Current issues of ACP Journal Club are published in Annals of Internal Medicine


Low ankle brachial pressure index was associated with increased risk for stroke and death

ACP J Club. 1997 May-Jun;126:79. doi:10.7326/ACPJC-1997-126-3-079

Source Citation

Leng GC, Fowkes FG, Lee AJ, et al. Use of ankle brachial pressure index to predict cardiovascular events and death: a cohort study. BMJ. 1996 Dec 7;313:1440-4.



To determine whether an association exists between the ankle brachial pressure index and the risk for cardiovascular events and death.


5-year cohort analytic study of participants in the Edinburgh Artery Study.


11 general practices in Edinburgh, Scotland, UK.


1592 men and women aged 55 to 74 years randomly selected from age-sex registers.

Assessment of risk factors

At the baseline assessment, a questionnaire was used to collect data on smoking history, history of diabetes, and angina. A clinical examination included blood pressure assessment, measurement of ankle systolic pressures, 12-lead electrocardiography, and a laboratory assessment of lipid levels and response to a glucose load.

Main outcome measures

Incidence of fatal and nonfatal cardiovascular events and all-cause mortality.

Main results

At baseline, 90 participants (6%) had an ankle brachial pressure index ≤ 0.7, 288 (18%) had an index ≤ 0.9, and 566 (36%) had an index ≤ 1.0. During the 5-year follow-up period, 144 participants (9%) had a myocardial infarction (MI); 55 of these particpants died. A stroke occurred in 50 participants (3%), and 22 strokes were fatal. Of the 203 deaths, 89 (44%) were caused by cardiovascular factors. After adjustment for age, sex, coronary disease, and diabetes, participants with a baseline ankle brachial pressure index ≤ 0.9 had an increased risk for stroke (relative risk [RR] 1.98, 95% CI 1.05 to 3.77), cardiovascular death (RR 1.85, CI 1.15 to 2.97), and all-cause mortality (RR 1.58, CI 1.14 to 2.18). The likelihood ratios for fatal and nonfatal cardiovascular events at 5 years were 1.76 for indexes ≤ 0.9 and 3.07 for indexes ≤ 0.7. The ability to predict subsequent events was enhanced when the ankle brachial pressure index was combined with other risk factors. For example, the positive predictive value for a future cardiovascular event in a participant who smoked, had hypertension, and had a normal cholesterol level was 25%, but this value increased to 44% in participants who also had a low ankle brachial pressure index (≤ 0.9) and decreased to 16% in those who had a higher index.


A low ankle brachial pressure index was associated with an increased risk for stroke, cardiovascular death, and all-cause mortality.

Source of funding: British Heart Foundation.

For article reprint: Professor F.G. Fowkes, Wolfson Unit for Prevention of Peripheral Vascular Diseases, Department of Public Health Sciences, University of Edinburgh, Edinburgh EH8 9AG, Scotland, UK. FAX 44-131-650-6904.


In essence, a low ankle brachial pressure index equals peripheral artery disease. It is not surprising that a low pressure index is associated with a high risk for other manifestations of cardiovascular disease, such as MI and stroke. The strength of this study by Leng and colleagues is that it puts these associations into a clinically meaningful context. The important question is this: In routine clinical practice, is there added value in using measurement of ankle brachial pressure index together with conventional predictors of MI and stroke?

Leng and colleagues have shown that in patients with none or only 1 of the risk factors for a future cardiovascular event, the index provides additional predictive information. Severe peripheral artery disease with gangrene mainly affects patients who smoke or have diabetes (and particularly patients with both characteristics). It would have been useful to know the marginal value of the pressure index when diabetes is added to the other 3 conventional risk indicators.

The virtue of the ankle brachial pressure index is that it is simple, quick, and inexpensive and can easily be introduced into routine clinical practice. The problem with the pressure index, however, is that its sensitivity as a predictor of cardiovascular events is low (30% in this study). On the basis of the figures Leng and colleagues provide, eighteen 55- to 74-year-old Scottish men and women would have a pressure index ≤ 0.9 if 100 were screened. Of the 18, 3 would have a cardiovascular event (MI, stroke, or any vascular death) in the next 5 years. According to the data provided by Leng and colleagues, 2 of the 3 patients with an event would already have been identified by a tobacco history, blood pressure, and serum lipid levels. During the 5 years, nearly 3 times as many cardiovascular events (8 events) would occur among the 82 persons who had a normal pressure index.

Arithmetic exercises such as these show the problems involved in screening and predictions; the calculations would give similarly modest results if the added value of other risk indicators was assessed. The important message is that no single risk indicator for cardiovascular events is powerful enough on its own. The ankle brachial pressure index is one of the many fragments that, when put together with other bits of clinical information, helps to provide a more detailed clinical picture of the patient at risk for MI and stroke.

Kjell Asplund, MD, PhD
University HospitalUmeå, Sweden