The biggest part of our self-image as human beings is determined by the unlimited function of our brain. This becomes obvious especially when we get older and physical shortcomings amplify. Unfortunately, some medical conditions not only augment physical shortcomings that are apparent on first sight, but also attack the function of the brain, an organ that has nothing or at least little to do with the primary disease.
Diabetes mellitus—a disease affecting millions of people worldwide—is the prototype of such a multiorgan dysfunction syndrome compromising neurocognitive function via poorly understood pathways (1). Not only long-term diabetes but also short-term dysregulations of blood glucose are linked to profound permanent neurocognitive impairment (e.g., when they are experienced during critical illness) (2). Is this an effect of glucose per se—too much or too little of it—or of insulin or both?
No doubt, brain cells largely rely on glucose as fuel and glycopenia is thus directly deleterious. Also, hyperglycemia is a hazard to our central nervous system as it destroys neurons (3). This is mediated both directly via glucose toxicity and indirectly via more long-term consequences such as vasculopathy. So, it is easy to believe that dysglycemia might be causal for neurocognitive impairment, leaving very little space to imagine treatment strategies effective enough to tackle the disease when the dysfunctions have become apparent. Besides accepted effects of glucose, the other big player in glucose regulation—insulin—was long neglected as the nervous system was considered to be insulin independent. This notion surely does not hold true (4). Although we poorly understand the pathophysiology, diabetes, brain insulin resistance, and low levels of insulin in the cerebrospinal fluid (CSF) are closely linked to mental disorders such as Alzheimer disease (5).
Without relevant systemic effects in doses around 40 IU, insulin can be given intranasally, thus avoiding the blood-brain barrier. Insulin directly migrates into the CSF and into the brain where it acts as a neuromodulator to influence, among others, a couple of metabolic pathways, systemic blood glucose, and cognitive function by affecting not only brain cell metabolism and energy homeostasis but also regional brain perfusion, synapse formation, and neuronal survival (6). Indeed, increasing insulin levels by intranasal insulin overcomes brain insulin resistance and thus improves neurocognitive performance (e.g., in Alzheimer patients) (7) (Fig. 1).
In this issue of Diabetes, Zhang et al. (8) present their work on unraveling the underlying mechanisms behind intranasal insulin and neurocognitive function. In their cohort of patients with diabetes, they showed that a single dose of 40 IU of intranasal insulin not only improved neurocognitive performance in test batteries but also restored the connectivity of hippocampal regions and the mediofrontal cortex back to a level comparable to healthy subjects. Certainly, cognition is a rather complex process depending on the connection of various neural networks and regions in the brain than a located spot in the human central nervous system. However, we still do not know the exact pathways of insulin influence on the activation and connection of the networks among each other that appear to be crucial for neurocognitive performance. Perfusion might be a cornerstone but not the exclusive key to understand the phenomenon.
The study by Zhang et al. opens further perspectives. Rather than being a single-target intervention, increasing CSF insulin levels by intranasal insulin appears to restore complex neural networking in the direction of normalization. It is tempting to speculate about future applications of this simple and inexpensive intervention. Besides Alzheimer disease, other conditions characterized by CSF insulin deficiency could be targets. Can we counteract postoperative neurocognitive impairment when stress-induced dysglycemia might contribute to the condition? What is it about critical illness−induced neurocognitive dysfunction, postoperative delirium, or drug-induced dysglycemia?
Certainly, although Zhang et al. give some evidence, we are left with reasonable limitations. We still do not know whether the effects will be reproducible in a larger cohort and whether patients’ long-term outcomes will be improved. It is probable that the effect might fade away or trigger deleterious side effects in the short or long run. Although we get a hint of the mechanisms with resting-state MRI, we wonder what the picture will be in functional imaging. Moreover, we still have to discover the underlying pathways.
The proof-of-concept study of Zhang et al. (8) might open new perspectives for patients with an impairment of the neurocognitive function but is far from forcing us to treat every patient with intranasal insulin. It calls for prospective trials in larger populations to evaluate the effect on daily life and investigations to broaden the indication and to further unravel the underlying mechanisms. Maybe intranasal insulin might help us to boost our mental activities to implement bottom-up research.
See accompanying article, p. 1025.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.