Prevalence of peripheral artery disease varies significantly depending upon the method of calculating ankle brachial index

Abstract

Objective

Peripheral artery disease (PAD) identifies individuals at high risk for future cardiovascular disease (CVD) warranting aggressive risk reduction therapies. PAD can be diagnosed noninvasively by calculating the ankle brachial index (ABI), a ratio of ankle and arm blood pressures. We examined the existing various methods of calculating ABI and the resulting estimates of PAD prevalence.

Methods

We analyzed data from three National Health and Nutrition Examination Surveys. PAD prevalence using three different methods of calculating ABI was determined in 5376 participants, aged ≥ 40 years without prior history of CVD. ABI was defined as an ankle brachial index of less than 0.9. Statistical analysis was performed using SPSS V15.0.

Results

PAD prevalence among asymptomatic adults without CVD increased significantly during the 6-year time period (1999-2004), regardless of the method used for determining ABI. However, across the National Health and Nutrition Examination Survey assessments, ABI method significantly affected calculated PAD prevalence. Differences in calculated PAD prevalence correspond to approximately 2.2 million persons who would be reclassified as having or not having PAD.

Conclusion

The calculated prevalence of asymptomatic PAD varies significantly by the ABI method used. Further study is required to determine the most accurate method of performing ABI.

Keywords

ankle brachial index cardiovascular disease diagnostic test peripheral artery disease

Introduction

Atherosclerosis, a systemic vascular disease, involves multiple vascular territories and leads to symptomatic coronary heart disease (CHD), cardiovascular disease (CVD), and peripheral artery disease (PAD). PAD not only causes symptomatic claudication, limb pain, and amputation, but it also serves as a marker of subclinical coronary artery disease and is associated with an elevated risk of cardiovascular morbidity and mortality [1–5]. Identifying asymptomatic persons with PAD is clinically important, because the presence of PAD identifies a person as being at high risk for suffering a future cardiovascular disease event. Such a person will require more aggressive risk factor modification and will have lower LDL treatment goals compared with an asymptomatic person without PAD.

PAD can be diagnosed noninvasively by calculating the ankle brachial index (ABI), a ratio of the ankle and arm blood pressures. By convention, PAD is present when the ABI is less than 0.90 [6]. Estimates of PAD prevalence in the United States based on an ABI less than 0.90 varies substantially ranging from 3 to 30% in US adult populations [7–10]. There are at least three methods for computing the ABI, and the current American Heart Association/American College of Cardiology guidelines on PAD describe only one particular technique [6]. The National Health and Nutrition Examination Survey (NHANES) database is one of the most commonly cited surveys from which PAD prevalence data is obtained and uses a method for performing ABI different from that described by the American Heart Association/American College of Cardiology guidelines. It is possible that the prevalence of PAD in the United States is significantly under- or over-estimated depending upon the method of calculating the ABI. The purpose of this study was to compare the effects of three methods of calculating ABI on prevalence estimates of PAD in the adult asymptomatic US population.

Materials and methods

The effect of three methods of calculating ABI on PAD prevalence was assessed using the continuous NHANES data of 6 years (1999-2004). The prevalence of PAD, as defined by an ABI less than 0.9, and associated cardiac risk factors were determined in 5376 NHANES asymptomatic participants aged 40 years and above.

NHANES is a nationally representative survey of the US civilian noninstitutionalized population conducted by the National Center for Health Statistics of the Centers for Disease Control and Prevention. NHANES participants are interviewed in their homes to obtain information on health history, health behaviors, and risk factors. Those participants then undergo a physical examination at a mobile examination center. The procedures followed to select the representative sample and conduct the interview and examinations are carefully outlined (National Center for Health Statistics Analytic Guidelines, 2006). Informed consent was obtained from all participants. This study was approved by the Lehigh Valley Hospital Institutional Review Board.

Calculation of ankle brachial index

NHANES participants aged 40 years and above were asked to participate in the lower extremity disease examination. Individuals were excluded from the exam if they had a bilateral amputation or weighed over 400 pounds (owing to equipment limitations). Some participants who were eligible for the exam might not have received the exam for various reasons (e.g.: casts, ulcers, dressings or other conditions interfering with testing, equipment failure, or participant refusal). The lower extremity disease examination was performed by trained health technicians in a specially equipped room in the mobile examination center following a prescribed protocol.

The prevalence of PAD was determined by calculating the ABI using three different methods: the NHANES method, the high ankle pressure (HAP) method, and the low ankle pressure (LAP) method. Systolic blood pressures (SBP) were obtained in the left and right brachial, dorsalis pedis, and posterior tibial arteries. To calculate the right ABI using the NHANES method, the mean SBP in the right ankle (dorsalis pedis and posterior tibialis) was divided by the mean blood pressure in both arms. The left ABI was determined in a similar manner using the left ankle pressure. To calculate the right ABI using the HAP method, the higher of the SBPs of the two right ankle arteries were divided by the higher of the two brachial SBPs. The left ABI using the HAP method was performed in a similar manner using the left dorsalis pedis and posterior tibial pressures. In the LAP method, the ABI was calculated by dividing the LAP between the two ankle arteries of a particular side by the higher of the two brachial blood pressures.

Clinical characteristics and definitions

Individuals were classified as asymptomatic if they answered ‘no’ to all of the following questions in the NHANES examination questionnaire: Has a doctor told you that you have had an MI? Has a doctor told you that you have had a stroke? Has a doctor told you that you have coronary heart disease? Has a doctor told you that you have had angina?

Age and race/ethnicity were assessed using the NHANES demographic questionnaire. Age was categorized as 40-59, 60-69, and 70+ years. Race/ethnicity was classified as nonHispanic white, nonHispanic black, and Mexican American. Hypertension was defined as a mean SBP of ≥ 140 mmHg, mean diastolic blood pressure ≥ 90 mmHg, physician diagnosis, or medication use. Individuals were defined as diabetic if they reported ever having had a doctor tell them that they had diabetes mellitus or medication use. Hypercholesterolemia was defined as a total cholesterol level of 240 mg/dl (6.216 mmol/l), physician diagnosis, or medication use. Height and weight measures were used to compute body mass index (BMI, kg/m²). BMI was then categorized into less than 25.0, 25.0-29.9, and ≥ 30.0 kg/m².

Kidney filtration function, glomerular filtration rate (GFR), was calculated using the abbreviated Modification of Diet in Renal Disease Study formula based on serum creatinine, age, sex, and race. A GFR greater than 90 (ml/min 1.73/m²) was defined as normal, mildly decreased kidney function was defined as 60-90 (ml/min 1.73/m²), and a low kidney function was defined as a GFR less than 60 (ml/min 1.73/m²). Smoking status was determined using answers to the questions, ‘Have you smoked at least 100 cigarettes in your life?’ and ‘Do you now smoke?’

Statistical analysis

The prevalence of PAD was determined for each ABI calculation method. SPSS version 15.0 statistical software (SPSS Inc., Chicago, Illinois, USA) was used for all analyses. All prevalence estimates were weighed; the sample weights account for the unequal probabilities of selection from the complex NHANES sampling and the oversampling of selected population subgroups. Prevalence rates among the three NHANES surveys were compared using Pearson's χ² analyses. Trend analysis was assessed using linear regression methods. Significance was set at a less than 0.05.

Results

The prevalence of PAD among asymptomatic adults as determined by the three different methods of calculating ABI is presented in Table 1. During the 6-year time period (1999-2004), the prevalence of PAD increased significantly regardless of the method used to determine ABI. Using the NHANES method, the adjusted prevalence increased from 3.7% in 1999-2000 to 4.2% in 2001-2002 and to 4.6% in 2003-2004 (P = 0.030). The adjusted prevalence of PAD as determined by the LAP and HAP methods for calculating ABI also increased over the 6-year time period (P = 0.028 and 0.029, respectively). The adjusted prevalence rates increased over the 6-year study for each method of calculating ABI, which implies that the absolute number of persons with PAD would differ significantly. Over the 6-year study period, the NHANES method identified fewer individuals as having PAD than either the LAP or HAP methods. The resulting differences in calculated PAD prevalence over the 6-year period corresponds to approximately 2.2 million persons who would be reclassified as having or not having PAD based on the ABI calculation method. Over the same time period, the absolute number of women who could be potentially classified as having or not having PAD based on either LAP, HAP, or NHANES methods would be 7.4, 6.0, or 5.5 million women, respectively.

Table 1

Prevalence of peripheral arterial disease (ankle brachial index < 0.90) in adults aged 40 years and above with no known history of CVD from NHANES by the method of ABI calculation


PAD (ABI < 0.90)	NHANES 1999-2000 (n = 73610141) (%)	NHANES 2001-2002 (n = 88 080811) (%)	NHANES 2003-2004 (n = 57 981 176) (%)	P value (trend)

LAP	4.8	5.4	5.4	0.028
HAP	4.0	4.3	4.4	0.029
NHANES	3.7	4.2	4.6	0.030

NHANES: ABI of right side: mean SBP (mmHg) of the right ankle/mean of the two brachial SBP (mmHg). ABI of left side: mean SBP (mmHg) of the left ankle/mean of the two brachial SBP (mmHg). HAP: ABI of right side: higher of the right ankle SBP (mmHg)/higher of the two brachial SBP (mmHg). ABI of left side: higher of the left ankle SBP (mmHg)/higher of the two brachial SBP (mmHg). LAP: ABI of right side: lower of the right ankle SBP (mmHg)/higher of the two brachial SBP (mmHg). ABI of left side: lower of the left ankle SBP (mmHg)/higher of the two brachial SBP (mmHg). ABI, ankle brachial index; CVD, cardiovascular disease; HAP, high ankle pressure; LAP, low ankle pressure; NHANES, National Health and Nutrition Examination Surveys; PAD, peripheral artery disease; SBP, systolic blood pressure.

Table 2

Prevalence of peripheral arterial disease (ABI < 0.90) in the US adults by the method of ABI calculation


	NHANES 1999-2000 (%)	NHANES 2001-2002 (%)	NHANES 2003-2004 (%)	P value

Female
LAP	5.5	7.0	6.9	0.093
HAP	4.5	5.6	5.8	0.072
NHANES	4.1	5.3	6.3	0.109
P value	0.689	0.379	0.864
Male
LAP	3.9	3.7	3.9	0.407
HAP	3.4	3.0	2.7	0.501
NHANES	3.3	3.1	2.8	0.460
P value	0.929	0.864	0.868
Race/ethnicity
NH White
LAP	4.7	5.4	4.8	0.795
HAP	4.0	4.4	3.7	0.903
NHANES	3.5	4.4	4.1	0.750
P value	0.841	0.675	0.796
NH Black
LAP	7.3	8.6	11.1	0.370
HAP	6.7	6.9	10.3	0.344
NHANES	7.5	6.4	10.8	0.318
P value	0.861	0.775	0.966
MA/other Hispanic
LAP	4.5	3.4	6.1	0.039
HAP	2.6	2.5	4.8	0.024
NHANES	3.2	2.3	3.6	0.081
P value	0.665	0.730	0.843

ABI, ankle brachial index; HAP, high ankle pressure; LAP, low ankle pressure; MA, Mexican American; NHANES, National Health and Nutrition Examination Surveys; NH White, New Hampshire White; Other Hispanic, other Hispanic as identified by the NHANES survey.

Table 3

Prevalence of peripheral arterial disease (ABI < 0.90) in selected risk factors by the method of ABI calculation


	NHANES 1999-2000 (%)	NHANES 2001-2002 (%)	NHANES 2003-2004 (%)	P value

Diabetes mellitus
LAP	8.5	9.9	7.3	0.627
HAP	7.8	7.8	6.6	0.510
NHANES	8.7	8.3	7.6	0.388
P value	0.800	0.894	0.914
Hypertension
LAP	7.0	11.1	7.5	0.250
HAP	6.0	9.0	6.3	0.410
NHANES	6.6	8.8	6.7	0.785
P value	0.751	0.759	0.869
Hypercholesterolemia
LAP	5.3	5.9	5.8	0.269
HAP	4.4	4.7	4.5	0.405
NHANES	4.0	4.5	4.6	0.487
P value	0.821	0.560	0.739
Smoking status (current)
LAP	7.0	7.1	5.6	0.900
HAP	5.9	5.7	5.3	0.691
NHANES	5.6	5.8	5.5	0.610
P value	0.756	0.700	0.990
BMI (< 25 kg/m²)
LAP	4.9	5.2	5.6	0.724
HAP	4.9	4.5	5.1	0.773
NHANES	4.6	4.2	4.6	0.543
P value	0.969	0.827	0.927
BMI (25-30kg/m²)
LAP	3.8	4.0	5.2	0.140
HAP	3.5	2.6	3.8	0.141
NHANES	3.7	2.4	3.8	0.148
P value	0.917	0.550	0.880
BMI (> 30kg/m²)
LAP	5.8	6.6	5.6	0.350
HAP	3.6	5.5	4.5	0.099
NHANES	2.7	5.7	5.7	0.035
P value	0.307	0.682	0.836
GFR >90 (ml/min 1.73/m²)
LAP	3.8	3.4	4.3	0.147
HAP	3.1	2.4	3.7	0.104
NHANES	2.5	2.4	3.1	0.147
P value	0.645	0.597	0.826
60 ≤ GFR ≤ 90 (ml/min 1.73/m²)
LAP	6.3	5.7	5.0	0.504
HAP	5.9	4.3	4.0	0.182
NHANES	6.2	4.3	4.7	0.099
P value	0.898	0.530	0.833
GFR < 60 (ml/min 1.73/m²)
LAP	20.4	16.1	12.0	0.251
HAP	19.4	16.1	8.2	0.195
NHANES	21.7	15.5	9.9	0.325
P value	0.922	0.998	0.818

ABI, ankle brachial index; GFR, glomerular filtration rate; HAP, high ankle pressure; LAP, low ankle pressure; NHANES, National Health and Nutrition Examination Surveys; NH White, New Hampshire White.

The prevalence of PAD stratified by sex, race/ethnicity, and by method of ABI calculation is presented in Table 2. During the6-yeartimeperiod(1999-2004), the prevalence of asymptomatic PAD increased significantly in women, but decreased in men, regardless of the method used to determine ABI.

The prevalence of PAD stratified by the type of associated cardiovascular risk factors and by the method of ABI calculation is presented in Table 3. The prevalence of most cardiovascular risk factors in persons with asymptomatic PAD decreased over the 6-year study. In participants with a BMI greater than 30 kg/m², however, the prevalence of PAD increased. The magnitude of the increase varied according to the method used for calculating the ABI.

Discussion

The major finding of this study is that the method used for calculating ABI substantially influences estimates of PAD prevalence among asymptomatic adults without CVD. Although the prevalence of PAD among asymptomatic adults without CVD increased significantly during the 6-year time period (1999-2004) regardless of the method used to determine ABI, the NHANES method consistently identified fewer individuals with PAD than either the LAP or HAP methods. As the presence of asymptomatic PAD identifies an individual as being at significantly higher risk of suffering a myocardial infarction or stroke compared with an individual without PAD, knowing accurate prevalence estimates for asymptomatic PAD has significant public health and screening implications.

Lower extremity PAD is a common syndrome affecting a large part of the adult population worldwide [8, 11, 12]. Prevalence estimates can vary depending upon the specific population studied and whether or not symptomatic or asymptomatic persons are included [7, 8, 13, 14]. Variance in estimates of PAD prevalence may also be because of different methods used for calculating the ABI. This study is the first that compares the effects of three methods of calculating ABI on the prevalence of PAD. Other researchers have reported the effects of different methods of performing ABI on the sensitivity and diagnostic accuracy of identifying patients with PAD. Schroder et al. [15] evaluated the diagnostic accuracy of LAP method for detecting ABI in 216 patients undergoing arterial duplex ultrasonography. In their study, the LAP method of calculating ABI was superior to that of the HAP method, with a sensitivity and specificity of 0.89 and 0.93, respectively, for the LAP method versus 0.68 and 0.99 for the HAP method [11]. Niazi et al. [16] reported similar findings of the LAP ABI with higher sensitivity and diagnostic accuracy for identifying PAD compared with the HAP method in 107 patients undergoing digital subtraction angiography.

Results from our study show that the method of calculating ABI substantially influenced the prevalence of PAD in asymptomatic persons. As patients were not followed up over time, our study was unable to define the prognostic implications related to different methods of calculating the ABI. Moreover, our study did not utilize angiography to assess the diagnostic accuracy of a specific method of measuring ABI.

Conclusion

This analysis from the NHANES database shows that the method used for calculating ABI substantially influences estimates of PAD prevalence among asymptomatic adults without CVD. As the presence of asymptomatic PAD identifies an individual as being at significantly higher risk of suffering a myocardial infarction or stroke compared with an individual without PAD, accurate prevalence estimates for asymptomatic PAD would have significant public health and screening implications. These findings highlight the need for a consistent and unified method for calculating ABI and that further study is required to determine the most accurate method of performing ABI.

References

1 Diehm

Lange

Darius

Pittrow

Von Stritzky

Tepohl

. Association of low ankle brachial index with high mortality in primary care. Eur Heart J 2006; 27:1743–1749.

Wild

Byrne

Smith

Lee

Fowkes

FG.

Low ankle-brachial pressure index predicts increased risk of cardiovascular disease independent of the metabolic syndrome and conventional cardiovascular risk factors in the Edinburgh Artery Study. Diabetes Care 2006; 29:637–642.

3 Li

Luo

Yang

Zheng

Hasimu

. Risk factors of peripheral arterial disease and relationship between low ankle-brachial index and mortality from all-cause and cardiovascular disease in Chinese patients with type 2 diabetes. Circ J 2007; 71:377–381.

Murabito

Evans

Larson

Nieto

Levy

Wilson

PWF.

The ankle-brachial index in the elderly and risk of stroke, coronary disease, and death; the Framingham Study. Arch Intern Med 2003; 163:1939–1942.

5 Zheng

Sharrett

Chambless

Rosamond

Nieto

Sheps

. Associations of ankle-brachial index with clinical coronary heart disease, stroke and preclinical carotid and popliteal atherosclerosis: the Atherosclerosis Risk in Communities (ARIC) Study. Atherosclerosis 1997; 131:115–125.

6 Hirsch

Haskal

Hertzer

Bakal

Creager

Halperin

. ACC/AHA 2005 practice guidelines for the management of patients with peripheral arterial disease (lower extremity, renal, mesenteric, and abdominal aortic). J Am Coll Cardiol 2006; 47:1–192.

Ostchega

Paulose-Ram

Dillon

Hughes

JP.

Prevalence of peripheral arterial disease and risk factors in persons aged 60 and older: data from the National Health and Nutrition Examination Survey 1999–2004. J Am Geriatrics Soc 2007; 55:583–589.

Selvin

Erlinger

TP.

Prevalence of and risk factors for peripheral arterial disease in the United States: results from the National Health and Nutrition Examination Survey, 1999–2000. Circulation 2004; 110:738–743.

9 Gregg

Sorlie

Paulose-Ram

Eberhardt

Wolz

. Prevalence of lower-extremity disease in the US adult population ≧ 40 years of age with and without diabetes: 1999–2000 National Health and Nutrition Examination Survey. Diabetes Care 2004; 27:1591–1597.

10.

10 Centers for Disease Control and Prevention (CDC). Lower extremity disease among persons age ≧ 40 years with and without diabetes United States, 1999–2002. MMWR Morb Mortal Wkly Rep 2005; 54:1158–1160.

11.

11 Guan

Guo

Liu

. Prevalence and risk factors of peripheral arterial disease in diabetic patients over 50 years old in China. Chin Med Sci J 2007; 22:83–88.

12.

12 Gimeno

Hirai

Harima

Kikuchi

Simony

de Barros

Jr . Fat and fiber consumption are associated with peripheral arterial disease in a cross-sectional study of a Japanese-Brazilian population. Circ J 2008; 72:44–50.

13.

De Vinuesa

Ortega

Martinez

Goicoechea

Campdera

Luno

Subclinical peripheral arterial disease in patients with chronic kidney disease: prevalence and related risk factors. Kidney Int Suppl 2005; 93:S44–S47.

14.

14 Periard

Cavassini

Taffe

Chevalley

Senn

Chapuis-Taillard

. High prevalence of peripheral artery disease in HIV infected persons. Clin Infect Dis 2008; 46:761–767.

15.

15 Schroder

Diehm

Kareem

Ames

Pira

Zwettler

. A modified calculation of ankle-brachial pressure index is far more sensitive in the detection of peripheral arterial disease. J Vasc Surg 2006; 44:531–536.

16.

Niazi

Khan

Easley

KA.

Diagnostic utility of the two methods of ankle brachial index in the detection of peripheral arterial disease of lower extremities. Catheter Cardiovasc Interv 2006; 68:788–792.