I drink coffee/espressos each day, probably too much. Just recently after not having my usual Americanos I had a massive headache before bed. Advil did nothing so I figured it must be a caffeine withdrawal headache. They're hell and there's no way I can sleep with one so I made a shot of espresso and downed it quickly. Within 30 mins my headache went from a jackhammer pounding my skull to a distant woodpecker, and after another 30 mins I was asleep and slept fine. The big confounding factor with caffeine studies might be tolerance.
I had a similar experience on a long flight. I figured I'd avoid caffeine so I could sleep better, but I just woke up a few hours into the flight with a bad headache. Headed back to the galley to get some coffee, then slept halfway well.
Excedrin or the generic OTC "migraine" medications are good for this: a bit of caffeine, aspirin, and acetaminophen. Easier to carry around than having to scout out coffee and a bit more general purpose than caffeine pills alone.
The big confounding factors with caffeine is that nutrition based studies are extremely hard to run to the point where they are almost as bad as sociology studies.
Controlling for all the factors is nearly impossible.
Age, life style, genetics and everything else in between introduces so many variables that you simply can’t estimate or predict their impact on the study.
Even in your specific example it might be caffeine dependency or it might be something else completely like giving up 3-4 americanos a day and not making up for the liquid intake loss which left you more dehydrated in the evening.
