Updated

A review of the best evidence for interventions to prevent declining brain power finds that only one - mental exercise - consistently makes a difference.

The analysis of clinical trial results for assorted drugs, supplements and activities still can't say, however, whether the brain training programs that do seem to sharpen mental function also improve people's daily lives or lower their risk of developing dementia.

"All we know is you will do better on certain (cognitive) tests. Whether that delays dementia...remains to be seen," said Dr. Raza Naqvi, the study's lead author and a researcher at the University of Toronto.

Mild cognitive impairment may affect as many as a quarter of people over age 70, according to Naqvi and his colleagues. And perhaps 10 percent of those seniors progress to more serious dementia each year, the researchers write in the Canadian Medical Association Journal.

Numerous products and activity programs claim to be able to slow mental decline, but nothing has proven a surefire way to preserve brainpower.

"Most physicians have a feeling that there isn't anything out there that has shown strong evidence, otherwise we would all be out there promoting it ourselves," said Naqvi.

To see whether any approach fits that bill, or at least seems promising, Naqvi and his colleagues gathered results from all the randomized controlled trials - the gold standard for research - they could find comparing the mental functioning of adults given a particular treatment to others who received no intervention.

Participants in all the trials were 65 years or older and had no mental decline at the beginning of the experiment.

Drugs, hormone therapy (in both men and women), vitamins and supplements including gingko and omega 3s mostly showed no benefits. Indeed, most of the trials involving estrogen replacement therapy for women showed greater cognitive decline among women taking the hormones.

One study of the drug donepezil, which goes by the brand name Aricept and is approved for use to slow the progression of dementia, found improvements in users' ability to recall facts, but experiments with other medications showed no benefits.

"At this time, none of the medications or pharmacologic therapies have any evidence to support their use," Naqvi said. "I don't want researchers to give up - I think it's important to critically look at potential therapies - but I don't think the evidence is there to recommend a lot of these."

Studies that involved physical activity yielded mixed results; of the three that Naqvi's group reviewed, one found no benefit to memory, but an improvement in mental processing known as "executive function."

Another study found some memory benefits, but no other cognitive improvements, while a third saw no change in the performance on tests given to participants who went through an exercise program.

Only mental training - for which there were three studies - yielded positive results every time.

One large study that included more than 2,800 people offered one of three mental training programs focused on memory, reasoning or processing speed.

The participants randomly assigned to the memory group, for instance, went through 10 hour-long training sessions that taught methods for remembering written material, such as word lists.

Two years after the training programs, people who participated in a mental exercise performed better on related tasks than others who did not participate.

In other words, the memory group did better on memory tests than people who received no special training, while the reasoning group did better on reasoning tests.

The researchers also tested how well people performed on everyday tasks, such as finding a number in a phone book or preparing a meal, and found signs that people performed better if they had been through a mental training program.

"There was some evidence then that...specifically our speed and reasoning interventions had begun to transfer to everyday function," said Michael Marsiske, an associate professor at the University of Florida who was involved in the mental exercise study.

But he added that although mental exercises help boost performance on cognitive tests, "it's too early yet to say whether they actually prevent dementia or decline."

Marsiske said that the new review might overstate the benefits of cognitive training as a result of Naqvi and his colleagues selecting only studies that used the most rigorous experimental methods to include in their analysis.

Naqvi said it's difficult to compare the mental exercises used in the studies to those that are available to consumers, because each one is designed differently.

"Regardless, I still recommend to my patients to remain mentally and cognitively active as long as possible in whatever way is stimulating to them," Naqvi said.