Your Brain Tells Your Liver to Spike Blood Sugar When You're Stressed
A newly discovered amygdala–liver neural circuit drives stress hyperglycemia independently of cortisol or adrenaline — and chronic stress breaks it.
Summary
Researchers at Mount Sinai have identified a direct brain-to-liver signaling pathway that explains why blood sugar spikes during stress. When mice encounter a threat, neurons in the medial amygdala activate a chain of signals that reaches the liver and triggers rapid glucose production via gluconeogenesis. Remarkably, this happens independently of cortisol, adrenaline, glucagon, or insulin — the hormones traditionally thought to drive stress hyperglycemia. The circuit also suppresses appetite during stress. Repeated stress exposure disrupts this pathway, producing diabetes-like dysregulation of blood glucose. The findings, published in Nature, reveal a previously unknown amygdala–liver axis that may help explain why chronic psychological stress is a major risk factor for type 2 diabetes.
Detailed Summary
Stress reliably raises blood sugar — a response long attributed to cortisol and adrenaline flooding the bloodstream. But a landmark study published in Nature from the Icahn School of Medicine at Mount Sinai reveals a fundamentally different mechanism: a direct neural circuit running from the medial amygdala (MeA) through the ventromedial hypothalamus (VMH) to the liver that drives rapid hyperglycemia during stress, entirely independently of classical stress hormones. This discovery reframes how we understand the brain's control of glucose metabolism and its connection to metabolic disease.
The researchers began by characterizing metabolic responses to acute stress in C57Bl/6 mice. Both physical restraint stress (30 minutes) and social stress (exposure to a conspecific male's odor in a territorialized cage) rapidly elevated blood glucose and impaired glucose tolerance. Restraint stress increased plasma corticosterone, adrenaline, glucagon, and glycerol, and upregulated hepatic G6pc (glucose-6-phosphatase) gene expression — a key enzyme in gluconeogenesis — without altering plasma insulin or noradrenaline. Critically, even 5-minute exposures to either stressor were sufficient to raise blood glucose and corticosterone, demonstrating the speed of the response.
Using FOS immunostaining and in vivo fiber photometry with GCaMP8s calcium sensors, the team showed that MeA neurons were rapidly and robustly activated by multiple stressor types — restraint, footshock, social odor, and even a visual threat (an approaching robotic 'robobug'). Importantly, MeA calcium signals rose before blood glucose increased, establishing the amygdala as an upstream driver rather than a downstream responder. MeA activity was not altered by novel clean cage exposure or locomotion, ruling out movement as a confound.
Chemogenetic activation of MeA neurons using excitatory hM3DGq DREADDs (CNO 3 mg/kg, intraperitoneal) significantly elevated blood glucose in unstressed mice without raising plasma corticosterone or adrenaline. This hyperglycemic effect persisted even when corticosterone synthesis was blocked with metyrapone, confirming that the MeA-driven glucose response is independent of HPA axis activation. Whole-body viral tracing identified a polysynaptic pathway from MeA neurons projecting to the VMH, which in turn connects to the liver via autonomic relays, promoting hepatic gluconeogenesis. Inhibition of MeA-to-VMH (MeA^VMH) neurons blunted stress-induced hyperglycemia, while their activation recapitulated it — along with hypophagia, mirroring the full metabolic stress response.
Perhaps most clinically significant, repeated stress exposure disrupted this circuit's normal function, producing a diabetes-like state of dysregulated glucose homeostasis. This provides a mechanistic link between chronic psychological stress and the development of type 2 diabetes — a relationship long observed epidemiologically but poorly understood at the circuit level. The study is currently limited to mouse models, and whether an analogous amygdala–liver axis operates in humans remains to be established. Nonetheless, the identification of this hormonally independent neural pathway opens new avenues for targeting stress-related metabolic dysfunction through brain-based interventions.
Key Findings
- Acute restraint stress (30 min) rapidly elevated blood glucose and impaired glucose tolerance in food-restricted C57Bl/6 mice, with effects detectable after as little as 5 minutes of stress exposure
- Restraint stress upregulated hepatic G6pc (glucose-6-phosphatase) gene expression in fed mice, indicating increased gluconeogenic capacity, without significant changes in Pck1 or liver glycogen
- MeA calcium signals (GCaMP8s fiber photometry) rose before blood glucose increased during stress, establishing MeA as an upstream driver of the glycemic response
- Chemogenetic activation of MeA neurons (hM3DGq DREADD, CNO 3 mg/kg) significantly raised blood glucose without increasing plasma corticosterone or adrenaline, and the effect persisted after corticosterone synthesis blockade with metyrapone
- Whole-body viral tracing identified a polysynaptic MeA→VMH→liver circuit that drives hepatic gluconeogenesis independently of adrenal or pancreatic glucoregulatory hormones
- MeA^VMH neuron activation also produced hypophagia, recapitulating the full metabolic stress phenotype (hyperglycemia + appetite suppression)
- Repeated stress exposure disrupted MeA circuit control of blood glucose, producing diabetes-like dysregulation of glucose homeostasis
Methodology
The study used C57Bl/6 mice subjected to multiple acute stressor paradigms (restraint, social odor/territorialized cage, footshock, visual threat). Neural activity was measured via fiber photometry with GCaMP8s calcium sensors; circuit mapping used whole-body viral tracing. Chemogenetic manipulation employed AAV-hSyn-hM3DGq-mCherry DREADDs with CNO (3 mg/kg i.p.) versus AAV-hSyn-mCherry controls. Hormonal independence was confirmed using the corticosterone synthesis inhibitor metyrapone (50 mg/kg s.c.). Statistical analyses and exact sample sizes are reported in Supplementary Data Table 1.
Study Limitations
The study was conducted entirely in mice, and direct translation of the MeA^VMH–liver circuit to human physiology has not yet been demonstrated. The chemogenetic tools used (DREADDs) activate broad neuronal populations rather than specific cell types, which may not fully replicate the selectivity of endogenous stress signaling. The paper does not report conflicts of interest disclosures in the provided text.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
