Effects of acute psychological stress on blood cell-free mitochondrial DNA (cf-mtDNA): A crossover experimental study

In response to acute stress, prior studies have found an increase in circulating cell-free mitochondrial DNA (cf-mtDNA) and pro-inflammatory cytokines, highlighting two potential inter-related mechanisms by which stressors can get under the skin. However, prior studies lacked a resting control condition to isolate the effect of psychological stress from other aspects related to laboratory procedures. Here, we conducted a crossover experimental trial examining responses to a socio-evaluative stressor under laboratory conditions. 72 volunteers (age 20-50, 48% women) were tested on two occasions, counterbalanced, separated by at least a month. On one occasion, they were exposed to a 5-min socio-evaluative stressor (speech task), and on the other occasion, rested for the same period. Blood samples were obtained at 10 timepoints from pre- to 2 hours post-exposure to assess neuroendocrine (cortisol, catecholamines), pro-inflammatory cytokine (IL-6, IL-10, TNF-[a]), and both plasma and serum cf-mtDNA responses. Compared to the control visit, the stressor significantly increased anxiety, heart rate, blood pressure, cortisol, and norepinephrine (ps<0.05-0.0001), confirming the psychobiological impact of the stressor. Unexpectedly, IL-6 and plasma cf-mtDNA increased (time effect p<0.0001) in both the stress and control conditions. While no significant effect of time was found for serum cf-mtDNA, plasma cf-mtDNA showed a bi-phasic response with an initial 22-24% increase at 5-10 min (g=0.07, stress-control visits), followed by a decrease and another 70-81% increase from 45 to 75 min (g=0.59 (stress visit), g=0.41 (control visit)). There were no significant associations between the pro-inflammatory and cf-mtDNA responses, pointing to their independent regulation. While mood, cardiovascular, and neuroendocrine reactivity were selectively induced by socio-evaluative stress, IL-6 and blood cf-mtDNA increased across both the stress and control conditions, suggesting that these biomarkers may reflect non-specific responses to the laboratory protocols (e.g., blood draw) rather than to socio-evaluative stress itself.