A large network meta-analysis published in JAMA has ranked antihypertensive drug combinations by tolerability, finding that regimens containing angiotensin II receptor blockers (ARBs) were the best tolerated, with lower rates of treatment discontinuation due to side effects than even placebo. The analysis involved 159,362 participants across 716 double-blind randomised controlled trials of at least four weeks duration.
Approximately 1.4 billion people were living with hypertension globally in 2024, according to the WHO, yet fewer than one in five have the condition adequately controlled. Poor tolerability is a recognised barrier to adherence, with studies estimating that between 30% and 80% of patients newly prescribed antihypertensive medication discontinue treatment within the first year. Common reasons include headaches, dizziness, peripheral oedema, and cough.
Using network meta-analysis and SUCRA (Surface Under the Cumulative Ranking Curve) statistical modelling, researchers were able to rank treatments that had not been directly compared in the same trials. ARBs featured in four of the five best-tolerated treatment options. The combination of an ARB with a calcium channel blocker (CCB) was the most tolerated pairing overall. CCBs used as monotherapy were significantly more likely to lead to treatment discontinuation. Combinations of a beta-blocker with a diuretic were also associated with higher discontinuation rates.
An additional finding was that most antihypertensive agents, except CCBs, were associated with fewer headaches than placebo, possibly because CCBs promote cerebral vasodilation, which may itself trigger headache.
The authors note that current pharmacological imaging tools do not provide sufficient resolution to make this distinction, and suggest that SSP could inform the design of future trials by identifying structural barriers to drug efficacy earlier in the development process.
Source: Wang N et al. Adverse Effects and Treatment Discontinuation of Blood Pressure-Lowering Drugs and Combinations. JAMA (2026). DOI: 10.1001/jama.2026.6214
