Animal studies suggest that insulin action in the brain is involved in the regulation of peripheral insulin sensitivity. Whether this holds true in humans is unknown. Using intranasal application of insulin to the human brain, we studied the impacts of brain insulin action on wholebody insulin sensitivity and the mechanisms involved in this process. Insulin sensitivity was assessed by hyperinsulinemic-euglycemic glucose clamp before and after intranasal application of insulin and placebo in randomized order in lean and obese men. After insulin spray application in lean subjects, a higher glucose infusion rate was necessary to maintain euglycemia compared with placebo. Accordingly, clamp-derived insulin sensitivity index improved after insulin spray. In obese subjects, this insulin-sensitizing effect could not be detected. Change in the high-frequency band of heart rate variability, an estimate of parasympathetic output, correlated positively with change in whole-body insulin sensitivity after intranasal insulin. Improvement in whole-body insulin sensitivity correlated with the change in hypothalamic activity as assessed by functional magnetic resonance imaging. Intranasal insulin improves peripheral insulin sensitivity in lean but not in obese men. Furthermore, brain-derived peripheral insulin sensitization is associated with hypothalamic activity and parasympathetic outputs. Thus, the findings provide novel insights into the regulation of insulin sensitivity and the pathogenesis of insulin resistance in humans.

Aims/hypothesis Impaired insulin sensitivity is a major factor leading to type 2 diabetes. Animal studies suggest that the brain is involved in the regulation of insulin sensitivity. We investigated whether insulin action in the human brain regulates peripheral insulin sensitivity and examined which brain areas are involved. Methods Insulin and placebo were given intranasally. Plasma glucose, insulin and C-peptide were measured in 103 participants at 0, 30 and 60 min. A subgroup (n012) was also studied with functional MRI, and blood sampling at 0, 30 and 120 min. For each time-point, the HOMA of insulin resistance (HOMA-IR) was calculated as an inverse estimate of peripheral insulin sensitivity.