A patient starts an antidepressant and, weeks later, notices a faint ringing that was not there before. The doctor is skeptical. The drug is supposed to help. But a study published in early 2026 in the Proceedings of the National Academy of Sciences suggests the patient may be onto something: a specific serotonin-releasing circuit in the brainstem can actively drive tinnitus-like perception in mice, and activating it makes the phantom sound worse, not better.
The finding lands in uncomfortable territory. Selective serotonin reuptake inhibitors, or SSRIs, are among the most widely prescribed medications on the planet, taken by tens of millions of people for depression and anxiety. Tinnitus, the perception of sound without an external source, affects roughly 10 to 15 percent of adults, according to estimates from the National Institute on Deafness and Other Communication Disorders. The overlap between those two populations is enormous. And until now, the prevailing clinical assumption has been that raising serotonin levels should broadly help mood without meaningfully affecting hearing. This study challenges that assumption at the circuit level.
What the OHSU team found
The research comes from the Vollum Institute at Oregon Health & Science University, led by senior authors Laurence Trussell and Zheng-Quan Tang. Using optogenetics, a technique that uses light to switch targeted neurons on or off, the team zeroed in on serotonin-producing cells in a brainstem structure called the dorsal raphe nucleus (DRN) that send projections directly to the dorsal cochlear nucleus (DCN), a region responsible for early-stage sound processing.
When the researchers activated that serotonergic pathway, mice showed increased tinnitus-like behavior on a modified auditory startle test, a standard assay used to detect phantom sound perception in rodents. When they inhibited the same pathway, the tinnitus-like responses dropped. The result was strikingly specific: toggling serotonin release in this single circuit was sufficient to turn the phantom-sound behavior up or down.
That precision matters because it moves the conversation beyond vague claims that “serotonin affects hearing.” Prior electrophysiology work had already established that serotonin shapes how DCN neurons integrate auditory and multisensory signals, and that the DCN receives dense serotonergic input from the brainstem. A separate 2011 study demonstrated that mice with behavioral signs of tinnitus show hyperactivity in the DCN driven by weakened inhibitory signaling. The new optogenetics data add serotonin as a second, independent driver of that same hyperactivity, placing the DCN at a junction where multiple chemical signals can tip auditory processing toward phantom perception.
A signal in the adverse-event data
The laboratory findings do not exist in a vacuum. A disproportionality analysis of the FDA’s Adverse Event Reporting System (FAERS) database, covering voluntary reports filed between 2004 and 2023, has flagged elevated reporting rates for tinnitus associated with antidepressant use. FAERS data are not controlled clinical evidence; reports are filed voluntarily by patients, clinicians, and manufacturers, and they lack audiometric confirmation or precise dosing timelines. But the statistical signal is consistent with the biological story emerging from the mouse work.
Taken alone, the FAERS signal could reflect reporting bias. Patients on antidepressants may be more attuned to bodily sensations, or more likely to mention symptoms to a prescriber who then files a report. Paired with a plausible causal mechanism in animals, however, the two lines of evidence reinforce each other. A biological explanation makes the population-level signal harder to dismiss as coincidence.
What the study cannot tell us yet
The most important caveat is the species gap. Every piece of circuit-level evidence so far comes from mice. No human neuroimaging or post-mortem study has confirmed that the DRN-to-DCN serotonergic projection behaves identically in people with tinnitus. Mouse and human auditory brainstems share structural features, but translating optogenetic findings into clinical predictions requires more than anatomical analogy.
The experiments also tested acute activation and inhibition of the pathway. Chronic serotonin elevation, which more closely mirrors what happens during weeks or months of SSRI therapy, has not been studied in behaving animals within this circuit framework. Whether sustained exposure produces the same tinnitus-driving effect, or whether the brain compensates over time, remains an open question.
Individual variability adds another layer of uncertainty. Not everyone on SSRIs reports tinnitus, and not everyone with tinnitus takes antidepressants. Genetic differences in serotonin receptor expression, prior noise exposure, pre-existing hearing loss, and co-occurring psychiatric conditions could all influence how strongly a given person’s DCN responds to elevated serotonin. The OHSU study was not designed to address those human differences, so it cannot explain why some patients appear more susceptible than others.
No prescribing guidelines have been updated in response to these findings, and no clinical trials have directly compared tinnitus outcomes in patients taking SSRIs versus alternative antidepressants that do not primarily target serotonin. Formal clinical guidance, if it comes, will likely require human data that do not yet exist.
How to weigh the evidence
The PNAS optogenetics study sits at the top of the evidence hierarchy for basic neuroscience. It uses a controlled experimental design in which a single variable, serotonin release in one circuit, was manipulated while a specific behavioral output was measured. That kind of causal test is the gold standard in preclinical research, even though it cannot be replicated in living human brains for ethical and technical reasons.
The supporting electrophysiology literature provides a plausible mechanism. Serotonin does not simply amplify sound signals in the DCN; it alters how the region weighs competing inputs from auditory and non-auditory sources. That reweighting could explain why some patients perceive sounds that have no external origin, or why existing tinnitus becomes more intrusive during periods of heightened stress or arousal, states in which serotonin activity fluctuates.
But it is equally important not to overextend early-stage findings. SSRIs have decades of clinical evidence supporting their effectiveness against severe depression and anxiety, conditions that themselves can amplify the perceived loudness and distress of tinnitus. Stopping medication abruptly can trigger withdrawal symptoms and psychiatric relapse. For patients currently taking SSRIs who also experience tinnitus, this research does not justify changing medication without a clinician’s involvement.
Where the research points next
The study’s most lasting contribution may be the target it hands to drug developers. By identifying a specific DRN-to-DCN circuit that modulates tinnitus-like behavior, the OHSU team opens the door to more selective interventions. Future compounds might block particular serotonin receptor subtypes in the DCN, or dampen the activity of the serotonergic fibers projecting there, without broadly suppressing serotonin in mood-regulating regions of the brain. That kind of precision pharmacology does not exist yet for tinnitus, but the circuit map now provides a starting point.
Non-drug approaches could also benefit. Neuromodulation techniques such as transcranial magnetic stimulation might eventually be calibrated to influence DCN excitability, counteracting the hyperactivity linked to tinnitus without touching serotonin directly. Behavioral therapies that reduce stress and anxiety could still play a complementary role by lowering the brain’s overall arousal state, potentially making the DCN less reactive to serotonergic input even if the underlying circuit vulnerability remains.
For patients right now, the practical step is straightforward: anyone who starts or adjusts an SSRI and notices new or worsening ringing in the ears should document the timing, the dose, and any other medications or noise exposures, and bring that information to a clinician promptly. Clinicians treating patients with significant noise-exposure history or pre-existing hearing loss may want to ask about auditory symptoms more routinely when prescribing serotonergic drugs.
The picture that emerges from this research is not one of a simple villain. Serotonin remains essential for mood regulation, sleep, and dozens of other functions. But the OHSU findings make clear that its effects are not uniform across the brain. In at least one auditory circuit, more serotonin appears to mean more phantom sound. As researchers work to confirm whether the same holds true in humans, both patients and clinicians will need to navigate that complexity with the evidence available, incomplete but, as of May 2026, considerably sharper than it was a year ago.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
