Neuromelanin-Sensitive MRI Contrast and Chronic Depression in Young Women

Abstract

Importance: Whether chronic depression can be distinguished from nonchronic depression by mesolimbic dopamine hypofunction is a long-standing but untested hypothesis.

Objective: To determine whether cumulative mesolimbic dopamine function, assessed with neuromelanin-sensitive magnetic resonance imaging (NM-MRI) contrast, is associated with chronic depression in young women.

Design, setting, and participants: In the Adolescent Development of Emotions and Personality Traits (ADEPT) prospective cohort study, depressive disorders in women were assessed by diagnostic interview at regular intervals from ages 14 to 22 years (between July 2, 2012, and March 5, 2021). The current analysis includes a subset of ADEPT participants (aged 20-24 years) who completed a cross-sectional NM-MRI imaging study conducted between June 15, 2019, and May 31, 2021. Data were analyzed from March 27, 2024, to June 30, 2025.

Main outcomes and measures: Course of depression was rated using semistructured interviews. The nonparametric permutation test (whole-mask voxelwise analysis) of NM-MRI contrast in 2060 voxels within the substantia nigra and ventral tegmental area complex was used to test the association of cumulative mesolimbic dopamine function with chronic depression.

Results: Of the 166 young women contacted for this study, 140 had no MRI contraindications and were eligible. A total of 118 women underwent imaging, and 13 were excluded for MRI artifacts. Therefore, 105 women (mean [SD] age, 21.6 [0.91] years) were included in this analysis. There were 9 women (8.6%) with chronic depression (mean [SD], 45.6 [21.7] months), 28 (26.7%) with nonchronic depression (mean [SD], 6.0 [5.4] months), and 68 (64.8%) with no lifetime history of depression. Women with chronic depression exhibited significantly lower bilateral midbrain NM-MRI contrast than both women with nonchronic depression (795 negative-sign suprathreshold voxels, corrected P = .005) and women with no lifetime history of depression (692 negative-sign suprathreshold voxels, corrected P = .01), who did not differ from each other. Furthermore, lower bilateral midbrain NM-MRI contrast findings were associated with lower trait extraversion, a personality trait that reflects blunted reward sensitivity and is implicated in chronic depression, acquired at time of MRI (1061 positive-sign suprathreshold voxels, corrected P = .002), as well as from an earlier assessment at age 14 years prior to the onset of depression (1054 positive-sign suprathreshold voxels, corrected P = .001).

Conclusions and relevance: In this cohort study of young women, chronic depression was associated with reduced NM-MRI contrast, consistent with cumulative mesolimbic dopamine hypofunction. NM-MRI contrast findings were normal in young women with nonchronic depression. These findings support the importance of distinguishing between chronic and nonchronic forms of illness in parsing the heterogeneity of depressive disorders, and they suggest that low dopamine may play a role in the etiopathophysiology of chronic depression.

Figure 1.. Neuromelanin-Sensitive Magnetic Resonance Imaging (NM-MRI) Contrast in Women Aged 20 to 24 Years

A, The grand mean NM-MRI contrast image from all participants at coronal (top left), axial (bottom), and sagittal (top right) views at Montreal Neurological Institute (MNI) coordinates 84, 106, 56; 95, 106, 56; and 83, 110, 56. B, Grand mean NM-MRI contrast image as in A with the subnuclei mask overlaid from coronal (top left), axial (bottom), and sagittal (top right) views at MNI coordinates 84, 106, 56; 95, 106, 56; and 83, 110, 56, respectively. Blue indicates substantive nigra pars reticulata (SNr); green, substantia nigra pars compacta (SNc); red, parabrachial pigmented nucleus (PBP); and purple, ventral tegmental area (VTA). C, Mean NM-MRI contrast of voxels at locations attributed to each subnucleus. Error bars were calculated as 1 times the SE. Other SN includes remaining voxels in the SN-VTA mask that had low probability of inclusion in the other 4 subnuclei. Dots are the mean NM-MRI contrast for VTA, SNr, SNc, PBP, and other SN ROIs for each participant. NS indicates not significant.

Figure 2.. Neuromelanin-Sensitive Magnetic Resonance Imaging (NM-MRI) Contrast and Chronicity of Depression

A, Heat map of t statistics depicting the association between log₁₀-transformed depression months and voxels in the substantia nigra (SN) and the ventral tegmental area (VTA) controlling for age overlaid on the average NM-MRI contrast image from all participants at Montreal Neurological Institute (MNI) coordinates 87, 108, 56 from axial (left side), coronal (middle), and sagittal (right side) views. B, Group differences in the mean NM-MRI contrast in SN-VTA mask (region of interest [ROI]) by depression group (eTable 2 in Supplement 1). Each dot is one participant’s mean NM-MRI contrast. The box is drawn around the median of the lower half quartile and the median of the upper half quartile. The error bars are drawn to the first and last value, excluding outliers. C and D, Scatterplot showing the mean NM-MRI contrast extracted from suprathreshold voxels in Figure 1A using a leave-one-out approach plotted against log₁₀-transformed depression months (C) and extraversion (D) for visualization purposes. Each dot is one participant's data. The line is the ordinary lease squares regression. It demonstrates the slope of the association between the two variables

Figure 3.. Proportion of Significant Voxels in Subregion

Reference dashed lines are drawn at the proportion of supraheight threshold voxels in whole-mask voxelwise analysis (2060 voxels). Bars that surpass the dashed lines indicate that more voxels in that subnuclei mask were supraheight threshold than expected by chance, given the total number of supraheight threshold in mask. Bars below the dashed lines indicate that fewer voxels in that subnuclei mask were supraheight threshold than expected by chance, given total number of supraheight threshold in mask. For instance, the reference line is drawn at 0.37 (756 of 2060) for log₁₀-transformed depression months, which is the proportion of negative-sign supraheight threshold voxels from voxelwise analysis. The dashed line corresponds to the positive-sign supraheight threshold voxels for extraversion variables, the negative-sign supraheight threshold voxels for depression months, and the supraheight threshold voxels for group comparisons (no lifetime history of depression > chronic depression, no lifetime history of depression > nonchronic depression, and nonchronic depression > chronic depression). Voxels at the ventral tegmental area (VTA) coordinates, as well as the substantia nigra pars compacta (SNc) and parabrachial pigmented nucleus (PBP) coordinates, were disproportionately supraheight threshold for chronic depression and extraversion. Voxels in the substantive nigra pars reticulata (SNr) subnuclei coordinates were not disproportionately supraheight threshold for chronic depression and extraversion. Other SN indicates remaining voxels in the SN-VTA mask that had low probability of inclusion in the other 4 subnuclei. MRI indicates magnetic resonance imaging; W1, wave 1.