In the excellent review article of renovascular hypertension by Mann and Pickering [1], the captopril test is proposed as a diagnostic tool to differentiate between renovascular and nonrenovascular hypertension, particularly in patients with a moderate index of clinical suspicion (5% to 15%). In patients where the prevalence is estimated to be 1% or less, the test is not recommended because of a low positive predictive value of 6.4% or less [1].

We evaluated the usefulness of the captopril test [2] at our nonreferral general medical outpatient clinic in 141 consecutive patients who had newly detected hypertension or treatment-resistant hypertension. All patients had normal renal function. Patients with a positive test underwent renal angiography. A renal artery stenosis with 75% or more luminal narrowing was considered significant [3]. Thirteen of 141 patients had a positive test, and 1 of them showed renal artery stenosis at angiography (prevalence, 0.7%). The positive predictive value of 7.7% agrees well with the estimates of Mann and Pickering [1]. The only variable we could determine to be associated with a positive captopril test was a positive smoking history. Eight of thirteen patients (62%) with a positive test were smokers, as opposed to 38 of 113 (29%) of those with a negative test (P = 0.04; chi-square test). The incidence of false-positive test results was higher in smokers (7 of 45, 16%) than in nonsmokers (5 of 97, 5%; P = 0.04).

The number of cigarettes smoked daily was documented in the charts of 24 patients. Patients smoking 20 or more cigarettes per day had significantly higher baseline and stimulated plasma renin activity values and a greater increase in those values after captopril than nonsmokers; patients smoking 20 or fewer cigarettes per day had plasma renin activity values similar to nonsmokers (Table 1).

We conclude that smoking may interfere with the results of the captopril test. Further, smoking may limit the utility of the test as a screening procedure.