Why Women and Men Use Energy Differently During Exercise

Women burn more fat and less carbohydrates than men during aerobic exercise. This has been replicated in enough studies that it's not controversial. What is still being figured out is the mechanism, which matters because mechanism is what lets you build interventions, models, and predictions.

During my research term at the University of Waterloo, I helped build a mathematical model to trace the causal chain. The model showed something surprising: many of the sex differences are in the liver. And the implications go well beyond exercise physiology.

The quantitative difference

During moderate-intensity aerobic exercise (60% of your VO₂max), the difference in macronutrient oxidation rates is large. After 60 minutes of exercise:

Women: 56% energy from carbohydrates, 44% energy from fat
Men: 68% energy from carbohydrates, 32% energy from fat

This is not explained by body composition. The sex difference persists even when fat mass is controlled. Women's metabolic machinery is configured to preferentially oxidize fat, independent of how much fat they're carrying.

The liver as the control point

In the model, the key control point was hepatic GIR resistance. GIR is the glucagon-to-insulin ratio, and it's the primary hormonal signal that drives the liver to break down glycogen and release glucose during exercise.

Here's what happens hormonally. During exercise, glucagon rises in both sexes, but women have higher absolute glucagon levels throughout. Women also have drastically lower epinephrine: men have 67% higher epinephrine at rest and 175% higher after 60 minutes of exercise. The GIR is higher in women both at rest and during exercise.

You'd expect higher GIR to drive more glycogen breakdown. But in the female model, the liver doesn't respond as aggressively. The female liver is resistant to GIR-mediated glycogenolysis. It spares hepatic glycogen instead of dumping glucose into the bloodstream.

In the model, the result is lower arterial glucose availability, which shifts substrate utilization systemwide.

The cascade

Lower arterial glucose means skeletal muscle has to run on something else. In the female model, that something else is plasma free fatty acids (FFAs) released from adipose tissue. In the male model, skeletal muscle relies more on intramyocellular lipids (fat stored inside the muscle cells) and on circulating glucose.

This is not just a quantitative shift. It's a qualitative difference in metabolic architecture. Women mobilize fat from adipose stores and transport it through the bloodstream to working muscles. Men use local fat depots and circulating glucose.

The redox mechanism links it together. Lower glycogenolysis reduces glucose availability systemwide. Lower glucose means less pyruvate production via glycolysis. Less pyruvate oxidation means less NADH production and higher NAD⁺/NADH ratio. That elevated redox ratio directly activates FFA oxidation in skeletal muscle, heart, liver, and adipose tissue.

The chain: hepatic GIR resistance → glycogen sparing → reduced arterial glucose → lower glycolytic flux → elevated NAD⁺/NADH → FFA oxidation activation.

Why this matters outside of exercise

If the female liver responds differently to glucagon during exercise, it probably responds differently to other metabolic signals too. Drug metabolism. Glucose regulation during fasting. Insulin resistance. Ketone production. Dietary interventions.

Most metabolic models in the literature are parameterized on male subjects or male-dominated datasets. The default assumption has been that sex differences are minor, hormonal adjustments to a shared baseline. This model suggests otherwise. The female metabolic control architecture is not the male architecture with estrogen added. It's a different configuration with different setpoints, different feedback loops, different tissue-level priorities.

The popular shorthand is that women are not small men. The mechanistic version is more interesting: female metabolism is not just a scaled version of male metabolism. The data supports that. The liver, which is regarded as the best example of a sexually dimorphic non-reproductive organ, shows thousands of sex-specific gene expression patterns. Estrogen regulates hepatic gluconeogenesis, glycogenolysis, fatty acid metabolism, and cholesterol metabolism through estrogen receptor signaling.

The data problem

When we parameterized this model, the hardest part was finding data. Female-specific metabolic rate data is sparse. Studies that control for menstrual cycle phase, training status, body composition, and dietary intake simultaneously are rare. Most exercise physiology datasets are male-only or mixed-sex with no sex stratification.

This is where simulation comes in as a powerful tool. In the absence of female-specific data, we can fill in the gaps and make predictions about what we don't know. All in an attempt to build a better understanding of the world.