Our goal was to use 2-channel frequency domain near-infrared spectroscopy (NIRS) to investigate the hemodynamic and metabolic mechanisms underlying hyperglycemia-associated long-term memory impairment. We hypothesized that prefrontal cortex (PFC) oxygen saturation (%Sat) and perfusion (tHb, i.e. total hemoglobin) would decrease due to hyperglycemia during learning, and then increase during recall. During learning, participants’ blood glucose was manipulated with beverages containing either 47.4 mg saccharine control (CON, n = 10), or 50 g dextrose + 23.7 mg saccharine (GLC, n = 10). In the Symbol-Digit Modalities Test (SMDT) participants matched nine symbols to corresponding digits (1-9 inclusive), completing 105 learning and 15 testing trials on day 1 and 15 testing trials on day 2. From learning to recall, CON SMDT performance was unchanged, but GLC SMDT performance was decreased 11% (P = 0.0173). There were significant interactions (2-way ANOVA) between the CON-GLC treatment effects and the learning-recall effects for both PFC perfusion and oxygen saturation. Specifically, comparing learning to recall, CON exhibited no tHb differences but for GLC there was a large tHb decrease during learning with a partial recovery toward CON values during recall (P = 0.0012); and, comparing learning to recall, CON exhibited a large %Sat decrease but GLC exhibited a large %Sat increase (P = 0.021). We speculate that, during learning, after overnight fasting (CON) the PFC demands more hemodynamic and metabolic resources and “works” harder, but with readily available sugar (GLC) the PFC exhibits decreased “effort.”
|