Subscribe to eTOC

Sniff Test for Unresponsive Patients Offers New Tool for Prognosis

Article In Brief

Response to a “sniff test” may offer prognostic clues on the potential for recovery in people who are unconscious after a traumatic brain injury.

Patients who are unresponsive following a severe brain injury but who measurably reduce their inhalation of strong odors are far more likely to eventually regain consciousness than are those whose breathing does not change, according to the results of a study published in Nature.

The so-called “sniff response” test was highly specific but not perfectly sensitive: 100 percent of those whose respiration changed in response to the strong odors later regained consciousness, but so did 35 percent of those who did not.

Still, the finding offers an accessible, inexpensive bedside tool for clinicians to assess the likelihood of recovery, said neurologists who specialize in disorders of consciousness but who were not involved in the current study. They offered high praise for the paper, calling it “incredibly important” and a “landmark.”

“The idea that something as simple yet important as smelling can help clinicians understand these very challenging but often misdiagnosed patients is really great,” said Steven Laureys, MD, PhD, director of the Brain Clinic at the University Hospital of Liège in Belgium and professor in the Coma Science Group & GIGA Consciousness Research Unit at the University of Liège, who was not involved with the study.

The authors of the paper hypothesized that they would see different reactions to the two strong odors they used—increased inhalation to the pleasant odor of shampoo or decreased inhalation in response to the unpleasant odor of rotting fish.

“We knew from previous studies that odors are effective in moderating respiration,” said the first author of the study, Anat Arzi, PhD, a neurobiologist who is the Blavatnik Fellow in the department of psychology at the University of Cambridge. “In our patient group, we saw decreased respiration in response to both odors, but no distinction between the two kinds.”

Dr. Arzi, who was a postdoctoral fellow at the Weizmann Institute of Science in Rehovot, Israel, at the time of the study, has previously demonstrated with colleagues there that strong odors can be used to induce learned responses in sleeping individuals. In 2014, she reported that a single night of aversive conditioning, by linking profoundly unpleasant odors with the smell of cigarette smoke during sleep, significantly reduced the number of cigarettes smoked in the following days.


“It doesnt require expensive equipment. It may be more generalizable for institutions around the world.”—DR. BRIAN L. EDLOW

The senior author of those papers and the new one was Noam Sobel, PhD, head of the department of neurobiology at the Weizmann Institute.

“Olfaction has a unique interaction with consciousness,” Dr. Arzi said. “At the neuroanatomical level, there is a direct pathway to the cortex that bypasses the thalamus. No other sense has this. And I don't have to ask you for your response; I can just measure your respiration.”

She added: “It's such a simple test. If you have a spirometer, you can present an odor and just look to see if the respiration changes. In some patients you can see it by eye, with others you need a statistical measure. It is fast, simple, accessible, affordable and can be done at the bedside.”

Study Details

The study sought to detect both sensory-driven, automatic changes in nasal airflow in response to odors, and a cognitively-driven response reflecting situational understanding or learning. For instance, in addition to presenting the patients with pleasant or unpleasant odors, at times they would tell the patients that they would be presented with an odorant but would present only an empty jar beneath their nose. If they then modified their nasal airflow, that would imply possible awareness.

The study involved 43 patients with disorders of consciousness, diagnosed using standard clinical measures to determine whether the patient was in a minimally conscious state (MCS), in which the patient showed inconsistent but reproducible evidence of consciousness, or unresponsive wakefulness syndrome (UWS) also known as vegetative state, in which the patient showed no signs of consciousness.

The researchers conducted 73 sessions with 31 MCS patients, and 73 sessions with 24 UWS patients. (Sixteen of the 24 patients transitioned from UWS to MCS during the study.) Immediately following each testing session, the state of consciousness of the patient was assessed using standard bedside measures. The investigators conducted 146 sessions (between one to 12 sessions per patient), with intervals between the sessions ranging from one to ten weeks.

At the group level, nasal inhalation volume was significantly reduced, by about 10 percent, in response to both shampoo and rotten fish among patients with MCS, but not among those with UWS.


“Olfaction has a unique interaction with consciousness. At the neuroanatomical level, there is a direct pathway to the cortex that bypasses the thalamus. No other sense has this. And I dont have to ask you for your response; I can just measure your respiration.”—DR. ANAT ARZI

To assess individual-level responses, the researchers set a threshold of a 15 percent change in nasal airflow volume as a measure of whether or not the patient showed a sniff response. They found that 20 of 31 patients in an MCS displayed a sniff response during at least one session, compared with 10 of 24 with UWS.

“Remarkably,” the paper reported, “we observed that all 10 patients with UWS who had a sniff response in one session or more later transitioned to MCS. Thus, a sniff response in UWS indicated transition to MCS with 100 percent specificity and 62.5 percent sensitivity. Furthermore, in four of the patients the sniff response preceded any other sign of consciousness recovery by days to months (2.5 months, around 2 months, about 1.5 months, and 2 days). This suggests that sniff responses are informative for the prognosis of a patient at the single-patient level.”


“The idea that something as simple yet important as smelling can help clinicians understand these very challenging but often misdiagnosed patients is really great.”—DR. STEVEN LAUREYS


“This would be complicated to implement outside of a research setting. I dont think the average office practice or hospital will be ready to do this.”—DR. NICHOLAS D. SCHIFF

The study also found that the sniff response was useful in predicting survival at a mean of 14.7 months (range 14-64). Only two of 24 patients who had a sniff response soon after their injury had died by the time of follow-up, compared with 12 of 19 who did not have a sniff response.

“All patients with a sniff response ultimately showed signs of consciousness and all patients with UWS who remained unconscious did not have a sniff response,” the paper concluded. “This places the sniff response among the most-suitable analyses for the estimation of the recovery of consciousness in patients with disorders of consciousness.”

Expert Commentary

Brian L. Edlow, MD, director of the Laboratory for NeuroImaging of Coma and Consciousness at Massachusetts General Hospital, and assistant professor of neurology at Harvard Medical School, called the study a “landmark.”

“It's a groundbreaking, innovative contribution to our field, for several reasons,” Dr. Edlow said. “It provides additional evidence beyond fMRI and EEG that there are diagnostic tests to identify patients who may have consciousness that evades detection on the bedside behavioral exam. It also suggests we may have a new tool now to predict recovery of consciousness. But what is unique about this test, and may lead to its biggest clinical impact, is the ease of performing it at the bedside. It doesn't require expensive equipment. It may be more generalizable for institutions around the world.”

Based on the single-center results, Dr. Laureys said, “Now we need multicenter studies to show just how sensitive and specific this measure is at the single-patient level.”

Nicholas D. Schiff, MD, the Gerald B. Katz Professor of Neurology and Neuroscience at Weill Cornell Medicine in New York, agreed that larger studies of the method are necessary.

“If these findings get reproduced in a thousand or ten thousand patients, it will change practice forever,” Dr. Schiff said.

For now, however, given the complex testing protocol used in the current study, Dr. Schiff said he does not think it is ready for clinical application.

“This would be complicated to implement outside of a research setting,” Dr. Schiff said. “I don't think the average office practice or hospital will be ready to do this.”

Although the method's sensitivity was only 65 percent, the sniff test could still eventually be useful as an initial screen, Dr. Schiff said.

“If somebody doesn't have motor responses or other signs that allow a behavioral assessment, but they do have a sniff response, that could serve as a quick screen for sending someone to an MRI or a PET scan for a more thorough evaluation.”

Joseph T. Giacino, PhD, director of rehabilitation neuropsychology at Spaulding Rehabilitation Hospital and associate professor in the department of physical medicine and rehabilitation at Harvard Medical School, said the sniff test may have benefits over other methods in which patients are asked to imagine various activities while an fMRI is imaging their brain.

“There's no cognitive effort required for the sniff-response test,” he said. “In the active imaging paradigm, you have to imagine you're playing tennis, walking around your house, tapping your fingers. That's hard to do. Even healthy volunteers don't have an easy time with it. You really want tasks that require low cognitive effort but are still distinct from a purely sensory response. They accomplished that.”

Even so, Dr. Giacino said, “The holy grail is not predicting recovery of consciousness, but the level of functional recovery. That's what families want to know: Is he going to be able to care for himself? Will he be able to go back to work? We're not even close to being able to predict that.”

Ultimately, Dr. Giacino said, a variety of tests might be necessary to offer the most accurate prognosis for severely brain-injured patients.

“As we continue to get more tools in the chest, we will have a multi-modal approach,” he said. “We have the standardized coma recovery scare, the fMRI-based tests, PET scans, functional EEG, and now we'll add the sniff test. We might get a variety of permutations in the findings: anything from all five positive to all five negative or two positive and three negative. We'll need to know how to reconcile the disparities that might result.”

Promising as the results are, Dr. Laureys bemoaned the scarcity of rehabilitation programs that can maximize the capacity of brain-injured patients to recover.

“We have these very smart paradigms and high-impact papers, but in the end, what we need to see is a better qualify of life for these patients,” Dr. Laureys said. “It is still very difficult for families to find a center and the rehabilitation the patient deserves.”


Dra. Arzi, Giacino, Edlow, and Laureys had no disclosures.

Link Up for More Information

• Arzi A, Rozenkrantz L, Gorodisky L, et al. Olfactory sniffing signals consciousness in unresponsive patients with brain injuries Nature 2020;Epub 2020 Apr 29.
    • Braiman C, Fridman EA, Conte MM, et al. Cortical response to the natural speech envelope correlates with neuroimaging evidence in severe brain injury Curr Biol 2018; 28(23):3833–3839.e3.
    • Edlow BL. Covert consciousness: Searching for volitional brain activity in the unresponsive Curr Biol 2018;28(23):R1345–R1348.
    • Demertzi A, Tagliazucchi E, Dehaene S, et al. Human consciousness is supported by dynamic complex patterns of brain signal coordination Sci Advances 2019;5(2): eaat7603.
      • Claasen J, Doyle K, Matory A, et al. Detection of brain activation in unresponsive patients with acute brain injury N Engl J Med 2019;380(26):2497–2505.