COVID-19 and Stress-Related Disorders

The article by He et al. (14) presents data linking brain function to individual differences in the development of stress-related anxiety responses occurring during the COVID-19 pandemic. Using an initial sample of 589 undergraduate students in which functional MRI and behavioral data were collected prior to the COVID-19 pandemic, the researchers defined a pre-COVID “functional connectome” that later predicted COVID-19–related levels of anxiety. The functional connectome was characterized by using machine-learning techniques on whole-brain resting functional imaging data, and major components of this network comprised regions of the prefrontal, insular, and anterior cingulate cortices, as well as the thalamus, putamen, and hippocampus. Interestingly, while these functional brain measures predicted individual differences in COVID-19 anxiety, they were not effective in predicting pre-COVID levels of anxiety collected immediately after the scanning session. To further validate the findings, the researchers tested the defined functional connectome in additional samples, confirming its utility in predicting pandemic-related anxiety as well as in differentiating patients with generalized anxiety disorders from healthy control subjects. Taken together, these findings shed light onto preexisting neural vulnerabilities that may interact with stressful experiences to result in the development of psychopathology. However, it should be noted that while the defined functional connectome was able to predict stress-related levels of anxiety, the strength of the associations between the brain data and anxiety were relatively small. In her editorial, Dr. Lianne Schmaal from the University of Melbourne (15) discusses the psychosocial determinants of COVID-19–related anxiety and the implications of the neuroimaging findings that predict stress-related, but not baseline, levels of anxiety.

Interactions between comorbid medical and psychiatric illnesses are well known, and in this issue of the Journal, Castro and coauthors present data suggesting an interaction between mood disorders and COVID-19 that affects morbidity and mortality (10). In this report, the researchers examined electronic medical record data from close to 3,000 individuals who were admitted for COVID-19 to hospitals in the Eastern Massachusetts area during the winter and spring of 2020. The findings demonstrate that hospitalized COVID-19 individuals with a history of a mood disorder are at greater risk of death while in the hospital and have more complicated recoveries after discharge. Specifically, the researchers found that a history of mood disorder was associated with an approximately twofold increase in the likelihood of dying after 12 days of hospitalization and a twofold increase in the likelihood of being discharged to a nursing or rehabilitation facility. Of note, the effects of mood disorders on mortality were not found to be significant for the first 12 days of hospitalization. It is also important to note that the associations between mood disorder and morbidity and mortality remained significant when controlling for other associated variables, such as sociodemographic and clinical features. The pathophysiological processes underlying the interaction between mood disorders and COVID-19 outcomes are completely unknown. If replicated, these findings have important clinical implications and underscore the importance of taking psychiatric history into account in the assessment of risk and prognosis in COVID-19 patients.

C-reactive protein (CRP) is a protein made in the liver that is associated with inflammation. Its production is stimulated by interleukin-6, and via its downstream effects, it activates the complement system facilitating macrophage phagocytosis. Although numerous studies have demonstrated that depression is associated with increased blood levels of CRP, these findings can be difficult to interpret because depression is frequently comorbid with other conditions that influence CRP levels, such as obesity and chronic inflammatory states. In this issue of the Journal, Pitharouli et al. (16) use data from the extremely large UK Biobank to explore the genetic and nongenetic factors that could be contributing to the association between depression and CRP levels. In this case-controlled study, the authors compared data from 26,894 individuals who were considered to have a history of major depression with 59,001 control subjects. In a finding consistent with other reports, CRP levels were significantly higher in individuals with a history of depression compared with individuals with no history of psychiatric illness. When controlling for body mass index (BMI), smoking, and alcohol consumption, as well as early-life trauma and socioeconomic status, the strength of the relation between depression and CRP levels was weakened but remained significant. This suggests that while other factors are associated with CRP levels, there is an overall direct linkage between depression and CRP and thus inflammatory processes. In relation to genetics, the authors also found that the polygenic risk score for depression was predictive of CRP levels; however, this relation was no longer significant after taking BMI scores and smoking behavior into account. This implies that the genetic components underlying the risk to develop depression also influence eating and smoking behaviors in ways that in turn influence CRP levels. In her editorial, Dr. Brenda Penninx from Vrije Universiteit Medical Center in Amsterdam further discusses the meaning of these findings, including the significant but relatively weak relation between depression and CRP levels (17). She notes that some individuals in the sample had CRP levels in much higher ranges, supporting the idea that there is likely a subset of patients in which alterations in immune and inflammatory processes may be a more prominent contributor to pathophysiology.

In this article, Shmulewitz and coauthors (18) used data from the National Epidemiologic Survey on Alcohol and Related Conditions to assess changes in alcohol use in the U.S. population from 2001–2002 to 2012–2013. Findings demonstrated significant increases in the levels of high- and very high-risk drinking over this time period. For men, very high-risk drinking was defined as more than 7.1 drinks per day, and for women, very high-risk drinking was considered to be more than 4.3 drinks per day. The overall prevalence of very high-risk drinking in 2012–2013 was found to be 3.5%, which was an overall increase of 0.9% from 2001–2002. It is noteworthy that the largest increase during this period occurred in women, at 1.4%, compared with a 0.5% increase in men. In addition to these findings and not surprisingly, the researchers demonstrated that mental health conditions were associated with moderate and high-risk drinking. For example, in the 2012–2013 cohort, 57.8% of the very high-risk drinkers were reported to have alcohol use disorder (AUD) and 7.1% to have a depressive or anxiety disorder. The associations between high-risk drinking and anxiety and depression were markedly higher in women compared with men. The findings from this study are an important reminder of the high and increasing prevalence of AUD in the United States and of the need to consider psychiatric comorbidities in the assessment and treatment of AUD.

Preclinical and clinical research demonstrates the involvement of reward and negative affect processes in mediating substance use and addiction (19), and the article by King and coauthors in this issue (20) provides insights into the importance of individual differences in the pleasurable effects of alcohol. This study uses longitudinal data from a sample of 184 subjects to understand how subjective and physiological responses to alcohol relate to the development of AUD. Subjects between 21 and 35 years of age who were deemed to be either low- or high-risk drinkers were entered into the study and initially assessed in the laboratory for responses to ingested alcohol and placebo. The laboratory testing was repeated 5 and 10 years later. The researchers found that after 10 years of follow-up, 21% of the participants met criteria for AUD. Notably, these individuals had larger positive subjective responses (wanting and liking) to alcohol administration when initially tested, and over the ensuing 10 years, their positive responses to alcohol increased. In contrast, individuals who did not develop AUD had lower initial positive responses to alcohol, and these responses did not increase over time. The groups did not differ in their physiological responses or in their alcohol breath levels in response to alcohol consumption. These results highlight the importance of early subjective responses to alcohol as an indicator of risk. Furthermore, the results link the development of AUD to sensitization of affective processes reflected by increased pleasure over time associated with alcohol consumption. Drs. Lara Ray and Erica Grodin, along with ReJoyce Green, all from the University of California, Los Angeles, provide an insightful editorial that discusses the interpretation of these findings in relation to the sample’s demographic and clinical characteristics (21). These authors emphasize that while the King et al. findings support reward-related processes in the development of AUD, they do not directly refute mechanisms of addiction that are related to negative affect. Considerable data support repeated addictive drug use as a means to overcome the dysphoric states that occur in individuals with addiction between episodes of drug use (19).

The papers in this issue of the Journal span topics related to COVID-19, anxiety, depression, and alcohol use. Stress, which could not be more relevant to our present times, is a common theme that underpins the symptoms and disorders discussed in this issue. In relation to mood disorders, the Pitharouli et al. article (16) leverages the valuable UK Biobank to provide insights into the multiple factors that underlie the linkages between depression, CRP, and inflammatory processes. While altered inflammatory processes have long been hypothesized to be relevant to depression, COVID-19 is an illness that elicits a robust immune response, and this is one of the likely mechanisms that facilitates the development of COVID-19–related neuropsychiatric symptoms. The interaction between COVID-19 and mental illness is bidirectional. In addition to the effects of COVID-19 on brain function resulting in neuropsychiatric symptoms, psychiatric illnesses themselves appear to influence COVID-19 outcomes. For example, the data presented by Castro et al. (10) demonstrate that mood disorders are associated with increased COVID-19 morbidity and mortality. In a very interesting article relevant to stress and anxiety disorders, He et al. (14) use prepandemic functional brain data to make predictions related to COVID-19–associated anxiety. This study provides early evidence for the possibility of using functional brain measures to help predict individual differences in vulnerability and resilience when individuals must endure chronic stress.

Increased substance use is also commonly associated with stress. Consistent with this, recent reports demonstrate increased alcohol use related to the COVID-19 pandemic (22), with suggestions that this is more likely to occur in individuals with anxiety and depressive symptoms (23). The interactions among stress, psychiatric symptoms, and alcohol use are not surprising because addictive substances typically have the dual short-term reinforcing and “self-treatment” effects of both decreasing dysphoria and increasing positive affect. The article by King et al. in this issue (20) sheds further light onto understanding the risk factors involved in developing AUD, as the findings link increased positive subjective responses to alcohol over time to the development of AUD.

Taken together, the papers in this issue provide valuable insights into understanding mood disorders, anxiety, and AUD. These disorders are all stress related, and their incidence has increased during the pandemic. It is highly likely that this trend will continue and possibly worsen in the aftermath of the pandemic. As a profession, we need to be prepared. In our individual practices, we need to be increasingly aware of the COVID-19–related factors that influence the clinical presentation of our patients and take them into account as we formulate treatment plans. At a population health level, we must implement strategies that will help individuals build resilience to cope with trauma associated with COVID-19 and develop intervention strategies that can reach individuals who do not currently have access to mental health care resources.