Central obesity but not generalised obesity (body mass index) predicts high prevalence of fatty liver (NRFLD), in recently detected untreated, IGT and type 2 diabetes Indian subjects.

Recently detected 152 IGT and 158 type 2 diabetes patients aged between 30 and 69 years, never treated with any antidiabetic drug, lipid lowering agent, angiotension converting enzyme (ACE) inhibitor and/or angiotensin receptor blocker (ARB) were evaluated. One hundred and sixty asymptomatic, willing, healthy, normoglycaemic spouse or unrelated attendants accompanying the patients were also selected. History of significant alcohol intake or hepatotoxic drugs and positive serologic findings for hepatitis B and C viruses were excluded. The prevalence of fatty liver was 64% in type 2 DM, 52% in impaired glucose tolerance (IGT) and 20% in normal glucose tolerance (NGT) subjects. Subjects with fatty liver had significantly higher body mass index (BMI), waist, waist hip ratio, waist to height ratio, triglyceride, fasting insulin, insulin resistance (HOMA-IR), along with significantly lower HDL-C and quantitative insulin-sensitivity check index (QUICKI) compared to those without fatty liver. Similar findings were noted in subjects with fatty liver in the subgroups of NGT, IGT and diabetes as well, for the above mentioned parameters. There was no significant difference in total cholesterol, LDL-C between subjects with or without fatty liver. However, analysis by the multivariate regression technique revealed that QUICKI (but not HOMA-IR), waist circumference and waist to height ratio had a significant association (p < 0.01 for all groups and for all these three parameters except in the NGT group for waist to height ratio p < 0.05) with development of fatty liver. BMI on the other hand was not significantly associated with the fatty liver in the multivariate regression (p = 0.067).