Music in depression: Neural correlates of emotional experience in remitted depression

Abstract

Aim: To investigate neural and behavioral correlates of emotional experiences as potential vulnerability markers in remitted depression.

Methods: Fourteen remitted participants with a history of major depression and fourteen closely matched healthy control participants took part in the study. We used two psychiatric interviews (Hamilton Depression Rating Scale, Montgomery-Asberg Depression Rating Scale) and one self-report scale (Beck Depression Inventory) to assess remission. Healthy control participants were interviewed by an experienced psychiatrist to exclude those who showed any current or lifetime psychiatric or neurological disorders. To explore psychosocial and cognitive-interpersonal underpinnings of potential vulnerability markers of depression, early life stress, coping styles and alexithymia were also assessed. We induced pleasant and unpleasant emotional states using congruent combinations of music and human emotional faces to investigate neural and behavioral correlates of emotional experiences; neutral stimuli were used as a control condition. Brain responses were recorded using functional magnetic resonance imaging. Behavioral responses of pleasantness, arousal, joy and fear were measured via button-press inside the resonance imaging scanner.

Results: The mean age of the sample was 54.9 (± 11.3) years. There were no differences between remitted depressed (RD) (n = 14; 9 females and 5 males) and healthy participants (n = 14; 8 females and 6 males) regarding age, current degree of depression, early life stress, coping styles and alexithymia. On a neural level, RD participants showed reduced activations in the pregenual anterior cingulate cortex (pgACC) in response to pleasant [parameter estimates: -0.78 vs 0.32; t(26) = -3.41, P < 0.05] and unpleasant [parameter estimates: -0.88 vs 0.56; t(26)= -4.02, P < 0.05] emotional stimuli. Linear regression analysis revealed that pgACC activity was modulated by early life stress [β = -0.48; R(2) = 0.23, F(1,27) = 7.83, P < 0.01] and task-oriented coping style [β = 0.63; R(2) = 0.37, F(1,27) = 16.91, P < 0.001]. Trait anxiety modulated hippocampal responses to unpleasant stimuli [β = 0.62; R(2) = 0.38, F(1,27) = 15.95, P < 0.001]. Interestingly, in their reported experiences of pleasantness, arousal, happiness and fear in response to pleasant, unpleasant and neutral stimuli, RD participants did not differ significantly from healthy control participants. Adding trait anxiety or alexithymia as a covariate did not change the results.

Conclusion: The present study indicates that, in euthymic individuals, depression history alters neural correlates, but not the subjective dimension of pleasant and unpleasant emotional experiences.

Keywords: Anterior cingulate cortex; Early life stress; Emotion; Functional magnetic resonance imaging; Mood disorders; Music; Remission.

Figure 1

Reduced pg anterior cingulate cortex activation in response to pleasant > neutral stimuli in remitted depressed patients as compared to healthy controls. Left side corresponds to left hemisphere, MNI coordinates [-6 44 10], height and extend threshold: P < 0.05, false discovery rate corrected on peak-level, k = 62.

Figure 2

Hippocampal activation in response to unpleasant > neutral stimuli explained by trait anxiety in remitted depression [linear regression analysis; β = 0.62; R² = 0.38, F(1,27) = 15.95, P < 0.001]. Left side corresponds to left hemisphere, MNI coordinates [-32 -31 -3], k = 46. The result remains significant (P < 0.05) after excluding the two participants with the strongest hippocampal responses from the analysis. STAI: State and Trait Anxiety Inventory.