Several conditions associated with aging, such as atherosclerotic blockages of the arteries and cataracts can be treated nowadays, but for osteoarthritis, a condition that affects millions in this country alone, little can be done save the administration of anti-inflammatory agents and pain relievers.

Osteoarthritis of the knees can be particularly unfortunate since afflicted individuals typically cut down their physical activity in response, and this can exacerbate or cause other health conditions like diabetes. 

A cure does not appear imminent, but a new technique developed by Lior Shamir and colleagues at the NIH might just be able to predict who will get the condition 20 years prior to symptom onset.

This would be good since people found to be at risk could be encouraged to lose weight, exercise properly, and alter their diet to ward off the onset of the painful condition.

Shamir’s group digitized 200 x-rays that were taken in the 1980s for a project designed to document human aging. All radiographs had been read as normal at the time.

Since the images were obtained, many of these people developed osteoarthritis of the knees.

The researchers sorted the original x-rays into 2 groups: one group included “normal” pictures from the people that eventually developed osteoarthritis while the other had pictures from people that did not.

The researchers then looked for subtle structural alterations of the bone and cartilage in both groups of x-rays.

From the inter-group differences they detected, they built a computer algorithm designed to predict which individuals would eventually develop osteoarthritis and which would not.

Their algorithm proved to be accurate 72% of the time, an astonishing finding considering that the x-rays had originally been read as normal and that they were obtained 20 years before symptom onset.

The write-up is in Osteoarthritis and Cartilage.