Pharmacological conditioning of insulin effects in patients with diabetes type-2 and healthy controls: randomized controlled trial

Abstract

Aims/hypothesis. Pharmacological conditioning, where an association is formed between a drug and a conditioned stimulus (CS, e.g., smell), can influence physiological processes. Evidence exists that it is possible to condition insulin effects in animals and healthy humans but it is unclear whether these effects can be observed in patients as well. The aim of this study was to investigate the effects of pharmacological conditioning with intranasal insulin on blood glucose, insulin, c-peptide, hunger and memory in patients with diabetes type-2 and age- and sex-matched healthy controls. Exploratively, we examined sex differences in these effects. We hypothesized that conditioning with insulin would trigger endogenous insulin release, decrease blood glucose and hunger, and improve memory. Methods. In a randomized double-blind controlled trial, thirty-two patients with diabetes type-2 (17 males, mean age=68.8, SD=11.86) and thirty-two healthy controls (17 males, mean age=67.8, SD=6.12) were randomly assigned to a conditioned or a control group. On day 1, participants in the conditioned group received 6 administrations of 20 units of intranasal insulin together with a CS (smell of rosewood oil) while the control group received a placebo with the CS. On day 2, participants in both groups received a placebo spray with the CS. Glucose, insulin and c-peptide were repeatedly measured in blood on both days. Hunger was measured with a self-report question, a mobile approach-avoidance task and a taste test. Memory was measured with an auditory verbal learning test.Results. Intranasal insulin prevented the drop of glucose levels in patients but not in healthy controls (B=0.03, SE=0.02, p=0.027), and decreased C-peptide levels in healthy controls but not patients (B=0.01, SE=0.001, p=0.008). Conditioned insulin effects also prevented the drop of blood glucose levels but only in male participants (for both healthy and patient groups) (B=0.001, SE=0.0003, p=0.024). Moreover, conditioning significantly decreased hunger in healthy participants (B=0.31, SE=0.09, p< 0.001). No effects of intranasal insulin or conditioning were found on insulin, approach-avoidance tendencies, calories eaten and memory. Conclusions/interpretation. Our study provides further evidence that conditioning with intranasal insulin might modify blood glucose levels and decrease hunger and that its effects differ depending on health status and sex. The unexpected direction of intranasal insulin effects on blood glucose might potentially be explained by age as our study population was older than participants from previous research. Conditioning with intranasal insulin might be beneficial for groups suffering from intensive hunger, but seems not be particularly suitable for blood glucose reduction. Diabetes type-2 patients might benefit from conditioning with glucose lowering drugs.