Using data-driven analysis k-means based on fat content in the liver, pancreas, muscle, and visceral bed, we identified four fat distribution clusters: hepatic steatosis, pancreatic steatosis, trunk myosteatosis, and steatopenia. In comparisons with the steatopenia cluster, the adjusted hazard ratios for incident T2D were 4.

The clusters were replicated in German individuals without diabetes who underwent MRI and metabolic phenotyping. The distribution of the glucose area under the curve across the four clusters found in Germany was similar to the distribution of T2D risk across the four clusters in Japan.

In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript. To compare the relationships between markers of total and regional adiposity with muscle fat infiltration in type 1 diabetic and type 2 diabetic subjects and their respective nondiabetic controls, and to document these relationships in type 1 diabetic subjects.

In total, 86 healthy, with type 1 diabetes, type 2 diabetes or control subjects. Each diabetic group was matched for age, sex and body mass index with its respective nondiabetic control group.

Measures of body composition hydrodensitometry , fat distribution waist circumference, abdominal and mid-thigh computed tomography scans and blood lipid profiles were assessed.

Low attenuation mid-thigh muscle surface correlated similarly with markers of adiposity and body composition in all groups, regardless of diabetes status, except for visceral adipose tissue and waist circumference.

In addition, in well-controlled type 1 diabetic subjects mean HbA 1c of 6. This study suggests that the relationship of central adiposity and muscle adiposity is modulated by diabetes status and is stronger in the insulin resistant diabetes type type 2 diabetes.

In well-controlled nonobese type 1 diabetic subjects, the relationship between muscle fat accumulation and insulin sensitivity was also maintained. This is a preview of subscription content, access via your institution.

Goldstein BJ. Insulin resistance as the core defect in type 2 diabetes mellitus. Am J Cardiol ; 90 : 3G—10G. Article CAS Google Scholar. Goodpaster BH, Thaete FL, Kelley DE. Composition of skeletal muscle evaluated with computed tomography. Ann NY Acad Sci ; : 18— Itani SI, Ruderman NB, Schmieder F, Boden G.

Lipid-induced insulin resistance in human muscle is associated with changes in diacylglycerol, protein kinase C, and IkappaB-alpha. Diabetes ; 51 : — Gan SK, Kriketos AD, Poynten AM, Furler SM, Thompson CH, Kraegen EW et al.

Insulin action, regional fat, and myocyte lipid: altered relationships with increased adiposity. Obes Res ; 11 : — Bushberg JT, Seibert JA, Leidholdt EMJ, Boone JM. X-Ray computed tomography.

In: Passano WM eds. The Essentials of Medical Imaging. Williams et Wilkins: Baltimore, MD, , pp — Google Scholar. Goodpaster BH, Brown NF.

Skeletal muscle lipid and its association with insulin resistance: What is the role for exercise? Exerc Sport Sci Rev ; 33 : — Article Google Scholar. Goodpaster BH, Thaete FL, Simoneau JA, Kelley DE.

Subcutaneous abdominal fat and thigh muscle composition predict insulin sensitivity independently of visceral fat. Diabetes ; 46 : — Yki-Jarvinen H, Helve E, Koivisto VA. Hyperglycemia decreases glucose uptake in type I diabetes. Diabetes ; 36 : — Rossetti L, Smith D, Shulman GI, Papachristou D, DeFronzo RA.

Correction of hyperglycemia with phlorizin normalizes tissue sensitivity to insulin in diabetic rats. J Clin Invest ; 79 : — Rossetti L, Giaccari A, DeFronzo RA.

Glucose toxicity. Diabetes Care ; 13 : — Perseghin G, Lattuada G, Danna M, Sereni LP, Maffi P, De Cobelli F et al. Insulin resistance, intramyocellular lipid content, and plasma adiponectin in patients with type 1 diabetes.

Am J Physiol Endocrinol Metab ; : E—E E-pub August Ebeling P, Essen-Gustavsson B, Tuominen JA, Koivisto VA. Intramuscular triglyceride content is increased in IDDM. Diabetologia ; 41 : — The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus.

American diabetes association: clinical practice recommendations Diabetes Care ; 25 Suppl 1 : S1—S Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man.

Diabetologia ; 28 : — Callaway WC, Chumlea WC, Bouchard C, Himes JH, Lohman TG, Martin AD et al. In: Lohman TG, Roche AF, Martorel R eds. Anthropometric Standardization Reference Manual.

Human Kinetic Books: Champaign, IL, , pp 39— Meneely E, Kaltreider N. Curr Obes Rep. Shiel F, Persson C, Furness J, Simas V, Schram B. Dual energy X-ray absorptiometry positioning protocols in assessing body composition: a systematic review of the literature.

J Sci Med Sport. American Diabetes Association. Classification and Diagnosis of Diabetes: Standards of Medical Care in Diabetes Diabetes Care. Use WHOECoPSt, Anthropometry Io. Physical status: the use and interpretation of anthropometry.

Report of a WHO Expert Committee. World Health Organ Tech Rep Ser. PubMed Abstract Google Scholar. Dickey RA, Bartuska DG, Bray GW, Callaway W, Kennedy FP. Google Scholar. Holman RR, Paul SK, Bethel MA, Matthews DR, Neil H. N Engl J Med. Després J. Abdominal obesity as important component of insulin-resistance syndrome.

Klein S, Sheard NF, Pi-Sunyer X, Daly A, Wylie-Rosett J, Kulkarni K, et al. Weight management through lifestyle modification for the prevention and management of type 2 diabetes: rationale and strategies. Am J Clin Nutr. Bray G. Pathophysiology of obesity. Lee DH, Keum N, Hu FB, Orav EJ, Rimm EB, Willet WC, et al.

Comparison of the association of predicted fat mass, body mass index, and other obesity indicators with type 2 diabetes risk: two large prospective studies in US men and women. Eur J Epidemiol. Boyko EJ, Fujimoto WY, Leonetti DL, Newell-Morris L.

Visceral adiposity and risk of type 2 diabetes: a prospective study among Japanese Americans. Neeland IJ, Turer AT, Ayers CR, Powell-Wiley TM, Vega GL, Farzaneh-Far R, et al. Dysfunctional adiposity and the risk of prediabetes and type 2 diabetes in obese adults.

Björntorp P. Neuroendocrine abnormalities in human obesity. An KH, Han KA, Sohn TS, Park IB, Min KW. Body fat is related to sedentary behavior and light physical activity but not to moderate-vigorous physical activity in type 2 diabetes mellitus.

Diabetes Metab J. Merlotti C, Morabito A, Ceriani V, Pontiroli AE. Prevention of type 2 diabetes in obese at-risk subjects: a systematic review and meta-analysis. Acta Diabetol. Hamman RF, Wing RR, Edelstein SL, Lachin JM, Bray GA, Delahanty L, et al.

Effect of weight loss with lifestyle intervention on risk of diabetes. Zurlo FS, Lillioja S, Puente ED, Nyomba BL, Ravussin E. Low ratio of fat to carbohydrate oxidation as predictor of weight gain: study of h RQ. Am J Physiol. Robinson SL, Hattersley J, Frost GS, Chambers ES, Wallis GA.

Maximal fat oxidation during exercise is positively associated with hour fat oxidation and insulin sensitivity in young, healthy men. J Appl Physiol. Pernicova I, Korbonits M. Metformin—mode of action and clinical implications for diabetes and cancer.

Nat Rev Endocrinol. Stumvoll M, Nurjhan N, Perriello G, Dailey G, Gerich JE. Metabolic effects of metformin in non-insulin-dependent diabetes mellitus.

Pi-Sunyer X, Astrup A, Fujioka K, Greenway F, Halpern A, Krempf M, et al. A randomized, controlled trial of 3. Haider D, Schaller G, Kapiotis S, Maier C, Luger A, Wolzt M. The release of the adipocytokine visfatin is regulated by glucose and insulin.

Colagiuri S, Lee CMY, Wong TY, Balkau B, Borch-Johnsen K. Glycemic thresholds for diabetes-specific retinopathy: implications for diagnostic criteria for diabetes. Monami M, Dicembrini I, Kundisova L, Zannoni S, Nreu B, Mannucci E.

A meta-analysis of the hypoglycaemic risk in randomized controlled trials with sulphonylureas in patients with type 2 diabetes. Diabetes Obes Metab. Abdelmoneim AS, Eurich DT, Light PE, Senior PA, Seubert JM, Makowsky MJ, et al. Cardiovascular safety of sulphonylureas: over 40years of continuous controversy without an answer.

Barnett A. Complementing insulin therapy to achieve glycemic control. Adv Ther. Mingrone G, Panunzi S, Gaetano AD, Guidone C, Rubino F. Bariatric surgery versus conventional medical therapy for type 2 diabetes. Citation: Sun J, Liu Z, Zhang Z, Zeng Z and Kang W The Correlation of Prediabetes and Type 2 Diabetes With Adiposity in Adults.

Received: 19 November ; Accepted: 04 March ; Published: 11 April Copyright © Sun, Liu, Zhang, Zeng and Kang. This is an open-access article distributed under the terms of the Creative Commons Attribution License CC BY.

The use, distribution or reproduction in other forums is permitted, provided the original author s and the copyright owner s are credited and that the original publication in this journal is cited, in accordance with accepted academic practice.

No use, distribution or reproduction is permitted which does not comply with these terms. Importance of Body Composition Analysis in Clinical Nutrition.

Design, Setting, and Participants Genome-wide association studies GWAS for WHR combined data from the UK Biobank cohort and summary statistics from previous GWAS data collection: Specific polygenic scores for higher WHR via lower gluteofemoral or via higher abdominal fat distribution were derived using WHR-associated genetic variants showing specific association with hip or waist circumference.

Associations of polygenic scores with outcomes were estimated in 3 population-based cohorts, a case-cohort study, and summary statistics from 6 GWAS data collection: Exposures More than 2. Main Outcomes and Measures BMI-adjusted WHR and unadjusted WHR GWAS ; compartmental fat mass measured by dual-energy x-ray absorptiometry, systolic and diastolic blood pressure, low-density lipoprotein cholesterol, triglycerides, fasting glucose, fasting insulin, type 2 diabetes, and coronary disease risk follow-up analyses.

Conclusions and Relevance Distinct genetic mechanisms may be linked to gluteofemoral and abdominal fat distribution that are the basis for the calculation of the WHR.

These findings may improve risk assessment and treatment of diabetes and coronary disease. The distribution of body fat is associated with the propensity of overweight individuals to manifest insulin resistance and its associated metabolic and cardiovascular complications.

While many studies support this assertion and plausible mechanisms have been proposed, WHR can also be increased by a reduction in its denominator, the hip circumference.

Evidence from several different forms of partial lipodystrophy 6 , 7 and functional studies of peripheral adipose storage compartments 8 - 10 suggests that a primary inability to expand gluteofemoral or hip fat can also underpin subsequent cardiometabolic disease risk.

Emerging evidence from the analysis of common genetic variants associated with greater insulin resistance but lower levels of hip fat suggests that similar mechanisms may also be relevant to the general population.

In this study, large-scale human genetic data were used to investigate whether genetic variants related to body fat distribution via lower levels of gluteofemoral hip fat or via higher levels of abdominal waist fat are associated with type 2 diabetes or coronary disease risk.

A multistage approach was adopted Table 1. In stage 1, genome-wide association studies GWAS of WHR with and without adjustment for BMI were performed to identify genetic variants associated with fat distribution.

Stage 1 included data from participants of European ancestry in the UK Biobank study and summary statistics from previously published GWAS of the Genetic Investigation of Anthropometric Traits GIANT Consortium. Stage 2 included data from participants of European ancestry in the UK Biobank and summary statistics from GIANT.

In stage 3, associations of polygenic scores with compartmental fat mass measured by dual-energy x-ray absorptiometry DEXA were estimated in participants of European ancestry in the UK Biobank, Fenland, and EPIC-Norfolk studies.

In stage 4, associations of polygenic scores with 6 cardiometabolic risk factors and with risk of type 2 diabetes and coronary artery disease were estimated using data from participants of European ancestry in the UK Biobank, the EPIC-InterAct case-cohort study, and summary statistics from 6 previously published GWAS.

All studies were approved by local institutional review boards and ethics committees, and participants gave written informed consent.

The UK Biobank data collection: is a prospective population-based cohort study of people aged 40 to 69 years who were recruited from to from 22 centers located in urban and rural areas across the United Kingdom. Fenland data collection: is a prospective population-based cohort study of people born from to and recruited from to from outpatient primary care clinics in Cambridge, Ely, and Wisbech United Kingdom.

EPIC-Norfolk data collection: is a prospective population-based cohort study of individuals aged 40 to 79 years and living in Norfolk County rural areas, market towns, and the city of Norwich in the United Kingdom at recruitment from outpatient primary care clinics in to EPIC-InterAct data collection: is a case-cohort study nested within the European Prospective Investigation into Cancer and Nutrition EPIC , a prospective cohort study.

Summary statistics from 11 GWAS published by research consortia between and were used in the different stages of the study eMethods 1 and eTable 1 in the Supplement.

Detailed descriptions of study design, sources of data, and participants in each stage are in Table 1 and Table 2 , and eMethods 1 and eTables in the Supplement.

Outcomes of the study were WHR stages 1 and 2b , hip and waist circumference stage 2a , compartmental body fat masses stage 3 , 6 cardiometabolic risk factors systolic and diastolic blood pressure, fasting glucose, fasting insulin, triglycerides, and LDL-C; stage 4 , and 2 disease outcomes type 2 diabetes and coronary disease; stage 4.

In stages 1 and 2, WHR was defined as the ratio of the circumference of the waist to that of the hip, both of which were estimated in centimeters using a Seca cm tape measure.

BMI-adjusted WHR was obtained by calculating the residuals for a linear regression model of WHR on age, sex, and BMI. In stage 3, compartmental fat masses were measured in grams by DEXA, a whole-body, low-intensity x-ray scan that precisely quantifies fat mass in different body regions.

In the UK Biobank, DEXA measures were obtained using a GE-Lunar iDXA instrument. In the Fenland and EPIC-Norfolk studies, DEXA scans were performed using a Lunar Prodigy advanced fan beam scanner GE Healthcare. Participants were scanned by trained operators using standard imaging and positioning protocols.

All the images were manually processed by one trained researcher, who corrected DEXA demarcations according to a standardized procedure as illustrated and described in eFigure 1 and eMethods 1, respectively, in the Supplement.

In brief, the arm region included the arm and shoulder area. The trunk region included the neck, chest, and abdominal and pelvic areas. The abdominal region was defined as the area between the ribs and the pelvis, and was enclosed by the trunk region. The leg region included all of the area below the lines that form the lower borders of the trunk.

The gluteofemoral region included the hips and upper thighs, and overlapped both leg and trunk regions. The upper demarcation of this region was below the top of the iliac crest at a distance of 1. DEXA CoreScan software GE Healthcare was used to determine visceral abdominal fat mass within the abdominal region.

In stage 4, the risk factors included systolic and diastolic blood pressures, defined as the values of arterial blood pressure in mm Hg measured using an Omron monitor during the systolic and diastolic phases of the heart cycle.

For disease outcomes analyses in the UK Biobank in stage 4, binary definitions of prevalent disease status and a case-control analytical design were used in line with previous work.

Controls were participants who 1 did not self-report a diagnosis of diabetes of any type, 2 did not take any diabetes medications, and 3 did not have an electronic health record of diabetes of any type. In EPIC-InterAct, the outcome was incident type 2 diabetes.

Incident type 2 diabetes case status was defined on the basis of evidence of type 2 diabetes from self-report, primary care registers, drug registers medication use , hospital record, or mortality data.

Participants with prevalent diabetes at study baseline were excluded from EPIC-InterAct. Controls were participants who did not meet any of these criteria. In stage 1, GWAS analyses were performed in the UK Biobank using BOLT-LMM, 27 which fits linear mixed models accounting for relatedness between individuals using a genomic kinship matrix.

Restriction to individuals of European ancestry, use of linear mixed models UK Biobank , and adjustment for genetic principal components and genomic inflation factor GIANT were used to minimize type I error.

Quality measures of genuine genetic association signal vs possible confounding by population stratification or relatedness included the mean χ 2 statistic, the linkage-disequilibrium score LDSC regression intercept, and its attenuation ratio eMethods 2 in the Supplement , as recommended for genetic studies of this size using linear mixed model estimates.

A forward-selection process was used to select independent genetic variants for stage 2. Full details about genetic analyses are in eMethods 2 in the Supplement. In stage 2, polygenic scores capturing genetic predisposition to higher WHR were derived by combining the independent genetic variants from stage 1 or subsets of the variants as described below , weighted by their association with BMI-adjusted WHR in stage 1.

A general polygenic score for higher WHR was derived by combining all genetic variants. A fourth polygenic score was derived by combining genetic variants not included in the waist- or hip-specific polygenic scores. The statistical performance of these polygenic scores was assessed by estimating the proportion of the variance in BMI-adjusted WHR accounted for by the score variance explained and by the F statistic eMethods 4 in the Supplement.

The F statistic is a measure of the ability of the polygenic score to predict the independent variable BMI-adjusted WHR. Values of F statistic greater than 10 have been considered to provide evidence of a statistically robust polygenic score.

In stages 3 and 4, associations of polygenic scores with DEXA phenotypes, cardiometabolic risk factors, and outcomes were estimated in each study separately and results were combined using fixed-effect inverse-variance weighted meta-analysis.

In individual-level data analyses, polygenic scores were calculated for each study participant by adding the number of copies of each contributing genetic variant weighted by its association estimate in SD units of BMI-adjusted WHR per allele from stage 1.

Association of polygenic scores with outcomes were estimated using linear, logistic, or Cox regression models as appropriate for outcome type and study design. Regression models were adjusted for age, sex, and genetic principal components or a genomic kinship matrix to minimize genetic confounding.

In UK Biobank disease outcomes analyses, prevalent disease status was defined as a binary variable and logistic regression was used to estimate the odds ratio OR of disease per 1-SD increase in BMI-adjusted WHR due to a given polygenic score.

In EPIC-InterAct, Cox regression weighted for case-cohort design was used to estimate the hazard ratio of incident type 2 diabetes per 1-SD increase in BMI-adjusted WHR due to a given polygenic score. In summary statistics analyses, estimates equivalent to those of individual-level analyses were obtained using inverse-variance weighted meta-analysis of the association of each genetic variant in the polygenic score with the outcome, divided by the association of that genetic variant with BMI-adjusted WHR.

They also assume a linear relationship of the polygenic score with continuous outcomes linear regression , with the log-odds of binary outcomes logistic regression , or with the log-hazard of incident disease Cox regression.

All of these assumptions were largely met in this study eMethods 5, eTable 4, and eFigures in the Supplement. Meta-analyses of log-ORs and log—hazard ratios of disease assumed that these estimates are similar, an assumption that was shown to be reasonable in a sensitivity analysis conducted in EPIC-InterAct eMethods 5 and eFigure 7 in the Supplement.

In stages 3 and 4, associations with continuous outcomes were expressed in standardized or clinical units of outcome per 1-SD increase in BMI-adjusted WHR corresponding to 0. Associations with disease outcomes were expressed as ORs for outcome per 1-SD increase in BMI-adjusted WHR due to a given polygenic score.

Absolute risk increases ARIs for disease outcomes were estimated using the estimated ORs and the incidence of type 2 diabetes or coronary disease in the United States eMethods 5 in the Supplement.

All reported P values were from 2-tailed statistical tests. In addition to deriving specific polygenic scores, the independent association of gluteofemoral or abdominal fat distribution with outcomes was studied using multivariable genetic association analyses adjusting for either of these 2 components of body fat distribution eMethods 6 and eFigure 8 in the Supplement.

Adjusting for abdominal fat distribution measures was used as a way of estimating the residual association of the polygenic score with outcomes via gluteofemoral fat distribution, while adjusting for gluteofemoral fat distribution measures as a way of estimating the residual association via abdominal fat distribution eFigure 8 in the Supplement.

To obtain adjusted association estimates, multivariable-weighted regression models were fitted in which the association of the variant general polygenic score exposure with cardiometabolic risk factors or diseases outcomes was estimated while adjusting for a polygenic score comprising the same genetic variants but weighted for measures of abdominal fat distribution or measures of gluteofemoral fat distribution covariates.

This method was also used to conduct a post hoc exploratory analysis of the association of the hip-specific polygenic score with cardiometabolic disease outcomes after adjusting for visceral abdominal fat mass estimates.

Statistical analyses were performed using Stata version These genetic variants were used to derive polygenic scores for higher WHR Table 1. The general variant and variant polygenic scores were associated with higher visceral abdominal and lower gluteofemoral fat mass Figure 1 A; eFigure 15 in the Supplement.

The waist-specific polygenic score for higher WHR was associated with higher abdominal fat mass, but not with gluteofemoral or leg fat mass Figure 1 B. The hip-specific polygenic score for higher WHR was associated with lower gluteofemoral and leg fat mass, but did not show statistically significant associations with abdominal fat mass Figure 1 B.

Participants with higher values of the hip-specific polygenic score had numerically higher visceral abdominal fat mass, but the difference was not statistically significant when accounting for multiple tests Figure 1 B.

Both hip-specific and waist-specific polygenic scores for higher WHR were associated with higher systolic and diastolic blood pressure and triglyceride level, with similar association estimates for a 1-SD increase in BMI-adjusted WHR Figure 2 A.

While the hip-specific polygenic score was associated with higher fasting insulin and higher LDL-C levels, the waist-specific polygenic score did not have statistically significant associations with these traits Figure 2 A. Both the hip-specific and waist-specific polygenic scores were associated with higher odds of type 2 diabetes and coronary disease, similarly in men and women Figure 2 B and eTable 9 in the Supplement.

The hip-specific polygenic score had a statistically larger association estimate for diabetes than the waist-specific polygenic score per 1-SD increase in BMI-adjusted WHR OR, 2.

In a post-hoc multivariable analysis adjusting for visceral abdominal fat mass estimates, the hip-specific polygenic score showed a statistically significant association with higher odds of type 2 diabetes and coronary disease OR for diabetes per 1-SD increase in BMI-adjusted WHR due to the hip-specific polygenic score, 2.

The variant polygenic score showed associations with risk factors and disease outcomes similar to those observed for the variant general polygenic score eFigure 15 in the Supplement.

Sensitivity analyses supported the robustness of the main analysis to sex-specific associations, associations with height, or the possibility of false-positive associations in stage 1 or stage 2 eMethods 7 and eTables in the Supplement. In multivariable analyses adjusting for hip circumference estimates, the variant polygenic score had a pattern of association with compartmental fat mass, cardiometabolic risk factors, and disease outcomes, which was similar to that of the waist-specific polygenic score eFigures 8D and 17 in the Supplement.

The variant polygenic score remained associated with higher risk of type 2 diabetes and coronary disease even when adjusting for hip circumference and leg fat mass in the same model eTable 12 in the Supplement.

In multivariable analyses adjusting for waist circumference estimates, the variant polygenic score had a pattern of association with compartmental fat mass, cardiometabolic risk factors, and disease outcomes, which was similar to that of the hip-specific polygenic score eFigures 8C and 17 in the Supplement.

The variant polygenic score remained associated with higher risk of type 2 diabetes and coronary disease even when adjusting for waist circumference and visceral abdominal fat mass in the same model eTable 12 in the Supplement. In multivariable analyses adjusting for both waist and hip circumference estimates, the variant polygenic score was not associated with risk of type 2 diabetes or coronary disease eFigure 8B and eTable 12 in the Supplement.

This large study identified distinct genetic variants associated with a higher WHR via specific associations with lower gluteofemoral or higher abdominal fat distribution.

Both of these distinct sets of genetic variants were associated with higher levels of cardiometabolic risk factors and a higher risk of type 2 diabetes and coronary disease. While this study supports the theory that an enhanced accumulation of fat in the abdominal cavity may be a cause of cardiovascular and metabolic disease, it also provides novel evidence of a possible independent role of the relative inability to expand the gluteofemoral fat compartment.

Previous studies of approximately 50 genomic regions associated with BMI-adjusted WHR 16 have shown an association between genetic predisposition to higher WHR and higher risk of cardiometabolic disease, 26 , 35 mirroring the well-established BMI-independent association of a higher WHR with incident cardiovascular and metabolic disease in large-scale observational studies.

The results of this study support the hypothesis that an impaired ability to preferentially deposit excess calories in the gluteofemoral fat compartment leads to higher cardiometabolic risk in the general population. This is consistent with observations in severe forms of partial lipodystrophy 6 , 7 and with the emerging evidence of a shared genetic background between extreme lipodystrophies and fat distribution in the general population.

These associations may perhaps reflect the secondary deposition within ectopic fat depots, such as liver, cardiac and skeletal muscle, and pancreas, of excess calories that cannot be accommodated in gluteofemoral fat. It has been hypothesized that the association between fat distribution and cardiometabolic risk is due to an enhanced deposition of intra-abdominal fat generating a molecular milieu that fosters abdominal organ insulin resistance.

This study has several limitations. First, as this is an observational study, it cannot establish causality. Second, the discovery and characterization of genetic variants was conducted in a large data set but was limited to individuals of European ancestry.

While the genetic determinants of anthropometric phenotypes may be partly shared across different ethnicities, 16 , 39 , 40 further investigations in other populations and ethnicities will be required for a complete understanding of the genetic relationships between body shape and cardiometabolic risk.

Third, this study was largely based on population-based cohorts, the participants of which are usually healthier than the general population, and used analytical approaches that deliberately minimized the influence of outliers, in this case people with extreme fat distribution.

Genetic studies in people with extreme fat distribution may help broaden understanding of the genetic basis of this risk factor. Fifth, absolute risk increase estimates are based on incidence rates and ORs calculated in different populations and therefore assume that these populations are similar.

Seventh, this analysis focused on common genetic variants captured in both UK Biobank and GIANT and, by design, did not investigate the role of rare genetic variation or of other variants captured by dense imputation in the UK Biobank.

Eighth, there was a statistically significant difference in the association of hip- vs waist-specific polygenic scores with diabetes risk, with greater estimated magnitude of association for the hip-specific polygenic score.

However, given that the difference in absolute risk was small, this observation does not necessarily represent a strong signal of mechanistic difference or differential clinical importance in the relationship between the gluteofemoral vs abdominal components of fat distribution and diabetes risk.

Distinct genetic mechanisms may be linked to gluteofemoral and abdominal fat distribution that are the basis for the calculation of the waist-to-hip ratio.

Corresponding Authors: Claudia Langenberg, MD, PhD claudia. langenberg mrc-epid. uk , and Luca A. Lotta, MD, PhD luca. lotta mrc-epid. uk , MRC Epidemiology Unit, University of Cambridge, Cambridge CB20QQ, United Kingdom. Author Contributions: Dr Lotta had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Acquisition, analysis, or interpretation of data: Lotta, Wittemans, Zuber, Stewart, Sharp, Luan, Day, Li, Bowker, Cai, De Lucia Rolfe, Khaw, Perry, Scott, Burgess, Wareham, Langenberg. Critical revision of the manuscript for important intellectual content: Lotta, Wittemans, Zuber, Stewart, Sharp, Luan, Day, Li, Bowker, Cai, De Lucia Rolfe, Khaw, Scott, Burgess, Wareham, Langenberg.

Statistical analysis: Lotta, Wittemans, Zuber, Stewart, Sharp, Luan, Day, Li, Bowker, Cai, De Lucia Rolfe, Perry, Burgess, Langenberg. Obtained funding: Khaw, Wareham, Langenberg. Administrative, technical, or material support: De Lucia Rolfe, Khaw, Wareham, Langenberg.

Supervision: Lotta, Wareham, Langenberg. Dr Scott is an employee and shareholder in GlaxoSmithKline. No other disclosures were reported. Additional Contributions: This research was conducted using the UK Biobank resource and data from the EPIC-InterAct, Fenland, and EPIC-Norfolk studies. We gratefully acknowledge the help of the MRC Epidemiology Unit Support Teams, including the field, laboratory, and data management teams.

full text icon Full Text. Download PDF Top of Article Key Points Abstract Introduction Methods Results Discussion Conclusions Article Information References. Figure 1. Associations With Compartmental Fat Mass of Polygenic Scores for Higher Waist-to-Hip Ratio WHR.

View Large Download. Figure 2. Associations With Cardiometabolic Risk Factors and Disease Outcomes of Waist- or Hip-Specific Polygenic Scores for Higher Waist-to-Hip Ratio WHR. Table 1. Summary of the Study Design. Table 2. Participants of the UK Biobank Included in This Study a.

Data Sources, Study Design, Measurements, and Phenotype Definitions eMethods 2. Genetic Association Analyses eMethods 3. Selection of Subsets of Genetic Variants Associated With Higher WHR via a Specific Association With Higher Waist Circumference, or via a Specific Association With Lower Hip Circumference eMethods 4.

Assessment of Performance and Statistical Power of Polygenic Scores for Higher WHR eMethods 5. Assumptions and Interpretation of Association Analyses Between Polygenic Scores for Higher WHR and Outcome Traits eMethods 6.

Multivariable Genetic Association Analyses eMethods 7. Secondary and Sensitivity Analyses eTable 1. Participating Studies eTable 2. and UK Biobank Studies Who Underwent Detailed Anthropometric Measurements by Dual-Energy X-ray Absorptiometry eTable 3.

Characteristics of Participants of the EPIC-InterAct Study Included in the Analysis eTable 4. Difference in Age-, Sex- and BMI-Residualized WHR at Different Levels of the Distribution of Standardized BMI-Adjusted WHR Following the Inverse-Rank Normal Transformation eTable 5.

Standard Deviation Values Used to Convert Estimates Between Clinical and Standardized Units and Their Source eTable 6. List of the Independent Lead Genetic Variants Identified in Stage 1 Which Were Used to Derive Polygenic Scores for Higher WHR eTable 7.

Associations of Polygenic Scores for Higher WHR With Additional Continuous Phenotypes in Secondary Analyses eTable 8. Associations of Polygenic Scores for Higher WHR With Nondiabetic Hyperglycemia eTable 9.

Association of Polygenic Scores for Higher WHR With Risk of Type 2 Diabetes and Coronary Artery Disease in Men and Women From the UK Biobank Study eTable Results of Sensitivity Analyses eTable Associations of the Variant Polygenic Score for Higher WHR With Cardiometabolic Disease Outcomes in Multivariable Genetic Association Analyses Adjusting for Height eTable Associations of the Genetic Variants With Risk of Cardiometabolic Disease Outcomes in Multivariable Genetic Analyses eFigure 1.

Compartmental Body Fat Mass Measurement by Dual-Energy X-ray Absorptiometry eFigure 2. Statistical Power Calculations eFigure 3. Distribution of the Values of Polygenic Scores for Higher WHR in UK Biobank eFigure 4. Distribution of the Values of Standardized Systolic and Diastolic Blood Pressure Outcome Variables in UK Biobank eFigure 5.

Linear Association Between Polygenic Score for Higher WHR and Outcomes eFigure 6. Distribution of BMI-Adjusted WHR Variables in UK Biobank eFigure 7. Correlation of Estimates From Weighted Cox and Logistic Regression Models in EPIC-InterAct eFigure 8.

Schematic Representation of Multivariable Polygenic Score Association Analysis eFigure 9. Diagnostic Funnel Plots for the Association of the Genetic Variants Included in the Polygenic Scores for Higher WHR and Type 2 Diabetes or Coronary Disease eFigure Manhattan and Quantile-Quantile Plot for the Genome-Wide Association Analysis of BMI-Adjusted WHR eFigure Manhattan and Quantile-Quantile Plot for the Genome-Wide Association Analysis of Unadjusted WHR eFigure Associations of the Genetic Variants With BMI-Adjusted WHR in GIANT and UK Biobank eFigure Consistency of Stage 1 Associations After Exclusion of Cardiometabolic Disease Cases eFigure Associations With Hip, Waist Circumference and Body Mass Index of the Four Polygenic Scores for Higher WHR eFigure Associations With DEXA Variables, Cardio-metabolic Risk Factors and Disease Outcomes of the Variant Polygenic Score for Higher WHR eFigure Associations with Cardiometabolic Risk Factors and Disease Outcomes of the Variant Polygenic Score for Higher WHR eFigure Associations With Anthropometry, Cardio-metabolic Risk Factors and disease outcomes of Variant Polygenic Score for Higher WHR in Multivariable Genetic Association Analyses Adjusted for Genetic Associations With Hip or Waist Circumference eReferences.

Vague J. The degree of masculine differentiation of obesities: a factor determining predisposition to diabetes, atherosclerosis, gout, and uric calculous disease.

Am J Clin Nutr. doi: Obesity and the risk of myocardial infarction in 27, participants from 52 countries: a case-control study. Biggs ML, Mukamal KJ, Luchsinger JA, et al. Association between adiposity in midlife and older age and risk of diabetes in older adults.

Langenberg C, Sharp SJ, Schulze MB, et al; InterAct Consortium. Long-term risk of incident type 2 diabetes and measures of overall and regional obesity: the EPIC-InterAct case-cohort study.

PLoS Med. Stefan N, Häring HU, Hu FB, Schulze MB. Metabolically healthy obesity: epidemiology, mechanisms, and clinical implications. Lancet Diabetes Endocrinol. Garg A. Acquired and inherited lipodystrophies.

N Engl J Med. Compared with body mass index BMI , total percent fat TPF as well as the indexes of fat distribution by region trunk fat mass, android fat mass, gynoid fat mass and android to gynoid ratio are more important and accurate indicators in obesity evaluation.

Specifically, dual-energy x-ray absorptiometry DXA is widely considered a precise and accurate clinical technology for directly measuring fat mass and distribution nowadays 7 , 8. In recent years, a significant association between prediabetes or T2DM and adiposity has been established.

However, these studies used proxy measures for overall or abdominal obesity such as BMI or waist circumference without taking into account the composition of that mass and the correlation between diabetes status, blood glucose, HbA1c and fat mass measured by DXA in large samples are still limited.

Therefore, this study investigated the relationship of diabetes status and disease duration of T2DM with TPF and fat distribution in adults using a nationally representative sample. Furthermore, due to the different blood glucose, HbA1c and adiposity between individuals with and without prediabetes or T2DM, we also evaluated the associations of blood glucose and HbA1c with adiposity in the subgroup stratified by diabetes status.

National Health and Nutrition Examination Survey NHANES is a continuous surveillance survey conducted by the Centers for Disease Control and Prevention CDC and the National Center for Health Statistics NCHS to assess the health and nutritional status of the US population.

Data obtained from NHANES can be freely available to researchers worldwide. In our study, we pooled data from 10 2-year cycles of NHANES — A total of , participants were enrolled from the NHANES — database.

To minimize the number of participants with T1DM, participants with age of DM onset before the age of 30 were also excluded. Finally, 28, participants were analyzed after applying these exclusion criteria Figure 1. The National Center for Health Statistics Research Ethics Review Board reviewed and approved NHANES, and all participants signed written consents prior to participating in each year's survey.

De-identified data are accessible online. The outcomes of our study were TPF, trunk fat mass, android fat mass, gynoid fat mass and android to gynoid ratio, which were measured by well-trained technicians using dual energy X-ray absorptiometry DXA QDR Hologic Scanner Bedford, MA of the whole body.

TPF was calculated by the ratio of total fat mass to total fat and lean mass and multiplied by to produce a percentage. The regions of fat distribution were defined by the Hologic APEX software. Specifically, the trunk fat mass was defined as total fat mass minus fat in limbs and head.

The android area was defined as the lower trunk area bounded by two lines: the pelvic horizontal cut line on its lower side, and a line automatically placed above the pelvic line. The upper gynoid line was placed 1. DXA scans were not performed for participants with a self-reported use of radiographic contrast material barium in the last seven days before scans, weight above pounds or height above Exposures were prediabetes and T2DM status, disease duration in those with T2DM, serum glucose and HbA1c.

Disease durations in patients with T2DM were defined as the age at screening minus the age when which doctors told the participants they had DM. Serum glucose and Hb1Ac were obtained from the standard biochemistry profile and glycohemoglobin sections in the laboratory data part of NHANES.

Sociodemographic variables mainly included age, sex, race and ratio of family income to poverty threshold. Current insulin use and glucose-lowering medication intake were assessed by questionnaires.

Health behaviors variables included smoking status whether smoked at least cigarettes in life and vigorous work activity. Health-related variables were hypertension ever told by a doctor that you have high blood pressure , hypercholesterolemia ever told by a doctor that you have high cholesterol , BMI and waist circumference calculated during the study visit.

The NHANES sample weights were used as recommended by the NCHS. The associations between diabetes status, duration of T2DM, serum glucose, HbA1c and TPF or fat distribution were evaluated by multivariable linear regression models. The model was adjusted for age, sex, race, health behaviors smoking status, vigorous work activity , hypertension, and hypercholesterolemia.

Subgroup analyses stratified by sex and diabetes status were also performed. The weighted distributions of the characteristics according to diabetes status were shown in Table 1. In 7, prediabetic and T2DM patients, 6, Compared with non-diabetic participants, prediabetic, and T2DM participants were older, more had hypertension, hypercholesterolemia, smoked at least cigarettes in life and fewer had vigorous work activity.

Besides, prediabetes and T2DM group had higher BMI, waist circumference, total cholesterol, triglyceride, blood urea nitrogen, uric acid, creatinine, and lower ratio of family income to poverty, HDL cholesterol. The mean ± SD of disease duration of diabetes for T2DM group was 6.

Furthermore, the serum glucose, HbA1c, TPF, trunk fat mass, android fat mass, gynoid fat mass, and android to gynoid ratio in adults with prediabetes or T2DM were all higher than non-diabetic participants.

We found the direct associations of prediabetes and T2DM status with TPF and fat distribution compared with non-diabetes in fully adjusted models βs in sequence of TPF, trunk fat mass, android fat mass, gynoid fat mass, and android to gynoid ratio for prediabetes were: 2.

That is to say, after controlling for potential confounding factors, compared with those without DM, the TPF, trunk fat mass, android fat mass, gynoid fat mass and android to gynoid ratio for prediabetes were 2.

In the subgroup analysis stratified by sex, this direct association existed in both males and females after adjusting for confounders. These results are presented in Table 2.

After adjusting for sociodemographic covariates, health behaviors, hypertension, and hypercholesterolemia, we found direct associations of serum glucose and HbA1c with TPF, trunk fat mass, android fat mass, gynoid fat mass, and android to gynoid ratio.

And when stratifying by diabetes status, these direct associations still existed in participants without diabetes and with prediabetes. Other associations were of no statistical significance. The results are shown in Tables 3 , 4. When stratified by sex, the inverse associations of disease duration with TPF and gynoid fat mass were still statistically significant in females, but not in males.

Besides, the inverse associations of disease duration with trunk fat mass, android fat mass and android to gynoid ratio were of no statistical significance either. The results are shown in Table 5. Table 5. Associations of disease duration years with adiposity in patients with T2DM.

The results of our study showed that individuals with prediabetes and T2DM had significantly higher TPF, trunk fat mass, android fat mass, gynoid fat mass, and android to gynoid ratio compared with those without DM.

The fat mass decreased as the disease duration increased in patients with T2DM. Moreover, in participants without DM and with prediabetes, serum glucose and HbA1c were directly associated with TPF, trunk fat mass, android fat mass, gynoid fat mass, and android to gynoid ratio, while the inverse associations were observed in those with T2DM.

Prediabetes and T2DM are associated with the increased insulin resistance in target organs e. Although patients with prediabetes or T2DM are not necessarily obese, weight gain before DM develop is common Obesity is recognized as the most powerful environmental risk factor among several modifiable risk factors for diabetes 14 , which is associated with an increased insulin demand and increased likelihood of insulin resistance leading to prediabetes or hyperinsulinemia and ultimately T2DM 13 , Therefore, the results of our study that trunk fat mass, android fat mass, gynoid fat mass, android to gynoid ratio as well as TPF were higher in patients with prediabetic and T2DM than those without DM could be explained.

Lee's study found that the predicted fat mass and percent fat estimated by anthropometric prediction equations were also positively associated with the risk of T2DM A study of Japanese Americans found that greater visceral adiposity preceded the development of T2DM and also demonstrated an effect independent of fasting insulin, insulin secretion, glycemia, total and regional adiposity, and family history of diabetes Furthermore, the investigation of associations between adiposity phenotypes and risk for incident prediabetes and diabetes of obese adults found that visceral adiposity, increased liver fat, decreased lower body fat, insulin resistance, elevated triglycerides, and low adiponectin levels were associated with incident prediabetes and diabetes in obese individuals The pathologic process of this increased insulin resistance may include the following aspects: the accumulation of excess fat leading to the increase of plasma free fatty acid FFA levels in obese patients may interfere with muscle insulin sensitivity and the increased FFA of those intra-abdominal tissues drained by portal circulation may lead to high FFA in portal vein, which may inhibit the hepatic clearance of portal insulin in turn.

Besides, obesity may also cause the increased cortisol and androgen secretion leading to lower insulin sensitivity in muscle tissue and liver and physical and chronic psychologic stress may play an important role in exacerbating insulin resistance, prediabetes, and T2DM 11 , The present study also showed that TPF, trunk fat mass, android fat mass, and gynoid fat mass decreased as the disease duration of T2DM increased, although it was of no significance for trunk fat mass and android fat mass.

This may be because once an individual was diagnosed with T2DM, the use of anti-diabetic drugs, the intervention of diet, and exercise may cause weight loss and decrease in fat mass, and the elimination of blood sugar through urination at an extremely high sugar level may also make some contributions The oxidation of FFA during exercise was associated with insulin sensitivity, metabolic flexibility, and body fat mass 23 , There were studies showing that metformin can decrease food intake and body weight 25 , with weight loss preferentially involving adipose tissue Glucagon-like peptide-1 GLP-1 receptor agonists such as liraglutide had been proved to sustain weight loss in obese patients and were associated with the reversal of prediabetes to normoglycemia during 1—2 years of follow-up 1 and another study showed that 3.

In terms of insulin for the treatment of T2DM, a study by Haider showed that insulin or somatostatin infusion suppressed glucose-induced elevation of visfatin a novel insulin-mimetic adipocytokine 28 , which can reduce fat accumulation and insulin resistance in patients with T2DM.

All of these may coincide with the result of the decreased fat mass as the duration of T2DM increases in our study. The blood glucose and HbA1c are considered as measures of DM control and parameters in relation to the risk of complications for decades.

Of the two, HbA1c is more stable and convenient because of its absence of fasting In this study, we found inverse associations of serum glucose and HbA1c with adiposity in patients with T2DM but direct correlations in those with prediabetes and without DM, which may be due to some drugs used for patients with T2DM being associated with weight gain in addition to their function of lowering blood glucose and HbA1c.

In T2DM, there were patients taking insulin and 1, patients taking glucose-lowering pills, with a few in prediabetes also doing so For example, sulfonylureas, such as gliclazide and glimepiride, were reported to have an association with hypoglycemia 30 and weight gain at the same time of their actions on β-cells to stimulate insulin secretion Besides, thiazolidinediones, such as rosiglitazone and pioglitazone, were used clinically for improving of insulin sensitivity, might cause weight gain of up to 6 kg, which mainly because of fluid retention 4.

Although insulin is an effective treatment to control blood glucose, weight and reduce HbA1c of 1. What's more, the bariatric surgery for those patients with T2DM and obesity was effective for weight reduction but also risky for hyperglycemia Coincident with these findings, our study has implied that the simple measure of serum glucose, HbA1c or fat mass, and distribution may be insufficient to monitor the development and treatment effect of prediabetes and T2DM.

The main strengths of this study are the availability of a large, nationally representative population of US adults with data of fat mass and diabetes status from NHANES. Additionally, the availability of body composition measures by DXA offers additional information compared to the traditional measures of adiposity.

More importantly, a large enough sample size allowed us to make the subgroup based on diabetes status and showed the distinct but neglected pattern of serum glucose and HbA1c with TPF and fat distribution that had never been reported in the previous studies.

This study also has several limitations. First, it is a cross-sectional study, which limits the inference of a causal correlation between serum glucose, HbA1c and TPF, fat distribution among adults. So, further basic mechanism research and large sample prospective study are still needed to identify the exact mechanism between them.

Second, some NHANES participants were not eligible for a DXA scan because of excessive weight, height, or other reasons, so the estimates in this study might not fully represent the TPF and fat distribution in the general population.

Third, there remains the possibility of bias caused by other potential confounding factors that we did not adjust for. Furthermore, some associations of disease duration, serum glucose, and HbA1c with adiposity in patients with T2DM were of no clinical or statistical significance, which may mainly because the number of patients with T2DM was small, so study on a larger sample of patients with T2DM is eagerly needed.

Besides, TPF defined by DXA may have several limitations, so the body fat mass and distribution in prediabetes and T2DM requires in-depth research in multiple areas such as diagnostic criteria for obesity in T2DM patients using TPF measured by DXA, alteration in fat mass since prediabetes or T2DM is diagnosed and its detailed mechanisms.

Our study indicated that adults with prediabetes and T2DM had significantly higher TPF, trunk fat mass, android fat mass, gynoid fat mass, and android to gynoid ratio compared with those without DM, and the fat mass decreased as the disease duration of T2DM increased.

We also found inverse associations between serum glucose, HbA1c, and fat mass in patients with T2DM and direct association in those with prediabetes and without DM participants, which may give us a hint that just the measurements of serum glucose, HbA1c, or fat mass and distribution may be insufficient to monitor the development and treatment effect of T2DM.

The Ethics Review Board of the National Center for Health Statistics approved all NHANES protocols. JS and ZL contributed to data collection, statistical analysis, and writing and revising of the manuscript.

ZZh and ZZe contributed to statistical analysis. WK supervised the study and contributed to polishing and reviewing of the manuscript. All authors contributed to the article and approved the submitted version. This work was supported by the CSCO-ROCHE Research Fund No.

Y Roche, Beijing Xisike Clinical Oncology Research Foundation Y-HS, Wu Jieping Medical Foundation No. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

The authors thank the staff and the participants of the NHANES study for their valuable contributions. Tabák AG, Herder C, Rathmann W, Brunner EJ, Kivimäki M. Prediabetes: a high-risk state for diabetes development. doi: PubMed Abstract CrossRef Full Text Google Scholar.

Control CD Prevention. CDC, National Diabetes Fact Sheet: General information and National Estimates on Diabetes in the United States Cho NH, Shaw JE, Karuranga S, Huang Y, Fernandes JR, Ohlrogge AW, et al. IDF diabetes atlas: global estimates of diabetes prevalence for and projections for Diabetes Res Clin Pract.

Chatterjee S, Khunti K, Davies MJ.