INDEPENDENT THINKING. WHOLE-FOOD LIVING.NO ANIMALS / NO OIL
THENANODIET
← Blood pressure / Health Desk

Cocoa and blood pressure: a small effect with important trial caveats

A Cochrane review found a short-term reduction of about 2 mmHg, not evidence of fewer cardiovascular events.

Whole-food ingredients prepared without added oil

Illustrative food image, not a photograph of the study intervention.

A review of products, not one recipe

The 2017 Cochrane review included 35 trials with 40 treatment comparisons of cocoa or chocolate products against low-flavanol products or placebo. In total, the systolic analysis involved 1,804 mainly healthy participants. Trials lasted two to 18 weeks, averaging nine weeks. Active products supplied 30–1,218 mg of flavanols daily in 1.4–105 g of cocoa products, a very wide range of interventions. These details matter because the result is not a test of one standard cocoa recipe. Nor does the abstract establish that a familiar serving of any particular retail product reproduces the interventions used across this varied collection of trials.

Ried K, Fakler P, Stocks NP. The Cochrane database of systematic reviews, 2017. PMID 28439881 · DOI

The average effect was modest

The pooled systolic difference was −1.76 mmHg, with a 95% confidence interval from −3.09 to −0.43. The diastolic difference was also −1.76 mmHg, with an interval from −2.57 to −0.94; that analysis involved 39 comparisons and 1,772 participants. The authors summarized this as a small effect of about 2 mmHg. They rated the evidence moderate rather than high quality because unexplained heterogeneity remained. The result therefore supports a modest short-term average, not a dramatic or universal response. Preserving the effect size, duration and evidence rating gives readers a more useful picture than describing cocoa simply as a blood-pressure treatment.

Ried K, Fakler P, Stocks NP. The Cochrane database of systematic reviews, 2017. PMID 28439881 · DOI

Subgroups did not settle who benefits

Trials involving hypertensive participants showed a systolic reduction of about 4 mmHg, whereas trials involving normotensive participants did not show a significant difference. However, the test comparing baseline-pressure subgroups was only borderline, with P = 0.08. The authors explicitly cautioned that subgroup findings needed confirmation through direct randomized comparisons. That means the apparent pattern is a research lead, not a demonstrated rule for selecting responders. There was also a nonsignificant trend toward larger reductions in unblinded trials than double-blind trials. Neither observation removes the review's unresolved heterogeneity or justifies presenting a particular subgroup estimate as a reliable individual forecast.

Ried K, Fakler P, Stocks NP. The Cochrane database of systematic reviews, 2017. PMID 28439881 · DOI

Sponsorship and longer-term uncertainty

When the reviewers excluded trials with authors employed by the sponsoring industry, the estimated effect became slightly smaller. The abstract identifies this as an indication of some reporting bias. It also reports gastrointestinal complaints and nausea in about 1% of active-group participants versus 0.4% of controls. These details belong alongside the favorable pressure result rather than outside the story. Most importantly, the review called for long-term trials of clinical outcomes and chronic adverse effects. Its two-to-18-week pressure data do not establish fewer heart attacks or strokes. The supported conclusion remains a small, moderately certain short-term effect across heterogeneous cocoa-product trials.

Ried K, Fakler P, Stocks NP. The Cochrane database of systematic reviews, 2017. PMID 28439881 · DOI