Fish oil in knee osteoarthritis: a randomised clinical trial of low dose versus high dose

Randomized, 1:1, parallel-group clinical trial conducted at three Australian centres. Adults older than 40 years with knee osteoarthritis and regular knee pain were assigned to low-dose versus high-dose fish oil and followed for 24 months.

Participants were randomized to daily oral fish oil for 24 months. The high-dose arm received 15 mL/day fish oil supplying 4.5 g EPA+DHA per day (EPA 18%, DHA 12%); the low-dose arm received a 1:9 blend of fish oil and higholeic sunola oil supplying 0.45 g EPA+DHA per day, equivalent to about 1.5 standard 1 g fish oil capsules daily. Both arms had a 4-week run-in with citrus-flavoured sunola oil at 15 mL/day.

High-dose fish oil did not outperform low-dose fish oil in knee osteoarthritis and was not associated with better structural outcomes. At 2 years, the high-dose group had worse WOMAC pain and function than the low-dose group in the intention-to-treat analysis, with between-group differences of 3.3 for pain (p=0.009) and 8.5 for function (p=0.032); the per-protocol analyses were also significant in favor of low-dose fish oil. MRI outcomes showed no meaningful between-group difference in cartilage volume loss or bone marrow lesion change over 2 years. Adverse events were generally similar, but high-dose fish oil caused more gastrointestinal intolerance and higher discontinuation, and it was associated with greater weight gain.

The trial compared two active fish oil doses rather than fish oil versus placebo, so it cannot determine whether either dose is superior to no supplementation. Structural MRI analyses included only subsets of participants, and differential discontinuation from gastrointestinal intolerance in the high-dose arm may have affected longer-term estimates. Generalizability is limited to adults with symptomatic knee osteoarthritis recruited from Australian centres.