Sex-Divergent Effects of Magnesium L-Threonate Supplementation on Sleep Quality, Cognitive Performance, and Neuromuscular Function in Healthy Adults

This study was a randomized, double-blind, placebo-controlled, parallel-group trial examining the effects of six weeks of magnesium L-threonate (MgT) supplementation on cognitive performance, mental fatigue, neuromuscular function, sleep quality, and psychophysiological outcomes in healthy adults. Biological sex and weight status were examined as pre-specified moderating variables given evidence for sex-specific differences in magnesium homeostasis and the elevated prevalence of magnesium inadequacy in individuals with overweight or obesity.

Participants:

Healthy male and female adults aged 18-45 years were recruited from a local university and surrounding community. Eighty-one participants were enrolled and randomized; the final analytic sample comprised 54 participants (MgT: n = 28; PLA: n = 26) following exclusions for non-adherence, technical data malfunction, and multivariate outlier removal. The sample was 59% female with a mean age of 23 ± 6 years and mean BMI of 26.7 ± 6.3 kg/m².

Intervention:

Participants assigned to the intervention group consumed 2 g·day-¹ of magnesium L-threonate (Magtein®; AIDP Inc., City of Industry, CA, USA) divided into two daily doses (morning and evening) for six weeks. Participants assigned to the placebo group consumed visually identical capsules containing rice powder (2 g·day-¹) on the same dosing schedule. Supplement compliance was assessed via capsule count at study conclusion; participants demonstrating less than 85% adherence were excluded from the final analytic sample.

Testing Procedures:

Participants completed identical testing sessions at baseline (Week 0) and following six weeks of supplementation (Week 6). All sessions were conducted at the same time of day for each participant to minimize circadian variability. Prior to each session, participants were instructed to abstain from caffeine for at least 12 hours, alcohol for 24 hours, and strenuous exercise for 24 hours.

Testing was conducted in a standardized order as follows:

Psychophysiological Questionnaires: Participants completed the Pittsburgh Sleep Quality Index (PSQI; 0-21 scale), Perceived Stress Scale (PSS), State-Trait Anxiety Inventory (STAI; state and trait subscales), and Visual Analog Scale for Fatigue (VAS-F; 0-100 mm) prior to any physical or cognitive testing.

Neuromuscular Assessment - Pre-fatigue: Countermovement jump (CMJ) performance was assessed using dual force plates (VALD ForceDecks, VALD Performance, Brisbane, Australia). Variables of interest included jump height (cm), peak landing force (N), reactive strength index (m·s-¹), and concentric impulse (N·s).

Cognitive Battery - Pre-fatigue: Cognitive performance was assessed using a computerized battery administered via the SOMA platform (SOMA Technologies, Lucerne, Switzerland). The battery included four tasks: the Psychomotor Fatigue Test (PFTT), Incongruent Flanker Task, Incongruent Stroop Task, and Task-Switching Test. Primary cognitive outcomes included reaction time (RT; ms), processing speed (1000/RT; s-¹), coefficient of variation (%), and rate correct score (correct responses·s-¹) derived from each task.

Mental Fatigue Induction: Mental fatigue was induced via a standardized 20-minute time-load dual-back (TLDB) task. The VAS-F was administered immediately before and after the TLDB task to quantify changes in perceived fatigue. The NASA Task Load Index (NASA-TLX) was completed immediately following the TLDB task to assess subjective cognitive workload across six subscales: mental demand, physical demand, temporal demand, performance, effort, and frustration.

Cognitive Battery - Post-fatigue: The full cognitive battery was repeated immediately following the TLDB protocol to assess performance under conditions of experimentally induced cognitive fatigue.

Neuromuscular Assessment - Post-fatigue: CMJ testing was repeated following the post-fatigue cognitive battery to assess the effect of cognitive load on subsequent neuromuscular output.

Statistical Analysis All analyses were performed in R with statistical significance set at α = 0.05. Primary outcomes were analyzed across three structurally distinct approaches: (1) absolute outcomes and within-bout change scores (pre- to post-mental fatigue) at W0 and W6 were analyzed using linear mixed models with fixed effects of condition, time, the condition-by-time interaction, and a random intercept for participant; (2) six-week change scores (W6 - W0) were analyzed using ANOVA with fixed effects of condition, sex, weight status, and the corresponding two- and three-way interactions; and (3) the change in the within-bout mental fatigue response across the six-week period was analyzed using analogous ANOVA to isolate whether MgT altered the magnitude of the acute mental fatigue response over time. Significant interaction effects were further assessed using post hoc comparisons with Bonferroni-Holm adjustment. Unadjusted simple effects tests were conducted when omnibus interactions were significant but adjusted post hoc comparisons were not, and are reported as exploratory. All primary models were re-estimated following removal of values exceeding ±3 SD to evaluate robustness of findings.

Inclusion criteria

• Healthy male or female adults aged 18-45 years
• Free from diagnosed neurological, cardiovascular, metabolic, or psychiatric conditions
• Not currently taking medications or supplements known to influence cognitive or neuromuscular function
• Willing to maintain habitual diet, physical activity, and sleep routines throughout the study period

Exclusion criteria

• Pregnant or breastfeeding
• Diagnosed sleep disorder
• Initiated any new supplement or medication regimen within the past three months
• Failed to meet ≥85% supplementation adherence threshold during the study (post-enrollment exclusion)
• Technical data malfunction resulting in unrecoverable testing data (post-enrollment exclusion)
• Identified as a multivariate outlier (values exceeding ±3 SD from the group mean on primary outcomes; post-enrollment exclusion)