During adulthood, cognitive executive functions such as working memory and inhibitory control are increasingly challenged by sustained demands. Acute intake of dietary compounds like caffeine or certain amino acids may help support cognitive performance by modulating neuronal activity and cerebral blood flow. Functional near-infrared spectroscopy (fNIRS) is a promising technique to investigate these effects, providing real-time, non-invasive measurements of cortical hemodynamics linked to neuronal activation.

This thesis aims to develop methodological approaches using fNIRS to assess the cognitive effects of acute nutritional interventions in healthy adults. It begins with a review of current methodologies applied in this context to refine protocols and better control confounding factors in fNIRS measurements. These refinements informed a study that analysed the relationship between cortical hemodynamic responses and cognitive performance using multivariate modelling applied to data from 36 participants performing working memory (N-back) and inhibitory control (Stroop) tasks under varying cognitive loads. This combined approach of protocol refinement and modelling was applied to assess the acute effects of caffeine, both alone and in combination with L-arginine. This investigation was conducted in a double-blind, randomized, incomplete crossover study where 36 participants performed the N-back task under varying cognitive demands.

Results showed that the relationship between cortical activation and performance varied with task demand and was modulated by acute nutritional interventions. In the first study, during the inhibitory control task, low cognitive load was associated with increased activation in fronto-occipito-temporal regions, which correlated with better performance, reflecting enhanced attentional control for detecting subtle incongruencies. Conversely, under high cognitive load, decreased activation in these regions was linked to improved accuracy, suggesting more efficient conflict resolution when stimulus saliency was stronger. For the working memory task, higher cognitive loads elicited stronger activation and higher accuracy, while lower loads corresponded to reduced activation but comparable performance, indicating reduced reliance on attention and working memory when encoding fewer items. By manipulating cognitive demand, the second study identified specific fNIRS activation patterns related to cognitive performance during nutritional interventions. Notably, caffeine was associated with increased fronto-temporo-parietal activation and slower but accurate responses during the high-load working memory task, suggesting enhanced controlled processing to meet elevated cognitive demand. In contrast, the combination of caffeine and L-arginine resulted in stronger activation and faster responses during the low-load working memory task, reflecting heightened attentional engagement that facilitated processing speed in line with task demands.

In summary, this thesis advances the methodological and scientific foundations for using fNIRS in acute nutrition interventions. Future research should combine predictive modelling with improved fNIRS resolution to better capture the effects of dietary components on cortical and subcortical activity, paving the way for personalized, mechanism-based strategies to enhance cognitive performance.