A new study suggested that dogs could detect and respond to human stress through smell alone, potentially influencing their cognitive processes and behavior.

In the study, published in Scientific Reports, researchers from the University of Bristol and Cardiff University conducted a two-phase experiment to examine how human odor samples influenced dogs' performance on a spatial judgment task. In the first phase, they collected sweat and breath samples from human volunteers during a stressful activity (the Trier Social Stress Test) and a relaxing activity (watching nature videos). They confirmed stress levels through measures of salivary cortisol, heart rate, heart rate variability, and self-reported anxiety.

In the second phase, 18 dogs completed a cognitive bias test under three odor conditions: no odor, stress odor, and relaxed odor. The researchers trained the dogs to approach a bowl in a "positive" location containing food and avoid a bowl in a "negative" location with no food. They then presented bowls in ambiguous locations and measured the dogs' approach behavior.

The researchers found that exposure to the odor of stressed humans caused the dogs to exhibit more pessimistic responses in a cognitive bias test compared with exposure to relaxed human odors or no odor.

Among the key findings were:

• When exposed to human stress odor in the third test session, dogs were significantly less likely to approach a bowl placed in the "near-negative" ambiguous location compared to the no-odor condition (hazard ratio [HR] = 0.42, P = .035).
• Dogs showed improved learning of positive and negative locations with repeated testing, with the effect strongest when stress odor was present in session three.
• In the presence of stress odor, dogs were more likely to approach the positive location (HR = 3.66, P < .001) and less likely to approach the negative location (HR = 0.33, P < .001) compared with no odor.
• The order of odor presentation influenced results, with effects strongest when stress odor was introduced after relaxed odor.

The researchers recruited 26 dog-owner pairs, with 18 dogs completing all three test sessions. Dogs ranged from 8 months to 10 years old and included various breeds. Human odor samples were collected from three volunteers showing consistent stress responses across physiological and self-report measures.

In each test session, dogs completed a training phase to learn positive and negative bowl locations, followed by a testing phase with ambiguous bowl locations. The researchers measured latency to approach bowls and used Cox proportional hazards models to analyze the likelihood of approach behavior.

They controlled for potential confounding factors such as individual dog characteristics, handler relaxation, and sample donor. A reward control test confirmed dogs were not using olfactory cues to determine food presence.

The study had limitations, including a relatively small sample size and the lack of a control group exposed only to blank odor samples across all sessions. The researchers noted that both odor effects and learning effects from repeated testing likely influenced results.

These findings suggested that human emotional states may be communicated to dogs through olfactory cues alone, potentially affecting their cognitive processes and risk assessment. This had implications for dog welfare and working performance, particularly in environments where human stress levels may be high.

The researchers concluded that further studies may be needed to fully understand the mechanisms by which human odors influence canine cognition and behavior. They emphasized the importance of considering olfactory aspects of dogs' environments in training and working contexts.

Full ethics declarations can be found within the study