Ketamine rapidly improves depression and anxiety, but insomnia follows distinct pathway

Seventy-one patients with treatment-resistant depression arrived at a research clinic carrying the weight of three overlapping burdens: a depression that had resisted standard medications, anxiety that wouldn't lift, and insomnia that kept them awake at night. Within twenty-four hours of receiving a single intravenous dose of ketamine—0.5 milligrams per kilogram of body weight—all three symptoms improved measurably. But the mechanism behind that improvement, researchers discovered, was not uniform. Ketamine did not simply turn down the volume on all three problems at once. Instead, it appeared to work through separate pathways, particularly for sleep.

The study, conducted under the auspices of the National Institutes of Health, measured depression using the Montgomery-Åsberg Depression Rating Scale, anxiety using the Hamilton Anxiety Rating Scale, and insomnia using sleep-related items from the Hamilton Depression Rating Scale. The results were unambiguous: depression scores dropped by an average of 11.72 points, anxiety by 7.87 points, and insomnia by 0.89 points—all statistically significant. Every symptom domain moved in the same direction. The drug worked.

What happened next, however, revealed something more intricate. The researchers examined whether improvements in one symptom predicted improvements in another. They found that when a patient's anxiety got better, their depression tended to improve as well—the two symptoms were tightly linked, with a correlation of 0.61. But insomnia did not follow this pattern. Patients whose sleep improved did not necessarily see their mood lift, and those whose mood improved did not necessarily sleep better. The insomnia response operated on its own track, independent of the depression and anxiety changes.

At baseline, before any treatment, insomnia had been moderately correlated with anxiety, and anxiety with depression—the three symptoms clustered together as they often do in clinical practice. After ketamine, those baseline relationships persisted, but the changes diverged. This suggested that ketamine was not simply reducing overall psychological distress, which would have lifted all three symptoms together. Instead, it appeared to engage distinct biological mechanisms for sleep versus mood.

When researchers looked at sleep architecture itself—the actual structure of sleep as measured by polysomnography—they found short-term increases in total sleep time, in slow-wave sleep (the deepest, most restorative stage), and in sleep efficiency. Sleep latency, the time it takes to fall asleep, decreased. These are the hallmarks of improved sleep quality, independent of mood. Patients whose insomnia was most severe at the start showed the largest absolute reductions in insomnia symptoms, suggesting that baseline severity predicted response.

The finding carries practical weight for clinicians. If insomnia in treatment-resistant depression operates through a separate neurobiological pathway than mood symptoms, then future treatments might be designed to target these pathways independently or in combination, tailored to which symptoms dominate a particular patient's presentation. A person whose primary burden is sleeplessness might benefit from a different approach than one whose primary burden is despair, even though both conditions often travel together.

The study was registered with ClinicalTrials.gov under identifier NCT00088699 and was funded by the NIH's Intramural Research Program. One of the lead researchers, Dr. Zarate, holds patents related to ketamine's use in depression and suicidal ideation, though he has assigned those rights to the U.S. government. The work opens a door to understanding not just whether ketamine works, but how it works—and for whom it might work best.