TY - JOUR
T1 - Prediction of type 2 diabetes mellitus using noninvasive MRI quantitation of visceral abdominal adiposity tissue volume
AU - Wang, Meng
AU - Luo, Yanji
AU - Cai, Huasong
AU - Xu, Ling
AU - Huang, Mengqi
AU - Li, Chang
AU - Dong, Zhi
AU - Li, Zi Ping
AU - Feng, Shi Ting
N1 - Funding Information:
Funding: This study was supported by grants from the National Natural Science Foundation of China (No. 81801761, 81771908, 81571750, and 81770654).
Publisher Copyright:
© Quantitative Imaging in Medicine and Surgery. All rights reserved.
PY - 2019
Y1 - 2019
N2 - Background: The correlation between visceral adipose tissue volume (VATV), hepatic proton-density fat fraction (PDFF), and pancreatic PDFF has been previously studied to predict the presence of type 2 diabetes mellitus (T2DM). This study investigated VATV quantitation in patients with T2DM, prediabetes, and normal glucose tolerance (NGT) using MRI to assess the roles of VATV, hepatic, and pancreatic PDFF in predicting the presence of T2DM. Methods: Forty-eight patients with a new clinical diagnosis of T2DM (n=15), prediabetes (n=17), or NGT (n=16) were included and underwent abdominal magnetic resonance imaging (MRI) scanning with the iterative decomposition of water and fat with echo asymmetry and least square estimation image quantification (IDEAL-IQ) sequencing. VATV was obtained at the level of the 2nd and 3rd lumbar vertebral bodies (VATV L2 and VATV L3) where the sum of VATV L2 and VATV L3 (total VATV) were computed, respectively. Also, pancreatic and hepatic fat content was quantified by measuring the PDFF. The receiver operating characteristic (ROC) curve and binary logistics regression model analysis were employed to evaluate their ability to predict the presence of T2DM. Results: The VATV L2, VATV L3, and total VATV values of the T2DM group were significantly higher than the prediabetes and NGT groups (P<0.05). There was no statistically significant difference between the values of VATV L2, VATV L3, and total VATV between the prediabetes and NGT groups (P>0.05). The ROC curve showed the areas under the curve for VATV L2, VATV L3, total VATV, hepatic PDFF, and pancreatic PDFF were 0.76, 0.80, 0.80, 0.79, and 0.75, respectively, in predicting the presence of T2DM (P<0.01). The ROC curves of VATV L2, VATV L3, total VATV, hepatic PDFF, and pancreatic PDFF failed to predict the presence of prediabetes and NGT (P>0.05). The binary logistics regression model analysis revealed that only VATV L3 was independently associated with the incidence of T2DM (P=0.01 and OR =1.01). The sensitivity, specificity, and total accuracy were 80.00%, 88.20%, and 84.40%, respectively. Conclusions: Compared with hepatic PDFF, pancreatic PDFF, VAVT L2, and total VATV, VAVT L3 was the better predictor of T2DM.
AB - Background: The correlation between visceral adipose tissue volume (VATV), hepatic proton-density fat fraction (PDFF), and pancreatic PDFF has been previously studied to predict the presence of type 2 diabetes mellitus (T2DM). This study investigated VATV quantitation in patients with T2DM, prediabetes, and normal glucose tolerance (NGT) using MRI to assess the roles of VATV, hepatic, and pancreatic PDFF in predicting the presence of T2DM. Methods: Forty-eight patients with a new clinical diagnosis of T2DM (n=15), prediabetes (n=17), or NGT (n=16) were included and underwent abdominal magnetic resonance imaging (MRI) scanning with the iterative decomposition of water and fat with echo asymmetry and least square estimation image quantification (IDEAL-IQ) sequencing. VATV was obtained at the level of the 2nd and 3rd lumbar vertebral bodies (VATV L2 and VATV L3) where the sum of VATV L2 and VATV L3 (total VATV) were computed, respectively. Also, pancreatic and hepatic fat content was quantified by measuring the PDFF. The receiver operating characteristic (ROC) curve and binary logistics regression model analysis were employed to evaluate their ability to predict the presence of T2DM. Results: The VATV L2, VATV L3, and total VATV values of the T2DM group were significantly higher than the prediabetes and NGT groups (P<0.05). There was no statistically significant difference between the values of VATV L2, VATV L3, and total VATV between the prediabetes and NGT groups (P>0.05). The ROC curve showed the areas under the curve for VATV L2, VATV L3, total VATV, hepatic PDFF, and pancreatic PDFF were 0.76, 0.80, 0.80, 0.79, and 0.75, respectively, in predicting the presence of T2DM (P<0.01). The ROC curves of VATV L2, VATV L3, total VATV, hepatic PDFF, and pancreatic PDFF failed to predict the presence of prediabetes and NGT (P>0.05). The binary logistics regression model analysis revealed that only VATV L3 was independently associated with the incidence of T2DM (P=0.01 and OR =1.01). The sensitivity, specificity, and total accuracy were 80.00%, 88.20%, and 84.40%, respectively. Conclusions: Compared with hepatic PDFF, pancreatic PDFF, VAVT L2, and total VATV, VAVT L3 was the better predictor of T2DM.
KW - Magnetic resonance imaging (MRI)
KW - Quantitation
KW - Type 2 diabetes mellitus (T2DM)
KW - Visceral adipose tissue volume (VATV)
UR - http://www.scopus.com/inward/record.url?scp=85068545211&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85068545211&partnerID=8YFLogxK
U2 - 10.21037/qims.2019.06.01
DO - 10.21037/qims.2019.06.01
M3 - Article
C2 - 31367561
AN - SCOPUS:85068545211
SN - 2223-4292
VL - 9
SP - 1076
EP - 1086
JO - Quantitative Imaging in Medicine and Surgery
JF - Quantitative Imaging in Medicine and Surgery
IS - 6
ER -