Alarm pheromones are airborne chemical signals released by an individual into the environment, which provide unconscious olfactory cues with respect to the donor’s state of fear, to members of the same species. While alarm pheromones’ existence and neurobiological effects are well established for many non-human mammals, only very recently has there been direct evidence for a human alarm pheromone, findings that suggest there may be a hidden biological component to human social dynamics, in which emotional stress is, quite literally, 'contagious.'  Since stress is known to deleteriously affect almost every facet of health, it is important to understand whether chronic stress may have an unaccounted-for social component, particularly in closed-space environments such as space missions, submarines, cockpits, emergency rooms, and the offices of individuals writing scientific grant proposals under deadline. 


Our research on human alarm pheromones was explicitly designed to replicate the rodent research on alarm pheromones as closely and as rigorously as possible while still remaining non-invasive. Since the animal literature relies upon predator models that are not appropriate for clinical research, we were forced to be creative in designing a paradigm that would reliably and ethically induce real fear. Our solution was to obtain “fear” sweat samples from individuals participating in their first-time tandem skydives (4km/13,000 ft. with 60s free-fall), with time-matched exercise sweat as a control.


Our initial research, conducted during  2005-2006, used sweat samples obtained during an acute emotional stressor and a time-locked exercise condition to identify candidate compounds for the pheromone using gas chromatography-mass spectroscopy. Our results also allowed us to iteratively refine our sweat collection procedures to ensure minimal contamination via bacteria. Without bacteria, our sweat samples contain almost no consciously-detectable odor, which is critical since it means that the stimuli are not consciously aversive during experiments that test the sweat’s effects on the brain. In fact, subjects were unable to consciously distinguish the two conditions (stress sweat vs. exercise sweat) from one another, and rated both as having low intensity and neutral valence on Likert scales designed to assess this question.


In 2006, our laboratory provided the first neurobiological evidence for the existence of a human alarm pheromone, providing a rigorous basis for the anecdotal claim that emotional stress can be detected between unrelated individuals by unconscious olfactory cues. After extracting sweat for each donor subject under both stress and control conditions, we then presented the extracts to a separate group of unrelated subjects during their fMRI scans, to determine if the fear, but not exercise, sweat activated areas of the brain associated with the excitatory response to emotional arousal (amygdala, hypothalamus), which they did. Our follow-up fMRI research in 2007 both replicated the original effect using different sweat samples, different fMRI subjects, and a different scanner, as well as extended the original study, to investigate both the four-way interactions between donor sex-detector sex (male skydivers, male fMRI subjects, male skydivers, female fMRI subjects, female skydivers, male fMRI subjects, female skydivers, female fMRI subjects), as well as comparison between the neural pathways activated by unconsciously aversive odors (alarm pheromones) as compared to the neural pathways activated by consciously aversive odors (rotten eggs). Our analyses showed no gender interactions, and that the pathways were, in fact, distinct.  In 2008 we showed, using a sensitive psychometric behavioral experiment, that stress sweat increases accuracy in the discrimination of threat from non-threat.


The broad, long-term objectives of the research are to determine whether human alarm pheromones play a potentially significant role in the group dynamics of fear, anxiety, and stress. If our initial research was designed to ask the fundamental question: “do chemosensory cues to others’ stress exist?” our current research in now builds directly upon this work, to systematically ask more detailed questions about how they affect the brain and what their potential impact on humans might be.  To complement this work, our laboratory conducts research on human reproductive pheromones, oxytocin/vasopressin, social hierarchies, and the biological basis for evolutionary kin selection, to further explore the degree to which our cognitive processing is unconsciously affected by chemosensory signals.