Longitudinal Measures Of Adolescent Anxiety And Their Relationship To Adult Amygdala Function

• In Event: Poster Session 05
  In Poster Session: PS 05 Topic - Developmental Psychopathology

Abstract

Anxiety disorders make up the most prevalent category of mental disorders in the United States (Kessler, Chiu, Demler, & Walters, 2005), and are associated with negative effects in quality of life and daily functioning (e.g. Mendlowicz & Stein, 2000). Significant biological sex differences have been observed to emerge in middle childhood, putting females at greater risk (Klein & Pine, 2002; Hankin, 2009). As such, understanding the biological correlates of anxiety, as well as its trajectory across the lifespan, is of great importance. Anxiety is often associated with the amygdala, a brain region responsible for threat processing and fear (Tye et al., 2011). The present study uses self-reported anxiety data from the Minnesota Twin and Family Study, a large-scale longitudinal study of twins beginning at age 11 and continuing in 3-year intervals into adulthood. Furthermore, due to the addition of functional magnetic resonance imaging (fMRI) measures at an adult time point, we can assess whether trajectories of anxiety correspond to brain functioning in adulthood. Research questions: 1) How does subjective anxiety change across adolescence? 2) Does biological sex moderate the trajectory of anxiety symptoms? 3) Is adult amygdala reactivity predicted by concurrent anxiety symptoms, or their trajectory of change? We hypothesize that anxiety will peak in early adolescence, corresponding with increased emotionality and psychopathology during that time (Guyer, Silk, & Nelson, 2013), and that the baseline and peak anxiety scores will be higher for female respondents. Analysis of amygdala reactivity will be exploratory. Although the amygdala is considered central to anxiety, threat, and fear conditioning, a longitudinal relationship between anxiety and amygdala reactivity has yet to be established (Rauch, Shin, & Wright, 2003). Analysis: We use two population-based samples of twins (N1=1,482, N2=998). Participants were given the Spielberger State-Trait Anxiety Inventory (STAI-C or STAI; Spielberger, 1983). In adulthood, participants completed an amygdala activating fMRI task (Hariri, Bookheimer, & Mazziotta, 2000). Longitudinal multilevel modeling will be conducted, excluding individuals with fewer than three timepoints. Non-independence will be accounted for by modeling each twin pair as an indistinguishable dyad, such that twin order is unimportant (Kashy, Donnellan, Burt, & McGue, 2008). Statistical packages “nlme” and “dyadr” will be used in R for dyadic modeling. Sex and amygdala reactivity are included as time-invariant variables. Akaike Information Criterion (AICc) will determine the best-fitting model. Image preprocessing was conducted using FMRIB’s Software Library (FSL; www.fmrib.ox.ac.uk/fsl/) and included motion correction, slice timing correction, skull stripping, spatial and temporal filtering, and grand-mean scaling. A mask for the amygdala was defined anatomically in standard space by the Harvard-Oxford Subcortical Atlas and constrained using a post-interpolation threshold of 0.7. The present study offers a unique opportunity to examine associations between adolescent anxiety symptomatology and adult emotional reactivity in a dataset with robust sample size. Repeated measures during a key period of vulnerability can provide insight into development of psychopathology. Feasibility for these analyses is high, as data collection and image preprocessing is complete and model construction is underway.

Authors

• Shreya Lakhan-Pal, University of Minnesota - Twin Cities
  Presenting Author

• Stephen M Malone, University of Minnesota
  Non-Presenting Author

• Ruskin H Hunt, University of Minnesota, Twin Cities
  Non-Presenting Author

• William G Iacono, University of Minnesota
  Non-Presenting Author

• Kathleen M. Thomas, University of Minnesota - Twin Cities
  Non-Presenting Author