Your brain’s insulin sensitivity isn’t constant—it shifts across the menstrual cycle

Most of us learn about insulin as a blood-sugar hormone. But insulin is also a signal to the brain. It shapes appetite, reward, satiety (the feeling of fullness), and decision-making.

A recent study examined whether the menstrual cycle changes how the brain responds to insulin, and whether this could help explain cycle-linked shifts in food desire (Hummel et al., 2026). Have you not heard of luteal brownie cravings?! Instagram certainly has.

And by the way, the short answer was yes.

The authors show that insulin’s effects on large-scale brain networks flip between the follicular and luteal phases. In the luteal phase, food cues trigger stronger responses in memory and reward circuits. This means that the same food, say a bar of chocolate, engages the brain’s desire a lot more if you’re nearing your period (luteal/PMS) than if your period just finished (follicular). This indicates that hormone changes help determine when and how strongly insulin can influence brain networks.

How they studied “brain insulin”

The study included 15 healthy women with regular menstrual cycles. Each participant took part twice: once in the follicular phase and once in the luteal phase. This setup, known as a within-subject cross-over design, meant that each woman was compared to herself at different points in the cycle rather than to other participants. This is an important control because it makes cycle-related insulin changes easier to detect and acknowledges that every woman is different, but that the changes are still shared regardless of our different baselines.

At each visit, participants had a brain scan using functional magnetic resonance imaging (fMRI), a non-invasive method that measures changes in blood flow to show which brain regions are active and how they communicate at rest. Scans were taken before and about 30 minutes after insulin was delivered through a nasal spray, which allows insulin to reach the brain directly while minimizing effects on the rest of the body.

About 40 minutes after insulin, participants completed a task that involved viewing images of different foods.

To confirm cycle phase and link hormone levels to brain activity, blood samples were taken to measure several reproductive hormones, including estradiol (the primary form of estrogen) and progesterone.

Food cues hit differently in the luteal phase

After insulin, participants viewed images of sweet and savory foods. Activity in the hippocampus (DMN) and dorsal striatum (SN) —regions involved in memory, reward, and decision-making, was higher in the luteal phase, particularly in response to sweet foods.

The standout finding? This effect was not driven by hunger. Food desire ratings were higher in the luteal phase, but hunger itself did not differ significantly. This points away from basic energy needs and toward changes in motivational salience and learned reward.

In other words, food cues carried more pull in the luteal phase, even when hunger was unchanged.

Insulin also hits your brain networks in opposite ways based on your cycle phase

The authors also looked at resting-state functional connectivity, which refers to how strongly different regions within a brain network communicate at rest.

They examined three major networks:

• Default Mode Network (DMN): involved in memory, internal thought, and goal-directed behavior
• Salience Network (SN): helps prioritize what matters and coordinate responses
• Somatosensory network (SMN): involved in bodily sensation and readiness to act

In the follicular phase, insulin increased connectivity in the hippocampus (part of the DMN) and the putamen (part of the salience network), while decreasing connectivity in the supplementary motor area (SMA).

In the luteal phase, this pattern reversed. Insulin reduced connectivity in the hippocampus and putamen and increased connectivity in the SMA.

Together, these findings suggest a phase-specific shift in how the brain responds to insulin.

The bigger picture

Taken together, the findings point to cycle-linked “metabolic brain states”: temporary, hormone-linked patterns of brain activity that reflect how the brain is processing metabolic signals like insulin. Again, showing that many of the effects that we associate with hormones, such as cravings, are instead brain-based signals impacted but not fully described by hormonal shifts.

In the follicular phase, insulin more strongly supports networks involved in cognition, salience, and satiety. In the luteal phase, those effects weaken or reverse, while food cues (especially sweet ones) become more motivating, even without increased hunger.

While this study is small and lacks a placebo control, it adds something concrete to the conversation. The findings reinforce that the brain is the control center for metabolic behavior, and that cycle-related shifts in food desire are driven by measurable changes in brain function rather than cravings fueled by imagination or lack of restraint.