- 2026
- NeuroImage. Clinical 50
- Ravi Bansal
- David J Hellerstein
- Siddhant Sawardekar
- Ying Chen
- Bradley S Peterson
Study Design
- Type
- Randomized Controlled Trial (RCT)
- Population
- 80 adults with persistent depressive disorder (41 duloxetine, 39 desvenlafaxine) and 35 healthy controls
- Methods
- Two double-blind RCTs of duloxetine and desvenlafaxine, with DTI and clinical assessments at baseline and trial completion
- Blinding
- Double-blind
- Funding
- Unclear
Background
Prior studies have reported abnormalities in brain volumes and white matter microarchitecture in persistent depressive disorder(PDD). Randomized controlled trials(RCTs) yoked with brain MRIs are needed to discern whether medications change tissue microstructure in patients with persistent depressive disorder(PDD).Methods
We conducted two double-blind, RCTs of duloxetine and desvenlafaxine for treating PDD and collected prospective, longitudinal diffusion tensor imaging(DTI) data and clinical assessments at baseline and upon trial completion in 41 duloxetine and 39 desvenlafaxine patients. Data were combined across the two trials to increase statistical power. DTI data were also acquired once in 35 healthy controls. Fractional anisotropy and average diffusivity maps were computed to assess tissue microstructure. We hypothesized that medications would normalize gray and white matter microstructure within the limbic system.Results
Duloxetine and desvenlafaxine had unique treatment effects on tissue microstructure in the dorsal prefrontal cortex, especially in the superior frontal gyrus, superior longitudinal fasciculus, and anterior corona radiata. Duloxetine and desvenlafaxine also had common treatment effects in the limbic system: tissue microarchitecture normalized in medication-treated patients but deviated further away from values for healthy participants in placebo-treated patients. Furthermore, symptom severity significantly mediated the treatment effects on tissue microstructure, suggesting that abnormal tissue microstructure at baseline is at least in part a neuroplastic compensatory response to symptoms.Conclusions
Medication may have reduced the need for compensatory response because it reduces symptom severity, whereas placebo sustains the need for compensation. The unique and common effects of duloxetine and desvenlafaxine on neurotransmitter systems are likely responsible for their spatially unique and common effects in altering tissue microstructure.Trial registration
ClinicalTrials.gov identifiers: NCT01537068 (Desvenlafaxine study, 2/16/2012); and NCT00360724 (Duloxetine study, 08/03/2006).