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


Running was not associated with increased progression or incidence of osteoarthritis of the knee or spine

ACP J Club. 1993 Sept-Oct;119:57. doi:10.7326/ACPJC-1993-119-2-057

Source Citation

Lane NE, Michel B, Bjorkengren A, et al. The risk of osteoarthritis with running and aging: a 5-year longitudinal study. J Rheumatol. 1993 Mar;20:461-8.



To determine the effect of running and aging on the development of osteoarthritis of the knees, hands, and lumbar spine.


5-year cohort study.


A community in California, USA.


Participants lived near Stanford, were 50 to 70 years of age, were high school graduates, and spoke English as their primary language. Runners were members of the 50-Plus Runners Association. Nonrunners were selected from the Lipid Research Clinic study participants. Of 41 runners and 57 nonrunners, 41 pairs were formed, matched for age, sex, years of education, and occupation. 73 participants and 33 matched pairs (30 men) completed the study.

Assessment of risk factors

Participants were interviewed and had physical examinations, including radiographs, at baseline and then filled out yearly questionnaires to collect data on musculoskeletal injuries and medical, dietary, and exercise histories.

Main outcome measures

Radiographs completed in 1984 and 1989 measured osteoarthritis progression for hands, knees, and lumbar spine. Radiographs were read blindly by 3 observers and scored according to the San Francisco Conference on Radiographic Assessment of the Progression of Osteoarthritis. Second and third distal and proximal interphalangeal joints and the first carpal-metacarpal joints were assessed in the hands.

Main results

Mean scores of 3 observers were used. At baseline, runners had a trend toward lower mean weight (68 vs 73 kg) and a lower disability score (0.02 vs 0.04). Runners also spent more time exercising (mean 300 vs 90 min/wk, P < 0.001) and running (mean 236 vs 21 min/wk, P < 0.001). After 5 years both groups had doubled their disability scores (0.04 vs 0.09). Runners, compared with nonrunners, weighed less (68 vs 73 kg, P < 0.05) and spent more time exercising (mean 304 vs 142 min/wk, P < 0.001) and running (mean 185 vs 14 min/wk, P < 0.001). For weight-bearing joints, runners had an increase in their spur score for knee radiographs from 1.3 to 1.5 (15%), whereas nonrunners had an increase from 0.9 to 1.4 (44%). Spur scores for lumbar radiographs increased 1.6 units in each group (P < 0.001 for both). The groups did not differ for development of knee or hand osteoarthritis.


5-year radiographic progression of osteoarthritis occurred in both runners and nonrunners who were over 50 years old. Running was not associated with more rapid progression or increased incidence of osteoarthritis of the knee or the spine.

Source of funding: National Institutes of Health.

For article reprint: Dr. N.E. Lane, Division of Rheumatology, Box 0868, University of California at San Francisco, San Francisco, CA 94143-0868, USA. FAX 415-648-8425.


The study by Lane and colleagues places in perspective the challenges of studying the effect of activity on the natural history of osteoarthritis. The major challenge for studies of the relation between running and osteoarthritis is the choice of control group. Paucity of limiting physical and psychological endurance attributes makes membership in a long-distance runner's association a self-selection process—not so for participants in a Lipid Research Clinic study (weight and, perhaps, disability score variation show this). Because 15% of the community controls were runners, the study was not a true comparison of runners and nonrunners.

Interobserver variation in recognition of joint erosions is difficult to interpret because osseous erosion is not part of osteoarthritis (1). The authors further added sclerosis, narrowing, and spurs to obtain statistical significance. The validity of adding such phenomena, however, does not appear to be documented. Although osteophytes and cysts seem to be the best markers for presence of osteoarthritis, joint space narrowing and subchondral sclerosis have been suggested as the best "markers" for progression of the disease (2). No statistically significant variations in joint space or sclerosis were noted. Disc margin osteophytes represent spondylosis deformans, not osteoarthritis (1). Only zygapophyseal joint spurs might be considered evidence of osteoarthritis, and these were not assessed.

Given the rather minor statistical variation among the groups, loss of 8 pairs (20%) from the study group is of concern, although matched-pair analysis maximizes power. For all sites, the best predictors of progression of osteoarthritis were the participant's weight and osteoarthritis status at onset. Presence of osteoarthritis perhaps predicts further progression with age. The study suggests that runners appear able to maintain good knee mechanics; however, they are a self-selected group, and it may be that persons with good knee mechanics are more likely to be runners.

Bruce Rothschild, MD
Northeastern Ohio Universities College of MedicineRootstown, Ohio, USA


1. Rothschild BM, Martin L. Paleopathology: Disease in the Fossil Record. Boca Raton, Florida: CRC Press; 1992.

2. Croft P, Cooper C, Wickham C, Coggon D. Defining osteoarthritis of the hip for epidemiologic studies. Am J Epidemiol. 1990;132:514-22.