The research, led by Xinya Li and colleagues including Anxin Wang, drew on the Third China National Stroke Registry, one of the largest and most rigorously maintained stroke databases in the world. Between 2015 and 2018, the registry enrolled consecutive adult patients admitted to participating hospitals with a diagnosis of ischemic stroke, in which a blood clot blocks flow to part of the brain, or transient ischemic attack, a temporary interruption of blood supply that often precedes a full stroke. From this population, the investigators identified 2,859 patients who left the hospital with a prescription for either amlodipine or nifedipine and who could be followed for five years afterward.

The question the team posed was deceptively simple: did patients taking one drug fare better than patients taking the other? But answering it in an observational setting, where doctors choose medications rather than a computer randomizing them, requires sophisticated statistical machinery. The researchers’ primary yardstick was the modified Rankin Scale, or mRS, a standard clinical tool that grades functional outcomes after stroke on a spectrum from zero, meaning no symptoms at all, to six, indicating death. Rather than collapsing this scale into crude categories, the team analyzed the full distribution of scores using ordinal logistic regression, a method that preserves the ordered nature of disability levels and yields an odds ratio describing the odds of being at a better rather than a worse point on the scale.

Because patients prescribed amlodipine might differ systematically from those given nifedipine, in age, sex, body mass index, blood pressure on admission, or timing of their stroke, the researchers applied inverse probability of treatment weighting, a technique that statistically reweights the two groups to mimic the balance achieved in a randomized trial. After this adjustment, the analytic cohort comprised 1,847 patients on amlodipine and 1,058 on nifedipine, with baseline characteristics aligned across the two groups. The models also adjusted for age, gender, body mass index, admission systolic blood pressure, and the date of symptom onset, guarding against confounding from calendar-time changes in stroke care.

The results pointed in one direction. Patients discharged on amlodipine showed significantly better functional outcomes at five years than those on nifedipine, with an odds ratio of 0.76 for worse mRS scores, a 95 percent confidence interval of 0.65 to 0.89, and a p-value below 0.001. Put more concretely, 51.76 percent of amlodipine users, 956 patients, had at least some residual disability or worse, an mRS score above zero, at five years, compared with 59.74 percent of nifedipine users, 632 patients. The corresponding odds ratio of 0.71, with a confidence interval of 0.61 to 0.83, indicates that the odds of avoiding disability were meaningfully higher in the amlodipine group. Sensitivity analyses, which test whether the findings survive alternative statistical assumptions, produced consistent results throughout.

What could explain such a divergence between two drugs that lower blood pressure through the same basic mechanism? Calcium channel blockers work by blocking L-type calcium channels in vascular smooth muscle, preventing calcium ions from entering these cells and thereby relaxing arterial walls. Nifedipine is a first-generation dihydropyridine with a short half-life in its immediate-release form, producing rapid vasodilation that can trigger reflex activation of the sympathetic nervous system, surges of heart rate, and oscillating blood pressure. Amlodipine, by contrast, has a long half-life of roughly 30 to 50 hours, producing gradual, sustained vasodilation with minimal peak-to-trough fluctuation. For a brain already injured by ischemia, the stability of perfusion pressure may be critical; sharp swings in blood pressure can stress impaired cerebral autoregulation, the mechanism by which brain blood vessels maintain constant flow despite systemic changes. The authors note that these pharmacological differences, along with amlodipine’s longer duration of action and smoother pharmacokinetic profile, could plausibly underlie the observed advantage, though the study design cannot prove causation.

The stakes of this question are enormous. Stroke remains a leading cause of death and long-term disability worldwide, and China bears a particularly heavy burden, with ischemic stroke accounting for the majority of the roughly four million new strokes recorded there each year. Secondary prevention after a stroke or TIA hinges on controlling vascular risk factors, and blood pressure management is the single most powerful modifiable lever. Because amlodipine and nifedipine are among the most commonly prescribed antihypertensives for these patients, and because they are often treated as clinically equivalent members of the same drug class, even a modest real difference in five-year functional outcomes would translate into substantial population-level effects if prescribing patterns shifted toward the better-performing agent.

The authors are careful to frame their findings as hypothesis-generating rather than practice-changing. As an observational study, it inherits the inherent limitations of registry data: physicians chose which drug to prescribe, and despite the statistical balancing achieved through weighting, unmeasured differences between the groups could still have influenced outcomes. The exclusive reliance on registry records limits the granularity of clinical information available, and the loss to follow-up at five years was relatively high, a common challenge in long-term cohort studies that can bias results if attrition differs between groups. These caveats do not erase the signal, which was robust across sensitivity analyses, but they mean the findings stop short of the evidentiary standard that typically guides prescribing decisions.

The path forward, the researchers argue, runs through randomized controlled trials, in which patients with ischemic stroke or TIA would be assigned by chance to amlodipine or nifedipine and followed prospectively. Such trials would eliminate the confounding that shadows any observational comparison and could also probe intermediate mechanisms, including ambulatory blood pressure variability, sympathetic activation, and markers of cerebral small-vessel health. Until then, the study stands as a striking reminder that drugs within the same class are not interchangeable, and that the quiet, long-acting pharmacology of amlodipine may offer stroke survivors something nifedipine does not: a steadier road to recovery measured not in days or months, but in the fullness of five years of life after the brain has been injured.

Subject of Research: Comparative effectiveness of amlodipine versus nifedipine at discharge after ischemic stroke or transient ischemic attack

Article Title: Comparative effectiveness of amlodipine versus nifedipine prescribed at discharge in patients with ischemic stroke or transient ischemic attack: A nationwide 5-year follow-up registry study

Article References: Li, X., Xia, X., Gao, M., Xu, Q., Zhang, Y., Jing, J., Zhai, Y., Meng, X., & Wang, A. (2026). Comparative effectiveness of amlodipine versus nifedipine prescribed at discharge in patients with ischemic stroke or transient ischemic attack: A nationwide 5-year follow-up registry study. PLOS Medicine, 23(10), e1005277. https://doi.org/10.1371/journal.pmed.1005277

Image Credits: AI Generated

DOI: 10.1371/journal.pmed.1005277

Keywords: amlodipine, nifedipine, calcium channel blockers, ischemic stroke, transient ischemic attack, blood pressure, functional outcome, modified Rankin Scale, stroke registry, secondary prevention, pharmacology, observational study