Depressive Symptoms and Subcortical Volumes Predict Changes in Each Other in Adolescent Girls

• In Event: Poster Session 13
  In Poster Session: PS 13 Topic - Biological Processes: Neuroscience and Genetics

Abstract

Adolescence is marked both by the emergence of sex differences in depression vulnerability and a sex-differentiated maturation of striatal regions involved in depression (Newhouse & Albert, 2015; Salk et al., 2017). Alongside their three-fold greater risk of depression, adolescent girls consistently display smaller volumes in the nucleus accumbens (NAcc), caudate nucleus, and putamen - controlling for brain size -, as well as declines in caudate nucleus, putamen, and NAcc volumes not observed in adolescent boys (Herting et al., 2018). Despite this concurrence, the patterns of longitudinal relations between depressive symptoms and the volumes of these three striatal regions remain unclear.

To address this knowledge gap, 232 anti-depressant naïve adolescent girls from the Pittsburgh Girls Study-Emotions (PGS-E) sub-study were assessed for depressive symptom severity using the Adolescent Symptom Inventory and underwent an annual magnetic resonance imaging scan four times at ages 16 to 19 years. The volumes of the NAcc, caudate nucleus, and putamen were reconstructed, segmented and labeled as the three striatal regions of interest (ROIs) using FreeSurfer v5.3.0 (http://surfer.nmr.mgh.harvard.edu/). Three competing forms of structural equation models - cross-lagged panel, bivariate latent growth curve and dual change score - were used to approximate the longitudinal relations between these ROI volumes and depressive symptoms over the four-year study period to determine the existence of different change patterns between depressive symptoms and each ROI’s volume across the examined age range.

The relations between depressive symptoms and NAcc and putamen volumes were best modeled using a bivariate growth curve model, whereas relations between depressive symptoms and caudate volumes were best modeled using a dual change score model. On average, NAcc and caudate volumes decreased as girls aged, as indicated by the mean values of the latent slopes’ (μ = -.100, p < .001and μ = -.085, p = .021, respectively) significant differences from zero. Differing from NAcc and caudate nucleus volumes, putamen volumes evinced no significant longitudinal changes. Greater initial values in NAcc volumes predicted decreases in depressive symptoms across the examined age range, ρ = -.056, p = .024, whereas greater depressive symptoms predicted relatively greater subsequent NAcc volumes, ρ = .024, p = .046. Caudate volumes were not a significant predictor of later changes in depressive symptoms, γ = -.324, p = .144, however, greater depressive symptoms predicted greater subsequent caudate nucleus volumes, γ = .360, p = .007. No significant associations were found between depressive symptoms and putamen volumes, despite the overall pattern of association between these two variables over time mirroring those between depressive symptoms and NAcc and caudate nucleus volumes.

Notably, these findings raise the possibility that, between the ages of 16 and 19, larger NAcc volumes and attenuated patterns of caudate nucleus and NAcc volume decline result from compensatory mechanisms functioning to limit depressive symptoms in adolescent girls. Whereas the volumetric declines observed in the caudate nucleus and NAcc are consistent with prior work, the relation of putamen volume patterns to prior findings should be explored further.

Authors

• Sarah I Hudson, University of California - Davis
  Presenting Author

• Grant S. Shields, University of California, Davis
  Non-Presenting Author

• Veronika Vilgis, University of California, Davis
  Non-Presenting Author

• Erika E Forbes, University of Pittsburgh
  Non-Presenting Author

• Alison E. Hipwell, University of Pittsburgh
  Non-Presenting Author

• Kate Keenan, University of Chicago
  Non-Presenting Author

• Amanda E Guyer, University of California - Davis
  Non-Presenting Author