Subcortical brain alterations in major depressive disorder: findings from the ENIGMA Major Depressive Disorder working group

Abstract

The pattern of structural brain alterations associated with major depressive disorder (MDD) remains unresolved. This is in part due to small sample sizes of neuroimaging studies resulting in limited statistical power, disease heterogeneity and the complex interactions between clinical characteristics and brain morphology. To address this, we meta-analyzed three-dimensional brain magnetic resonance imaging data from 1728 MDD patients and 7199 controls from 15 research samples worldwide, to identify subcortical brain volumes that robustly discriminate MDD patients from healthy controls. Relative to controls, patients had significantly lower hippocampal volumes (Cohen's d=-0.14, % difference=-1.24). This effect was driven by patients with recurrent MDD (Cohen's d=-0.17, % difference=-1.44), and we detected no differences between first episode patients and controls. Age of onset ⩽21 was associated with a smaller hippocampus (Cohen's d=-0.20, % difference=-1.85) and a trend toward smaller amygdala (Cohen's d=-0.11, % difference=-1.23) and larger lateral ventricles (Cohen's d=0.12, % difference=5.11). Symptom severity at study inclusion was not associated with any regional brain volumes. Sample characteristics such as mean age, proportion of antidepressant users and proportion of remitted patients, and methodological characteristics did not significantly moderate alterations in brain volumes in MDD. Samples with a higher proportion of antipsychotic medication users showed larger caudate volumes in MDD patients compared with controls. This currently largest worldwide effort to identify subcortical brain alterations showed robust smaller hippocampal volumes in MDD patients, moderated by age of onset and first episode versus recurrent episode status.

Figure 1

Cohen's d-effect sizes 95% CI and for differences in subcortical brain volumes between major depressive disorder (MDD) patients and healthy control subjects. Effect sizes were corrected for age, sex and intracranial volume (ICV). The effect size for ICV was corrected for age and sex. *P<0.05 corrected. CI, confidence interval.

Figure 2

(a) Cohen's d-effect sizes 95% CI for differences in subcortical brain volumes between recurrent major depressive disorder (MDD) patients and healthy control subjects (striped pattern) and between first episode MDD patients and healthy controls (no pattern). (b) Cohen's d-effect sizes 95% CI for differences in subcortical brain volumes between early onset (⩽21) MDD patients and healthy control subjects (no pattern) and between later onset (>21) MDD patients and healthy controls (striped pattern). Effect sizes were corrected for age, sex and intracranial volume (ICV). *P<0.05 corrected, **P<0.05. CI, confidence interval.