Abstract
Background Recently, a newly proposed data-driven approach for classifying
diabetes has challenged the status quo of the classification of adult-onset
patients with diabetes. This study investigated the association between liver
injury and diabetes, classified by data-driven cluster analysis, as liver injury
is a significant risk factor for diabetes.
Methods We enrolled 822 adult patients with newly diagnosed diabetes.
Two-step cluster analysis was performed using six parameters, including age at
diagnosis, body mass index, hemoglobin A1C, homoeostatic assessment model 2
estimates about insulin resistance (HOAM2-IR) and beta-cell function (HOMA2-B),
and glutamic acid decarboxylase antibodies (GADA) positivity. Patients were
allocated into five clusters. Serum alanine aminotransferase (ALT) and aspartate
aminotransferase (AST) activity were compared as indicators of liver injury
among clusters.
Results Serum ALT and AST activities were significantly different among
clusters (P=0.002), even among those without GADA positivity
(P=0.004). Patients with severe insulin-resistant diabetes
(SIRD) and mild obesity-related diabetes (MOD) had a more severe liver injury.
Gender dimorphism was also found for serum ALT and AST activities among
subgroups. Female patients had better liver function than males with SIRD and
MOD.
Conclusions We verified the feasibility of a newly proposed diabetes
classification system and found robust and significant relationship and gender
differences between serum ALT and AST activities and diabetes in some specific
subgroups. Our findings indicate that more attention should be paid to diabetes
subgroups when studying risk factors, indicators, or treatment in diabetic
research.
Key words
cluster analysis - diabetes mellitus - liver diseases - gender difference
