Survival and Functional Status 20 or More Years After First Stroke
The Framingham Study
Background and Purpose—We examined the 20-or-more-year survival and functional levels of 148 stroke survivors and 148 age- and sex-matched control subjects from the Framingham Study Cohort, whom we originally studied in 1972–1974 to ascertain the survival and disability status of stroke survivors compared with that of controls.
Methods—This long-term evaluation was done with use of data from the 1993–1995 Framingham Study Cohort Examination 23 on the 10 stroke survivors and 20 control subjects still living to identify and compare the host characteristics and functional status of each group. The survival curves for both stroke survivors and controls were derived from the ongoing Framingham Study database.
Results—Twenty-plus-year stroke survivors experienced a greater mortality than age- and sex-matched controls (92.5% and 81%, respectively). The slopes of the two survival curves were essentially the same. Functional status (eg, walking and independence in activities of daily living) of stroke survivors, however, compared very favorably with that of the control subjects. Stroke survivors were more likely to be female and to have a number of comorbidities, including elevated blood pressures, greater use of medications, less use of alcohol, and less depressive symptomology.
Conclusions—In the Framingham cohort, 20-plus-year stroke survivors showed greater mortality than age- and sex-matched control subjects; functionally, however, the groups were very similar and in general quite independent.
From 1972 to 1974, we studied all of the then-living stroke survivors in the original Framingham Study cohort and an equal number of age- and sex-matched control subjects who were free of stroke to determine the frequencies of various types of disability.1 2 This group comprised 148 persons in each category (stroke survivors and age- and sex-matched controls). That study has been extensively cited, because it (1) measured the frequencies of various disabilities through objective methods and produced normative data on the prevalence of specific disabilities in stroke survivors, (2) documented that psychosocial disabilities after stroke were more frequent than physical ones, (3) established that the frequency of each type of disability was greater in stroke survivors than in control subjects, and (4) showed that stroke survivors had a significant amount of cardiovascular and other comorbidities that accounted for a portion of their disabilities. The ages of the stroke survivors at that time were 55 to 87 (mean, 73.8) years, and the mean ages at most recent stroke were 64.3 years for the 66 men and 65.7 years for the 82 women.
During 1993–1995 (Framingham Study Cohort Examination 23; December 13, 1993, through June 13, 1995), approximately 20 years later, we had the opportunity to review this original group of stroke survivors and age- and sex-matched control subjects. We examined the survival patterns of both stroke survivors and controls and the functional status of both groups.
Subjects and Methods
The Framingham Study cohort, 5209 men and women between 30 and 62 years of age at entry in 1948 to 1950, has been examined every 2 years since the initiation of the study. The details of the study design, implementation, and diagnostic criteria have been published previously.3 At each examination the participant receives a physical examination with laboratory testing. Documentation of interim illnesses and hospitalizations are reviewed. The records and examinations for all cardiovascular and cerebrovascular events are reviewed and confirmed by a panel of investigators.
Physical evaluation and documentation for events and functional status in the stroke survivors and control subjects were abstracted from Framingham Study Cohort Examination 23, which took place between 1993 and 1995. Standardized tests used to document health status at Examination 23 included the Mini-Mental State Examination4 for cognitive function, the Center for Epidemiologic Studies-Depression Scale5 for depressive symptomology, and the Katz Index6 and Rosow-Breslau scale7 for physical functioning.
As noted in the introduction, our previous disability study1 2 was carried out from April 1972 through March 1975. At that time, 354 cases of stroke had been documented; 155 of these patients were still living. Seven stroke survivors (3 men and 4 women) refused to be examined. The remaining 148 (95% of the eligible total) were the actual participants in the study. Evaluations were done at least 6 months after the most recent stroke to minimize the possible effects of a changing neurological status.
Each of the 148 stroke survivors evaluated was matched by sex and age with a stroke-free member of the same cohort (except for one 80-year-old whose control was aged 82). Control subjects were evaluated in the same manner as stroke survivors. The data from the functional assessments were subsequently combined with those from the regular medical and neurological examinations of the same subjects by the Framingham Study physicians.
The significance of the difference between the frequencies for each specific variable in the stroke survivor and control groups was determined by the χ2method. Nine types of functional disability were documented. The final set of variables involved the presence or absence, in both stroke and control subjects, of 14 comorbid conditions. The choice of comorbid processes to be documented was determined by the preexisting protocol of the Framingham Study.
For this 20-year follow-up study, the biennial records of the 148 original 1972–1974 stroke survivors and those of the 148 age- and sex-matched controls were retrieved from The Framingham Heart Study Archives. Serendipitously, those subjects still living had just been evaluated in 1993–1995 (during Framingham Study Cohort Examination 23), and thus data on the current status of the entire study group still living were available. The charts of the stroke survivors and controls of the 1972–1974 cohort who were still living were reviewed to obtain information on their functional status in the 1993–1995 examination cycle of the overall Framingham Stroke Study. The survival curves for both stroke survivors and controls were derived from the Framingham Study database.
One hundred forty-seven pairs of 1972–1974 stroke survivors and stroke-free control subjects were available for analysis in the database (one pair from the original 148 was deleted on retrospective review). Table 1⇓ shows the survival status of the 147 pairs.
As shown, only 10 (6.8%) of the 1972–1974 stroke survivors were still living in the 1993–1995 period, compared with 20 (13.6%) of the age- and sex-matched control subjects free of stroke in 1972–1974. The difference in frequencies of survival status is significant at the P=.007 level. The differential mortality between the two groups is also displayed in Fig 1⇓.
The survival curves for the 1972–1974 stroke survivors and controls were essentially of the same configuration with a greater level of survival, at each point in time, for the controls (Fig 1⇑). The age at first stroke of the 147 1972–1974 survivors is shown in Table 2⇓. The 10 1972–1974 stroke survivors who were still living in 1993–1995 had a mean age at first stroke of 56.0±3.4 years as opposed to 68.8±1.7 years for the 26 who died of stroke and 65.9±0.9 years for those dying of other causes. The younger age at first stroke of the 20-year stroke survivors is noteworthy. Table 2⇓ also shows that the mean total number of years of survival after first stroke for the 1972–1974 stroke survivors who died before 1993–1995 was 14.3 (range, 0.7 to 38.3) years. The 26 1972–1974 stroke survivors who had died of stroke lived a mean of 10.7 (range, 2.5 to 31.4) years. Of the 26 individuals who died of stroke, 22 had a least one recurrent stroke and 4 died of the original stroke. In contrast, the 106 1972–1974 stroke survivors who died of causes other than stroke lived for a mean of 14.8 years. The 1972–1974 stroke survivors still living in 1993–1995 had lived an average of 27.4 (range, 20 to 38) years after their first stroke.
The host characteristics and functional status of the 10 1972–1974 stroke survivors still living in 1993–1995 (Framingham Study Cohort Examination 23) are shown in Tables 3⇓ and 4⇓. One of the 10 died early in 1995, before follow-up evaluation (Examination 23) was performed. Table 3⇓ shows the host characteristics of the remaining 9. Their mean age was 80.3 (range, 74 to 90) years, 7 were women, and less than half were married. The mean number of years since first stroke was 27.4 (range, 20 to 38); 5 first strokes had been atherothrombotic brain infarctions, 3 had been subarachnoid hemorrhages, and 1 was classified as other. Of the 3 survivors who had the SAH stroke subtype, 2 did not have surgical treatment for the hemorrhages. Five had possible coronary artery disease, and their diastolic blood pressures ranged from 54 to 98 mm Hg (systolic blood pressures ranged from 106 to 225 mm Hg). None were smokers or drank alcohol. All were taking some type of medication, and 8 of the 9 drank 1 to 2 cups of coffee per day.
As shown in Table 4⇑, the functional status of the 9 remaining stroke survivors was excellent except for 1 patient who was demented (requiring total care) and living in a nursing home. The other 8 could still walk (although 6 required a walker or cane), all were independent in activities of daily living (2 with assistive devices) and instrumental activities of daily living, and all were still driving cars. In addition, 8 of these very-long-term stroke survivors had normal cognitive function, and only 1 showed depressive symptoms. Only 2 of the 8 had sustained falls during the previous year. Seven of the 8 were living in their own homes, 5 alone and 2 with a spouse.
Table 5⇓ presents selected comparative characteristics of the 9 stroke survivors and the 20 members of the age- and sex-matched control group. Although the numbers are too small to justify the use of tests for statistically significant differences, the relative frequencies of several characteristics are worthy of note. There were more women than men in the stroke survivor group compared with the control group (78% and 35%, respectively). The stroke group showed higher systolic and diastolic blood pressures, more comorbidity (56% to 5%), greater use of medications, less use of alcohol, and less depressive symptomology. There was little difference in walking and independence in activities of daily living.
Very-long-term outcome studies of stroke survivors are rare. In our initial 1972–1974 study of this cohort, 67% of the cases were at 20 or more months after first stroke,1 2 but some were as close as 6 months after. The 1972 study by Moskowitz and colleagues8 presented outcome data on 518 stroke patients collected over a 3-year period. These study subjects were only 2 years post-onset. Both studies indicated that poststroke disability (as opposed to survival) was not as severe as previously believed (confirming the previous work of Rankin9 and Adams and Merrett10 ).
The cohort of 675 stroke patients in the Oxfordshire Community Stroke Project was evaluated at 2 to 6.5 years after stroke.11 The overall survival rate was 51% at this time. Functional status of the survivors was not given. Scmidt and colleagues12 followed 1 538 stroke patients in Moscow for 7 years. At that time 76.5% were dead, but 81% of the 7-year survivors were described as “independent.” A 10-year follow-up study from Poland13 reviewed 195 cases. At the end of the 10 years, 135 (69%) had died, 15 were discharged home, and 44 (26 females and 18 males) were “rehabilitated.” Motor improvement was described as “very good” in 21 patients. One was lost to follow-up.
The Finnish study by Tuomilehto and colleagues14 looked at psychosocial and health status after 14 years in a cohort of 1 241 persons with stroke in 1972–1974. Of these, 241 (19.4%) were still living (80% at home or with relatives), and 2 of 3 were functionally independent. Ten to fifteen percent had symptoms of depression, but half of the survivors considered their health status to be good.
The Shibata Study15 examined a cohort of 2302 subjects for 15.5 years. Seventy-eight had strokes during this period, for an incidence rate of 4.36 per 1000 person-years, but only risk factors (not outcomes) were reported. Matsumoto and colleagues16 also did a 15-year follow-up study in Rochester, Minn, but only survival (not functional status) was measured at that interval.
To our knowledge, ours is the first study in the English-language literature on stroke survivors performed over 20 years after ictus, the mean interval between onset and assessment for the survivors being 27.4 (range, 20 to 38 years). As shown in this study and others, the progressive mortality in stroke survivors is inexorable and greater than that of stroke-free controls, following a similar pattern at a somewhat higher level. The fortunate few, however, who live for long periods of time appear to enjoy relatively good functional status. In our data, this may be related to the previously noted fact that the 20-year stroke survivors had their first stroke at a younger age than both those stroke survivors who died of stroke and those who died of other causes. In many instances, however, we found the functional status of the stroke survivors to be better than that of the age- and sex-matched controls, who were free of stroke when the 20-year period of observation began. It will be of great interest to see whether other controlled studies of very-long-term stroke survivors replicate these findings.
This study was supported by grants from the National Institute of Neurological Disorders and Stroke (2-R01-NS-17950-15) and the Framingham Heart Study (supported by NIH/NHLBI contract NO1-HC-38038).
- Received October 28, 1997.
- Revision received December 29, 1997.
- Accepted January 13, 1998.
- Copyright © 1998 by American Heart Association
Shurtleff D. Some characteristics related to the incidence of cardiovascular disease and death: Framingham Study, 18-year follow-up. In: Kannel WB, Gordon T, eds. The Framingham Study: An Epidemiological Investigation of Cardiovascular Disease. Washington, DC: Dept of Health, Education, and Welfare; 1974. DHEW publication NIH 74–599 (section 30).
Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas. 1977;1:385–401.
Katz S, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW. Studies of illness in the aged: the index of ADL: a standardized measure of biological and psychological function. JAMA. 1963;21:914–919.
Rosow I, Breslau N. A Guttman Health Scale for the aged. J Gerontol. 1966;21:556–559.
Adams GF, Merrett JD. Prognosis and survival in the aftermath of hemiplegia. Br Med J. 1961;1:309–314.
Dennis MS, Burn JPS, Sandercock PAG, Bamford JM, Wade DT, Warlow CP. Long-term survival after first-ever stroke: The Oxfordshire Community Stroke Project. Stroke. 1993;24:796–800.
Scmidt EV, Smirnov VE, Ryabova VS. Results of the seven-year prospective study of stroke patients. Stroke. 1988;19:942–949.
Wisniewska-Roszkowska K, Jedynecki A, Ziolkowski W. 10 years observation and rehabilitation of stroke disability: a longitudinal study. Gerontol Clin. 1975;17:67–71.
Tuomilehto J, Nuottimaki T, Salmi K, Aho K, Kotila M, Sarti C, Rastenyte D. Psychosocial and health status in stroke survivors after 14 years. Stroke. 1995;26:971–975.
Nakayama T, Date C, Yokoyama T, Yoshiike N, Yamaguchi M, Tanaka H. A 15.5 year follow-up study in stroke in a Japanese provincial city: The Shibata Study. Stroke. 1997;28:45–52.
Matsumoto N, Whisnant JP, Kurland LT, Okazaki H. Natural history of stroke in Rochester, Minnesota, 1955 through 1969: an extension of a previous study, 1945 through 1954. Stroke. 1973;4:20–29.