Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables

doi:10.1016/S2213-8587(18)30051-2

The Lancet Diabetes & Endocrinology

Volume 6, Issue 5, May 2018, Pages 361-369

, , , , , , , , , , , , , , , , , , , …

https://doi.org/10.1016/S2213-8587(18)30051-2 Get rights and content

Summary

Background

Diabetes is presently classified into two main forms, type 1 and type 2 diabetes, but type 2 diabetes in particular is highly heterogeneous. A refined classification could provide a powerful tool to individualise treatment regimens and identify individuals with increased risk of complications at diagnosis.

Methods

We did data-driven cluster analysis (k-means and hierarchical clustering) in patients with newly diagnosed diabetes (n=8980) from the Swedish All New Diabetics in Scania cohort. Clusters were based on six variables (glutamate decarboxylase antibodies, age at diagnosis, BMI, HbA_1c, and homoeostatic model assessment 2 estimates of β-cell function and insulin resistance), and were related to prospective data from patient records on development of complications and prescription of medication. Replication was done in three independent cohorts: the Scania Diabetes Registry (n=1466), All New Diabetics in Uppsala (n=844), and Diabetes Registry Vaasa (n=3485). Cox regression and logistic regression were used to compare time to medication, time to reaching the treatment goal, and risk of diabetic complications and genetic associations.

Findings

We identified five replicable clusters of patients with diabetes, which had significantly different patient characteristics and risk of diabetic complications. In particular, individuals in cluster 3 (most resistant to insulin) had significantly higher risk of diabetic kidney disease than individuals in clusters 4 and 5, but had been prescribed similar diabetes treatment. Cluster 2 (insulin deficient) had the highest risk of retinopathy. In support of the clustering, genetic associations in the clusters differed from those seen in traditional type 2 diabetes.

Interpretation

We stratified patients into five subgroups with differing disease progression and risk of diabetic complications. This new substratification might eventually help to tailor and target early treatment to patients who would benefit most, thereby representing a first step towards precision medicine in diabetes.

Funding

Swedish Research Council, European Research Council, Vinnova, Academy of Finland, Novo Nordisk Foundation, Scania University Hospital, Sigrid Juselius Foundation, Innovative Medicines Initiative 2 Joint Undertaking, Vasa Hospital district, Jakobstadsnejden Heart Foundation, Folkhälsan Research Foundation, Ollqvist Foundation, and Swedish Foundation for Strategic Research.

Introduction

Diabetes is the fastest increasing disease worldwide and a substantial threat to human health.¹ Existing treatment strategies have been unable to stop the progressive course of the disease and prevent development of chronic diabetic complications. One explanation for these shortcomings is that diagnosis of diabetes is based on measurement of only one metabolite, glucose, but the disease is heterogeneous with regard to clinical presentation and progression.

Diabetes classification into type 1 and type 2 diabetes relies primarily on the presence (type 1 diabetes) or absence (type 2 diabetes) of autoantibodies against pancreatic islet β-cell antigens and age at diagnosis (younger for type 1 diabetes). With this approach, 75–85% of patients are classified as having type 2 diabetes. A third subgroup, latent autoimmune diabetes in adults (LADA; affecting <10% of people with diabetes), defined by the presence of glutamic acid decarboxylase antibodies (GADA), is phenotypically indistinguishable from type 2 diabetes at diagnosis, but becomes increasingly similar to type 1 diabetes over time.² With the introduction of gene sequencing in clinical diagnostics, several rare monogenic forms of diabetes were described, including maturity-onset diabetes of the young and neonatal diabetes.3, 4

Existing treatment guidelines are limited by the fact they respond to poor metabolic control when it has developed, but do not have means to predict which patients will need intensified treatment. Evidence suggests that early treatment is crucial for prevention of life-shortening complications because target tissues seem to remember poor metabolic control decades later (so-called metabolic memory).5, 6

A refined classification could provide a powerful tool to identify at diagnosis those at greatest risk of complications and enable individualised treatment regimens in the same way as genetic diagnosis of monogenic diabetes guides clinicians to optimal treatment.⁷ With this aim, we present a novel diabetes classification based on unsupervised, data-driven cluster analysis of six commonly measured variables and compare it metabolically, genetically, and clinically to the current classification in four separate populations from Sweden and Finland.

Research in context

Evidence before this study

National guidelines maintain information about diabetes classification, but this classification has not been much updated during the past 20 years, and very few attempts have been made to explore heterogeneity of type 2 diabetes. We searched PubMed up to Jan 1, 2017, using the Medical Subject Heading terms “diabetes mellitus”, “type 2”, and “classification”. We identified several calls from expert groups for a revised classification, but few efforts to subgroup type 2 diabetes, none of which have been implemented in the clinic.

Added value of this study

In this study, a data-driven cluster analysis of six simple variables measured at diagnosis in adult patients with newly diagnosed diabetes (n=14 755) identified five replicable clusters of patients with significantly different characteristics and risk of diabetic complications. These included a cluster of very insulin-resistant individuals with significantly higher risk of diabetic kidney disease than the other clusters, a cluster of relatively young insulin-deficient individuals with poor metabolic control (high HbA_1c), and a large group of elderly patients with the most benign disease course.

Implications of all the available evidence

This new substratification could change the way we think about type 2 diabetes and help to tailor and target early treatment to patients who would benefit most, thereby representing a first step towards precision medicine in diabetes.

Section snippets

Study populations

We used data from five cohorts: All New Diabetics in Scania (ANDIS), the Scania Diabetes Registry (SDR), All New Diabetics in Uppsala (ANDIU), Diabetes Registry Vaasa (DIREVA), and Malmö Diet and Cancer CardioVascular Arm (MDC-CVA).

The ANDIS project aims to recruit all incident cases of diabetes within Scania County in Sweden (about 1 200 000 inhabitants). All health-care providers in Scania were invited; the current registration covered the period from Jan 1, 2008, to Nov 3, 2016, during which

Results

We first analysed the ANDIS cohort, consisting of 14 652 patients with newly diagnosed diabetes from Sweden, 932 (6·4%) of whom were registered before age 18 years and were not included in our analysis of adult diabetes. Of the 13 720 adult patients, 204 (1·5%) had type 1 diabetes, 723 (5·3%) had LADA, 162 (1·2%) had secondary diabetes (coexisting pancreatic disease), and 519 (3·8%) were unclassifiable because of missing data. The remaining 12 112 (88·3%) patients were considered to have type 2

Discussion

Taken together, the results of our study suggest that this new clustering of patients with adult-onset diabetes is superior to the classic diabetes classification because it identifies patients at high risk of diabetic complications at diagnosis and provides information about underlying disease mechanisms, thereby guiding choice of therapy. By contrast with previous attempts to dissect the heterogeneity of diabetes,²³ we used variables reflective of key aspects of diabetes that are monitored in

References (27)

F Vaziri-Sani et al.
A novel triple mix radiobinding assay for the three ZnT8 (ZnT8-RWQ) autoantibody variants in children with newly diagnosed diabetes
J Immunol Methods
(2011)
M Martinell et al.
Prevalence and risk factors for diabetic retinopathy at diagnosis (DRAD) in patients recently diagnosed with type 2 diabetes (T2D) or latent autoimmune diabetes in the adult (LADA)
J Diabetes Complications
(2016)
C Hennig
Cluster-wise assessment of cluster stability
Comput Stat Data An
(2007)
L Gnudi et al.
Diabetic nephropathy: perspective on novel molecular mechanisms
Trends Endocrinol Metab
(2016)
GI Welsh et al.
Insulin signaling to the glomerular podocyte is critical for normal kidney function
Cell Metab
(2010)
SH Lee et al.
Estimating missing heritability for disease from genome-wide association studies
Am J Hum Genet
(2011)
Worldwide trends in diabetes since 1980: a pooled analysis of 751 population-based studies with 4·4 million participants
Lancet
(2016)
T Tuomi et al.
Antibodies to glutamic acid decarboxylase reveal latent autoimmune diabetes mellitus in adults with a non-insulin-dependent onset of disease
Diabetes
(1993)
P Froguel et al.
Familial hyperglycemia due to mutations in glucokinase. Definition of a subtype of diabetes mellitus
N Engl J Med
(1993)
K Yamagata et al.
Mutations in the hepatocyte nuclear factor-1alpha gene in maturity-onset diabetes of the young (MODY3)
Nature
(1996)

MA Reddy et al.

Epigenetic mechanisms in diabetic complications and metabolic memory

Diabetologia

(2015)

M Brownlee

The pathobiology of diabetic complications: a unifying mechanism

Diabetes

(2005)

ER Pearson et al.

Switching from insulin to oral sulfonylureas in patients with diabetes due to Kir6.2 mutations

N Engl J Med

(2006)

Cited by (1388)

Cardiovascular risk reduction in older people with type 2 diabetes mellitus-a comprehensive narrative review
2024, Diabetes Research and Clinical Practice
Metabolic targets are controversial in older people with type 2 diabetes due to functional heterogeneity and morbidity burden. Tight blood pressure and metabolic control appears beneficial in fit individuals who are newly diagnosed with type 2 diabetes and have fewer comorbidities. The benefits of low blood pressure and tight metabolic control is attenuated with the development of comorbidities, especially frailty. Guidelines consider frail older people as one category and recommend relaxed targets. However, sarcopenic obese frail individuals may benefit from tight targets and intensification of therapy due to their unfavourable metabolic profile, accelerated diabetes trajectory and high cardiovascular risk. In addition, the early use of sodium glucose transporter-2 inhibitors and glucagon like peptide-1 receptor agonists may be beneficial in this frailty phenotype due to their cardio-renal protection, which is independent of glycaemic control, provided they are able to engage in resistance exercise training to avoid loss of muscle mass. In the anorexic malnourished frail individual, early use of insulin, due to its weight gain and anabolic properties, is appropriate. In this phenotype, targets should be relaxed with deintensification of therapy due to significant weight loss, decelerated diabetes trajectory and increased risk of medication side effects.
Diabetes subgroups and sociodemographic inequalities in Mexico: a cross-sectional analysis of nationally representative surveys from 2016 to 2022
2024, The Lancet Regional Health - Americas
Differences in the prevalence of four diabetes subgroups have been reported in Mexico compared to other populations, but factors that may contribute to these differences are poorly understood. Here, we estimate the prevalence of diabetes subgroups in Mexico and evaluate their correlates with indicators of social disadvantage using data from national representative surveys.
We analyzed serial, cross-sectional Mexican National Health and Nutrition Surveys spanning 2016, 2018, 2020, 2021, and 2022, including 23,354 adults (>20 years). Diabetes subgroups (obesity-related [MOD], severe insulin-deficient [SIDD], severe insulin-resistant [SIRD], and age-related [MARD]) were classified using self-normalizing neural networks based on a previously validated algorithm. We used the density-independent social lag index (DISLI) as a proxy of state-level social disadvantage.
We identified 4204 adults (median age: 57, IQR: 47–66, women: 64%) living with diabetes, yielding a pooled prevalence of 16.04% [95% CI: 14.92–17.17]. When stratified by diabetes subgroup, prevalence was 6.62% (5.69–7.55) for SIDD, 5.25% (4.52–5.97) for MOD, 2.39% (1.95–2.83) for MARD, and 1.27% (1.00–1.54) for SIRD. SIDD and MOD clustered in Southern Mexico, whereas MARD and SIRD clustered in Northern Mexico and Mexico City. Each standard deviation increase in DISLI was associated with higher odds of SIDD (OR: 1.12, 95% CI: 1.06–1.12) and lower odds of MOD (OR: 0.93, 0.88–0.99). Speaking an indigenous language was associated with higher odds of SIDD (OR: 1.35, 1.16–1.57) and lower odds of MARD (OR 0.58, 0.45–0.74).
Diabetes prevalence in Mexico is rising in the context of regional and sociodemographic inequalities across distinct diabetes subgroups. SIDD is a subgroup of concern that may be associated with inadequate diabetes management, mainly in marginalized states.
This research was supported by Instituto Nacional de Geriatría in Mexico.
The hidden epidemic: Uncovering incidental fatty liver disease and its metabolic comorbidities by datamining in a hospital data lake – A real-world cohort study
2024, Diabetes Research and Clinical Practice
To identify individuals with incidental fatty liver disease (FLD), and to evaluate its prevalence, metabolic co-morbidities and impact on follow-up.
We leveraged the data-lake of Helsinki Uusimaa Hospital district (Finland) with a population of 1.7 million (specialist and primary care). A phrase recognition script on abdominal imaging reports (2008–2020) identified/excluded FLD or cirrhosis; we extracted ICD-codes, laboratory and BMI data.
Excluding those with other liver diseases, the prevalence of FLD was 29% (steatosis yes/no, N=61,271/155,521; cirrhosis, N=3502). The false positive and negative rates were 5–6%. Only 1.6% of the FLD cases had the ICD code recorded and 32% had undergone full clinical evaluation for associated co-morbidities. Of the 35–65-year-old individuals with FLD, 20% had diabetes, 42% prediabetes and 28% a high liver fibrosis index. FLD was independently predicted by diabetes (OR 1.56, CI 1.46–1.66, p = 2.3 * 10^-41), BMI (1.46, 1.42–1.50, p = 1.7 * 10^-154) and plasma triglyceride level (1.5, 1.43–1.57, p = 3.5 * 10^-68). Alanine aminotransferase level mildly increased (1.12, 1.08–1.16, p = 2.2 * 10^-9) and high age decreased the risk (0.92, 0.89–0.94, p = 4.65*10^-09). Half of the cases had normal ALT.
The incidental radiological finding of FLD is reliable and associated with metabolic risks but largely ignored, although it should lead to metabolic and hepatic follow-up.
Association of dietary patterns with diabetes-related comorbidities varies among diabetes endotypes
2024, Nutrition, Metabolism and Cardiovascular Diseases
Differences of dietary pattern adherence across the novel diabetes endotypes are unknown. This study assessed adherence to pre-specified dietary patterns and their associations with cardiovascular risk factors, kidney function, and neuropathy among diabetes endotypes.
The cross-sectional analysis included 765 individuals with recent-onset (67 %) and prevalent diabetes (33 %) from the German Diabetes Study (GDS) allocated into severe autoimmune diabetes (SAID, 35 %), severe insulin-deficient diabetes (SIDD, 3 %), severe insulin-resistant diabetes (SIRD, 5 %), mild obesity-related diabetes (MOD, 28 %), and mild age-related diabetes (MARD, 29 %). Adherence to a Mediterranean diet score (MDS), Dietary Approaches to Stop Hypertension (DASH) score, overall plant-based diet (PDI), healthful (hPDI) and unhealthful plant-based diet index (uPDI) was derived from a food frequency questionnaire and associated with cardiovascular risk factors, kidney function, and neuropathy using multivariable linear regression analysis. Differences in dietary pattern adherence between endotypes were assessed using generalized mixed models.
People with MARD showed the highest, those with SIDD and MOD the lowest adherence to the hPDI. Adherence to the MDS, DASH, overall PDI, and hPDI was inversely associated with high-sensitivity C-reactive protein (hsCRP) among people with MARD (β (95%CI): −9.18 % (−15.61; −2.26); −13.61 % (−24.17; −1.58); −19.15 % (−34.28; −0.53); −16.10 % (−28.81; −1.12), respectively). Adherence to the PDIs was associated with LDL cholesterol among people with SAID, SIRD, and MOD.
Minor differences in dietary pattern adherence (in particular for hPDI) and associations with markers of diabetes-related complications (e.g. hsCRP) were observed between endotypes. So far, evidence is insufficient to derive endotype-specific dietary recommendations.
Clinicaltrials.gov: NCT01055093.
Unique Challenges in Diabetic Foot Science
2024, Clinics in Podiatric Medicine and Surgery
A comprehensive evaluation of the phenotype-first and data-driven approaches in analyzing facial morphological traits
2024, iScience
The phenotype-first approach (PFA) and data-driven approach (DDA) have both greatly facilitated anthropological studies and the mapping of trait-associated genes. However, the pros and cons of the two approaches are poorly understood. Here, we systematically evaluated the two approaches and analyzed 14,838 facial traits in 2,379 Han Chinese individuals. Interestingly, the PFA explained more facial variation than the DDA in the top 100 and 1,000 except in the top 10 phenotypes. Accordingly, the ratio of heterogeneous traits extracted from the PFA was much greater, while more homogenous traits were found using the DDA for different sex, age, and BMI groups. Notably, our results demonstrated that the sex factor accounted for 30% of phenotypic variation in all traits extracted. Furthermore, we linked DDA phenotypes to PFA phenotypes with explicit biological explanations. These findings provide new insights into the analysis of multidimensional phenotypes and expand the understanding of phenotyping approaches.

View all citing articles on Scopus

^†: Contributed equally

View full text

ArticlesNovel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables

Summary

Background

Methods

Findings

Interpretation

Funding

Introduction

Section snippets

Study populations

Results

Discussion

J Immunol Methods

J Diabetes Complications

Comput Stat Data An

Trends Endocrinol Metab

Cell Metab

Am J Hum Genet

Worldwide trends in diabetes since 1980: a pooled analysis of 751 population-based studies with 4·4 million participants

Lancet

Antibodies to glutamic acid decarboxylase reveal latent autoimmune diabetes mellitus in adults with a non-insulin-dependent onset of disease

Diabetes

Familial hyperglycemia due to mutations in glucokinase. Definition of a subtype of diabetes mellitus

N Engl J Med

Mutations in the hepatocyte nuclear factor-1alpha gene in maturity-onset diabetes of the young (MODY3)

Nature

Epigenetic mechanisms in diabetic complications and metabolic memory

Diabetologia

The pathobiology of diabetic complications: a unifying mechanism

Diabetes

Switching from insulin to oral sulfonylureas in patients with diabetes due to Kir6.2 mutations

N Engl J Med

Articles
Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables