Association between admission ECG changes and long-term mortality in patients with an incidental myocardial infarction: Results from the KORA myocardial infarction registry

BACKGROUND

The aim of this study was to examine the predictive value of specific changes in admission ECG on long-term outcome in acute myocardial infarction (AMI).

METHODS

From 2000 until 2017 all AMI cases (n = 9,689) in the study area of Augsburg, Germany, were prospectively recorded. For this study, all patients with a first-time AMI, who survived the first 28 days, were considered. Median observational time was 6.7 years (IQR: 3.6-10.9). Each case was assigned to one of the following groups according to the admission ECG: 'ST-segment elevation', 'ST-segment depression', 'T-wave inversion', 'unspecific changes', 'normal ECG' and 'bundle branch block' (BBB). Multivariable adjusted COX regression models were calculated to compare long-term all-cause mortality.

RESULTS

The final regression model revealed a significantly higher mortality among patients with BBB (HR: 1.52 [1.34-1.73], p-value: < 0.001) and 'ST-segment depression' (HR: 1.16 [1.03-1.29], p-value: 0.01252) compared to the STEMI group (reference group). The 'normal ECG' group (HR: 0.76 [0.66-0.87], p-value: < 0.001) on the other hand was associated with significantly lower long-term mortality. The 'T-wave inversion' group (HR: 1.08 [0.96-1.21]) and the 'unspecific changes' group (HR: 1.05 [0.94-1.17]) did not differ significantly from the STEMI group.

CONCLUSION

ST-segment depressions and BBB admission ECGs go along with higher long-term mortality in AMI patients compared to STEMI cases. This should be taken into account by physicians when treating patients with NSTEMIs. Only the complete absence of AMI-related ECG changes predicts a more favorable outcome.

Integrative analysis of clinical and epigenetic biomarkers of mortality

DNA methylation (DNAm) has been reported to be associated with many diseases and with mortality. We hypothesized that the integration of DNAm with clinical risk factors would improve mortality prediction. We performed an epigenome-wide association study of whole blood DNAm in relation to mortality in 15 cohorts (n = 15,013). During a mean follow-up of 10 years, there were 4314 deaths from all causes including 1235 cardiovascular disease (CVD) deaths and 868 cancer deaths. Ancestry-stratified meta-analysis of all-cause mortality identified 163 CpGs in European ancestry (EA) and 17 in African ancestry (AA) participants at p < 1 × 10^-7 , of which 41 (EA) and 16 (AA) were also associated with CVD death, and 15 (EA) and 9 (AA) with cancer death. We built DNAm-based prediction models for all-cause mortality that predicted mortality risk after adjusting for clinical risk factors. The mortality prediction model trained by integrating DNAm with clinical risk factors showed an improvement in prediction of cancer death with 5% increase in the C-index in a replication cohort, compared with the model including clinical risk factors alone. Mendelian randomization identified 15 putatively causal CpGs in relation to longevity, CVD, or cancer risk. For example, cg06885782 (in KCNQ4) was positively associated with risk for prostate cancer (Beta = 1.2, P_MR  = 4.1 × 10^-4 ) and negatively associated with longevity (Beta = -1.9, P_MR  = 0.02). Pathway analysis revealed that genes associated with mortality-related CpGs are enriched for immune- and cancer-related pathways. We identified replicable DNAm signatures of mortality and demonstrated the potential utility of CpGs as informative biomarkers for prediction of mortality risk.

Genome-wide studies reveal factors associated with circulating uromodulin and its relationships to complex diseases

Uromodulin (UMOD) is a major risk gene for monogenic and complex forms of kidney disease. The encoded kidney-specific protein uromodulin is highly abundant in urine and related to chronic kidney disease, hypertension, and pathogen defense. To gain insights into potential systemic roles, we performed genome-wide screens of circulating uromodulin using complementary antibody-based and aptamer-based assays. We detected 3 and 10 distinct significant loci, respectively. Integration of antibody-based results at the UMOD locus with functional genomics data (RNA-Seq, ATAC-Seq, Hi-C) of primary human kidney tissue highlighted an upstream variant with differential accessibility and transcription in uromodulin-synthesizing kidney cells as underlying the observed cis effect. Shared association patterns with complex traits, including chronic kidney disease and blood pressure, placed the PRKAG2 locus in the same pathway as UMOD. Experimental validation of the third antibody-based locus, B4GALNT2, showed that the p.Cys466Arg variant of the encoded N-acetylgalactosaminyltransferase had a loss-of-function effect leading to higher serum uromodulin levels. Aptamer-based results pointed to enzymes writing glycan marks present on uromodulin and to their receptors in the circulation, suggesting that this assay permits investigating uromodulin's complex glycosylation rather than its quantitative levels. Overall, our study provides insights into circulating uromodulin and its emerging functions.

Genetically regulated gene expression and proteins revealed discordant effects

Background

Although gene-expression (GE) and protein levels are typically strongly genetically regulated, their correlation is known to be low. Here we investigate this phenomenon by focusing on the genetic background of this correlation in order to understand the similarities and differences in the genetic regulation of these omics layers.

Methods and results

We performed locus-wide association studies of 92 protein levels measured in whole blood for 2,014 samples of European ancestry and found that 66 are genetically regulated. Three female- and one male-specific effects were detected. We estimated the genetically regulated GE for all significant genes in 49 GTEx v8 tissues. A total of 7 proteins showed negative correlations with their respective GE across multiple tissues. Finally, we tested for causal links of GE on protein expression via Mendelian Randomization, and confirmed a negative causal effect of GE on protein level for five of these genes in a total of 63 gene-tissue pairs: BLMH, CASP3, CXCL16, IL6R, and SFTPD. For IL6R, we replicated the negative causal effect on coronary-artery disease (CAD), while its GE was positively linked to CAD.

Conclusion

While total GE and protein levels are only weakly correlated, we found high correlations between their genetically regulated components across multiple tissues. Of note, strong negative causal effects of tissue-specific GE on five protein levels were detected. Causal network analyses revealed that GE effects on CAD risks was in general mediated by protein levels.

A randomization-based causal inference framework for uncovering environmental exposure effects on human gut microbiota

Statistical analysis of microbial genomic data within epidemiological cohort studies holds the promise to assess the influence of environmental exposures on both the host and the host-associated microbiome. However, the observational character of prospective cohort data and the intricate characteristics of microbiome data make it challenging to discover causal associations between environment and microbiome. Here, we introduce a causal inference framework based on the Rubin Causal Model that can help scientists to investigate such environment-host microbiome relationships, to capitalize on existing, possibly powerful, test statistics, and test plausible sharp null hypotheses. Using data from the German KORA cohort study, we illustrate our framework by designing two hypothetical randomized experiments with interventions of (i) air pollution reduction and (ii) smoking prevention. We study the effects of these interventions on the human gut microbiome by testing shifts in microbial diversity, changes in individual microbial abundances, and microbial network wiring between groups of matched subjects via randomization-based inference. In the smoking prevention scenario, we identify a small interconnected group of taxa worth further scrutiny, including Christensenellaceae and Ruminococcaceae genera, that have been previously associated with blood metabolite changes. These findings demonstrate that our framework may uncover potentially causal links between environmental exposure and the gut microbiome from observational data. We anticipate the present statistical framework to be a good starting point for further discoveries on the role of the gut microbiome in environmental health.

Environmental influences on the human gut microbiome are still to be discovered or better understood. In this paper, we contribute to the field of microbiome research and environmental epidemiology by suggesting a stage-based causal inference framework relying on the foundations of the Rubin Causal Model. A particularity of the framework is the use of randomization-based inference, which we value to be a necessary exploratory inference method when tackling untapped research questions. To illustrate the framework, we explore the effects of two inhaled environmental exposures previously hypothesized to be linked with gastrointestinal diseases and the gut microbiome: air pollution exposure and cigarette smoking.

DNA methylation signature of chronic low-grade inflammation and its role in cardio-respiratory diseases

We performed a multi-ethnic Epigenome Wide Association study on 22,774 individuals to describe the DNA methylation signature of chronic low-grade inflammation as measured by C-Reactive protein (CRP). We find 1,511 independent differentially methylated loci associated with CRP. These CpG sites show correlation structures across chromosomes, and are primarily situated in euchromatin, depleted in CpG islands. These genomic loci are predominantly situated in transcription factor binding sites and genomic enhancer regions. Mendelian randomization analysis suggests altered CpG methylation is a consequence of increased blood CRP levels. Mediation analysis reveals obesity and smoking as important underlying driving factors for changed CpG methylation. Finally, we find that an activated CpG signature significantly increases the risk for cardiometabolic diseases and COPD.

MO048: Genome-wide studies reveal factors associated with circulating uromodulin and its relations with complex diseases

BACKGROUND AND AIMS

UMOD is a major risk gene for monogenic and complex forms of kidney disease. The encoded kidney-specific protein uromodulin is the most abundant protein in urine and related to chronic kidney disease, hypertension and pathogen defense. Through basolateral release from kidney epithelial cells, uromodulin also reaches the blood, where its function is largely unknown. To gain insights into potential systemic roles, we performed genome-wide screens of circulating uromodulin in seven cohorts using two complementary assays.

METHOD

Separate genome-wide association study meta-analyses for circulating uromodulin were conducted for the antibody-based assay (five cohorts, N = 13 985) and the aptamer-based SOMAscan assay (two cohorts, N = 18 070). Genome-wide significant loci were placed into their functional genomic context using RNA-seq, ATAC-seq and Hi-C data generated from primary human kidney tissue. An array of downstream genetic analyses was then performed for significant loci, including fine-mapping, colocalization analyses and gene-by-gene interaction analyses. The B4GALNT2 p.Cys466Arg allele was expressed in MDCK cells and studied by immunofluorescence and Western blotting analyses.

RESULTS

We detected and replicated 13 genome-wide significant loci (P &lt;5e−8; 12 novel). At the UMOD locus, functional genomics data of primary human kidney tissue highlighted an upstream regulatory variant with differential accessibility and UMOD transcription in uromodulin-synthesizing kidney cells. Shared association patterns with complex traits, including chronic kidney disease and blood pressure, placed the PRKAG2 locus in the same context as UMOD. Experimental validation of another locus, B4GALNT2, showed that the detected p.Cys466Arg variant of the encoded N-acetylgalactosaminyltransferase has a loss-of-function effect leading to higher serum uromodulin levels. Lastly, our results point to enzymes writing glycan marks present on uromodulin and to their receptors in the circulation.

CONCLUSION

This study provides human genetic evidence of new pathway members of uromodulin and delivers novel insights into its determinants and systemic role in the circulation.

Setting the European environment and health research agenda -under-researched areas and solution-oriented research

BACKGROUND

The aim of the EU-funded HERA (health and environment research agenda) project is to set priorities for the future European research agenda in the environment, climate and health nexus. We report results from a European researcher's perspective and identify research areas that have been inadequately investigated to date.

METHODS

An online survey was completed by European researchers to assess, evaluate and visualise research gaps. These research gaps were identified for 21 predefined areas within 3 main categories: i) classical environment and health paradigm; ii) problem or sector-based research areas and approaches and iii) holistic research areas and concepts. All research gaps were then evaluated by expert groups with the pre-defined criteria and systematically summarized. For areas identified within the survey as under-reported, additional input was sought from a range of key selected experts. The EU project database Cordis was utilized to verify that these areas were under-researched.

RESULTS

Between May and July 2019, 318 respondents from 38 countries reported 624 research gaps. The main areas for attention identified were: urban environments; chemicals; and climate change, (combined n = 313 gaps). Biodiversity loss and health; transport, mobility, sustainable solutions and health; energy transition and health; waste and the circular economy and health; ethics and philosophy and health were areas that were acknowledged as under-researched (combined n = 27 gaps). These under-researched areas were identified as having certain commonalities, they: i) mostly fell in the category "problem or sector based approaches"; ii) they are essential for developing and implementing solutions; and iii) require trans-disciplinary and cross-sectoral collaboration.

CONCLUSIONS

Currently attention is given to topical and highly researched areas in environmental health. In contrast, this paper identifies key topics and approaches that are under-researched, yet, are critical for the implementation of the EU Green Deal, related strategies and action plans, and require further investigation and investment. The findings reveal the imperative to foster solutions-oriented, trans-disciplinary and participatory research and its implementation through changes in research funding and research structures.

Multi-ancestry genetic study of type 2 diabetes highlights the power of diverse populations for discovery and translation

We assembled an ancestrally diverse collection of genome-wide association studies (GWAS) of type 2 diabetes (T2D) in 180,834 affected individuals and 1,159,055 controls (48.9% non-European descent) through the Diabetes Meta-Analysis of Trans-Ethnic association studies (DIAMANTE) Consortium. Multi-ancestry GWAS meta-analysis identified 237 loci attaining stringent genome-wide significance (P < 5 × 10-9), which were delineated to 338 distinct association signals. Fine-mapping of these signals was enhanced by the increased sample size and expanded population diversity of the multi-ancestry meta-analysis, which localized 54.4% of T2D associations to a single variant with >50% posterior probability. This improved fine-mapping enabled systematic assessment of candidate causal genes and molecular mechanisms through which T2D associations are mediated, laying the foundations for functional investigations. Multi-ancestry genetic risk scores enhanced transferability of T2D prediction across diverse populations. Our study provides a step toward more effective clinical translation of T2D GWAS to improve global health for all, irrespective of genetic background.

The effect of LPA Thr3888Pro on lipoprotein(a) and coronary artery disease is modified by the LPA KIV-2 variant 4925G>A

BACKGROUND AND AIMS

High lipoprotein(a) [Lp(a)] concentrations are associated with increased coronary artery disease (CAD) risk. Lp(a) is regulated mainly genetically by the LPA gene but involved genetic variants have not been fully elucidated. Improved understanding of the entanglements of genetic Lp(a) regulation may enhance genetic prediction of Lp(a) and CAD risk. We investigated an interaction between the well-known LPA missense SNP rs41272110 (known as Thr3888Pro) and the frequent LPA splicing mutation KIV-2 4925G>A.

METHODS

Effects on Lp(a) concentrations were investigated by multiple quantile regression in the German Chronic Kidney Disease (GCKD) study, KORA-F3 and KORA-F4 (n_total = 10,405) as well as in the UK Biobank (UKB) 200k exome dataset (n = 173,878). The impact of the interaction on CAD risk was assessed by survival analysis in UKB.

RESULTS

We observed a significant SNP-SNP interaction in all studies (p = 1.26e-05 to 3.03e-04). In quantile regression analysis, rs41272110 as a predictor shows no impact on Lp(a) (β = -0.06 [-0.79; 0.68], p = 0.879), but in a joint model including both SNPs as predictors, rs41272110 is associated with markedly higher Lp(a) (β = +9.40 mg/dL [6.45; 12.34], p = 4.07e-10). Similarly, rs41272110 shows no effect on CAD in UKB (HR = 1.01 [0.97; 1.04], p = 0.731), while rs41272110 carriers not carrying 4925G>A show an increased CAD risk (HR = 1.10 [1.04; 1.16], p = 6.9e-04). This group corresponds to 4% of the population. Adjustment for apolipoprotein(a) isoforms further modified the effect estimates markedly.

CONCLUSIONS

This work emphasizes the complexity of the genetic regulation of Lp(a) and the importance to account for genetic subgroups in Lp(a) association studies and when interpreting genetic cardiovascular risk profiles.

Association of C-Terminal Pro-Endothelin-1 with Mortality in the Population-Based KORA F4 Study

Introduction

Endothelin-1 and its prohormone C-terminal pro-endothelin-1 (CT-proET-1) have been linked to metabolic alterations, inflammatory responses and cardiovascular events in selected study populations. We analyzed the association of CT-proET-1 with cardiovascular events and mortality, carotid intima-media-thickness as surrogate for early atherosclerotic lesions, biomarkers of subclinical inflammation and adipokines in a population-based study.

Methods

The cross-sectional and prospective analyses used data from the KORA F4 study with a median follow-up time of 9.1 (8.8–9.4) years. Data on CT-proET-1 and mortality were available for 1554 participants, data on the other outcomes in subgroups (n = 596–1554). The associations were estimated using multivariable linear regression and Cox proportional hazard models adjusted for sex, age, body mass index, estimated glomerular filtration rate, arterial hypertension, diabetes, low-density and high-density lipoprotein cholesterol, current and former smoking and physical activity. The Bonferroni method was used to correct for multiple testing.

Results

In the fully adjusted model, CT-proET-1 was associated with cardiovascular (hazard ratio (HR) per standard deviation increase: 1.66; 95% confidence interval (CI): 1.10–2.51; p = 0.017) and all-cause mortality (HR: 2.03; 95% CI 1.55–2.67; p < 0.001), but not with cardiovascular events, and was inversely associated with the intima-media thickness (β: −0.09 ± 0.03; p = 0.001). CT-proET-1 was positively associated with five out of ten biomarkers of subclinical inflammation and with two out of five adipokines after correction for multiple testing. After inclusion of biomarkers of subclinical inflammation in the Cox proportional hazard model, the association of CT-proET-1 with all-cause mortality persisted (p < 0.001).

Conclusion

These results emphasize the complexity of endothelin-1 actions and/or indicator functions of CT-proET-1. CT-proET-1 is a risk marker for all-cause mortality, which is likely independent of vascular endothelin-1 actions, cardiovascular disease and inflammation.

Undiagnosed Impaired Glucose Tolerance and Type-2 Diabetes in Acute Myocardial Infarction Patients: Fequency, Characteristics and Long-Term Mortality

Background

In this study we investigated the prevalence of undiagnosed impaired glucose tolerance and type-2-diabetes (T2D) among patients with acute myocardial infarction (AMI) and prospectively analyzed whether these patients have a higher long-term mortality.

Methods

The analysis was based on 2,317 AMI patients aged 25–84 years from the population-based Myocardial Infarction Registry Augsburg, recruited between 2009 and 2014 and followed-up until 2019 (median follow-up time 6.5 years [IQR: 4.9–8.1]). AMI patients with a diagnosis of diabetes were divided into a high (>7.0%) and a low HbA1c group (≤7.0%) according to HbA1c values at admission. The remaining patients (without known diabetes) were grouped into normal (<5.7%), elevated (5.7–6.4%), and high (≥6.5%) HbA1c groups. In a multivariable-adjusted COX regression analysis, the association between HbA1c groups and long-term mortality was investigated. Linear regression models were used to identify AMI patients with elevated HbA1c values by means of personal characteristics.

Results

At admission, 29.5% of all patients reported a diagnosis of diabetes. Of all patients without known diabetes, 5.4% had HbA1c values of ≥ 6.5 and 37.9% had HbA1c values between 5.7 and 6.4%. The fully adjusted Cox regression model showed a non-significant trend toward higher long-term mortality for AMI patients with increased HbA1c values (HbA1c 5.7–6.4% HR: 1.05 [0.79–1.38], HbA1c > 6.5% HR: 1.34 [0.77–2.31]). A linear regression model including the variables admission serum glucose, BMI, age, sex and type of infarction (STEMI, NSTEMI) showed only poor prediction of HbA1c values (R²: 11.08%).

Conclusion

A fairly high number of AMI patients without known diabetes have elevated HbA1c values. Though we could not prove a higher risk of premature mortality in these patients, early detection and adequate therapy might lead to reduced diabetes-associated complications and improve long-term outcomes.

A Family and a Genome-Wide Polygenic Risk Score Are Independently Associated With Stroke in a Population-Based Study

BACKGROUND

Positive family history and genetic risk scores have been shown to independently capture those individuals with high risk for stroke. The aim of our study was to evaluate the amount of shared information between family history and genetic risk and to investigate their combined effect on the association with prevalent and incident stroke cases.

METHODS

We obtained a family risk score (FamRS), weighted for disease onset and family size as well as genome-wide polygenic risk score (PGS) including over 3.2 million single-nucleotide polymorphisms in the population-based prospective KORA F3 (Cooperative Health Research in the Region of Augsburg) study (n=3071) from Southern Germany. FamRS and PGS were evaluated separately and combined. The measures were once treated as continuous variables but also divided in the highest 20%, 10%, 5%, and 1% percentiles. Odds ratios via logistic regression and hazard ratios via Cox regression were estimated. A stroke event was defined as a hospitalization for stroke that was self-reported in a standardized interview by certified and supervised personnel.

RESULTS

The FamRS outperformed other simplified family measures such as affected parents or number of affected family members. FamRS and PGS were not correlated, and no individuals were observed with both very high FamRS and very high PGS (top 1% percentile). In a combined model, both FamRS and PGS were independently from each other associated with risk of stroke, also independent of other traditional risk factors (p [FamRS]=0.02, p [PGS]=0.005). Individuals in the top 1% of either FamRS or PGS were found to have >5-fold risk for stroke (odds ratios, 5.82 [95% CI, 2.08-14]; P=0.0002). The results for incident stroke events showed the same trend but were not significant.

CONCLUSIONS

Our study shows that a family risk score and PGS capture different information concerning individual stroke risk. Combining the risk measures FamRS and PGS increases predictive power, as demonstrated in a population-based study.

Genome-Wide Characterization of a Highly Penetrant Form of Hyperlipoprotein(a)emia Associated With Genetically Elevated Cardiovascular Risk

Lp(a) (lipoprotein [a]) is a highly atherogenic lipoprotein strongly associated with coronary artery disease (CAD). Lp(a) concentrations are chiefly determined genetically. Investigation of large pedigrees with extreme Lp(a) using modern whole-genome approaches may unravel the genetic determinants underpinning this pathological phenotype.

Population-based cohort imaging: skeletal muscle mass by magnetic resonance imaging in correlation to bioelectrical-impedance analysis

BACKGROUND

Skeletal muscle mass is subjected to constant changes and is considered a good predictor for outcome in various diseases. Bioelectrical-impedance analysis (BIA) and magnetic resonance imaging (MRI) are approved methodologies for its assessment. However, muscle mass estimations by BIA may be influenced by excess intramuscular lipids and adipose tissue in obesity. The objective of this study was to evaluate the feasibility of quantitative assessment of skeletal muscle mass by MRI as compared with BIA.

METHODS

Subjects from a population-based cohort underwent BIA (50 kHz, 0.8 mA) and whole-body MRI including chemical-shift encoded MRI (six echo times). Abdominal muscle mass by MRI was quantified as total and fat-free cross-sectional area by a standardized manual segmentation-algorithm and normalized to subjects' body height² (abdominal muscle mass indices: AMMI_MRI ).

RESULTS

Among 335 included subjects (56.3 ± 9.1 years, 56.1% male), 95 (28.4%) were obese (BMI ≥ 30 kg/m² ). MRI-based and BIA-based measures of muscle mass were strongly correlated, particularly in non-obese subjects [r < 0.74 in non-obese (P < 0.001) vs. r < 0.56 in obese (P < 0.001)]. Median AMMI_Total(MRI) was significantly higher in obese as compared with non-obese subjects (3246.7 ± 606.1 mm² /m² vs. 2839.0 ± 535.8 mm² /m² , P < 0.001, respectively), whereas the ratio AMMI_Fat-free /AMMI_Total (by MRI) was significantly higher in non-obese individuals (59.3 ± 10.1% vs. 53.5 ± 10.6%, P < 0.001, respectively). No significant difference was found regarding AMMI_{Fat-free(MRI)} (P = 0.424). In analyses adjusted for age and sex, impaired glucose tolerance and measures of obesity were significantly and positively associated with AMMI_Total(MRI) and significantly and inversely with the ratio AMMI_{Fat-free(MRI)} /AMMI_Total(MRI) (P < 0.001).

CONCLUSIONS

MRI-based assessment of muscle mass is feasible in population-based imaging and strongly correlated with BIA. However, the observed weaker correlation in obese subjects may explain the known limitation of BIA in obesity and promote MRI-based assessments. Thus, skeletal muscle mass parameters by MRI may serve as practical imaging biomarkers independent of subjects' body weight.

Novel associations between inflammation-related proteins and adiposity: A targeted proteomics approach across four population-based studies

Chronic low-grade inflammation has been proposed as a linking mechanism between obesity and the development of inflammation-related conditions such as insulin resistance and cardiovascular disease. Despite major advances in the last 2 decades, the complex relationship between inflammation and obesity remains poorly understood. Therefore, we aimed to identify novel inflammation-related proteins associated with adiposity. We investigated the association between BMI and waist circumference and 72 circulating inflammation-related proteins, measured using the Proximity Extension Assay (Olink Proteomics), in 3,308 participants of four independent European population-based studies (KORA-Fit, BVSII, ESTHER, and Bialystok PLUS). In addition, we used body fat mass measurements obtained by Dual-energy X-ray absorptiometry (DXA) in the Bialystok PLUS study to further validate our results and to explore the relationship between inflammation-related proteins and body fat distribution. We found 14 proteins associated with at least one measure of adiposity across all four studies, including four proteins for which the association is novel: DNER, SLAMF1, RANKL, and CSF-1. We confirmed previously reported associations with CCL19, CCL28, FGF-21, HGF, IL-10RB, IL-18, IL-18R1, IL-6, SCF, and VEGF-A. The majority of the identified inflammation-related proteins were associated with visceral fat as well as with the accumulation of adipose tissue in the abdomen and the trunk. In conclusion, our study provides new insights into the immune dysregulation observed in obesity that might help uncover pathophysiological mechanisms of disease development.

Physical Activity, Incidence, and Progression of Age-Related Macular Degeneration: A Multicohort Study

PURPOSE

To investigate the impact of physical activity (PA) on the incidence or progression of age-related macular degeneration (AMD) in the general population.

DESIGN

Meta-analysis of longitudinal cohort studies.

METHODS

We included 14,630 adults with no or early AMD at baseline from 7 population-based studies and examined associations of PA with AMD incidence and progression using multistate models (MSM) per study and subsequent random effects meta-analysis. Age effects were assessed using meta-regression. The main outcome measure was the hazard ratio (HR) for incident early or progression to late AMD.

RESULTS

At baseline, mean age was 60.7 ± 6.9 to 76.4 ± 4.3 years, and prevalence of early AMD was 7.7% (range, 3.6%-16.9%) between cohorts. During follow-up, 1461 and 189 events occurred for early and late AMD, respectively. In meta-analyses, no or low to moderate PA (high PA as reference) was associated with an increased risk for incident early AMD (HR, 1.19; 95% CI, 1.01-1.40; P = .04), but not for late AMD. In subsequent meta-regression, we found no association of age with the effect of PA on incident AMD.

CONCLUSIONS

Our study suggests high levels of PA to be protective for the development of early AMD across several population-based cohort studies. Our results establish PA as a modifiable risk factor for AMD and inform further AMD prevention strategies to reduce its public health impact.

Simple cardiovascular risk stratification by replacing total serum cholesterol with anthropometric measures: The MORGAM prospective cohort project

To assess whether anthropometric measures (body mass index [BMI], waist-hip ratio [WHR], and estimated fat mass [EFM]) are independently associated with major adverse cardiovascular events (MACE), and to assess their added prognostic value compared with serum total-cholesterol. The study population comprised 109,509 individuals (53% men) from the MORGAM-Project, aged 19–97 years, without established cardiovascular disease, and not on antihypertensive treatment. While BMI was reported in all, WHR and EFM were reported in ∼52,000 participants. Prognostic importance of anthropometric measurements and total-cholesterol was evaluated using adjusted Cox proportional-hazards regression, logistic regression, area under the receiver-operating-characteristic curve (AUC_ROC), and net reclassification improvement (NRI). The primary endpoint was MACE, a composite of stroke, myocardial infarction, or death from coronary heart disease. Age interacted significantly with anthropometric measures and total-cholesterol on MACE (P ≤ 0.003), and therefore age-stratified analyses (<50 versus ≥ 50 years) were performed. BMI, WHR, EFM, and total-cholesterol were independently associated with MACE (P ≤ 0.003) and resulted in significantly positive NRI when added to age, sex, smoking status, and systolic blood pressure. Only total-cholesterol increased discrimination ability (AUC_ROC difference; P < 0.001). In subjects < 50 years, the prediction model with total-cholesterol was superior to the model including BMI, but not superior to models containing WHR or EFM, while in those ≥ 50 years, the model with total-cholesterol was superior to all models containing anthropometric variables, whether assessed individually or combined. We found a potential role for replacing total-cholesterol with anthropometric measures for MACE-prediction among individuals < 50 years when laboratory measurements are unavailable, but not among those ≥ 50 years.

Associations of carotid intima media thickness with gene expression in whole blood and genetically predicted gene expression across 48 tissues

Carotid intima media thickness (cIMT) is a biomarker of subclinical atherosclerosis and a predictor of future cardiovascular events. Identifying associations between gene expression levels and cIMT may provide insight to atherosclerosis etiology. Here, we use two approaches to identify associations between mRNA levels and cIMT: differential gene expression analysis in whole blood and S-PrediXcan. We used microarrays to measure genome-wide whole blood mRNA levels of 5647 European individuals from four studies. We examined the association of mRNA levels with cIMT adjusted for various potential confounders. Significant associations were tested for replication in three studies totaling 3943 participants. Next, we applied S-PrediXcan to summary statistics from a cIMT genome-wide association study (GWAS) of 71 128 individuals to estimate the association between genetically determined mRNA levels and cIMT and replicated these analyses using S-PrediXcan on an independent GWAS on cIMT that included 22 179 individuals from the UK Biobank. mRNA levels of TNFAIP3, CEBPD and METRNL were inversely associated with cIMT, but these associations were not significant in the replication analysis. S-PrediXcan identified associations between cIMT and genetically determined mRNA levels for 36 genes, of which six were significant in the replication analysis, including TLN2, which had not been previously reported for cIMT. There was weak correlation between our results using differential gene expression analysis and S-PrediXcan. Differential expression analysis and S-PrediXcan represent complementary approaches for the discovery of associations between phenotypes and gene expression. Using these approaches, we prioritize TNFAIP3, CEBPD, METRNL and TLN2 as new candidate genes whose differential expression might modulate cIMT.

Cis-epistasis at the LPA locus and risk of cardiovascular diseases

AIMS

Coronary artery disease (CAD) has a strong genetic predisposition. However, despite substantial discoveries made by genome-wide association studies (GWAS), a large proportion of heritability awaits identification. Non-additive genetic effects might be responsible for part of the unaccounted genetic variance. Here, we attempted a proof-of-concept study to identify non-additive genetic effects, namely epistatic interactions, associated with CAD.

METHODS AND RESULTS

We tested for epistatic interactions in 10 CAD case-control studies and UK Biobank with focus on 8068 SNPs at 56 loci with known associations with CAD risk. We identified a SNP pair located in cis at the LPA locus, rs1800769 and rs9458001, to be jointly associated with risk for CAD [odds ratio (OR) = 1.37, P = 1.07 × 10-11], peripheral arterial disease (OR = 1.22, P = 2.32 × 10-4), aortic stenosis (OR = 1.47, P = 6.95 × 10-7), hepatic lipoprotein(a) (Lp(a)) transcript levels (beta = 0.39, P = 1.41 × 10-8), and Lp(a) serum levels (beta = 0.58, P = 8.7 × 10-32), while individual SNPs displayed no association. Further exploration of the LPA locus revealed a strong dependency of these associations on a rare variant, rs140570886, that was previously associated with Lp(a) levels. We confirmed increased CAD risk for heterozygous (relative OR = 1.46, P = 9.97 × 10-32) and individuals homozygous for the minor allele (relative OR = 1.77, P = 0.09) of rs140570886. Using forward model selection, we also show that epistatic interactions between rs140570886, rs9458001, and rs1800769 modulate the effects of the rs140570886 risk allele.

CONCLUSIONS

These results demonstrate the feasibility of a large-scale knowledge-based epistasis scan and provide rare evidence of an epistatic interaction in a complex human disease. We were directed to a variant (rs140570886) influencing risk through additive genetic as well as epistatic effects. In summary, this study provides deeper insights into the genetic architecture of a locus important for cardiovascular diseases.

Circulating Interleukin-6 Levels and Incident Ischemic Stroke: A Systematic Review and Meta-analysis of Prospective Studies

BACKGROUND AND OBJECTIVES

Human genetic studies support a key role of interleukin-6 (IL-6) in the pathogenesis of ischemic stroke. However, there are only limited data from observational studies exploring circulating IL-6 levels as a risk factor for ischemic stroke. We set out to perform a systematic review and meta-analysis of aggregate data on cohort studies to determine the magnitude and shape of the association between circulating IL-6 levels and risk of incident ischemic stroke in the general population.

METHODS

Following the PRISMA guidelines, we systematically screened the PubMed search engine from inception to March 2021 for population-based prospective cohort studies exploring the association between circulating IL-6 levels and risk of incident ischemic stroke. We pooled association estimates for ischemic stroke risk with random-effects models and explored nonlinear effects in dose-response meta-analyses. Risk of bias was assessed with the Newcastle-Ottawa Scale (NOS). We used funnel plots and trim-to-fill analyses to assess publication bias.

RESULTS

We identified 11 studies (n = 27,411 individuals; 2,669 stroke events) meeting our eligibility criteria. Mean age of all included participants was 60.5 years and 54.8% were female. Overall, quality of the included studies was high (median 8 out of 9 NOS points, interquartile range 7-9). In meta-analyses, 1 SD increment in circulating log-transformed IL-6 levels was associated with a 19% increase in risk of incident ischemic stroke over a mean follow-up of 12.4 years (relative risk 1.19; 95% confidence interval 1.10 to 1.28). A dose-response meta-analysis showed a linear association between circulating IL-6 levels and ischemic stroke risk. There was only moderate heterogeneity and the results were consistent in sensitivity analyses restricted to studies of low risk of bias and studies fully adjusting for demographic and vascular risk factors. The results also remained stable following adjustment for publication bias.

DISCUSSION

Higher circulating IL-6 levels in community-dwelling individuals are associated with higher long-term risk of incident ischemic stroke in a linear pattern and independently of conventional vascular risk factors. Along with findings from genetic studies and clinical trials, these results provide additional support for a key role of IL-6 signaling in ischemic stroke.

Endoscopic grading systems for nasal polyps: are we comparing apples to oranges?

Endoscopic grading of nasal polyps (NP) is typically a coprimary endpoint in clinical trials evaluating treatments for chronic rhinosinusitis with nasal polyps (CRSwNP). However, a consensus on the most effective way to grade nasal polyps has not been reached. Different scales have been used, hampering the interpretation of data across trials. This review compares the characteristics of NP grading systems used in registration trials for approved NP treatments. These fundamental differences in grading systems make quantitative comparison of outcomes between trials inaccurate and potentially misleading. In lieu of a universal grading system, reporting the baseline distribution of polyp grades (unilateral and/or summed/total grades), as well as changes from baseline over time by baseline grade may help improve interpretability of outcomes and reduce inaccuracy when attempting cross-trial comparisons and making therapeutic decisions.

From Atherosclerosis to Myocardial Infarction: A Process-Oriented Model Investigating the Role of Risk Factors

Mathematical models are able to reflect biological processes and to capture epidemiologic data. Thus, they may help elucidate roles of risk factors in disease progression. We propose to account for smoking, hypertension, and dyslipidemia in a previously published process-oriented model that describes the development of atherosclerotic lesions resulting in myocardial infarction (MI). The model is sex-specific and incorporates individual heterogeneity. It was applied to population-based individual risk factors and MI rates (Cooperative Health Research in the Region of Augsburg (KORA) study) together with subclinical atherosclerotic lesion data (Pathobiological Determinants of Atherosclerosis in Youth (PDAY) study). Different model variants were evaluated, testing the association of risk factors with different disease processes. Best fits were obtained for smoking affecting a late-stage disease process, suggesting a thrombogenic role. Hypertension was mainly related to complicated, vulnerable lesions. Dyslipidemia was consistent with increasing the number of initial lesions. By accounting for heterogeneity, individual hazard ratios differ from the population average. The mean individual hazard ratio for smoking was twice the population-based hazard ratio for men and even more for women. Atherosclerotic lesion progression and MI incidence data can be related in a mathematical model to illuminate how risk factors affect different phases of this pathological process.

Whole-Body MRI-Derived Adipose Tissue Characterization and Relationship to Pulmonary Function Impairment

Background: Specification of adipose tissues by whole-body magnetic resonance imaging (MRI) was performed and related to pulmonary function parameters in a population-based cohort. Methods: 203 study participants underwent whole-body MRI and pulmonary function tests as part of the KORA (Cooperative Health Research in the Augsburg Region) MRI study. Both visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) were derived from the T1-Dixon sequence, and hepatic adipose tissue from the proton density fat fraction (PDFFhepatic). Associations between adipose tissue parameters and spirometric indices such as forced vital capacity (FVC), forced expiratory volume in 1 s (FEV1) and Tiffeneau-index (FEV1/FVC) were examined using multivariate linear regression analysis excluding cofounding effects of other clinical parameters. Results: VAT (β = −0.13, p = 0.03) and SAT (β = −0.26, p < 0.001), but not PDFFhepatic were inversely associated with FEV1, while VAT (β = −0.27, p < 0.001), SAT (β = −0.41, p < 0.001), and PDFFhepatic (β = −0.17, p = 0.002) were inversely associated with FVC. PDFFhepatic was directly associated with the Tiffeneau index (β = 2.46, p < 0.001). Conclusions: In the adjusted linear regression model, VAT was inversely associated with all measured spirometric parameters, while PDFFhepatic revealed the strongest association with the Tiffeneau index. Non-invasive adipose tissue quantification measurements might serve as novel biomarkers for respiratory impairment.

Incidence, progression and risk factors of age-related macular degeneration in 35-95-year-old individuals from three jointly designed German cohort studies

Objective

To estimate age-related macular degeneration (AMD) incidence/progression across a wide age range.

Methods and analysis

AMD at baseline and follow-up (colour fundus imaging, Three Continent AMD Consortium Severity Scale, 3CACSS, clinical classification, CC) was assessed for 1513 individuals aged 35–95 years at baseline from three jointly designed population-based cohorts in Germany: Kooperative Gesundheitsforschung in der Region Augsburg (KORA-Fit, KORA-FF4) and Altersbezogene Untersuchungen zur Gesundheit der Universität Regensburg (AugUR) with 18-year, 14-year or 3-year follow-up, respectively. Baseline assessment included lifestyle, metabolic and genetic markers. We derived cumulative estimates, rates and risk factor association for: (1) incident early AMD, (2) incident late AMD among no AMD at baseline (definition 1), (3) incident late AMD among no/early AMD at baseline (definition 2), (4) progression from early to late AMD.

Results

Incidence/progression increased by age, except progression in 70+-year old. We observed 35–55-year-old with 3CACSS-based early AMD who progressed to late AMD. Predominant risk factor for incident late AMD definition 2 was early AMD followed by genetics and smoking. When separating incident late AMD definition 1 from progression (instead of combined as incident late AMD definition 2), estimates help judge an individual’s risk based on age and (3CACSS) early AMD status: for example, for a 65-year old, 3-year late AMD risk with no or early AMD is 0.5% or 7%, 3-year early AMD risk is 3%; for an 85-year old, these numbers are 0.5%, 21%, 12%, respectively. For CC-based ‘early/intermediate’ AMD, incidence was higher, but progression was lower.

Conclusion

We provide a practical guide for AMD risk for ophthalmology practice and healthcare management and document a late AMD risk for individuals aged <55 years.

Differential effects of lung inflammation on insulin resistance in humans and mice

BACKGROUND

The rates of obesity, its associated diseases, and allergies are raising at alarming rates in most countries. House dust mites (HDM) are highly allergenic and exposure often associates with an urban sedentary indoor lifestyle, also resulting in obesity. The aim of this study was to investigate the epidemiological association and physiological impact of lung inflammation on obesity and glucose homeostasis.

METHODS

Epidemiological data from 2207 adults of the population-based KORA FF4 cohort were used to test associations between asthma and rhinitis with metrics of body weight and insulin sensitivity. To obtain functional insights, C57BL/6J mice were intranasally sensitized and challenged with HDM and simultaneously fed with either low-fat or high-fat diet for 12 weeks followed by a detailed metabolic and biochemical phenotyping of the lung, liver, and adipose tissues.

RESULTS

We found a direct association of asthma with insulin resistance but not body weight in humans. In mice, co-development of obesity and HDM-induced lung inflammation attenuated inflammation in lung and perigonadal fat, with little impact on body weight, but small shifts in the composition of gut microbiota. Exposure to HDM improved glucose tolerance, reduced hepatosteatosis, and increased energy expenditure and basal metabolic rate. These effects associate with increased activity of thermogenic adipose tissues independent of uncoupling protein 1.

CONCLUSIONS

Asthma associates with insulin resistance in humans, but HDM challenge results in opposing effects on glucose homeostasis in mice due to increased energy expenditure, reduced adipose inflammation, and hepatosteatosis.

Low emission zones reduced PM10 but not NO2 concentrations in Berlin and Munich, Germany

Low emission zones (LEZs) aiming at improving the air quality in urban areas have been implemented in many European cities. However, studies are limited in evaluating the effects of LEZ, and most of which used simple methods. In this study, a general additive mixed model was utilized to account for confounders in the atmosphere and validate the effects of LEZ on PM₁₀ and NO₂ concentrations in two German cities. In addition, the effects of LEZ on elemental carbon (EC) and total carbon (TC) in Berlin were also evaluated. The LEZ effects were estimated after taking into account air pollutant concentrations at a reference site located in the regional background, and adjusting for hour of the week, public holidays, season, and wind direction. The LEZ in Berlin, and the LEZ in combination with the heavy-duty vehicle (HDV) transit ban in Munich significantly reduced the PM₁₀ concentrations, at both traffic sites (TS) and urban background sites (UB). The effects were greater in LEZ stage 3 than in LEZ stages 2 and 1. Moreover, compared with PM₁₀, LEZ was more efficient in reducing EC, a component that is considered more toxic than PM₁₀ mass. In contrast, the LEZ had no consistent effect on NO₂ levels: no effects were observed in Berlin; in Munich, the combination of the LEZ and the HDV transit ban reduced NO₂ at UB site in LEZ stage 1, but without further reductions in subsequent stages of the LEZ. Overall, our study indicated that LEZs, which target the major primary air pollution source in the highly populated city center could be an effective way to improve urban air quality such as PM mass concentration and EC level.

Epigenetic scores for the circulating proteome as tools for disease prediction

Protein biomarkers have been identified across many age-related morbidities. However, characterising epigenetic influences could further inform disease predictions. Here, we leverage epigenome-wide data to study links between the DNA methylation (DNAm) signatures of the circulating proteome and incident diseases. Using data from four cohorts, we trained and tested epigenetic scores (EpiScores) for 953 plasma proteins, identifying 109 scores that explained between 1% and 58% of the variance in protein levels after adjusting for known protein quantitative trait loci (pQTL) genetic effects. By projecting these EpiScores into an independent sample (Generation Scotland; n = 9537) and relating them to incident morbidities over a follow-up of 14 years, we uncovered 137 EpiScore-disease associations. These associations were largely independent of immune cell proportions, common lifestyle and health factors, and biological aging. Notably, we found that our diabetes-associated EpiScores highlighted previous top biomarker associations from proteome-wide assessments of diabetes. These EpiScores for protein levels can therefore be a valuable resource for disease prediction and risk stratification.

Association of circulating MR-proADM with all-cause and cardiovascular mortality in the general population: Results from the KORA F4 cohort study

Background and aim

Despite its vasodilatory effect, adrenomedullin and its surrogate mid-regional pro-adrenomedullin (MR-proADM) have been found to be positively associated with all-cause and cardiovascular mortality. However, the underlying mechanisms thereof remain unclear and the associations were mostly shown in geriatric cohorts or in patients with chronic diseases. Therefore, we aimed to investigate the possible involvement of abdominal obesity, selected adipokines, and biomarkers of subclinical inflammation in the association of MR-proADM with mortality in a population based study cohort.

Methods

Prospective analysis of the KORA F4 study; median follow-up 9.1 (8.8–9.4) years. Complete data on MR-proADM and mortality was available for 1551 participants, aged 56.9±12.9 years (mean±SD). Correlation and regression analyses of MR-proADM with overall (BMI) and abdominal obesity (waist circumference), selected adipokines and biomarkers of subclinical inflammation. Cox proportional hazard models on the association of MR-proADM with all-cause and cardiovascular mortality with adjustment for cardiovascular risk factors and selected biomarkers in study subgroups (n = 603–1551).

Results

MR-proADM associated with all-cause (HR (95%CI): 2.37 (1.72–3.26) and 2.31 (1.67–3.20)) and cardiovascular mortality (4.28 (2.19–8.39) and 4.44 (2.25–8.76)) after adjustment for traditional cardiovascular risk factors including BMI or waist circumference, respectively. MR-proADM was further associated with four out of seven examined adipokines (leptin, retinol-binding protein-4, chemerin, and adiponectin) and with five out of eleven examined biomarkers of subclinical inflammation (high-sensitivity C-reactive protein, interleukin-6, myeloperoxidase, interleukin-22, and interleukin-1 receptor antagonist) after multivariable adjustment and correction for multiple testing. However, only IL-6 substantially attenuated the association of MR-proADM with all-cause mortality.

Conclusions

We found an association of MR-proADM with (abdominal) obesity, selected adipokines, and biomarkers of subclinical inflammation. However, the association of MR-proADM with mortality was independent of these parameters. Future studies should investigate the role of IL-6 and further characteristics of subclinical inflammation in the association between MR-proADM and all-cause mortality.

TIGER: technical variation elimination for metabolomics data using ensemble learning architecture

Large metabolomics datasets inevitably contain unwanted technical variations which can obscure meaningful biological signals and affect how this information is applied to personalized healthcare. Many methods have been developed to handle unwanted variations. However, the underlying assumptions of many existing methods only hold for a few specific scenarios. Some tools remove technical variations with models trained on quality control (QC) samples which may not generalize well on subject samples. Additionally, almost none of the existing methods supports datasets with multiple types of QC samples, which greatly limits their performance and flexibility. To address these issues, a non-parametric method TIGER (Technical variation elImination with ensemble learninG architEctuRe) is developed in this study and released as an R package (https://CRAN.R-project.org/package=TIGERr). TIGER integrates the random forest algorithm into an adaptable ensemble learning architecture. Evaluation results show that TIGER outperforms four popular methods with respect to robustness and reliability on three human cohort datasets constructed with targeted or untargeted metabolomics data. Additionally, a case study aiming to identify age-associated metabolites is performed to illustrate how TIGER can be used for cross-kit adjustment in a longitudinal analysis with experimental data of three time-points generated by different analytical kits. A dynamic website is developed to help evaluate the performance of TIGER and examine the patterns revealed in our longitudinal analysis (https://han-siyu.github.io/TIGER_web/). Overall, TIGER is expected to be a powerful tool for metabolomics data analysis.

Concept of the Munich/Augsburg Consortium Precision in Mental Health for the German Center of Mental Health

The Federal Ministry of Education and Research (BMBF) issued a call for a new nationwide research network on mental disorders, the German Center of Mental Health (DZPG). The Munich/Augsburg consortium was selected to participate as one of six partner sites with its concept "Precision in Mental Health (PriMe): Understanding, predicting, and preventing chronicity." PriMe bundles interdisciplinary research from the Ludwig-Maximilians-University (LMU), Technical University of Munich (TUM), University of Augsburg (UniA), Helmholtz Center Munich (HMGU), and Max Planck Institute of Psychiatry (MPIP) and has a focus on schizophrenia (SZ), bipolar disorder (BPD), and major depressive disorder (MDD). PriMe takes a longitudinal perspective on these three disorders from the at-risk stage to the first-episode, relapsing, and chronic stages. These disorders pose a major health burden because in up to 50% of patients they cause untreatable residual symptoms, which lead to early social and vocational disability, comorbidities, and excess mortality. PriMe aims at reducing mortality on different levels, e.g., reducing death by psychiatric and somatic comorbidities, and will approach this goal by addressing interdisciplinary and cross-sector approaches across the lifespan. PriMe aims to add a precision medicine framework to the DZPG that will propel deeper understanding, more accurate prediction, and personalized prevention to prevent disease chronicity and mortality across mental illnesses. This framework is structured along the translational chain and will be used by PriMe to innovate the preventive and therapeutic management of SZ, BPD, and MDD from rural to urban areas and from patients in early disease stages to patients with long-term disease courses. Research will build on platforms that include one on model systems, one on the identification and validation of predictive markers, one on the development of novel multimodal treatments, one on the regulation and strengthening of the uptake and dissemination of personalized treatments, and finally one on testing of the clinical effectiveness, utility, and scalability of such personalized treatments. In accordance with the translational chain, PriMe's expertise includes the ability to integrate understanding of bio-behavioral processes based on innovative models, to translate this knowledge into clinical practice and to promote user participation in mental health research and care.

Association of renin and aldosterone with glucose metabolism in a Western European population: the KORA F4/FF4 study

INTRODUCTION

Primary aldosteronism is associated with impaired glucose tolerance. Whether plasma aldosterone and/or renin concentrations are associated with type 2 diabetes and continuous measures of glucose metabolism in the general population is still under debate.

RESEARCH DESIGN AND METHODS

The analyses included 2931 participants of the KORA F4 study at baseline and 2010 participants of the KORA FF4 study after a median follow-up of 6.5 years. The associations of active plasma renin and aldosterone concentrations with type 2 diabetes and continuous measures of glucose metabolism were assessed using logistic and linear regression models. Results were adjusted for sex, age, body mass index (BMI), estimated glomerular filtration rate, potassium, use of ACE inhibitors, angiotensin receptor blockers, beta blockers, diuretics and calcium channel blockers.

RESULTS

Cross-sectionally, renin was associated with type 2 diabetes (OR per SD: 1.25, 95% CI 1.10 to 1.43, p<0.001), fasting glucose, 2-hour glucose, insulin, proinsulin, HOMA-B (homeostasis model assessment of beta cell function) and HOMA-IR (homeostasis model assessment of insulin resistance) (all p values <0.001). Aldosterone was not associated with type 2 diabetes (OR: 1.04, 95% CI 0.91 to 1.19; p=0.547) but with insulin, proinsulin and HOMA-IR (all p values <0.001). The aldosterone-renin ratio was inversely associated with type 2 diabetes and several measures of glucose metabolism. Longitudinally, neither renin (OR: 1.12, 95% CI 0.92 to 1.36) nor aldosterone (OR: 0.91, 95% CI 0.74 to 1.11) were associated with incident type 2 diabetes. Renin was inversely associated with changes of insulin concentrations.

CONCLUSIONS

In the KORA F4/FF4 study, renin and aldosterone were not associated with incident type 2 diabetes and largely unrelated to changes of measures of glucose metabolism. Cross-sectionally, aldosterone was associated with surrogate parameters of insulin resistance. However, these associations were not independent of renin.

Genetic variation influencing DNA methylation provides insights into molecular mechanisms regulating genomic function

We determined the relationships between DNA sequence variation and DNA methylation using blood samples from 3,799 Europeans and 3,195 South Asians. We identify 11,165,559 SNP-CpG associations (methylation quantitative trait loci (meQTL), P < 10-14), including 467,915 meQTL that operate in trans. The meQTL are enriched for functionally relevant characteristics, including shared chromatin state, High-throuhgput chromosome conformation interaction, and association with gene expression, metabolic variation and clinical traits. We use molecular interaction and colocalization analyses to identify multiple nuclear regulatory pathways linking meQTL loci to phenotypic variation, including UBASH3B (body mass index), NFKBIE (rheumatoid arthritis), MGA (blood pressure) and COMMD7 (white cell counts). For rs6511961 , chromatin immunoprecipitation followed by sequencing (ChIP-seq) validates zinc finger protein (ZNF)333 as the likely trans acting effector protein. Finally, we used interaction analyses to identify population- and lineage-specific meQTL, including rs174548 in FADS1, with the strongest effect in CD8+ T cells, thus linking fatty acid metabolism with immune dysregulation and asthma. Our study advances understanding of the potential pathways linking genetic variation to human phenotype.

Associations of endogenous androgens and sex hormone-binding globulin with kidney function and chronic kidney disease

INTRODUCTION

The role of endogenous androgens in kidney function and disease has not been extensively explored in men and women.

RESEARCH DESIGN AND METHODS

We analyzed data from the observational KORA F4 study and its follow-up examination KORA FF4 (median follow-up time 6.5 years) including 1293 men and 650 peri- and postmenopausal women, not using exogenous sex hormones. We examined the associations between endogenous androgens (testosterone [T], dihydrotestosterone [DHT], free T [fT], free DHT [fDHT], and T/DHT), with estimated glomerular filtration rate (eGFR) at baseline and follow-up, prevalent, and incident chronic kidney disease (CKD) adjusting for common CKD risk factors.

RESULTS

At baseline, 73 men (5.7%) and 54 women (8.4%) had prevalent CKD. Cross-sectionally, no significant associations between androgens and kidney function were observed among men. In women, elevated T (β=-1.305, [95% CI -2.290; -0.320]) and fT (β=-1.423, [95% CI -2.449; -0.397]) were associated with lower eGFR. Prospectively, 81 men (8.8%) and 60 women (15.2%) developed incident CKD. In women, a reverse J-shaped associations was observed between DHT and incident CKD (P_non-linear=0.029), while higher fDHT was associated with lower incident CKD risk (odds ratio per 1 standard deviation=0.613, [95% CI 0.369; 0.971]. Among men, T/DHT (β=-0.819, [95% CI -1.413; -0.226]) and SHBG (P_non-linear=0.011) were associated with eGFR at follow-up but not with incident CKD. Some associations appeared to be modified by type 2 diabetes (T2D).

CONCLUSION

Suggestive associations are observed of androgens and SHBG with kidney impairment among men and women. However, larger well-phenotyped prospective studies are required to further elucidate the potential of androgens, SHBG, and T2D as modifiable risk factors for kidney function and CKD.

Long-term exposure to low ambient air pollution concentrations and mortality among 28 million people: results from seven large European cohorts within the ELAPSE project

BACKGROUND

Long-term exposure to ambient air pollution has been associated with premature mortality, but associations at concentrations lower than current annual limit values are uncertain. We analysed associations between low-level air pollution and mortality within the multicentre study Effects of Low-Level Air Pollution: A Study in Europe (ELAPSE).

METHODS

In this multicentre longitudinal study, we analysed seven population-based cohorts of adults (age ≥30 years) within ELAPSE, from Belgium, Denmark, England, the Netherlands, Norway, Rome (Italy), and Switzerland (enrolled in 2000-11; follow-up until 2011-17). Mortality registries were used to extract the underlying cause of death for deceased individuals. Annual average concentrations of fine particulate matter (PM_2·5), nitrogen dioxide (NO₂), black carbon, and tropospheric warm-season ozone (O₃) from Europe-wide land use regression models at 100 m spatial resolution were assigned to baseline residential addresses. We applied cohort-specific Cox proportional hazard models with adjustment for area-level and individual-level covariates to evaluate associations with non-accidental mortality, as the main outcome, and with cardiovascular, non-malignant respiratory, and lung cancer mortality. Subset analyses of participants living at low pollutant concentrations (as per predefined values) and natural splines were used to investigate the concentration-response function. Cohort-specific effect estimates were pooled in a random-effects meta-analysis.

FINDINGS

We analysed 28 153 138 participants contributing 257 859 621 person-years of observation, during which 3 593 741 deaths from non-accidental causes occurred. We found significant positive associations between non-accidental mortality and PM_2·5, NO₂, and black carbon, with a hazard ratio (HR) of 1·053 (95% CI 1·021-1·085) per 5 μg/m³ increment in PM_2·5, 1·044 (1·019-1·069) per 10 μg/m³ NO₂, and 1·039 (1·018-1·059) per 0·5 × 10^-5/m black carbon. Associations with PM_2·5, NO₂, and black carbon were slightly weaker for cardiovascular mortality, similar for non-malignant respiratory mortality, and stronger for lung cancer mortality. Warm-season O₃ was negatively associated with both non-accidental and cause-specific mortality. Associations were stronger at low concentrations: HRs for non-accidental mortality at concentrations lower than the WHO 2005 air quality guideline values for PM_2·5 (10 μg/m³) and NO₂ (40 μg/m³) were 1·078 (1·046-1·111) per 5 μg/m³ PM_2·5 and 1·049 (1·024-1·075) per 10 μg/m³ NO₂. Similarly, the association between black carbon and non-accidental mortality was highest at low concentrations, with a HR of 1·061 (1·032-1·092) for exposure lower than 1·5× 10^-5/m, and 1·081 (0·966-1·210) for exposure lower than 1·0× 10^-5/m.

INTERPRETATION

Long-term exposure to concentrations of PM_2·5 and NO₂ lower than current annual limit values was associated with non-accidental, cardiovascular, non-malignant respiratory, and lung cancer mortality in seven large European cohorts. Continuing research on the effects of low concentrations of air pollutants is expected to further inform the process of setting air quality standards in Europe and other global regions.

FUNDING

Health Effects Institute.

Association of serum uromodulin with adipokines in dependence of type 2 diabetes

BACKGROUND

The renal tubular glycoprotein uromodulin is associated with obesity and type 2 diabetes, but the underlying mechanisms are elusive. We investigated the association of serum uromodulin with adipokines and tested the effect modification by diabetes status.

METHODS

The associations of serum uromodulin with eight adipokines were assessed in 795-1080 participants of the KORA F4 study aged 62-81 years using linear regression models adjusted for sex, age, BMI, estimated glomerular filtration rate and diabetes. Significant associations were assessed for effect modification by diabetes status. We further tested using logistic regression whether adjustment for the significant adipokines affected the association of uromodulin with type 2 diabetes.

RESULTS

Serum uromodulin was inversely associated with chemerin and retinol-binding protein-4 after multivariable adjustment (p < 0.001) and Bonferroni correction for multiple testing. No significant association was observed between uromodulin and the other adipokines (leptin, adiponectin, secreted frizzled-related protein 5, progranulin, omentin-1 and vaspin) after correcting for multiple testing. The association of uromodulin with chemerin and retinol-binding protein-4 was stronger in participants with type 2 diabetes than in participants without diabetes (p for interaction < 0.05). However, inclusion of chemerin and retinol-binding protein-4 in logistic regression models did not attenuate the association of serum uromodulin with diabetes.

CONCLUSIONS

Serum uromodulin was inversely associated with the predominantly pro-inflammatory adipokines chemerin and retinol-binding protein-4. The associations were stronger in participants with type 2 diabetes compared to participants without diabetes. However, the association of serum uromodulin with type 2 diabetes was independent of chemerin and retinol-binding protein-4.

Epigenome-wide association study of serum urate reveals insights into urate co-regulation and the SLC2A9 locus

Elevated serum urate levels, a complex trait and major risk factor for incident gout, are correlated with cardiometabolic traits via incompletely understood mechanisms. DNA methylation in whole blood captures genetic and environmental influences and is assessed in transethnic meta-analysis of epigenome-wide association studies (EWAS) of serum urate (discovery, n = 12,474, replication, n = 5522). The 100 replicated, epigenome-wide significant (p < 1.1E-7) CpGs explain 11.6% of the serum urate variance. At SLC2A9, the serum urate locus with the largest effect in genome-wide association studies (GWAS), five CpGs are associated with SLC2A9 gene expression. Four CpGs at SLC2A9 have significant causal effects on serum urate levels and/or gout, and two of these partly mediate the effects of urate-associated GWAS variants. In other genes, including SLC7A11 and PHGDH, 17 urate-associated CpGs are associated with conditions defining metabolic syndrome, suggesting that these CpGs may represent a blood DNA methylation signature of cardiometabolic risk factors. This study demonstrates that EWAS can provide new insights into GWAS loci and the correlation of serum urate with other complex traits.

Meta-analyses identify DNA methylation associated with kidney function and damage

Chronic kidney disease is a major public health burden. Elevated urinary albumin-to-creatinine ratio is a measure of kidney damage, and used to diagnose and stage chronic kidney disease. To extend the knowledge on regulatory mechanisms related to kidney function and disease, we conducted a blood-based epigenome-wide association study for estimated glomerular filtration rate (n = 33,605) and urinary albumin-to-creatinine ratio (n = 15,068) and detected 69 and seven CpG sites where DNA methylation was associated with the respective trait. The majority of these findings showed directionally consistent associations with the respective clinical outcomes chronic kidney disease and moderately increased albuminuria. Associations of DNA methylation with kidney function, such as CpGs at JAZF1, PELI1 and CHD2 were validated in kidney tissue. Methylation at PHRF1, LDB2, CSRNP1 and IRF5 indicated causal effects on kidney function. Enrichment analyses revealed pathways related to hemostasis and blood cell migration for estimated glomerular filtration rate, and immune cell activation and response for urinary albumin-to-creatinineratio-associated CpGs.

N-glycosylation of immunoglobulin G predicts incident hypertension

OBJECTIVES

Glycosylation of immunoglobulin G (IgG) is an important regulator of the immune system and has been implicated in prevalent hypertension. The aim of this study is to investigate whether the IgG glycome begins to change prior to hypertension diagnosis by analysing the IgG glycome composition in a large population-based female cohort with two independent replication samples.

METHODS

We included 989 unrelated cases with incident hypertension and 1628 controls from the TwinsUK cohort (mean follow-up time of 6.3 years) with IgG measured at baseline by ultra-performance liquid chromatography and longitudinal BP measurement available. We replicated our findings in 106 individuals from the 10 001 Dalmatians and 729 from KORA S4. Cox regression mixed models were applied to identify changes in glycan traits preincident hypertension, after adjusting for age, mean arterial pressure, BMI, family relatedness and multiple testing (FDR < 0.1). Significant IgG-incident hypertension associations were replicated in the two independent cohorts by leveraging Cox regression mixed models in the 10 001 Dalmatians and logistic regression models in the KORA cohort.

RESULTS

We identified and replicated four glycan traits, incidence of bisecting GlcNAc, GP4, GP9 and GP21, that are predictive of incident hypertension after adjusting for confoundes and multiple testing [hazard ratio (95% CI) ranging from 0.45 (0.24-0.84) for GP21 to 2.9 (1.5-5.68) for GP4]. We then linearly combined the four replicated glycans and found that the glycan score correlated with incident hypertension, SBP and DBP.

CONCLUSION

Our results suggest that the IgG glycome changes prior to the development of hypertension.

The power of genetic diversity in genome-wide association studies of lipids

Increased blood lipid levels are heritable risk factors of cardiovascular disease with varied prevalence worldwide owing to different dietary patterns and medication use1. Despite advances in prevention and treatment, in particular through reducing low-density lipoprotein cholesterol levels2, heart disease remains the leading cause of death worldwide3. Genome-wideassociation studies (GWAS) of blood lipid levels have led to important biological and clinical insights, as well as new drug targets, for cardiovascular disease. However, most previous GWAS4-23 have been conducted in European ancestry populations and may have missed genetic variants that contribute to lipid-level variation in other ancestry groups. These include differences in allele frequencies, effect sizes and linkage-disequilibrium patterns24. Here we conduct a multi-ancestry, genome-wide genetic discovery meta-analysis of lipid levels in approximately 1.65 million individuals, including 350,000 of non-European ancestries. We quantify the gain in studying non-European ancestries and provide evidence to support the expansion of recruitment of additional ancestries, even with relatively small sample sizes. We find that increasing diversity rather than studying additional individuals of European ancestry results in substantial improvements in fine-mapping functional variants and portability of polygenic prediction (evaluated in approximately 295,000 individuals from 7 ancestry groupings). Modest gains in the number of discovered loci and ancestry-specific variants were also achieved. As GWAS expand emphasis beyond the identification of genes and fundamental biology towards the use of genetic variants for preventive and precision medicine25, we anticipate that increased diversity of participants will lead to more accurate and equitable26 application of polygenic scores in clinical practice.

Rare variant analysis in eczema identifies exonic variants in DUSP1, NOTCH4 and SLC9A4

Previous genome-wide association studies revealed multiple common variants involved in eczema but the role of rare variants remains to be elucidated. Here, we investigate the role of rare variants in eczema susceptibility. We meta-analyze 21 study populations including 20,016 eczema cases and 380,433 controls. Rare variants are imputed with high accuracy using large population-based reference panels. We identify rare exonic variants in DUSP1, NOTCH4, and SLC9A4 to be associated with eczema. In DUSP1 and NOTCH4 missense variants are predicted to impact conserved functional domains. In addition, five novel common variants at SATB1-AS1/KCNH8, TRIB1/LINC00861, ZBTB1, TBX21/OSBPL7, and CSF2RB are discovered. While genes prioritized based on rare variants are significantly up-regulated in the skin, common variants point to immune cell function. Over 20% of the single nucleotide variant-based heritability is attributable to rare and low-frequency variants. The identified rare/low-frequency variants located in functional protein domains point to promising targets for novel therapeutic approaches to eczema.

Genetic studies of eczema to date have mostly explored common genetic variation. Here, the authors perform a large meta-analysis for common and rare variants and discover 8 loci associated with eczema. Over 20% of the heritability of the condition is attributable to rare variants.

Association of glycated hemoglobin A1c levels with cardiovascular outcomes in the general population: results from the BiomarCaRE (Biomarker for Cardiovascular Risk Assessment in Europe) consortium

BACKGROUND

Biomarkers may contribute to improved cardiovascular risk estimation. Glycated hemoglobin A1c (HbA1c) is used to monitor the quality of diabetes treatment. Its strength of association with cardiovascular outcomes in the general population remains uncertain. This study aims to assess the association of HbA1c with cardiovascular outcomes in the general population.

METHODS

Data from six prospective population-based cohort studies across Europe comprising 36,180 participants were analyzed. HbA1c was evaluated in conjunction with classical cardiovascular risk factors (CVRFs) for association with cardiovascular mortality, cardiovascular disease (CVD) incidence, and overall mortality in subjects without diabetes (N = 32,496) and with diabetes (N = 3684).

RESULTS

Kaplan-Meier curves showed higher event rates with increasing HbA1c levels (log-rank-test: p < 0.001). Cox regression analysis revealed significant associations between HbA1c (in mmol/mol) in the total study population and the examined outcomes. Thus, a hazard ratio (HR) of 1.16 (95% confidence interval (CI) 1.02-1.31, p = 0.02) for cardiovascular mortality, 1.13 (95% CI 1.03-1.24, p = 0.01) for CVD incidence, and 1.09 (95% CI 1.02-1.17, p = 0.01) for overall mortality was observed per 10 mmol/mol increase in HbA1c. The association with CVD incidence and overall mortality was also observed in study participants without diabetes with increased HbA1c levels (HR 1.12; 95% CI 1.01-1.25, p = 0.04) and HR 1.10; 95% CI 1.01-1.20, p = 0.02) respectively. HbA1c cut-off values of 39.9 mmol/mol (5.8%), 36.6 mmol/mol (5.5%), and 38.8 mmol/mol (5.7%) for cardiovascular mortality, CVD incidence, and overall mortality, showed also an increased risk.

CONCLUSIONS

HbA1c is independently associated with cardiovascular mortality, overall mortality and cardiovascular disease in the general European population. A mostly monotonically increasing relationship was observed between HbA1c levels and outcomes. Elevated HbA1c levels were associated with cardiovascular disease incidence and overall mortality in participants without diabetes underlining the importance of HbA1c levels in the overall population.

Educational Level, but Not Income or Area Deprivation, is Related to Macrovascular Disease: Results From Two Population-Based Cohorts in Germany

Objectives: An inverse relationship between education and cardiovascular risk has been described, however, the combined association of education, income, and neighborhood socioeconomic status with macrovascular disease is less clear. The aim of this study was to evaluate the association of educational level, equivalent household income and area deprivation with macrovascular disease in Germany.

Methods: Cross-sectional data from two representative German population-based studies, SHIP-TREND (n = 3,731) and KORA-F4 (n = 2,870), were analyzed. Multivariable logistic regression models were applied to estimate odds ratios and 95% confidence intervals for the association between socioeconomic determinants and macrovascular disease (defined as self-reported myocardial infarction or stroke).

Results: The study showed a higher odds of prevalent macrovascular disease in men with low and middle educational level compared to men with high education. Area deprivation and equivalent income were not related to myocardial infarction or stroke in any of the models.

Conclusion: Educational level, but not income or area deprivation, is significantly related to the macrovascular disease in men. Effective prevention of macrovascular disease should therefore start with investing in individual education.

Multi-ancestry genome-wide gene-sleep interactions identify novel loci for blood pressure

Long and short sleep duration are associated with elevated blood pressure (BP), possibly through effects on molecular pathways that influence neuroendocrine and vascular systems. To gain new insights into the genetic basis of sleep-related BP variation, we performed genome-wide gene by short or long sleep duration interaction analyses on four BP traits (systolic BP, diastolic BP, mean arterial pressure, and pulse pressure) across five ancestry groups in two stages using 2 degree of freedom (df) joint test followed by 1df test of interaction effects. Primary multi-ancestry analysis in 62,969 individuals in stage 1 identified three novel gene by sleep interactions that were replicated in an additional 59,296 individuals in stage 2 (stage 1 + 2 Pjoint < 5 × 10-8), including rs7955964 (FIGNL2/ANKRD33) that increases BP among long sleepers, and rs73493041 (SNORA26/C9orf170) and rs10406644 (KCTD15/LSM14A) that increase BP among short sleepers (Pint < 5 × 10-8). Secondary ancestry-specific analysis identified another novel gene by long sleep interaction at rs111887471 (TRPC3/KIAA1109) in individuals of African ancestry (Pint = 2 × 10-6). Combined stage 1 and 2 analyses additionally identified significant gene by long sleep interactions at 10 loci including MKLN1 and RGL3/ELAVL3 previously associated with BP, and significant gene by short sleep interactions at 10 loci including C2orf43 previously associated with BP (Pint < 10-3). 2df test also identified novel loci for BP after modeling sleep that has known functions in sleep-wake regulation, nervous and cardiometabolic systems. This study indicates that sleep and primary mechanisms regulating BP may interact to elevate BP level, suggesting novel insights into sleep-related BP regulation.

Meta-analysis of epigenome-wide association studies of carotid intima-media thickness

Common carotid intima-media thickness (cIMT) is an index of subclinical atherosclerosis that is associated with ischemic stroke and coronary artery disease (CAD). We undertook a cross-sectional epigenome-wide association study (EWAS) of measures of cIMT in 6400 individuals. Mendelian randomization analysis was applied to investigate the potential causal role of DNA methylation in the link between atherosclerotic cardiovascular risk factors and cIMT or clinical cardiovascular disease. The CpG site cg05575921 was associated with cIMT (beta = -0.0264, p value = 3.5 × 10^-8) in the discovery panel and was replicated in replication panel (beta = -0.07, p value = 0.005). This CpG is located at chr5:81649347 in the intron 3 of the aryl hydrocarbon receptor repressor gene (AHRR). Our results indicate that DNA methylation at cg05575921 might be in the pathway between smoking, cIMT and stroke. Moreover, in a region-based analysis, 34 differentially methylated regions (DMRs) were identified of which a DMR upstream of ALOX12 showed the strongest association with cIMT (p value = 1.4 × 10^-13). In conclusion, our study suggests that DNA methylation may play a role in the link between cardiovascular risk factors, cIMT and clinical cardiovascular disease.

Natriuretic Peptides and Risk of Type 2 Diabetes: Results From the Biomarkers for Cardiovascular Risk Assessment in Europe (BiomarCaRE) Consortium

OBJECTIVE

Natriuretic peptide (NP) concentrations are increased in cardiovascular diseases (CVDs) but are associated with a lower diabetes risk. We investigated associations of N-terminal pro-B-type NP (NT-proBNP) and midregional proatrial NP (MR-proANP) with incident type 2 diabetes stratified by the presence of CVD.

RESEARCH DESIGN AND METHODS

Based on the Biomarkers for Cardiovascular Risk Assessment in Europe (BiomarCaRE) Consortium, we included 45,477 participants with NT-proBNP measurements (1,707 developed type 2 diabetes over 6.5 years of median follow-up; among these, 209 had CVD at baseline) and 11,537 participants with MR-proANP measurements (857 developed type 2 diabetes over 13.8 years of median follow-up; among these, 106 had CVD at baseline). The associations were estimated using multivariable Cox regression models.

RESULTS

Both NPs were inversely associated with incident type 2 diabetes (hazard ratios [95% CI] per 1-SD increase of log NP: 0.84 [0.79; 0.89] for NT-proBNP and 0.77 [0.71; 0.83] for MR-proANP). The inverse association between NT-proBNP and type 2 diabetes was significant in individuals without CVD but not in individuals with CVD (0.81 [0.76; 0.86] vs. 1.04 [0.90; 1.19]; P multiplicative interaction = 0.001). There was no significant difference in the association of MR-proANP with type 2 diabetes between individuals without and with CVD (0.75 [0.69; 0.82] vs. 0.81 [0.66; 0.99]; P multiplicative interaction = 0.236).

CONCLUSIONS

NT-proBNP and MR-proANP are inversely associated with incident type 2 diabetes. However, the inverse association of NT-proBNP seems to be modified by the presence of CVD. Further investigations are warranted to confirm our findings and to investigate the underlying mechanisms.

Validating biomarkers and models for epigenetic inference of alcohol consumption from blood

Background

Information on long-term alcohol consumption is relevant for medical and public health research, disease therapy, and other areas. Recently, DNA methylation-based inference of alcohol consumption from blood was reported with high accuracy, but these results were based on employing the same dataset for model training and testing, which can lead to accuracy overestimation. Moreover, only subsets of alcohol consumption categories were used, which makes it impossible to extrapolate such models to the general population. By using data from eight population-based European cohorts (N = 4677), we internally and externally validated the previously reported biomarkers and models for epigenetic inference of alcohol consumption from blood and developed new models comprising all data from all categories.

Results

By employing data from six European cohorts (N = 2883), we empirically tested the reproducibility of the previously suggested biomarkers and prediction models via ten-fold internal cross-validation. In contrast to previous findings, all seven models based on 144-CpGs yielded lower mean AUCs compared to the models with less CpGs. For instance, the 144-CpG heavy versus non-drinkers model gave an AUC of 0.78 ± 0.06, while the 5 and 23 CpG models achieved 0.83 ± 0.05, respectively. The transportability of the models was empirically tested via external validation in three independent European cohorts (N = 1794), revealing high AUC variance between datasets within models. For instance, the 144-CpG heavy versus non-drinkers model yielded AUCs ranging from 0.60 to 0.84 between datasets. The newly developed models that considered data from all categories showed low AUCs but gave low AUC variation in the external validation. For instance, the 144-CpG heavy and at-risk versus light and non-drinkers model achieved AUCs of 0.67 ± 0.02 in the internal cross-validation and 0.61–0.66 in the external validation datasets.

Conclusions

The outcomes of our internal and external validation demonstrate that the previously reported prediction models suffer from both overfitting and accuracy overestimation. Our results show that the previously proposed biomarkers are not yet sufficient for accurate and robust inference of alcohol consumption from blood. Overall, our findings imply that DNA methylation prediction biomarkers and models need to be improved considerably before epigenetic inference of alcohol consumption from blood can be considered for practical applications.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13148-021-01186-3.