Effects of aspirin on risks of vascular events and cancer according to bodyweight and dose: analysis of individual patient data from randomised trials

Peter M. Rothwell (Lead / Corresponding author), Nancy R. Cook, J. Michael Gaziano, Jacqueline F. Price, Jill F. F. Belch, Maria Carla Roncaglioni, Takeshi Morimoto, Ziyah Mehta

Research output: Contribution to journalArticle

101 Citations (Scopus)
128 Downloads (Pure)

Abstract

Background: A one-dose-fits-all approach to use of aspirin has yielded only modest benefits in long-term prevention of cardiovascular events, possibly due to underdosing in patients of large body size and excess dosing in patients of small body size, which might also affect other outcomes. Methods: Using individual patient data, we analysed the modifying effects of bodyweight (10 kg bands) and height (10 cm bands) on the effects of low doses (≤100 mg) and higher doses (300–325 mg or ≥500 mg) of aspirin in randomised trials of aspirin in primary prevention of cardiovascular events. We stratified the findings by age, sex, and vascular risk factors, and validated them in trials of aspirin in secondary prevention of stroke. Additionally, we assessed whether any weight or height dependence was evident for the effect of aspirin on 20-year risk of colorectal cancer or any in-trial cancer. Results: Among ten eligible trials of aspirin in primary prevention (including 117 279 participants), bodyweight varied four-fold and trial median weight ranged from 60·0 kg to 81·2 kg (p<0·0001). The ability of 75–100 mg aspirin to reduce cardiovascular events decreased with increasing weight (p interaction=0·0072), with benefit seen in people weighing 50–69 kg (hazard ratio [HR] 0·75 [95% CI 0·65–0·85]) but not in those weighing 70 kg or more (0·95 [0·86–1·04]; 1·09 [0·93–1·29] for vascular death). Furthermore, the case fatality of a first cardiovascular event was increased by low-dose aspirin in people weighing 70 kg or more (odds ratio 1·33 [95% CI 1·08–1·64], p=0·0082). Higher doses of aspirin (≥325 mg) had the opposite interaction with bodyweight (difference p interaction=0·0013), reducing cardiovascular events only at higher weight (p interaction=0·017). Findings were similar in men and women, in people with diabetes, in trials of aspirin in secondary prevention, and in relation to height (p interaction=0·0025 for cardiovascular events). Aspirin-mediated reductions in long-term risk of colorectal cancer were also weight dependent (p interaction=0·038). Stratification by body size also revealed harms due to excess dosing: risk of sudden death was increased by aspirin in people at low weight for dose (p interaction=0·0018) and risk of all-cause death was increased in people weighing less than 50 kg who were receiving 75–100 mg aspirin (HR 1·52 [95% CI 1·04–2·21], p=0·031). In participants aged 70 years or older, the 3-year risk of cancer was also increased by aspirin (1·20 [1·03–1·47], p=0·02), particularly in those weighing less than 70 kg (1·31 [1·07–1·61], p=0·009) and consequently in women (1·44 [1·11–1·87], p=0·0069). Interpretation: Low doses of aspirin (75–100 mg) were only effective in preventing vascular events in patients weighing less than 70 kg, and had no benefit in the 80% of men and nearly 50% of all women weighing 70 kg or more. By contrast, higher doses of aspirin were only effective in patients weighing 70 kg or more. Given that aspirin's effects on other outcomes, including cancer, also showed interactions with body size, a one-dose-fits-all approach to aspirin is unlikely to be optimal, and a more tailored strategy is required. Funding: Wellcome Trust and National Institute for Health Research Oxford Biomedical Research Centre.

Original languageEnglish
Pages (from-to)387-399
Number of pages13
JournalLancet
Volume392
Issue number10145
Early online date12 Jul 2018
DOIs
Publication statusPublished - 4 Aug 2018

    Fingerprint

Keywords

  • Age Factors
  • Aged
  • Aspirin/administration & dosage
  • Body Height
  • Body Weight
  • Cardiovascular Diseases/epidemiology
  • Colorectal Neoplasms/epidemiology
  • Death, Sudden/epidemiology
  • Female
  • Humans
  • Kaplan-Meier Estimate
  • Male
  • Platelet Aggregation Inhibitors/administration & dosage
  • Proportional Hazards Models
  • Randomized Controlled Trials as Topic/statistics & numerical data
  • Risk Factors
  • Sex Factors
  • Stroke/prevention & control

Cite this