Born to yawn? Understanding yawning as a warning of the rise in cortisol levels: randomized trial

Yawning consistently poses a conundrum to the medical profession and neuroscientists. Despite neurological evidence such as parakinesia brachialis oscitans in stroke patients and thermo-irregulation in multiple sclerosis patients, there is considerable debate over the reasons for yawning with the mechanisms and hormonal pathways still not fully understood. Cortisol is implicated during yawning and may link many neurological disorders. Evidence was found in support of the Thompson cortisol hypothesis that proposes cortisol levels are elevated during yawning just as they tend to rise during stress and fatigue.

To investigate whether saliva cortisol levels rise during yawning and, therefore, support the Thompson cortisol hypothesis.

We exposed 20 male and female volunteers aged between 18 and 53 years to conditions that provoked a yawning response in a randomized controlled trial. Saliva samples were collected at the start and again after the yawning response, or at the end of the stimuli presentations if the participant did not yawn. In addition, we collected electromyographic data of the jaw muscles to determine rest and yawning phases of neural activity. Yawning susceptibility scale, Hospital Anxiety and Depression Scale, General Health Questionnaire, and demographic and health details were also collected from each participant. A comprehensive data set allowed comparison between yawners and nonyawners, as well as between rest and yawning phases. Collecting electromyographic data from the yawning phase is novel, and we hope this will provide new information about neuromuscular activity related to cortisol levels. Exclusion criteria included chronic fatigue, diabetes, fibromyalgia, heart conditions, high blood pressure, hormone replacement therapy, multiple sclerosis, and stroke. We compared data between and within participants.

In the yawning group, there was a significant difference between saliva cortisol samples (t10 = -3.071, P = .01). Power and effect size were computed based on repeated-measures t tests for both the yawning and nonyawning groups. There was a medium effect size for the nonyawners group (r = .467) but low power (36%). Results were similar for the yawners group: medium effect size (r = .440) and low power (33%).

There was significant evidence in support of the Thompson cortisol hypothesis that suggests cortisol levels are elevated during yawning. A further longitudinal study is planned to test neurological patients. We intend to devise a diagnostic tool based on changes in cortisol levels that may assist in the early diagnosis of neurological disorders based on the data collected.

International Standard Randomized Controlled Trial Number (ISRCTN): 61942768; http://www.controlled-trials.com/ISRCTN61942768/61942768 (Archived by WebCite at http://www.webcitation.org/6A75ZNYvr).