Cyclists: adjusting saddle height and maintaining bone density


As of September 2017, new Sweat Science columns are being published at Check out my bestselling new book on the science of endurance, ENDURE: Mind, Body, and the Curiously Elastic Limits of Human Performance, published in February 2018 with a foreword by Malcolm Gladwell.

- Alex Hutchinson (@sweatscience)


Two studies in the March issue of  the Journal of Strength and Conditioning Research that may be of interest to cyclists:

First (and simplest), another study about cyclists and bone density. Lots of previous cross-sectional studies have found that cyclists seem to have lower bone density than non-athletes and people who do impact activities like running. This one, from UC San Diego, actually followed 19 male masters cyclists and 18 matched non-athletes for seven years. Sure enough, the cyclists started with lower bone density, and declined faster during the study. By the end of the study (when both groups had an average age of 57), 17 of the cyclists had osteopenia and six had more serious osteoporosis; in the control group, 11 had osteopenia and one had osteoporosis. The message: get some impact activity, do some strength training, and get your bone density checked periodically.

Second is a study about optimizing saddle height (using “saddle” rather than “seat” makes me snicker, but that’s what they use in the paper, so I’ll stick with it!) for both performance and injury prevention. Apparently there are two standard, well-studied approaches to setting saddle height. The Hamley method, based on research in the 1960s, recommends setting the distance between pedal and saddle as 109% of inseam, measured from ischium (hip bone) to floor. That’s based on optimizing performance. The Holmes method, on the other hand, suggests setting knee angle at the bottom of the stroke to between 25 and 35 degrees to avoid injury.

A useful aside in the paper, derived from the Holmes research: pain in the front of the knee usually means the saddle is too high low, pain in the back of the knee means it’s too low high. [Update March 6: Thanks to commenter Phil for catching the fact that the journal article had it backwards!]

Anyway, the problem is that these two methods don’t always coincide, mainly because people have very different ratios of upper leg to lower leg length. In this study, setting the seat at 109% of inseam led to knee angles ranging from 19 to 44 degrees, and only three of the 11 subjects (who were well-trained cyclists) fell into the 25-35 degree injury reduction zone. So which is best?

The researchers looked at anaerobic power (30 second sprints) and economy (15-minutes at 70% VO2max) for three different settings: 109% inseam, 25 degree knee angle, and 35 degree knee angle. Surprisingly, the 25 degree knee outperformed 35 degrees AND 109%, particularly for economy. These results in well-trained cyclists echoed earlier studies by the same group in casual cyclists. So they conclude that 25 degree knee angle is the best way to set saddle height, since it’s within the “minimize injury” range and also appears to optimize performance. Obviously if you’re a Tour de France racer, you’re going to optimize bike position in a much more sophisticated way, but this seems like a useful rule of thumb for the rest of us.

12 Replies to “Cyclists: adjusting saddle height and maintaining bone density”

  1. Saddle is the correct term.

    Thanks for the article! Now if I can only figure out how to measure a 25° leg angle…

  2. Nice to know that you don’t have to sacrifice efficiency for comfort, but as a practical matter I have always found it necessary to adjust the saddle to the most comfortable position. What’s the point of being efficient if you can’t ride because you’re injured?

    Even if people did have identical ratios of upper to lower leg length, an inseam ratio wouldn’t work because effective leg length depends on cleat position, which varies by rider preference. The front-of-knee/back-of-knee observation is generally right but in extreme cases you can be fooled by fore-and-aft position. For the record, here is how to fit a bike for general road use:

    1. Move the saddle back until you can hold your position at least momentarily when you remove your hands from the bars. This should be in your lowest normal riding position – in the hooks or on the drops if you are going to ride any distance.

    2. Position the saddle height until it is comfortable with the actual shoes, cleats, and pedals you are going to wear. The 25 degree rule quoted is a good guideline.

    3. Because the seat tube is sloping, if you change the saddle height you will shift the fore-and-aft position, so you will have to iterate steps 1 and 2 until you converge to an acceptable tolerance; this is pretty tight, about 1 or 2mm.

    4. Now select a stem length such that you are comfortable riding on the tops in your selected saddle position; that usually means your back makes about a 45 degree angle. If the necessary stem is unavailable because it would be too long or short, or if your weight is too far forward or back on the bike, this frame is no good for you; you need a different size or geometry.

    5. Because stem lengths come in increments of 1cm, you will be lucky if the stem you selected in step 4 is exactly right; probably you had to choose the closest fit. In this case, go back to steps 1 & 2 and fine-tune your saddle position.

    You may notice that I haven’t said anything about stand-over height. Although it is the first (and sometimes only) thing bike shops check, it is by far the least important aspect of bike fit. Yes, you need to be able to get on and off the bike safely, and to stop and put a foot on the ground at red lights, but if your bike fits in other respects it is very unlikely to cause a problem in this way. In particular, there is no reason whatsoever to think you need to be able to stand over the top tube with both feet on the ground; in years of daily commuting and riding hundreds of kilometers recreationally each week, I have never wanted to do this.

  3. Hmm. I also didn’t mention bar height in my previous comment. Most riders have a preferred bar height at a given point in time, so you can treat this as fixed when you fit your bike. However, this preference changes over time for various reasons. If you lower your bars, you move your weight forward in a given position, so all things being equal you will have to move your saddle back a bit; you may even need a longer stem in some cases. The opposite would be true, of course, if you moved your bars up. To some extent, your ability to unload your hands is a function of core strength, so that is another variable.

    And of course, this fitting advice is no good for a time-trialist; they need to have a much more forward position on the bike.

  4. The stuff about bone density makes intuitive sense. Cycling is so catabolic – it’s just impossible to keep weight on when you pile up the distance, and that lost weight has to come from somewhere. Why wouldn’t your body cannibalize the tissues least used?

    On the other hand, I find it difficult to lift weights during the summer cycling season. To work around this, I try to have a distinct anabolic phase in winter when I lift a lot. I am hoping that as long as my “seasonally-adjusted” weight is fixed I am OK. It would be interesting to know whether any studies support or puncture this hope …

  5. Thanks for the detailed instructions, Phil! I knew that proper fitting procedures were above my pay grade, so I’m glad to defer to an expert. 🙂 And I certainly agree that comfort overrides other considerations — though it’s good to have rules of thumb (e.g. 25 degrees) as a starting point for people who have so little experience on a road bike that they don’t really know what “comfortable” is supposed to feel like.

    Sarah, the study did indeed use a goniometer. That being said, the abstract you linked to doesn’t make goniometers sounds that bad. They have a systematic “error” of 13.9 degrees compared to the computer system, but that’s irrelevant as long as you compare apples to apples (i.e. the rule of thumb was devised using goniometers, so you SHOULD use a goniometer to implement the rule of thumb).

    What matters is the non-systematic error — i.e. how repeatable the measurements are. In the abstract you linked to, the mean error was a much more reasonable 5.07 degrees between any two repeated measurements, compared to 1.15 degrees with the computer system. So if you take multiple measurements and average them, you’re going to be within much less than 5 degrees, which seems about as accurate as one can hope for. And that is, in fact, what they did in the new study: “All measurements were taken by the primary investigator and repeated multiple times to ensure accuracy.”

  6. > A useful aside in the paper, derived from the Holmes research: pain in the front of the knee usually means the saddle is too high, pain in the back of the knee means it’s too low.

    Could this be a typo? My understanding is that anterior discomfort is usually the result of the saddle being too low or too far forward, e.g. patellar tendonitis caused by increased tension at TDC, while posterior discomfort is usually the result of over extension near BTC.

  7. @Phil
    Hi Phil: This is one of those questions where I don’t have any actual knowledge of my own to fall back on, so all I can do is paste from the paper:

    “The Holmes method is recommended for injury prevention (10). A saddle that is set too high can lead to anterior knee pain and a saddle height that is set too low could lead to posterior knee pain (10,13). To avoid either extreme, Holmes et al. (10) recommended using a saddle resulting in a knee angle between 25 and 35 degrees, with the pedal located at the bottom of the pedal stroke.”

    So I can at least say that I didn’t make a typo! 🙂 Of course, it’s possible that the paper got it wrong — or that it’s simply more complex than any simple rule of thumb can capture. Anyone else have votes either way?

  8. @alex
    I’ll take “the paper got it wrong” for $400 Alex. 🙂

    I don’t have access to the the source you cited but in an earlier paper, Comparing methods for setting saddle height in trained cyclists (Journal of Exercise Physiology, V.8, No. 1, 2005, p. 52), Peveler states: Patellar tendonitis is a common overuse injury in cyclists and is usually associated with a low saddle height. A saddle that is too low can cause over-compression of the knee, resulting in anterior knee pain. Due to this, it has been recommended that cyclists with patellar tendonitis move their saddle height closer to a knee angle of 25° at the bottom of the pedal stroke (5,6,10,11). Biceps tendonitis is another common over-use injury in cyclists and can be caused by a saddle height that is too high. If the saddle is too high it can cause posterior knee pain by over-extension in the dead spot at the bottom of the stroke, which puts a heavy strain on the biceps femoris.

  9. @Phil
    Phil, you’re absolutely right. I checked yet another paper by Peveler, this one from 2008, to get the deciding vote. He has almost exactly the same introductory paragraphs as in the new paper, except that he swapped “anterior” and “posterior”:

    “A saddle height that is set too low can result in anterior knee pain due to increased compression in the knee joint through the top of the pedal stroke and during the follow through to the bottom. A saddle height that is set too high can lead to posterior knee pain due to overextension of the knee at the bottom of the stroke.”

    My guess is that he was trying to do some minor rewriting so that the new paper didn’t look like a complete cut-and-paste job, and switched the order of “too high” and “too low” but forgot to switch the order of “posterior” and “anterior.” Or something like that.

    Anyway, the point is: thanks very much for catching that! I’ve updated the main text.

  10. In the studies on cyclists and bone density, are they looking specifically at road cycling? I’m curious about mountain biking, which feels to me a much higher impact activity.

  11. @Heather
    Interesting point, Heather. I believe most of these studies looked at road cyclists. I just did a quick search and saw this study (, which found that mountain biking does indeed provide an “osteogenic stimulus” that road biking doesn’t. And having recently gone for the my first “real” mountain bike ride (on a bike with no shocks), I have no trouble believing that!

Comments are closed.