Current issues of ACP Journal Club are published in Annals of Internal Medicine


Acetaminophen, high-dose ibuprofen, and low-dose ibuprofen were equally effective for symptoms of osteoarthritis of the knee

ACP J Club. 1991 Nov-Dec;115:78. doi:10.7326/ACPJC-1991-115-3-078

Source Citation

Bradley JD, Brandt KD, Katz BP, Kalasinski LA, Ryan SI. Comparison of an antiinflammatory dose of ibuprofen, an analgesic dose of ibuprofen, and acetaminophen in the treatment of patients with osteoarthritis of the knee. N Engl J Med. 1991 Jul 11;325:87-91.



To compare the short-term risks and benefits of acetaminophen, low-dose ibuprofen, and high-dose ibuprofen for the symptomatic treatment of osteoarthritis of the knee.


Randomized, double-blind, controlled, 4-week trial.


Patients were recruited from the medicine, rheumatology, and orthopedic clinics of a university health center and from the community.


Patients were ≥ 30 years of age, with knee pain and radiographic evidence of mild or moderate osteoarthritis. Patients were excluded if they had myofascial, neurologic, or vascular disease affecting the legs, or underlying inflammatory arthritis; a contraindication to the study medications; a history of trauma, surgery, or corticosteroid injection of the knee within 3 months; or the inability to walk unassisted. Of the 195 patients who were randomized, 184 (94%) were included in the intention-to-treat analysis.


All patients had a washout period of 3 to 7 days. "Rescue analgesia" (with propoxyphene napsylate) was permitted. 61 patients were assigned to acetaminophen, 4000 mg/d; 62 patients to low-dose ibuprofen, 1200 mg/d; and 61 patients to high-dose ibuprofen, 2400 mg/d. 2 tablets were to be taken 4 times/d after meals.

Main outcome measures

The Health Assessment Questionnaire (HAQ) disability score and expanded pain scales; "50-ft-walk time"; global symptomatic change.

Main results

144 patients (78%) completed 4 weeks of treatment; 12% were < 75% compliant, and 9% discontinued treatment for adverse events or for uncontrolled pain. All 3 groups showed improvement from baseline scores on all the scales. Pain at rest decreased for both ibuprofen groups compared with the acetaminophen group (P = 0.05), but there were no other differences among the groups. 14 of 61patients (23%) in the high-dose ibuprofen group had gastrointestinal side effects (including 2 patients with occult blood in their stools) compared with 7 of 62 patients (11%) in the low-dose ibuprofen group {absolute risk difference 12%, 95% CI -2% to 25%, P = 0.09}* and 10 of 61 patients (16%) on acetaminophen {absolute risk difference 7%, CI -8% to 21%, P = 0.36}.*


Short-term changes in pain and disability did not differ for patients with knee osteoarthritis assigned to acetaminophen, high-dose ibuprofen, or low-dose ibuprofen.

Source of funding: National Institute of Arthritis and Musculoskeletal and Skin Diseases.

Address for article reprint: Dr. K.D. Brandt, Rheumatology Division, Indiana University School of Medicine, 541 Clinical Drive, Room 492, Indianapolis, IN 46202-5103, USA.

*Numbers calculated from data in article.


This study addresses an important question because nonsteroidal anti-inflammatory drugs, sometimes in high doses, are often prescribed for osteoarthritis. It was found that over a 4-week period, acetaminophen, low-dose ibuprofen, and high-dose ibuprofen were similarly effective in reducing pain and improving physical function in mild-to-moderate, symptomatic osteoarthritis.

Care must be used when prescribing either acetaminophen or ibuprofen for patients with osteoarthritis, because these patients are often elderly, with comorbid conditions, and associated renal or hepatic dysfunction or both. Even in this short-term study there was a trend toward greater gastrointestinal side effects and an increase in serum creatinine concentration with high-dose ibuprofen. Given the lower costs and toxicity of acetaminophen, these results suggest that it may be preferred first-line therapy for mild-to-moderate osteoarthritis. Patients on acetaminophen, however, did have an increase in serum aspartate aminotransferase concentration. The results also suggest that if nonsteroidal anti-inflammatory drugs, such as ibuprofen, are prescribed, a lower analgesic dose regimen may be a reasonable altemative to higher doses. The availability and lower cost of over-the-counter ibuprofen (and also possibly naproxen in the future) make lower-dose therapy an even more attractive alternative to high-dose therapy.

As the authors suggest, caution must be used when extrapolating these results. The long-term risks and benefits of these 3 regimens are more difficult to compare. Although long-term use of low-dose or high-dose ibuprofen may be associated with increased toxicity, acetaminophen may not be as effective. The present study showed a trend toward greater improvement in the walking-pain subscale score of the HAQ with high-dose and low-dose ibuprofen compared with acetaminophen. The comparable efficacy of long-term acetaminophen needs to be shown, particularly in patients with more severe disease or a more marked inflammatory component.

C. Kent Kwoh, MD
University of PittsburghSaint Margaret Memorial HospitalPittsburgh, Pennsylvania, USA