Risk Factors for Angiotension‐Converting Enzyme Inhibitor–Associated Cough

Abstract

The article by Fernandez-Rando et al1 published in the Journal of Clinical Pharmacology continues to create intriguewith the definition of risk factors for angiotensinconverting enzyme inhibitor (ACEI)-associated cough. Their study found risk factors to include female sex, sleep apnea, and use of thiazide diuretics. A review of past similar studies, however, finds considerable variation in such associations. For example, some studies have not found any greater association with asthma.2,3 Some studies have not found an association of female sex with cough.2,4 Further study of the latter, which was said to support a link between female sex and cough, did not reach statistical significance for this variable.5 Yet others have found a positive association with smoking.6,7 The aforementioned studies, among others, clearly indicate a lack of consensus for risk factors and ACEI-associated cough. Despite the latter, clinical practice guidelines for use of this class of pharmacologic agents cite some such risk factors.8 If validated risk factors exist, it would potentially be of value to create clinical predictive rules to determine which patient may be harmed from ACEI therapy.9 To determine how any such predictors could be used for a given patient population in the lower mainland of British Columbia, Canada, a pilot observational review for patients with validated ACEI cough (n = 40) and controls (n = 200) was structured to assess risk factors including type of ACEI, age, sex, ethnicity, comorbidities, and other coincident pharmacotherapy. Female sex (P= .49) and age (P= .18) were not risk factors in this multiethnic population. Specific ethnicity, and Asian profile in particular, were not associated, nor was the combination of female sex and Asian heritage (P = .84). There were no associations for asthma (P = .70) or smoking (P = 1.0). Controls were more likely to be prescribed calcium-channel blockers (P = .002), βblockers (P= .008), anticoagulants (P= .001), and oral diabetic medication (P = .002). The above results continue to raise doubt about reliable and clinically useful risk factors for the prescription of ACEIs and for the prevention of such a side effect. Contemporary studies that have examined genetic polymorphisms in the search for defining risk do not often maintain sufficient inclusion for concern of the many potentially applicable demographic or clinical variables.10 Many studies are complicated by a lack of understanding for the underlying prevalence of chronic cough in control populations and by the delay in the onset of cough after initiation of ACEI treatments. It would appear necessary, then, to continue with largescale multivariate analyses that would combine various genetic marker determinations and clinical and demographic patient status. Even when a risk factor is identified in a statistically significant manner, the margin of difference between affected and unaffected patientsmay not yield itself to realistic clinical predictive rules or practice guidelines. It is evident that solid conclusions are yet to be had in this endeavor.