| Major depressive disorder (MDD) accounts for significant disability and economic loss worldwide. Hippocampus is critical to its etiology and treatment and is critical in hypothalamic-pituitary-adrenocortical axis feedback, which is altered in two-thirds of patients with MDD. Moreover, hippocampal neurogenesis appears necessary for antidepressant efficacy. Neuroimaging reveals ∼4% smaller hippocampal volume in recurrent or chronic (>2 years duration) MDD. The microscopic structural correlates underlying this smaller volume are yet unknown.;Postmortem tissues (n = 17 MDD-control pairs) were obtained at autopsy with informed consent of next-of-kin, who gave structured interviews for retrospective psychiatric assessment. Formalin-fixed left temporal lobes were sectioned (40 mum) and every 10th section stained for Nissl substance; other sections were archived in 70% ethanol for future use. Every 6th Nissl-stained section was selected throughout rostrocaudal length of hippocampus. Total volume of hippocampal formation was calculated. Then total numbers of pyramidal neurons (CA1, CA2/3, hilus), granule cells (DG-gcl), and glial cells were estimated stereologically.;Smaller hippocampal volume in recurrent/chronic MDD (n = 11) was not seen here. However, in recurrent/chronic MDD, total volume decreased with duration of MDD (r = -0.693, p < 0.027). Moreover, hippocampal tail volume decreased with percent of life with untreated MDD (n = 11; r = -0.697, p < 0.018). Total hippocampal length was elevated in MDD (p < 0.021). There was no difference in total number or density of pyramidal neurons/granule cells or glial cells in CA1, CA2/3, hilus, or DG-gcl. However, CA1 pyramidal neuron density increased with duration in recurrent/chronic MDD (r = 0.846, p < 0.002). DG-gcl granule cell number decreased with MDD duration in suicide (n = 10; r = -0.638, p < 0.048) or with no antidepressant prescription (n = 10; r = -0.739, p < 0.037). Granule cell (r = 0.971, p < 0.002) and glial cell numbers (r = 0.980, p < 0.001) increased with age in those taking antidepressants (n = 6).;While Nissl staining allows quantification of glial cells per se, immunohistochemistry is needed to identify specific glial cell types. Astrocytes contribute to neurovascular coupling, transmitter uptake, and structural integrity and influence neurogenesis. In light of astrocytic alterations in prefrontal cortex (PFC) in MDD, second aim here was to determine whether similar alterations occur in hippocampus.;Three sections from body of hippocampus in the same subjects were retrieved from ethanol and labeled for glial fibrillary acidic protein (GFAP). Adjacent Nissl-stained sections aided hippocampal field delineation. Astrocyte density was assessed stereologically in CA1, CA2/3, and hilus. GFAP-ir area fraction was measured in CA1, CA2/3, and DG.;Although neither measure differed between MDD and controls in CA1, both astrocyte density (r = 0.638, p < 0.006) and area fraction (r = 0.525, p 50 years; n = 22) versus younger subjects (n = 12) (p < 0.029). Adjusted for age, area fraction was selectively lower in females with MDD (n = 5) in DG (p < 0.043); n.s. trend toward same in CA2/3 (p < 0.064). Finally, adjusted for age, astrocyte density in hilus was lower in MDD subjects not taking antidepressants (n = 11) and restored to control levels in those taking antidepressants (p < 0.005).;Astrocytic alterations in MDD observed here were generally consistent with those in PFC. Age-related alterations may contribute to the greater Alzheimer risk in those with history of MDD. Meanwhile, selectively lower GFAP area fraction in women with MDD likely reflects altered organizational effects of ovarian steroid hormones and thus may help explain the greater risk for MDD in women with early menarche. Finally, apparent restoration of hilar astrocyte density by antidepressants is consistent with a role for astrocytes in the etiology and treatment of MDD and may reflect altered expression of factors responsive to glucocorticoids or direct effects of serotonin on astrocytes.;Decreasing total hippocampal volume with increasing CA1 pyramidal neuron density with duration in recurrent/chronic MDD lends support to the neuropil hypothesis of MDD. Reduced neuropil may reflect effects of elevated glucocorticoids, reduced neurotrophins, or altered endocannabinoid or pro-inflammatory signaling. Increasing DG granule cell and glial cell numbers with age in antidepressant-treated MDD likely reflect interaction of antidepressant drugs with glucocorticoids, neurotrophins, and endocannabinoid signaling. Finally, hippocampal tail volume decreasing with percent of life in untreated MDD has cognitive implications related to differences in anterior versus posterior hippocampal connectivity.;Overall, these results suggest hippocampal neuropathology in MDD involves a neuropil deficit, not neuron loss. Astrocytic pathology may contribute significantly to this deficit. |
