Is long term paracetamol effective for arthritis knee pain

The short version: paracetamol offers a tiny, clinically meaningless benefit for knee arthritis pain, and the evidence for long-term use is essentially absent.

The best evidence on paracetamol for osteoarthritis comes from a 2015 systematic review in the BMJ [1]. It pooled data from randomized placebo-controlled trials and found that paracetamol produced a pain reduction so small it was clinically insignificant. The effect size was 0.21, which means it barely beat placebo. For context, NSAIDs like ibuprofen or meloxicam produce a pain reduction about twice as large, around 10 mm on a 100 mm scale versus paracetamol's 4.7 mm [5]. The review also found that paracetamol increased the risk of abnormal liver function tests [1]. A 2011 n-of-1 trial comparing celecoxib to sustained-release paracetamol found that most patients could not tell the difference between the two, but celecoxib performed slightly better overall [6]. A 2021 trial found that a bioavailable turmeric extract was equivalent to paracetamol for WOMAC pain and function scores after six weeks, and had fewer adverse events [4]. That means a spice extract matched the drug.

The evidence for long-term use is almost nonexistent. The trials that exist run for weeks, not months or years. The 2015 BMJ review is the highest-quality source, and it found no evidence that paracetamol changes the course of the disease or provides sustained benefit [1]. The 2011 study on ibuprofen/paracetamol combinations found modest short-term benefits but also a significant drop in hemoglobin in the combination group, suggesting hidden bleeding [2]. The 2018 narrative review on NSAIDs explicitly states that NSAIDs are clearly superior to paracetamol after four weeks [5]. So the drug does not work well, and we have no idea what happens if you take it every day for a year.

The mechanism question matters here. Paracetamol is a weak COX inhibitor with no anti-inflammatory effect. It does not address the underlying pathology of osteoarthritis, which involves mechanical wear, low-grade inflammation, and cartilage degradation. Taking it long term does nothing to slow the disease. Meanwhile, the liver toxicity risk is real, especially at doses above 3 grams per day, and the risk of gastrointestinal bleeding, though lower than with NSAIDs, is not zero. A 2011 commentary even argued that paracetamol's apparent effect in trials may be partly due to its use as a rescue medication, inflating the placebo response [3]. The drug may be doing less than the trials claim.

My call: paracetamol is not worth taking long term for knee arthritis. The benefit is too small to feel, the long-term safety data does not exist, and safer, more effective options exist, including physical therapy, NSAIDs used carefully, and even turmeric extract. Confidence: high.