Proton magnetic resonance spectroscopy (¹H-MRS) is a non-invasive neuroimaging technique that quantifies brain metabolites in vivo, offering unique insights into neurochemical processes underlying mental health disorders [1]. Unlike structural MRI, which depicts anatomical features, ¹H-MRS captures biochemical signatures linked to neuronal integrity, glial activity, and neurotransmission. Key metabolites—N-acetylaspartate (NAA), choline-containing compounds (Cho), creatine (Cr), myo-inositol (Myo-In), and glutamate/glutamine (Glx)—reflect neuronal viability, membrane turnover, osmotic regulation, and excitatory signaling [2, 3, 4, 5, 6]. These markers have been implicated in the pathophysiology of mood and anxiety disorders, making them clinically relevant targets for early detection and intervention.

Recent MRS meta-analyses have reported alterations in key metabolites—including reduced NAA and variable changes in Cho and Myo-In—in individuals with depression and anxiety disorders, suggesting disruptions in neuronal integrity, membrane turnover, and glial activity [7, 8, 9, 10]. These findings highlight the relevance of basal ganglia neurochemistry to affective processes. Emerging evidence further suggests that similar patterns of neurochemical variability may also be detectable in healthy individuals, potentially contributing to normal emotional regulation and stress reactivity [11, 12]. Identifying such patterns in nonclinical populations could enable early risk stratification and preventive strategies—an approach aligned with precision psychiatry [13, 14].

The basal ganglia, traditionally associated with motor control, are increasingly recognized for their role in emotional regulation and reward processing [15, 16]. Functional imaging and connectivity studies highlight their involvement in affective circuits and stress-related disorders [17]. Furthermore, ¹H-MRS investigations have demonstrated metabolite alterations in the basal ganglia among individuals with major depressive disorder and bipolar disorder, reinforcing its potential as a biomarker region [11].

Young adulthood represents a critical developmental stage characterized by heightened stress reactivity and mood fluctuations, and university students frequently report elevated psychological distress [18, 19]. Investigating basal ganglia neurochemistry in this population can reveal early neurochemical signatures of vulnerability, supporting targeted interventions before clinical onset [11]. This study examines whether basal ganglia metabolite ratios (NAA/Cr, Cho/Cr, Myo-In/Cr, Glx/Cr) predict mood and anxiety scores in healthy young women using 3T ¹H-MRS. By linking neurochemical variability to affective states, we aim to advance non-invasive biomarkers for mental health monitoring and bridge the gap between basic neuroscience and clinical application.

Mood and anxiety scores demonstrated substantial variability across participants. Mood scores ranged from 0 to 66 (mean = 28.4, SD = 12.7), while anxiety scores ranged from 0 to 63 (mean = 26.1, SD = 11.9). Both distributions approximated normality, confirming that the sample captured a broad spectrum of affective variability (Table 1).

Representative ¹H-MRS spectra, as seen in Fig. 2, showed clear separation of major metabolite peaks (NAA, Cho, Cr, Myo In, Glx), and LCModel fitting provided reliable quantification. showed clear separation of major metabolite peaks (NAA, Cho, Cr, Myo-In, Glx), and LCModel fitting provided reliable quantification. All spectra met quality control thresholds (Cramér–Rao lower bounds $\le$20%, linewidth $\le$0.1 ppm). Metabolite ratios were normally distributed across participants (Table 2).

Fig. 2.

Representative ¹H-MRS spectrum from the basal ganglia at 3T. Peaks for N-acetylaspartate (NAA), choline (Cho), creatine (Cr), myo-inositol (Myo-In), and glutamate/glutamine (Glx) are shown, with LCModel fitting applied for metabolite quantification. In this scanner-integrated display, the LCModel fit is shown as a red line, the original spectral data appear as a white dotted plot, and the residuals are displayed as a blue trace at the bottom.

Significant positive correlations were observed for Cho/Cr and Myo-In/Cr with both mood and anxiety scores (all p 0.01). NAA/Cr showed small but statistically significant negative correlations with mood and anxiety (p 0.05). Glx/Cr was not significantly associated with either outcome (Table 3).

Regression analyses identified Cho/Cr and Myo‑In/Cr as significant positive predictors of mood scores, while NAA/Cr showed a small negative contribution. Glx/Cr was not a significant predictor. The overall model accounted for 31% of the variance in mood scores (Table 4).

For anxiety, Cho/Cr and Myo-In/Cr emerged as significant positive predictors, whereas NAA/Cr showed a small negative contribution. Glx/Cr was not a significant predictor. The overall model explained 28% of the variance in anxiety scores (Table 5).

Principal component analysis (PCA) identified two latent neurochemical components. The first component, characterized by high loadings for Cho/Cr and Myo-In/Cr, explained 41.3% of the variance and was positively associated with both mood and anxiety scores. The second component, dominated by NAA/Cr and Glx/Cr loadings, explained 28.6% of the variance and showed small inverse associations with emotional scores. Together, these components accounted for 69.9% of the total variance, suggesting distinct biochemical pathways underlying affective variability.

This study demonstrates that specific basal ganglia metabolite ratios measured via 3T ¹H-MRS are significantly associated with mood and anxiety scores in healthy participants. The positive associations between Cho/Cr and Myo-In/Cr ratios and both emotional measures suggest that increased membrane turnover and glial activity may contribute to affective variability, even in nonclinical populations. These findings indicate that neurochemical markers derived from MRS are linked to emotional states, supporting their relevance for understanding subclinical affective fluctuations and informing preventive mental health strategies. In healthy individuals, higher scores on these scales are best interpreted as reflecting normal variability in emotional reactivity rather than clinical pathology, although some participants may exhibit subthreshold levels of distress that fall below diagnostic thresholds.

While some clinical studies have reported elevated Cho/Cr or Myo-In/Cr ratios in mood and anxiety disorders, the broader literature is mixed and varies across brain regions, populations, and methodological approaches. Our findings therefore should be interpreted as complementary to, rather than fully aligned with, prior clinical work. By examining a homogeneous, nonclinical sample, this study highlights that neurochemical variability is detectable even in healthy individuals, although the extent to which these patterns parallel those observed in clinical populations remains uncertain. Similarly, although emotional regulation may be conceptualized along a continuum, the present results do not directly establish continuity with the neurophysiological subtypes described in studies such as Drysdale et al. [23], which focus on functional connectivity rather than metabolite profiles.

The small inverse relationship between NAA/Cr and mood or anxiety scores suggests that reduced neuronal integrity may contribute to emotional vulnerability, although the effect size was modest. The lack of strong associations with Glx/Cr may reflect the complexity of glutamatergic signaling, which involves multiple receptor subtypes and region-specific dynamics. Sarawagi et al. [24] emphasize that glutamate variability in healthy individuals may require task-based paradigms to reveal functional relevance, while Godlewska et al. [15] report significant glutamatergic alterations in depressed patients using high-field MRS, contrasting with our nonsignificant Glx findings.

The identification of two latent neurochemical components through PCA—one reflecting glial/membrane activity (Cho, Myo-In) and the other neuronal integrity (NAA, Glx)—provides a biologically grounded framework for interpreting affective variability. Elevated Cho and Myo-In may indicate heightened membrane turnover and glial activation, processes implicated in stress reactivity and neuroinflammation, as noted by Munhoz et al. [25] and Sălcudean et al. [26]. Conversely, lower NAA may reflect subtle inefficiencies in neuronal mitochondrial function, which could predispose individuals to affective fluctuations. These observations suggest that emotional health may be shaped by a balance between glial activity and neuronal integrity.

From a translational perspective, our findings support the potential of ¹H-MRS as a non-invasive biomarker tool for identifying individuals at risk for emotional dysregulation. Current psychiatric diagnostics rely heavily on self-reported and behavioral assessments; neurochemical profiling could provide objective biological markers to complement these measures, as emphasized by Miller et al. [27]. This is particularly relevant for precision psychiatry frameworks advocating integration of neuroimaging biomarkers into screening protocols, as highlighted by Williams and Whitfield Gabrieli [28]. However, given the methodological constraints of the present study, including reliance on acquisition-time MRS quality thresholds, these translational implications should be considered preliminary. It is possible that applying more stringent CRLB thresholds or excluding outlier spectra could alter the strength or direction of the observed associations, and future work should incorporate these refinements to validate the robustness of our findings.

The focus on a homogeneous cohort of young women is both a strength and a limitation. While it minimizes confounding variables such as age and sex, it raises questions about generalizability. Future studies should examine whether similar neurochemical–emotional associations are present in males, older adults, and culturally diverse populations, as suggested by Trofimova and Gaykalova [29]. Expanding to longitudinal designs would clarify whether elevated Cho/Cr and Myo-In/Cr ratios predict future emotional difficulties or represent transient states. Multimodal imaging approaches, including functional magnetic resonance imaging (fMRI) and diffusion tensor imaging (DTI), could complement MRS findings and provide deeper insights into the neural circuits underlying mood and anxiety, as proposed by Etkin et al. [30]. Finally, these findings carry implications for university health programs. Psychological distress is disproportionately high among students, as shown by Granieri et al. [18], and our results suggest that neurochemical variability may contribute to this vulnerability. Integrating neuroimaging biomarkers into student wellness initiatives could help identify individuals at risk for mood and anxiety difficulties, guiding targeted support and resilience-building interventions. This aligns with global calls for precision mental health approaches that combine biological, psychological, and social indicators, as emphasized by Comai et al. [14]. Future work may also benefit from integrating susceptibility-based measures such as quantitative susceptibility mapping (QSM), which index iron content, myelination, and microstructural properties of the basal ganglia, and may provide complementary information to metabolite-based markers such as Cho and Myo-In.

This study demonstrates that neurochemical variability in the basal ganglia, as measured by 3T ¹H-MRS, is associated with mood and anxiety scores in healthy young women. Elevated Cho/Cr and Myo-In/Cr ratios may reflect neurochemical processes linked to affective variability, suggesting that glial and membrane-related metabolites contribute to emotional functioning even in nonclinical populations.

By employing a homogeneous cohort and standardized imaging protocols, we minimized several potential confounders and improved the interpretability of metabolite–emotion associations. While these findings provide preliminary insight into neurochemical correlates of affective states, they should be interpreted cautiously given the methodological constraints and the mixed results reported in the broader clinical literature.

The potential translational value of MRS-based neurochemical profiling for mental health monitoring is promising but remains exploratory. Future research should validate these observations in larger and more diverse samples, incorporate longitudinal designs to clarify temporal relationships, and apply more rigorous MRS quality-control procedures. Such work will be essential for determining whether metabolite patterns can reliably inform early identification of emotional vulnerability and contribute to emerging precision mental health approaches.

Data are available from the corresponding author upon reasonable request.

HH conceived and designed the study. MS and HA contributed to data acquisition and imaging protocol optimization. MGH, AIA, and NAA performed statistical analysis and contributed to data interpretation under the supervision and guidance of the principal investigator. SS Contributed to data collection and interpretation, and participated in manuscript preparation and final approval. HH supervised the project, integrated all contributions, and finalized the manuscript. All authors contributed to editorial changes in the manuscript. All authors reviewed and approved the final version of the manuscript and agree to be accountable for all aspects of the work.

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of Princess Nourah bint Abdulrahman University (IRB Log Number: 23 0257). Written informed consent was obtained from all participants prior to enrolment.

We would like to thank all the volunteers who generously participated in this study. Their contribution and willingness to take part were essential to the completion of this research.

This work was funded by the Deanship of Scientific Research at Princess Nourah bint Abdulrahman University, through the Research Groups Program Grant no. (RGP-1444-0062).

The authors declare no conflict of interest.