Sign up to receive special offers and news about Neal Stephenson.

Privacy Policy

Notes on 416 Days of Treadmill Desk Usage

Neal Stephenson - March 2015

Since I’ve advocated for the use of treadmill desks in Some Remarks and evangelized it to people I know, I thought it behooved me to publish some data and some reflections on it, born of experience. I have been using a treadmill desk for a few years, but in mid-January of 2014 I began keeping track of my daily mileage on a spreadsheet. As of early March 2015 I have enough data (416 days’ worth) to amass some simple statistics.

While its beneficial effects certainly outweigh its downside, it would be less than honest to claim that use of a treadmill while working is completely benign. During the first half of 2014 I began to experience discomfort in my left leg, typically in the buttock and thigh but sometimes extending down to the knee and ankle. This was clearly associated with walking on the treadmill. Interestingly, however, it was completely absent when walking out of doors. I could walk in a normal stride outdoors for many miles without having any trace of this problem, but even a short stint on the treadmill brought it back. I began to curtail my treadmill mileage by reducing the speed, typically to the minimum value of 0.5 miles per hour.

In mid-June of 2014 I consulted a physical therapist who observed me walking on a treadmill in her office and pointed out that I was tottering from side to side in a Frankenstein-like gait. The obvious remedy was to start walking normally, planting one foot ahead of the other. In the meantime she prescribed some stretching exercises.

A bit of experimentation showed that increasing the treadmill’s speed produced a longer, more normal stride. While the leg pain didn’t go away entirely (and is still with me to some degree) I was able to log more miles while experiencing a significant reduction in discomfort. Typically I now run the treadmill at 1.8 mph, almost quadruple the speed to which I had reduced it in June 2014.

Here’s a plot of all the raw numbers, showing miles per day, from mid-January 2014 through the first week of March 2015. The vertical red line in June is the day I went to the physical therapist:

There are outliers to both ends of the scale, evident on a histogram


The zeroes are simply days when I didn’t use the treadmill, typically because of travel--a two-week vacation in August stands out clearly on the first graph. The big numbers--basically anything more than 5 miles a day--reflect days when I went for a long walk outdoors and decided to log the miles into the spreadsheet. Those aren’t actually treadmill numbers.

Since I’m interested here in treadmill usage, I produced a smaller dataset that only reflects days on which I walked more than 0.75 miles but less than 4.5 miles--a somewhat arbitrary range chosen to grab the part of the data that looks vaguely like a normal statistical distribution:


The total number of such days is 256, accounting for about 62% of the days in the whole span.

The scatterplot for those days is still too noisy to make out any clear patterns:


As an experiment I applied a 10-day moving average to the data series. Here’s what the result looks like:


Again, the red line shows the date of the visit to the physical therapist. Up to that point my mileage shows a gradual, irregular decline as I dial back the speed. Immediately afterward it shoots up to about double the previous low as I crank up the speed. Then it drops again, probably because of various complications in late summer that kept me away from my desk, and then it begins a long irregular climb, with numbers in the range of 2.5 to 3 miles per day being typical recently.

The total number of miles walked is 735. According to a thing I found on the Internet, that represents 57000 calories burned for someone of my weight walking at my typical pace. Averaged over 416 days that is a somewhat underwhelming 137 calories per day, though more recently as mileage has rebounded it has looked more like 230 calories per day. According to people who research this kind of thing, however, the main health benefit of walking isn’t so much burning calories as it is changing other metabolic factors with long-term health implications.


2.5 to 3 miles a day is easily sustainable for me, a 55 year old man in generally good health.


A faster pace (1.8 mph) turns out to create less wear and tear on the body, because it induces a more natural gait.


People who do this a lot need to pay attention to gait, posture, shoes, and other factors that have a bearing on joint and muscle health. This seems like common sense, but anecdotally I’ve heard from a number of people who overlooked it.


The number of calories burned isn’t all that large.