Abstract

The current study describes the age- and sex-specific incidence rates and risk factors for asymptomatic and symptomatic peripheral arterial occlusive disease (PAOD) among 2,327 subjects and the incidence of intermittent claudication in asymptomatic PAOD subjects. The study population was selected from 18 general practice centers in the Netherlands. PAOD was assessed with the ankle-brachial blood pressure index, and intermittent claudication was assessed with a modified version of the Rose questionnaire. After 7.2 years, the overall incidence rate for asymptomatic PAOD, using the person-years method, was 9.9 (95% confidence interval (CI): 7.3, 18.8) per 1,000 person-years at risk. The rate was 7.8 (95% CI: 4.9, 20.3) for men and 12.4 (95% CI: 7.7, 24.8) for women. For symptomatic PAOD, the incidence rate was 1.0 (95% CI: 0.7, 7.5) overall, 0.4 (95% CI: 0.3, 10.0) for men, and 1.8 (95% CI: 1.0, 10.3) for women. Multivariate analyses showed that increasing age, smoking, hypertension, and diabetes mellitus were the most important risk factors. The overall incidence rate for intermittent claudication among PAOD subjects who were asymptomatic at baseline was 90.5 per 1,000 person-years at risk (95% CI: 36.4, 378.3). The incidence of asymptomatic PAOD was higher than the incidence of symptomatic PAOD, with women developing PAOD more often than men. In the development of preventive strategies, modification of atherosclerotic risk factors, such as smoking, hypertension, and diabetes, should be the main goals.

All epidemiologic cross-sectional studies show that in the general population asymptomatic peripheral arterial occlusive disease (PAOD) is more common than symptomatic PAOD (13). Less than half of all PAOD patients have symptoms of intermittent claudication. PAOD is part of a systemic illness caused by atherosclerosis, and its prevalence is associated with increasing age (4, 5).

Population-based studies indicate that PAOD has a relatively benign course in the legs (6, 7). However, claudicants show more cardiovascular comorbidity and have at least a twofold risk of fatal ischemic heart disease and cerebrovascular accidents compared with the general population (810), which results in high health care costs (1113). Concomitant cardiovascular disease and cardiovascular mortality is even more frequent in studies where the ankle-brachial index, the ratio of ankle systolic blood pressure to arm systolic blood pressure, has been used to indicate PAOD. The ankle-brachial index is currently the most common and useful diagnostic instrument for detecting PAOD, including asymptomatic cases (4, 5, 14, 15). Ankle-brachial index is inversely associated with the manifestation of severe disease and cardiovascular mortality (10, 16). Both intermittent claudication and asymptomatic PAOD seem to be associated with higher cardiovascular morbidity and mortality. However, data on the prognosis of persons with asymptomatic PAOD are scarce (6, 17).

Despite the relatively decreased life expectancy of asymptomatic and symptomatic PAOD patients due to cardiovascular morbidity, there have been only a few incidence studies of PAOD. Furthermore, there are almost no prospective studies on risk factors for asymptomatic PAOD or on the development of intermittent claudication symptoms when asymptomatic disease is already present (10).

In this paper, we report the age- and sex-specific incidence of asymptomatic and symptomatic PAOD in a population-based prospective study. In addition, we present the incidence of intermittent claudication in the subgroup of asymptomatic PAOD subjects at baseline. We also investigate possible risk factors for the development of asymptomatic and symptomatic PAOD.

MATERIALS AND METHODS

The Limburg PAOD Study started in 1988 (1). The study population was selected from a source population of 26,620 subjects aged 40–78 years from 18 general practice centers in the Netherlands. In the Netherlands, every individual is registered at a general practice center; therefore, the registered population of these general practices is a segment of the general population. A screening questionnaire was mailed to these subjects with five questions, mainly about risk factors possibly relevant to PAOD, so that responders could be characterized according to a “prior risk score” between zero and 5. Thus, the study population included subgroups corresponding to parts of the clinical spectrum relevant to the general practice. A stratified sampling procedure based on the subjects' prior risk scores was used to compose the study population. Two random samples were taken: one from the category with a score of zero and one from the category with a score of 1–5. From the remaining subjects with a score between 1 and 5, those with the highest risk scores were selected. A total of 5,301 subjects were invited, and 3,649 respondents (53 percent female) participated in the study. The response rate was 68.9 percent, and no statistically significant differences between responders and nonresponders were found with regard to age and gender.

The population's characteristics have been extensively described elsewhere (1). All participants gave their informed consent. The Medical Ethics Committee of the University Hospital Maastricht approved the study. For the current study, baseline data were related to measurements made after follow-up; therefore, those who died before follow-up measurement (n = 481) were not included in these analyses, but their baseline data were compared with the data of the included subjects.

Baseline measurements

Each participant completed a questionnaire that included questions about age, gender, symptoms of intermittent claudication, current and former smoking habits, physical activities, and the occurrence of atherosclerosis in the family. Subsequently, a physical examination was performed by a general practitioner/physician in one of the general practice centers. A practice assistant measured height, weight, and ankle-brachial index with a pocket Doppler device (Huntleigh Mini Dopplex D500 8Mhz; Huntleigh Healthcare Ltd., Luton, Bedfordshire, United Kingdom) and a sphygmomanometer. Systolic blood pressures were recorded on both arms and both ankles (posterior tibial artery or dorsalis pedis artery) with the subject lying in a supine position. The ankle-brachial indexes were calculated as the ratio of the systolic blood pressure of each ankle to the highest arm systolic blood pressure. To achieve uniformity of Doppler assessments, the practice assistants were specially trained (18). An ankle-brachial index less than 0.95 for at least one leg, measured twice at weekly intervals, was considered indicative of PAOD. In addition, relevant data on the presence of hypertension (guideline: five instances of diastolic blood pressure of 95–104 mmHg or three instances of diastolic blood pressure ≥105 mmHg), hypercholesterolemia (guideline: mean of three measurements ≥6.5 mmol/liter), and diabetes mellitus (guideline: typical complaints and one instance of abnormal serum glucose (fasting level ≥6.7 mmol/liter, 2 hours postprandial level ≥11.1 mmol/liter) or two instances of abnormal serum glucose) were collected from the patient's medical record.

Intermittent claudication was defined as pain in the calf while walking which disappeared when the patient was standing still and which was not present when the patient was resting (typical intermittent claudication) (19). Other complaints involving the calf and complaints involving the foot, thigh, or buttocks showing this ischemic pattern were included in the definition of atypical intermittent claudication (1, 6). Asymptomatic PAOD was defined as an ankle-brachial index less than 0.95 without symptoms of intermittent claudication. Symptomatic PAOD was defined as an ankle-brachial index less than 0.95 with intermittent claudication symptoms (1). Overweight was defined as a body mass index (weight (kg)/height (m)2) greater than 30. Familial atherosclerosis was coded positive when the participant had a first- or second-degree relative under the age of 65 years who suffered from atherosclerotic disease. Daily work and leisure time activities were classified as heavy, moderate, or light. Participants who scored light for both daily work and leisure time activities were categorized as having a sedentary lifestyle.

Follow-up evaluation

All participants were traced by means of the general practice registers. The subjects who were still alive were invited to undergo reexamination between January 1995 and April 1997, according to the same procedures as those used for the baseline measurements. Participants who had switched from their original general practice during the follow-up period but still resided in the province of Limburg were located, contacted, and invited. Participants who could not attend the examination at a general practice center were visited at home.

Statistical analyses

Total and sex- and age-specific (three categories: 40–54, 55–64, and ≥65 years) incidence rates for asymptomatic and symptomatic PAOD were estimated by means of the person-years method (number of new cases divided by the person-years at risk of the initial population, excluding all PAOD subjects at baseline). Incidence rates were recalculated to the source population as appropriately weighted means per prior risk category, as follows. The sampling fractions from the source population were known at the beginning of the study. The study population was divided into four strata, and each stratum was attributed a corresponding weight. The incidence rates were calculated as weighted means of stratum-specific incidences. Weights were taken as the inverse of the sampling fractions. We programmed the calculations using the statistical programming language S-PLUS (20).

Because of small numbers of observed events, the normal approximation method for confidence intervals was not valid. Therefore, the method proposed by Waller et al. (21) was used, with an adaptation for the assumed Poisson distribution for the numbers of events.

At the end of the follow-up period, subjects could be classified into the following groups: subjects who were normal (without PAOD) both at baseline and at follow-up; subjects normal at baseline but asymptomatic at follow-up; subjects normal at baseline but symptomatic at follow-up; subjects asymptomatic at baseline and not symptomatic at follow-up; and subjects asymptomatic at baseline but symptomatic at follow-up. Person-years at risk were calculated with the approximation that subjects who developed the disease during the follow-up period were assumed to contribute half of the follow-up time to the person-years at risk.

Backwards multiple logistic regression analyses were carried out to investigate the associations of possible risk factors—aging, gender, smoking, hypertension, overweight, diabetes, hypercholesterolemia, and sedentary lifestyle—and the occurrence of atherosclerosis in the family (independent variables) with the incidence of asymptomatic and symptomatic PAOD (dependent variables). Multivariable analysis was also performed in the subgroup of PAOD subjects who were asymptomatic at baseline to investigate the prognostic importance of these variables in the transition from asymptomatic disease to symptomatic disease. In this subgroup, the dependent variable was the presence of symptoms of intermittent claudication at the end of the follow-up period.

RESULTS

After 7.2 years, 2,589 subjects (54.2 percent female) from the included cohort (n = 3,649) with a mean age of 64.1 years (range, 48.1–85.5 years) had participated in the follow-up measurements (49 percent of the original cohort that was invited at baseline and 71 percent of the cohort that was included at baseline).

During the follow-up period, 481 participants (13.2 percent of the included cohort) died, of whom 61.4 percent were males. From the included cohort, death occurred in 11 percent of the normal subjects at baseline, 25.2 percent in the asymptomatic PAOD subjects, and 31.2 percent in the claudicants at baseline. Comparison of the baseline variables (age, sex, ankle-brachial index, possible risk factors, and cardiovascular comorbidity) among those who died and those who completed follow-up showed that those who died were mainly older men with lower ankle-brachial index, more prevalent cardiovascular disease, and a more unfavorable risk factor profile. The main causes of death were cardiovascular disease and cancer. A detailed mortality follow-up is currently in progress.

Thus, 84.2 percent of the included cohort at baseline was followed (71 percent who attended the second examination and 13.2 percent who died during the follow-up period). The remaining group (15.8 percent) was not followed for two reasons: refusal to take part or being incapable of taking part because of severe illness or hospitalization. Nonresponders and responders did not differ significantly by age or gender. Furthermore, nonresponders did not have a worse risk factor profile at baseline than responders, except for persons with diabetes.

A total of 262 subjects were excluded because of the presence of PAOD at baseline or because PAOD could not be assessed. Overall, 2,327 subjects were included in the incidence analyses.

Normal group at baseline: incidence of asymptomatic and symptomatic PAOD

At the end of the follow-up period, 200 participants had asymptomatic PAOD and 46 had symptomatic PAOD. Table 1 shows the age- and sex-specific incidence rates of asymptomatic and symptomatic PAOD among the survivors. The highest rates were observed for asymptomatic PAOD. The overall incidence rate for asymptomatic PAOD was 9.9 per 1,000 person-years at risk (7.8 for male subjects and 12.4 for female subjects). For symptomatic PAOD, the overall incidence rate was 1.0 per 1,000 person-years at risk (0.4 for men and 1.8 for women).

TABLE 1.

Age- and sex-specific incidence rates for asymptomatic and symptomatic peripheral arterial occlusive disease (PAOD): Limburg PAOD Longitudinal Study, 1988 –1997 (n = 2,327)

Sex and age group (years) Overall PAOD
 
Asymptomatic PAOD
 
Symptomatic PAOD
 
No.* Rate/1,000 person-years 95% confidence interval No.* Rate/1,000 person-years 95% confidence interval No.* Rate/1,000 person-years 95% confidence interval 
Male          
 40 –54 32 1.7 0.9, 8.1 25 1.5 0.7, 7.8 0.2 0.1, 5.9 
 55 –64 49 1.5 1.1, 14.9 36 1.1 0.8, 14.4 13 0.4 0.2, 13.6 
 ≥65 29 17.8 7.3, 51.7 23 17.2 6.9, 51.1 0.5 0.2, 23.7 
 Total 110 8.2 5.3, 20.8 84 7.8 4.9, 20.3 26 0.4 0.3, 10.0 
Female          
 40 –54 31 5.9 2.9, 13.3 27 5.8 2.8, 13.1 0.1 0.06, 4.8 
 55 –64 37 9.1 4.4, 20.2 33 5.9 2.7, 15.7 3.2 1.3, 11.7 
 ≥65 68 22.9 10.1, 55.6 56 20.8 8.7, 53.0 12 1.9 0.6, 23.2 
 Total 136 14.2 9.1, 27.1 116 12.4 7.7, 24.8 20 1.8 1.0, 10.3 
Entire group 246 11.0 8.2, 20.0 200 9.9 7.3, 18.8 46 1.0 0.7, 7.5 
Sex and age group (years) Overall PAOD
 
Asymptomatic PAOD
 
Symptomatic PAOD
 
No.* Rate/1,000 person-years 95% confidence interval No.* Rate/1,000 person-years 95% confidence interval No.* Rate/1,000 person-years 95% confidence interval 
Male          
 40 –54 32 1.7 0.9, 8.1 25 1.5 0.7, 7.8 0.2 0.1, 5.9 
 55 –64 49 1.5 1.1, 14.9 36 1.1 0.8, 14.4 13 0.4 0.2, 13.6 
 ≥65 29 17.8 7.3, 51.7 23 17.2 6.9, 51.1 0.5 0.2, 23.7 
 Total 110 8.2 5.3, 20.8 84 7.8 4.9, 20.3 26 0.4 0.3, 10.0 
Female          
 40 –54 31 5.9 2.9, 13.3 27 5.8 2.8, 13.1 0.1 0.06, 4.8 
 55 –64 37 9.1 4.4, 20.2 33 5.9 2.7, 15.7 3.2 1.3, 11.7 
 ≥65 68 22.9 10.1, 55.6 56 20.8 8.7, 53.0 12 1.9 0.6, 23.2 
 Total 136 14.2 9.1, 27.1 116 12.4 7.7, 24.8 20 1.8 1.0, 10.3 
Entire group 246 11.0 8.2, 20.0 200 9.9 7.3, 18.8 46 1.0 0.7, 7.5 
*

Number of incident cases during 7.2 years.

Asymptomatic PAOD incidence rates increased with age among men (except for the age category 55–64 years) and women. The asymptomatic incidence rates for men aged 40–54, 55–64, and ≥65 years were 1.5, 1.1, and 17.2, respectively; for women, these figures were 5.8, 5.9, and 20.8, respectively. For symptomatic PAOD, the incidence rates also increased with age, except for women in the age category ≥65 years. The symptomatic incidence rates for men aged 40–54, 55–64, and ≥65 years were 0.2, 0.4, and 0.5, respectively; for women, these rates were 0.1, 3.2, and 1.9, respectively. In both men and women, the incidence of asymptomatic PAOD was higher than the incidence of symptomatic PAOD. A (nonsignificant) difference was observed between the PAOD incidence rates of men and women: Women developed asymptomatic disease three times more often than men, on average. Among those older than 54 years, the incidence rates of symptomatic PAOD were also higher (approximately sixfold) in women compared with men.

Normal group at baseline: risk factor pattern

Table 2 presents the results of the multiple logistic regression analyses. Older age, smoking, hypertension, diabetes, hypercholesterolemia, and a sedentary lifestyle were the most important risk factors for asymptomatic PAOD. Older age, smoking, hypertension, and diabetes were the most important predisposing factors for symptomatic PAOD.

TABLE 2.

Risk factors associated with the probability of asymptomatic and symptomatic peripheral arterial occlusive disease (PAOD) after 7.2 years*: Limburg PAOD Longitudinal Study, 1988–1997 (n = 2,327)

Risk factor Overall PAOD
 
Asymptomatic PAOD
 
Symptomatic PAOD
 
No. Odds ratio 95% confidence interval No. Odds ratio 95% confidence interval No. Odds ratio 95% confidence interval 
Age (years)          
 40–54 63 1.0  52 1.0  11 1.0  
 55–64 86 1.8 1.3, 2.6 69 1.7 1.2, 2.6 17 2.1 0.9, 4.8 
 ≥65 97 4.0 2.8, 5.9 79 3.8 2.5, 5.7 18 5.3 2.3, 12.2 
Smoking status          
 Nonsmoker 69 1.0  60 1.0  1.0  
 Ex-smoker 45 0.9 0.6, 1.4 37 0.8 0.5, 1.3 1.4 0.5, 3.7 
 Current smoker 131 2.2 1.5, 3.1 102 1.9 1.3, 2.8 29 4.3 1.9, 10.1 
Hypertension 126 1.7 1.3, 2.4 103 1.7 1.2, 2.3 23 1.9 1.0, 3.6 
Diabetes mellitus 38 2.1 1.4, 3.3 28 1.8 1.1, 2.9 10 3.8 1.8, 8.2 
Hypercholesterolemia 44 1.5 1.0, 2.3 36 1.6 1.1, 2.5 §  
Sedentary lifestyle 201 §  166 1.6 1.0, 2.5 35 §  
Risk factor Overall PAOD
 
Asymptomatic PAOD
 
Symptomatic PAOD
 
No. Odds ratio 95% confidence interval No. Odds ratio 95% confidence interval No. Odds ratio 95% confidence interval 
Age (years)          
 40–54 63 1.0  52 1.0  11 1.0  
 55–64 86 1.8 1.3, 2.6 69 1.7 1.2, 2.6 17 2.1 0.9, 4.8 
 ≥65 97 4.0 2.8, 5.9 79 3.8 2.5, 5.7 18 5.3 2.3, 12.2 
Smoking status          
 Nonsmoker 69 1.0  60 1.0  1.0  
 Ex-smoker 45 0.9 0.6, 1.4 37 0.8 0.5, 1.3 1.4 0.5, 3.7 
 Current smoker 131 2.2 1.5, 3.1 102 1.9 1.3, 2.8 29 4.3 1.9, 10.1 
Hypertension 126 1.7 1.3, 2.4 103 1.7 1.2, 2.3 23 1.9 1.0, 3.6 
Diabetes mellitus 38 2.1 1.4, 3.3 28 1.8 1.1, 2.9 10 3.8 1.8, 8.2 
Hypercholesterolemia 44 1.5 1.0, 2.3 36 1.6 1.1, 2.5 §  
Sedentary lifestyle 201 §  166 1.6 1.0, 2.5 35 §  
*

Results from backwards multiple logistic regression analyses (reduced models).

For definition, see “Materials and Methods.”

Referent category.

§

Dropped out of the reduced model.

From the reduced logistic regression models, predicted probabilities for asymptomatic and symptomatic PAOD incidence according to the baseline risk factor profile were calculated. These absolute probabilities are risk differences for subjects with one or more of the risk factors found significant in our study compared with the reference group (i.e., those without risk factors). The probability of developing both asymptomatic PAOD and symptomatic PAOD increased with the number of risk factors present at baseline. Table 3 illustrates this increasing probability for the development of overall asymptomatic and symptomatic PAOD on the basis of the four most significant risk factors (age ≥65 years, smoking, diabetes, and hypertension) present at baseline.

TABLE 3.

Probability (%) of asymptomatic and symptomatic peripheral arterial occlusive disease (PAOD) according to the number and combination of the four most significant risk factors present at baseline*: Limburg PAOD Longitudinal Study, 1988–1997 (n = 2,327)

Number and combination of risk factors at baseline Probability (%) of overall PAOD Probability (%) of asymptomatic PAOD Probability (%) of symptomatic PAOD 
Reference probability 
Age ≥65 years 
Age ≥65 years and hypertension 15 
Age ≥65 years and smoking 18 10 
Age ≥65 years and diabetes 18 10 
Age ≥65 years, hypertension, and diabetes 28 16 
Age ≥65 years, smoking, and hypertension 29 17 11 
Age ≥65 years, smoking, and diabetes 33 18 19 
Age ≥65 years, smoking, hypertension, and diabetes 47 27 31 
Number and combination of risk factors at baseline Probability (%) of overall PAOD Probability (%) of asymptomatic PAOD Probability (%) of symptomatic PAOD 
Reference probability 
Age ≥65 years 
Age ≥65 years and hypertension 15 
Age ≥65 years and smoking 18 10 
Age ≥65 years and diabetes 18 10 
Age ≥65 years, hypertension, and diabetes 28 16 
Age ≥65 years, smoking, and hypertension 29 17 11 
Age ≥65 years, smoking, and diabetes 33 18 19 
Age ≥65 years, smoking, hypertension, and diabetes 47 27 31 
*

Deduced from backwards multiple logistic regression models.

Asymptomatic group at baseline: incidence of intermittent claudication and risk pattern

Of the 177 asymptomatic PAOD subjects at baseline who underwent a follow-up examination, 27 developed intermittent claudication (15 with typical symptoms and 12 with atypical symptoms). In this group, the overall incidence rate for symptomatic PAOD was 90.5 (95 percent confidence interval: 36.4, 378.3) per 1,000 person-years at risk. PAOD subjects who were asymptomatic at baseline developed intermittent claudication symptoms more often than did the subjects who were normal at baseline.

Multiple logistic regression results showed the possibility that male sex (odds ratio (OR) = 1.6), hypertension (OR = 1.5), diabetes (OR = 1.7), and older age (OR = 2.7 and 3.2 for the two older age categories (55–64 and ≥65 years), respectively) were associated with the development of symptoms of intermittent claudication. No statistical significance was reached, possibly because of small numbers.

DISCUSSION

From the invited cohort, approximately 70 percent of the invitees attended the first examination (3,649 of 5,301 subjects) and 49 percent attended both the first and the second examinations. This may constitute a possible source of bias, even though the study data at baseline did not seem to suggest that this bias was major. Nevertheless, a certain loss of power is obvious, since there were fewer participants at the end of follow-up.

Misclassification of especially symptomatic PAOD could have occurred in the study population. Symptomatic subjects who adapted their lifestyle to avoid symptoms or subjects who had a sedentary lifestyle preventing symptom onset may have been classified as asymptomatic. In our data, it appeared that only 3.8 percent of the asymptomatic subjects reported very low mobility or no mobility. Thus, the occurrence of misclassification bias in our study was not likely to be substantial. Spuriously elevated ankle-brachial indexes due to calcified vessels, which are sometimes found, especially in diabetic patients, may also form a source of misclassification bias. In our data, only two persons with diabetes had an ankle-brachial index ≥1.50. Thus, possible misclassification of diabetics in our study population was, at most, fractional.

Medical care given to the PAOD respondents at baseline may also be a potential source of confounding. All of the medical data from the respondents were available for this study. Nevertheless, this potential form of confounding is inevitable.

The generalizability of our results to other populations may be limited. However, nearly all citizens of the Netherlands are registered at a general practice. Thus, by sampling study subjects from general practices, a study population can be compounded that is relatively equivalent to the general population.

The estimation of incidence rates relies on the time of disease occurrence and loss to follow-up. The exact time of disease occurrence was not known in our study, the average time lapse between visits 1 and 2 being 7.2 years. Even with the applied correction in the calculations, this probably could have led to an underestimation of the reported incidence rates.

A source of bias could be subjects who refused to participate in the follow-up measurements. To minimize this potential bias and to keep the nonresponse group as small as possible, we offered home visits to subjects who were not willing to visit their general practice. Subjects who missed their appointment were telephoned and invited to attend at a later time. The nonresponders might have had a different disease incidence than the responders because of a higher prevalence of diabetes at baseline. Quantitatively, however, this did not have any significant impact on the incidence rates.

The attrition of the cohort could also have influenced the incidence rate calculations. Mortality may censor the diagnosis of PAOD and consequently lead to conservative incidence estimations. Even though PAOD is not a short term fatal disease, subjects have a lower survival rate because of ischemic heart disease and cerebrovascular disease mortality in comparison with the general population (7, 911, 13). We gathered data on the causes of mortality and possible nonfatal cardiovascular events before death, using hospital correspondence and the patients' medical records. Among the 349 subjects who were normal at baseline and who died during follow-up, cardiovascular disease was the cause of death in approximately 50 percent. New cases of PAOD are especially probable in this group.

Finally, small numbers of events probably led to low statistical power and imprecise estimates in some of the incidence rate analyses.

Normal group at baseline: incidence of asymptomatic and symptomatic PAOD

We found that the incidence of asymptomatic and symptomatic PAOD increased with age in both men and women. Furthermore, the incidence of asymptomatic PAOD was much higher in the general population than that of symptomatic PAOD. Surprisingly, a trend was observed that women developed asymptomatic and symptomatic PAOD more often than men after 7.2 years of follow-up.

In several studies, both asymptomatic and symptomatic PAOD increased with age, which is consistent with our findings (8, 22, 23). Our incidence rates for asymptomatic PAOD are difficult to compare with those of other studies because of a lack of data and differences in diagnostic methods used. In the Basel Study, a cohort of 2,630 men followed for 5 years, asymptomatic PAOD, defined as pathologic findings at pulse palpation, auscultation, or electronic oscillography, was three times more frequent than symptomatic PAOD (22). This is in accordance with our results. In the Framingham Study, an annual symptomatic PAOD incidence rate of 1.8 was reported for women, which is comparable with our findings (23). However, the Framingham Study found a twofold male predominance, with an incidence rate similar to that of the Quebec Cardiovascular Study (3.6 and 4.1, respectively) (8, 23). The Edinburgh Artery Study presented the highest overall symptomatic PAOD incidence rates (15.5 per 1,000 person-years) (6), men being at a slightly higher risk than women.

Various explanations can be postulated for the differences between the results of these other studies and our study. Firstly, the Limburg PAOD study population was younger than the populations of the Framingham (23) and Edinburgh (6) studies. Secondly, the Edinburgh Artery Study (6) did not exclude asymptomatic PAOD subjects at baseline from the denominator in its incidence calculations for symptomatic PAOD, even though asymptomatic subjects have, as our study shows, an incidence rate of intermittent claudication that is much higher than that of subjects without PAOD. In fact, when we calculated the overall symptomatic PAOD incidence rate, including the asymptomatic subjects at baseline in our calculations, the incidence rate for symptomatic PAOD almost doubled. Thirdly, in the Quebec Study (8), 70 percent of the men were smokers as compared with 58.6 percent in our study, smoking being one of the most powerful risk factors for PAOD.

Normal group at baseline: risk factor pattern

To identify risk factors for disease, the focus should ideally be on incident cases. We demonstrated that significant risk factors for the development of asymptomatic PAOD were older age, smoking, hypertension, diabetes, hypercholesterolemia, and a sedentary lifestyle. Because of a lack of data, we cannot compare these results with those of other studies. Furthermore, we found that older age, smoking, diabetes, and hypertension were strongly associated with the development of symptomatic PAOD, which supports the results of other studies (8, 23, 24). In addition, the probability of developing both asymptomatic and symptomatic PAOD increased with the number of risk factors at baseline, in concordance with the Basel Study (22).

The roles of hypercholesterolemia and sedentary lifestyle were not evident for symptomatic PAOD. In some studies, hypercholesterolemia is reported to be a possible risk factor for symptomatic PAOD (24, 25), but this has not been confirmed in other studies (4, 5). At the time of our baseline data collection, measurement of hypercholesterolemia in general practices in the Netherlands was just starting to receive more attention. Thus, the weak association found in our study might reflect an underreporting of hypercholesterolemia in the general practice centers. The relation between physical activity and PAOD has not been thoroughly investigated to date. There are indications that a sedentary lifestyle is associated with PAOD (26). However, symptoms of intermittent claudication can be disguised by a sedentary lifestyle.

Asymptomatic group at baseline: incidence of intermittent claudication and risk pattern

To our knowledge, this is the first study to report on the incidence of intermittent claudication among asymptomatic PAOD subjects and to have investigated the prognostic role of several risk factors. We found that PAOD subjects who were asymptomatic at baseline developed intermittent claudication symptoms more often than normal subjects. This is not surprising, since asymptomatic and symptomatic PAOD are not different entities but atherosclerotic manifestations appearing at different points in time. In particular, asymptomatic male subjects seemed to be at high risk for developing symptoms compared with asymptomatic female subjects, which provides further evidence that PAOD might have a worse prognosis in men.

Older age, hypertension, and diabetes probably play a role in the progression of asymptomatic PAOD to symptomatic PAOD.

Implications

Asymptomatic PAOD subjects as well as symptomatic PAOD subjects have increased risks of cardiovascular disease mortality (10). The incidence of asymptomatic PAOD is higher in the general population than that of symptomatic PAOD. Once asymptomatic disease is present, the incidence rate of intermittent claudication is significantly elevated. Claudicants have a reduced quality of life because of limited independence and mobility (27), impaired functioning being the main reason for referral to a vascular specialist. One of the goals of public and primary health care is to maintain a good quality of life in the elderly. To achieve this goal with regard to intermittent claudication, detection of asymptomatic PAOD is important. Measuring ankle-brachial index with a Doppler device assists the general practitioner in case-finding. It is a reproducible and valid diagnostic tool that is not only appropriate for detection of asymptomatic PAOD in epidemiologic studies but also suitable for use in primary care (15). Our data suggest that, in primary care, cessation of smoking and adequate treatment of hypertension and diabetes are the principal preventive objectives in PAOD patients. Treatment of hypercholesterolemia seems to be less important.

Thus, among persons over 50 years of age with one or more conventional atherosclerotic risk factors, assessment of ankle-brachial index by the general practitioner could provide additional information on the patient's risk of developing manifest atherosclerotic disease.

Correspondence to Dr. Jurenne D. Hooi, Leusdenhof 279, 1108 DM Amsterdam Z-O, The Netherlands (e-mail: jdhooi@zonnet.nl).

This study was funded by grant 92170 from the Netherlands Heart Foundation.

The authors thank the participating physicians and the practice assistants for their contributions to this study.

REFERENCES

1.
Stoffers HE, Rinkens PE, Kester AD, et al. The prevalence of asymptomatic and unrecognized peripheral arterial occlusive disease.
Int J Epidemiol
 
1996
;
25
:
282
–90.
2.
Fowkes FG, Housley E, Cawood EH, et al. Edinburgh Artery Study: prevalence of asymptomatic and symptomatic peripheral arterial disease in the general population.
Int J Epidemiol
 
1991
;
20
:
384
–92.
3.
Criqui MH, Browner D, Fronek A, et al. Peripheral arterial disease in large vessels is epidemiologically distinct from small vessel disease.
Am J Epidemiol
 
1989
;
129
:
1110
–19.
4.
Vogt MT, Wolfson SK, Kuller LH. Lower extremity arterial disease and the aging process: a review.
J Clin Epidemiol
 
1992
;
45
:
529
–42.
5.
Balkau B, Vray M, Eschwege E. Epidemiology of peripheral arterial disease.
J Cardiovasc Pharmacol
 
1994
;
23(suppl)
:
S8
–16.
6.
Leng GC, Lee AJ, Fowkes FG, et al. Incidence, natural history and cardiovascular events in symptomatic and asymptomatic peripheral arterial disease in the general population.
Int J Epidemiol
 
1996
;
25
:
1172
–81.
7.
Newton Peabody C, Kannel WB, McNamara PM. Intermittent claudication: surgical significance.
Arch Surg
 
1974
;
109
:
693
–7.
8.
Dagenais GR, Maurice S, Robitaille NM, et al. Intermittent claudication in Quebec men from 1974–1986: The Quebec Cardiovascular Study.
Clin Invest Med
 
1991
;
14:2
:
93
–100.
9.
Leng GC, Fowkes FG, Lee AJ, et al. Use of ankle brachial pressure index to predict cardiovascular events and death: a cohort study.
BMJ
 
1996
;
313
:
1440
–3.
10.
Hooi JD, Stoffers HE, Knottnerus JA, et al. The prognosis of non-critical limb ischaemia: a systematic review of population-based evidence.
Br J Gen Pract
 
1999
;
49
:
49
–55.
11.
Lebrun T, Selke B, Sailly JC, et al. The introduction of economic criteria into the management of arterial disease: an illustration with regard to the socioeconomic consequences of peripheral occlusive arterial disease of the lower limbs.
J Cardiovasc Pharmacol
 
1994
;
23(suppl 3)
:
S17
–21.
12.
Gillum RF. Peripheral arterial occlusive disease of the extremities in the United States: hospitalization and mortality.
Am Heart J
 
1990
;
6
:
1414
–18.
13.
Drummond M. Socioeconomic impact of peripheral vascular disease.
Atherosclerosis
 
1997
;
131(suppl)
:
S33
–4.
14.
Fowkes FG. The measurement of atherosclerotic peripheral arterial disease in epidemiological surveys.
Int J Epidemiol
 
1988
;
17
:
248
–54.
15.
Stoffers HE, Kester AD, Kaiser V, et al. The diagnostic value of the measurement of the ankle-brachial systolic pressure index in primary health care.
J Clin Epidemiol
 
1996
;
49
:
1401
–5.
16.
McGrae McDermott M, Feinglass J, Slavensky R, et al. The ankle-brachial index as a predictor of survival in patients with peripheral vascular disease.
J Gen Intern Med
 
1994
;
9
:
445
–9.
17.
Criqui MH, Langer RD, Fronek A, et al. Mortality over a period of 10 years in patients with peripheral arterial disease.
N Engl J Med
 
1992
;
326
:
381
–6.
18.
Stoffers HE, Kaiser V, Kester AD, et al. Peripheral arterial occlusive disease in general practice: the reproducibility of the ankle-arm pressure ratio.
Scand J Prim Health Care
 
1991
;
9
:
109
–14.
19.
Rose GA, Blackburn H, Gillum RF, et al. Cardiovascular survey methods. 2nd ed. Geneva, Switzerland: World Health Organization, 1982:162–5.
20.
S-PLUS guide to statistical and mathematical analysis, version 3.3. Seattle, WA: MathSoft, Inc, 1995.
21.
Waller JL, Addy CL, Jackson KL, et al. Confidence intervals for weighted proportions.
Stat Med
 
1994
;
13
:
1071
–82.
22.
Widmer LK, Biland L. Incidence and course of occlusive peripheral artery disease in geriatric patients.
Int Angiol
 
1985
;
4
:
289
–94.
23.
Kannel WB, McGee DL. Update on some epidemiologic features of intermittent claudication: The Framingham Study.
J Am Geriatr Soc
 
1985
;
33
:
13
–18.
24.
Bowlin SJ, Medalie JH, Flocke SA, et al. Epidemiology of intermittent claudication in middle-aged men.
Am J Epidemiol
 
1994
;
140
:
418
–30.
25.
Ingolfsson IÖ, Sigurdsson G, Sigvaldason H, et al. A marked decline in the prevalence and incidence of intermittent claudication in Icelandic men 1969–1986: a strong relationship to smoking and serum cholesterol—The Reykjavik Study.
J Clin Epidemiol
 
1994
;
47
:
1237
–43.
26.
Housley E, Leng GC, Donnan PT, et al. Physical activity and risk of peripheral arterial disease in the general population: Edinburgh Artery Study.
J Epidemiol Community Health
 
1993
;
47
:
475
–80.
27.
Khaira HS, Hanger R, Shearman CP. Quality of life in patients with intermittent claudication.
Eur J Vasc Endovasc Surg
 
1996
;
11
:
65
–9.