This is also a myth based on a poorly designed study, and it is a common for someone to skim the study and come to a false conclusion.
There WAS an increase in the MTCI (malar temperature circulation index) of the cold-caffeine/caffeine pill administered group(s). That's a number indicating the difference in baseline temperature of your skin vs. the temperature after the studied element was introduced. For reference, redness was seen at MTCI of 1.4 or higher. MTCI ranged from 0.8 bare minimum to 2.2 for the average high values. Since 0.8 was the bare minimum seen, all agents had an effect, even the caffeine. These are low doses of caffeine -- a single cup of coffee is 90-120 mg, and the researchers stopped at 200 mg when administering caffeine pills. The caffeine did have an effect on increased blood flow to the cheeks at this dose. In fact, the MCTI was 0.8 to 1.1 in the caffeine treated participants, where 1.4 is the flushing threshold (per the researchers.) Naturally, if 200 mg brought the skin that close to flushing they ought to have tried 300 to 400 mg, which would have likely brought the MCTI to 1.6 to 2.2 -- very visible flushing. However, they did not try any higher doses, so we can not draw a conclusion.
What the study actually found is that in a tiny sample of 24 people, a 200 mg caffeine pill did increase malar skin temperature towards the flush point, but didn't cause visible flushing at a 200 mg dose. It also found that in some of the cohort, hot water alone caused a reaction just as bad as caffeine alone!
More specifically, the 200 mg caffeine pill group saw MTCI of 0.8 to 1.1. Cold coffee was nearly identical. The hot water people saw 1.1 to 3.6 MTCI. So, in some of the participants, the caffeine pill WAS exactly as bad as hot water. The worst reaction was, of course, the group administered caffeine + hot water.
The dose of caffeine used is very important, because the average daily caffeine intake among caffeine drinkers is much higher. Per Villanova University, "More than half of all American adults consume more 300 milligrams (mg) of caffeine every day, making it by far America's most popular drug."
Let's look at the actual numbers from the study:
(agent: MTCI)
Hot coffee: 1.4 - 2.2 (a few were higher)
Hot water: 1.1 - 2.2 (a few were higher)
Cold coffee: 1.1- 2.2
200 mg caffeine.: 0.8-1.1
So, essentially, we know caffeine has an effect, and at 200 mg, it's slightly lower than what would cause visible redness. It stands to reason that doubling the caffeine would cause a more pronounced effect, but the authors did not bother to try this obvious test of their hypothesis.
24 people is a tiny sample. Caffeine allergy is uncommon, and a 24 person cohort would easily miss these people. For example, my mother, from a single cup of black coffee, becomes dizzy and violently ill. That'd be a classic example of caffeine intolerance. No such people were part of the study.
If you had a 24 person study, you'd miss all sorts of genetic anomalies. You wouldn't even find a person with common peanut allergy in a group that small due to the indicience of same, so this study is only worth the fact that heat is a contributor to flushing. It does not tell us that caffeine is not.
Anecdotally, my interest in this crappy study from 1981 is because my sole rosacea trigger is caffeine. I'm not a coffee drinker. I developed rosacea from a few years of strong green tea (10 minute brews, multiple tablespoons of raw leaf per day.)
Hot drinks, and hot herbal teas without caffeine have absolutely no effect on my flushing.
However, even a half cup of COLD white tea, or a single sip of yerba mate, or COLD green tea sends me into an immediate reaction within 30 minutes, followed by a 12 hour period of facial itching and very hot cheeks. Every single time.
Caffeine reaction is very real. I know of two popular rosacea bloggers who accidentally removed caffeine from their diet (say, because of stomach upset) and had all flushing symptoms resolve after years of playing around with other factors and medications that had no effect.
