This is a very interesting tidbit from LifeHacker about caffeine:
The effectiveness of caffeine varies significantly from person to person, due to genetics and other factors in play. The average half-life of caffeine—that is, how long it takes for half of an ingested dose to wear off—is about five to six hours in a human body. Women taking oral birth control require about twice as long to process caffeine. Women between the ovulation and beginning of menstruation see a similar, if less severe, extended half-life. For regular smokers, caffeine takes half as long to process—which, in some ways, explains why smokers often drink more coffee and feel more agitated and anxious, because they’re unaware of how their bodies work without cigarettes.
Assuming you know some of the variables at play here and track over an extended period (at least 1 month) so that you derive a baseline, you can essentially use algebra to solve for what you don’t know about a person, in particular for a woman based on their observable behavior regarding caffeine consumption and reaction. I recall reading between ovulation and menstruation it’s about 35-40% longer or roughly 9-10 hours. The half-life of a pregnant woman is 9-11 hours [cite].
In practice this would require some dedication as you need to derive a baseline and it would still never be truly accurate, but interesting regardless. I’m sure you can add other statistics and metrics to help improve the accuracy but again it wouldn’t be terribly accurate. Regardless I bet it would surprise some people.
It just shows how easily information is revealed through our mundane activities regardless of how well people conceal it. Psychology, chemistry, medicine, and security all in this one beverage consumption.
I’m sure you can come up with a more male specific scheme as well as many more gender neutral schemes. I just ran across this more female specific example and found it rather interesting.