Trajectories of clinical characteristics, complications and treatment choices in data-driven subgroups of type 2 diabetes.

AIMS/HYPOTHESIS: This study aimed to explore the added value of subgroups that categorise individuals with type 2 diabetes by k-means clustering for two primary care registries (the Netherlands and Scotland), inspired by Ahlqvist's novel diabetes subgroups and previously analysed by Slieker et al. METHODS: We used two Dutch and Scottish diabetes cohorts (N=3054 and 6145; median follow-up=11.2 and 12.3 years, respectively) and defined five subgroups by k-means clustering with age at baseline, BMI, HbA1c, HDL-cholesterol and C-peptide. We investigated differences between subgroups by trajectories of risk factor values (random intercept models), time to diabetes-related complications (logrank tests and Cox models) and medication patterns (multinomial logistic models). We also compared directly using the clustering indicators as predictors of progression vs the k-means discrete subgroups. Cluster consistency over follow-up was assessed. RESULTS: Subgroups' risk factors were significantly different, and these differences remained generally consistent over follow-up. Among all subgroups, individuals with severe insulin resistance faced a significantly higher risk of myocardial infarction both before (HR 1.65; 95% CI 1.40, 1.94) and after adjusting for age effect (HR 1.72; 95% CI 1.46, 2.02) compared with mild diabetes with high HDL-cholesterol. Individuals with severe insulin-deficient diabetes were most intensively treated, with more than 25% prescribed insulin at 10 years of diagnosis. For severe insulin-deficient diabetes relative to mild diabetes, the relative risks for using insulin relative to no common treatment would be expected to increase by a factor of 3.07 (95% CI 2.73, 3.44), holding other factors constant. Clustering indicators were better predictors of progression variation relative to subgroups, but prediction accuracy may improve after combining both. Clusters were consistent over 8 years with an accuracy ranging from 59% to 72%. CONCLUSIONS/INTERPRETATION: Data-driven subgroup allocations were generally consistent over follow-up and captured significant differences in risk factor trajectories, medication patterns and complication risks. Subgroups serve better as a complement rather than as a basis for compressing clustering indicators.

An omics-based machine learning approach to predict diabetes progression: a RHAPSODY study.

AIMS/HYPOTHESIS: People with type 2 diabetes are heterogeneous in their disease trajectory, with some progressing more quickly to insulin initiation than others. Although classical biomarkers such as age, HbA1c and diabetes duration are associated with glycaemic progression, it is unclear how well such variables predict insulin initiation or requirement and whether newly identified markers have added predictive value. METHODS: In two prospective cohort studies as part of IMI-RHAPSODY, we investigated whether clinical variables and three types of molecular markers (metabolites, lipids, proteins) can predict time to insulin requirement using different machine learning approaches (lasso, ridge, GRridge, random forest). Clinical variables included age, sex, HbA1c, HDL-cholesterol and C-peptide. Models were run with unpenalised clinical variables (i.e. always included in the model without weights) or penalised clinical variables, or without clinical variables. Model development was performed in one cohort and the model was applied in a second cohort. Model performance was evaluated using Harrel's C statistic. RESULTS: Of the 585 individuals from the Hoorn Diabetes Care System (DCS) cohort, 69 required insulin during follow-up (1.0-11.4 years); of the 571 individuals in the Genetics of Diabetes Audit and Research in Tayside Scotland (GoDARTS) cohort, 175 required insulin during follow-up (0.3-11.8 years). Overall, the clinical variables and proteins were selected in the different models most often, followed by the metabolites. The most frequently selected clinical variables were HbA1c (18 of the 36 models, 50%), age (15 models, 41.2%) and C-peptide (15 models, 41.2%). Base models (age, sex, BMI, HbA1c) including only clinical variables performed moderately in both the DCS discovery cohort (C statistic 0.71 [95% CI 0.64, 0.79]) and the GoDARTS replication cohort (C 0.71 [95% CI 0.69, 0.75]). A more extensive model including HDL-cholesterol and C-peptide performed better in both cohorts (DCS, C 0.74 [95% CI 0.67, 0.81]; GoDARTS, C 0.73 [95% CI 0.69, 0.77]). Two proteins, lactadherin and proto-oncogene tyrosine-protein kinase receptor, were most consistently selected and slightly improved model performance. CONCLUSIONS/INTERPRETATION: Using machine learning approaches, we show that insulin requirement risk can be modestly well predicted by predominantly clinical variables. Inclusion of molecular markers improves the prognostic performance beyond that of clinical variables by up to 5%. Such prognostic models could be useful for identifying people with diabetes at high risk of progressing quickly to treatment intensification. DATA AVAILABILITY: Summary statistics of lipidomic, proteomic and metabolomic data are available from a Shiny dashboard at https://rhapdata-app.vital-it.ch .

Circulating metabolomic markers linking diabetic kidney disease and incident cardiovascular disease in type 2 diabetes: analyses from the Hong Kong Diabetes Biobank.

AIMS/HYPOTHESIS: The aim of this study was to describe the metabolome in diabetic kidney disease (DKD) and its association with incident CVD in type 2 diabetes, and identify prognostic biomarkers. METHODS: From a prospective cohort of individuals with type 2 diabetes, baseline sera (N=1991) were quantified for 170 metabolites using NMR spectroscopy with median 5.2 years of follow-up. Associations of chronic kidney disease (CKD, eGFR<60 ml/min per 1.73 m2) or severely increased albuminuria with each metabolite were examined using linear regression, adjusted for confounders and multiplicity. Associations between DKD (CKD or severely increased albuminuria)-related metabolites and incident CVD were examined using Cox regressions. Metabolomic biomarkers were identified and assessed for CVD prediction and replicated in two independent cohorts. RESULTS: At false discovery rate (FDR)<0.05, 156 metabolites were associated with DKD (151 for CKD and 128 for severely increased albuminuria), including apolipoprotein B-containing lipoproteins, HDL, fatty acids, phenylalanine, tyrosine, albumin and glycoprotein acetyls. Over 5.2 years of follow-up, 75 metabolites were associated with incident CVD at FDR<0.05. A model comprising age, sex and three metabolites (albumin, triglycerides in large HDL and phospholipids in small LDL) performed comparably to conventional risk factors (C statistic 0.765 vs 0.762, p=0.893) and adding the three metabolites further improved CVD prediction (C statistic from 0.762 to 0.797, p=0.014) and improved discrimination and reclassification. The 3-metabolite score was validated in independent Chinese and Dutch cohorts. CONCLUSIONS/INTERPRETATION: Altered metabolomic signatures in DKD are associated with incident CVD and improve CVD risk stratification.

Circulating small non-coding RNAs are associated with the insulin-resistant and obesity-related type 2 diabetes clusters.

AIM: To uncover differences in small non-coding RNAs (sncRNAs) in individuals with type 2 diabetes (T2D) categorized into five clusters based on individual characteristics, which may aid in the identification of those prone to rapid progression. MATERIALS AND METHODS: In the Hoorn Diabetes Care System (DCS) cohort, participants were clustered by age, body mass index (BMI), and glycated haemoglobin, C-peptide and high-density lipoprotein (HDL) cholesterol levels, yielding severe insulin-deficient diabetes, severe insulin-resistant diabetes (SIRD), mild obesity-related diabetes (MOD), mild diabetes, and mild diabetes with high HDL cholesterol clusters (n = 412). Utilizing plasma sncRNA-sequencing, we identified distinct cluster-specific sncRNAs. Validation was performed in a smaller DCS Hoorn dataset (n = 138). To elucidate their potential functions, we examined tissue expression, identified potential targets or (co-)regulated proteins, conducted gene set enrichment analyses on the targets through Reactome, and examined tissue expression of the (co-)regulated proteins. RESULTS: The insulin-resistant cluster exhibited aberrant expression of 10 sncRNAs, while the high BMI cluster featured eight differentially expressed sncRNAs. Multiple (co-)regulated proteins were identified for sncRNAs associated with both clusters. Proteins associated with both clusters showed enrichment for metabolism. Proteins that specifically and only associated with the SIRD cluster showed enrichment for immune-related signalling. Furthermore, MOD cluster-specific associated proteins showed enrichment for the complement system. CONCLUSIONS: Our research showed differential sncRNA levels among type 2 diabetes clusters. This may reflect and could deepen our understanding of molecular mechanisms, in development, progression, and risk factors for each cluster.

Genomics Research of Lifetime Depression in the Netherlands: The BIObanks Netherlands Internet Collaboration (BIONIC) Project.

In this cohort profile article we describe the lifetime major depressive disorder (MDD) database that has been established as part of the BIObanks Netherlands Internet Collaboration (BIONIC). Across the Netherlands we collected data on Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) lifetime MDD diagnosis in 132,850 Dutch individuals. Currently, N = 66,684 of these also have genomewide single nucleotide polymorphism (SNP) data. We initiated this project because the complex genetic basis of MDD requires large population-wide studies with uniform in-depth phenotyping. For standardized phenotyping we developed the LIDAS (LIfetime Depression Assessment Survey), which then was used to measure MDD in 11 Dutch cohorts. Data from these cohorts were combined with diagnostic interview depression data from 5 clinical cohorts to create a dataset of N = 29,650 lifetime MDD cases (22%) meeting DSM-5 criteria and 94,300 screened controls. In addition, genomewide genotype data from the cohorts were assembled into a genomewide association study (GWAS) dataset of N = 66,684 Dutch individuals (25.3% cases). Phenotype data include DSM-5-based MDD diagnoses, sociodemographic variables, information on lifestyle and BMI, characteristics of depressive symptoms and episodes, and psychiatric diagnosis and treatment history. We describe the establishment and harmonization of the BIONIC phenotype and GWAS datasets and provide an overview of the available information and sample characteristics. Our next step is the GWAS of lifetime MDD in the Netherlands, with future plans including fine-grained genetic analyses of depression characteristics, international collaborations and multi-omics studies.

Apolipoprotein-CIII O-Glycosylation Is Associated with Micro- and Macrovascular Complications of Type 2 Diabetes.

Apolipoprotein-CIII (apo-CIII) inhibits the clearance of triglycerides from circulation and is associated with an increased risk of diabetes complications. It exists in four main proteoforms: O-glycosylated variants containing either zero, one, or two sialic acids and a non-glycosylated variant. O-glycosylation may affect the metabolic functions of apo-CIII. We investigated the associations of apo-CIII glycosylation in blood plasma, measured by mass spectrometry of the intact protein, and genetic variants with micro- and macrovascular complications (retinopathy, nephropathy, neuropathy, cardiovascular disease) of type 2 diabetes in a DiaGene study (n = 1571) and the Hoorn DCS cohort (n = 5409). Mono-sialylated apolipoprotein-CIII (apo-CIII1) was associated with a reduced risk of retinopathy (ß = -7.215, 95% CI -11.137 to -3.294) whereas disialylated apolipoprotein-CIII (apo-CIII2) was associated with an increased risk (ß = 5.309, 95% CI 2.279 to 8.339). A variant of the GALNT2-gene (rs4846913), previously linked to lower apo-CIII0a, was associated with a decreased prevalence of retinopathy (OR = 0.739, 95% CI 0.575 to 0.951). Higher apo-CIII1 levels were associated with neuropathy (ß = 7.706, 95% CI 2.317 to 13.095) and lower apo-CIII0a with macrovascular complications (ß = -9.195, 95% CI -15.847 to -2.543). In conclusion, apo-CIII glycosylation was associated with the prevalence of micro- and macrovascular complications of diabetes. Moreover, a variant in the GALNT2-gene was associated with apo-CIII glycosylation and retinopathy, suggesting a causal effect. The findings facilitate a molecular understanding of the pathophysiology of diabetes complications and warrant consideration of apo-CIII glycosylation as a potential target in the prevention of diabetes complications.

Genome-wide association study of serum magnesium in type 2 diabetes.

People with type 2 diabetes have a tenfold higher prevalence of hypomagnesemia, which is suggested to be caused by low dietary magnesium intake, medication use, and genetics. This study aims to identify the genetic loci that influence serum magnesium concentration in 3466 people with type 2 diabetes. The GWAS models were adjusted for age, sex, eGFR, and HbA1c. Associated traits were identified using publicly available data from GTEx consortium, a human kidney eQTL atlas, and the Open GWAS database. The GWAS identified a genome-wide significant locus in TAF3 (p = 2.9 × 10-9) in people with type 2 diabetes. In skeletal muscle, loci located in TAF3 demonstrate an eQTL link to ATP5F1C, a gene that is involved in the formation of Mg2+-ATP. Serum Mg2+ levels were associated with MUC1/TRIM46 (p = 2.9 × 10-7), SHROOM3 (p = 4.0 × 10-7), and SLC22A7 (p = 1.0 × 10-6) at nominal significance, which is in combination with the eQTL data suggesting that they are possible candidates for renal failure. Several genetic loci were in agreement with previous genomic studies which identified MUC1/TRIM46 (Pmeta = 6.9 × 10-29, PQ = 0.81) and SHROOM3 (Pmeta = 2.9 × 10-27, PQ = 0.04) to be associated with serum Mg2+ in the general population. In conclusion, serum magnesium concentrations are associated with genetic variability around the regions of TAF3, MUC1/TRIM46, SHROOM3, and SLC22A7 in type 2 diabetes.