Loading ...

TOPLINE:

Automated analysis of CT-derived body composition parameters, especially the visceral fat index, can predict the new-onset risk for type 2 diabetes (T2D) better than traditional anthropometric and clinical risk models, finds a large study of CT data taken for other purposes.

METHODOLOGY:

• CT scans performed for clinical indications also contain data that can potentially predict cardiovascular diseases and all-cause mortality, which are traditionally diagnosed using conventional modalities, but the predictive ability of individual imaging parameters for cardiometabolic diseases remains largely unexplored.
• Researchers evaluated fully automated CT markers associated with T2D prevalence and related cardiometabolic comorbidities at baseline in a cross-sectional study of CT data from 32,166 Korean adults ( mean age, 44.6 years; 90% men) who underwent noncontrast fluorine-18 fluorodeoxyglucose PET-CT from 2012 to 2015 as part of a comprehensive health examination.
• In a subgroup cohort of 27,298 individuals who were initially free from diabetes, they evaluated CT-derived risk factors for new cases of T2D (median follow-up, 7.3 years).
• Fully automated CT markers included visceral and subcutaneous fat, muscle, bone density, and liver fat, all normalized to height, as well as aortic calcification.
• T2D was defined by fasting glucose, A1c, or current use of insulin or glucose-lowering medications. Associated comorbidities included metabolic syndrome, sarcopenia, osteoporosis, fatty liver, and coronary artery calcium.

TAKEAWAY:

• At baseline, the overall prevalence of T2D was 5.8% (6% in men; 3.9% in women).
• Visceral fat index, a CT-derived parameter, was a better measure of prevalent diabetes in both men and women than body mass index, a more traditional marker (P < .001); it yielded an area under the curve (AUC) of 0.70 (95% CI, 0.68-0.71) and 0.82 (95% CI, 0.78-0.85) in identifying prevalent diabetes in men and women, respectively.
• Over the follow-up period, 2456 participants developed incident diabetes, with visceral fat the single best imaging marker, and the combination of visceral fat, muscle area indices, liver fat fraction, and aortic calcification predicting incident diabetes with an AUC of 0.69 (95% CI, 0.68-0.71) in men and 0.83 (95% CI, 0.78-0.87) in women.
• The automated CT-derived markers also identified fatty liver, metabolic syndrome, coronary artery calcium scores > 100, sarcopenia, and osteoporosis effectively, with AUCs ranging from 0.80 to 0.95.

IN PRACTICE:

"Achieving more efficient and safer approaches through reduced radiation exposure and targeted multiorgan assessments remains a necessity, and caution is warranted when considering the clinical applicability of these findings for practice," the authors wrote.

SOURCE:

The study was led by Yoosoo Chang, MD, PhD, and Soon Ho Yoon, MD, PhD, from Kangbuk Samsung Hospital, Sungkyunkwan University (SKKU) School of Medicine, Seoul, Republic of Korea, and was published online with an accompanying editorial in Radiology.

LIMITATIONS: 

The study's generalizability might be limited as it focused on young and middle-aged Korean men, potentially not representing the broader population and underrepresenting women. Diagnosing T2D using a single measurement of fasting levels of glucose and levels of A1c diverges from typical clinical practices and may require repeat testing. The study did not analyze pancreatic fat, another predictor of diabetes.

DISCLOSURES:

The study was supported by SKKU, the National Research Foundation of Korea, and the UK National Institute for Health Research Southampton Biomedical Research Centre. Several authors disclosed financial relationships with Medical IP and Echosens.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.