Anesthesia May Not Work the Way We Thought It Did

iStock / iStock

You lie back, and a nurse fits a mask over your face. Somebody tells you to count backward from 100. Your eyelids grow heavy. The next thing you know, you’re waking up. We thought we knew why this happens, but new research published in the journal PLOS Computational Biology suggests we may have had it wrong.

The brains of people on general anesthesia are far quieter than those of folks who haven’t been drugged. Previous studies have suggested that this quieting happens when anesthesia interferes with conversations, or couplings, between different parts of our brain. Less information is exchanged, and the volume of the conversation drops.

It seemed like a solid enough explanation. But a team of German neuroscientists saw a possible flaw in the logic. The amount of information being exchanged often depends on the amount of information available, not on the strength of the connection.

To explore this puzzle further, they brought two female ferrets into the lab and hooked them up to brain activity monitors. (Ferret brains’ similarity to primates’ makes them a good lab substitute for humans, at least in initial studies.)

Both ferrets went through three rounds of anesthesia and recovery, receiving slightly more of the drug each time as the scientists watched their brains produce, process, and exchange information.

As in previous studies, the conversations in the ferrets’ brains were indeed more subdued while they were anesthetized. But it wasn’t interference that quieted their brains. The brain regions that ordinarily do the listening were just as active as usual. But the talkative brain regions seemed to have less to say. They were making and sending less information.

Lead author Patricia Wollstadt is a neuroscientist at the Brain Imaging Center at Goethe University Frankfurt. "The relevance of this alternative explanation goes beyond anesthesia research,” she said in a statement, "since each and every examination of neuronal information transfer should categorically take into consideration how much information is available locally and is therefore also transferable."