Ten healthy adults (five male, five female; mean age 33 ± 11 years, range 20 - 53 years) were enrolled from a wellness clinic population. Participation was voluntary, with individuals expressing interest in understanding the physiological effects of fasting. The study was conducted as an exploratory pilot investigation rather than a formal clinical trial. The inclusion criteria required absence of metabolic, neurological, or cardiovascular disease and no use of medications or supplements known to influence glucose regulation, ketogenesis, or cognitive function. All participants provided written informed consent.

All participants were free from metabolic disorders, neurological disease, cardiovascular illness, and any condition known to interfere with glucose regulation, ketone production, or cognitive performance. None were taking medications or supplements that could alter the metabolic or neurocognitive variables under study. Recruitment was achieved through local outreach and word-of-mouth referrals. Prior to enrollment, all participants provided informed consent in accordance with the ethical principles set forth in the Declaration of Helsinki.

The study employed a repeated-measures design, with each participant serving as their own control across three time points: baseline, 24 h, and 48 h of continuous fasting [14] . Baseline testing occurred following an overnight fast, with participants instructed to cease caloric intake one hour before arrival to ensure a minimally post-absorptive state. From baseline until the 48 h endpoint, participants abstained from all calories but were permitted water, black coffee, or unsweetened tea; artificial sweeteners and additives were prohibited due to potential effects on insulin and ketone production [15] . Follow-up sessions were scheduled at the same time of day as baseline to minimize circadian variability, and all assessments were conducted in a controlled environment. Fasting compliance was verified via time-stamped glucose and β-hydroxybutyrate (BHB) readings at each visit, with expected biomarker trajectories (falling glucose, rising BHB) confirming adherence.

Capillary blood glucose and β-hydroxybutyrate (BHB) concentrations were obtained at each time point using a handheld meter [16] . BHB, the primary circulating ketone body, is widely recognized as the most reliable peripheral biomarker of nutritional ketosis [17] . While the term ketones is commonly used in public discourse, physiologically this refers collectively to BHB, acetoacetate, and acetone. For the purposes of this study, the BHB measurement served as the definitive index of ketone availability to the brain. Blood samples were collected via single-use lancets, with values recorded to the nearest 0.1 mmol/L for BHB and 1 mg/dL for glucose [18] . The same testing device and operator were used for all measurements to maintain methodological consistency.

Participants were permitted to maintain habitual caffeine intake (black coffee or unsweetened tea). Daily caffeine use was recorded but not standardized across participants. This factor is acknowledged as a limitation of the study design.

Cognitive performance was evaluated using the Trail Making Test (TMT), a validated neuropsychological instrument sensitive to subtle cognitive changes in healthy individuals [19] . Part A (TMTA) required participants to connect a series of numbers in ascending order as quickly as possible, providing a measure of visual scanning ability, psychomotor processing speed, and sequencing efficiency [20] . Part B (TMTB) incorporated an additional executive demand by requiring alternation between numbers and letters in ascending sequence (e.g., 1-A-2-B), thereby taxing cognitive flexibility, divided attention, and set-shifting capacity [21] . In both tasks, performance was quantified as the time in seconds required to complete the sequence, with shorter times representing superior performance.

Neural processing speed and cortical activation were assessed via event-related potentials (ERPs) recorded during an auditory oddball paradigm using the WAVi Research Brain Measurement System [22] . This protocol reliably elicits the P300 component, a positive deflection in the ERP waveform occurring approximately 300 milliseconds after the presentation of a target stimulus [23] . The P300 latency reflects the speed of cognitive evaluation, while the amplitude, measured in microvolts, is interpreted as an index of cortical activation and attentional resource allocation [24] .

Recordings were obtained with participants seated comfortably, eyes open, and fixated on a visual point to minimize ocular artifacts [25] [26] . Data were band-pass filtered between 0.1 and 30 Hz, and epochs containing movement or electromyographic noise were excluded prior to waveform averaging. Latency values were measured in milliseconds from the onset of the auditory stimulus to the peak of the P300 waveform, and amplitude values were measured from baseline to the peak of the positive deflection.

An auditory oddball paradigm was used to elicit the P300 component. Data were band-pass filtered (0.1 - 30 Hz) and baseline corrected (−200 to 800 ms relative to stimulus onset). Epochs containing blinks, eye movements, or muscular artifacts were excluded, and at least 30 artifact-free trials were averaged for each participant.

Statistical analyses were performed using GraphPad Prism (version 10). Because this was an exploratory pilot study with a small sample, analyses focused on within-subject changes across baseline, 24 h, and 48 h. Repeated-measures ANOVA was used for glucose, β-hydroxybutyrate (BHB), Trail Making Test (TMTA and TMTB), and P300 parameters. All values are reported as mean ± SD, with significance set at p < 0.05 but interpreted cautiously given the pilot nature of the study.

The metabolic data demonstrated a clear and progressive shift from glucose-dominant metabolism toward ketone utilization over the course of the 48-hour fasting period. Mean fasting glucose at baseline was 104.5 ± 10.2 mg/dL, consistent with euglycemia following an overnight fast. By 24 hours, glucose had declined to 87.7 ± 8.6 mg/dL, representing an approximate 16% reduction, and by 48 hours, levels reached 69.2 ± 7.9 mg/dL, marking a total decrease of nearly 34% from baseline. This inverse relationship between glucose and ketones is clearly depicted in Figure 1 , illustrating the coordinated metabolic transition toward nutritional ketosis.

Concomitantly, β-hydroxybutyrate (BHB) levels rose sharply, reflecting the anticipated metabolic switch to fatty acid oxidation and ketone production [27] . The figure displays baseline BHB values averaged 0.27 ± 0.15 mmol/L, indicative of minimal ketosis. At 24 hours, concentrations increased to 1.09 ± 0.38 mmol/L, and by 48 hours, levels reached 2.73 ± 0.81 mmol/L, signifying entry into a robust state of nutritional ketosis [28] . This observation aligns with prior literature on sex-based differences in lipid mobilization and ketone production during prolonged fasting, though the present sample size precludes definitive conclusions [29] .

The Trail Making Test performance revealed a differential pattern across the two tasks, suggesting selective effects of fasting on cognitive domains. TMTA completion time, which primarily reflects processing speed and visual scanning, increased modestly from a baseline mean of 42.1 ± 6.8 seconds to 48.3 ± 7.2 seconds at 24 hours, and further to 50.2 ± 7.5 seconds at 48 hours. As shown in the left panel of Figure 2 , this mild slowing, while not reaching statistical significance in this small cohort, may reflect transient adaptation to reduced glucose availability in tasks heavily reliant on psychomotor speed.

In contrast, TMTB performance improved markedly over the same period. Baseline completion time averaged 95.4 ± 12.7 seconds. As the right panel of the figure shows, at 24 hours, participants completed the task in 82.5 ± 11.9 seconds, and by 48 hours, mean completion time had dropped to 73.2 ± 10.8 seconds, representing a 22% improvement relative to baseline [30] . Given that TMTB performance is strongly dependent on cognitive flexibility, working memory, and set-shifting capacity, this enhancement suggests that early nutritional ketosis may selectively support executive function [31] .

Electrophysiological markers remained stable across the fasting interval [32] . As shown in the left panel of Figure 3 , the P300 latency measured 272 ms at baseline, 266 ms at twenty-four hours, and 267 ms at forty-eight hours. Voltage similarly exhibited negligible net change (14.5 μV at baseline, 13.9 μV at twenty-four hours, 14.5 μV at forty-eight hours), which is detailed in the right panel of Figure 3 . The absence of a systematic latency prolongation argues against any fasting-related slowing of cortical stimulus evaluation, and the preserved voltage suggests stable attentional resource allocation and network engagement [33] . Together with improved TMTB performance, these findings indicate that neural efficiency was maintained, if not functionally optimized for executive control, during early ketosis.

Given the marked rise in ketones and the suggestion of sex differences at forty-eight hours, BHB trajectories rose separately for males and females [34] . As Figure 4 notes, males displayed a steeper slope from twenty-four to forty-eight hours, culminating in higher mean ketones at the study endpoint. Despite this metabolic divergence, behavioral and electrophysiological measures did not differ qualitatively by sex in this sample. This pattern implies that a wider physiological range of ketone exposure, at least within the bounds observed here does not necessarily translate to measurable differences in executive performance or P300 indices over forty-eight hours in healthy adults.