Association Between Physical Activity and Mortality in Patients With Claudication

Andrew W. Gardner; Odessa Addison; Leslie I. Katzel; Polly S. Montgomery; Steven J. Prior; Monica C. Serra; John D. Sorkin


Med Sci Sports Exerc. 2021;53(4):732-739. 

In This Article



Approval and Informed Consent. The Institutional Review Board at the University of Maryland Baltimore and the Research and Development committee at the Baltimore Veterans Affairs Maryland Health Care System (VAMHCS) approved this study. Written informed consent was obtained from each patient for the baseline assessments as previously described.[21]

Recruitment. Patients who had PAD and claudication (Fontaine stage II and Rutherford grade I)[2] and who were not currently exercising were recruited from the Vascular Clinic at the site of the Baltimore VAMHCS and from local newspaper and radio advertisements for possible enrollment into the study.

Medical Screening through History and Physical Examination

Screening. A total of 528 patients who volunteered to participate were screened for PAD between 1994 and 2002 in the Geriatrics, Research, Education, and Clinical Center at the Baltimore VAMHCS hospital.

Inclusion and Exclusion Criteria. Eligible patients were classified as having Fontaine stage II/Rutherford grade I PAD[2] defined by the following inclusion criteria: (a) a history of claudication and (b) an ABI at rest <0.90. Patients were excluded from this study for the following conditions: (a) absence of PAD; (b) inability to obtain an ABI measure due to noncompressible vessels; (c) asymptomatic PAD (Fontaine stage I/Rutherford grade 0); (d) rest pain or tissue loss PAD (Fontaine stage III or IV/Rutherford grade II or III); (e) exercise tolerance limited by factors other than claudication (e.g., severe coronary artery disease, dyspnea, or poorly controlled blood pressure); (f) active cancer, renal disease (serum creatinine concentration greater than 1.2 mg·dL−1), or liver disease; (g) not living independently at home; and (h) any missing baseline data. A total of 386 patients were included in the study and analyses.

Baseline Measurements

Medical History and Physical Examination. Demographic information, cardiovascular risk factors, comorbid conditions, claudication history, a complete metabolic panel blood draw, and a list of current medications were obtained during a medical history and physical examination. Based on this battery of baseline assessments, patients were coded on cardiovascular risk factors and comorbid conditions according to standard definitions.[22] Medication information was used to assist in defining cardiovascular risk factors, as patients who were taking medications to treat hypertension, dyslipidemia, and diabetes were defined as having those conditions. Patients were further characterized on the presence, severity, and symptoms of PAD.

ABI. After 10 min of supine rest, the ankle and brachial systolic blood pressures were obtained according to standard guidelines.[23] The ABI was calculated as ankle systolic pressure/brachial systolic pressure.

Anthropometry. Height was recorded from a stadiometer (SECA, Germany), and body weight was recorded from a balance beam scale (Health-O-Meter Inc., Bridgeview, IL) without shoes. From these measurements, body mass index (BMI) was calculated as weight (kg)/height (m)2.

Physical Activity Status. The Johnson Space Center (JSC) physical activity scale was used to assess the activity level of the participants over the preceding month.[11,13,24,25] The JSC physical activity scale has a strong, independent relationship with maximal oxygen uptake in men and women between the ages of 20 and 79 yr,[24,25] and it is associated with peak walking time,[11] ABI,[11] and all-cause mortality in patients with PAD.[13] This 8-point Likert scale consists of the following score choices:

  • 0 = avoid physical activities whenever possible

  • 1 = light physical activities done occasionally

  • 2 = moderate physical activities done regularly for less than 1 h·wk−1

  • 3 = moderate physical activities done regularly for more than 1 h·wk−1

  • 4 = heavy physical activities done regularly for less than 30 min·wk−1

  • 5 = heavy physical activities done regularly between 30 and 60 min·wk−1

  • 6 = heavy physical activities done regularly between 1 and 3 h·wk−1

  • 7 = heavy physical activities done regularly for more than 3 h·wk−1

To better assist patients in their selection, examples of light, moderate, and heavy activities were provided at the bottom of this scale. Light activities were defined as regular walking, household chores, or comparable activities, which may cause light sweating. Moderate activities were defined as fast walking, jogging, or comparable activities, which may cause moderate sweating. Heavy activities were defined as fast jogging, running, or comparable activities, which may cause heavy sweating.

The patients were asked to select the appropriate score (0 to 7), which best described their general level of physical activity for the previous month. Patients who selected a score of 0 were placed into the physically sedentary group, and those who selected a score of 1 were placed into the light-intensity physical activity group. Because the number deaths attributable to cardiovascular mortality was modest in patients with a physical activity score of 2 (n = 11) and with scores between 3 and 7 (n = 21), patients who selected a score of 2 or above were placed into a single moderate- to vigorous-intensity activity group.

Survival Status at Follow-up

Vital status of study participants through December 2014 was determined using the National Death Index (NDI) and the U.S. Department of Veterans Affairs and the U.S. Department of Defense Suicide Data Repository (SDR). Relevant records in the NDI and SDR were obtained by a search based on a combination of name, social security number, and date of birth. Date and cause of death (encoded using ICD9 or ICD10 codes) were obtained from death certificate information provided by NDI and SDR. Cardiovascular mortality included ICD9 codes 390 through 459 or ICD10 I00 through I99 for the primary cause of death. Survival status information was obtained on all of the patients in the study.

Statistical Analyses

A Cox proportional hazards analysis was used to model the relationship between physical activity and two outcome measures, all-cause mortality and cardiovascular mortality. Both analyses were adjusted for age (yr), sex (reference female), BMI (centered at 28.0 kg·m−2), ABI, smoking (pack-years), and five dichotomous variables: hypertension, diabetes, history of a lower extremity revascularization, history of a stroke, and dyslipidemia (reference level for all five was risk factor not present). The proportional hazard assumption was checked by examination of log–log survival curves; plots of the residuals from each model were checked. Adjusted Kaplan–Meier survival curves for all-cause mortality and cardiovascular disease-free survival were created, by sex, for a 68-yr-old patient, having a BMI of 28.0 kg·m−2, an ABI of 0.67, a 39.9 pack-year history of smoking, and who had no other major risk factor for PAD, i.e., no history of hypertension, diabetes, hyperlipidemia, stroke, or lower extremity revascularization. The values used for age, BMI, and ABI are the mean values for the study patients. The value used for pack-years is the median value. All other risk factors were set to zero to eliminate the association of these risk factors in the displays.