Every year, 300+ million people are rendered unconscious by molecules that simultaneously reshape neural oscillations, fragment brain connectivity, suppress metabolism by 40-60%, and depress virtually every organ system. This is what happens.
Unlike most drugs that target a single receptor, anesthetics work through a constellation of molecular targets. The two dominant systems are GABAergic inhibition and glutamatergic excitation.
The GABA-A receptor is a pentameric chloride channel — most commonly assembled as α1β2γ2. When anesthetics bind, inward chloride flux hyperpolarizes neurons, suppressing signal transmission throughout the CNS.
| Agent | Primary Target | Clinical Conc. | Primary Action | Key Feature |
|---|---|---|---|---|
| Propofol | GABA-A (β+/α−) | 1-3 µM | Potentiation + direct gating | 97-98% protein bound |
| Sevoflurane | GABA-A (α TMD) | 330 µM | Phasic + tonic inhibition | Low blood:gas = 0.65 |
| Ketamine | NMDA (PCP site) | 1-2 µM free | Open-channel block | Preserves airway reflexes |
| Isoflurane | GABA-A + Glycine | 280 µM | Potentiation + Na+ block | Coronary vasodilator |
| Xenon | NMDA (glycine site) | 1.9 mM | Competitive inhibition | No psychotomimetic effects |
| Etomidate | GABA-A (β2/β3) | 2-3 µM | Potentiation | Hemodynamic stability |
| Thiopental | GABA-A (α+/β−) | 25 µM | Prolongs opening duration | t½ = 5-26 hours |
The convergence point for all anesthetic classes is the disruption of organized information flow across brain networks. The thalamic switch hypothesis holds that unconsciousness results from hyperpolarization block of thalamocortical relay neurons.
PET data consistently shows the thalamus exhibits the highest relative metabolic suppression — 48-66% reduction under propofol, particularly the medial and intralaminar nuclei.
PCI measures integrated information via TMS-EEG. Consciousness threshold: 0.31
Propofol not only suppresses within-network DMN connectivity but abolishes the normally robust anticorrelation between default-mode and task-positive networks. The frontoparietal network changes first, followed by the DMN — consciousness loss proceeds from disrupted external awareness to disrupted self-referential processing.
Hypersynchronous oscillations decrease differentiation. Connectivity breakdown reduces integration. Result: Φ (integrated information) falls below the threshold required for consciousness.
Anesthetics prevent the explosive frontoparietal "ignition" that makes information globally accessible. Propofol selectively suppresses frontal→parietal feedback while preserving feedforward connectivity.
Different anesthetic agents produce distinct EEG patterns that reflect their unique molecular targets and circuit-level mechanisms. These signatures serve as the physiological basis for brain monitoring during anesthesia.
| Band | Frequency | Normal State | Under Anesthesia |
|---|---|---|---|
| Slow | < 1 Hz | Deep sleep | Deep anesthesia, cortical UP/DOWN states |
| Delta | 1-4 Hz | Deep sleep | All agents at surgical depth |
| Theta | 5-8 Hz | Drowsiness | Sevoflurane signature; ketamine |
| Alpha | 9-12 Hz | Relaxed wakefulness | Propofol frontal anteriorization |
| Beta | 13-25 Hz | Active cognition | Nitrous oxide; early sedation |
| Gamma | 26-80+ Hz | Higher cognition | Ketamine gamma bursts (paradox) |
General anesthesia produces dose-dependent depression of virtually every physiological system. The effects range from predictable hemodynamic changes to subtle immune modulation that may influence long-term outcomes.
Minimum Alveolar Concentration — the concentration at which 50% of subjects don't move to surgical incision. Lower MAC = higher potency.
How quickly plasma concentration drops 50% after stopping an infusion. Critical for predicting recovery.
The cognitive consequences of general anesthesia range from transient confusion to potentially persistent neurocognitive changes. Age is the strongest risk factor. The ISPOCD1 study found POCD in 25.8% of elderly patients at one week.
Meta-analysis of 8 RCTs (3,555 patients) found no significant difference in POCD between general and regional anesthesia — surgery itself, not anesthesia type, appears to be the primary driver.
At the cellular level, anesthetics trigger the intrinsic mitochondrial apoptotic pathway, particularly devastating during the brain growth spurt period.
| Trial | Design | Population | Key Finding | Conclusion |
|---|---|---|---|---|
| GAS | International RCT | 772 infants | IQ: 99.08 vs 98.97 | No difference at 5 years (single brief exposure) |
| PANDA | Sibling-matched | 105 sibling pairs | IQ: 111.0 vs 109.7 | No difference in IQ or secondary outcomes |
| MASK | Population cohort | 997 children | Multiple exposures ↓ processing speed | Cumulative risk with multiple exposures |
Despite monitoring advances, patients occasionally regain consciousness during surgery. The UK's NAP5 audit of >2.7 million anesthetics remains the definitive epidemiological study.
PTSD develops in up to ~50% of awareness victims. Symptoms include flashbacks, nightmares, avoidance of medical settings, and emotional distress persisting months to years. Paralysis was reported as the most distressing aspect.
66% of AAGA events occurred during induction or emergence — the dynamic phases when drug concentrations are changing rapidly.
The Meyer-Overton correlation (1899-1901) showed anesthetic potency correlates with lipid solubility across five orders of magnitude. But the deeper mystery persists: why do drugs targeting completely different receptors all produce the same end-state of unconsciousness?
Eisen, Miller, Fiete et al. demonstrated in Cell Reports (March 2026) that three mechanistically distinct anesthetics — propofol, ketamine, and dexmedetomidine — all produce an identical dynamic brain destabilization measurable via EEG.
The awake brain maintains a balance between stability and excitability ("knife's edge"). All agents destabilize this balance until consciousness collapses. A prototype closed-loop delivery system is in development.
| Feature | NREM Sleep | General Anesthesia | Coma |
|---|---|---|---|
| Reversibility | Endogenous, cyclical | Pharmacological, controllable | Variable, injury-dependent |
| EEG | Slow-delta, spindles, K-complexes | Agent-specific signatures | Low-amp delta, burst suppression |
| Arousal | Moderate threshold | Very high (intentional) | Extremely high or absent |
| Burst suppression | Never occurs | At deep planes | In severe cases |
| Emergence | Seconds to minutes | Minutes | Hours to years |
Emergence occurs at lower drug concentrations than induction. The brain resists state transitions. Loss of consciousness involves gradual increases in temporal autocorrelation, while recovery shows an abrupt "reboot" — pharmacokinetics alone cannot explain this asymmetry.
Orexinergic/noradrenergic systems must overcome deep-sleep-like state
Unconscious state has deeper attractor basin — harder to escape
Lateral hypothalamus orexin neurons determine emergence timing