Abstract
Individuals experiencing generalized anxiety disorder (GAD) show heightened attention to threat, as suggested by greater amplitude of the late positive potential (LPP). However, amplitude measurements do not fully capitalize on the high temporal resolution of EEG. Specifically, amplitude does not reflect the rate of change in the LPP over a window of interest, which may be important to understand LPP dynamics in individuals with GAD. Indeed, this rate of change of the LPP (i.e., LPP slope) may reflect attentional orienting. The current study leveraged multilevel models to examine the LPP in relation to GAD symptoms. We hypothesized that more-positive LPP slopes to threat images will be associated with GAD symptoms from 400 to 700 ms. Participants (N = 105) passively viewed blocks consisting of threatening or neutral images during EEG recording. Participant-level LPP slopes were estimated using and extracted from multilevel models, and the extracted slopes were examined. LPP slopes were reliable, but they only weakly correlated with mean amplitudes-suggesting LPP slopes may capture an attentional process that could be distinct from that captured by mean amplitudes. When considered as concurrent predictors of GAD, in an early window of the LPP (400-700 ms), the conjunction of the threat-LPP slope and the threat-LPP mean amplitude explained three times as much variance in GAD symptoms as mean amplitude did alone. During a later window of the LPP (700-2000 ms), more-negative LPP slope responses to threat were also related to GAD symptomology. Sensitivity analyses demonstrated that the relationship between the threat-LPP slope and GAD symptoms was largely robust to measurement confounds. Together, the current study is the first to identify that LPP slope is uniquely related to GAD symptoms. Our data further suggest that LPP slope is a unique measure of the broader LPP response that warrants further investigation.