After the success of polygenic risk scores (PRS) that embed a useful summary of genomic information in a comprehensive score, the wish to develop summary statistics for DNA methylation had become more pressing. Developing such a score faces challenges, as the score has to be specific and sensitive as well. Epidemiological research on DNA methylation and depression would benefit from such score.

Here, we test a score trained on incident depression (case-control), i.e., a list of published weights for particular CpGs, for its validity in the context of depression severity as measured using MADRS in our sample with depressed patients only.

DNA methylation was assessed using the Illumina Infinium MethylationEPIC 850k BeadChip on a sample of 119 patients with a diagnosis of MDD. After data cleaning, 113 participants were included in the analysis (Mage= 47 years, 57.98% women, MMADRS=27.7). Data processing was conducted using the RnBeads package. From the published reference for the overall sample, a list of 196 CpGs was provided, 170 of these were present in our dataset and used for the score. The list of non-smokers comprised 144 CpGs, of which 124 were available. The score per individual was built using M-values, using the formula: S(weight*DNA methylation value). The score was tested in association with depression and other typical confounders using multiple regression in R. Confounders included ancestry, BMI, age, sex, and 6 cell types. We tested both scores in our sample: smokers and non-smokers.

In contrast to our expectations, none of the regression analyses showed a significant association with depression (MADRS-score). Nonetheless, a significant association was seen with biological sex for both analysis (overall: p=0.036, non-smokers: p=0.026). A reduced model with only this predictor explained 5% and 4% of the variance of the summary score calculated (R2), respectively (overall: p=0.013; non-smokers: p=0.019). One of the ancestry components was marginally significant too in the non-smoker summary score (p=0.065). This was not the case anymore in the reduced model.

Our results show that caution is still in place when using methylation risk scores as specificity and sensitivity might not yet be optimized. The score built for depression incidence does not seem fitting for depression severity at this moment. The use of DNA methylation, a marker that is generally sensitive to confounding factors, for a risk score, might pose more challenges in the context of reliable summary statistics, in particular also for cross-trait examination, which is currently a typical use of polygenic risk scores.

None Declared

Research repeatedly linked inflammation with major depressive disorder (MDD). The presence of an inflammatory subtype of depression is supported by molecular findings as well as imaging reports. We investigated the cell type composition estimated by using epigenome-wide DNA methylation markers in a sample of depressed individuals showing high or low inflammation levels measured by hsCRP. We aimed to understand the connection between depression and inflammation, specifically differences in cell type compositions between high and low inflammation groups at baseline.

119 individuals with MDD were included for this analysis. Following quality control procedures, 113 participants were included in the analysis (Mage= 47 years, 57.98% women). The sample consisted of 37 individuals with high hsCRP (hsCRP > 1.5, Mage=45, MhsCRP=8.2, MMADRS=28, 70% women) and 76 individuals with low hsCRP (hsCRP < 1.5, Mage= 44, MhsCRP=0.99, MMADRS=28, 49% women).

The Illumina Infinium MethylationEPIC 850k BeadChip was used for analyzing whole blood derived DNA. Data processing and cell type estimation was conducted using the RnBeads package. We applied the Houseman method to estimate cell type composition through epigenome-wide DNA methylation signatures, resulting in six cell types: neutrophils, natural killer cells, B cells, CD4+ T cells, CD8+ T cells and monocytes. Comparisons between both groups were tested using ANOVA.

High and low hsCRP groups were compared for each of the six cell types estimated. A statistically significant difference was seen for monocytes (p=0.0316) and a trend for neutrophils (p=0.0742). The mean values for neutrophils in patients without inflammation were found to be 60%, while in patients with inflammation, it was 63%. For monocytes, the mean values for patients without inflammation and those with inflammation were 10% and 9.4%, respectively, with a smaller range (4.5%-14.3%) for individuals with inflammation as compared to patients without inflammation (5.3%-20.7%). None of the other four cell types showed a statistically significant difference.

We identified differences in the cell type composition between groups of depressed patients with high versus low inflammation. These results align with the existing body of knowledge reported in established academic literature. Our study emphasizes the role of specific cells like neutrophils and monocytes in inflammation and depression. These findings offer valuable insights for improving depression treatment strategies as inflammation state may be relevant for treatment response. We also show the merit of DNA methylation signatures for the profiling of patients’ inflammation status, i.e., immunomethylomics.

None Declared

Depression is characterised by negative views of the self. Antidepressant treatment may remediate negative self-schema through increasing processing of positive information about the self. Changes in affective processing during social interactions may increase expression of prosocial behaviours, improving interpersonal communications.

To examine whether acute administration of citalopram is associated with an increase in positive affective learning biases about the self and prosocial behaviour.

Healthy volunteers (n = 41) were randomised to either an acute 20 mg dose of citalopram or matched placebo in a between-subjects double-blind design. Participants completed computer-based cognitive tasks designed to measure referential affective processing, social cognition and expression of prosocial behaviours.

Participants administered citalopram made more cooperative choices than those administered placebo in a prisoner's dilemma task (β = 20%, 95% CI: 2%, 37%). Exploratory analyses indicated that participants administered citalopram showed a positive bias when learning social evaluations about a friend (β = 4.06, 95% CI: 0.88, 7.24), but not about the self or a stranger. Similarly, exploratory analyses found evidence of increased recall of positive words and reduced recall of negative words about others (β = 2.41, 95% CI: 0.89, 3.93), but not the self, in the citalopram group.

Participants administered citalopram showed greater prosocial behaviours, increased positive recall and increased positive learning of social evaluations towards others. The increase in positive affective bias and prosocial behaviours towards others may, at least partially, be a mechanism of antidepressant effect. However, we found no evidence that citalopram influenced self-referential processing.