Changes in depressive symptoms during the COVID-19 pandemic differ by personality type: Findings from The Irish Longitudinal Study on Ageing

Background

There is an established body of research showing that many health outomes vary by personality type (Friedman & Kern, 2014; Fry & Debats, 2009; Luchetti et al., 2016). Much of this research on personality and health operationalises personality using the the ‘Big 5’ model that proposes five disctinct personality types: openness, conscientiousness, extraversion, agreeableness, and neuroticism. In terms of the health outcomes studied, research on depression among older adults specifically, found that both the presence of a depression diagnosis and severity of depressive symptomology were associated with higher neuroticism, and lower extraversion and conscientiousness (Koorevaar et al., 2013). A review by Klein et al. (2011) showed that depression was associated with neuroticism/negative emotionality, extraversion/positive emotionality, and conscientiousness. Importantly, their review of different models that have been proposed to explain the association between personality and depression also suggests that personality type contributes to both the onset and course of depression.

The COVID-19 pandemic provides a unique context within which to re-examine the association between personality and depression. The Irish Longitudinal Study on Ageing has shown that the prevalence of clinically-significant depressive symptoms increased threefold among older adults during the pandemic meaning there was a stark increase in a very short period of time that cannot be explained solely by typical risk factors such as the loss of a spouse or functional decline (Briggs et al., 2018; Ward et al., 2023). Much of the increase observed during the pandemic was due to loneliness brought on by public health strategies employed to limit the spread of the SARS-CoV-2 virus that profoundly impacted many aspects of daily life and at times severely restricted social contacts (Ward et al., 2023). These restrictions were keenly felt by older adults whom, along with other at-risk groups, were required to stay at home for long periods of time to control case numbers and relieve pressure on health services. This sudden and dramatic impact of the pandemic on people's everyday lives constituted a stressful event that led to increased psychological distress, poorer self‐reported mental health, and increased depression and loneliness. (Armitage & Nellums, 2020; Bailey et al., 2021; Ettman et al., 2020; Green et al., 2021; Groarke et al., 2021; Hwang et al., 2020; Krendl & Perry, 2021; Pierce et al., 2020; Ward et al., 2021; Wu, 2020). In this context, the pandemic provides a unique context within which to re-examine the role of personality type in accounting for variation in the progression of depressive symptomology.

In response to the dramatic changes experienced during the pandemic, a number of recent studies have examined associations between personality and mental health during the pandemic. These found that higher neuroticism was associated with worse mental health and more negative appraisals of the pandemic (Kroencke et al., 2020; Li et al., 2022; Liu et al., 2021; Modersitzki et al., 2021; Mourelatos, 2023; Shokrkon & Nicoladis, 2021; Wang et al., 2023). However, findings regarding the relationship between extraversion and depressive symptoms have been mixed with some studies providing evidence of a positive relationship, (Liu et al., 2021; Modersitzki et al., 2021) and some a negative relationship between extraversion and depressive symptoms (Shokrkon & Nicoladis, 2021). Finally, higher openness has tended to be associated with less negative appraisals of the pandemic. (Modersitzki et al., 2021). Specific to an older population, findings from a study examining the relationships between personality types and depression among adults aged 60 and older living in China showed concienciousness, extraversion, and agreeableness were negatively related to depression, while neuroticism was found to be positively related to depression.

Numerous other factors have also been identifed that explain differences between invdividuals and groups accumulation of mental health difficulties during this period. Among these factors were changes to physical activity behaviours (Cindrich et al., 2021; Meyer et al., 2020), employment (McDowell et al., 2021a), and loneliness (McDowell et al., 2021b), each of which were associated with an increase in depressive symptoms during the pandemic. While many of these factors can focus on physical health and behaviours, and sociality, intrinsic individual-level characteristics or traits, including personality, may also add to our understanding of the variation observed in older adults’ psychological responses to the pandemic.

Many of the above studies conducted during the pandemic share a number of important limitations. For instance, participants were mostly first recruited during the pandemic and it was therefore not possible to account for levels of depressive sympromology prior to the pandemic (Krendl & Perry, 2021). Many of these studies also included convenience samples which substantially increases risk of bias and limit the generalisability of the findings (Pierce et al., 2020). To address these limitations, we use data from a large cohort of older adults first recruited in 2009 with depressive symptomology measured on multiple occasions before and once during the pandemic. Our hypothesis is that changes in depressive symptomology during the COVID-19 pandemic varied by personality type. This study aim was to investigate the Big 5 personality traits as potential sources of heterogeneity in changes in depressive symptoms associated with the COVID-19 pandemic among older adults while controlling for prior depressive symptoms and other covariates.

Methods

This study was done in adherence with STROBE guidelines (Elm et al., 2007). Ethical approval for the wider TILDA study was granted by the Faculty of Health Sciences Research Ethics Committee in Trinity College Dublin. TILDA adheres to the 1964 Helsinki declaration and its later amendments. Ethical approval specific to the COVID-19 study was obtained from the Irish National Research Ethics Committee COVID-19 on 17th June 2020, Application number: 20-NREC-COV-030-2.

Consent to participate: Explicit informed consent was obtained from all participants in each Wave. In all Waves, participants were provided with detailed information regarding the purpose of the study, their rights as participants, and the measures taken to ensure confidentiality and data protection in advance of participation. Participation was voluntary, and participants were free to withdraw from the study at any point. TILDA adheres to the 1964 Helsinki declaration and its later amendments. In Waves 2–5, written consent was obtained. In Wave 6, consent was obtained verbally via telephone due to COVID-19 restrictions and recorded by the interviewer. TILDA adheres to the 1964 Helsinki declaration and its later amendments.

Sample

Participants were included in our analyses if they: participated in and were aged 50 years or older during the COVID sub-study; completed the 60-item NEO-Five Factor Inventory at Wave; and completed the eight item Centre for Epidemiological Studies Depression Scale (CESD-8) at Waves 2, 3, 4, 5, and the COVID Wave. Participants with missing covariate data (n=38; 1.1%) were subsequently excluded, resulting in an analytic sample of 3,370 (Figure 1).

Figure 1. Flowchart showing how the final analytic sample was arrived at.

Results

Participant characteristics

The flowchart presented in Figure 1 shows how the final analytic sample was arrived. Participants who did not take part in the COVID-19 sub-study, did not complete the NEO Five Factor Inventory at Wave 2, did not have at least one CES-D8 measurement, and who were aged under 50 at the time of the COVID-19 sub-study were not included.

Participant characteristics at baseline are presented in Table 1. Analyses included 14,135 observations from 3,404 participants. The average age was 62.9 years and 56.7% were female. The mean personality scores for Neuroticism were 18.0±7.5; Extraversion: 28.6±5.8; Openness: 28.3±6.0; Agreeableness: 34.0±5.0; and Conscientiousness: 33.6±5.3. (Table 1)

Mean CESD-8 depression scores at each wave are presented in Figure 2 and were significantly higher in the COVID Wave compared to the previous three waves (d=0.65, 95%CI: 0.61–0.69). Internal consistency (α) calculated for the current sample for the each of the five factors was 0.81 (Neuroticism), 0.72 (Extraversion), 0.78 (Openness), 0.80 (Agreeableness), and 0.77 (Conscientiousness). (Figure 2)

Figure 2. Mean depressive symptoms and 95% confidence intervals measured using the eight item Center of Epidemiological Studies Depression Scale (CESD-8; range: 0–24) at Wave 3 (n=3,386), Wave 4 (n=3,310), Wave 5 (n=3,199), and the COVID Wave (n=3,076).

Results from the baseline model presented in Table 2 show that, in the COVID Wave, depressive symptoms were 0.32 points higher per 1-SD increase in neuroticism (b=0.32, 95%CI: 0.20–0.44; p<0.001) and 0.08 points lower per 1-SD increase in openness (b=-0.08, 95%CI: -0.27–-0.05; p=0.004) than would have been expected from the trends observed before the pandemic. Depressive symptoms did not differ by extraversion (p=0.199), agreeableness (p=0.368) or conscientiousness (p=0.168) in the COVID Wave from the trends observed before the pandemic (Table 2, Panel A). There was no three-way interaction between the COVID-19 period indicator, neuroticism and openness (p=0.128). (Table 2)

Table 2. Association between personality traits at Wave 2 and depression symptoms (Wave 2 to COVID Wave).

In Model 1 (Panel A), association between depressive symptoms during COVID-19 and the Big Five personality factors (presented below) were fitted along with the time variable and its squared term. In Model 2 (Panel B),Associations between depressive symptoms during COVID-19 and all covariates at Wave 2 (age, sex, education, smoking, alcohol consumption, physical activity, physical limitations, cardiometabolic conditions and other chronic conditions) were added to Model 1.

N=3,404	Model 1		Model 2
	B (95% CI)	P value	B (95% CI)	P value
Neuroticism	0.32 (0.20, 0.44)	<0.001	0.29 (0.16, 0.42)	<0.001
Extraversion	0.08 (-0.04, 0.20)	0.199	0.12 (0.00, 0.24)	0.045
Openness	-0.08 (-0.27, -0.05)	0.004	-0.14 (-0.25, -0.03)	0.014
Agreeableness	-0.05 (-0.16, 0.06)	0.368	-0.10 (-0.22, 0.02)	0.102
Conscientiousness	-0.09 (-0.21, 0.04)	0.168	-0.08 (-0.21, 0.02)	0.187
Age			0.04 (0.03, 0.06)	<0.001
Sex
Female			0.17 (-0.07, 0.41)	0.164
Education
Secondary			0.09 (-0.23, 0.40)	0.584
Tertiary			0.01 (-0.32, 0.34)	0.957
Smoking
Past			-0.04 (-0.26, 0.19)	0.754
Current			0.61 (0.26, 0.97)	0.001
Problem alcohol			-0.16 (-0.47, 0.16)	0.322
Antidepressant use			0.29 (-0.16, 0.73)	0.207
Physically active			-0.01 (-0.23, 0.21)	0.944
Physical limitations			-0.00 (-0.07, 0.06)	0.948
Cardiometabolic conditions
1–2			-0.13 (-0.47, 0.22)	0.467
≥3			2.67 (0.93, 4.42)	0.003
Other chronic conditions
1			0.15 (-0.09, 0.38)	0.219
≥2			0.86 (0.17, 1.54)	0.015
T2	-0.02 (-0.04, -0.01)	<0.001	-0.02 (-0.04, -0.01)	<0.001
Constant	2.87 (2.60, 3.13)	<0.001	2.79 (2.01, 3.56)	<0.001

Figure 3 shows the unstandardised regression coefficients for the change in CESD-8 depression scores symptoms in the COVID Wave, per 1-SD increase in each personality trait. The full results for Model 2 which included all of the covariates are presented in Table 2 (Panel B). In the COVID Wave, depressive symptoms were 0.29 points higher (b=0.29. 95%CI: 0.29–0.42; p<0.001) per 1-SD increase in neuroticism, 0.12 points higher (b=0.12. 95%CI: 0.00–0.24; p=0.045) per 1-SD increase in extraversion, and 0.14 points lower (b=-0.14, 95%CI: -0.25–-0.03; p=0.014) per 1-SD increase in openness than would have been expected from the trends observed before the pandemic. Depressive symptoms did not differ by agreeableness (p=0.145) or conscientiousness (p=0.752) in the COVID Wave from the trends observed before the pandemic. There were no three- or four-way interactions between the COVID-19 period indicator, neuroticism, extraversion, and openness (all p=0.213). (Figure 3)

Figure 3. Results from linear mixed models (unstandardised regression coefficients and associated 95% confidence intervals) illustrating change in depressive symptoms in the COVID Wave, per 1-SD increase in each personality trait, compared to what would have been expected from the trends observed before the pandemic.

*p=0.047; **p=0.011; ***p<0.001.

Discussion

Taking advantage of the unique context of the COVID-19 pandemic, this study provides further prospective evidence of the association between personality and both the onset and course of depressive symptoms among older adults. In this cohort of 3,404 community-dwelling older adults in Ireland, depressive symptoms were significantly higher during COVID-19 compared to what would have been expected from the trends observed among these same individuals in the ten years prior to the pandemic. Importantly, the magnitude of this COVID-19-related increase in depressive symptoms varied according to personality type. Higher levels of neuroticism and extraversion and lower levels of openness were associated with greater increases in depressive symptoms. These findings further our understanding of the association of personality with in mental health and have important implications for identifying at-risk individuals and subsequently tailoring prevention and treatment programs.

Neuroticism, the degree to which a person experiences the world as threatening and beyond his/her control, is the personality trait that has shown strongest associations with mental health outcomes in prior research. In the current study, it was also associated with the greatest increase in depressive symptoms, supporting previous evidence of its association with greater stress in response to both the threat of disease and also social restrictions (Liu et al., 2021) This is likely explained in part by the association of neuroticism with greater concerns about finances and relationships which were heightened even more than usual during the COVID‐19 pandemic (Aschwanden et al., 2021) Additional, it has also been proposed that people who score high in neuroticism may have focused more on COVID-19-related information and consequences than people who score lower in this trait (Khosravi, 2020).

Extraversion reflects positive mood, optimism, and the degree to which a person needs attention and social interaction. Those who score higher on extraversion typically report higher subjective well-being, as well as several of its facets (for example, happiness, life satisfaction, and quality of life) (Steel et al., 2008), although recent cross-sectional evidence showed that it was associated with greater perceived stress during COVID-19 (Liu et al., 2021). Similarly, in the current study extraversion was associated with a greater increase in depressive symptoms during COVID-19 compared to prior trends. The COVID-19 containment strategies employed in Ireland to protect the health of older adults were specifically designed to minimise in-person contact between individuals, which likely had a greater impact on people who scored higher on extraversion. Compounding this, 71% of adults aged ≥50 years had internet access throughout the pandemic and, among those who did, less than half used it for social activities like audio/video calls (43%) or social media (40%). It is therefore plausible that the COVID-19 containment strategies had a greater impact on loneliness and social isolation, two well-established risk factors for depressive symptoms (Domènech-Abella et al., 2019) among those with higher extraversion by inhibiting their need for social engagement.

Openness, the degree to which a person needs intellectual stimulation, change, and variety, has previously been shown to be among the personality traits associated with better resilience in coping with the COVID-19 pandemic (Fernández et al., 2020) and associated with lower COVID-19 fear and better sleep quality (Ahmed et al., 2021). Cross-sectional studies have also shown associations between openness and its facets with lower perceived stress and better psychological outcomes during COVID-19 (Ahmed et al., 2021; Modersitzki et al., 2021). Supporting these previous findings, in the current study, individuals who scored higher on openness reported smaller increases in depressive symptoms during the pandemic.

One possible pathway by which personality traits influenced pandemic-related changes in depressive symptoms is through their relationships with physical activity. Meta-analytic evidence supports positive associations between physical activity and extraversion, conscientiousness, and openness, and negative associations between physical activity and neuroticism (Sutin et al., 2016; Wilson et al., 2015) while physical activity has also been negatively impacted by the pandemic, and these negative changes associated with worse mental health (Meyer et al., 2020). Prior to the pandemic, a growing body of research had sought to elucidate the complex relationships between personality, physical activity, and mental health (McDowell et al., 2020; Moor & Geus, 2018; Wilson et al., 2016). Future research should consider the interrelations of personality and activity behaviours (e.g., sedentary behaviour and physical activity) and their association with COVID-19-related changes in mental health.

This study has a several key strengths. Firstly, the longitudinal study design allowed the examination of changes in depressive symptoms associated with the COVID-19 pandemic, expanding on findings from cross-sectional studies that could not account for mental health prior to the pandemic. Secondly, the sample was randomly selected from the national population, building on recent evidence that primarily relies on convenience samples. Nonetheless, there are several limitations to the current study. Although the CESD-8 has been well-validated in the TILDA cohort, the gold standard to establish a clinical diagnosis of depression is a diagnostic structural interview. Secondly, personality traits would be better examined using multiple data sources (for example peer reports, experience sampling measurements) rather than a single questionnaire; however, these additional data sources are not assessed in TILDA. Finally, this study is observational and, although we controlled for key potential confounders of the relationships of interest, it remains plausible that confounding still exists.

Conclusions

Depressive symptoms increased significantly among older adults during the COVID-19 pandemic compared to what would have been expected from the trends observed prior to the pandemic. People who score higher on the personality traits neuroticism and extraversion, and lower on openness, had reported the greatest increases in depressive symptoms. While we have demonstrated differences according to personalitgy type in the accumulation of deprsssive symptoms during the COVID-19 pandemic specifically, our findings support previous studies that examined this association prior to the pandemic. Taken together, these findings can inform our understanding of the association of personality with both the onset and course of depression and may therefore contribute to the early identification of at-risk individuals and the tailoring of prevention and treatment programs.