RT Journal
A1 Murphy WA
T1 THe relation of running to bone and joint disease
JF JAMA
JO JAMA
YR 1986
FD August 8
VO 256
IS 6
SP 715
OP 715
DO 10.1001/jama.1986.03380060041015
UL http://dx.doi.org/10.1001/jama.1986.03380060041015
AB To the Editor.— 
    The recent investigations of a possible linkage between running and osteoarthritis of the knee reported by Lane et al1 and Panush et al2 are interesting and important. The studies are similar in that each compares avid runners with controls (mainly nonrunners) and arrives at the general conclusion that running is not associated with osteoarthritis of the knee.Both groups of authors consider the possibility that a selection bias might be present but either rationalize or dismiss it. Lane et al identify three possible self-selection biases: (1) resistance to osteoarthritis in runners, (2) osteoarthritis preventing running in nonrunners, and (3) early cessation of running in previous runners due to pain. They indicate that potential bias should be minimized because comparison of a group that runs a lot against a control group should exaggerate differences between the groups. Is it possible that they underestimated the magnitude and