Is fish oil worth buying?

For living longer, probably not. Across 45 trials of 143,693 adults, more fish-type omega-3, mostly as pills, made little or no difference to the risk of dying, and the reviewers rated that evidence high certainty (Abdelhamid, 2020). Heart and circulation events showed no clear drop either. It may slightly lower deaths from coronary heart disease, on uncertain evidence, and it lowers triglycerides, a blood fat (Abdelhamid, 2020). The review didn't weigh cost. Pregnancy is a separate case, below (Middleton, 2018). Every number links to its source.

The short version

• In 45 trials of 143,693 adults, people given more fish-type omega-3 had a 3% lower risk of dying than people on their usual or a lower intake, which could be chance (Abdelhamid, 2020).
• Heart and circulation events had a 4% lower risk and strokes a 2% higher risk, and both could be chance (Abdelhamid, 2020).
• Deaths from coronary heart disease may have been slightly lower, a 10% lower risk, but the reviewers rated that evidence low certainty, so the drop may be smaller or none (Abdelhamid, 2020).
• Triglycerides fell by about 15%, more at higher doses, and the reviewers rated that high certainty (Abdelhamid, 2020).
• In a separate review of pregnancy, omega-3 as supplements or foods cut births before 37 weeks from 13.4% of women without it to 11.9% with it, across 26 trials with 10,304 women (Middleton, 2018).

The review of 86 trials

Abdelhamid and colleagues gathered 86 trials of 162,796 adults at varying risk of heart disease, mainly in high-income countries, each lasting 12 to 88 months. Most tested fish-type omega-3 (EPA and DHA) as pills, and some used foods rich in it or diet advice, compared with dummy pills, usual diet or a lower intake. Doses of fish-type omega-3 ran from 0.5 g a day to more than 5 g a day. A smaller set of trials tested plant omega-3 (ALA) and is reported separately (Abdelhamid, 2020).

Compared with usual or less omega-3, more fish-type omega-3 came with (Abdelhamid, 2020):

• a 3% lower risk of dying from any cause, which could be chance, in 45 trials that counted deaths, with 143,693 people, 11,297 of whom died (high certainty)
• an 8% lower risk of dying from heart and circulation disease, in 29 trials of 117,837 people, which the reviewers still counted as little or no effect (moderate certainty)
• a 4% lower risk of heart and circulation events, which could be chance, in 43 trials of 140,482 people (high certainty)
• a 2% higher risk of stroke, which could be chance, in 31 trials of 138,888 people (moderate certainty)
• a 10% lower risk of dying from coronary heart disease, in 24 trials of 127,378 people (low certainty, so the drop may be smaller or none)

The review found no difference in these effects between bigger and smaller doses or between shorter and longer trials. Triglycerides were the exception: they fell by about 15%, more at higher doses. Body fat, other blood fats and blood pressure showed little or no change, and so did serious medical events recorded during the trials (Abdelhamid, 2020).

Pregnancy

A separate 2018 Cochrane review tested omega-3 in pregnancy, as supplements or foods, against dummy pills or no added omega-3. Births before 37 weeks fell from 13.4% of women without it to 11.9% with it, across 26 trials with 10,304 women, and births before 34 weeks from 4.6% to 2.7%, across 9 trials with 5,204 women. The reviewers rated both high quality. Pregnancies running past 42 weeks probably rose, from 1.6% to 2.6% (Middleton, 2018).

How far to trust these numbers

• The review pooled randomised trials, but how sure the reviewers were differs by outcome. The finding on dying from any cause was rated high certainty; the coronary heart disease finding was low certainty. 28 of the 86 trials were at low risk of bias (Abdelhamid, 2020).
• The review found little evidence on eating fish, so it can't say whether oily fish protects the heart (Abdelhamid, 2020).
• The review's searches ended in 2019, so newer trials aren't counted (Abdelhamid, 2020).
• Triglycerides are a blood test. In a separate analysis in the same review, heart and circulation events showed no clear drop (Abdelhamid, 2020).

Sources

• Abdelhamid et al. Omega-3 fatty acids for the primary and secondary prevention of cardiovascular disease. Cochrane Database Syst Rev, 2020.
• Middleton et al. Omega-3 fatty acid addition during pregnancy. Cochrane Database Syst Rev, 2018.