Earlier quoted context omitted.
They measure obesity and find it 0.02 vs 0.01 for vitamin d, which seems a decent enough proxy for lifestyle for this kind of study so your underlying point is valid here. Actually i think the biggest missing control is race. I went in skeptical even though intuitively being deficient in vitamin d makes sense, this isn’t convincing but I agree with with their conclusion that it’s worth a bigger study. If someone read…
Race can be a proxy for the amount of melatonin, which influences vitamin D. It can also be related to many other things. How would you use race to interpret the data?
control for it if you are making claims about vitamin d