Blood Pressure Lowering Treatment Trialists' Collaboration. Pharmacological blood pressure lowering for primary and secondary prevention of cardiovascular disease across different levels of blood pressure: an individual participant-level data meta-analysis. Lancet. 2021;397:1625-1636. PMID 33933205¶
One-paragraph summary¶
Individual participant data from 344,716 people in 48 randomised trials of blood-pressure-lowering drugs versus placebo, versus other classes, or more versus less intensive regimens (each with ≥1,000 person-years of follow-up per group; trials in heart failure or acute settings excluded) were analysed stratified by prior cardiovascular disease and by seven baseline systolic categories from <120 to ≥170 mm Hg. Pre-randomisation mean pressures were 146/84 mm Hg in the 157,728 participants with prior cardiovascular disease and 157/89 mm Hg in the 186,988 without; 19.8% and 8.0% respectively had baseline systolic pressure below 130 mm Hg. Over a median 4.15 years, 42,324 participants (12.3%) had a major cardiovascular event. Per 5 mm Hg systolic reduction, the hazard ratio was 0.91 (95% CI 0.89–0.94) without prior cardiovascular disease and 0.89 (0.86–0.92) with it, and there was no reliable heterogeneity by baseline cardiovascular disease status or by baseline systolic category.
Key findings¶
- A fixed degree of pressure lowering delivers a roughly constant relative benefit — about 10% per 5 mm Hg systolic — irrespective of starting pressure or prior disease.
- Incidence of major cardiovascular events per 1,000 person-years: 31.9 (95% CI 31.3–32.5) comparator vs 25.9 (25.4–26.4) intervention in primary prevention; 39.7 (39.0–40.5) vs 36.0 (35.3–36.7) in secondary prevention.
- Relative effects were proportional to the intensity of systolic reduction achieved.
- The authors' explicit clinical corollary: indication should be framed by cardiovascular risk, not by the blood-pressure number, and clinicians should emphasise risk reduction rather than the pressure value itself.
- The companion age-stratified analysis (PMID 34461040) extends this to age, finding benefit persisting to 75–84 years with larger absolute reductions in older groups and recommending removal of age-based thresholds from guidelines.
Limitations¶
- Data come from trials published between 1972 and 2013; contemporary background therapy, event definitions and measurement practices differ.
- Very few participants had baseline systolic pressure below 120 mm Hg, so the lowest stratum is thinly populated and its confidence interval reflects that.
- Trials with heart failure populations and acute settings were excluded by design, so the conclusion does not extend to those.
- Individual-participant meta-analysis of published trials cannot correct for the measurement heterogeneity across contributing studies.
- Because relative effects are constant, absolute benefit varies enormously across the population — the paper does not itself resolve who should be treated, only how to think about it.
Why it matters¶
This is the strongest randomised evidence that the treatment threshold is a policy choice rather than a biological boundary, and it is the analytic foundation for risk-based rather than pressure-based treatment decisions. Every guideline argument about where to set a diagnostic line is downstream of this result. It also constrains the interpretation of the target trials: if benefit is proportional to reduction achieved and independent of starting point, then SPRINT-style results are not surprising, and disagreement between ACCORD-BP and later diabetes trials is more plausibly about power and event rates than about a biological interaction.
Cited by wiki pages¶
- overview
- risk and outcomes
- blood-pressure targets