Plasma brain-derived neurotrophic factor in earthquake survivors with full and partial post-traumatic stress disorder


STRESS HAS BEEN widely linked with brain-derived neurotrophic factor (BDNF) modifications but results from the few studies on post-traumatic stress disorder (PTSD) are inconsistent: lower BDNF plasma levels, negative results in cerebrospinal fluid and higher levels in serum have been reported.[1-3]

We investigated BDNF in survivors of the L'Aquila earthquake, along with the post-traumatic spectrum, considering not only full expression but also subthreshold manifestations, that is, partial PTSD.

Thirty-seven medically healthy consecutive outpatients (M/F 13/24; mean age, 45.7 ± 11.7 years) referred for anxiety/affective symptoms manifesting after the earthquake, were recruited 2 years after the event. All were treated with low doses of benzodiazepines and/or antidepressants.

Twenty-two healthy subjects recruited from among those accompanying the outpatients, and who were matched for age and gender (M/F 8/14; mean age, 42.4 ± 11.5 years) were enrolled.

Blood samples were taken between 08.00 and 09.00 AM and processed using ELISA protocol on plasma.

By means of the Structured Clinical Interview officially available in Italy (SCID-I/P), 13 patients were diagnosed as having full PTSD, and 13 as having partial PTSD, that is, fulfilled a subset of the DSM-IV criteria (mean age, 49.9 ± 10.2 years; M/F 3/10; mean age, 44.4 ± 13.6 years, M/F 4/9); 11 patients (mean age, 42.3 ± 10.3 years; M/F 6/5) were diagnosed as not having PTSD but instead were diagnosed as having depressive episode (n = 8) or panic attack (n = 3). No differences in gender distribution or age were observed.

BDNF level was as follows: full PTSD subjects 3473.3 ± 2073.2 pg/mL; partial PTSD, 6180.9 ± 2348.3 pg/mL; clinical sample without PTSD, 3650.6 ± 1648.2 pg/mL; controls, 6172.9 ± 3899.3 pg/mL. Kruskal–Wallis test indicated significant difference (χ2 = 17.20, d.f. = 3, P < 0.001); post-hoc planned Mann–Whitney test indicated lower BDNF for subjects with full-blown or no PTSD than subjects with partial PTSD (Z = 3.51, P < 0.0005 and Z = 3.10, P < 0.005) and controls (Z = 2.56, P < 0.01 and Z = 2.29, P < 0.05) and no difference between partial PTSD and controls.

Reduced BDNF in full PTSD is in agreement with Dell'Osso et al.[1] The BDNF finding in patients with partial PTSD is in line with Bonne et al., who found no difference between patients with moderate PTSD severity and controls.[2] Hauck et al. reported that BDNF was elevated in subjects at early stages of trauma showing less severe symptoms.[3] Together with the current finding that patients with partial PTSD had higher BDNF levels, we can hypothesize that the higher plasma level of BDNF would somehow have a neuroprotective role against the expression of full PTSD.

The sample size was relatively small, but all patients were triggered by a unique trauma, had a similar low drug dosage and none presented comorbid psychiatric diagnoses.