Predictive power of cardiovascular risk factors for detecting peripheral vascular disease.

Objectives: Peripheral vascular disease (PVD) is underdiagnosed in primary care due to the absence of established criteria to warrant diagnostic testing. Our goal was to establish a risk factor hierarchy to enable earlier diagnosis of PVD.

Methods: Data sets of 142 patients with abnormal ankle brachial indices (ABI) were randomly selected from our patient database to determine the prevalence of specific cardiovascular risk factors and demographic data. An ABI score <0.90 is diagnostic of PVD. Patients were stratified into mild (0.75-0.89), moderate (0.50-0.74), and severe disease (<0.50).

Results: Mean age was 69 [+ or -] 9.9, ABI 0.65 [+ or -] 0.16. Risk factor prevalence: diabetes, 42%; hypertension, 87%; tobacco use, 34%; hyperlipidemia, 53%; obesity, 24%; cardiovascular disease (CVD), 69%; stroke, 15%. Total risk factors per patient v = 3.2 [+ or -] 1.3. Disease severity stratifications: mild, n = 46 (age v = 68.6 [+ or -] 10.4, ABI v = 0.82 [+ or -] 0.05); moderate, n = 72 (age v = 69.9 [+ or -] 9.4, ABI v = 0.62 [+ or -] 0.07); severe, n = 24 (age v = 67.5 [+ or -] 10.9, ABI v = 0.40 [+ or -] 0.06). Independent variable mean differences: hypertension-CVD (P = 0.0002); CVD-hyperlipidemia (P = 0.002); hyperlipidemia-diabetes (P = 0.0008); diabetes-tobacco use (P = 0.001); tobacco use-obesity (P = 0.0003); obesity-stroke (P = 0.05). Independent variable mean differences were significant across disease severity (P < 0.00001).

Conclusions: Our study establishes the following hierarchy of cardiovascular risk factors as predictors of PVD: hypertension, cardiovascular disease, hyperlipidemia, diabetes, tobacco use, obesity, stroke.

Key Words: ankle brachial indices, peripheral vascular disease, risk factors

**********

Peripheral vascular disease is a highly prevalent condition that affects approximately 12 million persons in the United States, and has a strong link to increased morbidity and mortality rates. (1-4) A recent study demonstrated that atherosclerotic risk factors are very prevalent in patients with peripheral vascular disease, but these patients are treated less intensively with lipid-lowering agents, antiplatelet therapy, or hypertensive drugs. (5) Risk factors, which increase the risk for development of peripheral vascular disease, include hypertension, hyperlipidemia, tobacco use, diabetes, cardiovascular disease, obesity, and stroke. (6) Over the last fifty years, the incidence and prevalence of peripheral vascular disease have climbed steadily in the majority of westernized countries. It is estimated that up to 20 million people in the United States and Europe combined have some evidence of peripheral vascular disease. Although peripheral vascular disease affects a large segment of the adult population, with an age-adjusted prevalence of 12 to 20%, it is severely underdiagnosed in primary care practices, and undertreated in terms of risk factor modification. (7,8) The aim of this analysis was to develop a hierarchy of cardiovascular risk factors to enable early diagnosis of peripheral vascular disease.

Materials and Methods

Patients

Data from 142 patients (male = 65) from within our practice who were diagnosed with peripheral vascular disease between January 2002 and January 2004 were randomly selected from the patient database at the South Carolina Heart Center, Columbia, SC, for this cross-sectional, retrospective data analysis. Peripheral vascular disease was defined as having ankle brachial indices <0.90. This mode of selection eliminated reactivity issues, attrition, and compensatory factors, as study subjects had no knowledge of their participation at the time of data collection. Inclusion criteria for participants included having both a diagnosis of peripheral vascular disease and complete data records available as required for our analysis.

Ankle brachial indices

A cross sectional, retrospective data analysis including medical records review was conducted to establish data points for this study. Patient data were stratified into three tiers of severity based on ankle brachial indices values: mild disease (0.75-0.89), moderate disease (0.50-0.74), and severe disease (<0.50). The stratifications have been well established as valid cut points correlated to disease progression. (9)

All study patients had been previously tested to determine ankle brachial indices score. An 8-mHz Doppler device (Versa Lab LE, Nicolet Vascular Inc., Golden, CO) was used to measure each of the ankle brachial indices. The unit utilizes a broad-beam ultrasound probe designed specifically to facilitate ankle brachial indices measurement. Ankle brachial indices scores were obtained with the subject in the supine position by recording the systolic blood pressures in the upper extremities at the brachial arteries and in the lower extremities at the dorsalis pedis and posterior tibial arteries. The ankle brachial indices for each leg were calculated by dividing the greater of the two ankle systolic readings in that leg by the greater brachial pressure. Subjects were considered to have peripheral vascular disease if the ankle brachial indices score was 0.90 or less. The sensitivity of the ankle brachial indices testing is 90%, and the specificity is 98% for an angiographically defined stenosis of 50% or more in a major leg artery. (10-12)

Cardiovascular risk factors

Hypertension was designated if the patient had a systolic blood pressure of 140 mm Hg or above, a diastolic blood pressure of 90 mm Hg or above regularly within office reports, or were prescribed antihypertensive medications. Known cardiovascular disease was established from a review of the patient medical history. Hyperlipidemia was determined from the medical record as a total cholesterol concentration greater than 200 mg/dL, or past or current use of lipid-lowering agents. Presence of diabetes, both type 1 and type 2, tobacco use past or current, obesity, and prior stroke were determined from the individual medical records.

Statistical analysis

Study population data were examined using descriptive statistics, analysis of variance (ANOVA), and paired T-testing allowing for a comprehensive, detailed analysis of the variables among study patients. Statistical significance ([alpha] = 0.05) was assessed to determine if significant variance was observed in the study population as a whole, and stratified by disease severity for each of the measured variables. In addition, paired T-testing, ANOVA, and Bonferroni mean difference procedures were conducted on data stratified by sex to allow for consideration of predictive influence over risk factors observed. Both an analysis of variance procedure and Bonferroni mean differences test were used to determine association between the prevalence of the independent variables (risk factors). Statistical analyses were performed using the Statistix, version 7 (Analytical Software Corp., Tallahassee, FL) computer statistical application within the South Carolina Heart Center, Department of Investigation Initiated Research. Random heterogeneity of the subjects providing statistical data for this study was inherent, as they were not selected or recruited by the study investigator, but rather enrolled after the study data had been recorded, thus controlling for threats to internal validity.

Results

Complete data sets from 142 patients (male = 65) who had an abnormal ankle brachial indices tests at our practice between January 2002 and January 2004 were analyzed during this study. Study population data is summarized in Table 1. Population mean data was age 69 [+ or -] 9.9 years (range 43-91) and ankle brachial indices 0.65 [+ or -] 0.16 (range 0.21-0.89). The following cardiovascular risk factor prevalence rates were found in the study population as a whole: diabetes, n=59 (42%); hypertension n=124 (87%); tobacco use n=48 (34%); hyperlipidemia n=75 (53%); obesity n=34 (24%); cardiovascular disease n=98 (69%); stroke n=21 (15%).

There were 7 measured cardiovascular risk factors, and the mean number of factors observed per patient were v = 3.2 [+ or -] 1.3 (min = 0, max = 6).

Disease severity stratifications data (Table 2) are: mild disease n = 46 (age v = 68.6 [+ or -] 10.4, ABI v = 0.82 [+ or -] 0.05), moderate disease n = 72 (age v = 69.9 [+ or -] 9.4, ABI v = 0.62 [+ or -] 0.07), severe disease n = 24 (age v = 67.5 [+ or -] 10.9, ABI v = 0.40 [+ or -] 0.06).

There were 65 male (46%) and 77 female (54%) subjects with complete data sets during the study time parameters. The female cohort of this study evidenced statistically similar total risk factor prevalence when compared with the male cohort. Females were older in all three disease severity stratifications, and were represented by over 2:1 among those stratified with severe disease. Ankle brachial indices mean scores were remarkably similar between the sexes. Sex was analyzed against the covariables to determine any predictive influence. Hypertension was statistically more prevalent among females than males (P = 0.003). Both tobacco use and a prior diagnosis of cardiovascular disease were more prevalent among males (P = 0.02, P = 0.04).

Independent variable mean testing for significant differences (Table 3) were: hypertension-CVD (P = 0.0002); CVD-hyperlipidemia (P = 0.002); hyperlipidemia-diabetes (P = 0.0008); diabetes-tobacco use (P = 0.001); tobacco use-obesity (P = 0.0003); obesity-stroke (P = 0.05).

ANOVA results confirm that independent variable mean differences were significantly different from one another across all three disease severity classifications (mild, P < 0.00001; moderate, P < 0.00001; severe, P < 0.00001).

Discussion

Clinical Implications

Peripheral vascular disease is a disease of lifestyle with highly modifiable risks, yet the prevalence of the disease is widely misunderstood and early diagnosis often missed. Historically, tobacco use has been viewed as the primary modifiable risk factor among peripheral vascular disease patients. (13, 14) In addition, there has been a prevailing notion that peripheral vascular disease is a disease more prominent in males. (8, 15) Our retrospective analysis of 142 patients showed peripheral vascular disease prevalence among both sexes was equal, with a statistically similar risk factor prevalence. This study also provides a hierarchy of traditional cardiovascular risk factors, which is not currently available, and which can be used in primary practice to detect the presence of peripheral vascular disease. The restructuring of the prevalent risk factors and their importance as a prognostic tool are valuable for improving early detection of peripheral vascular disease. This risk factor hierarchy, in conjunction with the presence of peripheral vascular disease symptoms, should establish a rationale for ankle brachial indices testing. Ankle brachial indices testing is an inexpensive yet highly reliable method of determining peripheral vascular disease, and those at substantial risk as a result of these findings warrant testing. In addition, the results of this study are clinically useful to the cardiologist and rehabilitation practitioner in structuring a more aggressive treatment regimen to reduce risk factors that were previously underappreciated as indicators of peripheral vascular disease prevalence and severity, such as hypertension, cardiovascular disease, and hyperlipidemia.

Study Limitations

The principal goal of this retrospective study was to determine the prevalence of traditional cardiovascular risk factors among those patients within our practice that have been diagnosed with peripheral vascular disease. While adherence to established clinical criteria were employed to define each of the risk factors, we were not able to validate each of the criteria via biochemical markers for each condition. It would, logistically, be impractical to obtain fasting glucose and lipid panels in addition to other clinical verification measures in a retrospective analysis. As a result, the prevalence of cardiovascular risk factors is likely to be underestimated in some instances.

Conclusion

Our retrospective analysis supports the contention that peripheral vascular disease affects both sexes equally when similar risk factors are observed. These results also establish a cardiovascular risk factor hierarchy that can be used in primary practice to enable earlier diagnosis of peripheral vascular disease. Finally, this list of modifiable risk factors can be used in prevention efforts and treatment of peripheral vascular disease to promote a more aggressive approach to the reduction or elimination of these risk indicators, with the aim of delaying disease progression and assisting with treatment efforts.

A witty saying proves nothing.
--Voltaire

Table 1. Study population data (a)

Variable        No.  %   Mean   SD    Minimum  Maximum

ABI             142       0.65  0.16    0.21     0.89
Age             142      69.1   9.9    43       91
Diabetes         59  42
Hypertension    124  87
Tobacco use      48  34
Hyperlipidemia   75  53
Obesity          34  24
CVD              98  69
Stroke           21  15
Total Risk      142       3.23  1.3     0        6
  Factors

(a) ABI, ankle brachial indices; CVD, cardiovascular disease.

Table 2. Stratified disease severity data (a)

                                        Mean         Mean total
Severity  Gender  No.  Mean age  SD     ABI    SD    risk factors  SD

Mild      Male    19     66.4    10.3   0.83   0.04   3.11         1.1
Mild      Female  27     70.1    10.4   0.81   0.05   2.96         1.3
Moderate  Male    39     68.7    10.3   0.61   0.07   3.08         1.2
Moderate  Female  33     71.4     8.3   0.63   0.07   3.52         1.3
Severe    Male     7     65.0    10.5   0.40   0.09   4.14         1.2
Severe    Female  17     68.5    11.3   0.39   0.05   3.24         1.2

(a) ABI, ankle brachial indices.

Table 3. Risk factor mean differences

Variable        Mean   SD     P

Hypertension    1.87  0.33  0.0002
CVD             1.69  0.46  0.002
Hyperlipidemia  1.53  0.5   0.0008
Diabetes        1.42  0.49  0.001
Tobacco use     1.34  0.47  0.0003
Obesity         1.24  0.43  0.05
Stroke          1.15  0.36

Accepted July 11, 2004.

References

1. Criqui MH, Fronek A, Barrett-Connor E, et al. The prevalence of peripheral arterial disease in a defined population. Circulation 1985;71:510-515.

2. Centers for Disease Control and Prevention (CDC): National vital statistics reports. 1999;49(8).

3. Criqui MH, Denenberg JO, Langer RD, et al. The epidemiology of peripheral arterial disease: importance of identifying the population at risk. Vasc Med 1997;2:221-226.

4. Meijer WT, Hoes AW, Rutgers D, et al. Peripheral arterial disease in the elderly: the Rotterdam Study. Arterioscler Thromb Vasc Biol 1998;18:185-192.

5. Mukherjee D, Lingam P, Chetcuti S, et al: Missed opportunities to treat atherosclerosis in patients undergoing peripheral vascular interventions: insights from the University of Michigan peripheral vascular disease quality improvement initiative (PVD-Q[I.sup.2]). Circulation 2002;106:1909-1912.

6. Stamler J, Stemler R, Neaton JD. Blood pressure, systolic and diastolic, and cardiovascular risks. Arch Intern Med 1993;153:598-615.

7. Hiatt WR, Marshall JA, Baxter J. Diagnostic methods for peripheral arterial disease in the San Luis Valley Diabetes Study. J Clin Epidemiol 1990;43:597-606.

8. Hirsch AT, Criqui MH, Treat-Jacobson D, et al. Peripheral arterial disease detection, awareness, and treatment in primary care. JAMA 2001;286:1317-1324.

9. Daigle RJ. Techniques in Noninvasive Vascular Diagnostics: Protocol and Procedures Guideline Manual. Rolling Hills, CA, Academy Medical Systems, 1999, p 134.

10. Ouriel K, McDonnell AE, Metz CE, et al. Critical evaluation of stress testing in the diagnosis of peripheral vascular disease. Surgery 1982;91:686-693.

11. Yao ST, Hobbs JT, Irvine WT. Ankle systolic pressure measurements in arterial disease affecting the lower extremities. Br J Surg 1969;56:676-679.

12. Criqui MH, Denenberg JO, Bird CE, et al. The correlation between symptoms and non-invasive test results in patients referred for peripheral arterial testing. Vasc Med 1996;1:65-71.

13. Hirsch AT, Treat-Jacobson D, Lando HA, et al. The role of tobacco cessation, and lipid lowering therapies for the treatment of peripheral arterial disease. Vasc Med 1997;2:243-251.

14. Hughson WG, Mann JI, Garrod A. Intermittent claudication: prevalence and risk factors. BMJ 1978;1:1379-1381.

15. Aronow WS, Ahn C. Prevalence of coexistence of coronary artery disease, peripheral arterial disease, and atherothrombotic brain infarction in men and women 62 years of age. Am J Cardiol 1994;74:64-65.

RELATED ARTICLE: Key Points

* Peripheral vascular disease has been traditionally underdiagnosed in the primary care setting due to the absence of established criteria to warrant diagnostic testing.

* Risk factors which increase the likelihood of peripheral vascular disease include hypertension, hyperlipidemia, tobacco use, diabetes, cardiovascular disease, obesity, and stroke.

* The aim of this analysis was to develop a hierarchy of cardiovascular risk factors with predictive power for peripheral vascular disease.

* Our study establishes the following hierarchy among cardiovascular risk factors as predictors of peripheral vascular disease: hypertension, cardiovascular disease, hyperlipidemia, diabetes, tobacco use, obesity, stroke.

Jeffrey J. Fine, MS, Patrick A.X. Hall, MD, FACC, and J. Huger Richardson, MD, FACC

From the South Carolina Heart Center, Columbia, SC.

Reprint requests to Jeffrey Fine, MS, South Carolina Heart Center, Department of Investigator Initiated Research, 2001 Laurel Street, Columbia, SC 29063. Email: [e-mail]