Depression and anxiety in acute ischemic stroke involving the anterior but not paramedian or inferolateral thalamus

Abstract

Background and objectives: Emotional and cognitive deficits are prevalent in strokes involving the thalamus. In contrast to cognitive deficits, emotional deficits have not been studied prospectively in isolated thalamic stroke.

Methods: In 37 ischemic thalamic stroke patients (57.0 [50.0; 69.5] years [median (Q1; Q3)], 21 males, 5 anterior, 12 paramedian, 20 inferolateral vascular territory), and 37 non-stroke control patients matched for age and sex, we prospectively examined depression, anxiety, activities of daily living, and quality of life at 1, 6, 12, and 24 months post-stroke using the Hospital-Anxiety-and-Depression Scale (HADS), Nürnberger-Alters-Alltagsaktivitäten scale (NAA), and Short Form-36 (SF36) questionnaire. Voxel-based lesion-symptom mapping (VLSM) and lesion-subtraction analyzes were performed to determine associations between questionnaire scores and thalamic stroke topography.

Results: At 1 month post-stroke, anterior thalamic stroke patients had higher depression scores [8.0 (7.5; 10.5)] than paramedian [4.5 (1.0; 5.8)] and inferolateral [4.0 (1.0; 7.0)] thalamic stroke patients. Furthermore, anterior thalamic stroke patients had higher anxiety scores [11.0 (8.0; 14.5)] than their matched controls [2.5 (2.0; 2.5)], paramedian [4.5 (1.0; 5.8)] and inferior [4.0 (1.0; 7.0)] thalamic stroke patients. Depression and anxiety scores in anterior thalamic stroke patients remained high across the follow-up [depression: 9.0 (3.5; 13,8); anxiety:10.05 (2.8, 14.5)].Physical health assessed by SF36 was intact in anterior [1 month post-stroke: T-score = 55.9 (37.0; 57.6)] but reduced in inferolateral [44.5(32.4; 53.1)] thalamic stroke, whereas mental health was reduced in anterior thalamic stroke [32.0 (29.8; 47.3)].VLSM confirmed that voxels in the anterior thalamus around Montreal Neurological Institute (MNI) coordinates X = -8, Y = -12, Z = 2 were more often affected by the stroke in depressed (HADS-score ≥ 8) than non-depressed (HADS-score < 8) patients and voxels around coordinates X = -10, Y = -12, Z = 2 were more often affected in anxious (HADS-score ≥ 8) than non-anxious (HADS-score < 8) patients.

Conclusion: Anterior, but not paramedian or inferolateral thalamic stroke was associated with depression and anxiety. Even though our results are mostly significant in the left thalamus, this observation on stroke laterality might be confounded by the fact that the right hemisphere was underrepresented in our study.

Keywords: anxiety; brain infarct; depression; emotion; magnetic resonance imaging; voxel-based lesion-symptom mapping.

Figure 1

Overlap lesion map of all 37 thalamic stroke patients. The color bar indicates the frequency of lesion overlap in thalamic stroke patients. Numbers above each brain section indicate MNI Z axis. MNI, Montreal Neurological Institute, R, right; L, left.

Figure 2

Depression and anxiety assessed by the Hospital Anxiety and Depression Scale (HADS) at different time-points during the follow-up. Data are median (Q1;Q3) of (A) depression and (B) anxiety scores of thalamic stroke patients and controls., † p-value vs. matched controls, ‡ p-value vs. anterior thalamic stroke patients.

Figure 3

HADS depression evaluated by subtraction analysis and voxel-based symptom-lesion mapping (VLSM). (A) Subtraction analysis in which a lesion overlay of non-depressed patients (HADS score < 8) was subtracted from a lesion overlay of depressed patients (HADS score ≥ 8) demonstrating the lesion topography of depressed patients. The color bar indicates a relative differences value by how much more frequently (in percentage) each voxel was damaged in the depressed patient group than in non-depressed patient group (from purple = 0% = voxel was damaged equally in both groups to red = 100% = voxel was damaged in every patient in the depressed patient group but in none of the non-depressed patient group). (B) VLSM analysis for depressed (HADS score ≥ 8) vs. non-depressed (HADS score < 8) patients. The color bar demonstrates z-values of the Liebermeister tests. (C) Enlargement of (B) at Z coordinate 2 next to the lesion location mapping according to Morel’s anatomical thalamic atlas. Numbers above each brain indicate MNI Z coordinates. Only voxels significant at p ≤ 0.05 are shown. MNI, Montreal Neurological Institute, R, right; L, left.

Figure 4

HADS anxiety assessed by subtraction analysis and VLSM. (A) Subtraction analysis in which a lesion overlay of non-anxious patients (HADS score < 8) was subtracted from a lesion overlay of anxious patients (HADS score ≥ 8) demonstrating the lesion topography of anxious patients. The color bar indicates a relative differences value by how much more frequently (in percentage) each voxel was damaged in the anxious patient group than in non-anxious patient group (from purple = 0% = voxel was damaged equally in both groups to red = 100% = voxel was damaged in every patient in the anxious patient group but in none of the non-anxious patient group). (B) VLSM analysis for anxious (HADS score ≥ 8) vs. non-anxious (HADS score < 8) patients. The color bar demonstrates z-values of the Liebermeister tests. (C) Enlargement of (B) at Z coordinate 2 next to the lesion location mapping according to Morel’s anatomical thalamic atlas. Numbers above each brain indicate MNI Z coordinates. Only voxels significant at p ≤ 0.05 are shown. MNI, Montreal Neurological Institute, R, right; L, left.

Figure 5

NAA and SF36 evaluated by subtraction analysis. (A) Subtraction analysis in which a lesion overlay of patients without ADL impairment (NAA score < 33) was subtracted from a lesion overlay of patients with ADL impairment (NAA score ≥ 33) demonstrating the lesion topography of ADL impaired patients. The color bar indicates a relative differences value by how much more frequently (in percentage) each voxel was damaged in ADL impaired than in ADL non-impaired patients (from purple = 0% = voxel was damaged equally in both groups to red = 100% = voxel was damaged in every patient in the ADL impaired patient group but in none of the ADL non-impaired patient group). (B) + (C) Subtraction analysis where a lesion overlay of patients without physical/mental health impairment (SF36 > 40) was subtracted from a lesion overlay of patients with physical/mental health impairment (SF36 ≤ 40) demonstrating the lesion topography of physical/mental health-impaired patients. The color bar indicates a relative differences value by how much more frequently (in percentage) each voxel was damaged in physical/mental health impaired than in physical/mental health non-impaired patients (from purple = 0% = voxel was damaged equally in both groups to red = 100% = voxel was damaged in every patient in the physical/mental health impaired patient group but in none of the physical/mental health non-impaired patient group).