Sniffing Out Diseases

A telltale COVID-19 symptom underscores the scientific significance of smell—for both patients and practitioners.

An unexpected byproduct of the novel coronavirus is that the most underrated of the five senses is suddenly making a stink. 

One of the more innocuous of a long list of physically destructive symptoms is losing your sense of smell and taste. We’re still not completely sure why the virus knocks out your ability to tell if the milk has gone bad—a condition called anosmia—but it’s presumed to be because the virus often gains access to the body via the olfactory nerve. (More specifically, through the angiotensin-converting enzyme II, or ACE-2.) Though it seems banal, the results can actually be dangerous: Not only can those left without a working sense of smell not detect a gas leak, but there is a solid correlation between anosmia and depression; when you can’t stop and smell the roses, life can lose a certain amount of lustre. Your days become as bland as your dinner tastes.

Losing your sense of smell may be one of the telltale signs of a coronavirus infection—but the smell of the virus itself could also potentially help us detect it. 

There is a long and well-studied history of diseases that stink. Some of the conditions that we can smell include tuberculosis, many cancers, and diabetes

(Not-so-fun fact: If you have friends who have been trying out the keto diet, the high-fat low-carb combination supposedly favored by cavemen, you might have heard them or their partners complain about their bad breath. This is caused by the presence of ketones, which are formed by the excessive fats their body is attempting to break down instead of turning to sugar for energy. This is the same unhealthy indicator for the highly dangerous condition diabetic ketoacidosis.)  

But how does a disease have a “smell” in the first place? Well, it’s more likely that you do. Whenever you breathe out, you exhale volatile organic compounds (VOCs) such as ethanol and acetone. The combined levels of these different biomarkers can give indications on the state of your health or the presence of an abnormal compound in your body. Some diseases and bacterias also have their own unique chemical structure that makes them easy to detect, meaning the presence of it on your breath or in your bodily fluid likewise indicates its presence in your body. For example, some microbiologists think that bacteria have certain whiffs: E. coli has a floral odor; Pseudomonas, the bacteria that commonly infects people in hospitals, is similar to either grapes or tortillas, depending on who you ask; and the bacteria in abscesses, Streptococcus anginosus, supposedly has a butterscotch vibe. Infectious diseases also have a reported history of certain stenches: Diphtheria smells sweet, yellow fever is like entering an abattoir, whereas typhoid fever is more like a bakery.

So what does COVID-19 smell like?

There are two main ways scientists are proposing to sniff out COVID-19. The first is with the assistance of our canine companions. While us mere humans have 6 million smell receptors in our noses, dogs have up to 300 million. They also have a specialized nose architecture that circulates scents aerodynamically; one group of researchers who strapped a plastic dog nose to a commercial vapor detector saw a 16-fold improvement in its ability to spot explosives. 

Using their highly attuned snouts, dogs are being trained to detect COVID-19. “Scent detection dogs can accurately detect low concentrations of … VOCs associated with various diseases such as ovarian cancer, bacterial infections, and nasal tumors,” says Cynthia Otto, a professor of Working Dog Sciences (yes that’s a thing), who is leading a study at University of Pennsylvania’s School of Veterinary Medicine. “The potential impact of these dogs and their capacity to detect COVID-19 could be substantial.” A group of scientists in Europe have also developed a proof of concept that dogs could sniff out COVID-positive patients using sweat samples.

This isn’t the first time that we’ve used dogs to detect disease. In addition to the medical community’s increasing comfort with canines being used for early-stage cancer diagnosis, other studies have shown them effective in diagnosing a malarial infection just from a child’s socks and for sniffing out early-stage Parkinson’s disease. They’ve also proven beneficial in alerting owners to the pre-symptomatic beginnings of epileptic fits, severe migraines, or bouts of narcolepsy, where taking medication before onset is vital. 

Diphtheria smells sweet, yellow fever is like entering an abattoir, whereas typhoid fever is more like a bakery.

The second sniffable COVID detection test comes from a rising biotech company called Koniku. Oshiorenoya Agabi’s startup claims to be able to accurately detect disease with a whiff. His device, the jellyfish-looking Konikore, is one day intended to be used as a diagnostic device that you breathe into to check your vitals against many different conditions. “Most diseases if not all diseases lead to a perturbation of a body’s signature or ‘baseline metabolomic profile,’” Agabi says. “The ability to measure these breath volatile organic compounds … is a game-changing proposition.”

His team has been working with Airbus since 2017 to develop bomb-sniffing devices for airport security, and now he’s looking for the biomarkers of the coronavirus to be able to tell if someone boarding a plane is infected, too. Building off recent research that suggests COVID-19 is detectable in exhaled breath, Koniku is using a combination of synthetic biology and machine learning to detect the virus. They will likely use the technology “as a perimeter screening to create a sterile space or biothreat-free space,” Agabi says. “The goal is restoring consumer confidence and safety in shared spaces.”

The Konikore is basically the Transformers version of a bionic nose. We already have bionic eyes and ears that have reconnected disabled communities with certain senses, and for decades, scientists have been developing different appendages that can smell just as well as humans do—or even better.

The most common is the Cyranose 320, which was originally intended to be used on the Space Shuttle to monitor air quality in space. Once trained on a scent, the handheld device can spot the presence of that odor in the air. After successful deployment in the chemical manufacturing space, where it was used for safety measures, it was developed into a non-invasive diagnostic tool to detect disease biomarkers in a patient’s breath or body fluids. Just a couple of recent successes include its ability to detect urinary tract infections, ventilator-associated pneumonia, and lung cancer. In addition to the Cyranose, there are also “sniff cams,” a biosensor that can tell when your food is about to go off, and many more artificial noses coming onto the market.

Joy Milne attempts to detect Parkinson’s disease. Chris Watt/Alamy

Sometimes we don’t need fancy devices or dogs to serve as a diagnostic test. Occasionally, a “super smeller” will do the trick. This is the common term for someone who suffers from hyposmia, which is an enhanced sense of smell. And “suffering” really is the correct verb to use: While anosmia describes someone who can’t tell the difference between a durian and a jackfruit, someone with hyposmia would live life continuously assaulted with smells, both good and bad. 

One famous hyopsmia-ite is Joy Milne, who can successfully spot early-onset Parkinson’s with alarming accuracy using only her nose. Milne first noticed that her husband had started emitting a musky smell a full 12 years before he was diagnosed. But was it a fluke? She worked with a researcher at the University of Edinburgh in a blind study where she tried to guess which patients had Parkinson’s from the scents of their T-shirts. She was right about all but one: a person in the control group, who had previously not tested positive for Parkinson’s. 

Then he was diagnosed eight months later. 

Since then, scientists have worked with Milne to pinpoint the smelly compounds in patients’ sebum that indicate the beginning of the disease. This means they can now test for the presence of these compounds on the skin. As early detection in Parkinson’s is particularly difficult, many doctors are hoping a swab test developed alongside Milne will be a quick, non-invasive way to determine neurodegenerative onset. She is also working to distinguish the compounds that give Alzheimer’s, cancer, and tuberculosis their distinct smells. 

Before you get too excited, it’s highly unlikely that Milne will determine what COVID-19 smells like. We spend so much of our days wiping down groceries with Clorox and spritzing our hands with sanitizer—who would want to put their nose right in a Petri dish of the stuff?

That doesn’t mean that doctors won’t continue to use all of their senses when it comes to diagnoses. Why does sight reign supreme when it comes to scientific observation, anyway? We’re getting used to the idea of sound being used as a form of scientific measurement—think everything from the blip of gravitational waves to the tumor-busting wavelengths of focused ultrasounds—so smell could be next. 

All these examples point towards the burgeoning benefits of smell as an underappreciated natural diagnostic device. And that’s nothing to sniff at.

Go Deeper