As someone who has occasional wrist pain and is fluent in both QWERTY and Dvorak, the latter is significantly less painful to type with when my pain is present. I believe that that indicates that it is less strenuous.
Could it be that you get less wrist pain because Dvorak switches things up rather than it being better overall? With mouse use I find that a different grip can alleviate pain not because it's easier on the wrist, but because it's just different from the way I normally use it.
It's possible, but given my experience seems unlikely - I spend the majority of my time using Dvorak, and very little time using QWERTY, so there's not much switching going on. Also, most typing effort models (e.g. carpalx[1]) put Dvorak as significantly lower effort than QWERTY.
I think it'd be more likely to be the switching effect if you were going between two similar-effort layouts (e.g. Dvorak and Colemak).
>Could it be that you get less wrist pain because Dvorak switches things up rather than it being better overall?
No. It's because you don't have to do hand contortions with Dvorak the way you do with qwerty. If you want a demonstration, try typing "minimal" on both, and watch your hands carefully. On qwerty, it's almost all on one hand, and alternating between top and bottom rows. On Dvorak, it's alternating between hands (which is always better). You can see similar things with any English text of decent length: Dvorak makes you alternate between hands far better, and far more of your typing is on the home row (esp. since the vowels are all on the home row).
But is that really good? I can increase my sensivity very high on my mouse, where I barely have to move the mouse. However, this does cause me more pain than low sensitivity instead. High sensitivity makes me move the mouse almost entirely with the wrist, where low sensitivity uses the wrist and forearm.
I don't know if the same applies to typing, but I can see a mechanism where it is.