In January the pilgrimage to the gym begins, with most visitors hoping to drop a few pounds or gain some muscle. They will be happy to know that regular physical exercise can also improve brain health and memory, a “brain boost” that many studies over several decades have confirmed. While the specific mechanisms of how that improvement happens have been slow to come into focus, they are now becoming somewhat clearer.

Neurotrophic factors—chemicals that promote the survival and birth of neurons—seem to play some role. Researchers have known for some time that exercise produces these chemicals in rodents, and that more exercise leads to a larger hippocampus—a brain region important for memory. A team led by psychologist Kirk Erickson at the University of Pittsburgh tried to find out whether the same held true for humans, and in 2009 were able to prove that a higher level of cardiovascular fitness in older adults was associated with a larger hippocampus. Erickson’s 2011 study further showed that exercise caused an increase in the size of the hippocampus.

“We were able to demonstrate that an exercise regimen could modify the hippocampus in older adults, a group that typically shows deterioration and atrophy of this part of the brain,” Erickson says. Those results were later replicated in children, adolescents and midlife adults.

Erickson’s more recent work looks at whether the hippocampus volume gained through exercise leads to improved memory, and what kinds and intensities of exercise are most beneficial for the brain. His lab has also shown that under the influence of exercise, other regions of the brain grew as well, including the prefrontal cortex, which plays an important role in a group of processing and decision-making skills called executive function. Tests also suggest that executive function may get a boost from exercise.

The role of exercise in strengthening bones might also offer brain benefits. Regular weight-bearing exercise such as walking, running and weight-lifting can help counteract the bone weakening that comes with aging, and strong bones release a hormone called osteocalcin. In 2013, a team led by Gerard Karsenty, the Paul A. Marks Professor and chair of the department of genetics and development at Columbia University Medical Center, showed that mice lacking osteocalcin have major cognitive defects.

Karsenty’s unexpected finding sparked the curiosity of neuroscientist Eric Kandel of Columbia University, who won the Nobel Prize in Physiology or Medicine in 2000 for discovering the biological basis of memory in the sea slug Aplysia californica. Kandel set out to determine whether osteocalcin could also reverse memory loss in older mice.

This became part of a long-standing effort to investigate how age-related memory loss, which affects part of the hippocampus known as the dentate gyrus, develops. To that end, Kandel and his team gathered a sample of human brains and measured the activity of genes in that region. One gene, RbAp48, particularly piqued their interest. It is active only in the dentate gyrus, its activity decreases throughout a person’s life, and, most tellingly, it is part of a molecular switch that converts short-term memories to long-term ones.

Young mice normally have an abundance of RbAp48 protein in their brains, and older mice have about half as much. But when the activity of RbAp48 was inhibited in younger mice, Kandel’s team found that they behaved like older mice in laboratory learning tasks. Meanwhile, when an older mouse was given an injection of RbAp48 in the dentate gyrus, its age-related memory loss was reversed.

But the gene didn’t act alone. Kandel showed—through a series of tests in mice—that it was almost certainly osteocalcin, which is secreted by cells in the bone, that regulated the gene’s activity. Mice that were deprived of either RbAp48 or osteocalcin exhibited poorer recall, and healthy levels of osteocalcin correlated to better memory in both young and old animals.

On the other hand, in a paper recently published in the Journal of Experimental Medicine, Karsenty’s team was able to show that osteocalcin is needed for the beneficial effect on cognition that occurs when older mice are given plasma from younger mice. This is explained by another finding from the Karsenty lab: circulating levels of osteocalcin are much higher in younger mice than in older mice. In this study they also identified the receptor that osteocalcin uses for its functions in the brain—a finding that was confirmed by Kandel’s team.

Bone mass and osteocalcin both decrease throughout life, but exercise that increases bone mass can increase osteocalcin release. This, in turn, can help maintain cognitive health, Kandel concludes.

The effect of exercise in improving brain health may involve thousands of molecular cascades, says Erickson, and scientists will continue to look at each of these mechanisms, both alone and in combination with others. “The sum of all of these things may be what is really driving the effect,” he says.