Noninvasive modulation of predictive coding in humans: causal evidence for frequency-specific temporal dynamics.

Increasing evidence indicates that the brain predicts sensory input based on past experiences, importantly constraining how we experience the world. Despite a growing interest on this framework, known as predictive coding, most of such approaches to multiple psychological domains continue to be theoretical or primarily provide correlational evidence. We here explored the neural basis of predictive processing using noninvasive brain stimulation and provide causal evidence of frequency-specific modulations in humans. Participants received 20 Hz (associated with top-down/predictions), 50 Hz (associated with bottom-up/prediction errors), or sham transcranial alternating current stimulation on the left dorsolateral prefrontal cortex while performing a social perception task in which facial expression predictions were induced and subsequently confirmed or violated. Left prefrontal 20 Hz stimulation reinforced stereotypical predictions. In contrast, 50 Hz and sham stimulation failed to yield any significant behavioral effects. Moreover, the frequency-specific effect observed was further supported by electroencephalography data, which showed a boost of brain activity at the stimulated frequency band. These observations provide causal evidence for how predictive processing may be enabled in the human brain, setting up a needed framework to understand how it may be disrupted across brain-related conditions and potentially restored through noninvasive methods.

Glutamatergic and GABA-ergic abnormalities in First-episode depression. A 1-year follow-up 1H-MR spectroscopic study.

BACKGROUND: Previous magnetic resonance spectroscopic (MRS) studies have reported brain metabolic abnormalities in Major Depressive Disorder (MDD). Nevertheless, results have been inconsistent, focusing on fully developed major depression neglecting first episode patients (FED). Longitudinal studies have also been rare and with short follow-up periods. The aim of the current study was to investigate the differences between healthy controls and first episode patients at baseline, together with changes of metabolites after 1 year follow-up in the ventromedial prefrontal cortex. METHODS: 1H-MRS images were obtained from 64 healthy controls and 31 FED patients using a 3T Philips Achieva scanner and processed with TARQUIN software at baseline and after 1 year. Examined metabolites included Glx (corresponding to Glu+Gln-peak), Glu, NAAG, myo-Ins, Cr, GSH and GABA. Clinical improvement was assessed by HDRS-17 scale. Differences in the concentrations of metabolites were evaluated by MANOVA/MANCOVA and GLM repeated measures for longitudinal changes. RESULTS: FED patients had significantly decreased glutamate levels at baseline (p < 0.05) along with significantly elevated GABA (p < 0.01) compared to healthy controls. At the follow up, myo- Ins levels were significantly increased compared to baseline (p < 0.05) LIMITATIONS: The limited sample size, together with the unexpectedly high response rate after treatment (83%) might suggest decreased representativeness of the sample. CONCLUSIONS: Results indicate glutamatergic and GABAergic changes taking place within the ventromedial prefrontal region even at the early stage of depression prior to any medication treatment.

Association study of polymorphisms within inflammatory genes and methylation status in treatment response in major depression.

BACKGROUND: Although pharmacogenetics for major depressive disorder (MDD) is gaining momentum, the role of genetics in differences in response to antidepressant treatment is controversial, as they depend on multifactorial and polygenic phenotypes. Previous studies focused on the genes of the serotonergic system, leaving apart other pathological factors such as the inflammatory pathway. The main objective of the study was to assess whether treatment response might be associated with specific inflammation-related genetic variants or their methylation status. METHODS: 41 SNPs in 8 inflammatory genes: interleukin (IL) 1-ß, IL2, IL6, IL6R, IL10, IL18, tumor necrosis factor (TNF)-α and interferon (IFN)-γ were genotyped in 153 patients with MDD, who were evaluated with the Mausdley Staging Method to determine treatment response profiles. Pyrosequencing reactions and methylation quantification were performed in a PyroMark Q24 in 5 selected CpG islands of IL1- ß, IL6 and IL6R. Linear and logistic regression analyses were conducted, including age and gender as covariates using PLINK 1.07. RESULTS: Allelic distribution of IL1- ß rs1143643 was significantly associated with MSM scores (FDR corrected p = 0.04). Allelic distribution of IL6R rs57569414 showed a trend towards significance with MSM scores (p = 0.002; FDR corrected p = 0.07). Haplotype analyses showed associations between allelic combinations of IL1-ß and IL10 with treatment response (FDR corrected p < 0.01). Methylation percentage of treatment responders was only higher in an IL6R CpG island (p < 0.05). CONCLUSIONS: These exploratory findings suggest that IL1-ß and, marginally, IL6R polymorphisms may affect treatment response in major depression. If confirmed, these results may account for the heterogeneous phenotypes of major depression that underlie differences in treatment response.