You’ve probably noticed that your dog likes sniffing; it’s well known that dogs use their keen sense of smell to understand their world. We humans have been utilising this power for centuries for hunting, search and rescue, drug detection and even finding expensive truffles! You might have also read articles about dogs being used to smell for illness in people. Is this true? What diseases can dogs sniff out? And might we be seeing dogs in hospitals anytime soon?

Table of contents

How Does a Sense of Smell Work?

In dogs (as well as most mammals and many other creatures), the sense of smell, or olfaction, involves multiple parts of the body. Physically, a smell is made up of tiny particles that can be detected by the body and interpreted by the brain. 

Primary Olfaction

How this works is as follows: minute particles (usually molecular) of ‘smell’ interact with olfactory epithelium, a special tissue made up of cells called olfactory sensory neurones (a type of nerve cell). The olfactory epithelium is located within the nose. The act of sniffing draws molecule-laden air into the nose where the molecules are absorbed by the epithelium and then bind to receptors on the olfactory sensory neurones. This activates the neurones and sends an electrical signal directly to the brain via the olfactory nerve.

The signal arrives at the olfactory bulb of the brain. Some of the signals reaching the olfactory bulb are then sent to other parts of the brain as well; including areas involved in memory and emotions. This may be why certain smells can cause certain feelings in some people. In either case, the brain interprets the huge mix of signals from all the different molecules around as a specific smell, which we perceive. It is still unclear exactly how this works.

Secondary Olfaction

Dogs, as well as many other mammals and reptiles, also have a second olfactory system. This involves an area between the nose and the roof of the mouth called the vomeronasal organ. This organ also detects molecules of ‘smell’ and transmits the signal to the olfactory bulb of the brain. But the signals then progress to areas of the brain that may be involved in aggression and mating behaviour. Some animals (though not dogs) will part their lips in a particular way to help bring smells to the vomeronasal organ.

Interestingly, it is known that most dogs are ‘right-nostriled’, preferring to sniff first with the right nostril only. And only if the smell is familiar and not aversive will then use the left. It is believed this is because the right-side of the brain is responsible for processing new information, while the left results in emotional and behavioural responses. 

How Does a Dog’s Sense of Smell Compare to a Human’s?

It’s well known that a dog’s sense of smell far exceeds that of a human; with some claiming it may be 10 million times as sensitive. Let’s find out why dogs smell so well. 

Anatomy

To start, dogs have a much higher density and number of olfactory neurones than humans. This means they can detect molecules of smell at a much lower concentration than humans can. (Meaning they can detect smells so weak that we cannot). The olfactory mucosa itself is larger than in humans, as is the olfactory bulb and associated structures. 

Dogs also have a nasal anatomy where in each breath, just over 10% of the air flows to the olfactory mucosa, while the rest continues further to the lungs for respiration. The air leaves the lungs and the nose via the same pathway, meaning that 10% of air intended for olfaction remains in the nose for the whole in-out cycle of breathing, increasing the time that molecules can be detected and perceived as smell. 

Of course, the addition of the secondary vomeronasal organ in dogs greatly enhances their sense of smell as well. While it is believed by some that humans do possess a functioning vomeronasal organ, and studies have found similar tissue in the region it would be located, it is likely vestigial, as no studies have proven there are active sensory neurones in the area.

Genetics

Olfaction requires certain genes to be active for olfactory receptors to work. In humans, around 50% of these genes are ‘pseudogenes’ which do not code for receptors; while in dogs only 20% are pseudogenes. This means they are able to produce more variety of receptors to detect smell than humans. Interestingly, many of these genes vary dog-breed to breed. Certain breeds, like bloodhounds, have been selectively bred for their sense of smell, and thus are much better at smelling than the average dog, and this may be one reason why.

Breeding and Other Factors

However, dogs are much more varied compared to humans, meaning an individual dog’s sense of smell may vary. For example, dogs with longer noses have more olfactory mucosa, and thus smell better than dogs with short noses. This is especially true for brachycephalic dogs, who have very short noses. Larger dogs smell better than smaller dogs, likely for the same reason. Just like many humans, as a dog ages they are less able to detect smells as their olfactory epithelium degrades. They may, however, be better at associating memories with olfaction. As well as activated genes, some breeds may be better at olfaction due to their ability to learn, be trained and work with humans. One study has also found that female dogs have more active olfactory bulbs than male dogs; possibly indicating the females are better at smelling. 

Can Dogs Smell Out Illness in People?

There have actually been a large number of studies that have assessed a dog’s ability to smell a wide variety of diseases. It is known that various compounds are created during different metabolic states, including illness. These compounds are excreted in exhaled air, urine and faeces, and can even be found in the blood. The concentrations of these compounds are within the ability for a dog to detect. So yes, dogs can definitely smell illness in people! There’s too many to cover here, but let’s look into some of the most interesting studies.

Cancer

Perhaps the most uplifting findings have been dogs’ abilities to smell cancer. In various excretions and tissues, dogs have been able to distinguish healthy people from those with lung, breast, prostate, ovary, bladder, skin and colon cancer. One study managed to train ‘ordinary’ dogs to detect lung and breast cancer via exhaled breath with almost perfect results. Another study had similar results, with their three dogs able to correctly identify lung cancer in blood over 96% of the time. It was even shown in a different study that dogs could detect cancer cells placed onto the skin of healthy patients. 

Infectious Diseases

The ability to detect disease also extends to infectious diseases. The parasitic infection malaria, spread by mosquitoes, is one of the deadliest diseases in the world. Though it can easily be identified in blood, some malaria parasites are now no longer producing the protein markers detected in blood samples. New tests involve analysing a victim’s breath. It is believed that malaria may cause humans to produce chemicals that attract mosquitoes, introducing more malaria into the body to accelerate the disease. A small study utilising worn socks from children infected with malaria demonstrated that two dogs were able to distinguish between infected and non-infected children with a specificity of over 90%, though the sensitivity was lower (meaning the dogs had some false-positive reactions). Other studies have shown similar results for viral respiratory infections, including influenza and COVID-19.

Other Diseases

The possibilities also extend to neurological and metabolic diseases. Some dogs have been found to predict seizuring, narcolepsy and hypoglycaemia, even with little training in the latter case. This has both indicated that there may be specific markers undetectable to humans in these patients, and that dogs may be able to be utilised to help protect patients with these conditions. Some do argue that dogs are detecting subtle social clues, either in addition to or as well as smell, and that other factors may be involved in predicting these conditions. There was even one study that found dogs were able to identify individual aggression in a group of psychiatric patients; though it was again unclear how much of this ability was based on smell.

Veterinary Use

Slightly outside the scope of this article, there have also been studies into using dog noses for veterinary purposes, with dogs being trained to smell mite-infected animals in the Alps, worm-infected poo, the stage of heat in dairy cows, and, of course, heat in other dogs for breeding purposes (which, unsurprisingly, they’re really good at!). 

Practicality of Dogs for Medical Purposes

By now, it should be pretty clear that dogs have an amazing sense of smell, and this ability may be able to be used to help human medicine. But how practical is this? 

Controversies

Despite there being concrete evidence that humans produce specific chemicals during times of illness, and studies like those listed above do seem to indicate dogs have a good ability to detect these chemicals, some are still sceptical of the results. Firstly, the studies are generally small, with only a handful of dogs tested at a time. This doesn’t discredit the results, but larger studies are needed to prove the results have reliability. 

Some also claim that because dogs cannot tell us what they are smelling, but only react in a certain way, care must be taken not to over-interpret a response. Did the dog react because of a person’s cancer, or regardless of the disease?  Some argue any positive response must be validated by a second test. There is also the argument, we discussed above, that smell is only one component of a dog reacting to a disease, and other subtle changes also elicit a response – disease should not be diagnosed based on smell alone. 

Efficacy

There have been more uses of dogs outside of laboratory settings for detecting seizuring and hypoglycaemia, among other diseases, meaning anecdotal evidence seems to be more common. This has even led to companies and charities being set up to specifically provide dogs for people in need, similar to guide dogs for the blind. However, these companies emphasise that these dogs are highly trained and should not be relied on to predict these events; instead being better used as an aid to detection and for companionship to sufferers. 

It has been noted that people have attempted to train dogs themselves to detect their diseases, or even stories of dogs being sold at a premium as ‘detection dogs’, despite being untrained puppies. Not only is this false, it is dangerous, as disease may be missed by an untrained or incompletely trained dog. The point is that even if dogs can detect seizuring, hypoglycaemic episodes and more, after training, they are not infallible, responses can be missed or over-interpreted, and using them as the sole test could lead to harm to humans.

Practicality

The biggest hurdle, assuming dogs may have a genuine ability to detect certain diseases, is the practicality of their use. Unlike a microscope, dogs require food, shelter, water, training and care. Where does this all come from? Do health services pay for it? Do the dogs have owners, or are they owned by the medical profession? In the former case, does this create financial incentive to train dogs for medical purposes, and open a person up to litigation should their dog misdiagnose? In the latter, can a dog’s welfare be maintained if they are kept as a tool in a lab, not as a pet? And in all cases, would the costs of training and caring for a dog not be far better spent on cheaper tests, especially for cash-strapped national health services? 

And what about poorer countries who cannot afford these costs? Dogs are also more difficult to send abroad or into remote areas than testing kits, meaning fighting disease can be harder. Overall, though individual dogs may find specific roles within laboratory settings, we cannot realistically see dogs commonly working with medical staff in hospitals. 

What’s the Future of Dogs Smelling Illness?

Clearly this area of research is interesting, but still in its infancy. And although dogs are starting to be used in real-world settings to help patients with various diseases, most of the widespread applications are limited or impractical. So does this mean the field is nothing more than an interesting curiosity? Perhaps not.

Inspired by studies like those listed above, researchers at the Technion-Israel Institute of Technology have been able to create a device called the Nano-Nose, or Na-Nose, which uses nanoparticles of gold to detect chemicals undetectable by the human nose. The device has been able to detect chemicals in lung cancer patients’ breath with up to 95% accuracy; thus being able to diagnose the disease. In a sense, they have created an artificial nose far superior to a human nose, perhaps on par or exceeding that of a dog’s. It goes without saying a small device is far more practical than a dog, so may be more suited for medical use.

And we wonder if this is where the future of canine disease detection lies – not in actually using dogs to sniff out disease, but in helping identify specific chemicals that artificial technology can then diagnose in the real world. 

Sources and Further Reading: