Omega-3 from a glass of chocolate soymilk instead of a capsule: it reached the blood in 21 adults, but 'as good as the capsule' is a stretch

EPA and DHA are the two long-chain omega-3 fats that most people get from oily fish, and most people don't eat much oily fish. Capsules are the usual workaround. A team in the Department of Food Science and Technology at The Ohio State University wanted to know whether the same fats could ride in an ordinary drink instead, and chose soymilk, partly because it holds an oil emulsion well and partly because chocolate soymilk can hide the taste of algae oil. Their trial was published in Food & Function on 8 October 2026.

They screened 51 volunteers, randomised 23 and ended with 21: healthy adults with an average age of about 28, roughly half of them women, none of whom had taken omega-3 supplements in the previous three months. Everyone first spent two weeks drinking a plain chocolate soymilk with no algae oil in it. Then 10 people drank 12 US fluid ounces a day of the fortified version (200 mg of EPA and 320 mg of DHA), and 11 took two algae oil capsules a day (230 mg of EPA and 380 mg of DHA) for four weeks. All were asked to keep fish and other marine omega-3 out of their diet. Blood was drawn after an overnight fast at the start, at two weeks and at four weeks, and the team measured EPA and DHA as a share of the fats in red blood cells, in plasma and in a type of white blood cell.

Both routes worked. EPA and DHA were up in all three parts of the blood after two weeks and higher again after four, in both groups. People took what they were given: about 96% of the drinks and 97% of the capsules were consumed. The soymilk scored 6.0 out of 9 for overall liking on the first day and 6.1 on the last, which on that scale is 'like slightly', and its fishy-flavour rating stayed at 2 or below on a 10-point line. Some people reported nausea, gas, diarrhoea or bloating, mostly during the two run-in weeks on plain soymilk, and the paper says these had settled by the time the fortified drink began. There were no serious adverse events.

Now the comparison, which is the point of the trial. For EPA the two groups rose in step (P = 0.35). For DHA they did not: red blood cell DHA and plasma DHA both rose more in the capsule group (P = 0.01 for each). The authors point out that the capsules turned out to carry a little more of both fats than the drink, so they re-ran the analysis as change in blood level per milligram actually consumed, counting any stray fish people ate as well. On that basis the gap in red cell DHA was no longer statistically significant, with P = 0.08, and the abstract concludes that the soymilk is 'a comparable food-based alternative to traditional capsules'.

From here the reading is ours. 'No significant difference' and 'the same' are different claims, and this trial was not built to make the second one. The sample size was worked out to catch a 20% difference in red cell DHA, so anything smaller could slip through unseen, and a P of 0.08 from 21 people leans toward the capsule more than it rules a gap out. The dose correction is also the authors' second analysis; the model they set out with found a DHA difference. There are other reasons for caution in the methods: everyone knew which product they had, the paper states that allocation was not concealed, and the soymilk group started with a higher average BMI (25.6 against 22.8). The team's own models found that heavier people gained less EPA in their red cells, which may be part of why the drink group lagged, or may not. Nobody can tell from 21 people.

What the study does show is modest and still useful: algae oil mixed into soymilk gets into the blood, people will drink it daily for a month, and it doesn't taste of fish. That matters for vegetarians and vegans, for whom algae is the only direct source of EPA and DHA. What it does not show is a health effect of any kind. Nothing but blood fats was measured, and four weeks is too short even for those to level off, as the authors say; red blood cells take about four months to turn over. This is also a product the team developed, and nothing you can buy. The work was supported by the university's cancer centre and the US National Institutes of Health, and partly funded by the Ohio Soybean Council, which the paper says had no role in the product, the study design, the analysis or the writing. The authors declare no conflicts of interest.