The efficacy of low frequency repetitive transcial magnetic stimulation for treating auditory verbal hallucinations in schizophrenia: Insights from functional gradient analyses

Abstract

Background: Auditory Verbal Hallucinations (AVH) constitute a prominent feature of schizophrenia. Although low-frequency repetitive transcranial magnetic stimulation (rTMS) has demonstrated therapeutic benefits in ameliorating AVH, the underlying mechanisms of its efficacy necessitate further elucidation.

Objective: This study investigated the cortical gradient characteristics and their associations with clinical responses in schizophrenia patients with AVH, mediated through 1 Hz rTMS targeting the left temporoparietal junction.

Method: Functional gradient metrics were employed to examine the hierarchy patterns of cortical organization, capturing whole-brain functional connectivity profiles in patients and controls.

Results: The 1 Hz rTMS treatment effectively ameliorated the positive symptoms in patients, specifically targeting AVH. Initial evaluations revealed expanded global gradient distribution patterns and specific principal gradient variations in certain brain regions in patients at baseline compared to a control cohort. Following treatment, these divergent global and local patterns showed signs of normalizing. Furthermore, there was observed a closer alignment in between-network dispersion among various networks after treatment, including the somatomotor, attention, and limbic networks, indicating a potential harmonization of brain functionality.

Conclusion: Low-frequency rTMS induces alternations in principal functional gradient patterns, may serve as imaging markers to elucidate the mechanisms underpinning the therapeutic efficacy of rTMS on AVH in schizophrenia.

Keywords: Auditory verbal hallucinations; Brain connectivity patterns; Functional gradient; Repetitive transcranial magnetic stimulation; Schizophrenia.

Fig. 1

Clinical symptom changes between patients at baseline and posttreatment. Abbreviations: P, positive symptoms of PANSS; N, negative symptoms of PANSS; G, general symptoms of PANSS; AHRS, auditory hallucination rating scale; ns, non-significance. The error bar indicates the standard deviation; *p < 0.05, ***p < 0.001.

Fig. 2

The group-average principal functional gradient of patients (baseline and posttreatment) and controls. L, left; Right. The color bar indicates gradient score.

Fig. 3

Global histogram of the principal gradient among patients (baseline and posttreatment) and controls.

Fig. 4

Regional-parcel principal gradient score comparisons between patients at baseline and healthy controls. Left panel: Cortical surface maps showing regions with significant inter-group differences; Right panel: Bar graph representation of the principal gradient scores for the identified regions. The warm color clusters denote increased regions. The color bar indicates t values. See Table 2 for the specific meaning of the abbreviation of the cluster names. **p < 0.01, ***p < 0.001.

Fig. 5

Regional-parcel principal gradient score comparisons between patients after treatment relative to baseline. Left panel: Cortical surface maps showing regions with significant inter-group differences; right panel: Bar graph representation of the principal gradient scores for the identified regions. The cold clusters denote decrease regions. The color bar indicates t values. See Table 3 for the specific meaning of the abbreviation of the cluster names. **p < 0.01.

Fig. 6

Network-level comparisons of the principal gradient scores across groups. Left panel: Bar graph representation of the principal gradient scores for subnetworks among the groups; Right panel: The radar chart shows the gradient Z-score (with respect to controls) of subnetworks among the groups. VIS, visual network; SMN, somatomotor network; DAN, dorsal attention network; VAN, ventral attention network; LN, limbic network; FPC, frontoparietal network; DMN, default mode network. *p < 0.05.

Fig. 7

Subnetwork dispersion comparison of principal gradient scores among the groups. A, Within-network comparison of principal gradient scores and their gradient Z-score (with respect to controls) among the groups; B, Between-network dispersion comparison and their gradient Z-score (with respect to controls) among the groups. Abbreviations: VIS, visual network; SMN, somatomotor network; DAN, dorsal attention network; VAN, ventral attention network; LN, limbic network; FPC, frontoparietal network; DMN, default mode network. *p < 0.05.

Fig. 8

Spearman rank correlation between the changed principal gradient scores in the temporal lobe and improved clinical symptoms. Abbreviations: AHRS, auditory hallucinations rating scale.