DataSheet_1_Dexamethasone Suppression Test May Predict More Severe/Violent Suicidal Behavior.xlsx (21.41 kB)
Download file

DataSheet_1_Dexamethasone Suppression Test May Predict More Severe/Violent Suicidal Behavior.xlsx

Download (21.41 kB)
posted on 04.03.2020, 10:14 by Adrián Alacreu-Crespo, Emilie Olié, Sebastien Guillaume, Chloé Girod, Aurélie Cazals, Isabelle Chaudieu, Philippe Courtet

Several studies demonstrated that the hypothalamic-pituitary-adrenal (HPA) axis is dysregulated in suicide attempters. Prospective studies found that people with an abnormal response at the dexamethasone suppression test (DST) are more likely to commit suicide. However, whether DST may predict suicide attempts remains less clear. A possible strategy to address this question is to consider the suicide attempt lethality.


(1) To compare the pre- and post-DST cortisol levels in serious/violent suicide attempters and in non-serious/non-violent suicide attempters, and (2) to investigate whether cortisol level can predict new suicide attempts or their lethality.


The study included 70 recent suicide attempters (25 with a serious or violent attempt) who were followed for two years. Three saliva samples for cortisol measurement were obtained at 8a.m., 3p.m., and 9p.m. before the DST (pre-DST). Then, at 11 p.m., 1 mg of dexamethasone was given orally. The following day (post-DST), three saliva samples were collected at the same hours as before. The post-DST–pre-DST salivary cortisol Δ index was calculated for each collection time. The Risk-Rescue Ratio Scale (RRRS) and the Suicidal Intent Scale (SIS) were used to characterize the suicide attempt at inclusion and those occurring during the follow-up.


Post-DST cortisol level at 9 p.m. was higher in patients with an initial violent or serious suicide attempt than in non-violent/non-serious attempters (p < .010). Higher post-DST cortisol at 9p.m. was associated with lower RRRS rescue score and higher clinical impression of suicide severity at inclusion. Among the 66 patients who completed the follow-up, 26 attempted suicide again at least once. Higher pre-DST cortisol at 8a.m. predicted new suicide attempts during the follow-up (OR = 2.15 [1.11, 4.15]), and higher cortisol Δ index at 9p.m. was associated with higher suicide intent during the follow-up.


Our results suggest that HPA axis hyper-reactivity monitored with the DST is a marker of violent/serious suicide attempt with lower rescue possibility. Furthermore, higher changes between pre-DST and post-DST cortisol levels may predict higher suicide intent. These findings might help to characterize the biological features of nearest suicide phenotypes.