¹China Medical University Beigang Hospital, Division of Psychiatry, Yunlin, Taiwan, ²King's College London, Centre for Healthy Brain Ageing, London, United Kingdom, ³Asia University, School of Nursing, Taichung City, Taiwan, ⁴Harvard Medical School, Department of Psychiatry, Boston, MA, United States, ⁵Harvard Medical School, Depression Clinical and Research Program, Massachusetts General Hospital, Boston, MA, United States

Background: Device-based neurostimulations can be effective treatments for major depressive disorder (MDD), but they are usually costly and need multiple in-office sessions performed by clinicians. Near-infrared transcranial photobiomodulation (tPBM) could be a safe alternative. A recent review published in the Lancet evaluating emerging treatments for MDD depicted ‘evidence exists’ in support of tPBM for MDD, but ‘nearly all studies lacked a control group’ (Marwaha et al. Lancet. 2023). Our published study (Guu et al. J Affect Disord. 2025) was the first randomised controlled trial investigating the feasibility and preliminary efficacy of a self-administered wearable tPBM device for MDD. We found this self-operable treatment model safe and feasible. In addition, notable sleep improvements but only modest antidepressant effects were found in our study, which were associated with high placebo response in the sham arm. As depression and sleep disturbance are often intertwined with each other, and placebo effects commonly nullify the demonstrable therapeutic effects in clinical trials, understanding factors influencing the responses in mood and sleep symptoms may enhance treatment outcomes for MDD patients, and facilitate future clinical trial design.

Aims & Objectives: In this secondary analysis, we aimed to:
1. Eexplore potential mechanisms underlying the sleep enhancing effects of self-operable near-infrared transcranial photobiomodulation.
2. Identify factors that contribute to the placebo effects.

Method: We examined data from a randomised, double-blind, sham-controlled trial (N = 48). Patients with moderate to severe MDD were randomised to tPBM (n = 25) or sham (n = 23) for 8 weeks, followed by 4 weeks of follow-up period. We investigated participant characteristics such as socio-demographics, clinical features, symptomatology, post-treatment satisfaction level, and their association with the outcomes. Factors potentially contribute to the different outcomes on the mood and sleep symptoms were explored using linear mixed-effects models.

Results: In the sham arm, greater clinician-rated improvement in depression was linked to greater self-rated depression improvement (rated by Beck’s Depression Inventory (BDI) scores; β = –0.35, p < 0.001) and better sleep quality (rated by Pittsburgh Sleep Quality Index; β = –0.998, p = 0.002) across follow-ups, as well as greater self-reported symptom change (β = 4.09, p < 0.001) and device satisfaction (β = 2.37, p = 0.004) in the post-treatment interview. In the tPBM group, clinician-rated improvement was only significantly linked to BDI reduction (β = –0.34, p = 0.002). Improvement in sleep started to differ from week 4 onward, and among all the sub-symptoms of sleep, participants in tPBM group had significantly reduced sleep latency (p = 0.025) and better efficiency (p = 0.026) at week 8, with persistent sleep latency improvement at week 12 (p = 0.030).

Discussion & Conclusions: High satisfaction level of this self-administration model potentially contributed to the high placebo response in our study, while improvements in sleep latency and efficiency appeared to be the major contributors to the overall better sleep found only in the tPBM group. Our findings suggest potentially different antidepressant mechanisms between the two groups that await larger studies for validation. Future study could also verify whether 'self-operability' could ameliorate depressive symptoms through improving self-efficacy.