|Home | About | Journals | Submit | Contact Us | Français|
To determine the relationship between heart rate response during low-grade physical exertion (six-minute walk) with mortality and adverse cardiovascular outcomes in the elderly.
Participants in the Cardiovascular Health Study, who completed a six-minute walk test, were included. We used delta heart rate (difference between post-walk heart rate and resting heart rate) as a measure of chronotropic response and examined its association with 1) all-cause mortality and 2) incident coronary heart disease (CHD) event, using multivariable Cox regression models.
We included 2224 participants (mean age 77±4 years; 60% women, 85% white). The average delta heart rate was 26 beats/min. Participants in the lowest tertile of delta heart rate (<20 beats/min) had higher risk-adjusted mortality (hazard ratio [HR] 1.18; 95% confidence interval [CI][1.00, 1.40]) and incident CHD (HR 1.37; 95% CI[1.05, 1.78]) compared to subjects in the highest tertile (≥30 beats/min), with a significant linear trend across tertiles (P for trend <0.05 for both outcomes). This relationship was not significant after adjustment for distance walked.
Impaired chronotropic response during six-minute walk test was associated with an increased risk of mortality and incident CHD among the elderly. This association was attenuated after adjusting for distance walked.
Previous studies have shown a strong association between an impaired heart rate response to exercise and clinical outcomes in a broad range of clinical populations.1–8 However, chronotropic assessment in these studies requires the use of a standardized, maximal exercise treadmill test. This is particularly cumbersome among the elderly, in whom musculoskeletal disorders and co-morbidities frequently limit the ability to perform a maximal exercise stress test.9
Distance walked on a six-minute walk test is used to assess functional status in patients with a wide range of cardiopulmonary disorders. 10–14 To our knowledge, the relationship of heart rate response during the test, an easily measured parameter that requires only two pulse measurements, with cardiovascular outcomes has not been previously studied.
Accordingly, we examined the risk of all-cause mortality and cardiovascular outcomes in subjects with an impaired heart rate response during a six-minute walk test enrolled in the Cardiovascular Health Study (CHS), a prospective cohort study of determinants of cardiovascular risk in older adults sponsored by the National Heart, Lung and Blood Institute. We hypothesized that a lower increase in heart rate on the six-minute walk test would reflect chronotropic incompetence and be associated with worse outcomes later in life.
The CHS is a prospective, community-based, longitudinal study of older adults. The main purpose of the study is to evaluate risk factors for the development and progression of cardiovascular diseases in older adults. The design of CHS has been previously described in detail. 15, 16 Briefly, 5201 ambulatory, non-institutionalizedmen and women who were 65 years or older were recruited during 1989–1990 from a random selection of Medicare eligibility lists in 4 US communities: Forsyth County, North Carolina (Wake Forest University School of Medicine, Winston-Salem); Sacramento County, California (University of California, Davis); Washington County, Maryland (Johns Hopkins University, Hagerstown); and Allegheny County, Pennsylvania (University of Pittsburgh, Pittsburgh). To increase participation of the African-American community, an additional 687 participantswere enrolled during 1992–1993 using the same sampling and recruitment methods. Each center’s institutionalreview committee approved the study, and all participants gaveinformed consent.
Comprehensive information on health-related variables was collected at baseline and annually thereafter in a standardized fashion from study participants. Clinic examinations were performed annually from 1989–1990 (baseline) to 1998–1999, and again in 2005–2006. Telephone contact was made annually from 1989–1999 and 2005–2006 and twice per year from 2000–2004 and 2006–2007. Standardized questionnaires were administered at a baseline home interview, at annual clinic visits, and during telephone contacts providing updated information on clinical co-morbidities, medications, and cardiovascular events. In addition, participants underwent echocardiography in 1994–1995 (2 years prior to the six-minute walk test) with measurement of left ventricular (LV) systolic function and assessment of valve disease. Descriptions of data collection methods, including instruments and protocols, have been reported previously.15
At year 9, 3400 participants attended the follow-up visit and were screened for eligibility to perform the six-minute walk test. Subjects were excluded if they had any acute medical conditions, resting heart rate < 50 or > 110 beats per minute (bpm), seated SBP > 200 or seated DBP > 110 mm Hg, resting pulse oximetry < 90%, severe aortic stenosis, recent or active angina, recent myocardial infarction or revascularization (within 3 months), need for an ambulatory aid such as a wheelchair, crutches, walker or cane or if they refused to provide informed consent.
The six-minute walk test was conducted according to a standardized protocol.17, 18 The test was administered by trained personnel and was performed in an internal hallway with the 100-foot distance marked by colored tape on the floor. After demonstration by study staff, participants were instructed to walk from one end of the hallway to the other at their own pace covering as much ground as possible. Participants were allowed to stop and rest during the test, but were instructed to resume walking as soon as they were able to do so, with the period of rest included in the six-minute period. A mechanical counter was used to count the number of completed laps, and an electronic timer with a buzzer sounded 6 minutes after the walk started to signal the completion of the test.
Heart rate was measured by trained staff using a Nellcor® Pulse Oximeter N-3000 and semi-disposable oximeter sensors (Nellco® D-25) attached to the index finger of the non-dominant hand. Presence of good perfusion was verified, and baseline heart rate was recorded in the seated position at the end of one minute. Post walk heart rate was recorded immediately after completion of the test. Distance walked was recorded in feet using the length of the hallway and the number of laps covered including the final partial lap.
We calculated delta heart rate as post walk heart rate minus baseline heart rate. Since there are no previous cutoffs for delta heart rate on the six-minute walk test, we created tertiles of this variable (lowest tertile < 20 bpm, middle tertile 20–30 bpm, highest tertile > 30 bpm) and also examined it as a continuous variable.
Our primary outcomes were (1) all-cause mortality and (2) incident coronary heart disease, defined as a combination of incident myocardial infarction or death due to atherosclerotic coronary heart disease.
Follow-up visits were conducted by telephone every six months and in person annually. Deaths were identified by a review of obituaries, medical records, death certificates, Centers for Medicare and Medicaid Services health care–utilization database for hospitalizations and from household contacts; 100 percent complete follow-up for ascertainment of mortality status was achieved. Myocardial infarction was ascertained from hospital records and was indicated by a clinical history of cardiac symptoms, elevated cardiac enzyme levels, and serial electrocardiographic changes.19 A CHS outcome-assessment committee adjudicated all outcomes including the principal cause of death. All events until June 30, 2007 were included.
We examined the distribution of patient demographics, prevalent co-morbidities and exercise test parameters according to tertiles of delta heart rate. We generated descriptive summary statistics for each category using means ± standard deviation (SD) for continuous variables and proportions for categorical variables. We also examined trends in these characteristics across tertiles of delta heart rate using Mantel Haenszel test of trend and linear regression as appropriate. Finally, we used Cox proportional hazards models to evaluate the relationship between delta heart rate and our study outcomes.
Candidate variables for inclusion into the models were age, sex, race, educational attainment (less than high school, high school, college or professional), smoking status (current, past, never), co-morbidities (history of hypertension, diabetes, coronary artery disease, congestive heart failure, COPD, cancer, presence of angina [based on Rose angina questionnaire], arthritis), body mass index (based on height and weight measured during the office visit), use of nodal blocking drugs (current beta blocker or digoxin use), cystatin C (measure of glomerular filtration rate), echocardiographic variables20, 21 (qualitative assessment of LV systolic function [normal or depressed], presence of significant aortic valve stenosis [severe impairment in aortic valve excursion]). All of the above variables were assessed at the time of Year 9 follow up visit, except echocardiographic variables which were assessed at Year 7.
Variables for model inclusion were selected based on clinical and statistical criteria. All variables with a significant unadjusted association with the outcome (p < 0.10) were included in the models; age, sex, and race were forced into the models regardless of statistical significance. Initial models adjusted for age, sex, and race only (Model 1). Subsequent models adjusted for smoking status, body mass index, hypertension, diabetes, history of angina, coronary heart disease, congestive heart failure, chronic obstructive pulmonary disease, cancer, adjusted Cystatin C level, use of nodal blocking medications, presence of LV dysfunction and educational attainment (Model 2). Since distance walked during the six-minute walk test is known to be an important determinant of subsequent cardiovascular events;13, 22 we further explored its relationship with delta heart rate. We hypothesized that distance walked would be a mediator of the relationship between delta heart rate and outcomes. Therefore, we included distance walked as a variable in our models (Model 3). Furthermore, we examined whether the relationship between delta heart rate and outcomes differed according to distance covered on the test by including an interaction term of delta heart rate with distance. Similarly, we also examined for interactions with age, sex and race. Analysis of time-varying covariates did not demonstrate a violation of the proportional hazards assumption.
Because of the potential pharmacological effect of beta blockers and digoxin on both baseline and post walk heart rate, we performed sensitivity analyses where we excluded all patients who were documented as taking these medications and repeated the analyses outlined above. Next, we also examined the association between heart rate response and clinical outcomes using an alternative definition of heart rate response. This was done using a linear regression model of post walk heart rate on baseline heart rate. From this model, we derived residual scores (difference between observed post walk heart rate and predicted post walk heart rate). We defined heart rate responders as those with a residual score greater than 1 standard deviation above the mean and all other patients as heart rate non-responders. We assessed the relationship between heart rate responder status and clinical outcomes using methods described above.
All P values were based on 2-sided tests, and were considered statistically significant at P less than 0.05. All analyses were performed using SAS version 9.2 (SAS Institute Inc, Cary, NC). The authors had full access to the data and take responsibility for the integrity of the results. All authors have read and agree to the manuscript as written. The study was approved by the Institutional Review Board at University of Iowa.
Of the 3400 subjects who attended the Year 9 follow up visit, 892 (26.4%) were excluded based on pre-defined criteria. Of the remaining 2508 patients, 213 (8.5%) subjects did not perform the six-minute walk test. As expected, these subjects were older, had a higher prevalence of co-morbidities and depressed LV function (eTable 1). The reasons for not performing the six-minute walk test in these patients are listed in eTable 2. Our final sample included 2224 subjects who completed the six-minute walk test and had non-missing data on heart rate variables. The mean age was 77.4 years; 59.6% were women and 85.3% were white. The average distance covered during the test was 342.7 m with a mean delta heart rate of 26 beats per minute.
Table 1 shows the baseline characteristics of study participants according to tertiles of delta heart rate. Subjects in the lowest tertile of delta heart rate were more likely to be men, of older age, with lower body mass index, higher prevalence of smoking, and were more likely to be on beta blockers or digoxin (P < 0.01 for all). They also had higher baseline heart rates, achieved lower post walk heart rates and covered shorter distances during the test compared to subjects in the highest tertile of delta heart rate (P < 0.01 for all). Prevalence of other co-morbidities did not differ significantly between the groups.
During a mean follow up of 8.5 years, there were 998 deaths (44.9%), with 257 (11.5%) due to coronary heart disease. Among 2023 participants without previous myocardial infarction, there were 241 (10.8%) incident cases of myocardial infarction. A low delta heart rate was associated with increased risk of clinical outcomes (Table 2). Risk of adjusted all-cause mortality was higher in subjects in the lowest tertile (hazard ratio [HR] 1.18; 95% confidence interval [CI] [1.00, 1.41]) and middle tertile (HR 1.13; 95% CI [0.96, 1.34]) with a linear trend across tertiles (P for trend 0.04, Table 2). Likewise, risk of incident coronary heart disease was also higher in this group (lowest tertile: HR 1.37; 95% CI [1.05, 1.78]; middle tertile HR 1.29; 95% CI [0.99, 1.68]; P for trend 0.02; Table 2). When we included distance walked in the model (model 3), the association between delta heart rate and mortality was completely eliminated. There was no interaction of the effect of delta heart rate on clinical outcomes by distance walked, age, gender, race, and baseline heart rate (P for all interactions > 0.20). Importantly, subjects who did not complete the walk test had a significantly higher risk of mortality (adjusted hazard ratio 1.35; 95% CI [1.04, 1.75]; P value 0.02) compared to those who completed the test.
In sensitivity analyses, we excluded 465 patients on nodal blocking drugs (beta blockers or digoxin) at the time of the visit. The adjusted risk for all-cause mortality (lowest tertile HR 1.22, 95% CI [1.01, 1.48]; middle tertile HR 1.13, 95% CI [0.94, 1.37; P for trend 0.04) and incident coronary heart disease (lowest tertile HR 1.46, 95% CI [1.08, 1.99]; middle tertile HR 1.33, 95% CI [0.98, 1.817; P for trend 0.02) was similar to that observed in the entire cohort. Using an alternative definition of heart rate response (residual scores derived from a regression model), we found results that were qualitatively similar to our main results (eTable 2). Finally, arthritis of hip and knee were not associated with the study outcomes and their inclusion into the models did not materially change our findings (results not shown).
Among a population of community dwelling older adults, we found that a lower increase in heart rate during the six-minute walk test was associated with an increased risk of all-cause mortality and coronary heart disease outcomes and this risk was greatest among those with the lowest increase in heart rate (< 20 beats/min). Subjects who failed to complete the six-minute walk test were at the highest risk. The association with mortality was completely eliminated after adjustment for distance walked.
First reported nearly 40 years ago,7 the relationship between impaired heart rate response during exercise and clinical outcomes has been well established in a broad range of clinical populations.2, 4–6 Lauer et al found a low chronotropic index during exercise to be strongly associated with all-cause mortality, even after accounting for the presence of perfusion defects on stress imaging5 or abnormalities on stress echocardiography.8 Other studies have shown this measure to be superior to coronary angiography in predicting clinical outcomes.4, 6 Our prior work has shown that an impaired heart rate response during exercise is associated with adverse cardiovascular outcomes and atherosclerotic disease progression even among patients who have undergone revascularization with coronary artery bypass graft surgery.3
However, the above studies have relied on the use of a maximal exercise treadmill test to determine chronotropic response. Given the physical limitations that come with aging including arthritis and osteoporosis, many elderly patients are unable to complete a maximal exercise treadmill test.9 The six-minute walk test on the other hand is much simpler to perform, and requires only a measured hallway and stopwatch. Thus, our finding of a relationship between a low delta heart rate on the six-minute walk test carries particular relevance among older patients who may be otherwise limited in performing more sophisticated assessments like a cardiopulmonary exercise test. Moreover, our findings are consistent with a previous study that found delta heart rate on a submaximal treadmill test to be as strongly associated with mortality and CHD incidence as more sophisticated measures of chronotropic incompetence like chronotropic reserve index (which usually requires a maximal exercise test).23 Thus, our findings would suggest that an impaired heart rate response on the six-minute walk test can be useful in assessing the risk of adverse outcomes in this population.
The association between delta heart rate and clinical outcomes was not significant after adjustment for distance walked. This finding was not entirely unexpected, since impaired heart rate response to exercise would limit functional capacity and the total distance covered and thus, may be in the causal pathway.24 While low delta heart rate was not found to be independently associated with clinical outcomes after adjusting for total distance walked, this finding does not negate the biological association between heart rate response with exercise and clinical outcomes. Furthermore, measurement of heart rate is substantially simpler in clinical practice compared to distance walked; the latter requires a measured hallway free of interference or traffic, real-time accounting of the number of laps covered during a six-minute period which may be cumbersome in busy clinical practice. Thus, the six-minute walk provides easily obtainable data, including ability to complete the walk and heart rate at the beginning and end, which convey important prognostic information with little inconvenience, monitoring, or specialized equipment.
It is important to note that this finding was significant even after adjusting for a wide range of co-morbidities including LV function. In fact, LV function was preserved in a vast majority of participants (92%) in our study, suggesting the role of other factors in causing an impaired heart rate response. Previous studies have suggested an important role for autonomic dysregulation in mediating this relationship,25, 26 however the underlying mechanisms are still elusive.
The strengths of our study include a diverse, community-based, older population with extensive assessment of cardiovascular and overall health at baseline and over time, long-term follow-up and formally adjudicated clinical events. Potential limitations should also be noted. First, participants were not asked to discontinue beta blockers or digoxin prior to the walk test. However, we adjusted for use of these drugs in our analysis and also performed sensitivity analyses excluding patients taking these drugs; our overall study results were not significantly changed using either of these strategies. Second, detailed information in CHS allowed us to adjust for the confounding effect of a number of co-morbidities that may cause functional limitation. However, the potential for residual confounding may still remain. Third, since our study included only older subjects and a vast majority of them had normal LV function, these findings may not be generalized to younger or sicker patients. Finally, we acknowledge that when feasible, a maximal exercise test may yield superior data regarding heart rate response, functional capacity and overall prognosis particularly when combined with expired gas analysis, albeit at a higher cost.
To conclude, an impaired heart rate response on the six-minute walk test was associated with an increased risk of death and adverse coronary outcomes among older adults. It also was associated with functional capacity as measured by ability to walk longer distances, which appeared to account for its association with clinical events. Although delta heart rate was not superior to distance walked, it is substantially easier to measure which would make it attractive for prognostic assessment in busy clinical practice.
The Cardiovascular Health Study is supported by contracts N01-HC-85239, N01-HC-85079 through N01-HC-85086, N01-HC-35129, N01 HC-15103, N01 HC-55222, N01-HC-75150, N01-HC-45133, and grant HL080295 from the National Heart, Lung, and Blood Institute (NHLBI), with additional contribution from the National Institute of Neurological Disorders and Stroke (NINDS). Additional support was provided through AG-023629, AG-15928, AG-20098, and AG-027058 from the National Institute on Aging (NIA). A full list of principal CHS investigators and institutions can be found at http://www.chs-nhlbi.org/pi.htm. No external funds were used for this study