8 NOV 2026 — Two Brazilian trials tested the same pair of systolic blood-pressure goals and got different results. In both, the question was whether to push systolic pressure below 120 mm Hg rather than below 140. Among 4,368 people who had already had an ischaemic stroke or a transient ischaemic attack, the tighter target cut major cardiovascular events by about 21%. Among 9,476 adults with type 2 diabetes, it produced no statistically significant benefit. Both studies were presented on 7 November at the American Heart Association's Scientific Sessions in Chicago and published the same day in the New England Journal of Medicine.
The stroke trial, OPTIMAL-STROKE, ran at 30 centres in Brazil and followed patients for an average of 3.9 years. Over that time 236 people on intensive treatment (10.9%) had a cardiovascular event, against 297 (13.5%) on standard care, and recurrent strokes fell from 6.4% to 4.7%. The diabetes result was closer. In OPTIMAL-DIABETES, which enrolled at 33 sites, 286 patients (6.0%) on intensive treatment had an event over a median of four years, compared with 315 (6.7%) on standard care. That is a risk about 10% lower, short of significance. Kidney-related events went the other way, with 270 patients (5.7%) affected on intensive treatment against 175 (3.7%) on standard.
Six trials of the same two targets
At least four earlier trials compared a systolic goal below 120 with one below 140. The table sets the two OPTIMAL trials beside them, using each paper's own abstract for the older four and the heart association's releases for the new pair.
| Trial (journal, year) | Who was enrolled | Patients | Where | Systolic reached, intensive vs standard (mm Hg) | Primary outcome, intensive vs standard | Significant? |
|---|---|---|---|---|---|---|
| ACCORD-BP (NEJM, 2010) | Type 2 diabetes, high cardiovascular risk | 4,733 | US, Canada | 119.3 vs 133.5 at 1 year | 1.87% vs 2.09% a year; hazard ratio 0.88 | No (P=0.20) |
| SPRINT (NEJM, 2015) | High cardiovascular risk, no diabetes | 9,361 | US | 121.4 vs 136.2 at 1 year | 1.65% vs 2.19% a year; hazard ratio 0.75 | Yes; stopped early |
| ESPRIT (Lancet, 2024) | High cardiovascular risk, including 4,359 with diabetes and 3,022 with a previous stroke | 11,255 | China | 119.1 vs 134.8 during follow-up | 9.7% vs 11.1%; hazard ratio 0.88 | Yes (p=0.028) |
| BPROAD (NEJM, 2025) | Type 2 diabetes, aged 50 or older | 12,821 | China | 121.6 vs 133.2 at 1 year | 1.65 vs 2.09 events per 100 person-years; hazard ratio 0.79 | Yes (P<0.001) |
| OPTIMAL-DIABETES (NEJM, 2026) | Type 2 diabetes, aged 50 or older | 9,476 | Brazil | 122 vs 135 during follow-up | 6.0% vs 6.7%; about 10% lower risk | No |
| OPTIMAL-STROKE (NEJM, 2026) | Previous ischaemic stroke or TIA | 4,368 | Brazil | Not given in the release | 10.9% vs 13.5%; about 21% lower risk | Yes |
The rows are not directly comparable. Each trial defined its composite outcome a little differently (ESPRIT counted revascularisation, for example), measured blood pressure at different points and ran for different lengths of time. The OPTIMAL figures come from conference releases, which give approximate relative reductions rather than hazard ratios. Sources for the older four are the abstracts of ACCORD-BP, SPRINT, ESPRIT and BPROAD; the countries for the two North American trials are from their ClinicalTrials.gov records.
Small absolute gaps, and a kidney cost
Relative reductions flatter. Working from the release figures, we put the stroke trial's absolute gap at 2.6 percentage points over roughly four years (13.5 minus 10.9), or about 26 fewer people in every 1,000 having an event. The gap for recurrent stroke alone is smaller, at 1.7 points (6.4 minus 4.7).
In the diabetes trial the cardiovascular gap is 6.7 minus 6.0, or 0.7 points, and it missed statistical significance. The kidney gap runs the other way and is larger, at 5.7 minus 3.7, or 2.0 points, nearly three times the size (2.0 divided by 0.7 is about 2.9). The two are not equal harms. The release says the kidney events were mostly declines in kidney function, end-stage renal disease was similar between groups, and serious adverse events overall were similar too.
The stroke release also reports more kidney-related events with intensive treatment, without giving numbers. Gisele Sampaio, the stroke trial's principal investigator and a neurology professor in São Paulo, said reaching a lower reading safely "requires appropriate medical supervision and additional monitoring, particularly of kidney function".
Diabetes remains the contested group
Three trials now sit in the diabetes rows. ACCORD-BP and OPTIMAL-DIABETES both pointed to lower risk and missed significance; BPROAD, the largest, reported a clear benefit. ESPRIT, which enrolled people with and without diabetes, found no difference in effect by diabetes status. Every result in the table leans the same way, towards fewer events with the tighter target. They differ on whether the gap was big enough to rule out chance.
Otavio Berwanger, who chairs the steering committee for both OPTIMAL trials, interpreted the null result that way. He said it suggests "any cardiovascular benefit may be smaller than that observed in individuals without Type 2 diabetes", and that a target below 120 "should not necessarily be applied universally to all patients".
Both OPTIMAL trials have limits the authors name themselves. Because the diabetes trial was open-label, patients and doctors knew which arm they were in, and in the stroke trial not every intensive-arm patient reached the target. Both releases also caution that results from Brazil may not carry over to other health systems. The current 2025 American Heart Association and American College of Cardiology guideline already recommends keeping systolic pressure below 130 and encourages below 120 for people at higher cardiovascular risk.
Anyone already being treated for high blood pressure should not change that treatment on the strength of these trials without first speaking to their doctor.
View author profile → · Editorial policy
