Table 4 Diffusion tensor imaging studies excluded from meta-analysis conducted in patients with both major depressive disorder and late-life depression (age >65 years) using other forms of analysis such as regions of interest or tractography^a

Lu et al. [39]

Tractography

1.5 Tesla

13 non-colinear directions

23 MDD patients, 24 controls

Increased neural connections between areas involved in mood regulation in MDD subjects, such as right superior frontal gyrus to right pallidum and left superior parietal gyrus to right superior occipital gyrus

Increased neural connections in MDD patients in brain areas associated with mood regulation

Cullen et al. [38]

Tractography and voxel-based analysis

3.0 Tesla

30 non-colinear directions

14 healthy MDD adolescents, 14 controls

Reduced FA in WM tracts connecting subgenual ACC to amygdala in right hemisphere and right and left uncinate to supragenual cingulum

Altered frontolimbic neural pathways in adolescent depression

Shimony et al. [37]

ROI analysis superficial: superior, middle and inferior frontal gyri, medial and lateral orbital frontal, dorsal, ventral and ACC, mesial frontopolar cortex, motor cortex, medial temporal gyrus, fusiform gyrus, auditory cortex, somatosensory cortex and posterolateral intraparietal sulcus, occipital pole and visual cortex

Deep: ventral, dorsal, and posterior frontal lobes, temporal and parietal lobes

1.5 Tesla

6 directions

73 LLD subjects, 23 controls

Reduced FA in prefrontal regions compared to controls

DTI abnormalities may be correlated with reduced cognitive processing speed

Yang et al. [17]

ROI analysis: DLPFC, parahippocampal gyrus and genu and body of corpus callosum

1.5 Tesla

25 diffusion-weighted directions

31 LLD subjects, 15 controls

Significantly reduced FA values in superior and middle frontal gyrus and right parahippocampal gyrus

Microstructural abnormalities in temporal and frontal areas of brain are associated with LLD

Nobuhara et al. [10]

ROI analysis: frontal WM 8 mm above AC-PC, on AC-PC, 8 mm below AC, genu and splenium

Bilaterally: temporal WM, parietal WM and occipital WM

1.5 Tesla

6 non-colinear directions

13 LDD patients, matched controls

Significantly reduced FA in the WM of frontal, temporal and occipital brain regions and corpus callosum of LLD individuals

Frontal, temporal and orbitofrontal WM deficits may play a role in symptom severity in LLD

Li et al. [35]

ROI analysis: prefrontal WM at 4 mm inferior to and at 0, 4, 8, 12, 16 and 20 mm superior to AC-PC plane

1.5 Tesla

13 non-colinear directions

51 LDD individuals

Significantly reduced FA values in prefrontal WM at bilateral 20 mm, right

16 mm and right 12 mm above AC-PC. No significant correlation

between FA and illness course or severity

Prefrontal WM abnormalities may occur early in the course of MDD and may play a role in the pathophysiology

Murphy et al. [36]

Voxel-based analysis

1.5 Tesla

8 diffusion sensitization directions

51 LDD individuals, no controls

Reduced FA in WM lateral to ACC and posterior cingulate cortex and in prefrontal, insular and parahippocampal

regions

Alterations in the frontostriatal-limbic networks may be associated with

executive dysfunction of LLD as measured on the Stroop task

Bae et al. [34]

ROI analysis: WM of superior and middle frontal gyri of DLPFC, anterior corpus callosum and anterior limb of internal capsule

1.5 Tesla

6 diffusion-weighted directions

106 LLD subjects, 84 elderly euthymic controls

Significantly reduced FA in WM of the right ACC, bilateral superior frontal gyri, and left middle frontal gyrus

Reduced FA in dorsolateral prefrontal cortex and ACC suggests altered brain connections are associated with LLD

Taylor et al. [16]

ROI analysis bilaterally: periventricular WM, frontal WM, anterior parietal WM, posterior parietal WM and caudate thalamus

1.5 Tesla

6 non-colinear

directions

29 LLD patients and 20 controls

Depressed nonremitting subjects showed fewer changes in ACC

Nonremitting subjects showed fewer FA changes, which may reflect antidepressant failure