Dark and Light Side of Obesity: Mortality of Metabolically Healthy Obese People

Lev M Berstein


Expert Rev Endocrinol Metab. 2012;7(6):629-632. 

In This Article

Discussion & Expert Commentary

The authors denote that their results are consistent with some[2,3] but not all previous work,[4] and explain this by the differences in length of follow-up: studies with shorter follow-up (less than 15 years) were usually in line with the present study. Metabolically healthy participants at baseline could go on to acquire the abovedescribed metabolic risk factors over a longer observational period, while on the other hand it also cannot be excluded that people with 'standard' (unhealthy) obesity could lose some of these factors during prolonged follow-up; for example, due to the changes in lifestyle, including diet, physical activity, and so on.

Among some potential limits of the publication under evaluation, its authors acknowledge that they may not have measured accumulation of visceral fat as well as they did not study adipose tissue morphology and function. Of course, methodologically it is also important that blood samples were not taken after fasting, and such indices as glucose and insulin were not measured. Besides, not very precise categorization of participants on the basis (as the authors describe it) 'of two or more from any recommended five risk factors' could entail some additional bias. Obviously, looking at outcomes of different combinations of involved factors (apart from waist circumference or BMI only) would give useful supplementary information and could shed light into the whole problem, since the prevalence of the metabolically healthy obese phenotype in different populations depends substantially on the definition and methods used.[1,3]

Speaking about definitions and terminology, it is worthy to mention publications of Sims,[5] on the one hand, and Ruderman et al.[6] on the other hand, who in the 1980s invented such notions as MHO and metabolically obese normal weight (MONW), respectively, and gave their first definitions, which were somewhat specified by Karelis et al..[7] Hamer and Stamatakis[1] do not use the abbreviation MONW and prefer the term 'metabolically healthy nonobese', which (since the boundaries of BMI in this paper lie between 18.0 and 29.9 kg/m2) is in its last part actually not the same that 'normal weight'.

Although one of the main conclusions of the present publication[1] is that metabolic risk factors are more important than overall adiposity, and this conclusion is illustrated only by the fact that obesity is not always associated with its allegedly typical markers, one needs to remember that according to the Emerging Risk Factors Collaboration study obesity indices correlated independently with such laboratory risk factors as lipids, fasting glucose and CRP.[8] Accordingly, the clinical consequences of the two situations (obesity plus metabolic factors and obesity minus metabolic factors) need further comparison and exploration, especially when considering the various end points apart from cardiovascular and all-cause mortality.

An example of cancer and diabetes deserves to be mentioned here. Thus, except for our preliminary data showing that mortality among endometrial cancer patients during an average 9.6 ± 0.3 year-long follow-up period was higher not upon 'standard' obesity featuring insulin resistance, but in normal-weight or MHO patients, we are aware of only one study conducted in Cremona (Italy) where cancer mortality adjusted for age and sex was – in contrast to the aforementioned – lower among MHO subjects than among obese insulin-resistant people.[3] The evidence from a meta-analysis suggests that the ability of metabolic syndrome adjusted for BMI to predict the future risk of all-cause mortality, cardiovascular disease and diabetes varies significantly with population-attributable fractions from 6 to 7% for all-cause mortality, 12 to 17% for cardiovascular disease and 30 to 52% for diabetes.[9] At the same time, as was shown very recently, adults who were normal weight at the time of incident diabetes had higher total, cardiovascular and noncardiovascular mortality than adults who were overweight or obese.[10] On the other hand, according to Hamer and Stamatakis' observation[1] in metabolically healthy obese groups, notable distinctions between cardiovascular and all–cause mortality were not revealed (hazard ratio: 1.64 and 1.79, respectively, vs metabolically healthy nonobese people), Although it is well known that cardiovascular disease may compete with cancer as the leading cause of death[11] and there are important distinctions in predisposition to these two groups of pathology in both offspring and parents, starting from the fetal programming and birthweight level.[12,13]

On the whole, the data discussed above suggest that there is a spectrum of transitions between so-called 'normal weight' and different states encompassed by the notion of obesity. These transitions may have important clinical implications, supporting the idea proposed for many years ago by Albrink[14] that avoidance of both factors (excess body fat and metabolic disturbances based on the insulin resistance phenomenon) should be helpful in reaching better health. Nevertheless, as correctly pointed out by Hamer and Stramatakis,[1] the stratification of obese individuals based on their metabolic phenotype may be important to identify those who are to be prioritized for early lifestyle and, possibly, pharmacological intervention.