The idea that coffee reduces the risk of dementia is kind of silly, if you think about it.
First of all, coffee has many effects. Too many to narrow down. For example, coffee could reduce the hours spent sleeping, or might increase exercise. We also know that it is a diuretic, and that it makes the heart beat faster and increases blood pressure. And it's a stimulant.
All of that aside, consider the history of coffee consumption and the incidence of Alzheimer's disease and other forms of dementia. Per capita coffee consumption was about 20 gallons in 1910. It was also about 20 gallons in 2005. Coffee drinking peaked just after World War II, to just under 50 gallons per person per year. Then soft drink consumption rises, and coffee declines back to early 20th century levels.
So, since most dementia has a decades-long latency period, many of the people who are slipping away now were born at a time when coffee drinking was highest. Consequently, their parents drank a lot of coffee. They were still drinking more coffee growing up than people drink today.
So, one might ask, why is dementia and Alzheimer's increasing as the rate of coffee consumption drops, if coffee protects against dementia?
One obvious factor is that people are living longer now than ever before. Many people probably used to die from other causes before they had a chance to develop dementia.
In other words, the coffee hypothesis makes no epidemiological sense. The crude data seems to suggest that coffee drinking is linked to higher rates of dementia. But, as we all should know, linkage doesn't necessarily reveal any causal connection.
The only positive side to this research is that it encourages behavior that's already taking place. People do not have to give up coffee, the same way they're supposed to quit smoking and cut back on fatty foods.
Keep the java coming.