ABSTRACT
Objectives Use a longitudinal approach to study the impact of the Covid-19 pandemic on the emergence of symptoms of depression and/or anxiety in college freshmen. Define the interplay between genetic risk and psychosocial factors in shaping vulnerability or resilience to pandemic stress.
Methods University of Michigan freshmen were characterized at baseline using multiple psychological instruments. They were genotyped and polygenic risk score for depression (MDD-PRS) was calculated. Daily physical activity was captured. They were sampled at multiple time points throughout the freshman year on clinical rating scales, including GAD-7 and PHQ-9 for anxiety and depression, respectively. The 2019-2020 cohort (N=122) was compared to an earlier cohort (N=106) to assess the impact of the pandemic.
Results Across cohorts, 25%-57% of freshmen developed significant symptoms of anxiety or depression. In the 2019-2020 cohort, measures of anxiety and depression increased significantly after the onset of COVID-19. Physical activity was dramatically reduced by the pandemic and was associated with the emergence of mood symptoms. Low MDD-PRS subjects exhibited lower relative risk for depression/anxiety during a typical freshman year, but they were more negatively impacted by the pandemic than High MDD-PRS subjects. Conversely, a cluster of psychological indices at baseline predicted resilience in High MDD-PRS subjects who did not develop a mood disorder post-stress.
Conclusions The pandemic had a profound impact on college freshmen triggering depression and anxiety symptoms. Pandemic stress overrode the advantage conferred by “genetic resilience”. By contrast, “psychosocial resilience” was protective even in the face of high genetic risk and pandemic stress.
INTRODUCTION
Few longitudinal studies exist in a young population that define key determinants of stress vulnerability or resilience and capture the psychobiological characteristics of individuals before they transition to depression. Major Depressive Disorder (MDD) is highly genetically complex, with heritability estimates at 30-40% (1). Gene-by-environment interactions play a strong role, as episodes of depression often precede periods of stress (2, 3). Moreover, previous depressive episodes contribute to the likelihood of subsequent depression (4), which makes it more critical to predict and attempt to prevent the emergence of depression in young healthy individuals.
Several studies have described a polygenic risk score for depression (MDD-PRS) (5), and a few have related this score to clinical outcomes (6, 7). In youths, the MDD-PRS predicted depression severity and age of onset (7), consistent with Wray et al. (8). A recent study applied the MDD-PRS to medical interns (9), and pointed to internship stress as playing a key role in mediating the relationship between genetic risk and depression. A longitudinal study that looked at the effects of parental criticism and MDD-PRS on the development of depression in adolescents also illustrated gene-by-environment interactions (10).
This study focuses on college freshmen, as the first year of college is considered to be a stressful life event due to moving away from home, loss of existing peer group and the challenges of more demanding scholastic endeavors. Some individuals appear resilient whereas others appear more vulnerable to developing a mood disorder during their freshman year. This vulnerability is likely influenced by the interplay between genetic risk and sustained social stress (11, 12).
There is emerging evidence that in 2020, college freshmen were profoundly affected by COVID-19. During the beginning of the pandemic in China, when sampled at a single time point, the prevalence of anxiety was 23% and the prevalence of depression was 48% (13), with comparable levels observed in US college students (14). However, these studies did not systematically provide comparisons to previous rates of incidence, nor did they carry out any longitudinal follow-up.
Our Michigan Freshmen Study offers several methodological advantages for parsing the interplay between genetics and environment in vulnerability to depression and anxiety. First, we compare depression and anxiety symptoms during COVID-19 to a pre-pandemic cohort. This has enabled us to describe the relative impact of different levels of environmental stress on the development of mood disorders by contrasting the freshman experience alone versus its combination with pandemic stress. Secondly, we perform a thorough characterization of the subjects at baseline using multiple clinical and psychological rating scales and then use a longitudinal design to follow them throughout the academic year. Finally, we genotype subjects and compute MDD-PRS to analyze its interactions with stress in shaping the emergence of anxiety and depression. By contrast, previously published studies in college students have limited clinical measures and no genetic information (15, 16), or there are no measures pre-COVID (14).
While our primary dependent variables are measures of symptoms of depression and anxiety throughout the freshman year, we also provide our subjects with wearable devices that capture behavioral changes such as physical activity, as decreased activity has been implicated in poorer health outcomes, including depression (17). This combination of measures sheds light on the relative role of genetics and environment in shaping the impact of life stressors, including the pandemic, on a young population.
METHODS
Subjects
Subjects were college freshmen at the University of Michigan recruited via email, Facebook and posted flyers by the beginning of their freshmen year and followed until September of their sophomore year. Because of COVID-19, the study has been split into two parts: Phase 1 consists of the 2015-2019 cohort, and Phase 2 (the COVID-19 year) consists of the 2019-2020 cohort (the last follow-up was extended until December 2020 for this latter group).
The study design during the pre-Covid era included students from multiple ethnic backgrounds, and assessment of several biological variables, e.g., stress blood measures and a laboratory stress test, most of which had to be discontinued during the pandemic. Since the current analysis focuses on the impact of the pandemic and its interaction with genetic risk on mood outcomes, and since MDD-PRS was derived from individuals with European ancestry, it was necessary to confine our analysis to the Caucasian subset of subjects. Thus, Phase 1 included N=106 (59M, 47F) Caucasians and Phase 2 included N=122 (39M, 83F) Caucasians, with the Combined Phases including 228 (98M, 130F) subjects. COVID-19 represents an important variable for Phase 2, as starting in March 2020, all subjects were sent home from college. All subjects provided written informed consent after receiving a complete oral description of the study.
Data and Sample Collection
Phase 1
Subjects aged 18-19 were enrolled by recruitment after matriculation. Starting in August (baseline) they were given the following written questionnaires: NEO Personality Inventory-Revised (NEO PI-R), Risky Family Questionnaire (RFQ), Childhood Trauma Questionnaire (CTQ), General Anxiety Disorder-7 (GAD-7), Spielberger State-Trait Anxiety Inventory, Positive and Negative Suicide Ideation Inventory (PANSI), Physician Health Questionnaire-9 (PHQ-9), and the Morningness-Eveningness Questionnaire, as well as the Perceived Stress Scale-10 (PSS), Multidimensional Scale of Perceived Social Stress (MSPSS), Connor-Davidson Resilience scale (CD-RISC), 5-item Dispositional Positive Emotions subscale (Compassion), Pearlin Mastery scale (Mastery), and the 3-item Revised UCLA Loneliness scale.
At 3-, 6-, 9- and 12-months, the PHQ-9, GAD-7 and PANSI were repeated in Qualtrics. At 4-, 8-, and 12-months, the PSS, MSPSS, CD-RISC, Compassion, Mastery, and Loneliness were repeated by paper and pencil. A highly skilled psychiatric nurse conducted both the SCID-NP and the Family History Method for Research Diagnostic Criteria (FH-RDC) on each subject at the start of the study, and a LIFE interview at the one-year follow-up. All subjects wore Fitbits for sleep and activity tracking.
Phase 2
The design was similar to Phase 1 with the following exceptions: Questionnaires were moved to a RedCap database; starting in May 2020, subjects were asked to complete the repeating questionnaires monthly; and data collection time was extended until December 2020.
For both Phases
the main follow-up measures were GAD-7 for anxiety, and PHQ-9 for depression. The highest follow-up value was used to determine the presence of anxiety or depression during the school year. For classification purposes we used cut-off values of 10 as these were shown to have high sensitivity and specificity for both PHQ-9 and GAD-7 (18, 19).
Sample Collection and Processing
A blood sample was collected for genomic DNA at baseline in EDTA BD vacutainer tubes. During Phase 1 we collected blood samples at baseline and hair or saliva at follow-up time points. During the pandemic, we shifted to collecting salivary samples (Analyses delayed due to pandemic).
Genotyping and Quality Control
For Phase 1, DNA was genotyped on Illumina HumanExome OmniExpress-24 v1.2 or v1.3 chips (N=184). For Phase 2, DNA was genotyped on Illumina Human Infinium CoreExome-24+ v1.3 chips (N=216). The details of processing the results, combining across the two chip types and conducting quality controls are under Supplementary Data (9).
MDD-PRS Calculation
An additive model, using PRSice v.2.3.3, was used for MDD-PRS calculation, with the exception that a secondary analysis was not performed (9). We relied on the most recent MDD GWAS summary statistics including 23andMe (5). The MDD-PRS was then mean-centered and scaled to a standard deviation (s.d.) of 1.
Statistical analyses
All statistical analyses were performed in R v4.0.5. A median split was used to classify subjects as High or Low MDD-PRS. For the analysis of the fraction of subjects who differed on mood during a given year, a Fisher’s Exact test was used. For MDD-PRS, mood or activity data across months, fixed-effects ANOVAs were performed followed by a Wilcoxon test. Pearson correlations were run to assess associations between PRS, mood or other data. When comparing two groups on High and Low MDD-PRS, we used two-tailed Wilcoxon or Student’s t-test.
RESULTS
Incidence of Anxiety and Depression During the Freshmen Year
To determine incidence of anxiety and depression, we used the highest value at any follow-up and applied a cut-off of >10 for both GAD-7 and PHQ-9. Table 1 shows the percent of subjects who developed depression, anxiety, either depression or anxiety, and both depression and anxiety for Phase 1, Phase 2, and the Combined Phases. Phase 2 is shown for all time points collected as well as for only the time points that matched Phase 1. All follow-up time points were significantly different from their respective baseline values. In comparing Phase 2 to Phase 1 for matched time points, Phase 2 did not differ from Phase 1 on anxiety. However, during Phase 2, subjects exhibited significantly higher levels of depression (baseline: OR=0.19, p<0.05; 3-month: OR=0.32, p<0.05; 12-month: OR=0.27, p<0.05). After controlling for sex and PRS, Phase 2 still differed from Phase 1 for depression (p<0.005). Overall, 25-38% met criteria for a mood disorder during the school year when using matched time points, depending on the stress of the year. Anxiety and Depression scores by Phase for each time point is shown in Figures SF1A and SF1B.
Mood and Activity During COVID
As shown in Figure 1A and 1B, anxiety (p=4.6e-8) and depression (p=7.6e-8) were significantly increased post-COVID (March 2020). The average number of steps/day was greatly decreased post-COVID (p=2.11e-13) (Figure 1C). The post-COVID period was also characterized by more perceived stress (p<0.001) and significant changes in loneliness (p=0.001), negative thoughts (p=0.005), positive thoughts (p<0.05), significant other support (p<0.01) and mastery (p<0.05) relative to pre-COVID. Table ST1 shows differences between subjects who met criteria for anxiety or depression compared to those without anxiety or depression that manifested only post-COVID.
A) Anxiety Scores were higher post-COVID compared to pre-COVID. B) Depression Scores were higher post-COVID compared to pre-COVID. C) Average number of steps per day were lower post-COVID compared to pre-COVID. Means and S.E.M.
As shown in Figures 2A and 2B, the students were less active following COVID (March 2020). When we look at anxiety group by time, there was a main effect of month (F(15,838)=47.83, p<0.001), but no effect of group (F(1,94)=1.44, p=0.23), nor an interaction (F(15,838)=0.78, p=0.70). At follow-up, students who met criteria for anxiety did not differ on average steps/day than students without anxiety (p=0.20).
The pandemic led to a dramatic increase in daily activity across all groups. In addition, activity was modulated by the propensity to develop symptoms of either Anxiety or Depression. A) Average number of steps per day were consistently lower over time for individuals who develop Anxiety during the course of the year, compared to those without Anxiety. B) Average number of steps per day for the No Anxiety and Anxiety Groups collapsed across time. C) Average number of steps per day were consistently lower over time for individuals who develop Depression compared to those without Depression. D) Average number of steps per day for the No Depression and Depression Groups collapsed across time. Means and S.E.M.
As shown in Figures 2C and 2D, when we look at depression group by time, there was a nonsignificant trend for the groups to differ on activity (F(1,93)=3.57, p=0.06), and there was a main effect of month (F(15,838)=46.64, p<0.001), but no interaction (F(15,838)=0.81, p=0.67). At follow-up, the students who met criteria for depression had less average steps/day than students without depression (p<0.05).
Effect of Genes and Environment on Mood During COVID-19
We then asked how genetic risk modulated the stress response, and whether genetic risk played a similar role during the pandemic compared to its role during more typical freshman stress. We conducted two types of analyses-a median split and a correlational analysis between MDD-PRS and the emergence of clinical symptoms of depression and anxiety.
In the median split analysis, we contrasted subjects in the top half of the MDD-PRS score (“High MDD-PRS”) to those in the bottom half of the MDD-PRS score (“Low MDD-PRS”). Figure 3A and 3B show the depression scores by month for these two groups for Phase 1 and Phase 2. For Phase 1 (Figure 3A), there was a nonsignificant overall trend for subjects with high MDD-PRS to exhibit higher depression scores (PHQ-9) (F(1,107)=3.74, p=0.06), and a significant main effect of month (F(4,381)=3.51, p<0.01). There was also an interaction effect (F(4,381)=2.45, p<0.05), with significant differences emerging between High and Low MDD-PRS subjects for March (p<0.01), June (p<0.05) and September (p<0.05), with High MDD-PRS individuals consistently showing higher depression scores.
Monthly Anxiety and Depression Ratings in the High MDD-PRS and Low MDD-PRS Groups across study phases. A) Phase 1: Depression scores over time for High MDD-PRS versus Low MDD-PRS groups, showing higher levels in the High MDD-PRS group. B) Phase 2: Depression scores over time for High MDD-PRS and Low MDD-PRS groups, showing no separation between them. C) Phase 1: Anxiety scores over time for High and Low MDD-PRS groups, showing higher levels in the High MDD-PRS group. D) Phase 2: Anxiety scores over time for High and Low MDD-PRS groups, showing no separation between them. Means and S.E.M.
By contrast, during Phase 2 (Figure 3B), there were no significant differences in depression between high and low MDD-PRS groups (F(1,121)=0.0001, p=0.99), nor an interaction (F(10,993)=0.97, p=0.47). The main effect of month was, however, significant (F(10,993)=2.49, p<0.01). Thus, the effect of COVID-19 in Phase 2 was to eliminate the differential impact of genetic predisposition, and the low MDD-PRS group was no longer distinguishable from the high MDD-PRS group.
A similar picture emerged for anxiety responses between the High and Low MDD-PRS groups. Figure 3C and 3D show anxiety scores over months for Phase 1 and Phase 2, respectively. For Phase 1, there was a nonsignificant trend for PRS (F(1,115)=3.36, p=0.07), and a significant main effect of month (F(4,382)=3.97, p<0.005). Although the interaction was not significant (F(4,382)=1.91, p<0.11), Wilcoxon tests showed significant differences between High and Low MDD-PRS for March (p<0.01), June (p<0.05) and September of sophomore year (p<0.01), with High MDD-PRS individuals consistently exhibiting higher anxiety scores. For Phase 2, there was only a significant main effect of month (F(10,989)=3.12, p<0.001). Neither the main effect of group (F(1,121)=0.54, p=0.46), nor the interaction was significant (F(10,989)=0.78, p=0.64).
MDD-PRS and Stress Vulnerability and Resilience
In addition to the median split analyses, we correlated MDD-PRS separately to anxiety and depression for Phase 1, Phase 2, and the Combined Years.
Correlation of MDD-PRS with Depression and Relative Risk in Phase 1 vs. Phase 2
The MDD-PRS correlation with depression for Phase 1 (Figure 4A) and Phase 2 (Figure 4B) revealed an interesting pattern. There was a significant positive correlation between MDD-PRS and follow-up depression in Phase 1 (r=0.27, p<0.05). However, this was no longer the case for Phase 2 (r=-0.01, p=0.91). There was a nonsignificant trend for a correlation between MDD-PRS and follow-up depression when using both Phase 1 and Phase 2 combined (r=0.22, p=0.07). This is due to the loss of the relationship in Phase 2 under COVID stress.
A and B: For Depression, there was a significant correlation between MDD-PRS and follow-up depression during Phase 1 (A), but this correlation no longer existed during Phase 2 (B). C and D: For Anxiety, there was a significant correlation between MDD-PRS and follow-up anxiety during Phase 1 (C), but this correlation no longer existed during Phase 2 (D). Means and S.E.M.
We also compared the relative risk ratio for developing depression in subjects who had greater than +0.5 s.d. PRS-MDD versus those who had less than -0.5 s.d. PRS-MDD. During Phase 1, this relative risk was 2.7 (p=0.11), but this decreased to 0.9 during Phase 2. In Phase 1, 90% in the Low MDD-PRS group were protected against expressing depression symptoms under stress, compared to 74% in the High MDD-PRS group. In Phase 2, 74% in the Low MDD-PRS group were protected against expressing depression symptoms under stress, compared to 76% in the High MDD-PRS group, suggestive of loss of genetic protection in the Low MDD-PRS group.
Correlation of MDD-PRS with Anxiety and Relative Risk in Phase 1 vs. Phase 2
Figure 4C shows that there was a significant positive correlation between MDD-PRS and follow-up anxiety in Phase 1 (r=0.26, p<0.01). However, this was no longer the case during Phase 2 (r=0.05, p=0.58) (Figure 4D). There was still a significant positive correlation between MDD-PRS and follow-up anxiety for Combined Years (r=0.14, p<0.05).
For Phase 1, the relative risk ratio for anxiety at > +0.5 s.d. compared to <-0.5 PRS was 5 (p<0.05). For Phase 2, the relative risk was no longer significant at 0.9. In Phase 1, 94% in the Low PRS group were protected against severe anxiety symptoms compared to 68% in the High MDD-PRS group. This relationship was lost in the Phase 2 and Combined data. Thus, genetic predisposition significantly correlated with susceptibility to anxiety symptoms during Phase 1, but this correlation no longer existed during Phase 2 or in the Combined dataset.
Interestingly, the chronicity of mood symptoms also varied with genetic risk between Phases, as shown in Figure SF2. When analyzing baseline measures in the Combined Years dataset, resiliency and having a risky family were differentially related to MDD-PRS (Figures SF3A and SF3B).
Non-Genetic Resilience
It is notable that there are individuals who have high genetic risk and yet appear resilient to stress. Correlates of this “non-genetic resilience” were analyzed for subjects who never meet criteria for anxiety (Table ST2) or depression (Table ST3). First, baseline PHQ-9 and GAD-7 scores (even below the 10 cut-off) were highly predictive of future propensity to express significant levels of anxiety or depression symptoms. Other protective factors relate to: a) Family history: Low scores on the RFQ predicted lower vulnerability to anxiety and depression, and a low level of childhood trauma (CTQ) predicted lower vulnerability to depression; b) Stable psychological characteristics: Low Neuroticism scores, higher sense of Mastery, lower Loneliness all predicted lower anxiety and depression, and higher Resiliency scores on CD-RISC predicted lower depression; and c) State-related measures that may modify immediate reaction to stressors: More positive thoughts (PANSI+) protected against both anxiety and depression, whereas less perceived stress (PSS) protected against the emergence of anxiety symptoms.
DISCUSSION
This longitudinal study characterized the impact of the Covid-19 pandemic on depression and anxiety states in a cohort of young, healthy subjects facing the stress of entering college and compared them to a previous cohort facing college before the pandemic. Not only did our results highlight the strong impact of the pandemic on this population, but they provided a novel perspective on the interplay between genetic risk and levels of environmental stress in shaping stress reactivity. In particular, our analyses led us to characterize two categories of stress resilience--“genetic resilience” as indexed by a low biological risk for mood disorders (Low MDD-PRS), and “psychosocial resilience” which becomes evident in subjects with high genetic risk (High MDD-PRS) who successfully cope with significant life stressors.
Our findings support the premise that the freshman year of college represents a significant psychosocial stressor, with between 25%-57% of freshmen developing symptoms that meet criteria for either anxiety or depression, depending on the year and the sampling frequency. The highest levels of negative mood symptoms always took place towards the end of the freshman year.
The pandemic represented a significantly greater stressor, as we observed the highest levels of negative affect in the 2019-2020 cohort of freshmen including measures of depression, anxiety, stress, and loneliness. But perhaps the most drastic pandemic-induced change was the large decrease in physical activity as monitored through wearable devices. This reduction in physical activity was more pronounced in individuals who exhibited symptoms depression during the course of the pandemic year. A recent study found that a reduction in self-reported physical activity during COVID-19 was more apparent in individuals who had been more active prior to the pandemic (17), and that, in turn, decreased activity impacted mood.
We then asked how the response to stress was modulated by genetic risk for depression. During typical freshman years, both anxiety and depression symptoms were positively correlated with MDD-PRS, and High MDD-PRS predicted greater chronicity of affective dysregulation. However, under the pandemic, modulation by genetic risk all but disappeared. Indeed, it was the Low MDD-PRS group that exhibited the greater shift during COVID-19, showing more incidence of anxiety and depression symptoms. The pandemic increased not only the incidence but also the chronicity of affective symptoms in individuals with low genetic risk. Thus, under more extreme stress low genetic risk failed to act as a buffer against the emergence or persistence of a mood disorder, underscoring the limits of “genetic resilience”.
An interesting counterpoint is the existence of non-genetic form of resilience; High MDD-PRS subjects who did not develop anxiety or depression symptoms including during the pandemic. We identified a cluster of factors at baseline associated with this psychosocial resilience (Low PHQ-9, Low GAD-7), including less risky early environment (RFQ), a range of stable psychological traits (lower neuroticism, lower loneliness, greater sense of mastery) and a more positive affective style (more positive thoughts). It is conceivable that genetically vulnerable subjects had to develop coping strategies earlier and more actively than their more genetically resilient counterparts that served them well under the greater stress of the pandemic.
One benefit of our study is its longitudinal nature. The freshman year is not only inherently stressful, but also exhibits an accumulating pattern of distress, with the end of the school year exhibiting the highest levels of anxiety and depression. During Phase 1, the summer break provided a “reset” and students return to their sophomore year with decreased levels of negative affect. Remarkably, in the 2019-2020 cohort, the “summer reset” did not occur, raising the possibility of cumulative and long-lasting negative consequences on mood for this cohort
Other social factors may also be important. Social isolation leads to an increase in depression (20, 21), and this may be especially true in the 10-24 age bracket when the need for peer interaction is especially important (22). Identifying ways to decrease loneliness and improve mastery, resiliency and social support may be important for stress buffering in this age group. (23).
In summary, our work highlights the profound impact of the pandemic on a young healthy population, with the emergence of symptoms of clinical anxiety and depression coupled with a dramatic decrease in physical activity. It sheds light on the interplay between genetic risk and environmental stressors in shaping affective responses, by identifying the existence of “genetic resilience” that appears protective under moderate stress, and of “psychosocial resilience” that may become especially relevant under more extreme stress. This lays the groundwork for arriving at predictive algorithms that can define individual risk for developing mood or anxiety disorders before they become significant, and for devising more personalized interventions to enhance stress resilience.
Data Availability
Data produced in the present study will be made available following publication in a peer-reviewed journal upon reasonable request to the authors.
Financial Considerations
This work was supported by the Pritzker Neuropsychiatric Disorders Research Fund L.L.C. The authors are members of the Pritzker Neuropsychiatric Disorders Research Consortium, which is supported by the Pritzker Neuropsychiatric Disorders Research Fund L.L.C. A shared intellectual property agreement exists between this philanthropic fund and the University of Michigan, Stanford University, the Weill Medical College of Cornell University, the University of California at Irvine, and the HudsonAlpha Institute for Biotechnology to encourage the development of appropriate findings for research and clinical applications. The authors declare no biomedical financial interests and no conflicts of interest.
Acknowledgements
This study was supported by the Office of Naval Research (ONR) Grants N00014-09-1-0598, N00014-12-1-0366 and N00014-19-1-2149, the Hope for Depression Research Foundation, and the Pritzker Neuropsychiatric Disorders Research Consortium Fund LLC (http://www.pritzkerneuropsych.org). We would like to thank the Michigan Institute for Clinical and Health Research for help with this study (UL1TR002240).
Footnotes
Disclosures: Drs. Turner, Hagenauer, Lopez, Sen, Watson & Akil are members of the Pritzker Neuropsychiatric Disorders Research Consortium which is supported by the Pritzker Neuropsychiatric Disorders Research Fund L.L.C. A shared intellectual property agreement exists between this philanthropic fund and the University of Michigan, Stanford University, the Weill Medical College of Cornell University, the University of California at Irvine, and the HudsonAlpha Institute for Biotechnology to encourage the development of appropriate findings for research and clinical applications.
