Functional Decline in Peripheral Arterial Disease: Associations With the Ankle Brachial Index and Leg Symptoms

Context Among individuals with lower-extremity peripheral arterial disease (PAD), specific leg symptoms and the ankle brachial index (ABI) are cross-sectionally related to the degree of functional impairment. However, relations between these clinical characteristics and objectively measured functional decline are unknown.

Objective To define whether PAD, ABI, and specific leg symptoms predict functional decline at 2-year follow-up.

Design, Setting, and Participants Prospective cohort study among 676 consecutively identified individuals (aged ≥55 years) with and without PAD (n = 417 and n = 259, respectively), with baseline functional assessments occurring between October 1, 1998, and January 31, 2000, and follow-up assessments scheduled 1 and 2 years thereafter. PAD was defined as ABI less than 0.90, and participants with PAD were categorized at baseline into 1 of 5 mutually exclusive symptom groups.

Main Outcome Measures Mean annual changes in 6-minute walk performance and in usual-paced and fast-paced 4-m walking velocity, adjusted for age, sex, race, prior-year functioning, comorbid diseases, body mass index, pack-years of cigarette smoking, and patterns of missing data.

Results Lower baseline ABI values were associated with greater mean (95% confidence interval) annual decline in 6-minute walk performance (−73.0 [−142 to −4.2] ft for ABI <0.50 vs −58.8 [−83.5 to −34.0] ft for ABI 0.50 to <0.90 vs −12.6 [−40.3 to 15.1] ft for ABI 0.90-1.50, P = .02). Compared with participants without PAD, PAD participants with leg pain on exertion and rest at baseline had greater mean annual decline in 6-minute walk performance (−111 [−173 to −50.0] ft vs −8.67 [−36.9 to 19.5] ft, P = .004), usual-pace 4-meter walking velocity (−0.06 [−0.09 to −0.02] m/sec vs −0.01 (−0.03 to 0.003] m/sec, P = .02), and fastest-pace 4-meter walking velocity (−0.07 [−0.11 to −0.03] m/sec vs −0.02 [−0.04 to −0.006] m/sec, P = .046). Compared with participants without PAD, asymptomatic PAD was associated with greater mean annual decline in 6-minute walk performance (−76.8 (−135 to −18.6] ft vs −8.67 (−36.9 to 19.5] ft, P = .04) and an increased odds ratio for becoming unable to walk for 6 minutes continuously (3.63; 95% confidence interval, 1.58-8.36; P = .002).

Conclusions Baseline ABI and the nature of leg symptoms predict the degree of functional decline at 2-year follow-up. Previously reported lack of worsening in claudication symptoms over time in patients with PAD may be more related to declining functional performance to than lack of disease progression.

Cross-sectional studies demonstrate that distinct types of leg symptoms reported by patients with peripheral arterial disease (PAD) in the lower extremities are associated with varying degrees of functional impairment.^1,2 Severity of PAD, as measured by the ankle brachial index (ABI), is also associated with the degree of functional impairment.^2,3 However, relationships between the ABI, leg symptoms, and functional decline are unknown.

Observational studies in the 1960s and 1970s suggested that the natural history of lower-extremity disease in patients with PAD and intermittent claudication was benign.^4-6 In these series, just 15% to 30% of individuals with claudication reported symptomatic worsening over 5- to 10-year follow-up. Currently, many medical textbooks and review articles report that most persons with intermittent claudication have stabilization or improvement in their symptoms over time.^7-10 However, symptoms may not correlate with objective measures of functional decline. It is possible that the low rate of symptomatic worsening reported in previous research has misled clinicians about the true natural history of PAD. However, if patients with PAD reduce their activity to keep leg symptoms in check, patient-reported improvement or stabilization of leg symptoms may mask PAD-associated functional decline. Therefore, in a prospective study of men and women with and without PAD, we assessed relationships between the ABI and specific leg symptoms and changes in lower-extremity functioning at 2-year follow-up.

The protocol was approved by the institutional review boards of Northwestern University and Catholic Health Partners Hospital. Participants gave written informed consent. Our original protocol specified our aim to determine the associations between baseline ABI categories and decline in lower-extremity functioning. Based on recent cross-sectional data showing significant associations between specific leg symptoms and the degree of functional impairment in patients with PAD,¹ our aim to assess the association between specific leg symptoms and decline in lower-extremity functioning was added after funding was obtained.

Participants were aged 55 years and older. Participants with PAD were identified consecutively from patients who tested positive for PAD in 3 Chicago-area noninvasive vascular laboratories. Half of the non-PAD participants were identified consecutively from patients who tested negative for PAD. Remaining non-PAD participants were identified consecutively from patients aged 55 years and older in a general medicine practice at Northwestern University. Baseline visits occurred between October 1, 1998, and January 31, 2000. Follow-up visits were scheduled 1 and 2 years after baseline.

All participants underwent ABI testing at their baseline study visit. PAD was defined as ABI less than 0.90.^11-14

Exclusion criteria have been previously reported.¹ Patients with dementia, recent major surgery, or foot or leg amputations were excluded, as were nursing home residents, wheelchair-bound patients, non–English-speaking patients (because investigators were not fluent in languages other than English), and individuals with ABIs greater than 1.50.^1,11 Individuals with PAD diagnosed in the noninvasive vascular laboratory were excluded if their ABI at the baseline visit indicated absence of PAD. This occasionally occurred in patients with PAD who were revascularized after vascular laboratory testing or in individuals with ABI values of approximately 0.90, due to measurement variation. Patients with an ABI of 0.90 or greater and with prior lower-extremity revascularization (n = 16) were excluded since they could not clearly be classified as with or without PAD. Participants with PAD who underwent lower-extremity revascularization after baseline were excluded (n = 17) since revascularization may affect the natural history of lower-extremity functioning.

Using established methods, a hand-held Doppler probe (Nicolet Vascular Pocket Dop II; Nicolet Biomedical Inc, Golden, Colo) was used to obtain systolic pressures in the right brachial, dorsalis pedis, and posterior tibial arteries; left dorsalis pedis and posterior tibial arteries; and left brachial artery.^15,16 Appropriately sized cuffs were used and deflated at a rate of 2 mm Hg per second. Systolic pressures were obtained during deflation. Each pressure was measured twice: in the order listed and then in reverse order. The ABI was calculated in each leg by dividing the mean of the dorsalis pedis and posterior tibial pressures in each leg by the mean of the 4 brachial pressures.¹⁵ Average brachial pressures in the arm with highest pressure were used when 1 brachial pressure was higher than the opposite brachial pressure in both measurement sets and when the 2 brachial pressures differed by 10 mm Hg or more in at least 1 measurement set, since in such cases subclavian stenosis was possible.^15,16 The lowest leg ABI was used in analyses.

For participants with PAD, leg symptoms were classified into 1 of 5 groups based on responses to the San Diego Claudication Questionnaire,^1,17 which is derived from the Rose Claudication Questionnaire.¹⁸ Four groups had exertional leg symptoms, based on an affirmative response to the question, "Do you get pain in either leg or buttock on walking?" These participants were further classified as follows based on their responses to the San Diego Claudication Questionnaire: (1) intermittent claudication (exertional calf pain that does not begin at rest, causes the participant to stop walking, and resolves within 10 minutes of rest); (2) leg pain on exertion and rest (exertional leg pain that sometimes begins at rest); (3) atypical exertional leg pain/carry on (exertional leg symptoms that do not begin at rest and do not stop the individual while walking); and (4) atypical exertional leg pain/stop (exertional leg symptoms that do not begin at rest, stop the individual from walking, and do not involve the calves or resolve within 10 minutes of rest). A fifth group was defined as asymptomatic because they reported no pain in either leg or buttock while walking.

Algorithms developed for the Women's Health and Aging Study were used to document comorbid diseases.¹⁹ These algorithms combine data from patient report, physical examination, medical record review, medications, laboratory values, and a primary care physician questionnaire. American College of Rheumatology criteria were used to diagnose osteoarthritis of the knee or hip.^20,21 Comorbid diseases assessed were diabetes mellitus, angina, heart failure, myocardial infarction, stroke, arthritis of the knee, arthritis of the hip, hip fracture, spinal stenosis, disk disease, pulmonary disease, and cancer.^22-25

Functional measures were performed by a health interviewer blinded to the patients' ABI status.

Six-Minute Walk. The 6-minute walk measures walking endurance and correlates with physical activity levels in patients with PAD.²⁶ Six-minute walk performance predicts mortality in patients with heart failure and oxygen consumption in patients with pulmonary disease.^27,28 Corridor walking, as performed during the 6-minute walk, is more familiar and acceptable to older patients than treadmill walking, which can be associated with balance problems and anxiety.^29-32 Following a standardized protocol,^33,34 participants walked up and down a 100-ft hallway for 6 minutes after instructions to cover as much distance as possible.

Summary Performance Score. The summary performance score is a global measure of leg functioning that predicts mobility loss, nursing home placement, and mortality among community dwelling elderly individuals.^35,36 A score (scale, 0-4) was assigned for performance on time to rise 5 times from a seated position, standing balance, and 4-meter walking velocity. Individuals received a score of 0 for each task they were unable to complete. One to 4 scores for each task were assigned based on quartiles of performance for more than 6000 participants in the Established Populations for the Epidemiologic Study of the Elderly.^35,36 Scores were summed to obtain the summary performance score, ranging from 0 to 12.

Repeated Chair Rises. This test measures leg strength and balance.^35,36 Participants sat in a straight-backed chair with their arms folded across their chest and stood 5 times consecutively as quickly as possible. The time to complete 5 chair rises was measured.

Standing Balance. PAD is associated with pathology in lower-extremity nerves^37-40 and impaired standing balance.² Participants were asked to hold 3 increasingly difficult standing positions for 10 seconds each: standing with feet together side-by-side and parallel (side-by-side stand), standing with feet parallel with the toes of one foot adjacent to and touching the heel of the opposite foot (semi-tandem stand), and standing with one foot directly in front of the other (tandem stand).^35,36

Four-Meter Walking Velocity. Slower walking speed is associated with increased risks of mobility disability, loss of the ability to perform activities of daily living, and becoming homebound.^35,36,41 Walking velocity was measured with a 4-m walk performed at "usual" and "fastest" pace. Each walk was performed twice. The faster walk in each pair was used in analyses.^35,36

Other Measures. Height and weight were measured at each visit. Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters. Cigarette smoking was assessed by patient report.

Individuals for whom data collection forms indicated that the participant was unable to complete functional measures at follow-up due to wheelchair confinement, exhaustion, or other significant symptom were classified as too disabled to complete functional measures. The principal investigator (M.M.M.) made these decisions based on the data collection forms, blinded to all other participant characteristics. When no information was provided for the reason a participant refused to complete functional tests, those who met at least 2 of the following criteria were considered too disabled to walk: (1) the participant reported walking fewer than 5 blocks during the previous week; (2) the score for repeated chair rises equaled 0 or 1; and (3) the score for the standing balance test equaled 0 or 1. The criteria were defined prior to data analyses. Individuals who refused functional testing at follow-up and met 2 of these criteria were assigned the minimum value for each test not completed. The minimum value for each test was equivalent to the poorest performance among those who completed testing at the corresponding visit. We also examined the sensitivity of results by considering 2 other methods of handling missing functional assessments for participants who returned for testing but did not perform functional measures. In one method, a score of 0 was assigned for the missing data; in the other, the fifth percentile score among functional assessment completers was assigned. Results for all 3 methods were similar. Results incorporating the minimum score for handling missing functional assessments are reported herein.

Baseline characteristics between participants with and without PAD were compared using general linear models for continuous variables and χ² tests for categorical variables. In comparing change in functioning (eg, 6-minute walk distance) across different patient groups, a longitudinal or repeated-measures analysis of covariance (ANCOVA) was carried out using generalized estimating equations.⁴² Dependent variables for each analysis were the successive annual differences in each functional measure. For example, for the 6-minute walk, the dependent variable was defined as the successive differences in 6-minute walk distances (ie, the difference in distance from baseline to the first follow-up visit and the difference in distance from the first to the second annual follow-up visit). A repeated-measures ANCOVA adjusting for baseline covariates (sex, age, and race) and a time-dependent covariate representing functional performance at the immediately preceding visit were carried out on these successive differences. Analyses were repeated adjusting additionally for baseline comorbid diseases and for time-dependent covariates (BMI and pack-years of smoking). For analyses that excluded participants without PAD, ABI was also an independent variable.

Handling Missing Data. Under this initial generalized estimating equations–type analysis, statistically valid inference is guaranteed provided missing data caused by patient dropout is unrelated to observed or unobserved data (ie, any missing data are missing completely at random). As a safeguard against violations to this assumption that missing data are missing completely at random, we repeated the fully adjusted comparisons using a repeated-measures pattern-mixture ANCOVA model.^43,44 In this model, patients may be classified into possible patterns of missing data. Because data were analyzed using successive differences, there were only 2 possible patterns of missing differences, since patients who miss the first follow-up visit cannot be included in analyses even if they attended the second follow-up visit. Thus, one pattern consists of all patients with data at baseline and at both the first and the second follow-up visits. A second pattern consists of patients who completed the baseline visit and the first follow-up visit but missed the second follow-up visit. The different patterns of missing data were included as binary indicator covariates (centered about their means). By including patterns of missing data in analyses as centered covariates and averaging over these patterns using adjusted least-squares means, one can obtain an unbiased estimate of the marginal means, adjusting for covariates.⁴⁴ To determine the validity of our findings, analyses were repeated among all participants with baseline and year 2 data, even if year 1 data were missing. In these additional analyses, values for missing data from the first follow-up visit were imputed by averaging performance on functional assessments at baseline and at the second follow-up visit.

Among participants able to walk for 6 minutes without stopping at baseline, multiple logistic regression analyses were used to model the odds for becoming unable to walk for 6 minutes continuously at follow-up across baseline ABI categories (<0.50, 0.50 to <0.70, 0.70 to <0.90, and 0.90 to <1.10, with 1.10-1.50 as the reference group), adjusting for age, sex, race, and comorbid diseases. Because of collinearity, these analyses did not adjust for cigarette smoking, BMI, or diabetes. Individuals who returned for follow-up but did not attempt the 6-minute walk and met criteria defined above for "disabled" were classified as unable to walk for 6 minutes continuously. Fit of the logistic regression models was assessed using Hosmer and Lemeshow statistics. All models passed the goodness-of-fit test. Analyses were repeated to assess associations between baseline leg symptom categories and becoming unable to walk for 6 minutes continuously at follow-up, adjusting for age, sex, race, BMI, smoking, and comorbid diseases. Analyses were performed using SAS version 8.2 (SAS Institute Inc, Cary, NC).

Power Analyses. Based on the sample sizes of 63 for ABI less than 0.50 and 259 for ABI 0.90 to 1.50, and assuming that the correlation between any 2 repeated measurements was 0.7, we had 80% power to detect a minimum detectable difference in annual change for the functional measures between these 2 ABI groups of 0.15 SDs based on a 2-tailed test at α = .05. For comparisons between participants with ABI 0.50 to less than 0.90 and those with ABI 0.90 to 1.50, the minimum detectable difference in annual change of the functional measures was 0.089 SDs.

Figure 1 shows reasons for nonparticipation among patients identified for the study. Of 707 eligible participants who completed baseline testing, 676 (96%) completed the first follow-up visit and were included in analyses. Among the 31 participants who did not complete the first follow-up visit, 24 died prior to returning for the first visit. The remainder died prior to their second follow-up visit. Compared with the 676 participants, the 31 who did not complete the first follow-up visit had a lower mean (SD) ABI (0.69 [0.20] vs 0.82 [0.25], P = .004), a higher prevalence of diabetes (45.2% vs 24.1%, P = .03), and poorer performance on baseline functional measures. Compared with the 623 participants who completed all 3 visits, the 53 who missed the second follow-up visit had significantly poorer performance on baseline functional measures, were older (mean [SD] age, 72.9 [8.6] years vs 70.8 [8.3] years), included a lower proportion of men (45.3% vs 56.3%), and had lower baseline ABIs (0.77 [0.23] vs 0.83 [0.25]). Only the differences in baseline functional performance were statistically significant.

Table 1 shows characteristics of the study population. Among participants with PAD, average mean (SD) baseline ABI values ranged from 0.62 (0.14) for those with intermittent claudication to 0.71 (0.11) among those with exertional pain/carry-on. Values for ABI were not significantly different across leg symptom categories. Lower ABI values were associated with higher mortality at 2-year follow-up (11.1% for ABI <0.50, 5.9% for ABI 0.50 to <0.90, and 3.1% for ABI 0.90 to 1.50; P = .01).

Table 2 shows associations between baseline ABI and functional decline. Adjusting for confounders including comorbid diseases and patterns of missing data, participants with baseline ABI less than 0.50 and those with ABI 0.50 to less than 0.90 each had significantly greater annual decline in 6-minute walk performance compared with those with baseline ABIs of 0.90 or greater.

Participants with PAD having leg pain on exertion and rest and those with asymptomatic PAD each had significantly greater annual decline in 6-minute walk performance than did participants without PAD, adjusting for patterns of missing data and confounders including comorbid disease. Participants with PAD having pain on exertion and rest had significantly greater declines in usual- and fastest-pace 4-meter walking velocity than did participants without PAD (Table 3). Results in Table 2 and Table 3 were similar when analyses were repeated and included all participants with data from the baseline and the second follow-up visits, even when data from the first visit were missing.

Among 80 participants with PAD having no exertional leg symptoms at baseline, 38 (48%) remained asymptomatic at follow-up and the remainder developed exertional leg symptoms at the first or second follow-up visit. Participants with asymptomatic PAD who developed exertional leg symptoms had greater mean annual functional decline than those who remained symptomatic (–136 vs –42.9 ft for the 6-minute walk, P = .12; –0.02 vs –0.01 m/sec for usual-pace 4-meter walk, P = .78; −0.06 vs −0.04 m/sec for fastest-pace 4-meter walk, P = .33; –0.54 vs –0.36 for the summary performance score, P = .17).

Table 3 analyses were repeated among participants with PAD only (data not shown). In these analyses, the pain/carry on group served as the reference, because previous cross-sectional study shows that these patients with PAD have better functioning than other leg symptom groups.¹ In fully adjusted analyses, PAD participants with pain on exertion and rest had significantly greater decline on all outcomes than did the reference group. Respective mean annual declines were −121.0 vs −20.9 ft for the 6-minute walk (P = .03), −0.06 vs −0.02 m/sec for usual-pace walking velocity (P = .04); −0.07 vs −0.02 m/sec for fastest-pace walking velocity (P = .049); and −0.82 vs −0.24 for the summary performance score (P = .04).

Figure 2 shows associations between baseline ABI levels and becoming unable to walk for 6 minutes continuously at follow-up among the 470 participants who walked continuously for 6 minutes at baseline. Adjusting for confounders, participants with baseline ABIs of less than 0.50, 0.50 to less than 0.70, and 0.70 to less than 0.90 were each significantly more likely to become unable to walk continuously for 6 minutes, compared with participants with ABIs of 1.10 to 1.50 (Figure 2).

Among participants with PAD who walked continuously for 6 minutes at baseline, those with pain on exertion and rest, atypical exertional leg pain/stop, intermittent claudication, and those who were asymptomatic at baseline were significantly more likely to become unable to walk for 6 minutes continuously than were participants without PAD at baseline, adjusting for confounders (Figure 3).

Among 676 men and women age 55 years and older, participants with low ABI levels at baseline had significantly greater decline in walking endurance at 2-year follow-up, compared with those with normal baseline ABI levels. Participants with ABIs less than 0.50 at baseline had a nearly 13-fold increased risk of becoming unable to walk for 6 minutes continuously 2 years later, relative to participants with ABIs of 1.10 to 1.50. These findings were independent of confounders, including comorbid diseases and patterns of missing data.

Baseline leg symptoms among participants with PAD also predicted rates of functional decline. Participants with PAD having leg pain on exertion and rest experienced greater declines in walking endurance and walking speed than did individuals without PAD. Participants with asymptomatic PAD had significantly greater declines in 6-minute walk performance than did participants without PAD. Within the group with asymptomatic PAD, greater functional decline was observed among participants who developed exertional leg symptoms at follow-up, compared with those who remained asymptomatic. However, these differences were not statistically significant. Further study is needed to determine mechanisms of functional decline in patients with asymptomatic PAD.

Among participants with PAD, leg pain on exertion and rest was associated with significantly greater decline on all functional outcomes compared with the leg pain/carry-on group, adjusting for confounders including the ABI. Leg pain on exertion and rest was defined as exertional leg symptoms that sometimes begin at rest. Thus, this leg symptom group was not synonymous with critical limb ischemia.

To our knowledge, no prior studies have prospectively assessed relationships between the ABI, leg symptoms, and change in objectively measured functioning in a large cohort of men and women. Findings reported herein challenge standard thinking regarding the natural history of leg functioning in patients with PAD.^4-6 In previous studies, most patients with intermittent claudication reported improvement or stabilization in leg symptoms over 5 years of follow-up, implying a benign natural history of lower-extremity functioning in those with PAD.^4-6 However, stabilization or improvement in claudication symptoms does not necessarily indicate stabilization or improvement in lower-extremity performance. Claudication symptoms may lessen because of reductions in levels of physical activity. Our data suggest that previously described lack of worsening in claudication symptoms over time may be more related to declining functional performance than to lack of PAD progression.

Based on our findings, clinicians should consider patients with PAD at increased risk of functional decline compared with those without PAD. An ABI less than 0.50, leg pain on exertion and rest, and asymptomatic PAD are all associated with greater functional decline. Our findings regarding PAD and functional decline are particularly important given the high prevalence of undiagnosed and asymptomatic PAD.^45,46 Among men and women aged 55 years and older in a general medicine practice, 34 of 239 patients screened with the ABI (14%) had previously undiagnosed PAD.⁴⁶ Of those with previously undiagnosed PAD, 19 (56%) had no exertional leg symptoms. In the PARTNERS study of 6417 men and women in general medical practices who underwent screening with the ABI, 821 patients (11.8%) had newly diagnosed PAD.⁴⁵ Of these, 342 (41.6%) were asymptomatic. Our findings suggest that patients with asymptomatic PAD who develop leg symptoms are particularly likely to have undergone functional decline and that if one waits until a patient becomes symptomatic to screen for PAD, additional functional decline—which may or may not be reversible—will occur prior to the detection of PAD. Further study is needed to determine whether interventions can prevent functional decline in patients with asymptomatic PAD prior to the onset of leg symptoms. Although an ABI less than 0.90 is highly sensitive and specific for the presence of PAD,⁴⁷ the hand-held Doppler may not precisely detect ankle systolic pressures less than 30 mm Hg. It is also important to note that approximately 5% of patients with PAD will have an ABI greater than 0.90 due to calcification of lower-extremity arteries, resulting in falsely elevated lower-extremity pressures.⁴⁸

Our study has some weaknesses. First, specific explanations were not available for all participants who returned for follow-up testing but did not perform the functional measures. Our a priori classification scheme identified individuals likely to have been too disabled to perform functional testing at follow-up. Second, some participants did not complete all follow-up visits. Our statistical methods adjusted for missing data, which is expected to reduce the influence of missing data on our findings. Third, our data may not be generalizable to individuals who declined participation. Fourth, the group of participants with PAD having atypical exertional leg pain/carry-on was relatively small. We lacked statistical power to demonstrate significant differences in some outcomes between these participants with PAD vs those without PAD. And fifth, our data do not allow us to determine the degree to which comorbid diseases contributed to the nature of leg symptoms or degree of functional decline in participants with PAD. However, our findings regarding baseline ABI levels, leg symptoms, and functional decline in those with PAD were independent of comorbid diseases.

Reasons for the significant decline in 6-minute walk performance observed in the group with asymptomatic PAD, but not in PAD groups with intermittent claudication or in the atypical leg pain/stop group, cannot be discerned from data presented here. However, some patients with asymptomatic PAD may restrict their physical activity to prevent exertional leg symptoms.¹ If patients with asymptomatic PAD restrict their activity to a greater degree than other patients with PAD, this phenomenon may contribute to the greater declines in 6-minute walk performance observed in the asymptomatic group compared with other PAD symptom groups. Further study is needed.

In conclusion, the presence and severity of PAD are associated with significant decline in walking endurance over 2-year follow-up compared with individuals without PAD. ABI values and leg symptoms can be used to identify patients with PAD who are at highest risk of functional decline. Our findings underscore the importance of using the ABI to identify persons with PAD, since PAD is frequently undiagnosed or asymptomatic. Further study is necessary to develop treatments to prevent functional decline in patients with PAD who do not have classic intermittent claudication.