Clinical Trials Logo
NCT number NCT03239782
Study type Interventional
Source Clinical Nutrition Research Centre, Singapore
Status Active, not recruiting
Phase N/A
Start date March 29, 2016
Completion date November 1, 2017

Clinical Trial Summary

The prevalence of overweight and obesity in Singapore is approximately half of that in the United States, yet the incidence of type 2 diabetes is similar, and is expected to double in the near future. This indicates that metabolic dysfunction, particularly insulin resistance, is widely prevalent even among individuals who are considered normal-weight or lean by conventional measures, i.e. body mass index (BMI) and percent body fat. These individuals are often referred to as "metabolically-obese normal-weight" (MONW), and have increased risk for cardiometabolic disease despite their normal BMI and total body fat values. The prevalence of the MONW phenotype varies across populations and differs markedly among different ethnicities. However, our understanding of the complex interactions between ethnicity, body composition, and metabolic dysfunction and its reversal remains rudimentary. Previous attempts to characterize the MONW phenotype are confounded by the small but significant differences in BMI or percent body fat between groups (even if all subjects were lean, within the "normal" range), with MONW subjects being always "fatter" than the corresponding control subjects. There are no published studies that prospectively recruited groups of metabolically healthy and unhealthy lean individuals matched on BMI and percent body fat. Furthermore, although weight loss improves body composition and many of the cardiometabolic abnormalities in most obese patients, little is known about the possible therapeutic effects of calorie restriction in MONW subjects.

Accordingly, a better understanding of the MONW phenotype and the evaluation of therapeutic approaches for its reversal will have important implications for public health. By facilitating earlier identification of these subjects, who are more likely to go undiagnosed and thus less likely to be treated before clinically overt cardiometabolic disease develops, results from this study will allow for earlier and effective intervention.

Clinical Trial Description

By the year 2050, it is estimated that more than half of the Singapore population will be overweight or obese, defined as having a body mass index (BMI, calculated as the weight in kilograms divided by the square of height in meters) equal to or greater than 25 kg/m2 (Phan et al. 2014). This is likely responsible, at least in part, for the concomitant increase in obesity-related co-morbid conditions, particularly type 2 diabetes (Phan et al. 2014; Ni Mhurchu et al. 2006). The relationship between BMI and the risk for type 2 diabetes in populations from the Asia-Pacific region is linear within a wide range of BMI values (from ~21 kg/m2 to ~34 kg/m2), so that for every 2 kg/m2 increase in BMI (which corresponds to ~6 kg for a normal-weight person of average stature), the risk for developing type 2 diabetes rises by ~27 % (Ni Mhurchu et al. 2006). In Singapore, the prevalence of type 2 diabetes is expected to double from 7.3 % in 1990 to ~15 % in 2050, predominantly as a result of the fattening of the population, with the burden being greater for those of Indian descent than those of Chinese descent (Phan et al. 2014). This is expected to reduce productivity, inflate healthcare costs, and increase mortality among Singaporeans (Phan et al. 2014; Ma et al. 2003).

The prevalence of type 2 diabetes in Singapore is similar to that in the United States, even though the prevalence of overweight and obesity (BMI ≥25 kg/m2) in Singapore is approximately half that in the US (Yoon et al. 2006). This observation corroborates findings from many studies showing that markers of metabolic dysfunction (e.g. hyperglycemia, hyperinsulinemia, insulin resistance, dyslipidemia, and hypertension) are highly prevalent among Singaporean adults even at normal BMI values, i.e. even among people who are considered "normal-weight" or "lean" by conventional measures (Deurenberg-Yap et al. 1999; Deurenberg-Yap, Chew, et al. 2001). The existence of people who have normal body weight but also have metabolic dysfunction, and therefore greater risk for developing cardiometabolic disease, was recognized several decades ago (Ruderman et al. 1998; Ruderman, Schneider, and Berchtold 1981). At the extreme of this paradigm, even among members of the Calorie Restriction Society who undergo self-imposed calorie restriction for years based on the belief that this will help them ensure a long and healthy life, there are many individuals (~40 %) with impaired glucose tolerance, despite very low BMI and total body fat (Fontana, Klein, and Holloszy 2010). These individuals are often referred to as "metabolically-obese normal-weight" (MONW) or "metabolically-abnormal lean" or "metabolically-unhealthy lean" subjects. The prevalence of this phenotype ranges from 5 % to 45 % depending on the BMI and the metabolic criteria used for its definition, as well as the characteristics of the population (i.e. age, sex, and ethnicity) (Conus, Rabasa-Lhoret, and Peronnet 2007; Teixeira et al. 2015). Similar variability has been observed across Asia (Lee et al. 2011; Luo et al. 2015; Yoo et al. 2014; Jung et al. 2015; Indulekha et al. 2015). For example, among the Chinese, ~8 % of the population as a whole, or ~13 % of those who are considered lean by virtue of body fat percent (i.e. ≤25 % for men and ≤35 % for women), are metabolically unhealthy, defined as having three or more metabolic abnormalities characteristic of the metabolic syndrome (Luo et al. 2015). Among Indians, on the other hand, 15-25 % of the population (or 20-40 % of those who are considered lean by virtue of BMI, i.e. <25 kg/m2) satisfies the criteria for metabolic syndrome (Indulekha et al. 2015; Geetha et al. 2011). The MONW phenotype in Asians is associated with 3-fold greater risk for carotid atherosclerosis (i.e. cardiovascular disease) (Yoo et al. 2014) and 4.5-8.5-fold greater risk for developing type 2 diabetes (Luo et al. 2015). In fact, MONW subjects have increased risk for cardiometabolic disease (Luo et al. 2015; Yoo et al. 2014) and greater all-cause mortality (Choi et al. 2013) not only compared to metabolically-healthy lean subjects, but also compared to metabolically-healthy obese subjects. This underscores the importance of metabolic dysfunction independent of excess body weight and total adiposity.

The mechanisms responsible for the development of metabolic abnormalities in lean people are not entirely clear. The MONW phenotype can manifest early in life, e.g. during childhood (Guerrero-Romero et al. 2013), which corroborates the existence of genetic predisposition for metabolic dysfunction in the face of low BMI values (Yaghootkar et al. 2014). Previous studies have identified a number of factors associated with the MONW phenotype, including increased intra-abdominal (visceral) adipose tissue, increased liver and muscle fat content, increased fat cell size, adipose tissue inflammation, altered inflammatory and adipokine profiles, reduced skeletal muscle mass, lack of physical activity, and low cardio-respiratory fitness (Badoud et al. 2015; Dvorak et al. 1999; Ruderman et al. 1998; Conus, Rabasa-Lhoret, and Peronnet 2007; De Lorenzo et al. 2007; Karelis et al. 2004; Kim et al. 2013; Lee 2009; Oliveros et al. 2014; Teixeira et al. 2015; Di Renzo et al. 2006; Conus et al. 2004; Indulekha et al. 2015; Luo et al. 2015; Fontana, Klein, and Holloszy 2010). All of these factors have been directly or indirectly associated with insulin resistance (defined by a variety of methods), which is by far the commonest metabolic correlate of the MONW phenotype across all ethnicities, age groups, and sexes (Conus, Rabasa-Lhoret, and Peronnet 2007; Oliveros et al. 2014; Karelis et al. 2004; Ruderman et al. 1998). In fact, the greater prevalence of the MONW phenotype in Indians (Indulekha et al. 2015; Geetha et al. 2011) than in the Chinese (Luo et al. 2015) mirrors results obtained recently by our team, showing that among lean Singaporean men (BMI <25 kg/m2 or body fat ≤20 %), those of Indian descent have significantly lower insulin sensitivity, evaluated as the insulin-mediated glucose disposal rate during a hyperinsulinemic-euglycemic clamp procedure, compared to those of Chinese descent (Khoo et al. 2014). Similar results have been reported by other investigators in smaller groups of subjects (Liew et al. 2003) or when using simpler indices of insulin sensitivity (Khoo et al. 2011; Tai et al. 2000). Therefore, an insulin resistant glucose metabolism, broadly defined by subnormal responses to physiological insulin concentrations (Kahn 1978), is the hallmark of the MONW phenotype.

Owing to the lack of a consistent definition, there is some variability among studies in the phenotypic characterization of MONW subjects (Teixeira et al. 2015). This is further complicated by the small but significant differences in BMI and, more commonly, percent body fat between groups of metabolically healthy and unhealthy lean subjects, with MONW subjects being always somewhat "fatter" (even though within the "lean" range) (Ruderman et al. 1998; Di Renzo et al. 2006; Badoud et al. 2015; Luo et al. 2015; Indulekha et al. 2015; Dvorak et al. 1999; Conus et al. 2004; De Lorenzo et al. 2007). Likewise, BMI and body fat are typically greater in relatively insulin-resistant (e.g. Indian) than in relatively insulin-sensitive (e.g. Chinese) individuals in studies reporting on ethnic differences in insulin action among lean people (Khoo et al. 2014; Khoo et al. 2011). This in itself could be responsible for the differences observed in metabolic function. There is considerable (~2-fold range) variability between individuals in the percent body fat (Gallagher et al. 2000; Gallagher et al. 1996) and the insulin-mediated glucose disposal rate (a direct measure of whole-body insulin sensitivity) (Bradley, Magkos, and Klein 2012) for the same BMI value within the normal-weight range (i.e. BMI <25 kg/m2), so that people with the same BMI can have very different body fat and insulin sensitivity without this necessarily being associated with the presence or absence of generalized metabolic dysfunction. Even among lean and metabolically-healthy Asians, total body fat is a major correlate of insulin-mediated glucose disposal (Rattarasarn et al. 2003). It is thus possible that some of the reported differences between metabolically healthy and unhealthy lean subjects arise from normal variability and the differences in body fat between groups, rather than being an inherent characteristic of the MONW phenotype. In support of this possibility, when metabolically healthy and unhealthy lean subjects (defined as those having normal and impaired glucose tolerance, respectively) were retrospectively matched on total body fat, there were no differences between phenotypes in circulating concentrations of metabolic and inflammatory markers (i.e. high-density lipoprotein (HDL)-cholesterol, triglycerides, free fatty acids, C-reactive protein, adiponectin, and leptin) (Fontana, Klein, and Holloszy 2010). There are no studies that prospectively recruited groups of metabolically healthy and unhealthy lean individuals matched on BMI and percent body fat. A deeper understanding of the MONW phenotype, as proposed here, is important to dissect the metabolic abnormalities that are inherent to the phenotype from those merely associated with differences in total body fat. This will allow for proper identification and more efficient therapeutic targeting of MONW individuals, who are at greater risk for cardiometabolic disease.

Little is known about possible interventions for improving metabolic function in MONW subjects. It is well established that diet-induced weight loss can improve body composition and many of the cardiometabolic abnormalities in most obese patients (e.g. decreases total body fat, intra-abdominal adipose tissue, and ectopic fat deposition in liver and muscle; increases insulin sensitivity; improves blood lipid profile; and reduces blood pressure) (Dattilo and Kris-Etherton 1992; de Leiva 1998; Goldstein 1992; Kirk et al. 2009; Muscelli et al. 1997; Pi-Sunyer 1993; Pasanisi et al. 2001; Escalante-Pulido et al. 2003; Mazzali et al. 2006; Klein, Wadden, and Sugerman 2002), so that a moderate 10 % weight loss has become the cornerstone of obesity treatment (Jensen et al. 2014). However, MONW individuals are by definition lean, so recommending even moderate amounts of weight loss may not be a feasible therapeutic target (Miller and Parsonage 1975). It is therefore important to better understand the metabolic effects of smaller amounts of weight loss. Recently, the principal investigator conducted a randomized controlled trial to evaluate the effects of mild weight loss (5 % of initial body weight) on cardiometabolic function in non-Asian subjects with obesity and insulin resistance and found that even this small amount of weight loss decreases fat deposition in the liver and the intra-abdominal area, and increases insulin action in skeletal muscle, liver, and adipose tissue (Magkos et al. 2016). These results demonstrate that mild weight loss can improve many cardiometabolic abnormalities in metabolically-unhealthy obese subjects, but whether the same holds true for metabolically-unhealthy lean subjects is not known. A small, non-randomized, single-arm study in 7 lean, insulin-resistant offspring of parents with type 2 diabetes reported that modest ~6 % diet-induced weight loss reduced intra-myocellular lipid (i.e. fat within skeletal muscle fibers) content and increased insulin-mediated glucose disposal rate (both by ~30 % compared with baseline values), but did not significantly affect intra-abdominal adipose tissue volume or liver fat content (Petersen et al. 2012). It is thus not known whether mild diet-induced weight loss produces similar changes in body composition, fat distribution, and metabolic function in lean versus obese metabolically unhealthy subjects.

Study Design

Related Conditions & MeSH terms

See also
  Status Clinical Trial Phase
Recruiting NCT02744183 - Training in the Fasted State, Glucose Metabolism and Energy Balance N/A
Completed NCT02697305 - Incretin Effects of Branched Chain Amino Acids N/A
Completed NCT01737164 - Effect of Age on Glucose and Lipid Metabolism N/A
Terminated NCT00935766 - Effect of Fish Oil (Omega-3 Fatty Acids) on Arteries Phase 3
Active, not recruiting NCT03013296 - Delinieation of GIP's Effects During a Meal in Humans Using GIP Receptor Antagonisation (GA-4) N/A
Completed NCT01708681 - Lean Seafood Intake and Postprandial Metabolism N/A
Completed NCT01757756 - Multiple Dose Safety Toelrability, Pharmacokinetics and Midazolam Interaction In Healthy Overweight And Obese Subjects Phase 1
Active, not recruiting NCT01803776 - Physical Activity and Nutrition in Children Study N/A
Completed NCT02296372 - Accuracy, Feasibility and Acceptance of Continuous Glucose Monitoring in Critically Ill Patients N/A
Active, not recruiting NCT01396954 - Cirrhosis-Diabetes N/A
Completed NCT00192621 - Seronegatives and Metabolic Abnormalities Protocol 2 (SAMA002): Study to Compare the Effect of Kaletra and Combivir® in HIV-Negative Healthy Subjects Phase 4
Not yet recruiting NCT03305783 - The Influence of Cholecystectomy on Secretion of Incretin Hormones N/A
Recruiting NCT02782208 - Lipolytic Effects of GH in Hypopituitary Patients in Vivo N/A
Recruiting NCT02523131 - Home Testing of Day and Night Closed Loop With Pump Suspend Feature N/A
Recruiting NCT02102737 - Comparison of A New Technique of Measure of the Insulin Resistance By Scintigraphy With the Reference Technique Phase 2
Completed NCT01681173 - OPTIFIT-Optimal Fiber Trial for Diabetes Prevention N/A
Completed NCT00464529 - Accuracy and Safety of Real Time Continuous Glucose Monitoring Before, During and After Surgery N/A
Completed NCT00221156 - Acarbose and Secondary Prevention After Coronary Stenting Phase 3
Recruiting NCT02925299 - Day and Night Closed-loop in Young People With Type 1 Diabetes N/A
Recruiting NCT03041363 - Treatment Development of Triheptanoin (G1D) Phase 1