Maria Pilar Aparisi Gómez1, Federico Ponti2, Daniele Mercatelli2, Chiara Gasperini3, Alessandro Napoli4, Giuseppe Battista3, Stefano Cariani5, Giulio Marchesini6, Alberto Bazzocchi7. 1. Department of Radiology, Auckland City Hospital, Auckland, New Zealand; Department of Radiology, Hospital Nueve de Octubre, Valencia, Spain. 2. Diagnostic and Interventional Radiology, IRCCS Istituto Ortopedico Rizzoli, Bologna, Italy. 3. Department of Specialized, Diagnostic, and Experimental Medicine, University of Bologna, Sant'Orsola - Malpighi Hospital, Bologna, Italy. 4. Radiology Section, Department of Radiological, Oncological and Anatomopathological Sciences, "Sapienza" University of Rome, Rome, Italy. 5. Obesity Surgery Center, Department of Emergency-Urgency, General Surgery and Transplantation, University of Bologna, Sant'Orsola - Malpighi Hospital, Bologna, Italy. 6. Unit of Metabolic Diseases & Clinical Dietetics, University of Bologna, Sant'Orsola - Malpighi Hospital, Bologna, Italy. 7. Diagnostic and Interventional Radiology, IRCCS Istituto Ortopedico Rizzoli, Bologna, Italy. Electronic address: abazzo@inwind.it.
Abstract
OBJECTIVE: The aim of this study was to review the existence and types of correlations between body composition densitometric parameters and laboratory values associated to cardiometabolic risk. METHODS: We retrospectively analyzed data from 316 individuals in the weight range from normality to super-obesity, submitted to total body dual-energy x-ray absorptiometry (DXA) scans and routine biochemistry at S.Orsola-Malpighi Hospital from June 2010 to March 2014. The study included 182 women, 45.8 ± 13.4 y of age, with a body mass index (BMI) of 31.5 (± 11) kg/m2 (group F) and 134 men, 45.4 ± 13.6 y of age, with a BMI of 27.6 (± 7.8) kg/m2 (group M). All patients underwent whole-body scan (Lunar iDXA, GE Healthcare, Madison, WI, USA) and laboratory analysis (blood fasting glucose, total cholesterol, high-density lipoprotein cholesterol, tricylglycerides [TGs], aspartate aminotransferase, and alanine aminotransferase). Correlation between laboratory values and total body and regional fat mass (including visceral adipose tissue [VAT] and subcutaneous adipose tissue in the android region), and lean mass parameters were analyzed with linear and stepwise regressions analysis (significance limit, P < 0.05). Receiver operating characteristic curves were performed to assess the accuracy of the best-fit DXA parameter (VAT) to identify at least one laboratory risk factor. RESULTS: In both groups, BMI and densitometric parameters showed a linear correlation with fasting blood glucose and TG levels and an inverse correlation with high-density lipoprotein cholesterol (P < 0.05), whereas no correlation was observed with total cholesterol levels. The only densitometric parameter retained in the final model of stepwise multiple regression was VAT for fasting blood glucose (group F: β = 0.4627, P < 0.0001; group M: β = 0.6221, P < 0.0001) and TG levels (group F: β = 0.4931, P < 0.0001; group M: β = 0.1990, P < 0.0261) independently of BMI. The optimal cutoff points of VAT to identify the presence of at least one laboratory risk factor were >1395 g and >1479 cm3 for men and >1281 g and >1357 cm3 for women. CONCLUSIONS: DXA analysis of VAT is associated with selected laboratory parameters used for the evaluation of cardiometabolic risk and could be per se a helpful parameter in the assessment of clinical risk.
OBJECTIVE: The aim of this study was to review the existence and types of correlations between body composition densitometric parameters and laboratory values associated to cardiometabolic risk. METHODS: We retrospectively analyzed data from 316 individuals in the weight range from normality to super-obesity, submitted to total body dual-energy x-ray absorptiometry (DXA) scans and routine biochemistry at S.Orsola-Malpighi Hospital from June 2010 to March 2014. The study included 182 women, 45.8 ± 13.4 y of age, with a body mass index (BMI) of 31.5 (± 11) kg/m2 (group F) and 134 men, 45.4 ± 13.6 y of age, with a BMI of 27.6 (± 7.8) kg/m2 (group M). All patients underwent whole-body scan (Lunar iDXA, GE Healthcare, Madison, WI, USA) and laboratory analysis (blood fasting glucose, total cholesterol, high-density lipoprotein cholesterol, tricylglycerides [TGs], aspartate aminotransferase, and alanine aminotransferase). Correlation between laboratory values and total body and regional fat mass (including visceral adipose tissue [VAT] and subcutaneous adipose tissue in the android region), and lean mass parameters were analyzed with linear and stepwise regressions analysis (significance limit, P < 0.05). Receiver operating characteristic curves were performed to assess the accuracy of the best-fit DXA parameter (VAT) to identify at least one laboratory risk factor. RESULTS: In both groups, BMI and densitometric parameters showed a linear correlation with fasting blood glucose and TG levels and an inverse correlation with high-density lipoprotein cholesterol (P < 0.05), whereas no correlation was observed with total cholesterol levels. The only densitometric parameter retained in the final model of stepwise multiple regression was VAT for fasting blood glucose (group F: β = 0.4627, P < 0.0001; group M: β = 0.6221, P < 0.0001) and TG levels (group F: β = 0.4931, P < 0.0001; group M: β = 0.1990, P < 0.0261) independently of BMI. The optimal cutoff points of VAT to identify the presence of at least one laboratory risk factor were >1395 g and >1479 cm3 for men and >1281 g and >1357 cm3 for women. CONCLUSIONS: DXA analysis of VAT is associated with selected laboratory parameters used for the evaluation of cardiometabolic risk and could be per se a helpful parameter in the assessment of clinical risk.