Dogs’ Incredible Noses: Scientists Uncover Canine Abilities to Detect Pregnancy and Human Stress

Groundbreaking research led by canine scientist Dr. Catherine Reeve is shedding new light on the extraordinary olfactory and perceptive abilities of dogs, revealing their capacity to detect human pregnancy and discern stress through scent. The findings, discussed in detail on "The Pawsitive Post in Conversation" podcast with Zazie Todd and Kristi Benson, highlight how dogs may be far more attuned to subtle physiological and emotional changes in their human companions than previously understood. This work has significant implications for understanding the human-animal bond, preparing dogs for major life changes, and advancing the training of service animals.

The Unseen World of Canine Olfaction

Dogs possess an unparalleled sense of smell, a biological marvel that allows them to navigate and interpret their world in ways humans can scarcely imagine. With an estimated 220 million to 300 million olfactory receptors, compared to humans’ mere 5 million, and a brain area dedicated to smell that is proportionally 40 times larger, dogs’ noses are sophisticated biological instruments. Their ability to "smell" time, fear, illness, and even impending natural disasters has long been a subject of scientific and anecdotal interest. Dr. Catherine Reeve, a postdoctoral fellow at Mount Saint Vincent University with a PhD in experimental psychology from Dalhousie University, specializes in applied canine behavior, cognition, and particularly, olfaction and human-animal interaction. Her research delves into these profound capabilities, seeking to provide empirical evidence for phenomena often observed anecdotally by dog owners.

Investigating Canine Perception of Pregnancy

The inspiration for Dr. Reeve’s study into dogs’ detection of pregnancy stemmed from a common observation: many women reported changes in their dog’s behavior when they became pregnant. Despite the widespread nature of these accounts, a comprehensive scientific investigation into the phenomenon was notably absent from the literature. Driven by an interest in women’s health and canine olfaction, Dr. Reeve, alongside her students, embarked on a retrospective online survey to document and quantify these reported behavioral shifts.

Methodology and Key Findings of the Pregnancy Study

The study utilized an online survey, recruiting participants who had owned a dog during a past pregnancy. This retrospective approach, while offering a broad data collection, acknowledged potential limitations related to recall bias. The survey meticulously gathered participant and dog demographics, details about the pregnancy, and then presented respondents with five distinct categories of dog behaviors:

1. Attention or connection-seeking behaviors.
2. Guarding or stress/anxiety behaviors around unfamiliar people.
3. Stress or anxiety/fear-based behaviors towards the owner.
4. Stress or anxiety/fear-based behaviors towards other dogs.
5. General stress or anxiety behaviors.

Participants were asked to select behaviors their dog displayed before pregnancy and then to indicate if their dog’s behavior changed during pregnancy. If a change was reported, they then specified which behaviors were exhibited during pregnancy, allowing for a comparative analysis. A particularly insightful question probed whether changes were observed before the owner was even aware of their pregnancy.

The results were compelling. Approximately 65% of the 130 participants reported that their dog’s behavior changed during pregnancy, with an increase observed across all five behavior categories. Strikingly, about 27% of respondents indicated that their dog’s behavior shifted before they themselves knew they were pregnant. This suggests a subtle, non-visual detection mechanism at play, likely related to hormonal and physiological changes in the pregnant individual.

Further analysis identified specific predictors of behavioral change. Dogs that exhibited guarding or fear/anxiety behaviors around unfamiliar people before pregnancy were more likely to show increased behavioral changes during their owner’s pregnancy. This suggests a heightened sensitivity in already cautious dogs to shifts in their environment and human companions. Conversely, dogs with higher baseline levels of fear and anxiety towards other dogs were less likely to report changes in behavior during pregnancy, implying that their existing anxieties remained dominant and perhaps masked new reactions to the owner’s pregnancy. The study also explored the dog-owner relationship using the Monash Dog Owner Relationship Scale (MDORS), but found no direct correlation between the strength of this bond and the likelihood of reported behavioral changes during pregnancy.

Implications for Expectant Parents

While the study’s findings are based on owner perception and cannot serve as a reliable pregnancy test, they provide strong evidence that dogs are acutely aware of the profound physiological, hormonal, and environmental shifts that accompany pregnancy. Dr. Reeve emphasized the importance of these findings for expectant parents: "People might be assuming the dog isn’t aware, but actually the dog quite likely is aware that something is going on." This underscores the need for early preparation and proactive training to help dogs adapt to the impending arrival of a baby, ensuring a smoother transition for both canine and human family members. The changes in a household during pregnancy – from alterations in routine and scent profiles to shifts in the owner’s body language and emotional state – are significant, and dogs, as integral family members, experience these changes deeply.

Canine Detection of Human Stress Through Scent

Beyond pregnancy, Dr. Reeve’s research also delved into another profound aspect of canine perception: the ability to detect human stress through scent alone. This area of inquiry holds particular relevance for understanding the capabilities of service dogs, especially those trained for conditions like PTSD, where they often alert to invisible physiological shifts in their handlers. The challenge was to isolate scent as the primary cue, eliminating any visual or auditory behavioral signals from the human.

Methodology of the Stress Detection Study

To achieve this, Dr. Reeve’s team employed a rigorous, controlled protocol for collecting human breath and sweat samples. Participants were recruited, and their physiological data (blood pressure, heart rate, skin conductance) was monitored using a BioPac system. Self-reported stress levels were also collected using a visual analog scale.

The sample collection involved two phases:

1. Baseline Samples: Participants sat silently for three minutes. They then, wearing gloves, wiped sterile gauze on the back of their neck and exhaled deeply three times into a glass vial. Three baseline samples were collected.
2. Stress Samples: Participants were then subjected to a well-validated psychological stressor: mental arithmetic performed out loud. They were asked to count backward by 17 from a starting number like 9000, with researchers sternly correcting any errors. This task, designed to induce significant stress, was performed for three minutes. Immediately afterward, stress samples were collected using the same method as the baseline. Control blank samples were also prepared.

Crucially, participants did not eat or drink between baseline and stress sample collection, ensuring that any changes in scent were primarily attributable to the acute stress response. The study was conducted in Belfast, Northern Ireland, and faced the unique challenge of the COVID-19 pandemic, necessitating adaptive protocols for remote sample collection with sterile kits.

Training and Testing the Detection Dogs

The next phase involved training volunteer dogs to discriminate between these samples. The training protocol was a testament to the principles of positive reinforcement and careful shaping, a detail lauded by dog trainers like Kristi Benson. The goal was to teach dogs to perform a clear targeting behavior (a stand/stare or a sit) in front of the correct sample on a three-arm apparatus.

The training progressed in stages:

1. Initial Targeting: Dogs were introduced to the apparatus with a target sample containing food, alongside two blank samples. Rewards were given for showing interest and then for holding a specific alert behavior.
2. Food Removal: Gradually, the food was removed, leaving only the scent paired with food residue, and eventually, just the human breath and sweat sample.
3. Discrimination Training: Dogs learned to discriminate between samples from different people, and then between samples from the same person collected at different times of the day.
4. Formal Testing: The final testing phase was double-blind, meaning neither the dog’s handler nor the experimenter knew the location of the target sample. Dogs were first presented with a stress sample and two blanks. Once they consistently achieved 8 out of 10 correct trials, they progressed. In the critical phase, they were presented with a new stress sample from a person, alongside that same person’s baseline sample (collected just minutes before the stressor) and a blank.

Remarkable Accuracy in Stress Detection

The results were astounding. Four dogs, a diverse group including a Lurcher, a Terrier mix, and a Cocker Spaniel mix, demonstrated exceptional ability. They achieved an average accuracy of 93-94% in discriminating between a person’s stress sample and their baseline sample. The strict controls, including the short interval between sample collections and the cleaning of the apparatus between trials, provided robust evidence that the dogs were indeed detecting unique odor signatures associated with acute psychological stress. Dr. Reeve expressed profound impressiveness at the dogs’ performance: "They were absolutely phenomenal… I really didn’t expect it to go that well." The dogs clearly enjoyed the challenge, exhibiting wagging tails and keen engagement during their "work."

Broader Implications and Future Directions

These two studies collectively underscore the profound sensitivity of dogs to the internal states of humans. The ability to detect physiological changes associated with pregnancy, and to discern the subtle olfactory markers of stress, highlights a dimension of human-animal interaction that extends far beyond overt behavioral cues.

For service dog training, particularly for individuals with PTSD, these findings provide crucial empirical support. It suggests that these dogs are not merely reacting to visible signs of distress but are actively perceiving biochemical shifts in their handlers, enabling them to provide alerts even before a crisis becomes apparent. This understanding can refine training protocols and enhance the efficacy of these life-changing partnerships.

More broadly, this research enriches our understanding of canine cognition and olfaction. It reinforces the idea that dogs are not just pets but sentient beings deeply connected to our physiological and emotional landscapes. Dr. Reeve’s work emphasizes the need for continued scientific exploration into these areas, leveraging dogs’ natural abilities for various applications, including medical detection, search and rescue, and enhancing the human-animal bond.

The studies also subtly touch upon the ethical considerations inherent in such research, ensuring both human participants (e.g., managing stress during the arithmetic task) and canine participants (e.g., ensuring enjoyment and motivation during training) are treated with utmost care. The dedication of Dr. Reeve’s graduate students, Clara Wilson and Carrie Campbell, in adapting to pandemic challenges and maintaining high scientific rigor, was also a highlight of the discussion.

As Dr. Reeve noted, "they’re picking up on a lot of stuff." This research serves as a powerful reminder of the remarkable, often hidden, capacities of our canine companions, urging us to recognize and appreciate the depth of their perception and their invaluable role in our lives.

For those interested in delving deeper, Dr. Catherine Reeve’s work can be explored via the Mount Saint Vincent University website or her dog training venture, epictails.ca. The papers discussed are linked in the show notes of "The Pawsitive Post in Conversation."

This page contains affiliate links which means I may earn a commission on qualifying purchases at no cost to you.