Electrophysiological Correlates of Cognition in Depression

1. Background and Significance Depictions of psychiatric illnesses as stemming largely from biological and genetic vulnerabilities have increased substantially in recent years (Deacon, 2013; Lebowitz \& Appelbaum, 2019; Schomerus et al., 2012). These messages are disseminated by physicians, pharmaceutical companies, anti-stigma campaigns, researchers, and the popular media alike. Although at first blush messages emphasizing genetic susceptibility may seem helpful in reducing stigma, growing research points to serious unintended consequences (Haslam \& Kvaale, 2015). Specifically, when participants believe their depression is due to biogenetic abnormalities, they expect to suffer for longer periods of time (Kemp, Lickel, \& Deacon, 2014), endorse more depressive symptoms (Lebowitz \& Ahn, 2017), and feel they have less control over their mood (Lebowitz \& Ahn, 2018). Moreover, biogenetic messaging has no impact on stigma (Haslam \& Kvaale, 2015). Despite the accumulating self-reported evidence that biogenetic messaging may be harmful, nothing is known about how such messages impact neural correlates of self-reflection and cognitive control - two key processes thought to subserve adaptive self-regulation that may be disrupted among individuals with major depressive disorder (MDD, Pizzagalli, 2011). This study will fill this knowledge gap by comparing resting and task-related electroencephalography (EEG) between adults with MDD randomly assigned to receive either positive or negative information about their genetic susceptibility to depression. Completion of this project will characterize the neural impacts of widespread messages about the etiology of depression. This study may inform clinical decision making, public policy, and guidelines regarding how mental health is discussed.
2. Specific Aims:

Aim1: To examine the impact of biogenetic messaging on default-mode network (DMN) Hypothesis 1: The DMN refers to a network of functionally connected brain regions that are most active at rest and during retrospection (Buckner, Andrews-Hanna, \& Schacter, 2008; Raichle, 2015). The DMN has been consistently found to be overactive in the context of depressive disorders (Pizzagalli, 2011), especially in the context of elevated rumination. Capitalizing on approaches to probe DMN functionality using source-localized EEG activity implemented in the mentor's lab (Whitton et al., 2018) the investigators expect that the DMN will be increased following the positive (vulnerable) genetic feedback manipulation. This would indicate that biogenetic messaging increases potentially maladaptive rumination.

Aim 2: To examine the impact of biogenetic messaging on cognitive control Hypothesis 2: Cognitive control refers to a suite of functions that allow humans to monitor, detect, and respond to conflicting information and mistakes, and to mobilize internal resources to resolve such occurrences from happening in the future (Braver, 2012; Miller \& Cohen, 2001). One commonly studied facet of cognitive control is error monitoring, which refers to the ability to detect and respond to mistakes. The error-related negativity (ERN) is elicited 0-100ms following an error and the error positivity (Pe) is elicited 200-400ms post-error (Gehring, Liu, Orr, \& Carp, 2012). Post-error behavioral adjustments include post-error slowing and post-error improvement in accuracy. Previous research suggests that depressive symptoms correlate with ERN and Pe amplitudes (Compton et al., 2008; Holmes \& Pizzagalli, 2008; Olvet, Klein, \& Hajcak, 2010; Schroder, Moran, Infantolino, \& Moser, 2013). Induction of genetic messaging about intelligence increased the Pe amplitude but also reduced the correlation between Pe and post-error performance (Schroder, Moran, Donnellan, \& Moser, 2014). Accordingly, in the current study, the investigators expect the Pe to be increased and a reduced relationship between Pe and post-error behavior in the vulnerable genetic condition.

Aim 3: To evaluate self-reported motivation for treatment, expectancies, and preferences Hypothesis 3: Previous research has documented a cost in self-reported motivation and future expectancies following receiving biogenetic information about depression (Kemp et al., 2014; Lebowitz \& Ahn, 2017). The investigators expect to replicate these effects in a sample of individuals with MDD. The investigators expect that participants receiving vulnerable genetic feedback will 1) endorse poorer perceived control over their emotions, 2) expect to have depression for a longer period of time, 3) endorse a preference for pharmacotherapy versus psychotherapy and 4) view pharmacotherapy as more effective than psychotherapy.

1. Description of the Research Design Participants The sample will consist of 80 male and female unmedicated adults with MDD, aged 18-45. Participants will be recruited primarily through Cragslist ads, flyering, and contacting participants who were previously enrolled in studies at the Center for Depression, Anxiety and Stress Research. After passing an initial phone screen, participants will complete the Mini International Neuropsychiatric Interview (MINI; Sheehan et al., 1988). Exclusion criteria for all participants will include failure to meet EEG safety requirements, current drug use, history of alcohol and drug dependence, lifetime history of psychosis and bipolar disorder, and imminent suicidal ideation. After the interview, participants will be asked to complete the Beck Depression Inventory (BDI-II, Beck, Steer, \& Brown, 1996), the Quick Inventory of Depressive Symptoms (QIDS, Rush et al., 2003), the Ruminative Response Style Questionnaire (RRS, Treynor, Gonzalez, \& Nolen-Hoeksema, 2003), the Penn State Worry Questionnaire (PSWQ, Meyer, Miller, Metzger, \& Borkovec, 1990), the Positive And Negative Affective Schedule (PANAS, Clark \& Watson, 1991) and Visual Analogue Mood Scale (VAMS, Aitken, 1969).

Baseline EEG Assessment After participants are deemed eligible, they will complete the baseline EEG assessment. Participants will be fitted with a 96-channel EEG cap. The baseline EEG assessment consists of two tasks. First, resting EEG data will be collected (8 min) in which participants will sit calmly with their eyes open or closed (randomly alternated in one-minute intervals). The resting EEG allows for collection of DMN. Second, participants will perform a flanker task (20 min). The flanker task is a well-validated task in which participants view five horizontal arrows on the computer screen and respond as quickly and as accurately to the central (target) stimulus using a response pad. Participants will complete 30 practice trials to titrate task difficulty in the main blocks, and 350 test trials. The ERN, Pe, behavioral adjustments and VAMS will be recorded from this task.

Saliva Sample and Genetic "Testing" Following completion of the flanker task, participants will be informed they will be taking a saliva sample to determine their genetic susceptibility to depression. Using a previously validated procedure (Lebowitz \& Ahn, 2017, 2018), participants will be provided with a "saliva testing kit", which consists of a plastic box containing a glucose test strip (which participants are led to believe gauges salivary levels of 5-Hydroxyindoleacetic acid (5-HIAA) as part of a genetic susceptibility test) and a small amount of mouthwash (containing glucose) in a plastic container. Participants will be provided with instructions on the computer screen for how to complete the saliva testing themselves. Participants will rinse their mouths with mouthwash for seven seconds, spit the mouthwash into the box, and insert the test strip under their tongues for 10 seconds, and then wait for 30 seconds. The test strip will turn brown as the strip is sensitive to glucose. Participants will be given a computer prompt to indicate which color their test strip turned (brown or pink) and will be randomly assigned to receive computer feedback indicating that a brown test strip means they 1) have a genetic vulnerability to depression or 2) do not have such a vulnerability. The feedback consists of one paragraph describing 5-HIAA and its implications for depression based on past research. The research assistant (RA) will be blind to condition assignment.

Post-Manipulation EEG and Self-reported Assessment Immediately following the genetic test manipulation, participants will complete the PANAS to assess state affects and then repeat the resting EEG recording and flanker task. They will then complete a battery of self-report measures to gauge their hypothetical mental health treatment preferences and expectancies, and perceived willingness to engage in treatment in the future. They will also complete the VAMS, the Implicit Theories Questionnaire (Schroder, Dawood, Yalch, Donnellan, \& Moser, 2015), the Negative Mood Regulation Scale (Catanzaro \& Mearns, 1990), the Perceptions of Depression Scale(Deacon \& Baird, 2009), and the Prognostic Pessimism Scale (Lebowitz, Ahn, \& Nolen-Hoeksema, 2013). Participants will also complete a manipulation check to assess perceived credibility of the genetic testing.

Debriefing Procedure At the end of the session, all participants will be thoroughly debriefed. Following previously published procedures (Lebowitz \& Ahn, 2017), debriefing will entail the Co-I - who has a PhD in clinical psychology - explaining that no genetic testing actually took place. The Co-I will explain that the mouthwash consisted of glucose and that when exposed to glucose, the test strip turns brown. Participants will be shown both feedback screens (susceptible and non-susceptible feedback). The concept of randomized assignment will be discussed. Participants will be encouraged to ask questions during this period. Finally, participants will complete a short quiz consisting of items that ask whether or not genetic testing took place. Participants will be required to respond accurately; if they do not respond accurately after debriefing, the Co-I will again emphasize that no genetic testing took place until full comprehension is achieved.