Proton magnetic resonance spectroscopy (1H-MRS) enables noninvasive in vivo detection of biochemical and neurotransmitter alterations in brain neurons, offering potential for early diagnosis of Alzheimer's disease (AD). This study investigates metabolite distribution patterns in bilateral frontal white matter and imaging biomarkers of early cognitive function across preclinical and prodromal stages of the AD continuum. A cohort of 362 right-handed participants underwent comprehensive neuropsychological assessments. Single-voxel 1H-MRS was employed to acquire metabolite spectra from bilateral frontal white matter, complemented by venous blood analysis for AD-associated genes and toxic proteins. Statistical analysis revealed the following key findings. Normal Controls (NC): (1) Higher relative concentrations of N-acetyl-aspartate/N-acetyl aspartyl-glutamate (NAA+NAAG) and glycerol-phosphocholine/phosphocholine (GPC+PCho) were observed in left versus right frontal white matter. (2) Age-related decline: NAA+NAAG levels in left frontal white matter demonstrated progressive reduction from younger to older age groups. (3) The elderly group exhibited significantly lower glutamate/glutamine (Glu+Gln) concentrations in left frontal white matter compared to middle-aged and younger groups. Mild Cognitive Impairment (MCI): (1) Inverse correlation between Auditory Verbal Learning Test (AVLT) scores and right frontal NAA+NAAG content. (2) Positive associations between Shape Trail Test (STT-A/B) completion times and right frontal myo-inositol (mIns) levels. Left frontal Glu+Gln concentrations correlated positively with plasma biomarkers: amyloid β-protein (Aβ1-42), total tau (t-tau), and phosphorylated tau181 (p-tau181). These 1H-MRS-detected metabolite alterations in bilateral frontal white matter may reflect both physiological brain aging and AD-related pathological changes, suggesting their utility as potential diagnostic indicators for early-stage AD. The integration of metabolic profiling with established biomarkers could enhance predictive accuracy in the AD continuum.
Xie et al. (2026) studied this question.