Biology subjects

Penninx, B. W.

Publications and source records attributed to Penninx, B. W..

5 recordsLinked to original sources

bioRxiv2023

Multimodal Data Integration Advances Longitudinal Prediction of the Naturalistic Course of Depression and Reveals a Multimodal Signature of Disease Chronicity

The ability to individually predict disease course of major depressive disorder (MDD) is essential for optimal treatment planning. Here, we use a data-driven machine learning approach to assess the predictive value of different sets of biological data (whole-blood proteomics, lipid-metabolomics, transcriptomics, genetics), both separately and added to clinical baseline variables, for the longitudinal prediction of 2-year MDD chronicity (defined as presence of MDD diagnosis after 2 years) at the individual subject level. Prediction models were trained and cross-validated in a sample of 643 patients with current MDD (2-year chronicity n = 318) and subsequently tested for performance in 161 MDD individuals (2-year chronicity n = 79). Proteomics data showed best unimodal data predictions (AUROC = 0.68). Adding proteomic to clinical data at baseline significantly improved 2-year MDD chronicity predictions (AUROC = 0.63 vs AUROC = 0.78, p = 0.013), while the addition of other -omics data to clinical data did not yield significantly increased model performance. SHAP and enrichment analysis revealed proteomic analytes involved in inflammatory response and lipid metabolism, with fibrinogen levels showing the highest variable importance, followed by symptom severity. Machine learning models outperformed psychiatrists ability to predict two-year chronicity (balanced accuracy = 71% vs 55%). This study showed the added predictive value of combining proteomic, but not other -omic data, with clinical data. Adding other -omic data to proteomics did not further improve predictions. Our results reveal a novel multimodal signature of MDD chronicity that shows clinical potential for individual MDD disease course predictions from baseline measurements.

genomics↗

bioRxiv2020

Controlling for effects of confounding variables on machine learning predictions

Machine learning predictive models are being used in neuroimaging to predict information about the task or stimuli or to identify potentially clinically useful biomarkers. However, the predictions can be driven by confounding variables unrelated to the signal of interest, such as scanner effect or head motion, limiting the clinical usefulness and interpretation of machine learning models. The most common method to control for confounding effects is regressing out the confounding variables separately from each input variable before machine learning modeling. However, we show that this method is insufficient because machine learning models can learn information from the data that cannot be regressed out. Instead of regressing out confounding effects from each input variable, we propose controlling for confounds post-hoc on the level of machine learning predictions. This allows partitioning of the predictive performance into the performance that can be explained by confounds and performance independent of confounds. This approach is flexible and allows for parametric and non-parametric confound adjustment. We show in real and simulated data that this method correctly controls for confounding effects even when traditional input variable adjustment produces false-positive findings.

bioinformatics↗

bioRxiv2020

An integrative study of five biological clocks in somatic and mental health

Biological clocks have been developed at different molecular levels and were found to be more advanced in the presence of somatic illnesses and mental disorders. However, it is unclear whether different biological clocks reflect similar aging processes and determinants. In ~3000 subjects, we examined whether 5 biological clocks (telomere length, epigenetic, transcriptomic, proteomic and metabolomic clocks) were interrelated and associated to somatic and mental health determinants. Correlations between biological clocks were small (all r<0.2), indicating little overlap. The most consistent associations with the advanced biological clocks were found for male sex, higher BMI, metabolic syndrome, smoking and depression. As compared to the individual clocks, a composite index of all five clocks showed most pronounced associations with health determinants. The large effect sizes of the composite index and the low correlation between biological clocks, indicate that ones biological age is best reflected by combining aging measures from multiple cellular levels.

cell biology↗

bioRxiv2019

Predicting biological age based on the BBMRI-NL 1H-NMR metabolomics repository

The blood metabolome incorporates cues from the environment as well as the hosts genetic background, potentially offering a holistic view of an individuals health status. We have compiled a vast resource of 1H-NMR metabolomics and phenotypic data encompassing over 25,000 samples derived from 26 community and hospital-based cohorts. Using this resource, we constructed a metabolomics-based age predictor (metaboAge) to calculate an individuals biological age. Exploration in independent cohorts demonstrates that being judged older by ones metabolome, as compared to ones chronological age, confers an increased risk on future cardiovascular disease, mortality and functionality in older individuals. A web-based tool for calculating metaboAge (metaboage.researchlumc.nl) allows easy incorporation in other epidemiological studies. Access to data can be requested at bbmri.nl/samples-images-data. In summary, we present a vast resource of metabolomics data and illustrate its merit by constructing a metabolomics-based score for biological age that captures aspects of current and future cardio-metabolic health.

epidemiology↗

bioRxiv2019

Vitamin D: marker, cause or consequence of depression? An exploration using genomics

BackgroundTrials testing the effect of vitamin D or omega-3 fatty acids (n3-PUFA) supplementation on major depressive disorder (MDD) reported conflicting findings. These trials were boosted by epidemiological evidence suggesting an inverse association of circulating 25-hydroxyvitamin D (25-OH-D) and n3-PUFA levels with MDD. Observational associations may emerge from unresolved confounding, shared genetic risk, or direct causal relationships. We explored the nature of these associations exploiting data and statistical tools from genomics. MethodsResults from GWAS on 25-OH-D (N = 79366), n3-PUFA (N = 24925) and MDD (135458 cases, 344901 controls) were applied to individual-level data (>2,000 subjects with measures of genotype, DSM-IV lifetime MDD diagnoses and circulating 25-OH-D and n3-PUFA) and summary-level data analyses. Shared genetic risk between traits was tested by polygenic risk scores (PRS). Two-sample Mendelian Randomization (2SMR) analyses tested the potential bidirectional causality between traits. OutcomeIn individual-level data, PRS were associated with the phenotype of the same trait (PRS 25-OH-D p = 1.4e-20, PRS N3-PUFA p = 9.3e-6, PRS MDD p = 1.4e-4), but not with the other phenotypes, suggesting a lack of shared genetic effects. In summary-level data, 2SMR analyses provided no evidence of a causal role on MDD of 25-OH-D (p = 0.50) or n3-PUFA (p = 0.16), or for a causal role of MDD on 25-OH-D (p = 0.25) or n3-PUFA (p = 0.66). ConclusionsApplying genomics tools indicated that that shared genetic risk or direct causality between 25-OH-D, n3-PUFA and MDD is unlikely: unresolved confounding may explain the associations reported in observational studies. These findings represent a cautionary tale for testing supplementation of these compounds in preventing or treating MDD. Research in contextO_ST_ABSEvidence before this studyC_ST_ABSMeta-analyses of trials testing the effect of vitamin D or omega-3 fatty acids (n3-PUFA) supplementation on major depressive disorder (MDD) reported conflicting findings, including small clinical effect or no effect. These trials were boosted by epidemiological evidence suggesting an inverse association of circulating 25-hydroxyvitamin D (25-OH-D) and n3-PUFA levels with MDD. However, observational associations may emerge from different scenarios, including unresolved confounding, shared genetic risk, or direct causal relationships. Added value of this studyGenomics provides unique opportunities to investigate shared risk and causality between traits applying new statistical tools and results from genome-wide association studies (GWAS). In the present study we examined the nature of the association of 25-OH-D and n3-PUFA with MDD using the latest data and tools from genomics. We found no significant evidence of shared genetic risk or direct causality between vitamin D or n-3 PUFA and MDD; at this stage, unresolved confounding should be considered the most likely explanation for the association reported by observational studies. Implications of all the available evidenceFindings from the present study, in conjunction with previous conflicting evidence from clinical studies, represent a cautionary tale for further research testing the potential therapeutic effect of vitamin D and n3-PUFA supplementation on depression, as the expectations of a direct causal effect of these compounds on mood should be substantially reconsidered. Genomic tools could be efficiently employed to examine the nature of observational associations emerging in epidemiology, providing some indications on the most promising associations to be prioritized in subsequent intervention studies.

genomics↗