One Reason I'm So Optimistic About the Ability of Knee Cartilage to Heal ...

...is that (as I've written before), in this medical study, cartilage did appear to heal, and at a surprisingly high rate: 37 percent.

What's more, that figure almost certainly under represents the true rate. I'll get to why in a moment (the why is the reason for this post). The explanation is a bit wonky, but well worth the effort to understand.

First though, that tantalizing 37 percent. What does that mean? That 37 percent of all subjects with bad cartilage were totally healed by the end of the study (two years later)? Nope. Sorry, but nothing so dramatic. Rather, it means that 37 percent of the subjects experienced an improvement in a cartilage defect somewhere in their knee.

Ahah. Now we're getting somewhere. But still -- what's an "improvement in a cartilage defect"? Well, first, areas of knee cartilage were examined by MRI. Then a common scoring system was used to evaluate the tissue, where "0" is normal, "2" is some wearing away with at least 50 percent thickness remaining, and "4" is eroded to the bone (so you have five possible scores: 0, 1, 2, 3, 4).

So for your right knee, for example, you may have a spot that's graded "4" (no cartilage), and at the end of your tibia a place that's a "3," and then a location on the inside of your femur that's a "2." Two years later, the defects were reviewed again. A difference in the score was either an improvement or a worsening, depending on which way the number changed.

Up to this point, you may be thinking you've already read on this blog about the 37-percent-of-defects-healing study, so what's new? Answer: the 37 percent number is almost certainly not accurate. It's too low! "Why" is something I discuss in the book, but have never written about on my blog.

So let's delve into the reason. First, you have to appreciate a subtle truth about measuring rate of change: the amount of change that you find always depends on the accuracy of the measuring instrument.

Confused? Well, in Saving My Knees, I illustrate why this is true with a hypothetical world where you can determine people's heights using only devices that measure in increments of one foot. In this population, everyone has a height of 1 foot, 2 feet, 3 feet, etc. There are no inches.

Suppose you have three teenagers: Mary, Ted, and Fred. Mary (according to our Earth measuring tape) is 5' 6", Ted is 5' 9", and Fred is 5' 11". A study is undertaken to see what percentage of our three subjects grows over the course of 12 months. After a year, Mary is 5' 7", Ted is 5' 9.5", and Fred is 6' 1".

What percent of the population has grown, if we're measuring them with a device that has only feet, and no inches?

Only 33 percent! (Fred was "6 feet" before, because he was between 5 and 6 feet in height, so now that he's between 6 and 7 feet, he's considered 7 feet tall.) What about Mary, who grew an inch? No, she didn't grow at all using our rough measuring stick -- she went from 6 feet to 6 feet. No change. Same for Ted.

Seems crazy, huh? But that's what you get when using an imprecise measuring device.

Now what happens when you measure the difference in cartilage defects using only five grades? You're going to miss change that you would capture were you able to use 10, 20, or 50 grades. Consider: Cartilage at a given location that has worn 95 percent away is rated a "3." Two years later, if that same location is 55 percent worn away, that should be considered an improvement, right? But 55 percent still falls within the classification of a grade "3" defect. So, using this crude measuring system: no change!

If it were possible to make finer measurements with confidence, we'd expect to see a higher rate of change, all else being equal. Take a moment to ponder that. Once you fully grasp the implications, you'll never look at a study that involves rate of change the same again.

What I've explained above, you don't need an MD to understand. This isn't knowledge that's privileged to the cognoscenti of the field, who have examined thousands of knees. This is simple math and logic. But what's the main takeaway?

This: Cartilage is a dynamic tissue, often changing for the worse -- and for the better!

Insert thunderclap here. Because, honestly, too often changes in knee cartilage are seen as a one-way (downhill) process. But that doesn't accord with the reality discovered in this study (and there was another that found similar evidence that defects often improve).

Cartilage is constantly trying to heal: that's what this study suggests to me, when you explore the ramifications. And that should give you encouragement on the long road to recovery.