Madelyn G Jungbluth1*
1 Department of Psychiatry, University of Michigan, 4250 Plymouth Road, Ann Arbor, Michigan, 48109
Abstract
Individuals with Posttraumatic Stress Disorder (PTSD) often also experience depression. The way people learn and remember things, especially how to differentiate between threatening and safe contexts, may be intricately intertwined with both PTSD and depression. Previous research has utilized fear conditioning paradigms to examine how individuals with these disorders learn to associate and remember cues that predict threat and safety in different contexts. Past studies suggest that fear-based learning and memory are associated with the activation of the hippocampus and amygdala. The purpose of this study was to investigate how activation in these regions during fear memory is associated with the severity of depressive symptoms in participants with PTSD. We hypothesized that greater depressive symptoms in participants with PTSD would be associated with greater hippocampal and amygdala activation in the safe context and less activation in the threat context. We recruited adults with PTSD and varying levels of depressive symptoms to complete a contextual fear conditioning and memory paradigm during fMRI scanning. Results suggest that individuals with PTSD and more severe depressive symptoms exhibited lower hippocampal activation in the safe context, as well as greater hippocampal activation in the threat context. These findings suggest that higher depression scores are associated with an impaired ability to remember safety in the safe context and threat in the threat context. Future research should recruit more participants with PTSD and varying levels of depressive symptoms to more thoroughly investigate these relationships.
Introduction
Posttraumatic stress disorder (PTSD) is a psychiatric disorder that affects approximately 5% of adults in the United States each year (U.S. Department of Veterans Affairs 2023). Traumatic events, such as natural disasters, accidents, terrorist acts, combat or assaults, can lead to PTSD, which can manifest as intense, intrusive and disturbing thoughts or feelings related to the traumatic event that continue long after the event has occurred (Torres 2020b). These symptoms include nightmares, flashbacks, heightened startle responses, negative emotions, difficulty concentrating or sleeping, isolation or estrangement from others and avoidance of trauma reminders (Torres 2020b). In addition to PTSD, depressive symptoms can also develop following trauma (Torres 2020b). Depression is one of the most common mental illnesses, characterized by persistent sadness, loss of interest, changes in appetite and sleep patterns, feelings of worthlessness or guilt, difficulty focusing or concentrating and thoughts of death or suicide (Torres 2020a).
While depression or PTSD alone can cause significant distress and impairment in an individual’s life, they often co-occur. It is estimated that approximately half of people who suffer from PTSD also meet diagnostic criteria for major depressive disorder, and the presence of depression in people with PTSD can lead to worse health outcomes (Flory and Yehuda 2015). Individuals with comorbid PTSD and depression are more likely to experience reduced life satisfaction, lower quality of life and greater severity of symptoms compared to those with either disorder alone (Rosen et al. 2020).
Evidence from clinical neuroscience research suggests that activity in brain regions underlying cognitive and emotional processes, including the hippocampus and amygdala, is related to both PTSD and depression (Akiki et al. 2017; Yao et al. 2020). The hippocampus is involved in learning, memory and spatial navigation, whereas the amygdala is responsible for emotions, emotional regulation, motivation and fear conditioning (Anand and Dhikav 2012; Wright 2020). While the brain regions involved in these disorders are well established in scientific literature, less is known regarding the neural correlates underlying these disorders when they exist comorbidly. The current study aimed to investigate whether activation in the hippocampus and amygdala during memory for threat and safety was associated with depressive symptoms in individuals with PTSD.
One important similarity between PTSD and depression is impaired fear processing (Kemp et al. 2007). The way people learn to differentiate whether they are in a safe or a threatening situation relates to how fear is handled in the brain. Specifically, individuals with PTSD often exhibit fear in otherwise safe contexts or environments, such as during firework shows or after an accidental touch (Torres 2020b).
In one experiment examining fear conditioning and memory in PTSD, researchers utilized a contextual fear conditioning paradigm to investigate how well participants could differentiate between distinct threat and safety cues within different contexts (Garfinkel et al. 2014). Participants were fear-conditioned through the presentation of computerized images of library and office scenes, each of which contained lamps that illuminated pink or blue for each trial. One light cue (pink or blue) in one context (library or office) was paired with a 500ms shock delivered to the index and middle fingers, resulting in a “threat context,” while the shock was never paired with the light in the other context, resulting in a “safe context.” Individuals with PTSD did not differ from healthy controls in their ability to learn these associations, but they exhibited reduced memory performance the following day (Garfinkel et al. 2014).
Specifically, participants with PTSD exhibited impaired memory for safety, as shown by greater skin conductance response (SCR) and amygdala activation in the safe context (extinction recall), as well as impaired memory for threat, as shown by lower SCR and hippocampal and amygdala activation in the threat context (fear renewal). Essentially, participants with PTSD exhibited high ‘fear’ during the safe context (extinction recall), in which no aversive stimulus had been presented. However, they demonstrated a lack of expected ‘fear’ in the threat context (fear renewal), in which the aversive stimulus was presented before. These results suggest that individuals with PTSD are less able to effectively use context to differentiate between threat and safety cues (Garfinkel et al. 2014). Similarly, researchers in another study observed less hippocampal activation during extinction recall in participants with PTSD, suggesting that fear extinction may be impaired in individuals with PTSD and that the brain structures that underlie impairments in fear memory may contribute to the symptoms experienced by individuals with PTSD (Milad et al. 2009).
Few studies have investigated the association between depressive symptoms and fear conditioning and memory; however, prior research has studied the effects of depressive symptoms on other types of memory. In a study conducted by Milne et al. (2012), researchers observed attenuated hippocampal activation in participants with depression compared to healthy controls during a memory recollection task. These findings support the hypothesis that the disease burden of depression impacts hippocampal functioning and that memory recollection performance in individuals with depression may be associated with diminished activation in the hippocampus (Milne et al. 2012).
Taken together, findings from previous studies suggest that, individually, symptoms of PTSD and depression are associated with impairments in memory, specifically fear memory. However, while fear conditioning and memory in PTSD and depression have been studied extensively independently, no studies have looked at these processes in individuals with comorbid PTSD and depression.
The purpose of the current study was to investigate how functioning in the hippocampus and amygdala is associated with severity of depressive symptoms in participants with PTSD during a fear memory task. This study examined the latter portion of a two-day computerized contextual fear conditioning and memory paradigm. The present study examined solely memory because prior research has suggested that individuals with PTSD exhibit no impairment during learning stages (fear conditioning or extinction), but they do have deficits in extinction recall and fear renewal, which are the stages involved with memory (Garfinkel et al. 2014).
In order to examine this aim, we proposed the following two hypotheses. First, we hypothesized that greater hippocampus and amygdala activation would be associated with more severe depressive symptoms in individuals with PTSD during extinction recall. Second, we hypothesized that lower hippocampus and amygdala activation would be associated with more severe depressive symptoms in individuals with PTSD during fear renewal.
Materials and Methods
Participants
Right-handed adults aged 18-45 were recruited to participate in this study. We recruited participants from the greater Ann Arbor and Metro-Detroit community using University of Michigan’s UM Health Research recruitment website, social media advertisements and flyers posted around the campus and surrounding community. Participants included in this study were recruited in the context of a larger study examining learning and memory processes in PTSD (full results to be reported elsewhere). The analyses reported here represent secondary analyses of a subsample of the participants from the larger study. Participants were compensated $130 for completing all aspects of this study.
In order to be eligible for the study, participants must have been able to give informed consent, and they must have been free of other brain or cognitive disorders. Exclusion criteria included prior diagnosis of a brain disorder; cognitive impairment; learning disability; neurological disorder (MS, seizure, stroke, tumor, ADHD, etc.); serious head injury; inability to have an fMRI scan; active pregnancy; left-handedness or ambidextrousness; or a substance use disorder in the last year.
There were 19 participants with PTSD that completed all aspects of the study. 17 (89%) individuals in the cohort identified as women and two (11%) as men. 16 (84%) participants identified as white, one (5%) as black and two (11%) as other. The cohort had an age range of 19-45 (M = 26.8), and the cohort was composed of individuals with depression scores ranging from 0 (minimal) to 50 (severe; M = 22.2). However, four participants were excluded from final analysis. One participant was excluded due to bad tissue contrast in the fMRI scans, which prevented data processing; one participant was excluded due to missing behavioral data; one participant was excluded because they fell asleep during the tasks; and one participant was excluded due to not completing the fMRI portion of the study.
Following exclusions, we analyzed results from a cohort of 15 individuals who had data for at least one of our tasks. 13 (87%) individuals in the cohort identified as women, and two (13%) identified as men. 13 (87%) participants identified as white, and two (13%) participants identified as other. The cohort had an age range of 19-44 (M = 25.4), and participants exhibited depression scores ranging from 0 (minimal) to 50 (severe; M = 22.8).
All participants provided written informed consent prior to beginning study procedures, and approval for this study was obtained from the University of Michigan’s Institutional Review Board (HUM00121812). Dr. Elizabeth Duval was the authorized principal investigator for this study.
Measures
Participants underwent diagnostic screening to determine eligibility, and all participants included in this study had a primary diagnosis of PTSD based on the Clinician Administered PTSD Scale (CAPS) (Weathers et al. 2013). The CAPS is a 30-item diagnostic PTSD assessment interview that evaluates both current and lifetime PTSD diagnosis and symptom severity. Scores range from 0-80, and higher scores indicate greater PTSD symptom severity (Weathers et al. 2013).
Level of depressive symptoms was assessed with the Beck Depression Inventory (BDI) (Beck et al. 1961; Beck et al. 1988). The BDI is a 21-item self-report rating inventory that measures symptoms and behaviors characteristic of depression. Scores range from 0-63, and higher scores indicate greater depressive symptom severity (Beck et al. 1961; Beck et al. 1988).
MRI scanning was performed on a 3.0 Tesla GE Discovery MR750 System (Waukesha, WI) using a state-of-the art 32-channel radiofrequency coil and updated software (Discovery 20.0, Neuro-optimized gradients). T1-weighted anatomic images were acquired with a 3D MPRAGE sequence (FOV = 256 x 256mm, slice thickness = 1mm, 0mm gap). Functional scans consisted of gradient echo blood oxygen level dependent (BOLD) scans with standard parameters across scanners and tasks similar to the following: TR/TE = 2000/30ms, flip angle = 90 degrees, FOV = 192 x 192mm and slice thickness = 3mm.
Task and Procedure
Similar to other fear conditioning and extinction procedures, participants completed a series of fear conditioning and memory tasks over the course of two consecutive days. The first day consisted of fear conditioning and extinction learning via a computerized contextual fear conditioning and memory paradigm that occurred in a mock MRI scanner. The second day consisted of extinction recall and fear renewal via the same computerized contextual fear conditioning and memory paradigm but, this time, in a real fMRI scanner. This fear conditioning and memory paradigm was a modified version of a commonly used paradigm to assess fear learning and memory (Milad et al. 2007; Garfinkel et al. 2014). Number of trials, timing and rate of stimulus pairings were the same as in these prior studies. We utilized novel stimuli created in our lab that depicted different contexts (rooms) and cues (light colors). As prior studies have utilized a shock or a burst of white noise as the aversive unconditioned stimulus, we utilized a burst of white noise (Garfinkel et al., 2014; Javanbakht et al., 2016).
Task and Procedure: Day One
Participants were positioned in a mock MRI scanner to simulate a real MRI scan. A recording of an MRI scan was played during the duration of the tasks. A computer screen was viewed through mirrored goggles, and participants were given headphones and a button box (with each finger corresponding to a number 1-5) connected to their right hand. Before the fear conditioning paradigm began, participants completed a habituation task in order to become familiar with the stimuli for our task. For this portion, they saw two different contexts (living room and office scenes) that contained light cues. The light cues were presented as different colors of light (orange and blue) that turned on in a lamp fixture. Contexts and cues were counterbalanced across participants, resulting in 4 versions of the task corresponding to the unique combinations of contexts (living room or office) and cues (orange or blue light).
After habituation, the fear conditioning portion of our study began. In the fear conditioning portion, participants viewed one of the two contexts (office or living room) for 2-7 seconds. The light then turned on (either blue or orange) within the context for an additional 2-7 seconds (for a total of 9 seconds per trial). The light cues represented the conditioned stimuli (CS); one light cue was associated with the presentation of an aversive unconditioned stimulus (US; 500ms loud burst of white noise) in 60% of trials. Thus, participants learned that this light cue was predictive of threat (CS+). The other light cue was never associated with the US and thus was not predictive of threat (CS-; Figure 1a). For example, if the blue light in the living room was associated with the presentation of the burst of loud white noise (US), participants would learn that the blue light was predictive of threat in the living room. Therefore, they learned to exhibit a fear response when presented with the blue light. On the other hand, as the CS- was never paired with the US, it should not elicit a fear response. From these trials, participants learned to fear one light color, as it predicted the presence of the US, but not the other light color, as that signified safety (no US). The inter-trial interval (ITI) was a fixation cross, which was presented in jittered intervals of 4-12 seconds. There were three runs of this task, and 8 CS+ and CS- trials were presented in each run. This task lasted approximately 15 minutes in the mock scanner.
During the extinction portion of the task, participants saw the other context (office or living room; the option not presented during fear conditioning), as well as the two light cues. However, during this phase, neither light was paired with the US, which resulted in extinction of the CS+ (known as CS+E) (Figure 1b). As such, participants learned that the CS+ no longer signified threat in this new context. There were two runs of this task, and 8 CS+E trials were presented along with 8 CS- trials in random order within each run. This task lasted approximately 10 minutes in the mock scanner.
Figure 1. a. Example of stimulus presentations during fear conditioning. b. Example of stimulus presentations during extinction.
Task and Procedure: Day Two
The second day consisted of extinction recall and fear renewal to test participants’ memory of the information learned during fear conditioning and extinction. Unlike the first day of the study, this portion of the study occurred in a real fMRI scanner. MRI scans were completed to assess brain activation during extinction recall and fear renewal tasks. A high resolution structural scan was also completed. Most study procedures were identical to the first day. For the second day of the study, participants completed this task by looking at a computer screen through mirrored goggles in the fMRI scanner. Participants were instructed to remain as still as possible and to not fall asleep during the scanning.
During extinction recall, participants’ remembering that the CS+E was not predictive of threat in the extinction context (the safe context) was assessed by presenting each of the light cues in the context previously viewed during the extinction phase. This task was conducted to test whether participants remembered that this context is safe (since neither light cue predicted threat in this context). Similar to the other tasks, participants saw images of the living room or the office with both light cues (orange and blue) (Figure 2a). Participants were once again shown eight CS+E trials and eight CS- trials randomly presented in each of two runs. This task lasted about ten minutes in the fMRI scanner.
During fear renewal, the ability to remember that the CS+ was predictive of threat in the conditioning context (the threat context) was assessed by presenting each light cue (orange or blue) in the room previously viewed in the fear conditioning phase (Figure 2b). This was completed in order to test whether participants were able to remember that this context is threatening (as one light color predicted threat in this context). Once again, eight CS+E trials and eight CS- trials were randomly presented within each of two runs. This task took about ten minutes in the fMRI scanner to complete.
Figure 2. a. Example of a trial during extinction recall. b. Example of a trial during fear renewal.
Data Processing and Statistical Analyses
To process and analyze the MRI data, we used custom scripts from our lab to run Statistical Parametric Mapping (SPM12) software for MATLAB. Preprocessing of the functional images followed established methods and included slice-time correction, realignment and coregistration to the structural images, normalization to the Montreal Neurological Institute (MNI) standard brain and smoothing (with a 6mm kernel). Runs with more than 3mm of motion in any of the six planes (x, y, z, pitch, roll and yaw) were excluded from analysis. In order to test for preprocessing quality, we conducted quality checks to evaluate image alignment (registration), spatial correspondence (normalization) and motion. Scans that did not pass these preprocessing quality checks were excluded from further analysis.
To ensure none of the participants slept during our tasks, we checked for visual cortex activation during trials in which any stimulus was presented compared to implicit baseline (black screen with white fixation cross). Any participants without visual cortex activation during this contrast were excluded from further analysis.
To measure brain activation in the hippocampus and amygdala during extinction recall and fear renewal, activation was quantified for trials that contained a conditioned stimulus (CS) cue that predicted threat (CS+ trials); this was in comparison to trials that contained a cue that did not predict threat (CS- trials). In order to evaluate the relationship between hippocampus and amygdala activation and depressive symptom severity in individuals with PTSD, we used correlation analyses. The p-value threshold was set to 0.01 uncorrected. Therefore, results should be treated as preliminary. The dependent measures of the current study included significant BOLD responses in the hippocampus and amygdala.
Results
Extinction Recall (ER)
Neural Activation in the Hippocampus and Amygdala during CS+ vs CS- Trials
To establish activation in brain regions involved in memory for safety, we examined both hippocampal and amygdala activation during CS+ compared to CS- trials during extinction recall. This comparison revealed whether CS+ and CS- cues in the safe (extinction) context were associated with a difference in neural activity. We found that participants exhibited less activation in right (coordinates [x, y, z] = 36, -19, -10) and left hippocampus (coordinates [x, y, z] = -26, -18, -12) during CS+ compared to CS- trials (T = 3.71, p = .002; T = 2.95, p = .002, respectively) (Figure 3a). In other words, participants exhibited greater activation during trials in which the aversive US was not presented (CS-) within the safe (extinction) context. There was no statistically significant difference in amygdala activation during CS+ compared to CS- trials.
Figure 3a. Greater bilateral hippocampal activity was observed in CS- trials compared to CS+ trials
Relationship Between Depression Score and Hippocampal and Amygdala Activation
To examine whether activation in the hippocampus and amygdala during extinction recall was associated with depression score, we tested the hypothesis that participants with more severe depressive symptoms would exhibit greater threat reactivity and therefore greater hippocampus and amygdala activation during CS+ compared to CS- trials. Essentially, this analysis quantifies the neural activity in the hippocampus and amygdala during CS+ vs CS- trials in relation to depression score. Contrary to our hypothesis, we found that higher depression scores were associated with less activation of the right anterior hippocampus (coordinates [x, y, z] = 27, -13, -22; T = 3.13, p = .005) during CS+ compared to CS- trials (Figure 3b).
Figure 3b. Panel A
Figure 3b. Panel B
Unilateral hippocampal activity (panel A) is negatively correlated with BDI score (panel B) during extinction recall.
Fear Renewal (FR)
Neural Activation in the Hippocampus and Amygdala during CS+ vs CS- Trials
To establish activation in brain regions involved in memory for threat, we compared hippocampal and amygdala activation between CS+ and CS- trials during fear renewal. This comparison revealed whether CS+ and CS- cues in the threat (fear conditioning/renewal) context were associated with a difference in neural activity. We found no statistically significant hippocampal or amygdala activation during CS+ trials compared to CS- trials.
Relationship Between Depression Score and Hippocampal and Amygdala Activation
To examine whether activation in the hippocampus and amygdala during fear renewal is associated with depression score, we tested the hypothesis that participants with more severe depressive symptoms would exhibit lower threat reactivity and therefore lower hippocampus and amygdala activation during CS+ compared to CS- trials. Essentially, this analysis quantifies the neural activity in the hippocampus and amygdala during CS+ vs CS- trials in relation to depression score. Contrary to our hypothesis, we found that higher depression scores were associated with greater activation of the left posterior hippocampus (coordinates [x, y, z] = -24, -31, -4; T = 3.20, p = .004) during CS+ compared to CS- trials (Figure 4).
Discussion
Figure 4. Panel A
Figure 4. Panel B
Unilateral hippocampal activity (panel A) is negatively correlated with BDI score (panel B) during extinction recall.
Overview
The purpose of this study was to investigate how function in the hippocampus and amygdala underlying fear memory is associated with severity of depressive symptoms in participants with PTSD. We hypothesized that greater depressive symptom severity in participants with PTSD would be associated with greater abnormality during extinction recall and fear renewal, and we attempted to quantify this greater abnormality by measuring function in the hippocampus and amygdala using fMRI technology during fear memory tasks. Specifically, we hypothesized that greater hippocampus and amygdala activation would be associated with more severe depressive symptoms during extinction recall, and we hypothesized that less hippocampal and amygdala activation would be associated with more severe depressive symptoms during fear renewal. Since prior research has suggested that individuals with PTSD and individuals with depression exhibit exaggerated fear responses and an impaired ability to differentiate between threat and safety cues during fear conditioning and memory respectively, we hypothesized that comorbid PTSD and depressive symptoms would be associated with these same effects but to an exaggerated level (Rosen et al. 2020).
We first examined task-based effects irrespective of depressive symptoms to establish activation in the hippocampus and amygdala during CS+ compared to CS- trials for extinction recall and fear renewal. We found that participants with PTSD exhibited less bilateral hippocampal activation during extinction recall trials (the safe context) in CS+ trials compared to CS- trials. These findings suggest that participants with PTSD exhibit less activation in the hippocampus when processing cues that indicated threat in the past but no longer do. They also suggest that the hippocampus is differentially activated during CS+ and CS- trials in individuals with PTSD. Furthermore, these findings may suggest impairments in the complex neural networks underlying memory and emotion, as previous research has shown that individuals with PTSD exhibit deficits in episodic and emotional memory processing (Layton and Krikorian 2002). As these findings are consistent with prior research conducted by Milad et al. (2009), the findings of the current study suggest that extinction recall is impaired in individuals with PTSD, and that the brain structures, such as the hippocampus, which underlie these impairments may contribute to symptoms characteristic of PTSD. Throughout our analyses of the extinction recall portion of our study, we observed no differences in amygdala activation between CS+ and CS- trials in participants with PTSD.
After examining patterns of brain activation to CS+ compared to CS- trials in PTSD, we examined how activation in the hippocampus and amygdala was associated with depressive symptoms. The first hypothesis was that greater hippocampal and amygdala activation would be associated with more severe depressive symptoms during extinction recall. Contrary to our hypothesis, our results suggest that greater depressive symptom severity was associated with less activation of the right anterior hippocampus during CS+ compared to CS- trials. In other words, greater depression scores were associated with lower hippocampal activation in the safe context.
Although no prior studies have examined the correlation between comorbid depressive symptoms and fear memory in participants with PTSD, these findings are consistent with some of the prior literature regarding PTSD. As noted by Milad et al. (2009), reduced hippocampal activation is associated with impaired recall of extinction memory, indicating a correlation with impaired fear memory processing. In the case of the current study, it is possible that comorbid depressive symptoms in individuals with PTSD is associated with a fear understatement effect, in which these individuals may struggle to retain an extinguished fearful memory even in the safe context. While these findings do not support our hypothesis, they suggest that individuals with PTSD and higher depression scores exhibit lower levels of hippocampal activation during extinction recall and that this reduction in activation may be related to impairments in extinction recall performance (Milad et al. 2009). Furthermore, these findings suggest that comorbid depressive symptoms are associated with impaired fear memory processing in individuals with PTSD and that this is associated with activation in the hippocampus. Taken together, our findings and those of prior studies highlight the unknown association between comorbidities and fear conditioning and memory processing, as well as the potential association between dysfunction in the hippocampus, safety memory impairments and PTSD and depression symptoms. More research is needed in order to investigate these relationships.
Next, we examined task-based effects irrespective of depressive symptoms to establish activation in the hippocampus and amygdala during CS+ compared to CS- trials for fear renewal. We found no statistically significant hippocampal or amygdala activation during fear renewal trials (the threat context) in CS+ trials compared to CS- trials in participants with PTSD.
The second hypothesis predicted that lower hippocampal and amygdala activation would be associated with more severe depressive symptoms during fear renewal. Contrary to our hypothesis, our results suggest that higher depression scores were associated with greater activation of the left posterior hippocampus during CS+ compared to CS- trials during fear renewal. In other words, higher depression scores were associated with heightened threat reactivity and greater hippocampal activation in the threat context.
Although no prior studies have examined the correlation between comorbid depressive symptoms and fear memory in participants with PTSD, these findings are inconsistent with some of the prior literature regarding PTSD. For example, the study conducted by Garfinkel et al. (2014) observed lower hippocampal activation and threat reactivity during fear renewal tasks. One possible explanation for the current findings is that the presence of comorbid depressive symptoms in individuals with PTSD is correlated with a hypersensitivity to threat cues. As noted by Garfinkel and Liberzon (2009), hypersensitivity to threat, as well as hypervigilance and hyperarousal, are hallmark traits of PTSD. Therefore, in the presence of a threat context, individuals with PTSD may be more reactive to remembering the fear, as shown by greater function in the hippocampus during fear renewal.
Since our findings support the hypothesis that greater hippocampal activation during fear renewal is positively correlated with depressive symptom severity, they suggest that this hypersensitivity to threat, a hallmark symptom of PTSD, may be associated with depressive symptom severity (Garfinkel and Liberzon 2009). In other words, hyperreactivity to threat cues, as observed by increased hippocampal activation during fear renewal, may be associated with depressive symptom severity in participants with PTSD.
Another potential explanation could be a general deficit in contextual processing, which is a hippocampal-dependent process, in participants with comorbid depressive symptoms and PTSD (Garfinkel et al. 2014). More specifically, previous research has shown that individuals with PTSD exhibit a reduced ability to utilize safety cues, so it is possible that the greater hippocampal activation observed in participants with comorbid depressive symptoms and PTSD could be related to this overarching difficulty to differentiate between threat and safety contexts (Jovanovic et al. 2012).
Overall, the findings of the current study suggest that individuals with PTSD and higher depression scores exhibit higher levels of hippocampal activation during fear renewal and that this greater activation in the hippocampus may be associated with heightened threat reactivity during fear renewal tasks (Garfinkel et al. 2014). However, as our findings differ with some of the previous literature regarding PTSD and depression, our findings taken together with the prior literature emphasize the unknown relationship between comorbidities and fear memory processing capabilities. Therefore, more research is needed to investigate these relationships.
Future Directions
In the future, it would be useful to continue the present study with the goal of recruiting more participants with PTSD and comorbid depressive symptoms. One limitation of our study was the small sample size, as fMRI data from only 15 participants were included in final analyses. Additionally, approximately 90% of our sample identified as white, and approximately 90% identified as women. Therefore, the sample is too small and is not diverse enough to reflect the general population. Future studies should recruit a larger sample more representative of the general population, with participants identifying with diverse gender and racial identities. This would hopefully allow for a more precise estimate of statistical significance, as well as more generalizability to the greater population of individuals with comorbid PTSD and depression.
Another potential future direction could be to look at the relationship between hippocampal and amygdala activation during extinction recall and fear renewal in individuals with comorbid PTSD and anxiety symptoms. Similar to depressive symptoms, anxiety symptoms are also often comorbid with a PTSD diagnosis. As such, this avenue for future research could prove clinically beneficial (Sareen 2014). For example, this research could contribute to the overarching knowledge base underlying the neural correlates of PTSD and anxiety, and those contributions could impact the development of novel treatment methods for these comorbid conditions. Therefore, future studies could investigate whether anxiety symptoms, in addition to or without comorbid depressive symptoms, are related to impairments in fear conditioning and memory in individuals with PTSD.
Other research could also look at these relationships in relation to other comorbidities. PTSD is often comorbid with other mental illnesses, such as substance use disorders, and it would be interesting to analyze the association between fear conditioning and memory in individuals with comorbid PTSD and substance use disorders (Roberts et al. 2016).
Conclusions
In summary, the current study aimed to investigate how brain function in the hippocampus and amygdala during fear memory is associated with depressive symptoms in participants with PTSD. Results support the notion that participants with PTSD exhibit impaired extinction recall capabilities, as shown by lower bilateral hippocampal activation during presentation of the extinguished threat cue in the safety context. Additionally, the findings from this study suggest that there is a relationship between depressive symptom severity and brain function during fear memory, as shown by reduced hippocampal activation to a previously threatening cue within a safe context and greater hippocampal activation to a previously threatening cue within the threat context. Differences in hippocampal activation associated with depression may suggest impaired fear memory.
These results provide preliminary findings that contribute to our understanding of the neural mechanisms underlying fear conditioning and memory in people with PTSD and depression. This knowledge may aid in future studies investigating the association between comorbid mental illnesses and brain functioning. Additionally, these findings highlight the importance of research on the neural mechanisms of comorbid mental illnesses, as there may be a confounding or even exponential association between comorbid symptom severity and overall brain functioning. As PTSD, depression and a plethora of other mental illnesses impact the livelihoods and happiness of countless individuals around the world, understanding the neural functioning underlying comorbid mental health conditions has multiple potential clinical implications (Rosen et al. 2020). This research can contribute to treatments positioned to address multiple conditions, as well as overall wellbeing for individuals with comorbid mental health disorders.
Acknowledgments
I am sincerely grateful to Dr. Elizabeth Duval for her guidance throughout this entire process, as well as for her immense support throughout my undergraduate years. Her mentorship and advice has helped me discover my passion for research, and I am thankful for all the support she gave to make this project happen. To Liz, my time working in your lab has been the highlight of my college career, and I am thankful for all the wonderful opportunities you have given and encouraged me to pursue. I hope you know that you have played an instrumental role in my decision to pursue research in the future.
The data used for this project were collected as part of a larger, ongoing study. I assisted with participant recruitment and data collection. For the specific project presented in this article, I completed all aspects of developing the questions and hypotheses, reviewing the literature, data organization, preprocessing, analysis and interpretation of findings. This project was completed to fulfill the requirements for an undergraduate honors thesis at the University of Michigan, and a version of this manuscript has been archived in Deep Blue.
I would like to thank Dr. Hanjoo Kim, Samantha Goldberg and Kayla Smith for introducing me to data analysis in academic research, helping to answer my research and career based questions, for contributing to the overall project on which my article is based, and for teaching me how to run in-person study visits. Mike Angstadt and the Psychiatry Neuroimaging Methods Core developed analysis scripts that I used as templates for my analyses. Additionally, I would like to thank Amanda Hicks, Simone Wilson, Sonalee Joshi and Rachel John for stimulus development, recruitment of participants and collection and organization of data. I would also like to thank Maddie Sage for her work organizing and processing the fMRI data with me and for offering to collaborate and work together in the composition of our honors theses this year. Finally, thank you to all of the wonderful research assistants in the Duval Lab who fostered such a welcoming and supportive lab environment. I feel extremely lucky to have been a part of this community, and I am extremely grateful for the wonderful opportunity to complete this project.
This project was supported by the National Institutes of Mental Health (Duval, K23MH109762; Noll, S10OD026738).
References
Akiki, T. J., Averill, C. L., Wrocklage, K. M., Schweinsburg, B., Scott, J. C., Martini, B., Averill, L. A., Southwick, S. M., Krystal, J. H., and Abdallah, C. G. (2017) ‘The association of PTSD symptom severity with localized hippocampus and amygdala abnormalities’, Chronic Stress, 1, available: https://doi.org/10.1177%2F2470547017724069
Anand, K. S. and Dhikav, V. (2012) ‘Hippocampus in health and disease: An overview’, Annals of Indian Academy of Neurology, 15(4), 239-246, available: https://doi.org/10.4103%2F09722327.104323
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J. and Erbaugh, J. (1961) ‘An inventory for measuring depression’, Archives of General Psychiatry, 4(6), 561-571, available: https://doi.org/10.1001/archpsyc.1961.01710120031004
Beck, A. T., Steer, R.A. and Carbin, M. G. (1988) ‘Psychometric properties of the Beck Depression Inventory: Twenty-five years of evaluation’, Clinical Psychology Review, 8(1), 77100, available: https://doi.org/10.1016/0272-7358(88)90050-5
Flory, J. D. and Yehuda, R. (2015) ‘Comorbidity between posttraumatic stress disorder and major depressive disorder: Alternative explanations and treatment considerations’, Dialogues in
Clinical Neuroscience, 17(2), 141-150, available: https://doi.org/10.31887/DCNS.2015.17.2/jflory
Garfinkel, S. N. and Liberzon, I. (2009) ‘Neurobiology of PTSD: A review of neuroimaging findings’, Psychiatric Annals, 39(6), 370-381, available: https://doi.org/10.3928/0048571320090527-01
Garfinkel, S. N., Abelson, J. L., King, A. P., Sripada, R. K., Wang, X., Gaines, L. M. and Liberzon, I. (2014) ‘Impaired contextual modulation of memories in PTSD: An fMRI and psychophysiological study of extinction retention and fear renewal’, The Journal of Neuroscience, 34(40), 13435-13443, available: https://doi.org/10.1523/JNEUROSCI.428713.2014
Javanbakht, A., Duval, E. R., Cisneros, M. E., Taylor, S. F., Kessler, D., and Liberzon, I. (2016) ‘Instructed fear learning, extinction, and recall: Additive effects of cognitive information on emotional learning of fear’, Cognition and Emotion, 31(5), 980-987, available: https://doi.org/10.1080/02699931.2016.1169997
Jovanovic, T., Kazama, A., Bachevalier, J. and Davis, M. (2012) ‘Impaired safety signal learning may be a biomarker of PTSD’, Neuropharmacology, 62(2), 695-704, available: https://doi.org/10.1016/j.neuropharm.2011.02.023
Kemp, A. H., Felmingham, K., Das, P., Hughes, G., Peduto, A. S., Bryant, R. A. and Williams, L. M. (2007) ‘Influence of comorbid depression on fear in posttraumatic stress disorder: An fMRI study’, Psychiatry Research: Neuroimaging, 155(3), 265-269, available: https://doi.org/10.1016/j.pscychresns.2007.01.010
Layton, B. and Krikorian, R. (2002) ‘Memory mechanisms in posttraumatic stress disorder’, The Journal of Neuropsychiatry and Clinical Neurosciences, 14(3), 254-261, available: https://doi.org/10.1176/jnp.14.3.254
Milad, M. R., Wright, C. I., Orr, S. P., Pitman, R. K., Quirk, G. J. and Rauch, S. L. (2007) ‘Recall of fear extinction in humans activates the ventromedial prefrontal cortex and hippocampus in concert’, Biological Psychiatry, 62(5), 446-454, available: https://doi.org/10.1016/j.biopsych.2006.10.011
Milad, M. R., Pitman, R. K., Ellis, C. B., Gold, A. L., Shin, L. M., Lasko, N. B., Zeidan, M. A., Handwerger, K., Orr, S. P. and Rauch, S. L. (2009) ‘Neurobiological basis of failure to recall extinction memory in posttraumatic stress disorder’, Biological Psychiatry, 66(12), 1075-1082, available: https://doi.org/10.1016/j.biopsych.2009.06.026
Milne, A. M. B., MacQueen, G. M. and Hall, G. B. C. (2012) ‘Abnormal hippocampal activation in patients with extensive history of major depression: An fMRI study’, Journal of Psychiatry and Neuroscience, 37(1), 28-36, available: https://doi.org/10.1503/jpn.110004
Roberts, N. P., Roberts, P. A., Jones, N. and Bisson. J. I. (2016) ‘Psychological therapies for post-traumatic stress disorder and comorbid substance use disorder’, Cochrane Database of Systematic Reviews, 4, available: https://doi.org/10.1002/14651858.CD010204.pub2
Rosen, V., Ortiz, N. F. and Nemeroff, C. B. (2020) ‘Double trouble: Treatment considerations for patients with comorbid PTSD and depression’, Current Treatment Options in Psychiatry, 7, 258274, available: https://doi.org/10.1007/s40501-020-00213-z
Sareen, J. (2014) ‘Posttraumatic stress disorder in adults: Impact, comorbidity, risk factors, and treatment’, The Canadian Journal of Psychiatry, 59(9), 460-467, available:
https://doi.org/10.1177/070674371405900902
Torres, F. (2020 October) What is depression? [online], American Psychiatric Association, available: https://www.psychiatry.org/patients-families/depression/what-is-depression
Torres, F. (2020 August) What is posttraumatic stress disorder? [online], American Psychiatric Association, available: https://www.psychiatry.org/patients-families/ptsd/what-is-ptsd
U.S. Department of Veterans Affairs. (2023, February 3) How common is PTSD in adults? [online], U.S. Department of Veterans Affairs, available: https://www.ptsd.va.gov/understand/common/common_adults.asp#:~:text=About%206%20out%20of%20every,criteria%20for%20PTSD%20after%20treatment.
Weathers, F. W., Blake, D. D., Schnurr, P. P., Kaloupek, D. G., Marx, B. P. and Keane, T. M. (2013) The Clinician-Administered PTSD Scale for DSM-5 (CAPS-5) [assessment], available: www.ptsd.va.gov.
Wright, A. (2020 October 10) Chapter 6: Limbic system: Amygdala, McGovern Medical School at UTHealth, available: https://nba.uth.tmc.edu/neuroscience/m/s4/chapter06.html
Yao, Z., Fu, Z., Wu, J., Zhang, W., Yu, Y., Zhang, Z., Wu, X., Wang, Y., & Hu, B. (2020) ‘Morphological changes in subregions of hippocampus and amygdala in major depressive disorder patients’, Brain Imaging and Behavior, 14, 653-667. https://doi.org/10.1007/s11682-0180003-1