Nefeli Dimitropoulou1, Magdalena Beran1, Simone Eussen1, Cassy Dingena 1, Pieter Dagnelie1, Carla van der Kallen1, Annemarie Koster, Marleen van Greevenbroek1, Martijn Brouwers1, Miranda Schram1, Bastiaan de Galan1
1CARIM Institute for Cardiovascular Diseases, Maastricht University, the Netherlands
nefeli.dimitropoulou@maastrichtuniversity.nl
Background: Obesity and poor dietary habits, which are thought to explain high liver fat content in type 2 diabetes (T2D), increasingly affect individuals with type 1 diabetes (T1D). Whether diabetes per se affects liver fat accumulation is unknown. We therefore aimed to estimate the difference(s) in liver fat content in people with T1D, as compared to people with T2D and those with normal glucose metabolism (NGM) and investigate potential confounders that could explain differences in liver fat content between these groups.
Methods: Data were derived from The Maastricht Study, a population-based cohort study. We matched all participants with T1D to those with T2D and NGM in a 1:2:2 ratio, based on BMI, sex, age, and education. Magnetic Resonance Imaging (MRI) was used to estimate the liver fat content. Multiple linear regression was used to adjust for potential confounders, including cardiovascular risk, lifestyle, dietary factors, insulin resistance, and insulin treatment.
Results: We included individuals with T1D (n=29), T2D (n=58), and NGM (n=58). The median liver fat content in T1D was lower compared to T2D (2.1%, [IQR: 1.1,3.3] versus 4.9%, [2.2,9.6], p=0.001), but did not differ from that in individuals with NGM (2.6% [1.8,3.8], p=0.692). In the crude model, the difference of liver fat content (%) between T1D and T2D was -0.339 (95%CI: -0.509, -0.169), p<0.001). After full adjustments the difference in liver fat content attenuated between the two groups, in particular when adjusting for estimated glucose disposal rate (-0.229 (95%CI: -0.465, 0.07), p=0.057) and insulin treatment (-0.001 (95%CI: -0.356, 0.358), p= 0.995).
Discussion/Conclusion: Liver fat content in people with T1D is in line with that in individuals with NGM and considerably lower than in people with T2D. This difference between T1D and T2D is partially explained by insulin sensitivity and insulin treatment rather than obesity or dietary factors.