Funding agencies: This study was supported by The Swedish Research Council, the Swedish Heart and Lung foundation, the Marianne and Marcus Wallenberg Foundation, EXODIAB, Dalarna University and Uppsala University.
Hypertriglyceridemic waist phenotype is associated with decreased insulin sensitivity and incident diabetes in elderly men
Version of Record online: 13 JUN 2013
Copyright © 2013 The Obesity Society
Volume 22, Issue 2, pages 526–529, February 2014
How to Cite
Carlsson, A. C., Risérus, U. and Ärnlöv, J. (2014), Hypertriglyceridemic waist phenotype is associated with decreased insulin sensitivity and incident diabetes in elderly men. Obesity, 22: 526–529. doi: 10.1002/oby.20434
Disclosure: The authors declared no conflict of interest.
Author Contributions: JÄ and ACC conceived and designed the study; JÄ and ACC performed the statistical analysis; ACC, JÄ and UR interpreted the data; ACC drafted the manuscript; JÄ and UR critically revised the manuscript for important intellectual content; JÄ obtained the funding and supervised the study.
- Issue online: 3 FEB 2014
- Version of Record online: 13 JUN 2013
- Accepted manuscript online: 20 MAR 2013 02:33AM EST
- Manuscript Accepted: 17 FEB 2013
- Manuscript Received: 26 NOV 2012
- The Swedish Research Council
- Swedish Heart and Lung foundation
- Marianne and Marcus Wallenberg Foundation
- Dalarna University
- Uppsala University
Objective: To investigate the association between hypertriglyceridemic waist (HTGW) and insulin sensitivity (assessed by euglycemic clamp method), and the development of diabetes in a longitudinal community-based cohort of elderly men without diabetes at baseline.
Design and Methods: The present cross-sectional study comprised 1,026, 70-year-old men without diabetes. The gold standard euglycaemic–hyperinsulinaemic clamp technique was used. Six-year follow-up on diabetes status were available in n = 667. The HTGW phenotype was defined as having waist circumference ≥ 90 cm, and triglycerides ≥ 2 mmol L−1. The men were stratified into those having normal WC and TG (n = 299), one HTGW component (n = 606), and HTGW (n = 121).
Results: The association between insulin sensitivity and one HTGW component as well as HTGW was highly significant (P < 0.001) in the whole sample, as well as in individuals with high/low BMI (stratified at ≥25). In longitudinal analyses, participants with HTGW was associated with a more than fourfold increased risk for diabetes (Odds ratio 4.64, 95% CI 1.61–13.4, P = 0.004) compared to those with normal WC and TG.
Conclusion: The present study both confirm and extend previous research suggesting that the HTGW-phenotype portrays an increased glucometabolic risk, also in lean individuals.
With the increasing prevalence of obesity in populations throughout the world , a variety of measures have been proposed to identify those at increased cardio-metabolic risk for diseases such as cardiovascular disease and type 2 diabetes. These range from the simple anthropometric measures such as body mass index and waist circumference (WC) [2, 3], to more sophisticated methods such as the metabolic syndrome . Hypertriglyceridemic waist (HTGW) is a combination of abdominal obesity defined as an increased WC and elevated fasting triglycerides (TG), which may be a simpler tool for clinicians than the metabolic syndrome for assessing patient risk as only two components has to be measured, WC and TG. Moreover, HTGW has been shown to be equivalent to definitions of the metabolic syndrome for assessing cardiovascular risk . Less is known about the association between HTGW, insulin sensitivity and the risk of incident type 2 diabetes.
We hypothesized that HTGW is an insulin resistant state that increases diabetes risk. Accordingly, we aimed to investigate the association between HTGW and insulin sensitivity (as assessed by the gold standard euglycemic clamp method), and the development of diabetes in a longitudinal community-based cohort of elderly men without diabetes at baseline.
The analyses in the present study are based on the third and fourth examination cycles of The Uppsala Longitudinal Study of Adult Men (ULSAM) when participants were ∼71- and 77-years old, respectively (http://www.pubcare.uu.se/ULSAM) . After exclusion of 195 men with diabetes [defined as self-reported use of insulin or oral anti diabetic drugs in the questionnaire, or as having fasting plasma glucose above 7 mmol L−1 ] or unavailable covariates, the present study comprised 1,026 individuals (baseline investigation 1991-95). Follow-up data on diabetes status were available in 667 participants (1998–2002).
The euglycaemic–hyperinsulinaemic clamp technique according to De Fronzo  was used, with a slight modification to suppress hepatic glucose production , for estimation of in vivo sensitivity to insulin. Insulin (Actrapid Human; Novo, Copenhagen, Denmark) was infused in a primary dose for the first 10 min and then as a continuous infusion (56 mU min−1 per body surface area in meters squared, for 2 h to maintain steady-state hyperinsulinemia). The target plasma glucose level was 5.1 mmol L−1, maintained by measuring plasma glucose every 5 min. The glucose infusion rate during the last hour was used as a measure of insulin sensitivity (M value) and is given in mg/kg bw/min. The insulin sensitivity index (M/I ratio) is a measure of the tissue sensitivity to insulin expressed per unit insulin and was calculated by dividing M by the mean insulin concentration during the same period of the clamp. M/I thus represent the amount of glucose metabolized per unit of plasma insulin and is given in mg/kg bw/min per mU/L of insulin multiplied by 100. The calculation of the total body insulin sensitivity is based on the assumption that endogenous hepatic glucose production is entirely suppressed. Under euglycemic conditions it is known that almost 90% of this production is suppressed when the plasma insulin concentration is increased by 50 mU L−1 .
Waist circumference and serum triglycerides was assessed as previously described . The HTGW phenotype was defined as WC ≥ 90 cm and TG ≥ 2 mmol L−1. The men were stratified into those having:
- Normal WC and TG (n = 299)
- One HTGW component (either high WC or TG) (n = 606)
- HTGW (n = 121)
Linear regression analyses were used to assess cross-sectional associations between HTGW (independent variable) and insulin sensitivity (dependent variable), with adjustments for age (model A) as well as age and BMI (model B). We also performed these models in subsamples stratified at BMI ≥ 25 (low or normal weight/overweight or obesity) . Logistic regression was used to investigate the longitudinal association between HTGW and the development of type 2 diabetes.
The statistical software package STATA 11.2 (Stata, College Station, TX) was used for all analyses.
The baseline characteristics were stratified into BMI-groups (≥25, Table 1) as well as number of HTGW components (Table 2). The insulin sensitivity, waist circumference and triglycerides varied greatly in these different strata.
|n||BMI < 25||BMI ≥ 25|
|Age (years)||71 (0.6)||71 (0.6)|
|BMI (kg m−2)||23 (1.5)||28 (2.4)|
|Insulin sensitivity||6.6 (1.8)||4.7 (1.7)|
|Waist circumference (cm)||87 (6)||99 (8)|
|Triglycerides (mmol L−1)||1.2 (0.6)||1.5 (0.7)|
|N||Normal WC||One component||Elevated TG|
|Normal TG||Elevated WC, HTGW|
|Age (years)||71 (0.6)||71 (0.6)||71 (0.6)|
|BMI (kg m−2)||23 (1.9)||27 (2.7)||28 (3.3)|
|Insulin sensitivity||6.8 (1.7)||5.1 (1.7)||3.9 (1.4)|
|Waist circumference (cm)||84 (5)||97 (7)||101 (8)|
|Triglycerides (mmol L−1)||0.36 (0.4)||1.3 (0.4)||2.7 (0.7)|
The cross-sectional association between insulin sensitivity and one HTGW component as well as the HTGW phenotype was highly significant (P < 0.001) in all models tested (Table 3). The insulin sensitivity decrease (negative β-coefficients) with each added HTGW component. Moreover, the findings were consistent in when the study sample was stratified into individuals with low/normal weight, and overweight/obese individuals. During follow-up (median 6 years), 44 participants developed diabetes (numbers at risk 667). In these longitudinal analyses, participants with one HTGW component had a nonsignificant twofold increased diabetes risk, and participants with the HTGW-phenotype was associated with a more than fourfold increased risk for diabetes (Table 4). The association between presence of HTGW and incident diabetes was non-significant after adjustments for BMI as a continuous variable (odds ratio 2.66, 95% CI 0.29-8.91, P = 0.11).
|Participants with BMI < 25||Participants with BMI ≥ 25||All participants|
|Model A||Model B||Model A||Model B||Model A||Model B|
|β-coefficient (95% CI)||β-coefficient (95% CI)||β-coefficient (95% CI)||β-coefficient (95% CI)||β-coefficient (95% CI)||β-coefficient (95% CI)|
|One HTGW component||−0.92a (−1.28, −0.56)||−0.53a (−0.90, −0.15)||−1.41a (−1.89, −0.94)||−0.82a (−1.27, −0.37)||−1.78a (−2.02, −1.54)||−0.65a (−0.91, −0.39)|
|HTGW||−2.37a (−3.27, −1.47)||−1.91a (−2.80, −1.03)||−2.29a (−2.84, −1.75)||−1.45a (−1.97, −0.92)||−2.87a (−3.23, −2.51)||−1.37a (−1.75, −1.0)|
|Unadjusted association||Odds ratio||Confidence interval (P value)|
|Normal WC and TG||1||(referent)|
|One component||2.38||0.97–5.88 (0.06)|
Our findings suggest that HTGW identifies a group of men with impaired insulin sensitivity and increased risk of incident type 2 diabetes. The findings accord with associations found in cross-sectional studies: a Canadian study of healthy men where HTGW was associated with hyperinsulinemia as well as elevated apo B, and small dense LDL , and a study in an Inuit population which reported that HTGW was associated with prevalent diabetes . Moreover, among individuals with glucose intolerance and type 2 diabetes, individuals with HTGW compared to those without HTGW have been shown to develop cardiovascular disease earlier [5, 12]. However, to our knowledge the associations between HTGW, insulin sensitivity, and incident diabetes have not been reported previously.
Our findings may be explained by risk factor clustering, as both components of the HTGW phenotype are components of the metabolic syndrome , previously known as insulin resistance syndrome. HTGW has additionally been shown to capture increased visceral adiposity , which together with liver accumulation of fat could promote hepatic triglyceride secretion as well as insulin resistance. The mechanism is unclear, but is believed to involve altered hepatic free fatty acid metabolism, altered adipokine release from visceral fat , as well as various hormonal disturbances . In conjunction with visceral adiposity, liver fat accumulation may also induce both hypertriglyceridemia as well as insulin resistance . The metabolic disorders of visceral fat accumulation is also present in lean individuals who are classified as normal weight according to their BMI. Indeed, the present association between HTGW and insulin sensitivity appeared similar in participants above vs. below BMI 25 kg m−2.
Because of the changing of risk factors toward more abdominal obesity and subsequent insulin resistance [17, 18], a simple clinical tool is needed that identifies cardiometabolic risk in lean individuals A clinical limitation with the metabolic syndrome definitions is a fairly high residual risk in individuals without the syndrome, resulting in a risk of CVD that appears low, ∼1.5 ; which can be explained by a relatively high risk in the reference group, limiting the clinical utility of metabolic syndrome. In contrast to metabolic syndrome definitions that identify individuals with or without the trait , we have shown that hypertriglyceridaemic waist may be useful in identifying men and women at low, intermediate (with either a large waist or high triglycerides) and high risk to develop diabetes. Others have shown that the number of HTGW components gradually detect increased risks of downstream cardiovascular complications .
The strengths of the present study include the large community-based sample with gold standard clamp measurements of insulin sensitivity, as well as longitudinal data on incident diabetes. Our findings are limited due to the fact that they need confirmation in other age-groups, ethnic groups, and in women. Our elderly study population makes our analysis subject to potential survival bias and subject to healthy cohort bias. However, the development of diabetes late in life is a great public health concern.
Because of the limited number of participants with new onset diabetes during follow-up it was not possible to adequately address the issue of whether the strength of the association between HTGW and diabetes risk is different in normal weight individuals compared to overweight/obese individuals.
We warrant additional longitudinal studies where the prognostic ability of HTGW can be analyzed in different BMI-strata.
In conclusion, the present study both confirm and extend previous research suggesting that the HTGW-phenotype portrays an increased glucometabolic risk. The clinical relevance of our findings is uncertain and needs further investigation.
The study was investigator initiated and driven. Johan Ärnlöv is the guarantor of this study.