Metabolically Healthy Obesity: Epidemiology, Mechanisms, and Clinical Implications PDF

Personal View Metabolically healthy obesity: epidemiology, mechanisms, and clinical implications Norbert Stefan, Hans-Ulrich Häring, Frank B Hu, Matthias B Schulze Lancet D...

Personal View Metabolically healthy obesity: epidemiology, mechanisms, and clinical implications Norbert Stefan, Hans-Ulrich Häring, Frank B Hu, Matthias B Schulze Lancet Diabetes Endocrinol Obesity has become a worldwide epidemic that poses substantial health problems for both individuals and society. 2013; 1: 152–62 However, a proportion of obese individuals might not be at an increased risk for metabolic complications of obesity Published Online and, therefore, their phenotype can be referred to as metabolically healthy obesity. This novel concept of metabolically August 30, 2013 healthy obesity might become increasingly important to stratify individuals in the clinical treatment of obesity. http://dx.doi.org/10.1016/ S2213-8587(13)70062-7 However, no universally accepted criteria exist to define metabolically healthy obesity. Furthermore, many questions Department of Internal have been raised regarding the biological basis of this phenotype, the transitory nature of metabolically healthy Medicine IV, University obesity over time, and predictors of this phenotype. We describe the observational studies that gave rise to the idea of Hospital Tübingen, Tübingen, metabolically healthy obesity and the key parameters that can help to distinguish it from the general form of obesity. Germany (Prof N Stefan MD, We also discuss potential biological mechanisms underlying metabolically healthy obesity and its public health and Prof H-U Häring MD); Institute of Diabetes Research and clinical implications. Metabolic Diseases, Tübingen, Germany (N Stefan, Introduction important pathomechanisms of metabolic diseases and H-U Häring); German Center for The health consequences of obesity are well documented. also of certain types of cancer.14,15 In addition to body fat Diabetes Research, Neuherberg, Germany In particular, the worldwide increase in the incidence of distribution and insulin resistance, other metabolic risk (N Stefan, H-U Häring, type 2 diabetes, cardiovascular disease, and several types factors might also be useful in the characterisation of Prof M B Schulze DrPH); of cancer is thought to be largely attributed to the obesity metabolically healthy obesity in view of their well- Department of Nutrition, Harvard School of Public epidemic.1–4 Therefore, prevention and treatment of established association with risk, including lipid profiles, Health, Boston, MA, USA obesity to reduce risk of chronic diseases at the population blood pressure, inflammation, or physical fitness. (Prof F B Hu MD); Channing and individual level is crucial. Although the deleterious In this Personal View we describe observational data Division of Network Medicine, metabolic effects of obesity are widely recognised at that gave rise to the idea of metabolically healthy obesity. Department of Medicine, Brigham and Women’s Hospital population level, individual differences exist in metabolic We then discuss the key parameters that might help to and Harvard Medical School, responses to obesity. Findings from many studies show distinguish metabolically healthy obesity from the Boston, MA, USA (F B Hu); that a subgroup of obese individuals might be protected general form of obesity, such as smaller waist Department of Epidemiology, from metabolic complications of obesity or might be at circumference, increased physical fitness, decreased Harvard School of Public Health, Boston, MA, USA substantially lower risk than expected for their degree of insulin resistance, and low prevalence of metabolic risk (F B Hu); and Department of obesity. This subgroup has been described as having factors despite a high BMI. We also discuss potential Molecular Epidemiology, metabolically healthy obesity.5–-9 Many questions have biological mechanisms underlying this phenotype and German Institute of Human been raised regarding the biological basis, transitory its clinical implications. Nutrition Potsdam- Rehbruecke, Nuthetal, nature, and predictors of metabolically healthy obesity. Germany (M B Schulze) Findings from epidemiologic studies have shown that Observational data supporting the idea of Correspondence to: increased waist circumference is associated with metabolically healthy obesity Prof Matthias B Schulze, German mortality and cardiovascular disease independent of Individuals with metabolically healthy obesity are a subset Institute of Human Nutrition overall adiposity.4,10 Additionally, data from several small of individuals who meet the standard BMI cutoff point for Potsdam-Rehbruecke, Department of Molecular studies suggested that some obese people are not insulin obesity (≥30 kg/m²), but are regarded as metabolically Epidemiology, 14558 Nuthetal, resistant.5,11–13 This finding was unexpected, because healthy because they do not have other major cardiovascular Germany generally, a strong positive association exists between risk factors (figure 1). This subgroup is believed to be at [email protected] body-mass index (BMI) and insulin resistance. Insulin much lower risk of cardiovascular morbidity and mortality resistance is thought to represent one of the most Panel 1: Criteria to define metabolically healthy obesity in BMI epidemiological studies Normal weight Overweight Obese Absence of abdominal obesity on the basis of waist Metabolically Metabolically Metabolically healthy Metabolically healthy Metabolically healthy healthy normal weight overweight obese (MHO) circumference (men ≤102 cm, women ≤88 cm) abnormal Absence of metabolic syndrome components—eg, normal Metabolically Metabolically unhealthy Metabolically unhealthy Metabolically unhealthy unhealthy normal weight overweight obese (MUHO) blood pressure, normal lipid values, normal fasting glucose concentrations (at times also including normal C-reactive protein concentrations) Figure 1: Classification according to body fat on the basis of BMI and metabolic health Insulin sensitive on the basis of the homoeostatic model Absence and presence of major cardiovascular risk factors allows stratification of normal weight, overweight, and assessment of insulin resistance (HOMA-IR) obese individuals into metabolically healthy and metabolically unhealthy. Normal weight: BMI 18·5–24·9 kg m2; High level of cardiorespiratory fitness overweight: BMI 25·0–29·9 kg/m2; obese: BMI ≥30·0 kg/m2. BMI=body-mass index. 152 www.thelancet.com/diabetes-endocrinology Vol 1 October 2013  Personal View compared with obese individuals with major cardiovascular especially accumulation of body fat in the abdominal risk factors, who can consequently be judged as being region. However, findings from large prospective cohort metabolically at risk or metabolically unhealthy obese. The studies show that the association of risk of cardiovascular term metabolically healthy obesity implies that individuals disease and death with increasing waist circumference is with this phenotype are not at higher risk of cardiovascular stronger in non-obese individuals compared with obese disease than non-obese individuals. While the classification individuals.4,10,16 Waist circumference measurement of metabolically healthy obesity and metabolically might therefore be more useful for the identification of unhealthy obesity both require determination of BMI and high metabolic risk for normal weight and moderately cardiovascular risk factors at the same time, prospective obese patients than for severely obese patients. The studies are needed to demonstrate that individuals strong association between BMI and waist circumference classified as metabolically healthy obese are indeed makes it unlikely that both provide different answers protected against the cardiovascular complications of (figure 2). Measures that are less strongly related to BMI, obesity. Substantial evidence has accumulated from but show metabolic risk, might be more informative for prospective cohort studies in which sub-groups of obese characterisation of metabolically healthy obesity. Figure 2 individuals were compared with regard to their risk for shows that although insulin resistance (assessed by the cardiovascular disease or mortality, although how homoeostatic model assessment of insulin resistance subgroups have been defined varies largely across studies [HOMA-IR]) is strongly associated with BMI, it has much (panel 1). Here we summarise the evidence from larger variation compared with waist circumference at prospective studies on mortality. any given BMI among obese individuals. This results in a group of individuals with insulin resistance below the Body fat distribution and mortality in obese individuals population average, even though they are morbidly An obvious approach to define metabolically healthy obese. Although waist-to-hip ratio is less strongly obesity is a more detailed anthropometric character- associated with BMI than is waist circumference, this isation of obese individuals in addition to BMI. The ratio is not more informative in quantifying measurement of waist circumferences allows better cardiovascular risk among obese individuals than is characterisation of body fat distribution than BMI, waist circumference.4,10 Thus, the use of waist 6 Waist circumference HOMA-IR 5 4 Waist circumference and HOMA-IR in SD-units 3 Metabolically abnormal 2 1 0 –1 –2 –3 15 20 25 30 35 40 45 50 55 60 65 70 BMI Figure 2: Association between BMI, waist circumference, and insulin resistance in the Tübingen Family study and the Tübingen Lifestyle Intervention Program (n=2472)8,17 Association between BMI and waist circumference is strong. By contrast, for insulin resistance—estimated from the homoeostatic model assessment—large variation is recorded for any given BMI among obese individuals. HOMA-IR=homoeostatic model assessment of insulin resistance. www.thelancet.com/diabetes-endocrinology Vol 1 October 2013 153  Personal View circumference or waist to hip ratio alone is not sufficient define metabolically healthy obesity. Results from several to establish the metabolically healthy obesity phenotype. prospective studies show that only obese, unfit individuals, but not obese, fit individuals, are at higher mortality risk Physical fitness, activity, and mortality among obese than are normal weight fit individuals (table 1). Fitness is individuals measured by a treadmill exercise test and categories of fit Besides components of metabolic syndrome and insulin and unfit study participants are based on varying study- resistance, physical fitness is an alternative means to specific percentiles.24–28 However, whether fitness is Study Participants Definition of unhealthy Reference Outcome Relative risk (95% Relative risk (95% Adjustment metabolic phenotype group CI) metabolically CI) metabolically healthy obese unhealthy obese Studies using insulin resistance to define metabolic health Kuk et al, National 4602 men HOMA-IR ≥2·5 Normal weight Total mortality 2·58 (1·0–6·7) 3·09 (1·6–6·2) Age, sex, income, smoking status, 200918 Health and and women insulin sensitive ethnicity, and alcohol consumption Nutrition Examination Survey III, USA Arnlov Uppsala 1758 men HOMA-IR in top 25% of the Normal weight Total mortality 2·04 (1·25–3·32) 2·21 (1·64–2·99) Age, smoking, and LDL cholesterol et al, Longitudinal distribution in participants insulin sensitive Cardiovascular 1·80 (0·79–4·08) 2·87 (1·87–4·42) 201019 Study of Adult without diabetes (>3·43) mortality Men, Sweden Calori et al, Cremona 2011 men HOMA-IR ≥2·5 Nonobese Total mortality 0·99 (0·46–2·11) 1·40 (1·08–1·81) Age and sex 201120 Study, Italy and women insulin sensitive Cardiovascular 0·73 (0·18–3·00) 1·61 (1·10–2·36) mortality Bo et al, Asti 1658 men HOMA-IR >2·5 Normal weight Total mortality 1·57 (0·93–2·21) 1·51 (1·04–1·98) Age, sex and smoking 201221 (northwest and women insulin sensitive Cardiovascular 2·95 (1·03–3·98) 2·43 (1·57–3·29) Italy) mortality Durward National 4373 men HOMA-IR ≥2·5 Normal weight Total mortality 1·42 (0·6–3·2) 2·07 (1·3–3·4) Sex, age, income, education, race and et al, 201222 Health and and women insulin sensitive ethnicity, smoking status, alcohol Nutrition consumption, marital status, leisure Examination time physical activity, and Survey III, USA menopausal status in women Hinnouho Whitehall II 5269 men HOMA-IR in top 25% of the Normal weight Total mortality 1·08 (0·67–1·74) 2·14 (1·56–2·94) Age, sex, occupation, physical et al, 201323 Study, UK and women distribution metabolic Cardiovascular 1·04 (0·41–2·66) 2·63 (1·51–4·60) activity, smoking, alcohol, fruit and healthy mortality vegetable consumption, marital Matsuda index in lower 75% Normal weight Total mortality 2·30 (1·13–4·70) 1·57 (1·08–2·28) status, ethnicity of the distribution metabolic Cardiovascular 1·89 (0·43–8·33) 1·75 (0·89–3·41) healthy mortality Studies using cardiorespiratory fitness to define metabolic health Wei et al, Aerobics 25 714 men Metabolic equivalents Normal weight Total mortality 1·1 (0·8–1·5) 3·1 (2·5-3·8) Age and calendar year of baseline 199924 Center during maximum treadmill fit Cardiovascular 1·6 (1·0–2·8) 5·0 (3·6–7·0) examination Longitudinal exercise test; value