More older people today also have cardiovascular risk factor - high levels of blood pressure, blood sugar and cholesterol - that increase the risk of dementia. But more are taking medications for those conditions, so perhaps better control of those risk factors played a role in the decline.
The findings about obesity were especially puzzling. Compared with people of normal weight, overweight people and obese people had a 30 percent lower risk of dementia, the study found. Under-weight people had a risk 2.5 times as great. Yet the obesity is muddled because other studies have found that obesity in middle age increases dementia risk in old age.
Then there is the education. On average, older Americans in 2012 had one more year of schooling than older Americans in 2000. And years of education were associated with decreased dementia risk in this study, as in many others.
It is still not clear exactly why education would reduce the risk of dementia. There is the cognitive reserve hypothesis：that education changes developing brains in a good way, making them more resistant to dementia, and that people with more education have brains that are better able to compensate for dementia damage.
But education also is linked to more wealth. People with more education often live in environments that differ from those of people who have less schooling, and they tend to have better health over all. They also are less likely to smoke.
As for black Americans, the dementia risk was higher, but some possible reasons - less education, less wealth, more cardiovascular risk factors - did not fully explain the difference. One possibility is that they received a poorer quality of education, so each year of it offered less protection from dementia, said the study's lead author, Dr. Kenneth Langa.
In the end, much of what is happening with dementia rates defies explanation, said Dr. Langa, a professor of medicine at the University of Michigan in Ann Arbor, who also works in the Veterans Affairs health care system there.
Dr. Denis Evans, a professor of medicine at Rush University Medical Center in Chicago, urged caution in accepting the the conclusion that dementia rates were declining and, if they are, in accepting the possible explanations. Although he had nothing but praise for the abilities of the researchers, he noted that such studies were extremely difficult to do. Deciding if a respondent is demented can easily be inadvertently skewed, he said.
“It's very complex,” Dr. Evans said.
But the decline is consistent with what seems to be a long-term trend, despite researchers' failure to find any effective way for individuals to protect themselves from Alzheimer's, the most common form of dementia. Dr. Langa estimates that compared with the rate in the early 1990s, there has been 25 to 30 percent decrease in dementia rates among older Americans.
As for the future, that is hard to predict, Dr. Langa said.
Even with the lower prevalence of dementia, there will be many more older people in the United States over the next few decades, especially people age 85 and older, who are at highest risk. For that reason, the total number of people with dementia should rise, although not as much as had been estimated.
And there are forces acting against the continuation of the decline in dementia rates. More people reaching age 65 and above in the next few decades will have been obese in middle age, possibly increasing their risk of dementia. In addition, the trend toward more years of schooling seems to have levelled off, so the education effect might not be as much of a factor.
But researchers remain optimistic.
The study found associations, Dr. Hodes noted. “Now the real challenge,” he said, “is to see if we can generate evidence of what causes what.”
