Is dark chocolate good for you? What the trials and the metals show

Dark chocolate is neither a health food nor junk. The strongest test of its heart claims — a randomised trial in 21,442 older adults — used a concentrated cocoa flavanol extract, not chocolate, and it missed its main cardiovascular target [s1]. The blood-pressure effect that gets quoted is real but small, on the order of under 2 mmHg [s2], and it comes from flavanol doses a chocolate bar delivers unreliably. Set against that are the bar's sugar and calories, and contamination: in one multi-year analysis, 43 percent of dark chocolate products exceeded a strict lead limit [s3].

The flavanol claim, tested properly

Most of the good-news headlines about dark chocolate rest on the flavanols in cocoa, plant compounds that improve blood-vessel function in short, small studies. COSMOS was built to see whether that translates into fewer cardiovascular events. It randomised 21,442 US adults — 12,666 women aged 65 and older and 8,776 men aged 60 and older, free of major cardiovascular disease — to a cocoa extract supplement delivering 500 mg of flavanols a day, including 80 mg of (-)-epicatechin, or placebo [s1].

Over a median 3.6 years, the primary outcome — a composite of confirmed total cardiovascular events — occurred in 410 participants on cocoa extract and 456 on placebo, a hazard ratio of 0.90 (95% confidence interval 0.78 to 1.02, P = 0.11) [s1]. That crosses 1.0: the trial did not show a significant reduction. A secondary endpoint, cardiovascular death, was lower on cocoa extract (hazard ratio 0.73, 95% CI 0.54 to 0.98), and all-cause mortality was 0.89 (0.77 to 1.03) [s1]. The honest reading is the one the authors gave: cocoa extract "did not significantly reduce total cardiovascular events" but reduced cardiovascular death by 27 percent, a secondary finding that needs confirmation [s1]. And the intervention was a standardised extract, not a chocolate bar.

Blood pressure: real, small, dose-dependent

Cocoa's clearest measured benefit is on blood pressure, and a Cochrane review put a number on it. It pooled 35 trials — 40 treatment comparisons in 1,804 mostly healthy participants — giving cocoa products containing 30 to 1,218 mg of flavanols (mean 670 mg) in 1.4 to 105 grams of product per day [s2]. The pooled effect was a small but statistically significant reduction: systolic blood pressure fell by a mean of 1.76 mmHg (95% CI -3.09 to -0.43, P = 0.009) and diastolic by 1.76 mmHg (-2.57 to -0.94, P < 0.001) [s2].

Baseline mattered. Systolic pressure fell by 4 mmHg in people with hypertension (9 comparisons, 401 participants) and tended lower in those with prehypertension, but there was no significant effect in people with normal blood pressure (23 comparisons, 1,063 participants) [s2]. Two points follow. The average effect is modest even where it exists, and the flavanol doses used ranged up to 1,218 mg — a spread of 1.4 to 105 grams of product [s2] — because the flavanol content of chocolate is variable. A square of supermarket dark chocolate is not a calibrated 670 mg dose.

The contamination nobody markets

The downside dark chocolate's marketing omits is heavy metals. Cocoa plants take up cadmium from soil, and lead contamination is added largely after harvest. A study analysed 72 consumer cocoa-containing products bought across four rounds between 2014 and 2022, measured against California's Proposition 65 maximum allowable dose levels [s3]. Forty-three percent of products exceeded the Prop 65 level for lead and 35 percent for cadmium; none exceeded it for arsenic [s3].

The reassuring counterweight is in the same paper. Measured against the US Food and Drug Administration's interim reference level for lead, 97.2 percent — 70 of 72 products — fell below the limit, and the median concentration of each metal was below even the conservative Prop 65 levels, meaning the exceedances were driven by outlier products [s3]. The authors' read: contamination in more than half the products may pose no appreciable risk for the average person eating a single serving, but eating high-outlier products, or more than one serving a day alongside other dietary sources, could add up to exposure over the Prop 65 levels [s3]. One counterintuitive finding — "organic" products were significantly more likely to show higher cadmium and lead [s3].

So, good for you?

The defensible summary is that dark chocolate is a treat with a modest, dose-dependent effect on blood pressure and no proven reduction in cardiovascular events when the flavanols are delivered as a bar rather than a measured extract [s1][s2]. It also carries sugar, fat and calories, and a real but usually small heavy-metal load that is worth knowing about for anyone eating it daily or giving it to children [s3]. For blood pressure specifically, the interventions with better evidence are dietary patterns like the DASH approach, and for heart risk the food with the strongest cohort and trial backing is nuts, not chocolate. The pattern here is the same as for coffee: a food with a plausible mechanism and a small measured effect, oversold well beyond what the trials support.

This article is informational and is not medical advice.

Sources

• Effect of cocoa flavanol supplementation for the prevention of cardiovascular disease events: the COSMOS randomized clinical trial — The American Journal of Clinical Nutrition, 2022
• Effect of cocoa on blood pressure — Cochrane Database of Systematic Reviews, 2017
• A multi-year heavy metal analysis of 72 dark chocolate and cocoa products in the USA — Frontiers in Nutrition, 2024