The obesity epidemic; population levels of visceral adipose tissue and trends in body composition. Insights from The Tromsø Study

Abstract

<p><i>Background:</i> Simultaneously with the growing obesity epidemic, cardiometabolic risk have decreased in high-income countries. Do the most frequently used general and abdominal obesity measures (BMI and waist circumference) fail to represent actual obesity and obesity-related health risk? <p><i>Aims:</i> 1) present reference- and threshold values for DXA-derived visceral adipose tissue (VAT) in a general adult population, 2) compare VAT with anthropometric measures and their association with metabolic syndrome (MetS), and 3) study trends in total body fat-, lean-, and VAT mass during the last two decades. <p><i>Methods:</i> We included participants with measurements of VAT and cardiometabolic risk factors attending the population-based Tromsø Study. In paper Ⅰ and Ⅱ we included 3675 participants aged 40-84 years from Tromsø 7 (2015-2016). We presented threshold values of VAT and compared VAT with anthropometrics in association with MetS. In Paper III we included 1662, 901 and 3670 participants from the fifth (2001), sixth (2007-2008) and seventh (2015-2016) survey of the Tromsø Study, respectively, to study secular and longitudinal trends in fat-, lean-, and VAT mass. <p><i>Results:</i> The VAT thresholds were ≥1134 grams, index ≥0.44, and ≥40%, in women, and ≥1859 grams, index ≥0.55 and ≥61% in men. VAT correlated strongly with anthropometric measures. VAT was statistically stronger than anthropometrics in prediction of MetS, but the clinical differences in the prediction were minor. VAT and fat mass increased from 2001 to 2015-2016, with a larger increase between 2007-2008 to 2015-2016, and in the youngest birth-cohort (40-49 years in 2001, particularly in women). <p><i>Conclusion:</i> The threshold values are useful for comparison with future studies, clinics or other facilities using DXA-derived VAT. Due to the minor clinical differences between VAT and anthropometric measures in predicting MetS, and the similar time trends in VAT and anthropometrics, the anthropometric measures are regarded as satisfactory substitutes for VAT in population-based studies.