| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
1 Division of Epidemiology and Community Health, School of Public Health, University of Minnesota, Minneapolis.
2 Department of Medicine, University of Vermont, Burlington.
3 Department of Epidemiology, University of Washington, Seattle.
4 Collaborative Studies Coordinating Center, University of North Carolina, Chapel Hill.
5 Center on Aging and Health, Johns Hopkins University, Baltimore, Maryland.
Address correspondence to Aaron R. Folsom, MD, MPH, Division of Epidemiology and Community Health, School of Public Health, University of Minnesota, Suite 300, 1300 South Second St., Minneapolis, MN 55454-1015. E-mail: folsom{at}epi.umn.edu
 |
A |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Methods. We conducted a prospective cohort study in four U.S. communities involving 4859 participants 65 years old and older. At baseline, in 1989–1993, we assessed frailty based on weight loss, grip strength, feelings of exhaustion, walk time, and physical activity. Incident VTE (deep vein thrombosis or pulmonary embolus) through 2002 was identified by review of hospital records.
Results. Fifty-two percent of the sample was classified as having intermediate or definite frailty. After adjustment for age, race, sex, body mass index, and diabetes, the relative risk of total VTE (n = 150) for people who were frail compared with no frailty was 1.31 (95% confidence interval [CI], 0.93–1.84). The comparably adjusted relative risk for idiopathic VTE (n = 58) was 1.79 (95% CI, 1.02–3.13).
Conclusions. The incidence rates of idiopathic VTE was higher in community-dwelling older adults with baseline frailty compared with no frailty. Further studies of the clotting process in frailty may allow the development of an improved strategy to determine VTE risk in this vulnerable subset of older adults.
Rates of venous thrombosis and pulmonary embolism—collectively called venous thromboembolism (VTE)—also are highest in old age (4–6). A typical component of frailty—immobility—is widely considered to be a risk factor for VTE. Frailty has recently been linked to activation of coagulation and inflammatory pathways (7,8), and elevation of certain coagulation factor levels, particularly factor VIII (9), are VTE risk factors. Venous stasis, which may be associated with frailty and immobility, also activates the coagulation system (10). Yet, to our knowledge, no study has established a clear link between frailty in the outpatient setting and subsequent VTE. We hypothesized that frailty would be a risk factor for VTE occurrence in a prospective study of free-living older adults.
|
METHODS |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
As described in detail elsewhere (3), CHS defined frailty at baseline by combinations of weight loss (>10 lbs unintentionally in prior year); poor grip strength (lowest sex- and body mass index [BMI]–specific quintile); feeling exhausted (self-report); slow 15-foot walk time (lowest quintile); and low physical activity (lowest sex-specific quintile). Having 
3 criteria was labeled "definite frailty," and having 1–2 criteria was labeled "intermediate frailty." Frailty by this definition independently predicted increased falls, disability, subclinical cardiovascular disease, hospitalizations, and death in CHS (3,13).
During cohort follow-up, CHS staff telephoned or examined participants every 6 months. Hospitalizations were identified by participants or their proxies, and by periodic search of Medicare records. For every hospitalization, staff obtained discharge summaries and The International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) discharge codes. Validation of hospitalized VTE events was accomplished as an ancillary study, the Longitudinal Investigation of Thromboembolism Etiology (LITE) (5). Two physicians reviewed and adjudicated possible VTE events (venous thrombosis or pulmonary embolism) using standardized criteria, and they classified each VTE as idiopathic or secondary (occurring within 90 days of major trauma, surgery, marked immobility, or associated with active cancer or chemotherapy). Although the LITE methods article (5) describes validation through June 1997, we used identical methods and extended follow-up for CHS through 2002 for this report.
Our hypothesis was that frailty would be associated with incident VTE independent of other demographic and behavioral VTE risk factors in the LITE cohort, namely age, ethnicity, sex, BMI, and diabetes (9,14). We also sought, in a secondary analysis, to evaluate the potential mediating role of plasma factors VII and VIII, and CRP, which were measured at baseline in the whole CHS cohort. Because frailty could be affected, we excluded (sequentially), from a starting sample size of n = 5888, those persons with a history of Parkinson's disease (n = 47), stroke (n = 305), Mini-Mental State Examination scores < 18 (n = 71), and persons who were taking antidepressants (n = 179). To focus on incident VTE, we excluded participants who reported a history of VTE (n = 296) or who were taking warfarin at baseline (n = 36). We also excluded 95 persons with missing frailty information. After all exclusions, 4859 participants remained at risk for VTE. We compared levels of VTE risk factors among categories of frailty (no frailty, intermediate frailty, definite frailty) using analysis of covariance. We computed incidence rates of VTE over a mean follow-up of 9.3 years by frailty category using Poisson regression. We calculated hazard ratios (relative risks), adjusted for age, race, sex, and other covariates, using proportional hazards regression models.
|
RESULTS |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
|
As shown in Table 2, after adjustment for age, race, and sex (Model 1), the relative risks of VTE for the intermediate and definite frailty groups were approximately 40% higher than that for the no frailty group. For total frailty, the relative risk (95% confidence interval) was 1.41 (1.01–1.98). With further adjustment for BMI and diabetes (Model 2), this relative risk was 1.31 (0.93–1.84), and with further adjustment for habitual physical activity (Model 3), it was 1.37 (0.97–1.95). When analyses were restricted to idiopathic VTE, the relative risks were higher than that for all VTE: 1.83 (1.05–3.19) for Model 1 and 1.79 (1.02–3.13) for Model 2 comparing total frailty with no frailty.
|
|
|
DISCUSSION |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Frailty is a complex syndrome the definition of which has not reached full consensus. Investigators in CHS have shown that there is an increasing gradient of subsequent hospitalization, disability, and death across 3 frailty categories (3). Furthermore, compared to nonfrail CHS participants, the intermediate frailty group had a high probability of definite frailty in subsequent years (3). Thus, it is not surprising that incidence rates of VTE were similar for the two frailty groups.
Physical inactivity is a component of our frailty definition. However, it did not appear to be the primary explanation of why frailty and VTE are associated, because adjustment for physical inactivity did not greatly change the relative risk. This is because baseline physical activity level was only weakly, and not statistically significantly, associated with VTE incidence in this cohort (14).
Recent evidence has linked frailty with activation of the inflammatory and coagulation systems (7), and this activation may be associated with subsequent decline in function and mortality in elderly persons (8). Therefore, we hypothesized that any association between frailty and VTE might be mediated via these pathways. The findings from our supplementary analysis adjusting for factors VII and VIII and CRP (Table 2, Model 4) suggested that indeed these pathways could be important. However, factor VII and CRP contributed very little beyond factor VIII in explaining the association between frailty and VTE. Further, frailty seemed to be a somewhat weaker risk factor in those persons with higher factor VIII levels. Additional studies of the link between frailty, factor VIII, and VTE seem warranted, as well as any contribution of other inflammatory markers and cytokines (e.g., tumor necrosis factor-
) to the process.
Clinical perceptions that frail people are likely to develop VTE have exclusively been based on hospitalized ill patients. This study demonstrates that even community-dwelling frail persons are at increased risk of VTE. Limitations of this study include the small number of participants with definite frailty and the relatively few VTE events. We combined the definite and intermediate frailty groups, not only because their VTE rates were similar, but because the VTE risk factor levels were higher for the intermediate group than for the nonfrail participants. We also did not have D-dimer measurements at baseline for analysis on the whole cohort; elevated D-dimer has been linked to both frailty (7,8) cross-sectionally and VTE occurrence prospectively (15).
Summary
Community-dwelling older adults with intermediate to definite frailty have a moderately increased risk of idiopathic VTE. It is unclear whether this increase is directly related to lower levels of physical activity or to altered biology that influences clotting processes. The incidence rate of VTE of approximately 4 per 1000 person-years, though substantial, is too low to warrant VTE prophylaxis of outpatient frail elders. Further studies of the clotting process in frailty may allow the development of an improved strategy to determine VTE risk in this vulnerable subset of older adults.
|
A |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
We thank Laura Kemmis for manuscript preparation and acknowledge the long-term and valuable contributions of the investigators, staff, and participants of the CHS.
|
F |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Received January 31, 2006
Accepted April 2, 2006
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  G. Piazza, A. Seddighzadeh, and S. Z. Goldhaber Deep-Vein Thrombosis in the Elderly Clinical and Applied Thrombosis/Hemostasis, October 1, 2008; 14(4): 393 - 398. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
