The beetroot juice claim has an unusual property: the pooled number looks impressive and the reason it looks impressive is that somebody measured it in the wrong place. Two meta-analyses, published eleven years apart, describe the same effect — and then describe, in different words, why it is smaller than it appears.
The 2024 analysis: 5,31 mmHg, graded low
Grönroos and colleagues searched PubMed, SCOPUS, Medline Ovid, Cinahl, the Cochrane Library and Web of Science from inception to 13 April 2024, for randomised controlled trials comparing beetroot juice against placebo, water or no intake. Risk of bias was assessed with a Swedish agency instrument based on the Cochrane tool; pooling used random-effects models with meta-regression, and certainty was graded with GRADE.
The population was specific: adults with hypertension under European Society of Hypertension definitions, clinical BP ≥140/≥90 mmHg. Eleven trials, 349 patients.
The result:
- clinical systolic BP: −5,31 mmHg (95% CI −7,46 to −3,16), I² = 64 %, GRADE ⊕⊕○○
- clinical diastolic BP: no significant effect
- 24-hour BP: no significant effect
Read those last two lines again. The effect was in clinic systolic pressure and nowhere else — not in diastolic, and not in 24-hour monitoring. The authors rate certainty low and say the results should be interpreted with caution. The amount they describe as potentially effective is 200 to 800 milligrams of nitrate daily, with no sign of tolerance developing.
The 2017 analysis: same effect, same problem, spelled out
Ashor, Lara and Siervo asked a narrower question: what happens over more than a week? Thirteen trials, 325 participants, 7 to 65 per study, interventions of 1 to 6 weeks.
- systolic: −4,1 mmHg (−6,1 to −2,2), p < 0,001
- diastolic: −2,0 mmHg (−3,0 to −0,9), p < 0,001
Both significant. Then the sentence that matters:
Ten trials measured blood pressure at rest in a clinic. Six used 24-hour ambulatory monitoring. Three used daily home monitoring. The effect was only significant in resting clinical settings. No significant changes were observed with 24-hour ambulatory monitoring or with daily home monitoring.
The authors’ conclusion is unusually blunt for this literature: positive effects of medium-term dietary nitrate supplementation were observed only in clinical settings, and were not corroborated by more accurate methods.
Why the clinic and the monitor disagree
This is the part that gets skipped in every summary. A clinic reading is a measurement of a person sitting still in a building where they expect to have their blood pressure taken. Ambulatory monitoring is the same person, awake, moving, at home, across a day.
If a supplement moves a number that only moves when a person is sitting in a clinic, the honest description is that it moves a reading, not necessarily a physiology. That does not make it worthless. It makes it smaller than the pooled figure implies, and it changes who should be told to expect something.
The two meta-analyses are not in conflict. The 2024 one reports the clinic effect and reports the absence of the 24-hour effect in the same paragraph. The 2017 one explains why. Read together they say: roughly 4 to 5 mmHg systolic, in a clinic, with the authors of both papers declining to call the evidence strong.
The same question — who checked this number, and does the document behind it actually say what the label claims — is what the purity side of this industry runs into. The assay lines worth reading first walks through what an assay certificate should contain, what the percentage on the front of a vial is not, and what to do when the numbers in the report do not add up. Different substance, same problem: a measurement standing in for a physiology it cannot establish.
The 2013 analysis, and what it adds
Siervo and colleagues pooled 16 trials, all crossover designs, run between 2006 and 2012, with 254 participants, 7 to 30 per study, and interventions of 2 hours to 15 days.
- systolic: −4,4 mmHg (−5,9 to −2,8), p < 0,001
- diastolic: −1,1 mmHg (−2,2 to 0,1), p = 0,06
The diastolic result does not reach significance, and the authors say so plainly in the abstract by printing the p-value. They also report that meta-regression found an association between daily inorganic nitrate amount and systolic change (p < 0,05) — an amount-response signal — and then state that these findings need to be tested in long-term trials and in people at greater cardiovascular risk.
That was 2013. The 2024 analysis is partly the answer to that sentence, and the answer is: the effect holds up, the certainty is low, and the diastolic claim still is not there.
What this literature does not show
Not a 5 mmHg reduction in a person’s daily blood pressure. The 2024 analysis found nothing significant on 24-hour monitoring, and the 2017 analysis says the same in its conclusion.
Not a cardiovascular outcome. No trial here measured a heart attack, a stroke or a death. Blood pressure is a surrogate, and a surrogate measured in a clinic is a surrogate of a surrogate.
Not an effect on diastolic pressure in the modern analysis. Two of the three pooled analyses put diastolic at or past the line of significance, and the 2024 one — the largest and the most recent — found no significant effect at all.
Not tolerance-free indefinitely. The 2024 authors found no sign of tolerance developing over the included durations. That is not the same as a statement about years.
The sentence worth keeping
Four to five millimetres of systolic blood pressure, measured in a clinic, with the 24-hour evidence absent, the diastolic evidence absent, the certainty graded low by the authors, and a research programme that has not yet produced a single outcome trial. That is a real effect with an honest size attached to it.
