible, and
uncertain. Case subjects designated as having atheroembolic stroke
exhibited typical clinical features of stroke; cardioembolic case
subjects exhibited typical clinical features of stroke and a likely
cardiogenic source of embolus; lacunar cases fit 1 of the classic
clinical profiles of lacunar stroke; and intraparenchymal
hemorrhage cases were diagnosed by CT scan or MRI scan. Details
of syndrome criteria are presented in the Appendix.
Incidence rates of stroke were calculated as cases per 100 000
population. Rates were age-adjusted using the direct method and the
1970 US population as the standard. Rates were compared for statistical
significance. When all cell sizes used in the comparisons were greater
than 5, the normal theory test (2-sample test for binomial proportions)
was used, and for obtaining a confidence interval for the binomial
function. When any cells in the comparisons were <5, the exact method
was used.33 
statistic was used to measure the level
of agreement. Incidence rates were compared with other population-based
stroke studies.
Results
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
Appendix 1
References
Abstractors reviewed 1942 records as part of the case
ascertainment process. More than one fourth of the records (27.0%)
were determined to be ineligible because the event was a transient
ischemia attack only or other nonstroke cerebrovascular
disorder and/or the event was considered to be "old," ie, occurring
before 1990. The proportions of these ineligible cases were similar for
the 2 areas (Anderson, 24.7%; Pee Dee, 28.4%). The case ascertainment
process identified 1418 case subjects from the 2 areas (Table 1). The majority (60.6%) of the total
case subjects were ascertained from the Pee Dee area. Nearly all
(95.0%) of the case were identified from hospital records, with
the remainder as out-of-hospital deaths determined from death
certificate review. The proportion of hospitalized cases was similar
for both areas. The second component of the triage process consolidated
multiple admissions to 1 case subject and removed case subjects who
were not between the ages of 35 and 74 years, resulting in 674 (274
from Anderson, 400 from Pee Dee) cases eligible for review by the
neurologists.
View this table:
[in a new window]
Table 1. Chart and Record Review by Record Source and
Site ). Thirty-six percent were
determined to be probable, and 13% were possible. The proportions of
cases in the 3 classifications were similar for the 2 study areas.
Likewise, similar patterns were seen for the 4 race-sex groups. The
10% quality control re-review determined excellent interrater
agreement for stroke classification and subtype as identified by
>0.75.
View this table:
[in a new window]
Table 2. Final Classification Disposition for Eligible Cases
Subjects Aged 35–74
Years . Although the Pee Dee
area had significantly higher overall incidence rates, Pee Dee white
males were the only race-sex group to have had significantly greater
rates than for their Anderson counterparts. Similar patterns were
detected when probable and possible cases were considered (Table 4). The stroke rates for
probable and possible cases increased with age for all groups in both
areas. Although whites had higher numbers of strokes, black males and
females had significantly higher incidence rates than their white
counterparts. These higher rates were observed for each of the age
groups as well. Overall, the rates for the Pee Dee area were
statistically higher than for Anderson.
View this table:
[in a new window]
Table 3. Number, Rate (per 100 000 Population), and 95%
Confidence Interval for All-Criteria Stroke Case Subjects by Age, Race,
Sex, and Area
View this table:
[in a new window]
Table 4. Number, Rate (per 100 000 Population), and 95%
Confidence Interval for Probable and Possible Stroke Case Subjects by
Age, Race, Sex, and Area ). Although some variation was determined for subtypes
(atheroembolic stroke, cardioembolic stroke, lacunar stroke,
intraparenchymal hemorrhage, and indeterminate) the
distributions between the 2 areas were not statistically different.
However, statistically significant (P=0.021) variation in
subtypes was detected for the race-sex groups, with blacks having the
higher rates of intraparenchymal hemorrhage (Figure). In
particular, one fourth of the strokes in black males were classified as
intraparenchymal hemorrhage. Some variation of stroke type was
detected by the age groups <65 years and 
65 years (atheroembolic
stroke, 62% versus 55%; cardioembolic stroke, 7% versus 12%;
lacunar stroke, 11% versus 21%; and intraparenchymal
hemorrhage, 17% versus 10%, respectively).
View larger version (34K):
[in a new window]
Figure 1. Percent of probable/possible stroke by subtype categories
(atheroembolic, cardioembolic, lacunar, intraparenchymal
hemorrhage, and indeterminant) and race-sex categories. The
variation in subtypes was significant among the 4 race-sex groups
(P=0.021).
Discussion
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
Appendix 1
References
The variation in cerebrovascular disease incidence has been
recognized and reported from different areas of the
world.34 35 36 37 Nonetheless, estimates of
cerebrovascular disease have been limited in the Southeast United
States, and the purpose of this paper was to determine incidence rates
for stroke in specific populations of this region. Although this study
has the obvious limitations of a retrospective chart review, very
conservative criteria for stroke diagnosis were used to identify cases.
Likewise, interrater reliability for the classification of categories
was high, as indicated by >0.75. The protocols were adapted from
other population studies.28 29 30 31 32 Although case
management and diagnostic approaches are important
considerations in such a study, the similar proportions of cases
classified as strokes would suggest similar situations for the 2 South
Carolina areas included in this study.
The Pee Dee area had an overall 40% greater stroke incidence rate
that was statistically significant compared with the Anderson area. The
geographic difference was more dramatic in younger subjects: the Pee
Dee rates were 58% greater for subjects aged 55 to 64 years and 100%
greater for those aged 45 to 54 years, but only 13% greater for the
65- to 74-year age group. This contrast was particularly evident for
white males: the rates for the 2 areas were nearly identical for age
group 65 to 74 years, but rates were twice as high for Pee Dee white
males 55 to 64 years old and 3.6 times greater in the 45- to
54-year-old age group. The small cell size and wide confidence interval
are limitations of our study comparing the 2 areas by age-race-sex
groups. Nonetheless, a consistent pattern for the trends
appears to be present.
Appendix 1
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
Appendix 1
References
Criteria for Anderson and Pee Dee Stroke Study
I. Definition: A stroke is a neurological deficit with an onset of
72
hours and a duration of 24 hours due to ischemic infarction
or intraparenchymal hemorrhage. Subarachnoid
hemorrhage or intraparenchymal hemorrhage due to berry
aneurysm, trauma, neoplasm, or infection is not considered
stroke for purposes of this study.
II. Syndrome definitions
A. Atheroembolic stroke
III. Stroke diagnosis
1. Clinical signs—hemiplegia, hemiparesis, or hemianopsia or at least 2
or more of the following:
B. Cardioembolic stroke
a. Depression or level of consciousness
b. Disturbance of vision
c. Hemihypesthesias or hemiparesthesias
d. Speech or language impairment
e. Ataxia
f. Cranial nerve abnormalities
g. Neuropsychological dysfunction associated with a focal lesion
1. A (1) criterion
C. Lacunar stroke—focal syndrome with typical features for 1 of the
following:
2. Establishment of a likely cardiogenic source by history or examination
(must have 1 of the following):
a. Significant valvular disease
b. Atrial fibrillation
c. Myocardial infarction (within 6 weeks)
d. Recent cardiac surgery
e. Bacteria endocarditis
f. Atrial myxoma
1. Pure motor stroke
D. Intraparenchymal hemorrhage
2. Pure sensory stroke
3. Pure hemiballismus
4. Pure hemichorea
5. Ataxic hemiparesis
6. Dysasrthria/clumsy hand syndrome
1. A (1) criteria—clinical signs also may include sudden, severe headache
E. Stroke of indeterminate type
2. CT scan or MRI scan confirms hemorrhage
1. A (1) criteria
2. And 1 of the following:
a. Inadequate or conflicting information to make a determination
b. Presence of a significant comorbid disease or condition (eg, cancer or
lupus) makes etiology uncertain
A. Probable stroke
1. Meets clinical criteria for 1 of stroke syndromes
B. Possible stroke
2. CT scan or MRI scan consistent with clinical findings
1. Meets clinical criteria for 1 of stroke syndromes
C. Stroke uncertain—A (1) clinical criteria of uncertain validity
2. No CT scan or MRI scan available, or if available, no lesion that
corresponds to clinical findings
Acknowledgments
This work was supported in part by a grant from the American
Heart Association, South Carolina Affiliate, and the Young Scholar's
Award from the American Society of Hypertension.
References
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
Appendix 1
References
1.
Wolf PA, Dawber TR, Thoma HE, Colton T, Kannel WP.
Epidemiology of stroke. Adv
Neurol. 1977;16:5–19.[Medline]
[Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  M. M. Glymour, M. Avendano, and L. F. Berkman Is the 'Stroke Belt' Worn From Childhood?: Risk of First Stroke and State of Residence in Childhood and Adulthood Stroke, September 1, 2007; 38(9): 2415 - 2421. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. D. Bushnell, P. Hurn, C. Colton, V. M. Miller, G. del Zoppo, M. S.V. Elkind, B. Stern, D. Herrington, G. Ford-Lynch, P. Gorelick, et al. Advancing the Study of Stroke in Women: Summary and Recommendations for Future Research From an NINDS-Sponsored Multidisciplinary Working Group Stroke, September 1, 2006; 37(9): 2387 - 2399. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. El-Saed, L. H. Kuller, A. B. Newman, O. Lopez, J. Costantino, K. McTigue, M. Cushman, and R. Kronmal Geographic Variations in Stroke Incidence and Mortality Among Older Populations in Four US Communities Stroke, August 1, 2006; 37(8): 1975 - 1979. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J.P. Barker and D. T. Lackland Prenatal Influences on Stroke Mortality in England and Wales Stroke, July 1, 2003; 34(7): 1598 - 1602. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C D A Wolfe, A G Rudd, R Howard, C Coshall, J Stewart, E Lawrence, C Hajat, and T Hillen Incidence and case fatality rates of stroke subtypes in a multiethnic population: the South London Stroke Register J. Neurol. Neurosurg. Psychiatry, February 1, 2002; 72(2): 211 - 216. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. T. Lackland, B. M. Egan, H. E. Syddall, and D. J.P. Barker Associations Between Birth Weight and Antihypertensive Medication in Black and White Medicaid Recipients Hypertension, January 1, 2002; 39(1): 179 - 183. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Fang, M. H. Alderman, and J. V. Tu Trend of Stroke Hospitalization, United States, 1988-1997 Editorial Comment Stroke, October 1, 2001; 32(10): 2221 - 2226. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Engstrom, I. Jerntorp, H. Pessah-Rasmussen, B. Hedblad, G. Berglund, and L. Janzon Geographic Distribution of Stroke Incidence Within an Urban Population : Relations to Socioeconomic Circumstances and Prevalence of Cardiovascular Risk Factors Stroke, May 1, 2001; 32(5): 1098 - 1103. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. T. Lackland, H. E. Bendall, C. Osmond, B. M. Egan, and D. J. P. Barker Low Birth Weights Contribute to the High Rates of Early-Onset Chronic Renal Failure in the Southeastern United States Arch Intern Med, May 22, 2000; 160(10): 1472 - 1476. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. T. Lackland, B. M. Egan, and P. J. Jones Impact of Nativity and Race on "Stroke Belt" Mortality Hypertension, July 1, 1999; 34(1): 57 - 62. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. D. Rosamond, A. R. Folsom, L. E. Chambless, C.-H. Wang, P. G. McGovern, G. Howard, L. S. Copper, and E. Shahar Stroke Incidence and Survival Among Middle-Aged Adults : 9-Year Follow-Up of the Atherosclerosis Risk in Communities (ARIC) Cohort Stroke, April 1, 1999; 30(4): 736 - 743. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||