Serum adiponectin levels predict the effect of short-term dietary interventions on insulin sensitivity in humans

Fat-rich diets can acutely induce insulin resistance. Data from adiponectin knock-out mice suggest that this effect might be increased in the absence of adiponectin. In the present study we tested whether plasma adiponectin concentrations influence changes in insulin sensitivity induced by a short-term dietary intervention in humans.

We analysed data from 27 healthy, non-obese men with normal glucose tolerance. These men ate a diet high in fat and a diet high in carbohydrates for three days each.

The high-fat diet induced a significant drop in insulin sensitivity (determined by euglycaemic–hyperinsulinaemic clamp) compared to baseline (0.100±0.009 vs 0.083±0.007 µmol·kg⁻¹·min⁻¹·(pmol·l⁻¹), p=0.01). The drop in insulin sensitivity was more pronounced in subjects with low serum adiponectin (0.094±0.011 vs 0.077±0.010 µmol·kg⁻¹·min⁻¹·(pmol·l⁻¹), p=0.02) than in subjects with high serum adiponectin (0.103±0.011 vs 0.090±0.040 µmol·kg⁻¹·min⁻¹·(pmol·l⁻¹), p=0.16). In the whole group the high-carbohydrate, low-fat diet did not cause an increase in insulin sensitivity (0.095±0.007 vs 0.102±0.009 µmol·kg⁻¹·min⁻¹·(pmol·l⁻¹), p=0.06). However, insulin sensitivity was significantly increased in the subgroup with low serum adiponectin levels (0.084±0.013 vs 0.099±0.018 µmol·kg⁻¹·min⁻¹·(pmol·l⁻¹), p=0.01). In an additional multivariate analysis post-intervention insulin sensitivity was predicted by pre-intervention insulin sensitivity (p<0.001) and adiponectin concentrations (p=0.001).

These data indicate that the reduction in insulin sensitivity achieved by a short-term high-fat diet is more pronounced in non-obese subjects with low serum adiponectin. Thus it is possible that the restriction of dietary fat and a diet high in carbohydrates might be particularly effective in subjects with low adiponectin such as obese or Type 2 diabetic individuals.