Repeated physical activity builds endurance by triggering widespread adaptations throughout the human body, but a new study reveals that the brain actually directs these physiological changes. Researchers found that activating a specific cluster of brain cells immediately after a workout is required for mice to increase their physical stamina over time. This discovery flips the traditional view of exercise physiology by showing that structural changes in the brain drive muscle adaptations, rather than just reflecting them. The findings were published in the journal Neuron.
Physiology research on exercise adaptations often focuses on the periphery of the body. When an animal or a human engages in a workout routine, skeletal muscles, the cardiovascular system, and immune function all undergo structural and chemical changes. For years, researchers viewed the brain’s structural changes during exercise as a reflection of these peripheral adaptations. Under this model, muscle tissues release chemical signals into the bloodstream, and the rest of the body responds to that muscular stress.
However, the central nervous system maintains an expansive network for sensing and regulating the body’s daily energy demands. The ventromedial hypothalamus is a region deep in the brain that manages energy expenditure, feeding behavior, and blood sugar levels. Within this neural region sits a specific population of neurons that express a protein known as steroidogenic factor-1, or SF1. Because these cells regulate metabolism, researchers suspected they might play an active role in how the body handles the physical stress of a demanding workout.
University of Pennsylvania researcher Morgan Kindel, Providence College researcher Ryan J. Post, and a team of colleagues investigated whether these SF1 neurons respond to physical exertion. The scientists hypothesized that the brain actively coordinates the metabolic remodeling that leads to increased athletic stamina. They designed a series of small studies using mice to test how manipulating this tiny cluster of brain cells would affect overall physical endurance. The resulting data established a feed-forward loop between physical movement and brain plasticity.
The researchers first exposed a small group of mice to a single session of treadmill running and analyzed their brain tissue. They stained the tissue to measure the expression of a gene called Bdnf, which becomes more active when neurons fire. The analysis showed an increase in both the number of SF1 neurons expressing this gene and the total amount of the gene’s transcripts per cell. This initial test indicated that the exercise session successfully stimulated the neural population.
To see what happens when this neural activity is blocked, the team injected a specialized genetic tool into the brains of a small group of mice. This tool produced a tetanus toxin that prevented the SF1 neurons from releasing chemical signals to other cells. When placed on a treadmill stress test that gradually increased in speed, the mice with silenced SF1 neurons reached exhaustion faster than unmodified control mice. The researchers used indirect calorimetry, which measures oxygen intake and carbon dioxide output to track metabolism, and found that the modified mice burned through their carbohydrate stores prematurely rather than relying on fat reserves.
Next, the researchers put the mice through a three-week treadmill training program to see if they could build endurance. The control mice steadily improved their running times and maximum distances over the three weeks. The mice with silenced SF1 neurons failed to build this endurance capacity. An analysis of their leg muscles showed an absence of the genetic changes that typically optimize energy use after regular training, showing that silencing the brain cells halted the muscles’ ability to adapt.
To track how the neurons behaved in real time, the team used tiny head-mounted microscopes to monitor calcium levels in the brains of active mice. Calcium flows into neurons when they fire, providing a visual proxy for cellular activity. During a week of daily training, the researchers observed a specific subset of SF1 neurons that consistently became active right after the exercise ended. Over the course of the week, the repeated training increased both the number of these post-workout neurons and the intensity of their firing.
The researchers then extracted brain tissue from mice that had exercised for three weeks and measured the electrical properties of individual SF1 neurons. Compared to tissue from sedentary mice, the cells from the trained animals were more electrically excitable and fired spontaneous action potentials at a higher rate. High-resolution imaging of the cells’ branches, or dendrites, revealed that the exercised mice had twice as many tiny protrusions called dendritic spines. This structural growth means the neurons had physically adapted to accommodate more synaptic connections.
Because the live imaging showed peak activity right after running, the team tested whether post-workout firing is the primary driver of endurance. They used optogenetics, a technique that allows researchers to turn specific neurons off using targeted light. By delivering light through fiber optic implants, they inhibited SF1 neurons for exactly 15 minutes immediately following each daily treadmill session. This brief post-workout suppression blunted the endurance gains that the mice would normally achieve.
In a final test, the researchers used a different optogenetic setup to artificially stimulate the SF1 neurons for one hour immediately after daily exercise. Activating the cells during this recovery window enhanced the animals’ endurance gains beyond their normal physiological plateau. The artificial stimulation triggered an increase in blood sugar and overall energy expenditure, which the researchers suggest may help initiate the metabolic recovery signals that prepare muscles for future physical stress.
These experiments rely on mouse models running on treadmills, and rodent metabolism does not perfectly map onto human physiology. Human endurance adaptations take place over longer time scales and involve more varied types of physical training. The results show that activating these brain cells is necessary for building stamina in mice, but they do not establish exactly how the brain receives the message that the exercise has stopped.
Identifying the physical pathways that carry signals from fatigued muscles back up to the ventromedial hypothalamus remains a goal for future studies. Researchers will also need to determine if different types of exercise, such as weightlifting or high-intensity interval training, recruit these exact same neural pathways. Mapping out this brain-to-body loop could eventually inform the development of therapeutics for individuals who cannot safely engage in rigorous physical activity.
The study, “Exercise-induced activation of ventromedial hypothalamic steroidogenic factor-1 neurons mediates improvements in endurance,” was published in May 2026. It was authored by Morgan Kindel, Ryan J. Post, Kyle Grose, Louise Lantier, Eunsang Hwang, Jamie R.E. Carty, Lenka Dohnalova´, Lauren Lepeak, Hallie C. Kern, Rachael Villari, Nitsan Goldstein, Emily Lo, Albert Yeung, Lukas Richie, Bridget Skelly, Jenna Golub, Manmeet Rai, Teppei Fujikawa, Julio E. Ayala, Joel K. Elmquist, Christoph A. Thaiss, David H. Wasserman, Kevin W. Williams, Erik B. Bloss, and J. Nicholas Betley.