Review of Studies About Standing Desks | Hacker News Reader

Review of Studies About Standing Desks

fastcodesign.com

OT, but this image from the article:

http://b.fastcompany.net/multisite_files/fastcompany/imageca...

looks like hell made real on earth. Only thing missing are the demons forcing workers trying to escape back into the pit of anxiety.

I see no mention of the fact that lots of people who swore by treadmill desks also started swearing at them after about 6-9 months.

The problem is that your head is supposed to move when you are walking. When you are working at a computer, your head has a fixed point that you are trying to make stable. This is simply untenable posture and causes all manner of problems.

I wonder if that kind of problem holds for just standing, though.

Probably solved after the 4k VR displays arrive.

You would look rather ridiculous walking on a treadmill with a keyboard, wearing a Rift or something. But I think this could be the optimal setup.

Sadly not - I've got eye problems from staring at a screen 8 hours a day. Now I have to use an app which forces me to look out the window every 10 minutes and focus on something at a different distance.

I've had a treadmill desk for several years now and haven't noticed any issues with this. It's fine to walk when you're just reading or doing light tasks. I do need to sit when doing concentration heavy work though.

"did lose weight, though they experienced no changes in body composition, such as body mass index"

They got shorter from using standing desks? Only way you can lose weight but not change your BMI...

Taken literally, yes, but I'm guessing what it actually means is lost fat. Assuming similar mass was added in muscle, BMI would remain the same. (BMI not being a terribly useful metric...)

Actually, BMI _is_ a terribly useful metric (and it's unfortunate bashing, nowadays, is an echo chamber phenomenon in fitness circles around the web).

The study this is referring to has a small-ish n (36), so, no big deal here - it seems that they took body composition measurements (among many others) without a significant overhead, they even built a lab within the premises of the corporation they were doing the study in. And they got very rich data with it (although the difference in body fat didn't reach significance, others such as waist circumference did).

But imagine that they were doing a different study. Something like a cohort, longitudinal and distributed: studying various corporations which use standing desks during a long, long time (a decade?). This would imply many different people collecting data in different context and different times (and, effectively, bothering even more people). Building labs or ensuring similar equipment across the different sites would be very expensive and laborious - not to mention error-prone! Just imagine the overhead of getting body composition indicators - even something simple as asking people to measure their waist, but during a decade across N sites.

In this kind of large study, a seemingly simplistic but nonetheless standardized (and, in a way, normalized) indicator such as BMI is an extremely helpful tool. Imagine you want to do an analysis not by weight, but by by "obesity" ranges, or check if "obesity" status changed: getting body composition is nearly impossible, but using the BMI pre-established ranges (lean, normal, overweight, obese) will actually be feasible (you just need height and weight!). And yes, although a bodybuilder will be considered overweight or obese by BMI, remember this is a study with a large enough n - edge cases will eventually average each other and BMI will hold true for the sample. The BMI will be a cost effective (and feasible) solution to analyse something more than weight changes.

In that case BMI gets misused a lot. The bashing comes because whilst it may be useful on a large statistically significant population it is nearly useless at an individual level. One can take two people of identical BMI and one can be actually very fat and the other very slim.

Misused, by whom?

I trust my physician to know how to interpret the BMI values. I don't trust my insurance company to do the same, but they're eventually the ones with a large n (so to speak) - ethics aside, if they want a cost-effective way to factor obesity in their premium, BMI might be the way to go. They'll overestimate the premium for the bodybuilder, but they'll get right most of the time.

I'd expect most misuse to be by the media and random people in the Internet. In both cases, however, you should already be taking everything - not only BMI stuff - with a grain of salt.

P.S. Even at an individual level, BMI will hold true for the large majority. Even you're talking in a fitness forum, most people won't have enough muscle mass to affect the index that much.

I know a guy who is a hardcore triathlete. At one point a few years ago he was doing really well, sponsored athlete, over 210 pounds with only 7% body fat. He was at a regular check up and the nurse looked at the BMI chart, then looked at him and plainly said 'Maybe you should work out more'.

Just my anecdata that BMI is misused even in the medical community.

Talking about clueless - had a back issue, physical therapist gave me some good exercises. I asked "In winter I get soft, cycle all summer but with snow I need something else. I hate spin classes. What should I do?"

She gave me a dozen reps with a rope on my ankle, 4 lbs weight.

Now, 20 miles in a short ride is what, about 4000 reps, at quite a bit more than 4lbs. I mentioned this. She said "Lets just start out at 4 lbs and work up slowly".

Sometimes I just wonder.

The fact that it takes into account just height and weight is problematic. No consideration of bone structure or body frame.

At my absolutely lightest, when you could see every rib sticking out, my BMI was still comfortably in the range considered overweight. If I were to live through the Bataan death camps, I might get down to skin and bone enough to be in the normal range.

> The fact that it takes into account just height and weight is problematic. No consideration of bone structure or body frame.

That's a bug, not a feature, if you allow me to grab the expression from software. It's problematic as any other indicator has "problematic" features.

My point being: when you assess any indicator, you can't look only at it's ability to correctly reflect reality (validity). You'll also have to consider other stuff: is it cheap? can I trust people to use it correctly? can it be used consistently? Taking into account only height and weight will diminish validity, but give you positive answers to these three questions. It's a trade-off (and, IMHO, one that's being bashed excessively, considering the empirical evidence saying it's not that "invalid" at all).

Compare it with e.g. skinfold estimation of body composition: skinfold estimation is more valid, yet science shows us that it requires a highly trained specialist for it to be reliable (i.e. to do it consistently; and even with professionals, there's not much consistency). This is also problematic, but - again - it's a bug, not a feature.

> At my absolutely lightest, when you could see every rib sticking out, my BMI was still comfortably in the range considered overweight.

... and that's you being an outlier, or close. :) I'm one as well; also overweight, by BMI standard. But I have to consider that my care with nutrition and fitness should also be close to outlier in nature, regarding population in general.

Knowing this, you and I (and our physicians) will take it into account. And - if the need arises - use another metric. I do both skinfold estimation of body composition and waist circumference measurement but rely mostly on the latter for tracking progress (I'm not a professional and, of course, it's me measuring myself, so I choose the method where I might mess up less).

I'm not so sure that I'm really an outlier, to be honest. I think the data that the BMI calculations are based on is outdated. It's the same numbers you see from New York Life actuarial tables from the 40s and 50s, when big chunks of the population experienced malnutrition.

Quite honestly, I have no idea how someone could be 6'2" and only weigh 148 pounds, which is the bottom end of "normal" BMI, according to the NIH[1].

[1] http://www.nhlbi.nih.gov/health/educational/lose_wt/BMI/bmi_...

Although the formula is (way!) older, the BMI was imported to public health in the 70s and the categories were devised by the NIH in 1985, slightly revised in 1998.

I also feel the example you gave is extremely low, but less so if you consider the subject as a female, don't you think? It has a BMI of 19, which, on men, I feel most physicians would frown upon. Bear in mind national standards; I'm from Portugal, the normal here might be thinner.

As a curiosity, in France and Spain fashion shows can't use models with BMI > 18. Your example minus 8 lb wouldn't even be allowed in fashion shows, where people are expected to be ungodly underweight. As such, I think a physician would also interpret this as dangerously low value. Also, it seems to me that this is a great example of how BMI can be very useful! Waist circumference, percent of body fat or other body composition indicators would actually be worse to indicate a underweight person in this case..

Yes, that's a good point. My statement was a bit tongue in cheek. BMI can be useful if used appropriately, with its shortcomings in mind. It isn't nearly as useful as other metrics for the sort of individual comparison mentioned by the grandparent comment though.

I'm not an MD or dietician, but I'm kind of surprised that BMI hasn't been replaced by something as simple as height-to-waist ratio, which seems to me would be a potentially informative measure with higher correlation to body fat percentage than BMI.

Difference scientific fields will value different things. A computer scientist might prefer a fundamentally more "valid" metric, but someone who studies something more organic would rather have a less valid but more empirically grounded one.

You can't forget that there's really a ton of reported data behind BMI. So much that e.g. in children we know, per age, the percentiles which should be used as cut offs for considering obesity and triggering further analysis/intervention. But this is the individual level! As you certainly know, obesity is largely a public health issue, which makes BMI - which has lots of epidemiological data behind it - even more appropriate. There's really no better way, IMHO (and not only [1]), to track obesity as a public health problem.

[1] Hall, D. M. B., & Cole, T. J. (2006). What use is the BMI? Archives of Disease in Childhood, 91(4), 283–286. http://doi.org/10.1136/adc.2005.077339

Yeah, that was an over-simplification by the fasctodesign author(s). "No changes" doesn't mean quite the same as "no significant changes".

"The 18 participants in the intervention group showed a reduction in weight over a 3 month period but did not show any significant changes in body composition" [1]. It's a study with a small n, it's normal to fail to reach significance in some variables.

[1] Brittany T. MacEwen, Dany J. MacDonald, Jamie F. Burr, A systematic review of standing and treadmill desks in the workplace, Preventive Medicine, Volume 70, January 2015, Pages 50-58, ISSN 0091-7435, http://dx.doi.org/10.1016/j.ypmed.2014.11.011. (http://www.sciencedirect.com/science/article/pii/S0091743514...)

I have been doing this for about a week. I don't actually have a standing desk though. I just put the 24 inch display and my small mechanical keyboard on top of my tower cabinet (Which is already on a table). This set up places the display at my eye level and the keyboard at just below chest level. So I don't actually have a place support my hands, and I just go 'hands down' after typing something. And typing on the keyboard at a little below chest level feels really great. Because you don't have to bend your wrists up, while typing, as you would need to, if the keyboard was a bit lower (as shown in the pictures in this post).

It actually feels great. (I have a chair on stand by though). The biggest advantage I can see is that, it is easy to take a break and move around, when you are already standing.

I wonder how one compensates for the Hawthorne effect? https://en.wikipedia.org/wiki/Hawthorne_effect

I'd be curious about studies on something like programming. Many of these talk about cognitive function and typing ability, but they don't mention the actual tasks being done (that I can find). With all due respect to everyone else, programming is a different beast altogether. I worked in finance at a computer all day, and my requisite attention level for that job was significantly different than when I switched to programming full-time.

I worked on a personal project for about a week or so on a makeshift treadmill desk. I was somewhat more productive, not a magic bullet or anything, but it's definitely something you could get used to and derive advantages from.

I would bet that the studies focused on cognitive function and typing ability because they're objectively measurable, and much more repeatable than, say, measuring the significance or difficulty of the tasks.

look at the pictures of people using laptops with standup desks. Their head is pointed down, that is terrible for ergonomics.

Proper standup desks should have a different height for the keyboard and the monitor.

Hanging 3 kg weight with neck muscles all day kills the neck.

I made interesting observation in the work few months ago.

All tall people (> 1.8 m) tend to use monitors and keyboards every time they have opportunity. Smaller people are more comfortable using laptops in the office. Tall people have to bend their head downwards more and they have to give up terrible ergonomic position first.

Yes, I'm tall and I can confirm this.

I try to get the center of my displays at the height of my eyes.

But I have a convertible notebook, so I will probably buy a keyboard and use it in "tent-mode" in the future.

> In one year-long intervention that included lean, overweight, and obese workers, the 36 total participants lost an average of 3 to 7 pounds.

Now, that's a useless statement if I've ever seen one. The arithmetic mean is pretty much meaningless if the group varies that much.

Cool, nice to see. I've been wanting to convert my current sit/stand desk [1] to a treadmill setup for a while, but I'd like to work out a way to maintain the ability to relatively easily switch back to seated mode. Still need to think on that a bit.

[1] http://www.tempestblog.com/wp-content/uploads/2015/09/201508...

Reply on news.ycombinator.com