Major depressive episodes over the course of 7 years and hippocampal subfield volumes at 7 tesla MRI: The PREDICT-MR study

Highlights

• Number of major depressive episodes was associated with smaller subiculum volumes.

• Number of major depressive episodes was not associated with other subfield volumes.

• History or number of episodes was not associated with total hippocampal volume.

Abstract

Introduction

Smaller hippocampal volumes have been associated with major depressive disorder (MDD). The hippocampus consists of several subfields that may be differentially related to MDD. We investigated the association of occurrence of major depressive episodes (MDEs), assessed five times over seven years, with hippocampal subfield and entorhinal cortex volumes at 7 tesla MRI.

Methods

In this prospective study of randomly selected general practice attendees, MDEs according to DSM-IV-R criteria were assessed at baseline and after 6, 12, 39 and 84 months follow-up. At the last follow-up, a T2 (0.7 mm³) 7 tesla MRI scan was obtained in 47 participants (60±10 years). The subiculum, cornu ammonis (CA) 1 to 3, dentate gyrus&CA4 and entorhinal cortex volumes were manually segmented according a published protocol.

Results

Of the 47 participants, 13 had one MDE and 5 had multiple MDEs. ANCOVAs, adjusted for age, sex, education and intracranial volume, revealed no significant differences in hippocampal subfield or entorhinal cortex volumes between participants with and without an MDE in the preceding 84 months. Multiple episodes were associated with smaller subiculum volumes (B=−0.03 mL/episode; 95% CI −0.06; −0.003), but not with the other hippocampal subfield volumes, entorhinal cortex, or total hippocampal volume.

Limitations

A limitation of this study is the small sample size which makes replication necessary.

Conclusions

In this exploratory study, we found that an increasing number of major depressive episodes was associated with smaller subiculum volumes in middle-aged and older persons, but not with smaller volumes in other hippocampal subfields or the entorhinal cortex.

Introduction

Smaller hippocampal volumes are a common, but not a universal, finding in major depressive disorder (MDD) (Arnone et al., 2012, Du et al., 2012, McKinnon et al., 2009). The hippocampal formation is of importance in MDD for several reasons, as described in a recent review (MacQueen and Frodl 2011): 1) stressful life events are associated with the onset of MDD, but are also thought to be elicited by MDD (Kessler, 1997), and stress is also associated with smaller hippocampal volumes (Sapolsky et al., 1990, Sousa et al., 1999); 2) memory impairments that are dependent on hippocampal functioning, are often observed in MDD (Zakzanis et al., 1998); and 3) the hippocampal formation is part of a network that is often found to be disregulated in MDD (Frodl et al., 2008, Frodl et al., 2010, Zakzanis et al., 1998).

The hippocampal formation is not a homogeneous structure but consists of anatomically and functionally distinct subfields: the subiculum, cornu ammonis (CA) sections 1–4 and the dentate gyrus (Duvernoy et al., 2005). Tightly connected to the hippocampal formation lies the entorhinal cortex (Duvernoy et al., 2005). Because of their anatomical and functional differences, subfields of the hippocampal formation and the entorhinal cortex are likely to be differentially associated to aging and disease (Small et al., 2011). For instance, a recent review indicated that in MDD the subiculum is mainly reduced (Small et al., 2011), potentially because, according to animal studies, the subiculum has more cortisol binding sites (Reul and de Kloet, 1985) and receives a stronger serotonergic innervation from the raphe nucleus (Muller and Jacobs, 2010) compared to other subfields. Also the subiculum is connected to the basolateral nuclear group of the amygdala (Aggleton, 2000), which is also affected in MDD (Sheline et al., 1998). On the other hand post-mortem studies (Boldrini et al., 2009, Stockmeier et al., 2004) and stress research in animals (Sapolsky et al., 1990, Sousa et al., 1999) point to the involvement of CA3 and the dentate gyrus in MDD. However, it should be noted that these studies often did not include the subiculum in their analyses. Investigating whether hippocampal subfield volumes are differentially associated to MDD may provide insight in the neural substrate of MDD.

In recent years, the association between MDD and hippocampal subfield volumes has received increasing attention in MRI studies. The majority of these studies performed surface mapping (Ballmaier et al., 2008, Bearden et al., 2009, Cole et al., 2010, Gold et al., 2014, Posener et al., 2003, Sexton et al., 2012, Tae et al., 2011), a method in which inward deformations of the outer surface of the hippocampal formation are measured and reductions in subfield volumes are derived from these local deformations. Advantages of this method are that it can be performed at standard field strength (1.5 and 3 tesla (T) MRI) and that it is very sensitive to small changes. A limitation of this method is that the dentate gyrus, located within the hippocampal formation, cannot be assessed. Several of these studies reported inward deformations of the subiculum in patients with MDD compared to controls (Ballmaier et al., 2008, Bearden et al., 2009, Cole et al., 2010, Posener et al., 2003, Tae et al., 2011), but surface reductions in other subfields have also been reported (Ballmaier et al., 2008, Bearden et al., 2009, Cole et al., 2010, Tae et al., 2011). Fewer studies measured subfield volumes and/or entorhinal cortex volume (Gerritsen et al., 2011, Gold et al., 2010, Huang et al., 2013, Kook et al., 2012) and these studies reported inconsistent findings. These studies assessed volumes of hippocampal subfields at field strengths ranging from 1.5 to 4.7 T. Recent advances in ultra-high field imaging (e.g. 7 T MRI) have made it possible to visualize the hippocampal formation in great detail in three dimensions, with resolution and signal-to-noise ratio that is superior to lower field strength scanners. Recently, we developed a reliable protocol for the assessment of hippocampal subfield and entorhinal cortex volumes covering most of the longitudinal axis of the hippocampal formation using isotropic 0.7 mm 7 T MRI data (Wisse et al., 2012). In this study we will use this newly developed protocol to investigate the association of occurrence of major depressive episodes (MDEs), assessed five times over seven years, with hippocampal subfield and entorhinal cortex volumes at high field 7 T MRI.