The study was led by Marco Angrisani, senior economist at the USC Center for Economic and Social Research and associate professor (research) of economics at the USC Dornsife College of Letters, Arts and Sciences. Angrisani and his colleagues set out to answer a question that has troubled mental health researchers for decades: why do responses to large-scale societal stressors vary so dramatically from one person to the next? Their answer, drawn from one of the most extensive longitudinal datasets available, points to personality as a meaningful and measurable factor in shaping that heterogeneity. “There is a wide variety of mental health responses to large-scale societal stressors,” Angrisani said. “Our study shows that personality traits, and neuroticism in particular, play a role in shaping heterogeneous responses in a major crisis across individuals and could therefore inform better targeted interventions.”

The evidence base for these conclusions is unusually robust. The researchers followed 4,030 U.S. adults using data from the Understanding America Study, a nationally representative internet panel managed by the USC Center for Economic and Social Research. The analysis drew on more than 100,000 survey responses collected from 2019 through early 2024, a window that allowed the team to compare mental health before, during, and after the pandemic. That temporal span matters enormously. Cross-sectional studies conducted at a single point in time can show that distress was high during the pandemic, but they cannot reveal whether it rose, fell, or lingered afterward. By anchoring their measurements in pre-pandemic baselines and extending them well into the recovery period, the researchers could distinguish acute, transient distress from the kind of prolonged suffering that signals genuine vulnerability.

At the heart of the analysis were the so-called Big Five personality traits, the dominant framework in contemporary personality psychology. Neuroticism describes a tendency to experience worry, emotional sensitivity, and stronger reactions to stress or uncertainty. Extraversion captures the tendency to be outgoing, energetic, and motivated by social interaction. Openness reflects curiosity about new ideas and experiences and a willingness to consider change. Agreeableness denotes a tendency to be cooperative, trusting, and concerned about social harmony, while conscientiousness describes people who are organized, dependable, self-disciplined, and goal-oriented. Crucially, these characteristics exist on a continuum. Everyone possesses some degree of each trait, and a high or low score on any dimension does not constitute a disorder. The Big Five model is valued by researchers precisely because it reduces the enormous complexity of human temperament to five stable, measurable dimensions that can be tracked over time and correlated with real-world outcomes.

Mental distress in the study was measured with the PHQ-4, a brief validated questionnaire that assesses symptoms of anxiety and depression. A score of six or higher on the PHQ-4 indicates moderate-to-severe symptoms, although the instrument is a screening tool rather than a clinical diagnosis. Using this threshold, the researchers quantified how the probability of moderate or severe distress shifted across the pandemic for people at different points on the personality spectrum. The results were striking. Increases in the probability of moderate or severe distress were approximately 11 to 20 percentage points greater among participants who scored in the highest 25 percent for neuroticism compared with those in the lowest quarter. In practical terms, this means that highly neurotic individuals were dramatically more likely to cross the threshold into clinically significant distress as the crisis unfolded.

Just as important as the size of the effect was its persistence. The association between neuroticism and elevated distress held across young, middle-aged, and older adults, and it remained evident into the post-pandemic observation period. This durability is what distinguishes the finding from earlier work documenting short-lived spikes in anxiety during lockdowns. The pandemic has largely receded as an acute threat, yet the psychological shadow it cast appears to have lingered far longer for people high in neuroticism than for their more emotionally stable counterparts. For health systems still grappling with elevated demand for mental health services, that pattern suggests the pandemic’s toll was not merely an episode but, for a vulnerable segment of the population, a lasting shift.

The study also uncovered behavioral differences that may help explain the mechanism. Participants higher in neuroticism perceived greater health and financial risks during the crisis, adopted more protective behaviors, and reported exercising and relaxing less frequently while consuming more alcohol. These patterns sketch a coherent picture: people who react strongly to uncertainty tend to perceive danger more acutely, respond with heightened caution, and simultaneously withdraw from the very activities, such as exercise, relaxation, and moderate drinking habits, that buffer against psychological distress. Measurements of extraversion, openness, and agreeableness showed weaker or inconsistent associations with distress trajectories, while conscientiousness appeared modestly protective, particularly among adults aged 60 and older. That last finding hints that the self-discipline and structure characteristic of conscientious individuals may function as a psychological resource in older adults facing prolonged disruption.

The practical implications extend well beyond pandemic planning. The findings could help clinicians and emergency planners recognize who may need closer follow-up after disasters, public-health emergencies, or other prolonged disruptions. Public information campaigns could also be tested to ensure they inform people without unnecessarily increasing distress among those who are more vulnerable to it. Angrisani was careful to emphasize, however, that personality should be just one part of a broader risk assessment, considered alongside current symptoms, previous mental health conditions, and social circumstances, rather than serving as a label or an automatic trigger for treatment. Neuroticism is a risk factor, not a diagnosis, and it does not determine any individual’s mental health destiny. In the paper, the authors advocate for monitoring and stepped care: following people over time and offering treatments such as cognitive behavioral therapy or emotion-regulation training when significant, persistent symptoms emerge, rather than treating people solely because of a personality score.

Like all observational research, the study has limits that the authors acknowledge. It identifies associations and cannot prove that personality, especially neuroticism, caused the distress patterns observed. All personality and mental distress measures were self-reported, which introduces the possibility of reporting bias, and the findings may not generalize beyond the United States. Even so, the scale and duration of the dataset give the results considerable weight, and they point toward a concrete strategy for the future: crisis communication that emphasizes practical, controllable actions, combined with accessible mental health support, could help protect psychologically vulnerable groups when the next major disruption arrives. As climate events, geopolitical instability, and future pandemics test societies around the world, the USC-led study offers a deceptively simple insight with far-reaching consequences. The question of who will suffer most in a crisis may be partly answerable in advance, and answering it could mean the difference between reactive care and genuinely preventive support for the millions of people whose minds bear the longest scars of collective trauma.

Subject of Research: The influence of personality traits, particularly neuroticism, on the persistence of mental distress during and after the COVID-19 pandemic

Article Title: Personality traits influence level of lasting distress during a crisis, USC-led study finds

Article References: Personality traits influence level of lasting distress during a crisis, USC-led study finds. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: neuroticism, personality traits, Big Five, mental distress, COVID-19 pandemic, anxiety, depression, PHQ-4, Understanding America Study, stepped care, observational study, USC

The answer to the first question is a resounding yes, and the answer to the second is more surprising than anyone might have predicted. Among the 212 Mexican clinical psychologists surveyed, two factors stood out as significantly associated with lower levels of both anxiety and depression: being male, and having greater confidence in one’s own clinical skills. Just as striking were the factors that showed no significant relationship at all. Age, a personal history of psychotherapy, and whether the psychologist received clinical supervision all failed to predict symptom levels, challenging assumptions drawn from earlier studies across multiple countries.

The PHQ-4 itself is a marvel of compression. It consists of only four questions, two drawn from the PHQ-2 depression screener and two from the GAD-2 anxiety screener, each rated on a four-point scale of how often the respondent has experienced the symptom over the past two weeks. The depression items probe persistent low mood and loss of interest or pleasure, while the anxiety items ask about feeling nervous, anxious, or on edge, and about being unable to stop or control worrying. Because it takes less than a minute to complete, the instrument has become a workhorse of epidemiological research and primary care screening worldwide. Yet, until now, its psychometric performance specifically in clinical psychologists, a population under chronic emotional strain, had never been rigorously examined.

To test the instrument, the researchers turned to confirmatory factor analysis, a statistical technique that evaluates whether the pattern of responses matches a hypothesized underlying structure. Because the four items are rated on ordinal scales rather than continuous measurements, the team used the WLSMV estimator, which is specifically designed for categorical data. They then compared two competing models. The first treated all four items as measuring a single dimension of general psychological distress, an approach adopted by several previous studies in other populations. The second preserved the original design, separating anxiety items from depression items as two correlated but distinct factors.

The results were decisive. The unidimensional model failed badly, with a robust root-mean-square error of approximation of 0.49, far above the accepted threshold of 0.06, and a robust Tucker-Lewis index of just 0.45. The two-factor model, by contrast, fit the data almost perfectly: the comparative fit index and its robust version both reached 1.00, the robust RMSEA was 0.05, and the standardized root-mean-square residual was a mere 0.01. Every factor loading exceeded 0.80, indicating that each item strongly reflected its intended construct. Reliability, estimated with the categorical omega coefficient appropriate for ordinal items, came in at 0.78 for the anxiety subscale and 0.82 for the depression subscale, both comfortably within the range considered adequate for research and screening purposes.

One nuance deserves attention. The correlation between the latent anxiety and depression factors was high, at 0.73, confirming what clinicians have long observed: these two conditions travel together. This overlap explains why some research teams in other countries, including studies of healthcare workers in Taiwan and HIV care providers in China, have collapsed the PHQ-4 into a single distress score. The Mexican team argues, however, that maintaining the two-dimensional structure captures important nuances that a single score would erase, and their fit statistics give that argument considerable weight. The finding suggests that while anxiety and depression are deeply intertwined in this population, they remain distinguishable, and screening tools should treat them as such.

The second half of the study embedded the PHQ-4 within a full structural equation model, treating sex, age, psychotherapy history, clinical supervision, and confidence in clinical skills as potential predictors of the two latent symptom dimensions. This model also fit the data exceptionally well. Two paths emerged as statistically significant. Male psychologists reported lower levels of both anxiety and depression, a result consistent with broader epidemiological patterns in Mexico and beyond, where women consistently report more emotional distress. The authors caution that this gender gap must be interpreted carefully. More than 75 percent of Mexican clinical psychologists are women, and societal expectations may lead men to underreport emotional difficulties, while women in the profession may face compounded stressors from balancing professional and personal responsibilities.

The second significant predictor is arguably the study’s most actionable finding. Psychologists who expressed greater confidence in their own clinical skills reported significantly fewer symptoms of anxiety and depression. This echoes earlier research linking professional self-efficacy to mental health, including studies of psychology students in the United Arab Emirates and psychotherapists in Sweden, where low self-perceived competence predicted burnout. The mechanism is intuitive: a therapist who feels effective in daily work experiences a sense of mastery that buffers against the emotional toll of the job. Conversely, doubting one’s clinical abilities may amplify the rumination and worry that define anxiety, and the hopelessness that defines depression. Notably, receiving clinical supervision and having undergone personal psychotherapy, two interventions often assumed to protect therapists, showed no significant association with symptom levels in this sample, adding to a genuinely mixed international literature on these practices.

The implications reach well beyond Mexico. Clinical psychologists form the backbone of mental health care systems worldwide, and the COVID-19 pandemic exposed their vulnerability in stark terms. Studies cited in the paper found that 17 percent of Belgian psychologists experienced depression and 12 percent anxiety during the crisis, while 70 percent of Brazilian psychologists reported changes in their mental health. If confidence in clinical skills functions as a protective factor, then training programs, continuing education, and high-quality supervision designed to build genuine competence, rather than merely being offered supervision, could become explicit targets of mental health policy for the professionals themselves. The validated PHQ-4 offers an efficient way to monitor distress in this workforce, requiring almost no time burden while distinguishing anxiety from depression.

The authors are candid about the study’s limits. The sample was recruited through convenience sampling on social networks, was overwhelmingly female, and relied on self-selection, meaning psychologists experiencing distress may have been more motivated to respond. The cross-sectional design prevents any claim that confidence causes lower distress rather than the reverse. Still, the study marks a rare empirical examination of psychopathology measurement in clinical psychologists themselves, a population more often studied as assessors than as assessed. Future work, the team suggests, should test the PHQ-4 across cultures, explore bifactor models that separate general distress from specific symptoms, and follow professionals over time to untangle the direction of the relationship between self-efficacy and mental health. For now, the message is clear and quietly urgent: the healers need measuring too, and the four shortest questions in the field may be exactly the right ones to ask.

Subject of Research: Psychometric validation of the PHQ-4 anxiety and depression screening tool and associated risk and protective factors among clinical psychologists in Mexico

Article Title: Anxiety, Depression, and Associated Factors in Clinical Psychologists: Psychometric Evaluation of the PHQ-4

Article References: Avalos-Salcedo, L. J., González-Hernández, A. B., Blancas-Guillen, J., Ccoyllo-Gonzalez, L., Varela, S., & Valencia, P. D. (2025). Anxiety, Depression, and Associated Factors in Clinical Psychologists: Psychometric Evaluation of the PHQ-4. Trends in Psychology. https://doi.org/10.1007/s43076-025-00499-z

Image Credits: AI Generated

DOI: 10.1007/s43076-025-00499-z

Keywords: clinical psychologists, PHQ-4, anxiety, depression, psychometrics, confirmatory factor analysis, self-efficacy, mental health screening, Mexico, clinical supervision, psychological distress, structural equation modeling