Oral Presentation Australian & New Zealand Obesity Society 2016 Annual Scientific Meeting

Circulating bile acids are associated with insulin resistance and visceral and liver fat in human subjects (#29)

Ramy H Bishay 1 2 3 , Katherine T Tonks 1 2 4 , Jacob George 5 6 , Dorit Samocha-Bonet 4 , Donald J Chisholm 1 2 4 , David E James 7 , Jerry R Greenfield 1 2 4
  1. St Vincent's Hospital, Sydney, Darlinghurst, NSW, Australia
  2. St Vincent's Clinical School, University of New South Wales Medical School, Sydney, NSW, Australia
  3. St Vincent's and Mater Hospital, University of Notre Dame, Sydney, NSW, Australia
  4. Diabetes and Metabolism Research Program, Garvan Institute of Medical Research, Sydney, NSW, Australia
  5. Storr Liver Centre, Westmead Millennium Institute, Westmead Hospital, University of Sydney, Sydney, NSW, Australia
  6. Department of Gastroenterology and Hepatology, Westmead Hospital, Sydney, NSW, Australia
  7. School of Life and Environmental Sciences, Charles Perkins Centre, Sydney Medical School, University of Sydney, Sydney, NSW, Australia

INTRODUCTION Bile acids (BA) are purported to have a potential role in insulin resistance and obesity, although the exact mechanism remains elusive. We hypothesised that BA concentration is increased in obesity and/or insulin resistance. METHODS Seventy-one adult volunteers formed four groups based on BMI, homeostatic model assessment of insulin resistance (HOMA-IR) and a 75-g OGTT: lean insulin-sensitive (BMI≤25 kg/m2, HOMA-IR<2.0, n=19), overweight/obese insulin-sensitive non-diabetic (Obsen, BMI>25 kg/m2, HOMA-IR<1.5, n=11), overweight/obese insulin-resistant (Obres, BMI>25 kg/m2, HOMA-IR>3.0, n=20) and type 2 diabetes mellitus (T2DM, n=21). We measured insulin sensitivity by hyperinsulinaemic-euglycaemic clamp, body composition/central abdominal fat by dual energy X-ray absorptiometry, visceral fat area by computed tomography and fasting insulin, adiponectin and BA. RESULTS Neither total BA (r=0.012, p=0.91) nor cholic acid (CA, the predominant primary BA) (r=0.04, p=0.70) correlated with percent total body fat. However, there were significant associations between BA and CA levels and waist circumference (r=0.39, p=0.0008; r=0.34, p=0.0035), central abdominal (r=0.33, p=0.0057; r=0.28, p=0.01) and visceral fat (r=0.26, p=0.026; r=0.24, p=0.045), respectively. CA, but not total BA,  correlated with liver density (an inverse marker of hepatic fat, r=-0.25, p=0.03). Total BA inversely correlated with insulin sensitivity (glucose infusion rates corrected for fat-free mass [GIR/FFM], r=-0.35, p=0.003) and adiponectin levels (r=-0.24, p=0.04). In group comparisons, GIR/FFM was significantly lower, and visceral and liver fat significantly higher, in Obres compared to lean and Obsen subjects, despite similar total adiposity in Obres  and Obsen (data not shown). Consistent with correlation analyses, total BA concentration tended to be higher in Obres (1.35+1.1 mmol/L) versus Obsen (0.67+0.28 mmol/L) (p=0.057), but were similar between Obsen and lean (1.00+0.69 mmol/L). CONCLUSION Our data suggest that BA concentrations aligned closely with insulin resistance, central abdominal, visceral and liver fat in human subjects. Whether BA play an aetiolgoical role in insulin resistance is yet to be elucidated.