Reviews/ACE Inhibitors and Angiotensin Receptor Blockers/ACE inhibitor-induced cough is frequent
Safety · ACEARB-SAFE-01 · v2
Borderline: within 4 points of the Established/Likely boundary, so a re-grade could land either side.
Cough is a recognised side effect of ACE inhibitors, and several independent studies in this corpus measured how often it happens. A controlled cohort found cough in 12.3% of ACE-inhibitor patients versus 2.7% of controls [5], a large surveillance study reported 7.8% in men and 11.3% in women [11], a trial that stopped and restarted the drug found 10.96% to 22.86% across four agents [14], and a cross-sectional survey found 19.8% [23]. Designs that gave the drug under blinded conditions and watched the cough return confirm the drug is the cause [2][9]. One very large pooled analysis reports only 3.9%, but that figure counts patients who stopped the drug, not everyone who coughed [6]. Taken together the primary studies land inside the 5 to 20 percent range the claim states, which is why the panel treated both the causal link and the broad range as well supported.
Independent primary designs converge inside the 5 to 20 percent range: a controlled retrospective cohort at 12.3% versus 2.7% on controls [5], 47,351-patient surveillance at 7.8% and 11.3% [11], a dechallenge-rechallenge RCT spanning 10.96% to 22.86% [14], and a 19.8% cross-sectional prevalence [23]. Causation is secured by blinded rechallenge and controlled comparison, where cough falls sharply on an angiotensin-receptor blocker or diuretic [2] and recurs on reintroduction of the drug [9]. The one large pooled RCT analysis reports 3.9%, but defined as discontinuation due to cough rather than all cough [6]. Across different countries, designs and ascertainment methods the incidence estimates cluster within the claimed band. The contested point: The evidence divides on where the exact boundaries sit: the strongest single dataset reads 3.9% because it counts only patients who stopped the drug [6], sitting below the 5% floor, while the 20% ceiling is reached only in smaller studies [14][23]. The band holds as a broad bracket, but neither exact bound is pinned down by the largest primary cohorts.
All 28 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from ACE Inhibitors and Angiotensin Receptor Blockers, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The largest, methodologically cleanest dataset [6]/[15] reports 3.9% cough-driven discontinuation, below the 5% floor; if a reader conflates discontinuation with incidence the claim's lower bound is unsupported by the strongest study.
The exact '5-20%' boundary derives from excluded review literature ([18], [24]); among included primary studies the 20% ceiling is met or exceeded only by small studies ([14] 22.86%, [23] n=101 pilot), not by the large cohorts, so the upper bound rests on thin primary data.
Where it remains uncertain. The largest pooled analysis reports 3.9% [6], but that is cough-driven discontinuation rather than total incidence, and the 20% ceiling is met only in smaller studies [14][23].
Aleatory
Cough occurs in only a fraction of treated patients, and who coughs varies with sex and age — female gender and age above 65 raise the odds in [6] — so any population figure is a rate over variable individuals.
Epistemic
Studies measure different things — all cough [5], discontinuation [6][8], or prevalence [23] — so the corpus does not pin down a single incidence number, only a broad band.
Inferential
The claim specifies 'persistent dry cough', but several studies count any cough or cough-driven discontinuation rather than persistence specifically, a gap between what is measured and what is claimed.
The load-bearing papers from a corpus of 28, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Controlled cohort incidence (12.3%) sits inside the 5-20% band with a lipid-lowering control arm.
“The prevalence of ACE inhibitor cohort cough was 12.3% and 2.7% in the patients on LLDs, P < 0.0001.”
Panel auditInternal validity: medium · scope of inference: Adult patients on captopril, enalapril or lisinopril versus lipid-lowering controls.Key limitations: retrospective · cough grouped with respiratory reactions
Largest dataset; 3.9% discontinuation is a lower-bound subset of incidence, consistent with (not contradicting) a 5-20% incidence.
“In 27,492 patients with cardiovascular disease, 1076 patients discontinued ACE-inhibitor perindopril due to cough (3.9%)”
Panel auditInternal validity: high · scope of inference: ACE-inhibitor-naive cardiovascular-disease patients on perindopril, cough-driven discontinuation.Key limitations: discontinuation not incidence · single ACE inhibitor
Population-based case-control confirms the causal association under everyday conditions.
“cases were 3.6 times as likely as controls to have been exposed to ACE inhibitors (95% CI: 2.4-5.5)”
Panel auditInternal validity: medium · scope of inference: Adults in Dutch general practice with incident cough.Key limitations: odds ratio not incidence · case-control cannot give rate
Pharmacovigilance corroboration that cough is a genuine, dose-independent ACEi adverse reaction.
“Review of these reports showed that the cough was an adverse reaction to the drugs, occurred even with low dose treatment, and was severe enough to warrant withdrawal of the drugs in most of the cases reported.”
Panel auditInternal validity: low · scope of inference: Reported cases of captopril and enalapril cough in New Zealand surveillance.Key limitations: no denominator · selective reporting
4 of 28 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
Multiple independent primary designs establish both the causal link and an incidence bracketing 5-20%, with residual uncertainty confined to boundary values and population heterogeneity rather than the central claim.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
Incidence versus discontinuation reconciliation. The band's lower edge is contested because the largest dataset reports 3.9% discontinuation [6] while cohort incidence reads 12.3% [5]; the two metrics are never measured side by side in one population. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A single large cohort reporting both total cough incidence and cough-driven discontinuation in the same patients.
Open question: What is the incidence of specifically persistent dry cough, under a uniform duration-based definition, in an unselected treated population? Why it is still open: Included primary studies measure all cough [5], prevalence [23] or discontinuation [6][8] rather than persistence, so the exact phenotype in the claim is never isolated.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19