The finding is based on a study conducted by researchers at the French National Institute for Health and Medical Research, in Montpellier, France, who were led by Karen Ritchie, PhD.
The study involved 7,000 people whose cognitive abilities and caffeine consumption were evaluated over four years.
The researchers found that women age 65 and older who drank more than three cups of coffee (or the equivalent in tea) per day had less decline over time on tests of memory than women who drank one cup or less of coffee or tea per day.
These results remained the same even after the researchers adjusted for other factors that could affect memory abilities, such as age, education, disability, depression, high blood pressure, medications, cardiovascular disease, and other chronic illnesses.
Moreover, the benefits increased with age - coffee drinkers being 30 percent less likely to have memory decline at age 65 and rising to 70 percent less likely over age 80. Caffeine consumers did not seem to have lower rates of dementia.
Ritchie said researchers aren't sure why caffeine didn't show the same result in men. "Women may be more sensitive to the effects of caffeine. Their bodies may react differently to the stimulant, or they may metabolize caffeine," she said.
"Caffeine is a psychostimulant which appears to reduce cognitive decline in women. While we have some ideas as to how this works biologically, we need to have a better understanding of how caffeine affects the brain before we can start promoting caffeine intake as a way to reduce cognitive decline. But the results are interesting - caffeine use is already widespread and it has fewer side effects than other treatments for cognitive decline, and it requires a relatively small amount for a beneficial effect," she added.
Ritchie added that more research is now needed in this field. "We really need a longer study to look at whether caffeine prevents dementia; it might be that caffeine could slow the dementia process rather than preventing it," she said.
The study and its findings are published in the August 7, 2007, issue of Neurology, the medical journal of the American Academy of Neurology