Studies of the association between blood BDNF levels and delirium are very few and have yielded mixed results.

To investigate the blood BDNF levels in the occurrence and recovery of delirium.

Prospective, longitudinal study. Participants were assessed twice weekly with MoCA, DRS-R98, APACHE-II. BDNF levels of the same were estimated with ELISA method. Delirium has been define as per DRS-98R (cut-off>16) and recovery of delirium as at least two consequently assessments without delirium prior to discharge.

No differences in the levels of BDNF between those with delirium and those who never developed it. Excluding those who never developed delirium (n=140), we analysed the effects of BDNF and the other variables on delirium resolution and recovery. Of the 58 remained with delirium in the subsequently observations (max=8) some of them continue to be delirious until discharge or death (n=39) while others recovered (n=19). BDNF levels and MoCA scores were significantly associated with both delirium cases who became non-delirious (resolution) during the assessments and with overall recovery. BDNF (Wald χ²=11.652, df: 1 P=.001), for resolution. For recovery Wald χ²=7.155; df: 1, P=.007. No significant association was found for the other variables (APACHE-II, history of dementia, age or gender)

BDNF do not have a direct effect in the occurrence of delirium but for those delirious of whom the levels are increased during the hospitalisation they are more likely to recover from delirium.