- Email: [email protected]
Segmental arterial disease in the lower extremities: Correlates of disease and relationship to mortality
J CUDEpidmiol Vol.46,No. I I, pp.
12674276, Printed in Great Britain. All rights reserved
1993 Copyright 0
0895-4356/93 $6.00 + 0.00 1993 Pergamon Press Ltd
SEGMENTAL ARTERIAL DISEASE IN THE LOWER EXTREMITIES: CORRELATES OF DISEASE AND RELATIONSHIP TO MORTALITY MOLLY T. VOGT,‘* SIDNEY K. WOLF-SON’and LEWIS H. KULLER’ ‘Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, PA 15261 and ‘Department of Surgery and Neurological Surgery, School of Medicine, University of Pittsburgh and Montefiore University Hospital, Pittsburgh, PA 15261, U.S.A. (Received in revised form 13 April 1993)
Abstract-The purpose of this study was to identify the correlates of disease and mortality risk associated with the presence of stenosis in the major arterial segments of the lower limb in men and women evaluated in a peripheral vascular laboratory. A total
of 326 men and 249 women had unisegmental stenosis; 312 men and 275 women had multisegmental disease; 132men and 111 women had no apparent disease. Multivariate analysis indicated that current smoking and elevated systolic blood pressure were the key risk factors associated with isolated aortoiliac and femoropopliteal arterial disease in both men and women. A history of diabetes was significantly associated with tibioperoneal disease in men while elevated systolic blood pressure was the major correlate of distal disease in women in this patient population. The relative risk of mortality was elevated 2- to 7-fold in men and women with multi- and unisegmental disease involving the aortoiliac and femoropopliteal segments; presence of tibioperoneal disease did not significantly increase mortality relative to those who were free of disease. These results suggest that the etiology and mortality risk associated with atherosclerosis in the lower extremity may vary with the anatomic site and/or severity of the lesion. Atherosclerosis
Mortality
Risk factors
Vascular
INTRODUCTION
Atherosclerosis is a systemic disease in which plaques develop primarily at branches in the arteries and in areas where the blood flow is slow or turbulent. The terminal aorta and its branches supplying the lower extremity are frequently involved [ 1,2]. The atherosclerotic process in the vessels of the lower extremities usually occurs in conjunction with that in the coronary and cerebral arteries [3,4]. Life expectancy is decreased by about 10 years in patients with lower extremity disease [5] and the majority of these patients die from cardiac disease or stroke [6-l 11. *Author for correspondence.
disease
Lower limb atherosclerosis usually develops at several levels within the arteries but may also be restricted to a single localized region in a vessel. When stenosis is found at multiple sites the blood flow to the lower limbs is often severely compromised and limb ischemia may become a significant clinical problem. Such multisegmental disease is also associated with poorer overall cardiovascular health status and higher morbidity and mortality than in localized unisegmental disease [13-l 51. Several studies have suggested that the etiology of the disease process in the lower extremity may vary depending on the anatomic site of the lesion [4, 13, 14, 16, 171. Aortoiliac disease seems to occur in younger patients who smoke, while the more distal lesions in the infrapopliteal
1267
1268
MOLLYT. VOGTet al.
arteries are found primarily in diabetics [13, 14, 16-191. The correlates of disease are similar for men and women, although women are reported to be 8-10 years older before clinical symptoms become evident [12,20]. The majority of studies conducted to date have involved relatively small numbers of patients [19], and have not used multivariate analysis to adjust for confounding factors [13, 14, 191. The purpose of this study is to identify the characteristics in men and women associated with the development of stenotic lesions in the major arterial segments of the lower limb (aortoiliac, femoropopliteal and the tibioperoneal vessels); and to determine the relationship between the anatomical location of the lesion and subsequent mortality.
METHODS
Patients evaluated for lower extremity arterial disease in the Montefiore University Hospital Peripheral Vascular Laboratory, Pittsburgh, PA between 1 January 1977, and 31 December 1989, were included in this study if they were 50 years of age or older and had no history of arterial surgery or amputation involving the lower extremity (n = 1930). Patients were referred to the laboratory for a variety of reasons including complaints of leg and hip pain at rest or during exercise, presence of non-healing ulcers, gangrene, inability of primary physician to palpate pedal pulses, history of diabetes, painful feet and for evaluation prior to arterial bypass surgery. The majority of patients were 50 years of age or older and since several previous reports [14,21-23]have characterized peripheral vascular disease as being “particularly virulent” in the younger age groups it was decided to limit our study population to those 50 years or older. The relevant medical history of each patient was taken by a trained technologist as described previously [lo]. Information was collected on use of medication, tobacco use and past history of claudication, diabetes, stroke, transient ischemic attack, amputations, vascular and nonvascular surgery, cardiac disease and respiratory disease. Ischemic heart disease was defined as a history of myocardial infarction and/or coronary artery bypass grafting. Angina was defined as angina in the absence of known ischemic heart disease. Arterial evaluation of the lower extremities was carried out according to standardized procedures developed at the Montefiore University
Hospital. Patients were evaluated in the supine position with the appropriately sized standard blood pressure cuffs placed on the upper arm, thigh, calf, and ankle. Systolic blood pressure was measured in both arms and then sequentially on both legs; first the right side (ankle, calf, thigh, arm) and then on the left side. A Doppler ultrasonic instrument was used to detect flow in the distal arteries. The cuff was inflated until the flow was no longer detectable and then slowly deflated. The pressure at which the flow first returned to the artery being monitored was recorded. The ankle arm index (AAI) for each leg was calculated by dividing the ankle pressure (the average of the pressures recorded for the posterior and anterior tibia1 arteries) by the higher of the recorded brachial pressures. The lower of the AA1 values obtained for the two legs was used as the measure of disease severity in the subsequent data analysis. In patients for whom it was impossible to obtain both the left and right AA1 (due to pain or ulceration/gangrene at ankle) the value obtained on the other limb was used for analysis. A segmental lesion was defined as present when the AA1 was ~0.90 and there was a pressure gradient of at least 20 mmHg between the systolic pressures measured at adjacent segmental levels [24-281, i.e. an aortoiliac lesion was recorded as present when the brachial pressure minus the thigh pressure was 2 20 mmHg; a femoropopliteal lesion was present when the thigh pressure minus the below-knee pressure was 220 mmHg; and a tibioperoneal lesion when below-knee pressure minus the average of the tibia1 pressures was 2 20 mmHg. The nolesion control group of patients were those with an AA1 > 0.90 but c 1.50 and no lesions. Patients with any missing segmental blood pressure measurements were excluded from the analyses. Patients were categorized by anatomical level of disease and by the number of major arterial segments involved, i.e. isolated disease involving either the aortoiliac, femoropopliteal or tibioperoneal segments; multisegmental disease involving either the aortoiliac-femoropopliteal, aortoiliac-tibioperoneal, femoropopliteal-tibioperoneal or aortoiliac-femoropopliteal-tibioperoneal arteries. A total of 326 men and 249 women had unisegmental disease; 3 12 men and 275 women had multisegmental disease; and 132 men and 111 women had no apparent stenotic disease.
Segmental Arterial Disease in the Lower Extremities
Criqui and his colleagues have reported that ankle-arm blood pressure ratios GO.85 have over a 90% sensitivity and 99% specificity in detecting arteriographically positive disease [29,30]. Other investigators [31,32] have found that the cutpoint for AA1 may be as high as 0.97 with little loss of sensitivity or specificity. The reproducibility of sensitivity and quality of segmental blood pressure measurement is high in subjects with and without clinical peripheral vascular disease [29]. We have also found very good reproducibility of ankle arm blood pressure measurements among older individuals with systolic hypertension [33]. Fowkes [34,35] has noted that the variability in measuring the ankle systolic blood pressure and ankle arm index is similar to the measure of arm systolic blood pressure and that the “variability of taking measurements on different occasions by different observers has only a marginal effect on between subject variability” [34]. Similarly, the use of systolic blood pressure gradients of 20 mmHg or more in conjunction with AA1 measurements, has been shown to be highly specific (over 90%) and reasonably sensitive (86%) for detecting isolated segmental arterial disease as compared to arteriography [27,28]. Ascertainment of death The total patient list was compared with the death files maintained by the Department of Vital Statistics, Commonwealth of Pennsylvania. The files for years 1977-1989 were searched manually by matching name, social security number, date of birth and county of residence for each patient. The underlying cause of death was recorded and subsequently coded by a physician (LHK) blind to the medical history and vascular diagnosis of the patients. A random sample of patients (N = 200) for whom death certificates were not found, were followed up to assess the completeness of the death ascertainment. Among patients without lower extremity arterial disease, 4% of deaths were missed or occurred out of state; 7% were missed for those with disease. Statistical analysis The chi-square statistic was used to assess the differences for categorical variables; t-tests were used for continuous variables. Two-tailed p values were used in all tests and the 5% level was considered significant. No correction was made for multiple comparisons.
1269
In the cross-sectional analysis of the data, logistic regression uni- and multivariate methods were used to identify the baseline characteristics associated with the presence of arterial disease. The lifetable actuarial method was used to assess the difference between survival of the controls and those with disease and the p-value was determined using the logrank test [36,37]. The Cox proportional hazards model was used to determine the relative risk of mortality and to adjust for covariates and for the varying follow-up times [38]. For both logistic regression and Cox proportional hazards models the dichotomous dependent variable was the presence of a lesion at a given site/absence of lesion at all sites studied. Dummy variables were used to code for smoking status (never, former, current) and heart disease (none, angina, definite ischemic heart disease). Other independent variables (systolic blood pressure > 140 mmHg, diabetes, stroke, congestive heart failure) were entered into the models as dichotomous variables. Patient records with missing values were excluded from the analyses.
RESULTS
After excluding patients with incomplete data, a total of 638 men and 524 women were found to have evidence of stenosis in at least one arterial segment based on our criteria for the definition of disease. A relatively small percentage of patients had lesions in all three arterial segments (2% of men, 5% of women) while approximately half the patients had unisegmental disease (Table 1). The men and women with aortoiliac disease were of similar average age to patients with no lesions; and they were more likely to be current smokers when they were first evaluated in the laboratory. In addition, women with disease were more likely to have a history of ischemic heart disease and dyspnea and to be former smokers. The men, on the other hand, had elevated systolic blood pressure and were more likely to have a history of chronic obstructive pulmonary disease. Both men and women with femoropopliteal lesions had high blood pressure and were more likely to have a history of stroke and to be current smokers. The women with disease were older than those with no lesions and presence of disease was associated with a history of ischemic heart disease and congestive heart failure.
MOLLYT. VOGTet al.
1270
Table 1. Distribution of lesions Men
Women
Aortoihac-femoropopliteal
and tibioperoneal lesion in one or both legs (AI-FP-TP)
14 (2%)
26 (5%)
Aortoiliac-femoropopliteal either leg (AI-FP)
lesions in one or both legs; no tibioperoneal lesions in
114 (18%)
115 (22%)
49 (8%)
58 (11%)
57 (9%)
40 (8%)
Isolated aortoiliac lesion in one or both legs; no femoropopliteal or tibioperoneal lesions in either leg (AI)
127 (20%)
66 (13%)
Isolated femoropopliteal lesion in one or both legs; no aortoihac or tibioperoneal lesions in either leg (FP)
134 (21%)
138 (26%)
Isolated tibioperoneal lesion in one or both legs; no aortoiliac or femoropopliteal lesions in either leg (TP)
65 (10%)
45 (9%)
638 (100%)
524 (100%)
Aortoiliac-tibioperoneal either leg (AI-TP)
lesions in one or both legs; no femoropopliteal lesions in
Femoropopliteal-tibioperoneal either leg (FP-TP)
lesion in one or both legs; no aortoihac lesions in
Total with a lesion* *Numbers do no sum to the total due to the exclusions.
Patients with tibioperoneal disease were older than the control group and more likely to have congestive heart failure. A significantly higher proportion of men with distal lesions were diabetic while the women were also more likely to have a history of stroke and ischemic heart disease (Tables 2 and 3). The occurrence of femoropopliteal and tibioperoneal disease in women increased sharply with age while the more proximal disease was weakly related to increasing age. In men, presence of aortoiliac disease was inversely
associated with age, the disease being most prevalent in the W-59 year age group; more distal lesions showed a weak positive association with age (Table 4). After age-adjustment the major characteristic associated with the presence of isolated aortoiliac and femoropopliteal disease in both men and women was smoking status, with current smokers at an approximately 3- to S-fold greater risk than those who reported having stopped smoking (Table 4). Increased systolic blood pressure was a major correlate of disease in men, while
Table 2. Comparison of baseline characteristics of male patients with isolated aortoiliac, femoropopliteal or tibiopcroneal lesions and with no lesions
Age (yr) AA1 SBpt (mmHg) DBPj (mmHg) % SBPt > 140 % DBP$ > 90 Heart rate (beatslmin) % White race % Angina % Ischemic heart disease % CHF % Diabetes % Stroke % Dyspnea % COPD % Ever smoked % Currently smokes % formerly smoked
No lesions n = 132
Aortoihac lesions only n = 127
65.7 f 9.7 l.OS+O.ll 134.9 f 19.9 78.0 f 10.2 34.1 9.3 72.2 f 12.4
64.2 f 8.6 0.59 * 0.132 145.4 + 24.3* 79.5 f 11.6 55.1* 14.3 73.9 + 10.5
67.5 f 8.8 0.57 + 0.16* 141.7 + 23.4* 75.8 + 10.6 50.0’ 6.6 72.7 k 11.0
69.1 f 9.7+ 0.81 + 0.21* 136.9 + 22.0 77.2 + 10.9 40.0 8.1 72.1 + 10.9
80.3 15.9 15.2
88.7 16.5 20.5
86.2 11.9 22.4
78. I 20.0 20.0
5.3 26.7 6.1 9.1 7.6 74.4 26.4
6.3 16.6 10.0 8.7 15.7* 94.5* 59.8’
6.7 27.8 17.1* 11.2 11.2 87.8* 47.3*
18.5* 43.8* 11.9 9.2 9.2 69.2 20.0
48.1
34.6,
40.4
49.2
‘p < .05 compared with male patients with no lesions. tSystolic blood pressure. tDiastolic blood pressure.
Femoropopliteal lesions only n = 134
Tibioperoneal lesions only n =65
Segmental Arterial Disease in the Lower Extremities
1271
Table 3. Comparison of baseline characteristics of female patients with isolated aortoiliac, femoropopliteal or tibioperoneal lesions and with no lesions
Age 64 AA1 SBPt (mmHg) DBPJ (mmHg) % SBPt > 140 % DBPj > 90 Heart rate (beats/min) % White race % Angina % Ischemic heart disease % CHF % Diabetes % Stroke % Dyspnea % COPD % Ever smoked % Currently smokes % Formerly smoked
No lesions n = Ill
Aortoiliac lesions only n=66
64.4 f 9.3 1.07 kO.11 137.0 k 20.6 75.8 k 9.2 37.8 5.6 73.0 + 8.7
66.5 + 8.3 0.56 f 0.13’ 139.5 + 22.8 72.6 rl: 7.8+ 48.5 0 69.7 k lO.4*
72.0 k l0.6* 0.57 & 0.18* 144.8 + 25.4’ 75.1 * 13.3 55.1; 8.9 74.35 II.0
70.6 f 9.0* 0.73 +O.l8* 145.3 k 21.6’ 76.7 k II.2 64.4; 6.8 73.6 k 10.4
80.2 16.2 3.6
78.8 25.8 l3.6*
69.9 23.2 10.9;
82.2 17.8 20.0*
2.7 24.3 3.3 7.2 2.7 46.7 23.4
7.6 19.7 9.7 l8.2* 4.5 95.52 63.6*
10.9* 28.7 12.8’ 10.9 5.1 63.2* 34.6*
l5.6* 33.3 17.1* 6.7 4.4 35.7 II.9
23.4
31.8*
28.6
23.8
Femoropopliteal lesions only n = 138
Tibioperoneal lesions only n =45
*p < 0.05 compared with female patients with no lesion. tsystolic blood pressure. SDiastolic blood pressure.
a history of stroke, ischemic heart disease and CHF were associated with increased risk in women. Multivariate logistic regression analysis indicated that current smoking remained the strongest independent correlate of disease in both sexes after correction for the other covariates. A history of diabetes was strongly associated with the presence of tibioperoneal lesions in men, while in women a systolic blood pressure greater than 140 mmHg was the most important factor (Table 4). Survival in men and women with uni- and multisegmental disease was significantly decreased compared to those with no segmental lesions (Table 5). After adjusting for age, the relative risk for all-cause mortality was approximately doubled in men with a single lesion at any segmental level and in those with multisegmental disease. After adjustment for elevated systolic blood pressure, smoking status, a history of congestive heart failure, stroke, diabetes, angina and ischemic heart disease at baseline, the relative risks remained the same except for an increase in risk of mortality for those with aortoiliac-femoropopliteal or with aortoiliacfemoropopliteal-tibioperoneal disease (Table 6). Mortality risks in women were elevated 4- to 5-fold when those with aortoiliac-femoropo-
pliteal and aortoiliac-tibioperoneal lesions were compared with women with no detectable lesions. As in men, isolated tibioperoneal disease was associated with the lowest risk of mortality. The mean ankle arm index of patients with multisegmental disease was lower than that for those with unisegmental disease; the indices for the patients with isolated tibioperoneal lesions were markedly higher than those for any other patient group with lesions (Table 5). In order to examine the effects of severity of occlusive disease as well as the site of the stenotic lesion, the analysis was repeated including only those patients with an ankle arm index of 0.7 or less. The relative risk of mortality in each group was basically unchanged except that the risk of mortality in women with isolated tibioperoneal disease increased from 1.5 to 3.1.
DISCUSSION
Smoking is the key factor associated with the presence of isolated aortoiliac and femoropopliteal disease in the men and women studied here. This association is lower in those who have stopped smoking but remains elevated compared to lifelong non-smokers. Women with a history of smoking are particularly susceptible to aortoiliac disease. Smoking was not related to
MOLLYT. VOGTet al.
1272
Table 4. Odds ratios for risk factors associated with presence of isolated arterial lesions in male and female natients. 95% confidence interval shown in parentheses Men OR*
Women OR*
ORt
ORt
Aortoiliac lesion
Age (per 1Oyr) (0.69.81.1) (0.69.81.2)
(l.Ei 3) 491’ (47.1, 511.0) 34.9 (5.34, 228.0)
(l.iii.2)
Current smoker Former smoker (1.0, 6.7) SBPI > 140 mmHg
(2.d.b.3) 2.6 (1.0, 6.7)
c3.0g.:3. 1) 2.6 (1.0, 7.1)
(19.2, 474.0) 20.2 (4.8, 85.2)
(1$.5)
(l.Z.4)
(0.2q.so.9)
(0.::
(0.;;:.6)
(l.$.l)
(l$.O)
(1.?88.8)
(2.4’t.3) i.5
(2.8?5.3) 2.6 (1.2, 5.6)
(l.65.:6.0) i.6 (0.2, 2.1)
(O.d,.83.5)
Diabetes (0$.6)
.O)
Femoropopliteal lesion
Age (per 10 yr) (2.03’046) 9.5. (3.7, 24.2)
Current smoker Former smoker (O.:;L) SBP$ > 140 mmHg
(l.d.93 5) i.7 ’ (0.9, 3.4)
(1 .G.9)
(O.!Yl 7) d.7. (0.3, 1.9)
(0.6,.:.8)
(1.:.:.3)
(l.$.O)
(0.39.82.3)
(0.$82.6)
(O&24.7)
(0.49.81.6)
(0.49.92s
(0.49.92.4)
(0.2q.P.9)
(l.t$.l)
(0.!91.9)
(1.22,.86.3)
(1.t34.6)
(l$.S)
(OE7)
(l.75.Y4.4) b.8 (0.3, 2.2)
(l.d.83 0) i.3 ’ (0.7, 2.3)
Diabetes
(1$9)
(O.6;:.“’ (1.3,‘4.3) 1.3 (0.7, 2.5)
(o.d;94.2)
Tibioperoneal lesion
Age (per 10yr) Current smoker Former smoker SBP# > 140 mmHg Diabetes
*Age included in all models. tAdjusted for all other covariates shown, as well as history of angina, ischemic heart disease, stroke and congestive heart failure. $Systolic blood pressure.
Table 5. Five and ten-year survival in patients with atherosclerotic lesions at various sites* n
AI-FP-TP lesions AI-FP AI-TP FP-TP Isolated AI Isolated FP Isolated TP No lesions AI-FP-TP lesions AI-FP AI-TP FP-TP Isolated AI Isolated FP Isolated TP No lesions
14
wt
114 (58) 49 (18) 57 (24) 127 (47) 134 (52) 65 (25) 132 (31) 26 (11)
115 (62) 58 (27) 40 (17) 66 (20) 138 (55) 45 (12) 111 (14)
AAI
5 yr survival
1Oyr survival
Men 0.41 f 0.15 0.46 f 0.14 0.44 + 0.14 0.52 + 0.13 0.59 f 0.13 0.57 f 0.16 0.81 k 0.21 1.08 kO.11
0.62 0.52 0.68 0.60 0.66 0.62 0.68 0.75
0.32 0.40 0.36 0.52 0.40 0.40 0.65
0.48 0.39 0.44 0.55 0.82 0.57 0.66
0.48 0.32 0.44 0.41 0.58 0.36 0.66
0.90
0.78
Women 0.44 + 0.14 0.44 f 0.12 0.44+0.15 0.50 f 0.12 0.56 + 0.13 0.57 f 0.13 0.73 f 0.18 1.07Io.11
*Sites of lesions defined in Table 1 tnumber of patients that died during follow up.
-
o-Value 0.0020 0.0000 0.0603 0.0073 0.0462 0.0021 0.0100 ref 0.0000 O.OfMO 0.0000 0.0000
0.0015 0.0000 0.0087 ref
1273
Segmental Arterial Disease in the Lower Extremities Table 6. Relative risk of all-cause mortality in patientswith atheroscleroticlesionsat varioussites*comparedto with no lesions, 95% confidence interval shown in Parentheses
patients
Women
Men Site of lesion
Relative riskt
Relative riski
Relative risk$ AA1 < 0.7
Relative riskt
Relative riski
Relative risk2 AA1 < 0.7
AI-FP-TP AI-FP AI-TP FP-TP Isolated AI Isolated FP Isolated TP
2.6 (1.2, 2.5 (1.6, 1.6 (0.9, 1.7 (1.0, 1.7 (1.1, 2.1 (1.3, 1.8 (1.1,
7.2 (2.2, 3.5 (2.0, 1.4 (0.7, 1.8 (0.9, 2.0 (1.1, 2.6 (1.5, 1.3 (0.7,
7.2 (2.2, 3.5 (2.1, 1.4 (0.7, 2.0 (0.9, 2.1 (1.2, 2.4 (1.3, 1.3 (0.6,
3.3 5.2 4.2 2.5 2.9 3.2 2.1
2.4 (0.6, 5.3 (2.3, 4.5 (1.6, 2.5 (0.9, 2.8 (1.2, 2.1 (1.0, 1.5 (0.4,
2.4 5.3 4.0 2.6 4.5 2.4 3.1
5.6) 3.8) 2.8) 2.9) 2.7) 3.3) 3.1)
23.1) 6.1) 2.9) 3.8) 3.5) 4.5) 2.5)
23.1) 7.3) 2.8) 4.1) 3.8) 4.3) 2.8)
(1.4, (2.8, (2.1, (1.1, (1.5, (1.7, (0.9,
7.9) 9.8) 8.5) 5.5) 5.7) 5.8) 4.8)
10.1) 12.2) 12.8) 7.6) 6.7) 4.7) 5.2)
(0.6, (2.3, (1.4, (0.9, (1.6, (1.1, (0.9,
10.1) 12.3) 11.5) 7.6) 12.8) 5.2) 11.0)
*Sites of lesions defined in Table 1. TAdjusted for age. $Adjusted for age, systolic blood pressure, current and former smoking and a history of angina, ischemic heart disease, diabetes, stroke and congestive heart failure.
tibioperoneal disease in our patients and previous clinical observations [19] tend to support this finding. There have been no population-based studies to date which have considered etiologic factors in lower extremity arterial disease in relation to the anatomic site of the arterial occlusion. Calculations of the population attributable risk for smoking and lower extremity arterial disease, considering lesions at all levels, has varied from 14% to 78% [7] depending, in part, on the definition of disease and population characteristics. A study of normal and hyperlipidemic subjects in California [30] showed that smoking is a risk factor for large vessel lower extremity arterial disease in men but not in women. A more recent study of healthy elderly women found that current smoking increased the odds of disease 6-fold [39]. In addition, several clinical studies which focused on the assessment of patients for reconstructive vascular surgery of the lower limb, have reported that current smoking is strongly associated with both uni- and multisegmental disease involving aortoiliac and femoropopliteal lesions [40,41]. Multivariate analytic techniques were not used for data analysis in these studies and the effect of comorbid conditions and other risk factors were not considered. Our research indicates that elevated systolic blood pressure is an independent correlate of disease in each of the three arterial segments studied in women and of aortoiliac and femoropopliteal disease in men. Most studies, to date, have shown that elevated systolic blood pressure is associated with lower extremity arterial disease [3,39,42,43] although Gofin [44] and Criqui [30] did not find it to be a significant factor in women. None of these studies considered the relationship of elevated blood press-
ure to lesions at various levels in the arterial system. Diabetes is usually considered to be an important factor for development of lower extremity arterial disease with the occlusive process affecting the most distal vessels, primarily the posterior tibia1 and peroneal arteries [9, 13, 191. In this study, diabetes was a significant independent correlate of tibioperoneal disease in men but not women. Since diabetes is usually associated with diffuse and generalized atherosclerosis, many diabetics would be excluded from the study of risk factors for unisegmental disease. The low statistical power associated with the relatively small number of patients may explain our inability to show an association between diabetes and distal lesions in women (with 45 women having tibioperoneal disease, there is approximately 35% power to detect an odds ratio of 2.0). It is noteworthy that the percentage of patients with diabetes increases as the anatomic site of the lesion studied becomes more distal; this is true in both men and women but reaches statistical significance only in men. In addition, the diabetic patients may have a high degree of arterial calcification in the tibia1 arteries. Blood pressure measurements in rigid calcified arteries give artifactually high values, resulting in the classification of these patients as free of disease. We have attempted to minimize such misclassification of patients by excluding all patients with unusually high values of the ankle-arm index (greater than 1.50). Research studies to date have found that although the risk factors for atherosclerosis are similar in all vascular beds, the dominant risk factor seems to vary by anatomic location [41,42,45-511. For instance, elevated systolic blood pressure is the key predictor for stroke [45,46] while current smoking is a major risk
1274
MOLLY
T. VOGT et al.
factor for lower extremity arterial disease [47,50]. In addition, a number of clinical studies have suggested that even within the major arteries of the leg, various segments may react differently to stressors such as smoking and diabetes [13, 14, 16-191. The mechanism for such anatomically differential responses is unknown [19] although it may be related to arterial diameter, the histologic structure of the vessel wall, the turbulence of the blood flow through the various sections of the artery and/or specific biologic influence of the various risk factors on the atherosclerotic process [51]. Measurement of segmental limb blood pressures is a highly specific method of identifying isolated segmental disease but the sensitivity is considerably lower and varies depending on the anatomical location of disease [27,28]. Utilizing both the ankle-arm index and the segmental pressure gradients as in this study, the overall sensitivity is 86% for isolated segmental disease as compared to detection of disease by arteriography [271. The major sources of error noted by Rutherford [27] are errors in measurement of blood pressures due to arterial calcification as mentioned above; inability to detect stenoses that do not cause a change in pressure gradient at rest; inability to detect tibioperoneal disease that does not involve all three infrapopliteal arterial branches; and correct localization of stenoses that are under the blood pressure cuff. All of these sources of error would tend to bias our results toward an underestimation of effect size. Atherosclerosis in the lower extremity is associated with increased risk of mortality [7-l 11. The more severe the disease and the more segmental levels involved the greater the risk [7, l&13,52]. The location of the lesion did not affect mortality in some studies [15] but in others, distal disease was more indicative of early mortality [52-551. Our results indicate that crude survival is decreased in patients with lesions at every segmental level. After adjusting for other covariates, the all-cause mortality risk is significantly elevated in isolated aortoiliac and femoropopliteal disease and in aortoiliacfemoropopliteal-tibioperoneal, aortoiliac-femoropopliteal disease in both sexes and in aortoiliac-tibioperoneal disease in women. The lowest risk of dying was found in men and women with tibioperoneal disease and in men with femoropopliteal-tibioperoneal disease. The patients with unisegmental distal disease had ankle arm indices that were higher than
those with lesions in other locations, i.e. overall the arterial disease was less severe in these patients. After stratifying by disease severity (based on the ankle arm index) the women with tibioperoneal disease and an ankle arm index of 0.7 or less had an increased risk of mortality; no change in risk was found in the men. The differential mortality by site of lesion may therefore be related to disease severity rather than the site of the lesion. However, the prevalence of congestive heart failure is several fold higher among patients with distal disease compared to those with no lesions; although we have adjusted for this fact in our multivariate analysis, residual confounding may be affecting the results obtained. Kallero [53] and Martinez [54] have both reported that distal disease is associated with excess deaths due to myocardial disease. However, their patients had all undergone prior arterial reconstruction of proximal vessels and thus originally had multisegmental disease. In addition, the Ranch0 Bernard0 study of normal and hyperlipidemic men and women showed that posterior tibia1 arterial disease was associated with increased all cause and coronary heart disease mortality [1 11. Additional research is needed to resolve these apparent discrepancies. The results of this study must be interpreted with caution. Firstly, as noted earlier in this section, studies on patients referred to a particular clinical site are likely to be biased and the results may not be generalizable to other patient populations. Second, the controls used in the analyses are also patients and in poorer general health than men and women of similar age in the general population. Third, it is possible that some of patient deaths may have been missed, especially if they occurred outside Pennsylvania. Our detailed follow up study of 200 patients believed alive as of 31 December 1989, (see Methodology) indicated that a relatively small percentage of deaths had been missed and that underascertainment of death may have been more of a problem in patients with a low ankle-arm index than in those with no lesions; this would result in an underestimation of the relative risk of dying. Fourth, the patient records contained no information on the serum lipid levels. Most research to date has not found lipids to be a major risk factor for the development of arterial disease in the lower extremities [12, 30,44, 501. Similarly, the relationship between cholesterol levels and mortality in older populations seems to be much weaker than in
Segmental Arterial Disease in the Lower Extremities
middle-aged men and women [56]. Fifth, the small numbers in some of the patient groups studied means that the power to detect statistically significant associations is low. Finally, the models obtained from the data analysis have not been validated in a test population. Our patient population, although larger than that used in most other clinical studies, is not of sufficient size to split into two groups to permit model development and then model testing within the same population. In spite of these limitations, the results presented show that current smoking and high systolic blood pressure are the major independent factors related to the development of aortoiliac and femoropopliteal disease in this patient population. Lesions in these two arterial segments are associated with increased mortality. A history of diabetes is an independent correlate of more distal unisegmental disease in men but the presence of infrapopliteal lesions does not appear to significantly affect the risk of dying. The 5 year probability of dying for women with multisegmental disease is about 50% (Table 5) as compared to 2040% in those with unisegmental disease and 10% in women with no detectable arterial lesions. The trend is similar in men but much less pronounced. After adjusting for a variety of confounders (Table 6) these large increases in mortality related to severity and/or location of arterial lesion remain. Thus it is of considerable clinical interest to continue research to further delineate the role of smoking, hypertension, diabetes and impaired glucose tolerance as correlates of disease development in each of the major arterial segments of the lower extremity and to clarify the relationship between the site of the lesion, degree of occlusion, involvement of multiple arterial segments and subsequent morbidity and mortality. It will also be important to assess the effectiveness and efficacy of pharmacologic, dietary and lifestyle interventions on the development and course of disease, not only in clinical but also population-based studies. REFERENCES DeBakey ME, Lawrie GM, Glaeser DH. Patterns of atherosclerosis and their surgical significance. Ann Surg 1985; 201: 115-121. Ross R. The pathogenesis of atherosclerosis-an update. N Engl J Med 1986; 314: 488-500. Kannell WB, McGee DL. Update on some epidemiologic features of intermittent claudication: The Framingham study. J Am Cerirtr Sot 1985; 33: 13-18.
1275
4. Hertzer NR. The natural history of peripheral vascular disease: implications for its management. Circulation 1991; 83(Suppl.l):
112-19.
5.
Malone JM, Moore WS, Goldstone J. Life expectancy following aortofemoral arterial grafting. Surgery 1977; 81: 551-555. 6. Jelnes R, Gaardsting 0, Hougard Jensen K et al. Fate in intermittent claudication: outcome and risk factors. Br Mad J 1986; 293: 1137-1140. 7. Dormandy J, Mahir M, Ascady G et al. Fate of the patient with chronic leg ischaemia. J Cardiovasc Surg 1989; 30: 50-57.
8.
9.
10.
11.
12.
Kallero KS. Mortality and morbidity in patients with intermittent claudication as defined by venous occlusion plethysmography. A ten year f&ow-up study. J Chroo Dis 1981; 34: 455-462. Cronenwett JL, warner KG, Zelenock GB ef al. Intermittent claudication. Current results of nonoperative management. Arch Surg 1984; 119: 430-436. McKenna M, Wolfson S, Kuller L. The ratio of ankle and arm arterial pressure as an independent predictor of mortality. Atherosclerosis 1991; 87: 119-128. Criqui MH, Langer RD, Fronek A ef al. Mortality over a period of 10 years in patients with peripheral arterial-disease. N Eigl J Med 1992; 326: %-386. Voet MT. Wolfson SK. Kuller LH. Lower extremitv art&al disease and the aging process: a review. J Glib Epidemiol 1992; 45: 529-542.
13. Samson RH, Scher LA, Veith FJ. Comorbid segmental arterial disease. Surgery 1985; 97: 385-396. 14. McDaniel MD, Cronenwett JL. Basic data related to the natural history of intermittent claudication. Ann Vast Surg 1989; 3: 273-277. 15. Jonason T, Rinquist I. Mortality and morbidity in patients with intermittent claudication in relation to the location of the occlusive atherosclerosis in the leg. Angiology
1985; 36: 310-314.
16. Strandness DE, Priest RE, Gibbons GE. Combined clinical and pathological study of diabetic and nondiabetic peripheral arterial disease. Diabetes 1964; 13: 36G372.
17. Conrad MC. Large and small artery occlusion in diabetes with severe vascular disease. Circulation 1967; 36: 83-91.
18. Bild DE, Selby JV, Sinnock P, Browner WS, Braveman P, Showstack JA. Lower-extremity amputation in people with diabetes. Diabetes Care 1989; 12: 24-31.
19. Menzoian JO, LaMorte WW, Paniszyn CC et al. Symptomatology and anatomic patterns of peripheral vascular disease: differing impact of smoking and diabetes. Ann Vast Surg 1989; 3: 224228. 20. A. Widmer LK. Occlusive Peripheral Artery Disease. The Basle Study. Huber: Bern; 1980. 21. Evans WE, Hayes JP, Vermillion BD. Atherosclerosis in the younger patient. Am J Surg 1987; 154: 225-229. 22. Nunn DB. Symptomatic peripheral arteriosclerosis of patients under age 40. Am J Surg 1973; 39: 224-228. 23. McCready RA, Vincent AE, Schwartz RW et al. Atherosclerosis in the young: a virulent disease. Surgery 1984; 96: 863-868. 24. Carter SA. Role of pressure measurements in vascular disease. In: Bernstein EF, Ed. Noninvasive Diagnostic Techniuues in Vascular Disease. St Louis, MO: C.V. Mosby-Co; 1985: 513-544. 25. Raines J. Traad E. Noninvasive evaluation of oeriphera1 vascular disease. Med Clin N Am 1980: 64(2): 283-304. 26. Gundersen J. Segmental measurements of systolic blood pressure in the extremities including the thumb and the great toe. Acta Chir Stand 1972; 426: I-10.
1276 27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38. 39.
40.
41.
MOLLY T. VOG~et al. Rutherford RB, Lowenstein DH, Klein MF. Combined segmental systolic pressures and plethysmography to diagnose arterial occlusive disease of the legs. Am J Surg 1979; 138: 211-218. Barnes RW. Noninvasive diagnostic assessment of peripheral vascular disease. Circulation 1991; 83(Suppl.I): 120-127. Criqui MH, Langer RD, Fronek A, Feigelson HS, Klauber MR. Large vessel and isolated small vessel disease. In: Fowkes FGR, Ed. Epidemiology of Peripherai Vascular Disease. London: Springer-Verlag; 1991: 85-96. Criqui MH, Browner D, Fronek A er al. Peripheral arterial disease in large vessels is epidemiologically distinct from small vessel disease. An-analysis of risk factors. Am J Euidemioi 1989: 129: 1110-l 119. Hiatt WR, Marihall JA, Baxter J et al. Diagnostic methods for peripheral arterial disease in the San Luis Valley Diabetes Study. J Clin Epidemiol 1990; 43: 597-606. Carter SA. Clinical measurement of systolic pressure in limbs with arterial occlusive disease. JAMA 1969; 207: 18691874. Newman AB, Sutton-Tyrrell K, Rutan GH et al. Lower extremity arterial disease in elderly subjects with systolic hypertension. J Clin Epidemiol 1991; 44: 15-20. Fowkes FGR. Review of simple measuring techniques. In: Fowkes FGR, Ed. Epidemiology of Peripheral Vascular Disease. London: Springer-Verlag; 1991: 3-15. Fowkes FGR, Housley E, Cawood EHH et al. Edinburgh artery study: prevalence of asymptomatic and symptomatic peripheral arterial disease in the general population. Int J Epidemiol 1991; 20: 384392. Cutler SI, Ederer F. Maximum utilization of the life table method in analysing survival. J Chron Dis 1958; 8: 699-712. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966; 50: 163-170. Kalbfleisch JD, Prentice RL. The Statistical Analysis of Failure Time Data. New York: John Wiley; 1980. Vogt MT, Cauley JA, Kuller LH, Hulley SB. Prevalence and correlates of lower extremity arterial disease in elderly women. Am J Epidemiol 1993; 137: 559-568. Cronenwett JL, Davis JT, Gooch JB, Garrett HE. Aortoiliac occlusive disease in women. Surgery 1980; 88: 775-784. Haimovici H. Patterns of arteriosclerotic lesions of the lower extremity. Arch Surg 1967; 95: 918-933.
42. 43.
44.
45.
46.
47.
48.
49.
50.
51. 52.
53.
54.
55.
56.
Rosen AI, DePalma RG, Yona V. Risk factors in peripheral atherosclerosis. Arch Surg 1973; 107: 303-308. Schroll M, Munck 0. Estimation of peripheral arteriosclerotic disease by ankle blood pressure measurements in a population study of 60-year old men and women. J Cl&m Dis 1981;-34 2611269. Gofin R. Kark JD. Friedlander Y et al. Peribheral vascular disease in a middle-aged population sample. The Jerusalem lipid research clinic prevalence study. Isr J Med Sci 1987; 23: 157-167. Gordon T, Kannel WB. Predisposition to atherosclerosis in the head, heart and legs. JAMA 1972; 221: 661-666. Kannel WB, Wolf PA, Verte J, McNamara PM. Epidemiologic assessment of the role of blood pressure in stroke: the Framingham Study. JAMA 1970; 214: 301-310. Fowkes FGR. Aetiology of peripheral atherosclerosis: smoking seems especially important. Br Med J 1989; 298: 405-406. Sutton KC, Wolfson SK, Kuller LH. Carotid and lower extremity arterial disease in elderly adults with isolated systolic hypertension. Stroke 1987; 18: 817-822. Wiseman SA, Powell JT, Barber N er al. Influence of apolipoproteins on the anatomical distribution of arterial disease. Atherosclerosis 1991; 89: 231-237. Fowkes GFR, Hausley E, Riemersma RA et al. Smoking lipids, glucose intolerance and blood pressure as risk factors for peripheral atherosclerosis compared with ischemic heart disease in the Edinburgh Artery Study. Am J Epidemiol 1992; 135: 331-340. Badimon JJ, Fuster V, Chesebro JH, Badimon L. Coronary atherosclerosis: a multifactorial disease. Circulation i993; 87(Suppl.I1): 11-3-11-16. Howell MA. Colaan _ MP. Seeaer RW er al. Relationship of severity of lower limb peripheral vascular disease to mortality and morbidity: A six year foliowup study. J Vast Surg 1989; 9: 691697. Kallero KS, Bergqvist D, Cederholm C et al. Late mortality and morbidity after arterial reconstruction: The influence of arteriosclerosis in popliteal artery trifurcation. J Vast Sure 1985: 2: 541-546. Martinez BD, Hertzer NR, BeVeu EG. Influence of distal arterial occlusive disease on prognosis following aortobifemoral bypass. Surgery 1980; 6: 7955805. Burnham SJ, Johnson GJ, Gurri JA. Mortality risk for survivors of vascular reconstructive procedures. Surgery 1982; 92: 1072-1076. Kuller LH. Are risk factors for CHD the same at different ages? J Clin Epidemiol 1989; 42: 91-93.
12674276, Printed in Great Britain. All rights reserved
1993 Copyright 0
0895-4356/93 $6.00 + 0.00 1993 Pergamon Press Ltd
SEGMENTAL ARTERIAL DISEASE IN THE LOWER EXTREMITIES: CORRELATES OF DISEASE AND RELATIONSHIP TO MORTALITY MOLLY T. VOGT,‘* SIDNEY K. WOLF-SON’and LEWIS H. KULLER’ ‘Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, PA 15261 and ‘Department of Surgery and Neurological Surgery, School of Medicine, University of Pittsburgh and Montefiore University Hospital, Pittsburgh, PA 15261, U.S.A. (Received in revised form 13 April 1993)
Abstract-The purpose of this study was to identify the correlates of disease and mortality risk associated with the presence of stenosis in the major arterial segments of the lower limb in men and women evaluated in a peripheral vascular laboratory. A total
of 326 men and 249 women had unisegmental stenosis; 312 men and 275 women had multisegmental disease; 132men and 111 women had no apparent disease. Multivariate analysis indicated that current smoking and elevated systolic blood pressure were the key risk factors associated with isolated aortoiliac and femoropopliteal arterial disease in both men and women. A history of diabetes was significantly associated with tibioperoneal disease in men while elevated systolic blood pressure was the major correlate of distal disease in women in this patient population. The relative risk of mortality was elevated 2- to 7-fold in men and women with multi- and unisegmental disease involving the aortoiliac and femoropopliteal segments; presence of tibioperoneal disease did not significantly increase mortality relative to those who were free of disease. These results suggest that the etiology and mortality risk associated with atherosclerosis in the lower extremity may vary with the anatomic site and/or severity of the lesion. Atherosclerosis
Mortality
Risk factors
Vascular
INTRODUCTION
Atherosclerosis is a systemic disease in which plaques develop primarily at branches in the arteries and in areas where the blood flow is slow or turbulent. The terminal aorta and its branches supplying the lower extremity are frequently involved [ 1,2]. The atherosclerotic process in the vessels of the lower extremities usually occurs in conjunction with that in the coronary and cerebral arteries [3,4]. Life expectancy is decreased by about 10 years in patients with lower extremity disease [5] and the majority of these patients die from cardiac disease or stroke [6-l 11. *Author for correspondence.
disease
Lower limb atherosclerosis usually develops at several levels within the arteries but may also be restricted to a single localized region in a vessel. When stenosis is found at multiple sites the blood flow to the lower limbs is often severely compromised and limb ischemia may become a significant clinical problem. Such multisegmental disease is also associated with poorer overall cardiovascular health status and higher morbidity and mortality than in localized unisegmental disease [13-l 51. Several studies have suggested that the etiology of the disease process in the lower extremity may vary depending on the anatomic site of the lesion [4, 13, 14, 16, 171. Aortoiliac disease seems to occur in younger patients who smoke, while the more distal lesions in the infrapopliteal
1267
1268
MOLLYT. VOGTet al.
arteries are found primarily in diabetics [13, 14, 16-191. The correlates of disease are similar for men and women, although women are reported to be 8-10 years older before clinical symptoms become evident [12,20]. The majority of studies conducted to date have involved relatively small numbers of patients [19], and have not used multivariate analysis to adjust for confounding factors [13, 14, 191. The purpose of this study is to identify the characteristics in men and women associated with the development of stenotic lesions in the major arterial segments of the lower limb (aortoiliac, femoropopliteal and the tibioperoneal vessels); and to determine the relationship between the anatomical location of the lesion and subsequent mortality.
METHODS
Patients evaluated for lower extremity arterial disease in the Montefiore University Hospital Peripheral Vascular Laboratory, Pittsburgh, PA between 1 January 1977, and 31 December 1989, were included in this study if they were 50 years of age or older and had no history of arterial surgery or amputation involving the lower extremity (n = 1930). Patients were referred to the laboratory for a variety of reasons including complaints of leg and hip pain at rest or during exercise, presence of non-healing ulcers, gangrene, inability of primary physician to palpate pedal pulses, history of diabetes, painful feet and for evaluation prior to arterial bypass surgery. The majority of patients were 50 years of age or older and since several previous reports [14,21-23]have characterized peripheral vascular disease as being “particularly virulent” in the younger age groups it was decided to limit our study population to those 50 years or older. The relevant medical history of each patient was taken by a trained technologist as described previously [lo]. Information was collected on use of medication, tobacco use and past history of claudication, diabetes, stroke, transient ischemic attack, amputations, vascular and nonvascular surgery, cardiac disease and respiratory disease. Ischemic heart disease was defined as a history of myocardial infarction and/or coronary artery bypass grafting. Angina was defined as angina in the absence of known ischemic heart disease. Arterial evaluation of the lower extremities was carried out according to standardized procedures developed at the Montefiore University
Hospital. Patients were evaluated in the supine position with the appropriately sized standard blood pressure cuffs placed on the upper arm, thigh, calf, and ankle. Systolic blood pressure was measured in both arms and then sequentially on both legs; first the right side (ankle, calf, thigh, arm) and then on the left side. A Doppler ultrasonic instrument was used to detect flow in the distal arteries. The cuff was inflated until the flow was no longer detectable and then slowly deflated. The pressure at which the flow first returned to the artery being monitored was recorded. The ankle arm index (AAI) for each leg was calculated by dividing the ankle pressure (the average of the pressures recorded for the posterior and anterior tibia1 arteries) by the higher of the recorded brachial pressures. The lower of the AA1 values obtained for the two legs was used as the measure of disease severity in the subsequent data analysis. In patients for whom it was impossible to obtain both the left and right AA1 (due to pain or ulceration/gangrene at ankle) the value obtained on the other limb was used for analysis. A segmental lesion was defined as present when the AA1 was ~0.90 and there was a pressure gradient of at least 20 mmHg between the systolic pressures measured at adjacent segmental levels [24-281, i.e. an aortoiliac lesion was recorded as present when the brachial pressure minus the thigh pressure was 2 20 mmHg; a femoropopliteal lesion was present when the thigh pressure minus the below-knee pressure was 220 mmHg; and a tibioperoneal lesion when below-knee pressure minus the average of the tibia1 pressures was 2 20 mmHg. The nolesion control group of patients were those with an AA1 > 0.90 but c 1.50 and no lesions. Patients with any missing segmental blood pressure measurements were excluded from the analyses. Patients were categorized by anatomical level of disease and by the number of major arterial segments involved, i.e. isolated disease involving either the aortoiliac, femoropopliteal or tibioperoneal segments; multisegmental disease involving either the aortoiliac-femoropopliteal, aortoiliac-tibioperoneal, femoropopliteal-tibioperoneal or aortoiliac-femoropopliteal-tibioperoneal arteries. A total of 326 men and 249 women had unisegmental disease; 3 12 men and 275 women had multisegmental disease; and 132 men and 111 women had no apparent stenotic disease.
Segmental Arterial Disease in the Lower Extremities
Criqui and his colleagues have reported that ankle-arm blood pressure ratios GO.85 have over a 90% sensitivity and 99% specificity in detecting arteriographically positive disease [29,30]. Other investigators [31,32] have found that the cutpoint for AA1 may be as high as 0.97 with little loss of sensitivity or specificity. The reproducibility of sensitivity and quality of segmental blood pressure measurement is high in subjects with and without clinical peripheral vascular disease [29]. We have also found very good reproducibility of ankle arm blood pressure measurements among older individuals with systolic hypertension [33]. Fowkes [34,35] has noted that the variability in measuring the ankle systolic blood pressure and ankle arm index is similar to the measure of arm systolic blood pressure and that the “variability of taking measurements on different occasions by different observers has only a marginal effect on between subject variability” [34]. Similarly, the use of systolic blood pressure gradients of 20 mmHg or more in conjunction with AA1 measurements, has been shown to be highly specific (over 90%) and reasonably sensitive (86%) for detecting isolated segmental arterial disease as compared to arteriography [27,28]. Ascertainment of death The total patient list was compared with the death files maintained by the Department of Vital Statistics, Commonwealth of Pennsylvania. The files for years 1977-1989 were searched manually by matching name, social security number, date of birth and county of residence for each patient. The underlying cause of death was recorded and subsequently coded by a physician (LHK) blind to the medical history and vascular diagnosis of the patients. A random sample of patients (N = 200) for whom death certificates were not found, were followed up to assess the completeness of the death ascertainment. Among patients without lower extremity arterial disease, 4% of deaths were missed or occurred out of state; 7% were missed for those with disease. Statistical analysis The chi-square statistic was used to assess the differences for categorical variables; t-tests were used for continuous variables. Two-tailed p values were used in all tests and the 5% level was considered significant. No correction was made for multiple comparisons.
1269
In the cross-sectional analysis of the data, logistic regression uni- and multivariate methods were used to identify the baseline characteristics associated with the presence of arterial disease. The lifetable actuarial method was used to assess the difference between survival of the controls and those with disease and the p-value was determined using the logrank test [36,37]. The Cox proportional hazards model was used to determine the relative risk of mortality and to adjust for covariates and for the varying follow-up times [38]. For both logistic regression and Cox proportional hazards models the dichotomous dependent variable was the presence of a lesion at a given site/absence of lesion at all sites studied. Dummy variables were used to code for smoking status (never, former, current) and heart disease (none, angina, definite ischemic heart disease). Other independent variables (systolic blood pressure > 140 mmHg, diabetes, stroke, congestive heart failure) were entered into the models as dichotomous variables. Patient records with missing values were excluded from the analyses.
RESULTS
After excluding patients with incomplete data, a total of 638 men and 524 women were found to have evidence of stenosis in at least one arterial segment based on our criteria for the definition of disease. A relatively small percentage of patients had lesions in all three arterial segments (2% of men, 5% of women) while approximately half the patients had unisegmental disease (Table 1). The men and women with aortoiliac disease were of similar average age to patients with no lesions; and they were more likely to be current smokers when they were first evaluated in the laboratory. In addition, women with disease were more likely to have a history of ischemic heart disease and dyspnea and to be former smokers. The men, on the other hand, had elevated systolic blood pressure and were more likely to have a history of chronic obstructive pulmonary disease. Both men and women with femoropopliteal lesions had high blood pressure and were more likely to have a history of stroke and to be current smokers. The women with disease were older than those with no lesions and presence of disease was associated with a history of ischemic heart disease and congestive heart failure.
MOLLYT. VOGTet al.
1270
Table 1. Distribution of lesions Men
Women
Aortoihac-femoropopliteal
and tibioperoneal lesion in one or both legs (AI-FP-TP)
14 (2%)
26 (5%)
Aortoiliac-femoropopliteal either leg (AI-FP)
lesions in one or both legs; no tibioperoneal lesions in
114 (18%)
115 (22%)
49 (8%)
58 (11%)
57 (9%)
40 (8%)
Isolated aortoiliac lesion in one or both legs; no femoropopliteal or tibioperoneal lesions in either leg (AI)
127 (20%)
66 (13%)
Isolated femoropopliteal lesion in one or both legs; no aortoihac or tibioperoneal lesions in either leg (FP)
134 (21%)
138 (26%)
Isolated tibioperoneal lesion in one or both legs; no aortoiliac or femoropopliteal lesions in either leg (TP)
65 (10%)
45 (9%)
638 (100%)
524 (100%)
Aortoiliac-tibioperoneal either leg (AI-TP)
lesions in one or both legs; no femoropopliteal lesions in
Femoropopliteal-tibioperoneal either leg (FP-TP)
lesion in one or both legs; no aortoihac lesions in
Total with a lesion* *Numbers do no sum to the total due to the exclusions.
Patients with tibioperoneal disease were older than the control group and more likely to have congestive heart failure. A significantly higher proportion of men with distal lesions were diabetic while the women were also more likely to have a history of stroke and ischemic heart disease (Tables 2 and 3). The occurrence of femoropopliteal and tibioperoneal disease in women increased sharply with age while the more proximal disease was weakly related to increasing age. In men, presence of aortoiliac disease was inversely
associated with age, the disease being most prevalent in the W-59 year age group; more distal lesions showed a weak positive association with age (Table 4). After age-adjustment the major characteristic associated with the presence of isolated aortoiliac and femoropopliteal disease in both men and women was smoking status, with current smokers at an approximately 3- to S-fold greater risk than those who reported having stopped smoking (Table 4). Increased systolic blood pressure was a major correlate of disease in men, while
Table 2. Comparison of baseline characteristics of male patients with isolated aortoiliac, femoropopliteal or tibiopcroneal lesions and with no lesions
Age (yr) AA1 SBpt (mmHg) DBPj (mmHg) % SBPt > 140 % DBP$ > 90 Heart rate (beatslmin) % White race % Angina % Ischemic heart disease % CHF % Diabetes % Stroke % Dyspnea % COPD % Ever smoked % Currently smokes % formerly smoked
No lesions n = 132
Aortoihac lesions only n = 127
65.7 f 9.7 l.OS+O.ll 134.9 f 19.9 78.0 f 10.2 34.1 9.3 72.2 f 12.4
64.2 f 8.6 0.59 * 0.132 145.4 + 24.3* 79.5 f 11.6 55.1* 14.3 73.9 + 10.5
67.5 f 8.8 0.57 + 0.16* 141.7 + 23.4* 75.8 + 10.6 50.0’ 6.6 72.7 k 11.0
69.1 f 9.7+ 0.81 + 0.21* 136.9 + 22.0 77.2 + 10.9 40.0 8.1 72.1 + 10.9
80.3 15.9 15.2
88.7 16.5 20.5
86.2 11.9 22.4
78. I 20.0 20.0
5.3 26.7 6.1 9.1 7.6 74.4 26.4
6.3 16.6 10.0 8.7 15.7* 94.5* 59.8’
6.7 27.8 17.1* 11.2 11.2 87.8* 47.3*
18.5* 43.8* 11.9 9.2 9.2 69.2 20.0
48.1
34.6,
40.4
49.2
‘p < .05 compared with male patients with no lesions. tSystolic blood pressure. tDiastolic blood pressure.
Femoropopliteal lesions only n = 134
Tibioperoneal lesions only n =65
Segmental Arterial Disease in the Lower Extremities
1271
Table 3. Comparison of baseline characteristics of female patients with isolated aortoiliac, femoropopliteal or tibioperoneal lesions and with no lesions
Age 64 AA1 SBPt (mmHg) DBPJ (mmHg) % SBPt > 140 % DBPj > 90 Heart rate (beats/min) % White race % Angina % Ischemic heart disease % CHF % Diabetes % Stroke % Dyspnea % COPD % Ever smoked % Currently smokes % Formerly smoked
No lesions n = Ill
Aortoiliac lesions only n=66
64.4 f 9.3 1.07 kO.11 137.0 k 20.6 75.8 k 9.2 37.8 5.6 73.0 + 8.7
66.5 + 8.3 0.56 f 0.13’ 139.5 + 22.8 72.6 rl: 7.8+ 48.5 0 69.7 k lO.4*
72.0 k l0.6* 0.57 & 0.18* 144.8 + 25.4’ 75.1 * 13.3 55.1; 8.9 74.35 II.0
70.6 f 9.0* 0.73 +O.l8* 145.3 k 21.6’ 76.7 k II.2 64.4; 6.8 73.6 k 10.4
80.2 16.2 3.6
78.8 25.8 l3.6*
69.9 23.2 10.9;
82.2 17.8 20.0*
2.7 24.3 3.3 7.2 2.7 46.7 23.4
7.6 19.7 9.7 l8.2* 4.5 95.52 63.6*
10.9* 28.7 12.8’ 10.9 5.1 63.2* 34.6*
l5.6* 33.3 17.1* 6.7 4.4 35.7 II.9
23.4
31.8*
28.6
23.8
Femoropopliteal lesions only n = 138
Tibioperoneal lesions only n =45
*p < 0.05 compared with female patients with no lesion. tsystolic blood pressure. SDiastolic blood pressure.
a history of stroke, ischemic heart disease and CHF were associated with increased risk in women. Multivariate logistic regression analysis indicated that current smoking remained the strongest independent correlate of disease in both sexes after correction for the other covariates. A history of diabetes was strongly associated with the presence of tibioperoneal lesions in men, while in women a systolic blood pressure greater than 140 mmHg was the most important factor (Table 4). Survival in men and women with uni- and multisegmental disease was significantly decreased compared to those with no segmental lesions (Table 5). After adjusting for age, the relative risk for all-cause mortality was approximately doubled in men with a single lesion at any segmental level and in those with multisegmental disease. After adjustment for elevated systolic blood pressure, smoking status, a history of congestive heart failure, stroke, diabetes, angina and ischemic heart disease at baseline, the relative risks remained the same except for an increase in risk of mortality for those with aortoiliac-femoropopliteal or with aortoiliacfemoropopliteal-tibioperoneal disease (Table 6). Mortality risks in women were elevated 4- to 5-fold when those with aortoiliac-femoropo-
pliteal and aortoiliac-tibioperoneal lesions were compared with women with no detectable lesions. As in men, isolated tibioperoneal disease was associated with the lowest risk of mortality. The mean ankle arm index of patients with multisegmental disease was lower than that for those with unisegmental disease; the indices for the patients with isolated tibioperoneal lesions were markedly higher than those for any other patient group with lesions (Table 5). In order to examine the effects of severity of occlusive disease as well as the site of the stenotic lesion, the analysis was repeated including only those patients with an ankle arm index of 0.7 or less. The relative risk of mortality in each group was basically unchanged except that the risk of mortality in women with isolated tibioperoneal disease increased from 1.5 to 3.1.
DISCUSSION
Smoking is the key factor associated with the presence of isolated aortoiliac and femoropopliteal disease in the men and women studied here. This association is lower in those who have stopped smoking but remains elevated compared to lifelong non-smokers. Women with a history of smoking are particularly susceptible to aortoiliac disease. Smoking was not related to
MOLLYT. VOGTet al.
1272
Table 4. Odds ratios for risk factors associated with presence of isolated arterial lesions in male and female natients. 95% confidence interval shown in parentheses Men OR*
Women OR*
ORt
ORt
Aortoiliac lesion
Age (per 1Oyr) (0.69.81.1) (0.69.81.2)
(l.Ei 3) 491’ (47.1, 511.0) 34.9 (5.34, 228.0)
(l.iii.2)
Current smoker Former smoker (1.0, 6.7) SBPI > 140 mmHg
(2.d.b.3) 2.6 (1.0, 6.7)
c3.0g.:3. 1) 2.6 (1.0, 7.1)
(19.2, 474.0) 20.2 (4.8, 85.2)
(1$.5)
(l.Z.4)
(0.2q.so.9)
(0.::
(0.;;:.6)
(l.$.l)
(l$.O)
(1.?88.8)
(2.4’t.3) i.5
(2.8?5.3) 2.6 (1.2, 5.6)
(l.65.:6.0) i.6 (0.2, 2.1)
(O.d,.83.5)
Diabetes (0$.6)
.O)
Femoropopliteal lesion
Age (per 10 yr) (2.03’046) 9.5. (3.7, 24.2)
Current smoker Former smoker (O.:;L) SBP$ > 140 mmHg
(l.d.93 5) i.7 ’ (0.9, 3.4)
(1 .G.9)
(O.!Yl 7) d.7. (0.3, 1.9)
(0.6,.:.8)
(1.:.:.3)
(l.$.O)
(0.39.82.3)
(0.$82.6)
(O&24.7)
(0.49.81.6)
(0.49.92s
(0.49.92.4)
(0.2q.P.9)
(l.t$.l)
(0.!91.9)
(1.22,.86.3)
(1.t34.6)
(l$.S)
(OE7)
(l.75.Y4.4) b.8 (0.3, 2.2)
(l.d.83 0) i.3 ’ (0.7, 2.3)
Diabetes
(1$9)
(O.6;:.“’ (1.3,‘4.3) 1.3 (0.7, 2.5)
(o.d;94.2)
Tibioperoneal lesion
Age (per 10yr) Current smoker Former smoker SBP# > 140 mmHg Diabetes
*Age included in all models. tAdjusted for all other covariates shown, as well as history of angina, ischemic heart disease, stroke and congestive heart failure. $Systolic blood pressure.
Table 5. Five and ten-year survival in patients with atherosclerotic lesions at various sites* n
AI-FP-TP lesions AI-FP AI-TP FP-TP Isolated AI Isolated FP Isolated TP No lesions AI-FP-TP lesions AI-FP AI-TP FP-TP Isolated AI Isolated FP Isolated TP No lesions
14
wt
114 (58) 49 (18) 57 (24) 127 (47) 134 (52) 65 (25) 132 (31) 26 (11)
115 (62) 58 (27) 40 (17) 66 (20) 138 (55) 45 (12) 111 (14)
AAI
5 yr survival
1Oyr survival
Men 0.41 f 0.15 0.46 f 0.14 0.44 + 0.14 0.52 + 0.13 0.59 f 0.13 0.57 f 0.16 0.81 k 0.21 1.08 kO.11
0.62 0.52 0.68 0.60 0.66 0.62 0.68 0.75
0.32 0.40 0.36 0.52 0.40 0.40 0.65
0.48 0.39 0.44 0.55 0.82 0.57 0.66
0.48 0.32 0.44 0.41 0.58 0.36 0.66
0.90
0.78
Women 0.44 + 0.14 0.44 f 0.12 0.44+0.15 0.50 f 0.12 0.56 + 0.13 0.57 f 0.13 0.73 f 0.18 1.07Io.11
*Sites of lesions defined in Table 1 tnumber of patients that died during follow up.
-
o-Value 0.0020 0.0000 0.0603 0.0073 0.0462 0.0021 0.0100 ref 0.0000 O.OfMO 0.0000 0.0000
0.0015 0.0000 0.0087 ref
1273
Segmental Arterial Disease in the Lower Extremities Table 6. Relative risk of all-cause mortality in patientswith atheroscleroticlesionsat varioussites*comparedto with no lesions, 95% confidence interval shown in Parentheses
patients
Women
Men Site of lesion
Relative riskt
Relative riski
Relative risk$ AA1 < 0.7
Relative riskt
Relative riski
Relative risk2 AA1 < 0.7
AI-FP-TP AI-FP AI-TP FP-TP Isolated AI Isolated FP Isolated TP
2.6 (1.2, 2.5 (1.6, 1.6 (0.9, 1.7 (1.0, 1.7 (1.1, 2.1 (1.3, 1.8 (1.1,
7.2 (2.2, 3.5 (2.0, 1.4 (0.7, 1.8 (0.9, 2.0 (1.1, 2.6 (1.5, 1.3 (0.7,
7.2 (2.2, 3.5 (2.1, 1.4 (0.7, 2.0 (0.9, 2.1 (1.2, 2.4 (1.3, 1.3 (0.6,
3.3 5.2 4.2 2.5 2.9 3.2 2.1
2.4 (0.6, 5.3 (2.3, 4.5 (1.6, 2.5 (0.9, 2.8 (1.2, 2.1 (1.0, 1.5 (0.4,
2.4 5.3 4.0 2.6 4.5 2.4 3.1
5.6) 3.8) 2.8) 2.9) 2.7) 3.3) 3.1)
23.1) 6.1) 2.9) 3.8) 3.5) 4.5) 2.5)
23.1) 7.3) 2.8) 4.1) 3.8) 4.3) 2.8)
(1.4, (2.8, (2.1, (1.1, (1.5, (1.7, (0.9,
7.9) 9.8) 8.5) 5.5) 5.7) 5.8) 4.8)
10.1) 12.2) 12.8) 7.6) 6.7) 4.7) 5.2)
(0.6, (2.3, (1.4, (0.9, (1.6, (1.1, (0.9,
10.1) 12.3) 11.5) 7.6) 12.8) 5.2) 11.0)
*Sites of lesions defined in Table 1. TAdjusted for age. $Adjusted for age, systolic blood pressure, current and former smoking and a history of angina, ischemic heart disease, diabetes, stroke and congestive heart failure.
tibioperoneal disease in our patients and previous clinical observations [19] tend to support this finding. There have been no population-based studies to date which have considered etiologic factors in lower extremity arterial disease in relation to the anatomic site of the arterial occlusion. Calculations of the population attributable risk for smoking and lower extremity arterial disease, considering lesions at all levels, has varied from 14% to 78% [7] depending, in part, on the definition of disease and population characteristics. A study of normal and hyperlipidemic subjects in California [30] showed that smoking is a risk factor for large vessel lower extremity arterial disease in men but not in women. A more recent study of healthy elderly women found that current smoking increased the odds of disease 6-fold [39]. In addition, several clinical studies which focused on the assessment of patients for reconstructive vascular surgery of the lower limb, have reported that current smoking is strongly associated with both uni- and multisegmental disease involving aortoiliac and femoropopliteal lesions [40,41]. Multivariate analytic techniques were not used for data analysis in these studies and the effect of comorbid conditions and other risk factors were not considered. Our research indicates that elevated systolic blood pressure is an independent correlate of disease in each of the three arterial segments studied in women and of aortoiliac and femoropopliteal disease in men. Most studies, to date, have shown that elevated systolic blood pressure is associated with lower extremity arterial disease [3,39,42,43] although Gofin [44] and Criqui [30] did not find it to be a significant factor in women. None of these studies considered the relationship of elevated blood press-
ure to lesions at various levels in the arterial system. Diabetes is usually considered to be an important factor for development of lower extremity arterial disease with the occlusive process affecting the most distal vessels, primarily the posterior tibia1 and peroneal arteries [9, 13, 191. In this study, diabetes was a significant independent correlate of tibioperoneal disease in men but not women. Since diabetes is usually associated with diffuse and generalized atherosclerosis, many diabetics would be excluded from the study of risk factors for unisegmental disease. The low statistical power associated with the relatively small number of patients may explain our inability to show an association between diabetes and distal lesions in women (with 45 women having tibioperoneal disease, there is approximately 35% power to detect an odds ratio of 2.0). It is noteworthy that the percentage of patients with diabetes increases as the anatomic site of the lesion studied becomes more distal; this is true in both men and women but reaches statistical significance only in men. In addition, the diabetic patients may have a high degree of arterial calcification in the tibia1 arteries. Blood pressure measurements in rigid calcified arteries give artifactually high values, resulting in the classification of these patients as free of disease. We have attempted to minimize such misclassification of patients by excluding all patients with unusually high values of the ankle-arm index (greater than 1.50). Research studies to date have found that although the risk factors for atherosclerosis are similar in all vascular beds, the dominant risk factor seems to vary by anatomic location [41,42,45-511. For instance, elevated systolic blood pressure is the key predictor for stroke [45,46] while current smoking is a major risk
1274
MOLLY
T. VOGT et al.
factor for lower extremity arterial disease [47,50]. In addition, a number of clinical studies have suggested that even within the major arteries of the leg, various segments may react differently to stressors such as smoking and diabetes [13, 14, 16-191. The mechanism for such anatomically differential responses is unknown [19] although it may be related to arterial diameter, the histologic structure of the vessel wall, the turbulence of the blood flow through the various sections of the artery and/or specific biologic influence of the various risk factors on the atherosclerotic process [51]. Measurement of segmental limb blood pressures is a highly specific method of identifying isolated segmental disease but the sensitivity is considerably lower and varies depending on the anatomical location of disease [27,28]. Utilizing both the ankle-arm index and the segmental pressure gradients as in this study, the overall sensitivity is 86% for isolated segmental disease as compared to detection of disease by arteriography [271. The major sources of error noted by Rutherford [27] are errors in measurement of blood pressures due to arterial calcification as mentioned above; inability to detect stenoses that do not cause a change in pressure gradient at rest; inability to detect tibioperoneal disease that does not involve all three infrapopliteal arterial branches; and correct localization of stenoses that are under the blood pressure cuff. All of these sources of error would tend to bias our results toward an underestimation of effect size. Atherosclerosis in the lower extremity is associated with increased risk of mortality [7-l 11. The more severe the disease and the more segmental levels involved the greater the risk [7, l&13,52]. The location of the lesion did not affect mortality in some studies [15] but in others, distal disease was more indicative of early mortality [52-551. Our results indicate that crude survival is decreased in patients with lesions at every segmental level. After adjusting for other covariates, the all-cause mortality risk is significantly elevated in isolated aortoiliac and femoropopliteal disease and in aortoiliacfemoropopliteal-tibioperoneal, aortoiliac-femoropopliteal disease in both sexes and in aortoiliac-tibioperoneal disease in women. The lowest risk of dying was found in men and women with tibioperoneal disease and in men with femoropopliteal-tibioperoneal disease. The patients with unisegmental distal disease had ankle arm indices that were higher than
those with lesions in other locations, i.e. overall the arterial disease was less severe in these patients. After stratifying by disease severity (based on the ankle arm index) the women with tibioperoneal disease and an ankle arm index of 0.7 or less had an increased risk of mortality; no change in risk was found in the men. The differential mortality by site of lesion may therefore be related to disease severity rather than the site of the lesion. However, the prevalence of congestive heart failure is several fold higher among patients with distal disease compared to those with no lesions; although we have adjusted for this fact in our multivariate analysis, residual confounding may be affecting the results obtained. Kallero [53] and Martinez [54] have both reported that distal disease is associated with excess deaths due to myocardial disease. However, their patients had all undergone prior arterial reconstruction of proximal vessels and thus originally had multisegmental disease. In addition, the Ranch0 Bernard0 study of normal and hyperlipidemic men and women showed that posterior tibia1 arterial disease was associated with increased all cause and coronary heart disease mortality [1 11. Additional research is needed to resolve these apparent discrepancies. The results of this study must be interpreted with caution. Firstly, as noted earlier in this section, studies on patients referred to a particular clinical site are likely to be biased and the results may not be generalizable to other patient populations. Second, the controls used in the analyses are also patients and in poorer general health than men and women of similar age in the general population. Third, it is possible that some of patient deaths may have been missed, especially if they occurred outside Pennsylvania. Our detailed follow up study of 200 patients believed alive as of 31 December 1989, (see Methodology) indicated that a relatively small percentage of deaths had been missed and that underascertainment of death may have been more of a problem in patients with a low ankle-arm index than in those with no lesions; this would result in an underestimation of the relative risk of dying. Fourth, the patient records contained no information on the serum lipid levels. Most research to date has not found lipids to be a major risk factor for the development of arterial disease in the lower extremities [12, 30,44, 501. Similarly, the relationship between cholesterol levels and mortality in older populations seems to be much weaker than in
Segmental Arterial Disease in the Lower Extremities
middle-aged men and women [56]. Fifth, the small numbers in some of the patient groups studied means that the power to detect statistically significant associations is low. Finally, the models obtained from the data analysis have not been validated in a test population. Our patient population, although larger than that used in most other clinical studies, is not of sufficient size to split into two groups to permit model development and then model testing within the same population. In spite of these limitations, the results presented show that current smoking and high systolic blood pressure are the major independent factors related to the development of aortoiliac and femoropopliteal disease in this patient population. Lesions in these two arterial segments are associated with increased mortality. A history of diabetes is an independent correlate of more distal unisegmental disease in men but the presence of infrapopliteal lesions does not appear to significantly affect the risk of dying. The 5 year probability of dying for women with multisegmental disease is about 50% (Table 5) as compared to 2040% in those with unisegmental disease and 10% in women with no detectable arterial lesions. The trend is similar in men but much less pronounced. After adjusting for a variety of confounders (Table 6) these large increases in mortality related to severity and/or location of arterial lesion remain. Thus it is of considerable clinical interest to continue research to further delineate the role of smoking, hypertension, diabetes and impaired glucose tolerance as correlates of disease development in each of the major arterial segments of the lower extremity and to clarify the relationship between the site of the lesion, degree of occlusion, involvement of multiple arterial segments and subsequent morbidity and mortality. It will also be important to assess the effectiveness and efficacy of pharmacologic, dietary and lifestyle interventions on the development and course of disease, not only in clinical but also population-based studies. REFERENCES DeBakey ME, Lawrie GM, Glaeser DH. Patterns of atherosclerosis and their surgical significance. Ann Surg 1985; 201: 115-121. Ross R. The pathogenesis of atherosclerosis-an update. N Engl J Med 1986; 314: 488-500. Kannell WB, McGee DL. Update on some epidemiologic features of intermittent claudication: The Framingham study. J Am Cerirtr Sot 1985; 33: 13-18.
1275
4. Hertzer NR. The natural history of peripheral vascular disease: implications for its management. Circulation 1991; 83(Suppl.l):
112-19.
5.
Malone JM, Moore WS, Goldstone J. Life expectancy following aortofemoral arterial grafting. Surgery 1977; 81: 551-555. 6. Jelnes R, Gaardsting 0, Hougard Jensen K et al. Fate in intermittent claudication: outcome and risk factors. Br Mad J 1986; 293: 1137-1140. 7. Dormandy J, Mahir M, Ascady G et al. Fate of the patient with chronic leg ischaemia. J Cardiovasc Surg 1989; 30: 50-57.
8.
9.
10.
11.
12.
Kallero KS. Mortality and morbidity in patients with intermittent claudication as defined by venous occlusion plethysmography. A ten year f&ow-up study. J Chroo Dis 1981; 34: 455-462. Cronenwett JL, warner KG, Zelenock GB ef al. Intermittent claudication. Current results of nonoperative management. Arch Surg 1984; 119: 430-436. McKenna M, Wolfson S, Kuller L. The ratio of ankle and arm arterial pressure as an independent predictor of mortality. Atherosclerosis 1991; 87: 119-128. Criqui MH, Langer RD, Fronek A ef al. Mortality over a period of 10 years in patients with peripheral arterial-disease. N Eigl J Med 1992; 326: %-386. Voet MT. Wolfson SK. Kuller LH. Lower extremitv art&al disease and the aging process: a review. J Glib Epidemiol 1992; 45: 529-542.
13. Samson RH, Scher LA, Veith FJ. Comorbid segmental arterial disease. Surgery 1985; 97: 385-396. 14. McDaniel MD, Cronenwett JL. Basic data related to the natural history of intermittent claudication. Ann Vast Surg 1989; 3: 273-277. 15. Jonason T, Rinquist I. Mortality and morbidity in patients with intermittent claudication in relation to the location of the occlusive atherosclerosis in the leg. Angiology
1985; 36: 310-314.
16. Strandness DE, Priest RE, Gibbons GE. Combined clinical and pathological study of diabetic and nondiabetic peripheral arterial disease. Diabetes 1964; 13: 36G372.
17. Conrad MC. Large and small artery occlusion in diabetes with severe vascular disease. Circulation 1967; 36: 83-91.
18. Bild DE, Selby JV, Sinnock P, Browner WS, Braveman P, Showstack JA. Lower-extremity amputation in people with diabetes. Diabetes Care 1989; 12: 24-31.
19. Menzoian JO, LaMorte WW, Paniszyn CC et al. Symptomatology and anatomic patterns of peripheral vascular disease: differing impact of smoking and diabetes. Ann Vast Surg 1989; 3: 224228. 20. A. Widmer LK. Occlusive Peripheral Artery Disease. The Basle Study. Huber: Bern; 1980. 21. Evans WE, Hayes JP, Vermillion BD. Atherosclerosis in the younger patient. Am J Surg 1987; 154: 225-229. 22. Nunn DB. Symptomatic peripheral arteriosclerosis of patients under age 40. Am J Surg 1973; 39: 224-228. 23. McCready RA, Vincent AE, Schwartz RW et al. Atherosclerosis in the young: a virulent disease. Surgery 1984; 96: 863-868. 24. Carter SA. Role of pressure measurements in vascular disease. In: Bernstein EF, Ed. Noninvasive Diagnostic Techniuues in Vascular Disease. St Louis, MO: C.V. Mosby-Co; 1985: 513-544. 25. Raines J. Traad E. Noninvasive evaluation of oeriphera1 vascular disease. Med Clin N Am 1980: 64(2): 283-304. 26. Gundersen J. Segmental measurements of systolic blood pressure in the extremities including the thumb and the great toe. Acta Chir Stand 1972; 426: I-10.
1276 27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38. 39.
40.
41.
MOLLY T. VOG~et al. Rutherford RB, Lowenstein DH, Klein MF. Combined segmental systolic pressures and plethysmography to diagnose arterial occlusive disease of the legs. Am J Surg 1979; 138: 211-218. Barnes RW. Noninvasive diagnostic assessment of peripheral vascular disease. Circulation 1991; 83(Suppl.I): 120-127. Criqui MH, Langer RD, Fronek A, Feigelson HS, Klauber MR. Large vessel and isolated small vessel disease. In: Fowkes FGR, Ed. Epidemiology of Peripherai Vascular Disease. London: Springer-Verlag; 1991: 85-96. Criqui MH, Browner D, Fronek A er al. Peripheral arterial disease in large vessels is epidemiologically distinct from small vessel disease. An-analysis of risk factors. Am J Euidemioi 1989: 129: 1110-l 119. Hiatt WR, Marihall JA, Baxter J et al. Diagnostic methods for peripheral arterial disease in the San Luis Valley Diabetes Study. J Clin Epidemiol 1990; 43: 597-606. Carter SA. Clinical measurement of systolic pressure in limbs with arterial occlusive disease. JAMA 1969; 207: 18691874. Newman AB, Sutton-Tyrrell K, Rutan GH et al. Lower extremity arterial disease in elderly subjects with systolic hypertension. J Clin Epidemiol 1991; 44: 15-20. Fowkes FGR. Review of simple measuring techniques. In: Fowkes FGR, Ed. Epidemiology of Peripheral Vascular Disease. London: Springer-Verlag; 1991: 3-15. Fowkes FGR, Housley E, Cawood EHH et al. Edinburgh artery study: prevalence of asymptomatic and symptomatic peripheral arterial disease in the general population. Int J Epidemiol 1991; 20: 384392. Cutler SI, Ederer F. Maximum utilization of the life table method in analysing survival. J Chron Dis 1958; 8: 699-712. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966; 50: 163-170. Kalbfleisch JD, Prentice RL. The Statistical Analysis of Failure Time Data. New York: John Wiley; 1980. Vogt MT, Cauley JA, Kuller LH, Hulley SB. Prevalence and correlates of lower extremity arterial disease in elderly women. Am J Epidemiol 1993; 137: 559-568. Cronenwett JL, Davis JT, Gooch JB, Garrett HE. Aortoiliac occlusive disease in women. Surgery 1980; 88: 775-784. Haimovici H. Patterns of arteriosclerotic lesions of the lower extremity. Arch Surg 1967; 95: 918-933.
42. 43.
44.
45.
46.
47.
48.
49.
50.
51. 52.
53.
54.
55.
56.
Rosen AI, DePalma RG, Yona V. Risk factors in peripheral atherosclerosis. Arch Surg 1973; 107: 303-308. Schroll M, Munck 0. Estimation of peripheral arteriosclerotic disease by ankle blood pressure measurements in a population study of 60-year old men and women. J Cl&m Dis 1981;-34 2611269. Gofin R. Kark JD. Friedlander Y et al. Peribheral vascular disease in a middle-aged population sample. The Jerusalem lipid research clinic prevalence study. Isr J Med Sci 1987; 23: 157-167. Gordon T, Kannel WB. Predisposition to atherosclerosis in the head, heart and legs. JAMA 1972; 221: 661-666. Kannel WB, Wolf PA, Verte J, McNamara PM. Epidemiologic assessment of the role of blood pressure in stroke: the Framingham Study. JAMA 1970; 214: 301-310. Fowkes FGR. Aetiology of peripheral atherosclerosis: smoking seems especially important. Br Med J 1989; 298: 405-406. Sutton KC, Wolfson SK, Kuller LH. Carotid and lower extremity arterial disease in elderly adults with isolated systolic hypertension. Stroke 1987; 18: 817-822. Wiseman SA, Powell JT, Barber N er al. Influence of apolipoproteins on the anatomical distribution of arterial disease. Atherosclerosis 1991; 89: 231-237. Fowkes GFR, Hausley E, Riemersma RA et al. Smoking lipids, glucose intolerance and blood pressure as risk factors for peripheral atherosclerosis compared with ischemic heart disease in the Edinburgh Artery Study. Am J Epidemiol 1992; 135: 331-340. Badimon JJ, Fuster V, Chesebro JH, Badimon L. Coronary atherosclerosis: a multifactorial disease. Circulation i993; 87(Suppl.I1): 11-3-11-16. Howell MA. Colaan _ MP. Seeaer RW er al. Relationship of severity of lower limb peripheral vascular disease to mortality and morbidity: A six year foliowup study. J Vast Surg 1989; 9: 691697. Kallero KS, Bergqvist D, Cederholm C et al. Late mortality and morbidity after arterial reconstruction: The influence of arteriosclerosis in popliteal artery trifurcation. J Vast Sure 1985: 2: 541-546. Martinez BD, Hertzer NR, BeVeu EG. Influence of distal arterial occlusive disease on prognosis following aortobifemoral bypass. Surgery 1980; 6: 7955805. Burnham SJ, Johnson GJ, Gurri JA. Mortality risk for survivors of vascular reconstructive procedures. Surgery 1982; 92: 1072-1076. Kuller LH. Are risk factors for CHD the same at different ages? J Clin Epidemiol 1989; 42: 91-93.