Characterization of Incident Stroke Signs and Symptoms
Findings From the Atherosclerosis Risk in Communities Study
Background and Purpose— Although patterns of stroke occurrence and mortality have been well studied, few epidemiological data are available regarding the clinical characteristics of stroke events.
Methods— We evaluated hospitalized stroke events reported in the Atherosclerosis Risk in Communities (ARIC) Study to describe the clinical characteristics of incident stroke. Confirmed stroke cases (n=474) were evaluated for stroke symptoms (headache, vertigo, gait disturbance, convulsions) and stroke signs (hemianopia, diplopia, speech deficits, paresis, paresthesia/sensory deficits) and their univariate associations with race, sex, and stroke subtype.
Results— Over 9.2 years of follow-up, 402 (85%) ischemic and 72 (15%) hemorrhagic strokes occurred. Frequency of stroke symptoms (95% CIs) were as follows: headache (27.4%; 23.4% to 31.4%), gait disturbance (10.8%; 7.9% to 13.6%), convulsions (4.4%; 2.6% to 6.3%), and vertigo (2.1%; 0.8% to 3.4%). Speech deficits occurred in 24.0% (20.2% to 27.9%), hemianopia in 14.6% (11.4% to 17.7%), and diplopia in 5.5% (3.4% to 7.5%) of cases. Most cases involved paresis (81.6%; 78.1% to 85.1%), while fewer cases experienced sensory deficits (44.5%; 40.0% to 49.0%). Blacks were more likely than whites to experience paresis (85.4% versus 78.2%; P=0.044). Men were more likely than women to experience a gait disturbance (14.4% versus 6.7%; P=0.007). Persons with hemorrhagic strokes had a higher proportion of headaches (55.6% versus 22.4%; P=0.001) and convulsions (11.1% versus 3.2%; P=0.003) than those with ischemic events, while speech and sensory deficits were more common in ischemic strokes (26.1% versus 12.5%, P=0.013, and 49.0% versus 19.4%, P=0.001, respectively).
Conclusions— We present epidemiological data concerning the clinical characteristics of incident stroke in a population-based cohort. Although minor differences by race, sex, and stroke subtype were observed, data from additional follow-up are required to confirm observed variations.
Stroke remains a major public health concern in the United States, with more than 600 000 cases diagnosed annually.1 Most epidemiological data have focused more on mortality rates and associated risk factors and less on characterizing incident events.2–5⇓⇓⇓ Although mortality data are informative, they do not fully assess the qualitative nature of incident stroke: the signs, symptoms, sequelae, and associated deficits that represent the full clinical scope of a stroke event. These data are necessary to obtain a more complete understanding of the burden of stroke in a population and to better inform public health strategies, including public education and stroke prevalence assessments. We undertook an analysis of incident strokes reported in the Atherosclerosis Risk in Communities (ARIC) Study to better characterize the clinical presentation of incident stroke in a community-based setting.
Subjects and Methods
The ARIC Study has been described in greater detail elsewhere.6 Potential stroke events in ARIC were identified from a cohort of 15 792 individuals age 45 to 64 years were recruited during 1986–1989 from 4 communities in the United States: Washington County, Maryland; northwest suburbs of Minneapolis, Minnesota; Forsyth County, North Carolina; and Jackson, Mississippi. Only blacks were enrolled at the Jackson site, while all other sites reflected the demographic composition of the communities from which they were drawn. ARIC participants attended a clinic visit triennially and received a follow-up telephone call annually.
Details of the identification and classification of stroke events are described in greater detail elsewhere.7 Potential stroke events were identified from self-reported hospitalizations obtained during the annual follow-up and from ongoing community-wide hospital surveillance. Medical records for potential stroke events were forwarded to a single nurse abstractor at a central ARIC office who abstracted each record for number, type, and severity of neurological deficits and supporting angiographic, CT, MRI, spinal tap, or autopsy evidence. A computerized algorithm and physician reviewer were used to independently confirm the diagnosis of stroke, with disagreements adjudicated by a second physician reviewer. Strokes were further classified as ischemic (thrombotic or embolic), hemorrhagic, undetermined type, or fatal out-of-hospital stroke.
Of the 1719 eligible potential stroke events identified among ARIC participants during 1987–1997, we limited our consideration to the 873 first events with evidence in the medical record of any neurological signs lasting >24 hours or development of new signs or symptoms leading to or during hospitalization. Of these 873 events, 498 were validated as either definite or probable strokes. From these 498 we excluded fatal out-of-hospital strokes (n=10), strokes of undetermined type (n=7), patients who recently experienced head trauma (n=2), and those with other neurological conditions that may be associated with stroke signs and symptoms (n=5). The remaining 474 confirmed incident stroke events were available for these analyses.
The principal objective of our study was to characterize the signs and symptoms of stroke found on examination. Specific symptoms evaluated included headache at admission (and severity), vertigo, gait disturbance, and convulsions. Stroke signs evaluated included hemianopia, diplopia, and speech deficits (aphasia and/or dysphasia). Paresis or paresthesia/sensory deficits of the face, arms, or legs were also noted.
The frequency of stroke signs and symptoms among incident stroke events was determined and evaluated for univariate associations with race, sex, and stroke subtype with the use of χ2 analysis. All statistical calculations were performed with the use of SAS 7.0 software (SAS Institute Incorporated).
Of the 474 confirmed incident stroke events in the study cohort, 402 (84.8%) were ischemic and 72 (15.2%) were hemorrhagic. Subjects were predominantly male, white, and aged 62 years on average at the time of their stroke, and most were current or former users of tobacco and alcohol. Few subjects experienced acute clinical risk factors for stroke in the month before their stroke (Table 1).
More than 25% of subjects hospitalized with a confirmed stroke presented with a headache at admission, while vertigo, convulsions, and gait disturbance were less frequent. The most common stroke sign on presentation was a speech deficit (24.0%), followed by hemianopia (14.6%) and diplopia (5.5%). More than 80% of subjects presented with some paresis, most often of the arms (75.5%), although a majority reported paresis of the face (54.6%) and legs (68.6%); location of paresis was equally distributed between right and left sides. A near majority of subjects (44.5%) presented with some sensory deficit, most often of the arms (38.6%) and also, in decreasing frequency, of the legs (34.4%) and face (20.7%). A majority of sensory deficits involved the left side of the affected location (Table 2).
Evaluation of stroke signs and symptoms by sex indicated little variation. Although men had more frequent occurrence of gait disturbance (14.4% versus 6.7%; P=0.007) and women tended to have a higher frequency of headache and facial sensory deficits, sex was not associated with the occurrence of other stroke signs and symptoms (Table 2). In contrast, blacks were more likely to experience paresis of the face, arm, and leg, while whites tended to have more frequent occurrence of vertigo and a gait disturbance. Race was not associated with the occurrence of other stroke signs and symptoms (Table 2).
As expected, stroke signs and symptoms varied considerably by stroke subtype. Subjects with hemorrhagic strokes had a higher frequency of headaches (mostly severe) and convulsions at admission. In contrast, speech deficits were most common among ischemic strokes, while hemianopia and diplopia were comparable in occurrence. Hemorrhagic stroke events had the lowest frequency of paresis and sensory deficits in all sites (Table 2).
Our evaluation of incident strokes in the ARIC cohort found that paresis, speech, and sensory deficits were common among subjects hospitalized with confirmed stroke events. Male subjects were more likely to experience gait disturbances, and blacks were more likely to experience paresis in any location. Subjects with ischemic strokes were more likely to experience paresis, speech, and sensory deficits, while headaches were most common with hemorrhagic stroke events. Our data suggest that the clinical characteristics of stroke may vary for different population groups. These data provide a preliminary, detailed understanding of the specific functional burden of stroke in a population and can be used to better inform public health strategies, including public education and stroke prevalence assessments.
Use of the ARIC cohort provides several methodological advancements in the assessment of the clinical characteristics of incident stroke. The ARIC cohort was drawn from 4 US communities, including a substantial number of black participants, and thus represents a more population-based assessment than selected center data. ARIC used a clinically defined, validated diagnosis of stroke, which is an improvement from prior studies that relied solely on patient self-report, discharge codes from medical records, or administrative data.8 Finally, ARIC represents a recent evaluation of stroke events and associated clinical deficits, thus allowing for recent advances in stroke diagnosis, treatment, and management that may have reduced stroke-associated morbidity.2
Comparison of our findings with prior data is limited because few studies evaluate stroke signs in an epidemiological setting or report specific clinical deficits rather than composite severity scores such as the Barthel Index or Rankin Scale scores.9 Two studies, however, have reported stroke signs and symptoms with sufficient detail for comparative purposes. An analysis of stroke events in the northern Sweden World Health Organization Monitoring Trends and Determinants in Cardiovascular Disease (MONICA) cohort reported paresis in 48.9% of events,10 which is markedly lower than the 82% paresis frequency observed in ARIC. In contrast, stroke sign and symptom frequencies reported in an evaluation of health maintenance organization enrollees in Portland, Oregon, found similar rates for visual deficits (20%) and paresis (78%) but lower estimates of the prevalence of dysphasia (11%).11 Variations in stroke sign and symptom rates between studies may be expected because of their use of different populations: MONICA enrolled subjects as young as 35 years, while the Portland health maintenance organization study focused on elderly (aged ≥65 years) subjects. Differences in symptom ascertainment, evaluation, and documentation may have consequences for valid estimations of the frequency of stroke signs and symptoms given the 4.4 million Americans currently experiencing poststroke deficits.1 Future epidemiological studies should consider evaluation of stroke signs and symptoms with use of standardized measures to better evaluate the full clinical scope of stroke events.
Our study of stroke signs and symptoms was limited by the fact that we evaluated signs and symptoms reported on admission or during hospitalization and thus are unable to make conclusions regarding stroke deficits present at discharge. Our analysis was limited to hospitalized subjects. Subjects experiencing strokes resulting in out-of-hospital death are unavailable; we may therefore potentially underestimate the true incidence of stroke signs and symptoms, although the estimated size of this group (n=10 in the ARIC cohort) suggests that this bias is likely small. Similarly, subjects who experienced minor strokes that did not result in hospitalization would also be omitted, possibly biasing our results toward more severe stroke signs and symptoms. We consider it unlikely that a persistent (>24 hours) neurological deficit would not prompt a patient to seek medical attention, and we therefore estimate any such bias to be small. Finally, we did not account for differences in risk factors associated with race, sex, or stroke subtype that might confound differences in stroke signs and symptoms. However, we did not seek to determine whether race, sex, and stroke subtype were independent risk factors for specific stroke signs or symptoms, but rather we sought to characterize subpopulation differences in the clinical manifestations of a stroke event. Finally, data were obtained from a retrospective medical record review. Although we consider deficiencies in documentation of stroke-related signs and symptoms to be unlikely because of the clinical severity of the event, concerns of data reliability and validity cannot be precluded, and a prospective epidemiological study of stroke severity would be of value. In the interim, our study provides data that can inform public health strategies directed at assessing and reducing stroke severity and stroke events.
In conclusion, we present recent data on the frequency of signs and symptoms of incident stroke as reported in a population-based assessment. While most clinical characteristics of incident stroke did not appear to vary by sex, we found an increased frequency of stroke deficits, specifically paresis, among blacks. As expected, stroke signs and symptoms varied by stroke subtype. Data such as these may be used to more fully inform current epidemiological assessments of incident stroke events.
The authors thank the staff and participants in the ARIC Study for their important contributions. The ARIC Study was supported by contracts N01-HC-55015, N01-HC-55016, N01-HC-55018, N01-HC-55019, N01-HC-55020, N01-HC-55021, and N01-HC-55022 from the US National Heart, Lung, and Blood Institute. The authors thank Lisa Reeves-Brown for her assistance with computer programming and statistical analyses and the staff and participants of the ARIC Study for their important contributions.
- Received March 5, 2002.
- Revision received June 12, 2002.
- Accepted June 25, 2002.
- ↵2002 Heart and Stroke Statistical Update. Dallas, Tex: American Heart Association; 2002.
- ↵Wolf PA, D’Agostino RB, O’Neal MA, Sytkowsko P, Kase CS, Belanger AJ, Kannel WB. Secular trends in stroke incidence and mortality: the Framingham Study. Stroke. 1992; 23: 1551–1555.
- ↵Stegmayr B, Asplund K, Kuulasmaa K, Rajakangas A, Thorvaldsen P, Tuomilehto J. Stroke incidence and mortality correlated to stroke risk factors in the WHO MONICA Project: an ecological study of 18 populations. Stroke. 1997; 28: 1367–1374.
- ↵Sacco RL, Boden-Albala B, Gan R, Chen X, Kargman DE, Shea S, Paik MC, Hauser WA. Stroke incidence among white, black, and Hispanic residents of an urban community: the Northern Manhattan Stroke Study. Am J Epidemiol. 1998; 147: 259–268.
- ↵Broderick J, Brott T, Kothari R, Miller R, Khoury J, Pancioli A, Gebel J, Mills D, Minneci L, Shukla R. The Greater Cincinnati/Northern Kentucky Stroke Study: preliminary first-ever and total incidence rates of stroke among blacks. Stroke. 1998; 29: 415–421.
- ↵ARIC Investigators. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. Am J Epidemiol. 1989; 129: 687–702.
- ↵Rosamond W, Folsom A, Chambless L, Wang C, McGovern P, Howard G, Cooper L, Shahar E. Stroke incidence and survival among middle-aged adults: 9-year follow-up of the Atherosclerosis Risk in Communities (ARIC) cohort. Stroke. 1999; 30: 736–743.
- ↵Benesch C, Witter D Jr, Wilder A, Duncan P, Samsa G, Matchar D. Inaccuracy of the International Classification of Disease (IDC-9-CM) in identifying the diagnosis of ischemic cerebrovascular disease. Neurology. 1997; 49: 660–664.
- ↵Jamrozik K, Broadhurst RJ, Lai N, Hankey GJ, Burvill PW, Anderson CS. Trends in the incidence, severity, and short-term outcome of stroke in Perth, Western Australia. Stroke. 1999; 30: 2105–2111.
- ↵Stegmayr B, Asplund K, Wester PO. Trends in incidence, case-fatality rate, and severity of stroke in Northern Sweden, 1985–1991. Stroke. 1994; 25: 1738–1745.
- ↵Barker WH, Mullooly JP. Stroke in a defined elderly population, 1967–1985: a less lethal and disabling but no less common disease. Stroke. 1997; 28: 284–290.