Vascular Medicine

http://vmj.sagepub.com/ Sex differences in the prevalence of peripheral artery disease in patients undergoing coronary catheterization
Amir H Sadrzadeh Rafie, Marcia L Stefanick, Stacy T Sims, Tiffany Phan, Mamie Higgins, Andre Gabriel, Themistocles Assimes, Balasubramanian Narasimhan, Kevin T Nead, Jonathan Myers, Jeffrey Olin and John P Cooke Vasc Med 2010 15: 443 DOI: 10.1177/1358863X10388345 The online version of this article can be found at: http://vmj.sagepub.com/content/15/6/443

Published by:
http://www.sagepublications.com

On behalf of:

Society for Vascular Medicine

Additional services and information for Vascular Medicine can be found at: Email Alerts: http://vmj.sagepub.com/cgi/alerts Subscriptions: http://vmj.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://vmj.sagepub.com/content/15/6/443.refs.html

Downloaded from vmj.sagepub.com at COLUMBIA UNIV on April 4, 2011

Sex differences in the prevalence of peripheral artery disease in patients undergoing coronary catheterization
Amir H Sadrzadeh Rafie1, Marcia L Stefanick1, Stacy T Sims1, Tiffany Phan1, Mamie Higgins1, Andre Gabriel2, Themistocles Assimes1, Balasubramanian Narasimhan1, Kevin T Nead1, Jonathan Myers1, Jeffrey Olin2 and John P Cooke1

Vascular Medicine 15(6) 443450 The Author(s) 2010 Reprints and permission: sagepub. co.uk/journalsPermissions.nav DOI: 10.1177/1358863X10388345 http://vmj.sagepub.com

Abstract To determine whether there are sex differences in the prevalence of peripheral artery disease, we performed an observational study of 1014 men and 547 women, aged 40 years, referred for elective coronary angiography. Women were slightly older, more obese, had higher low-density lipoprotein cholesterol (LDL-C) levels and systolic blood pressure (BP), and were more likely to be African American. Women had higher high-density lipoprotein cholesterol (HDL-C) levels, lower diastolic BP, and were less likely to smoke or to have a history of cardiovascular disease. Women had less prevalent (62% vs 81%) and less severe coronary artery disease (CAD) (p < 0.001 for both) by coronary angiography, but more prevalent peripheral artery disease (PAD) as determined by the anklebrachial index (ABI) than men (23.6% versus 17.2%). Independent predictors of lower ABI were female sex, black race, older age, tobacco use, CAD, diabetes, and triglyceride level. In a full multivariable logistic regression model, women had a risk-adjusted odds ratio for PAD of 1.78 (95% CI 1.252.54) relative to men. Among patients referred for coronary angiography, women have less prevalent and less severe CAD, but more prevalent PAD, a sex difference that is not explained by traditional cardiovascular disease risk factors or CAD severity. Clinical Trial RegistrationURL: http://clinicaltrials.gov. Unique identifier: NCT00380185 Keywords coronary artery disease; gender; intermittent claudication; vascular disease

Introduction
Clinicians have long recognized that patients with similar cardiovascular risk factors may have strikingly different manifestations of atherosclerosis. There is individual heter ogeneity in the severity of atherosclerosis, and in its distri bution. Patients with the same risk factor profile may have subclinical disease, or present with hemodynamically sig nificant and symptomatic occlusive disease of the carotid, coronary and/or peripheral arteries. The Genetic determi nants of Peripheral Arterial Disease (GENEPAD) study, funded by the National Heart, Lung and Blood Institute, was designed to identify genetic determinants which preferen tially influence the susceptibility to hemodynamically signi ficant atherosclerosis in the peripheral arteries. Furthermore, the study examines the interaction of these determinants with accepted risk factors for atherosclerosis, and other pathophysiological mediators. The current paper describes the results of a subanalysis of the data gathered in this study, to determine the role of sex on the development of hemody namically significant lesions in the limb vessels. Sex differences in cardiovascular disease (CVD) and its comorbidities arise from a combination of biologic and

social factors, including gender biases in medical practice which influence diagnoses, clinical care, and outcomes.1 Whereas the increased ageadjusted prevalence of coronary artery disease (CAD) in men is welldocumented and widely accepted, there is less known about the prevalence of athero sclerotic occlusive arterial disease affecting the lower extrem ities (often termed peripheral artery disease or PAD). Despite the fact that PAD is a common condition affecting at least 8 million Americans,2 whether there are sex differences in the
1 2

Stanford University Medical Center, Stanford, CA, USA Mount Sinai Medical Center, New York, NY, USA

Corresponding author: John P Cooke Professor of Medicine Stanford University Medical Center 300 Pasteur Dr. Falk Cardiovascular Research Center Stanford, CA 94305-5406 USA Email: john.cooke@stanford.edu Dr Sonia Anand was the Guest Editor for this paper.

Downloaded from vmj.sagepub.com at COLUMBIA UNIV on April 4, 2011

444 prevalence of PAD is uncertain. It has been suggested that the prevalence of PAD is equal for men and women.37 By con trast, several populationbased studies have suggested that the prevalence of a low anklebrachial index (ABI) is more common in women.811 The current study was performed to determine if there is a sex difference in the prevalence of hemodynamically significant coronary and peripheral artery disease in patients undergoing coronary angiography.

Vascular Medicine 15(6) were measured by standard assays at the Stanford and Mt Sinai Hospital clinical laboratories, using an AU5400 Chemistry ImmunoAnalyzer (Olympus Inc.). Lowdensity lipoprotein cholesterol (LDL) was measured using an AU640 Chemistry ImmunoAnalyzer (Olympus Inc.).

ABI technique and coronary angiography procedures

The ABI (i.e. the ratio of ankle and arm systolic blood pres sure) was measured using previously established meth ods.9,12 In brief, prior to undergoing coronary angiogram, each participant rested in the supine position for 5 minutes, then, using a 5MHz Doppler ultrasound (Nicolet Elite 5MHz vascular model 110R Doppler; Nicolet Vascular, Golden, CO, USA), systolic pressures were measured in the posterior tibial, dorsalis pedis, and brachial arteries. Each pressure was measured twice in sequential and reverse order as listed. With an IV in one arm, only the contralateral arm was used for the brachial pressures. The ABI for each leg was calculated separately by dividing the higher of the two ankle pressures in that leg by the brachial pressure. If the ABI was less than 0.9 in either leg, the patient was con sidered to have PAD. The index leg was defined as the leg with the lower ABI. The ABI has been shown to have good sensitivity and specificity for PAD, as documented by direct comparisons to angiography.1214 Patients with a his tory of PADrelated revascularization, angioplasty or stent ing were also classified as having PAD, regardless of ABI. All coronary segments were interpreted visually by an experienced cardiologist blinded to participant details at each participating site. The left anterior descending, left circumflex (+ ramus), and right coronary arteries were the major coro nary arteries analyzed. Hemodynamically significant CAD was defined as > 60% stenosis15,16 or a history or evidence of one or more previous percutaneous transluminal coronary angioplasties (PTCAs) or grafts. Significant CAD was further expressed as one, two or threevessel or left main disease.

Methods Participant identification

The GenePAD study was approved by the Institutional Review Boards at Stanford University and Mount Sinai; reg istered at http://clinicaltrials.gov (NCT 00380185). Partici pants were recruited from patients undergoing an elective, nonemergent coronary angiogram for angina, shortness of breath, abnormal stress test, or known CAD at Stanford University and Mount Sinai Medical Centers between April 2004 and February 2008, for whom the PAD status was not known to the investigators at the time of informed consent and recruitment into the study. This crosssectional design allowed for two sharply defined phenotypes to emerge (i.e. patients with hemodynamically significant CAD alone (con trols) versus those with hemodynamically significant CAD and PAD (cases)). Furthermore, with this study design, con founding clinical covariates (e.g. traditional cardiovascular risk factors) would contribute minimally to the difference between the two phenotypes. This design would maximize our chances for discovery of novel factors that affect the dis tribution of atherosclerotic plaque. All participants provided written informed consent. Patients admitted for emergent catheterizations, or screen ing catheterizations prior to organ transplants were excluded. Additional exclusions included being younger than 40 years of age and having a history of radiation treat ment, known chronic infectious diseases such as HIV, hep atitis B or C, or a language barrier.

Statistical analysis
The primary statistical analyses focused on comparisons between men and women. Continuous variables were expressed as mean SD, and categorical variables were pre sented as number of subjects (frequencies). Baseline charac teristics of male and female participants were compared using unpaired ttests for continuous variables and chi square tests for categorical variables. Multivariate regres sion models and multivariate logistic regression techniques were developed to determine the relationships of sex, tradi tional risk factors and CAD with ABI and the diagnosis of PAD, respectively. The multivariate regression analyses in Tables 5 and 6 coded the presence or absence of PAD and ABI as the dependent variables, respectively. The independ ent variables, which are thus adjusted for in the analysis were as follows: age, female sex, BMI, systolic BP, diastolic BP, African American race, diabetes, triglyceride, HDL cholesterol, LDLcholesterol, packyears smoking and CAD. In this analysis, female sex, African American race, diabetes and CAD were coded as categorical variables. All statistical analyses were performed by SPSS software

Assessments of atherosclerotic risk factors

Information about demographic factors (e.g. age, ethnicity, medications for hypertension, hypercholesterolemia, dia betes and angina, medical and family history of CVD, life style, including smoking history (current or former) for patients with one or more packyears of smoking, and phys ical function) were obtained by detailed structured ques tionnaires administered by trained study personnel. Women were considered postmenopausal by selfreport or use of menopausal hormone therapy.

Blood analyses
Fasting blood (30 ml total) was collected from each patient through a venous or arterial femoral sheath or intravenous line in the arm, processed, then centrifuged at 3000 RPM for 20 minutes at 4C, and aliquots of EDTA plasma and serum were stored at 75C. Glucose, total and high density lipoprotein (HDL) cholesterol, and triglycerides

Downloaded from vmj.sagepub.com at COLUMBIA UNIV on April 4, 2011

Rafie AHS et al.

Table 1. Baseline characteristics Characteristic Age, years Systolic BP, mmHg Diastolic BP, mmHg Pulse rate, beats per minute Body mass indexa Glucose, mg/dl HDL-cholesterol, mg/dl LDL-cholesterol, mg/dl Triglyceride, mg/dl Smoking, no. (%) Currently Ever Cumulative pack-years Race, no. (%) White or European-American Black or African American Latina or Latino or Hispanic Asian or Asian-American Other races / ethnicities Entire cohort (n = 1561) 65.5 10.7 147 23 75 13 70 12 29 6 118 59 41 12 78 31 114 29 140 (9) 719 (46) 17 25 824 (56.1) 220 (15) 184 (12.5) 110 (7.5) 132 (9) Men (n = 1014) 65 10.5 146 22 76 12 69 12 28.6 5.6 119 62 38 11 74 29 115 80 100 (9.8) 512 (50.5) 19 27 571 (59.3) 114 (11.8) 109 (11.3) 74 (7.7) 95 (9.9) Women (n = 547) 66.4 11 149 25 73 14 73 13 29.6 6.7 117 50 46 13 85 33 112 77 40 (7.3) 207 (37.8) 12 24 253 (49.9) 106 (20.9) 75 (14.8) 36 (7.1) 37 (7.3)

445

p-value 0.01 0.009 < 0.001 < 0.001 0.004 0.5 < 0.001 < 0.001 0.5 0.3 0.01 < 0.001 0.001 < 0.001 0.06 0.7 0.1

Results are presented as mean standard deviation or number of subjects (% prevalence in parentheses). a Calculated as weight in kilograms divided by height in meters squared.

(version 15.0; Chicago, IL, USA). A twosided p-value less than 0.05 was considered statistically significant.

Results Baseline characteristics

Between April 2004 and February 2008, a total of 1561 eligi ble, consenting patients (1014 men, 547 women) underwent elective coronary angiography. Baseline patient charac teristics appear in Table 1. Women were slightly older than men (p = 0.01), with slightly higher systolic pressure, and higher HDL and LDL cholesterol levels. A higher proportion of women were African American or Hispanic (35.7% vs 23.2%, p < 0.001) and obese, defined as body mass index (BMI) 30 kg/m2 (40.5% vs 33.4% in men, p = 0.004). By contrast, a greater proportion of men had smoked in the past and cumulative packyears were higher in men. CVD history and risk factors at enrollment appear in Table 2. PAD was reported by 7% of men and 7.4% of women. A higher proportion of men (61.8%) reported prior CAD diagnoses, including myocardial infarction, coronary artery stenosis, coronary artery bypass graft, balloon angi oplasty (PTCA) or stenting, compared to women (52.8%, p = 0.001). There were no significant sex differences in selfreported hypertension, diabetes, hyperlipidemia, stroke, or abdominal aortic aneurysm.

Burden of coronary and peripheral artery diseases

Sex differences in the distribution of atherosclerotic risk factors by CAD and PAD status, alone and combined, appear

in Table 3. Patients without hemodynamically significant CAD or PAD comprised 24% of the study population. By contrast, 65% of men and 42% of women had CAD only (p < 0.001), and 16% of men and 20% of women had CAD and PAD (p = 0.06). Only 10 men (< 1%) and 20 women (3.7%) had PAD in the absence of hemodynamically signifi cant coronary disease (p < 0.001). As shown in Table 4, a significantly higher proportion of men (80.8%) were identified with CAD by coronary angi ography than women (61.6%, p < 0.001); furthermore, men were more likely to have either twovessel or left main or threevessel disease compared to women. By contrast, the average ABI was lower in women. Furthermore, the preva lence of PAD (as assessed by an ABI of 0.90 or a history of peripheral vascular surgery or endovascular interven tion) was greater in women than in men (23.6 vs 17.2%; p = 0.007). Notably, only 7% of men, and 7.4% of women, had a prior diagnosis of PAD, so that the majority of indi viduals with PAD were not diagnosed prior to our study. Of those individuals with CAD, 32% of women and 20% of men had PAD (of these, only 11% of women and 8% of men had a prior diagnosis of PAD). When an even more stringent definition of PAD was used, the sex difference was maintained. Specifically, 15.7% of women and 11% of men had an ABI of 0.80 (p = 0.01). Of those individuals with CAD, 22.2% of women and 13.4% of men had an ABI of 0.80 (p < 0.001). In a univariate logistic regression, factors predictive of PAD included older age, female sex, systolic BP, African American race, diabetes, triglyceride level, tobacco use, and CAD. As shown in Table 5, logistic regression

Downloaded from vmj.sagepub.com at COLUMBIA UNIV on April 4, 2011

446
Table 2. History of cardiovascular risk factors and diseases Characteristic Hypertensiona Diabetesa Hyperlipidemiaa Obesityb Coronary artery diseasec Heart attack Blockage of coronary arteries CABG, PTCA or stenting Stroke Carotid or subclavian atherosclerosis Carotid endarterectomy, angioplasty or stenting Peripheral artery disease (PAD) PAD-related surgery, angioplasty or stenting PAD-related amputation Abdominal aortic aneurysm (AAA) Surgical repair of AAA Total (n = 1561) 1262 (80.8) 472 (30.2) 1192 (76.3) 561 (35.9) 916 (58.9) 368 (23.6) 859 (55) 719 (46.1) 88 (5.6) 104 (6.7) 49 (3.1) 111 (7.1) 52 (3.3) 12 (0.8) 47 (3) 19 (1.2) Men (n = 1014) 813 (80.2) 292 (28.8) 793 (78.2) 339 (33.4) 627 (61.8) 257 (25.3) 596 (58.8) 503 (49.6) 55 (5.4) 69 (6.8) 37 (3.6) 71 (7) 30 (2.9) 9 (0.7) 37 (3.6) 14 (1.4)

Vascular Medicine 15(6)

Women (n = 547) 449 (82.1) 180 (32.9) 399 (72.9) 222 (40.5) 289 (52.8) 111 (20.3) 263 (48.1) 216 (39.5) 33 (6) 35 (6.4) 12 (2.2) 40 (7.4) 22 (4) 3 (0.5) 10 (1.8) 5 (0.9)

p-value 0.4 0.1 0.01 0.004 0.001 0.04 < 0.001 < 0.001 0.4 0.4 0.2 0.8 0.3 0.3 0.1 0.1

Results are presented as number of subjects (% prevalence in parentheses). a Based on the prescription of one or more medications for these conditions. b Obesity is defined as BMI 30 kg/m2. c Coronary artery disease is defined by history of myocardial infarction, and/or blockages of any coronary arteries with or without angina, coronary artery bypass graft (CABG), balloon angioplasty (PTCA) or stenting.

Table 3. Sex differences in the distribution of atherosclerotic risk factors Characteristic CAD only n = 883 Men n = 655 Proportion of study population (%) Age, years BMI, kg/m2 Systolic BP, mmHg Diastolic BP, mmHg Diabetes, no. (%) Glucose, mg/dl HDL-cholesterol, mg/dl LDL-cholesterol, mg/dl Triglyceride, mg/dl Creatinine, mg/dl Smoking-ever, no. (%) African American race, no. (%) 64.6 65 11 29 5.5 145 20 76 12 171 (26.1) 117 52 38 10 71 27 114 86 1.2 0.9 330 (50.4) 58 (8.8) Women n = 228 41.7 67 11 29 6.1 149 25 72 14 81 (35.5) 120 51 45 12 80 34 112 85 1.2 1.3 91 (39.9) 31 (13.6) p-value < 0.001 0.005 0.7 0.02 < 0.001 0.02 0.4 < 0.001 < 0.001 0.8 0.9 0.3 0.03 PAD only n = 30 Men n = 10 1 66.7 7 27.6 4.2 160 35 79 17 4 (40) 107 70 47 21 103 50 99 55 23 7 (70) 4 (40) Women n = 20 3.7 67.7 8 32.2 9 149 23 72 15 4 (20) 112 50 46 14 90 27 130 91 1.3 1.7 10 (50) 5 (25) p-value < 0.001 0.7 0.06 0.3 0.3 0.2 0.8 0.8 0.4 0.3 0.4 0.1 0.4 PAD and CAD n = 273 Men n = 164 16.2 68.7 9 28 5.5 151 25 76 12 80 (48.8) 137 104 38 12 73 29 124 74 1.7 1.9 92 (56.1) 28 (17.1) Women n = 109 19.9 71 10 29 6 154 25 76 15 54 (49.5) 132 64 44 13 84 33 130 77 1.3 1.2 50 (45.9) 31 (28.4) p-value 0.06 0.06 0.2 0.4 0.9 0.9 0.6 < 0.001 0.01 0.6 0.04 0.6 0.02

CAD, coronary artery disease; PAD, peripheral artery disease; BMI, body mass index; BP, blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein.

modeling revealed that independent predictors of PAD included older age, female sex, African American race, diabetes, triglyceride level, tobacco use, and CAD were independent predictors of PAD. BMI was inversely related to the risk of PAD.

Univariate regression revealed that age, female sex, African American race, diabetes, triglyceride level, tobacco use and CAD were inversely related to ABI, whereas BMI was positively related. As shown in Table 6, multiple regres sion modeling showed, in rank order, female sex (p < 0.001),

Downloaded from vmj.sagepub.com at COLUMBIA UNIV on April 4, 2011

Rafie AHS et al.

Table 4. Burden of coronary and peripheral artery diseases Characteristic Coronary artery diseasea One vessel disease Two vessel disease Left main or three vessel disease Index ABIb Peripheral artery diseasec Entire cohort n = 1561 1156 (74.1) 337 (21.6) 355 (22.7) 464 (29.7) 0.99 0.18 303 (19.4) Men n = 1014 819 (80.8) 207 (20.4) 258 (25.4) 354 (34.9) 1.02 0.18 174 (17.2) Women n = 547 337 (61.6) 130 (23.8) 97 (17.7) 110 (20.1) 0.95 0.19 129 (23.6)

447

p-value < 0.001 0.1 0.001 < 0.001 < 0.001 0.007

Results are presented as mean standard deviation or number of participants (% prevalence in parentheses). a Based on coronary angiography at the time of enrollment. b Anklebrachial index in the index leg. c Peripheral artery disease is defined as ABI < 0.9 or a history of arterial bypass of the lower extremities regardless of ABI measure.

Table 5. Multivariate logistic regression for diagnosis of peripheral artery disease Variable Coefficient Odds ratio 1.045 1.785 0.968 1.003 0.995 2.411 2.604 1.005 0.999 1.004 1.016 3.543 95% Confidence interval 1.0261.063 1.2512.546 0.9410.996 0.9941.011 0.9811.010 1.5813.676 1.8593.649 1.0021.008 0.9851.014 0.9981.009 1.0111.022 2.1525.832 p-value

In addition, the results of the sexspecific multivariate regression modeling in Table 6 were unchanged.

Discussion
< 0.001 0.001 0.03 0.5 0.5 < 0.001 < 0.001 < 0.001 0.9 0.2 < 0.001 < 0.001

Age Female sex BMI Systolic BP Diastolic BP African American race Diabetes Triglyceride HDL-cholesterol LDL-cholesterol Pack-years smoking Coronary artery disease

0.044 0.579 0.032 0.003 0.005 0.880 0.957 0.005 0.001 0.004 0.016 1.265

A sex difference in the distribution of occlusive arterial disease

In patients undergoing coronary angiography, we found that women were more likely to have PAD, despite less prevalent and less severe CAD. Consistent with our obser vations, a Swedish populationbased study reported a higher prevalence of severe PAD in women.8 Our study is also consistent with several prior studies of community dwelling populations that found higher prevalences of low ABI values among women as compared to men.911 It has been suggested that the sex differences may be partially accounted for by height (i.e. a greater augmentation in men of the ankle systolic pressure by pulse wave reflections is due to the greater distance from the heart to the ankle). However, in the population that we have studied, the aug mentation of the systolic pressure by pulse wave reflection would be minimized by increasing vascular stiffness with age and risk factors.17 Indeed, in each of these previous studies, after adjusting for height, there remains a signifi cant sex difference.911 Not all studies have observed a greater prevalence of PAD in women.47 The National Health and Nutrition Examination Survey (19992000) reported a prevalence of PAD (ABI < 0.9) of 4.3% among individuals aged 40 years.4 Prevalence was nearly equal in men and women overall, although there were differences by sex within specific age groups.4 The Cardiovascular Health study (a populationbased observational study in 5201 adults 65 years) observed PAD in 12.4% of subjects, with no sex dif ference.6 A Canadian study of 28,649 patients undergoing coronary angiography also reported no sex difference in the prevalence of PAD.7 However, a weakness of this study is that it relied upon medical history for the diagnosis of PAD; ABI was not measured. In the Rotterdam study, over 2000 subjects 55 years old underwent imaging or hemody namic assessment for coronary, carotid and peripheral

BMI, body mass index; BP, blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein.

black race (p < 0.001), older age (p < 0.001), tobacco use (p < 0.001), CAD (p < 0.001), diabetes (p < 0.001) and trig lyceride level (p = 0.002) as independent negative predictors of ABI in the entire cohort. After adjusting for traditional risk factors and CAD in a multivariable logistic regression model, women had a 1.78fold (95% CI 1.252.54) higher riskadjusted odds ratio for the diagnosis of PAD. Sex specific multivariate regression modeling, shown in Table 6, identified African American race, diabetes, packyears smoking and CAD as significant predictors of ABI in both men and women. In our population, a higher percentage of women were African American. Therefore, we recalculated the ABI for the total study population, and for males or females sepa rately, after excluding African Americans. When excluding African Americans, the sex difference in the ABI was main tained (i.e. 1.03 ( 0.17) vs 0.97 ( 0.18), male vs female).

Downloaded from vmj.sagepub.com at COLUMBIA UNIV on April 4, 2011

448
Table 6. Multivariate regression for ankle-brachial index Variable Entire cohort n = 1561 B coefficient Age Female sex BMI Systolic BP Diastolic BP African American race Diabetes Triglyceride HDL-cholesterol LDL-cholesterol Pack-years smoking Coronary artery disease 0.166 0.179 0.054 0.064 0.005 0.170 0.129 0.081 0.037 0.048 0.157 0.137 p-value < 0.001 < 0.001 0.04 0.04 0.9 < 0.001 < 0.001 0.002 0.2 0.07 < 0.001 < 0.001 Men n = 1014 B coefficient 0.155 0.069 0.070 0.037 0.163 0.139 0.080 0.025 0.088 0.145 0.147 p-value < 0.001 0.03 0.07 0.3 < 0.001 < 0.001 0.01 0.5 0.008 < 0.001 < 0.001

Vascular Medicine 15(6)

Women n = 547 B coefficient 0.197 0.025 0.060 0.035 0.184 0.117 0.082 0.047 0.009 0.205 0.124 p-value < 0.001 0.6 0.2 0.5 < 0.001 0.01 0.07 0.3 0.8 < 0.001 0.006

BMI, body mass index; BP, blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein.

artery disease. The prevalence of CAD (as assessed by the coronary calcium score) and in the prevalence of carotid artery disease (as assessed by duplex ultrasonography), were greater in men. By contrast, the investigators observed no sex difference in PAD prevalence, based on ABI or aor tic calcification.5 The virtue of this investigation was that it was a populationbased study. However, a quantitative assessment of coronary arterial occlusive disease was not obtained. These previous studies did not prospectively and quanti tatively assess arterial occlusive disease in both circulations. Our study adds the dimension of concomitant determination of CAD (by coronary angiography) and PAD (by ABI). These data permit us to compare the severity of disease in the coronary versus peripheral circulation. The intriguing new observation is that, with respect to atherosclerotic arte rial occlusive disease, the sex difference that exists in the coronary circulation is reversed in the peripheral arteries.

A sex difference in the recognition of PAD

We also confirm previous reports that PAD is often unrec ognized. Whereas just over 7% of men and women reported prior PAD diagnoses upon enrollment into our study, 17.2% of men and 23.6% of women were found to have PAD after measuring the ABI; thus, less than 42% of men and 32% of women with an ABI 0.90 had a prior diagnosis of PAD, despite extensive contact with cardiologists prior to their cardiac catheterization. Hirsch et al.12 reported underdiag nosis of PAD in a national cohort of adults aged 5069 years with a history of diabetes or cigarette smoking or aged 70 years or older. Of those with documented PAD in the study by Hirsch and coworkers, about 44% had not been previously diagnosed. Our reanalysis of their data by sex indicates that in those with isolated PAD, 50% of men and 59% of women had not been diagnosed. The fact that a substantial proportion of patients diagnosed with PAD by ABI are asymptomatic (i.e. they report no exertional leg

symptoms) may contribute to the underdiagnosis of PAD.18,19 Evidence that women have greater missed diag noses and more asymptomatic disease20,21 may have con tributed to the inconsistencies in the literature regarding sex differences in the prevalence of PAD. Underreporting of PAD is likely a common flaw in the existing literature, as well as medical practice. Indeed, in our own study, only 7% of men and 7.4% of women were reported to have PAD by patient or medical history. However, when the ABI was measured, 23.6% of women and 17.2% of men had an ABI 0.90. Intriguingly, despite an equal or even greater prevalence of PAD in women, men are more than twice as likely to be selected for lower extremity revascularizations, even after adjusting for limb salvage, age, ABI, comorbidity and smoking.22 In this regard, ageadjusted death rates for the ICD9 term disease of the arteries are higher in men than women.2 This term is inclusive of a number of vascular dis eases including aortic aneurysm and arterial embolism, as well as PAD. The greater death rate in men may be related to the greater likelihood that they have concomitant coro nary and/or carotid disease. Nevertheless, women with PAD have been reported to have greater walking impair ment from leg symptoms and greater leg pain on exertion and rest, and critical limb ischemia,8,21 compared to men with PAD. In a pilot study of 71 men and 26 women with PAD of similar severity, women reported decreased physi cal functioning, more bodily pain and greater mood distur bance compared to men. It was concluded that claudication and PAD had a great impact on women.23

The importance of recognizing PAD

Previous studies have demonstrated that PAD is a relative contraindication to the femoral approach for coronary angi ography.24 Patients with PAD are more likely to have com plications in the cardiac catheterization laboratory.25,26 Prior knowledge of PAD may influence the procedural approach

Downloaded from vmj.sagepub.com at COLUMBIA UNIV on April 4, 2011

Rafie AHS et al. to cardiac catheterization, so as to reduce the risk of complications. Therefore, the underdiagnosis of PAD in women is expected to increase their risk of complications related to cardiac catheterization. In those patients with diagnosed with significant CAD, 32% of the women, and 20% of the men, had PAD. Unfortunately, only about one third of these patients had been recognized as having PAD prior to the cardiac catheterization. Other cogent reasons to make the diagnosis of PAD include its impact on exercise and lifestyle issues. Patients with PAD are usually sedentary, in part due to arterial insuf ficiency, but also due to muscular weakness associated with the disease.27 An exercise prescription tailored to their limi tations may enhance their ability to follow recommenda tions. Finally, individuals with PAD need attention to foot care and footwear. Approximately 75% of amputations in patients with PAD are due to avoidable trauma.28 Unfortunately, public awareness of PAD is poor. Despite a high prevalence of PAD in the United States, only 23% of men and 28% of women expressed familiarity with PAD in a survey sample representative of the US adult population. The level of awareness of PAD is significantly less than that of CAD or cardiovascular risk factors.29

449 life and limbs, greater attention to its diagnosis is warranted in patients undergoing cardiac catheterization. Acknowledgements The authors are grateful to Mr Scott Reiff for his adminis trative assistance. Funding This study was supported by grants from the National Heart, Lung and Blood Institute (RO1 HL75774, 1K12 HL087746, 1P50HL083800) as well as a National Institutes of Health grant M01 RR 00070 (General Clinical Research Center, Stanford University School of Medicine) and the Stanford Cardiovascular Institute. References
1. Vaccarino V, Mallik S. Gender differences in the outcome of acute myocardial infarction. In: Legato MJ (ed.) Principles of gender-specific medicine. London: Elsevier Academic Press, 2003, p.224233. 2. NHLBI morbidity and mortality chartbook, 2007. Bethesda, MD: National Heart, Lung and Blood Institute, http://www. nhlbi.nih.gov/resources/docs/07chtbk.pdf (May 2007, accessed July 2009). 3. Hiatt WR. Medical treatment of peripheral arterial disease and claudication. N Engl J Med 2001; 344: 16081621. 4. Selvin E, Erlinger TP. Prevalence of and risk factors for peripheral arterial disease in the United States: results from the National Health and Nutrition Examination Survey, 19992000. Circulation 2004; 110: 738743. 5. Kardys I, Vliegenthart R, Oudkerk M, Hofman A, Witteman JCM. The female advantage in cardiovascular disease: do vascular beds contribute equally? Am J Epidemiol 2007; 166: 403412. 6. Newman AB, Siscovick DS, Manolio TA, et al. Anklearm index as a marker of atherosclerosis in the Cardiovascular Health Study. Cardiovascular Heart Study (CHS) Collaborative Research Group. Circulation 1993; 88: 837845. 7. Makowsky MJ, McAlister FA, Galbraith PD, et al Lower extremity peripheral artery disease in individuals with coro nary artery disease: prognostic importance, care gaps and impact of therapy. Am Heart J 2008; 155: 348355. 8. Sigvant B, WibergHedman K, Bergqvist D, et al. A populationbased study of peripheral arterial disease preva lence with special focus on critical limb ischemia and sex differences. J Vasc Surg 2007; 45: 11851191. 9. Hiatt WR, Hoag S, Hamman RF. Effect of diagnostic criteria on the prevalence of peripheral arterial disease: The San Luis Valley Diabetes Study. Circulation 1995; 91: 14721479. 10. McDermott MM, Liu K, Criqui MH, et al. Anklebrachial index and subclinical cardiac and carotid disease: The MultiEthnic Study of Atherosclerosis. Am J Epidemiol 2005; 162: 3341. 11. Aboyans V, Criqui MH, McClelland RL, et al. Intrinsic con tribution of gender and ethnicity to normal ankle brachial index values: the MultiEthnic Study of Atherosclerosis (MESA). J Vasc Surg 2007; 45: 319327. 12. Hirsch AT, Criqui MH, TreatJacobson D, et al. Peripheral arterial disease detection, awareness, and treatment in pri mary care. JAMA 2001; 286: 13171324. 13. Ouriel K, McDonnell AE, Metz CE, Zarins CK. Critical evaluation of stress testing in the diagnosis of peripheral vas cular disease. Surgery 1982; 91: 686693.

Study limitations
Several study limitations merit comment. We have studied patients undergoing cardiac catheterization, so the findings cannot be generalized beyond the population of patients referred for coronary angiogram. Furthermore, it is possible that the sex differences observed in this paper may be related to selection bias (i.e. perhaps more males than females with PAD died prior to angiography). In addition, we assessed the presence of disease in the coronary versus peripheral circulations with different techniques (i.e. ABI vs angiography). Although peripheral angiography may have provided more definitive information regarding the nature and severity of obstructive disease, it was avoided because of the increased risk associated with the additional contrast agent. Coronary angiography also has technical limitations.30 Studies with intravascular ultrasound (IVUS) have consistently shown that coronary angiograms under estimate atherosclerotic CAD.31 Thus, one possible expla nation of our results could be that coronary angiograms preferentially underestimate hemodynamically significant disease in women.

Summary
In conclusion, we find that among subjects referred for elective coronary angiography to evaluate suspected myo cardial ischemia, women had a higher prevalence of PAD than men, despite having a lower prevalence and severity of CAD. The sex difference in PAD is not explained by tradi tional cardiovascular risk factors or the severity of CAD, and the mechanisms underlying this sex difference are wor thy of further study. Notably, PAD was unrecognized in most patients with low ABIs undergoing elective cardiac catheterization, irrespective of gender. Because PAD is associated with a greater risk of complications involving

Downloaded from vmj.sagepub.com at COLUMBIA UNIV on April 4, 2011

450
14. Criqui MH, Denenberg JO, Bird CE, et al. The correlation between symptoms and noninvasive test results in patients referred for peripheral arterial disease testing. Vasc Med 1996; 1: 6571. 15. Atar D, Ramanujam PS, Saunamki K, Hauns S. Assessment of coronary artery stenosis pressure gradient by quantitative coronary arteriography in patients with coronary artery dis ease. Clin Physiol 1994; 14: 2335. 16. Tonino PA, Fearon WF, De Bruyne B, et al. Angiographic versus functional severity of coronary artery stenoses in the FAME study fractional flow reserve versus angiogra phy in multivessel evaluation. J Am Coll Cardiol 2010; 55: 28162821. 17. Swales JD. Textbook of hypertension. London: Blackwell Scientific, 1994, p. 85102. 18. McDermott MM, Greenland P, Liu K, et al. Leg symptoms in peripheral arterial disease: associated clinical characteristics and functional impairment. JAMA 2001; 286: 15991606. 19. McDermott MM, Kerwin DR, Liu K, et al. Prevalence and significance of unrecognized lower extremity peripheral arterial disease in general medicine practice. J Gen Intern Med 2001; 16: 384390. 20. Higgins JP, Higgins JA. Epidemiology of peripheral arterial disease in women. J Epidemiol 2003; 13: 114. 21. McDermott MM, Greenland P, Liu K, et al. Sex differences in peripheral arterial disease: leg symptoms and physical functioning. J Am Geriatr Soc 2003; 51: 222228. 22. Feinglass J, McDermott MM, Foroohar M, Pearce WH. Gender differences in interventional management of periph eral vascular disease: evidence from a blood flow laboratory population. Ann Vasc Surg 1994; 8: 343349.

Vascular Medicine 15(6)

23. Oka RK, Szuba A, Giacomini JC, Cooke JP. Gender differences in perception of PAD: a pilot study. Vasc Med 2003; 8: 8994. 24. HildickSmith DJ, Walsh JT, Lowe MD, et al. Coronary angiography in the presence of peripheral vascular disease: femoral or brachial/radial approach? Catheter Cardiovasc Interv 2000; 49: 3237. 25. Magnant JG, Cronenwett JL, Walsh DB, Schneider JR, Besso SR, Zwolak RM. Surgical treatment of infrainguinal arterial occlusive disease in women. J Vasc Surg 1993; 17: 6776. 26. Norman PE, Semmens JB, LawrenceBrown M, Holman CD. The influence of gender on outcome following peripheral vas cular surgery: a review. Cardiovasc Surg 2000; 8: 111115. 27. McDermott MM, Liu K, Greenland P, et al. Functional decline in peripheral arterial disease: associations with the ankle bra chial index and leg symptoms. JAMA 2004; 292: 453461. 28. Carmona GA, Hoffmeyer P, Herrmann FR, et al. Major lower limb amputations in the elderly observed over ten years: the role of diabetes and peripheral arterial disease. Diabetes Metab 2005; 31: 449454. 29. Hirsch AT, Murphy TP, Lovell MB, et al. Gaps in public knowledge of peripheral arterial disease: the first national PAD public awareness survey. Circulation 2007; 116: 20862094. 30. Johnson BD, Shaw LJ, Buchthal SD, et al. Prognosis in women with myocardial ischemia in the absence of obstruc tive coronary disease: results from the National Institutes of HealthNational Heart, Lung and Blood InstituteSponsored Womens Ischemia Syndrome Evaluation (WISE). Circulation 2004; 109: 29932999. 31. Nissen SE, Yock P. Intravascular ultrasound: novel patho physiological insights and current clinical applications. Circulation 2001; 103: 604616.

Downloaded from vmj.sagepub.com at COLUMBIA UNIV on April 4, 2011