For years, I’ve been convinced that exercise is the key to good health and longevity. It keeps all our systems up to scratch and, as a useful side effect, prevents things going wrong. For instance, if you ­exercise every day, you almost ­certainly won’t get type 2 diabetes.

I’ve often wondered why exercise is so good for us. What does it do to our bodies that’s so beneficial? And should we be “dosing” ourselves with it as you would a treatment?

This question has also vexed ­Jonathan Long, assistant professor of pathology at Stanford University, US.

He describes special exercise factors, or “exerkines”, produced by cells in our bodies when we exercise.

These, he suggests, could be harnessed as drugs and might lessen certain health problems such as obesity, heart disease and diabetes, as well as improve athletic performance.

Long and his team developed a novel technique to identify exerkines.

It also reveals which particular cell types produce which exerkines – key information for better understanding the role exercise plays in health.

Long’s method has revealed the myriad ways exercise changes the production of exerkines by 21 different cell types in mice. The team found nearly 200 different exerkines ­regulated by the 21 cell types.

“This means the effects of ­physical activity are very widespread across many tissues and organ systems,” Long says. “And we’re only just starting to understand that complexity.” Surprisingly, muscle, bone, and liver cells had only moderate responses to exercise when compared to another type of cell that’s found in many different tissues and organs.

“If we really want to understand the exercise response, we can’t just focus on muscle and bone and the other tissues we associate with exercise,” Long says. “We have to look much more widely.”

The team was also surprised to find liver cells produce several members of a family of proteins called carboxylesterases following exercise.

To understand their role better, Long’s team engineered mice that made heightened carboxylesterases levels in the liver without exercise.

These mice, it turned out, resisted weight gain on a high-fat diet and had improved endurance on a treadmill.

“These carboxylesterases are ­sufficient to confer some metabolic benefits of exercise without the animals themselves exercising,” Long says.

Just as the carboxylesterases show a metabolic benefit, might other exerkines have important anti-inflammatory effects or other beneficial effects on bones, the heart, immune system and brain? And are our levels of carboxylesterases bumped up by exercise? Could exercise in itself be medicine? Long thinks so.