Deep brain stimulation improves symptoms across all dimensions in treatment-resistant depression

Abstract

Deep brain stimulation (DBS) reduces depressive symptom scores in many patients with treatment-resistant depression (TRD). However, it is unclear whether the observed improvement is similar across various symptom dimensions (e.g., anhedonia, anxiety, insomnia) or if some require additional clinical attention. Using a retrospective chart review, we assessed the trajectory of HAM-D-17 and MADRS scores during vALIC or slMFB DBS treatment within different symptom dimensions (HAM-D-17: 1) affective/anhedonia, 2) somatic/anxiety, 3) insomnia; MADRS: 1) affective/anhedonia, 2) anxiety/vegetative, 3) hopelessness) after at least a 25 ​% symptom reduction (partial response) at any time during their treatment course (n ​= ​34 for HAM-D-17, n ​= ​25 for MADRS). Results showed that each of the assessed symptom dimensions was significantly reduced compared to baseline at each of the assessed time periods (last follow-up: 2-15 years) after (partial) DBS response onset, which occurred at a median of approximately 2.5 months. Additionally, there was a significant interaction effect between symptom dimension and time period (HAM-D-17: F (12,1655.46) ​= ​5.46, p ​< ​0.001; MADRS: F (12,938.73) ​= ​2.40, p ​< ​0.01). Model coefficients indicated that insomnia symptoms (HAM-D-17) and anxiety/vegetative symptoms (MADRS) improved at a slower rate than the other symptom dimensions. Additionally, higher baseline scores in the HAM-D-17 somatic/anxiety dimension were significantly associated with a larger percentage reduction in overall symptoms after DBS (n ​= ​39, F (1,32) ​= ​12.371, p ​< ​0.01). Our findings demonstrate that DBS for TRD effectively treats depressive symptoms in all dimensions, although insomnia symptoms may improve at a slower rate, and that patients with more anxiety symptoms, who typically tend to have worse pharmacological treatment outcomes, may particularly benefit from DBS.

Keywords: Deep brain stimulation; Symptom dimensions; Treatment-resistant depression.

Fig. 1

Estimated trajectory of HAM-D-17 scores in different symptom dimensions after (partial) DBS response onset. Displayed are the linear mixed model estimates, which reflect the estimated fraction of the HAM-D-17 total dimension scores (affective/anhedonia, somatic/anxiety, insomnia) present at different times after (partial) DBS response onset (0 ​= ​postoperative baseline). The model estimates are displayed at the mean number of days since the (partial) response onset of the assessments during each time period. For illustration purposes, after 6–12 months the x-axis is not shown according to scale (shaded purple). There was a significant interaction between HAM-D-17 dimension and time period (F (12,1655.46) ​= ​5.46, p ​< ​0.001). The (initial) reduction in symptom severity after (partial) response onset was larger in the affective/anhedonia dimension compared to the somatic/anxiety dimension and the insomnia dimension, and the insomnia dimension demonstrated a slower, more prolonged reduction.

Fig. 2

Estimated trajectory of MADRS scores in different symptom dimensions after (partial) DBS response onset. Displayed are the linear mixed model estimates which reflect the estimated fraction of the MADRS total dimension scores (affective/anhedonia, anxiety/vegetative, hopelessness) present at different times after (partial) DBS response (0 ​= ​postoperative baseline). The model estimates are displayed at the mean number of days since postoperative baseline of the assessments during each time period. For illustration purposes, after 6–12 months the x-axis is not shown according to scale (shaded red). There was a significant interaction between MADRS dimension and time period (F (12,938.73) ​= ​2.40, p ​< ​0.01). The (initial) reduction in symptom severity after (partial) response onset was larger in the affective/anhedonia dimension and the hopelessness dimension compared to the anxiety/vegetative dimension, and the anxiety/vegetative dimension demonstrated a slower, more prolonged reduction.