Sex-specific psychological risk profiles of CVD in the HUNT study: the role of neuroticism and extraversion

Abstract

Objective

The aim was to investigate psychological risk profiles of cardiovascular disease (CVD). Depression and anxiety have been linked to CVD, but research has not incorporated personality and sex-specific analyses are warranted. In this study, we examine the role of sex, neuroticism, extraversion, anxiety and depression on the risk of CVD.

Method

Using data from the HUNT-study and the mortality register, 32,383 (57.10% men) participants were followed for an average of 10.48 years. During this time, 142 died of myocardial infarction (MI) and 111 of stroke.

Results

Cox regression showed that depression (HR = 1.07, 95% CI = [1.00, 1.14]) and neuroticism (1.23 [1.08, 1.40]) were significantly related to an increased risk of MI. One standard unit increase in depression and neuroticism was associated with 1.22 [CI 1.01, 1.47] increase and 1.43 [CI 1.14, 0.78] increase in the risk of MI respectively. For stroke, there was no significant effect of anxiety, depression or personality. However, we found a significant interaction effect between sex and extraversion where higher extraversion was associated with greater risk of stroke for women only.

Conclusions

Both neuroticism and depression were related to MI. We observed an interaction between extraversion and sex with stroke, but the effect size was small. The role of extroversion as a risk factor for CVD remains inconclusive.

Keywords:

• Psychocardiology
• personality
• depression
• anxiety
• myocardial infarction
• stroke
• extraversion
• neuroticism

Introduction

Ischaemic heart disease (CHD) and stroke are leading causes of death globally (Wang et al., 2016). Blocking or interruption of the blood vessels supplying the heart can lead to CHD (Xu et al., 2015), of which one of the most common sub-types is myocardial infarction (MI). Obstruction in the blood vessels supplying the brain can lead to cerebrovascular disease, of which stroke is among the most common. According to guidelines (Visseren et al., 2022), persons with mental disorders are subject to special attention when it comes to the prevention of cardiovascular diseases (CVDs). Anxiety and depression are related to CVD (Gan et al., 2014; Wu & Kling, 2016), though depression more strongly than anxiety (Karlsen et al., 2021). Both angry and sadness rumination can cause cardiovascular reactivity in terms of increased heart rate and blood pressure (Busch et al., 2017). On the other hand, metabolic and inflammatory factors (i.e. diabetes, triglycerides and waist circumference and C-reactive protein) are prospectively associated with the onset of depression (Rudaz et al., 2017), and more evidence is needed on the role of psychosocial factors and whether they improve risk prediction beyond traditional risk factors (Visseren et al., 2022), especially concerning stroke (Graber et al., 2019). While anxiety and depression are distinct mental disorders, they have overlapping symptomology and are frequently comorbid (Brown et al., 2001; Jacobson & Newman, 2017). Anxiety and depression share neurobiological correlates (Neumann, 2020), and both are closely linked to dispositional tendencies to experience negative emotions, i.e. the personality trait neuroticism (Kotov et al., 2010). Meta-analysis supports the prospective association between neuroticism and mental disorders (Jeronimus et al., 2016). Symptoms of depression and anxiety often present trait-like properties in the general populations (Langvik & Hjemdal, 2015) implying that the role of dispositional factors should be considered when investigating the association between depression, anxiety and CVD. Different models have been suggested to explain the relationship between personality and affective disorders, including the common cause-model, where both personality (i.e. neuroticism) and mental disorders share a common genetic determinants (Ormel et al., 2013). The vulnerability model postulate that personality influence the susceptibility for developing mental disorders, while the spectrum-model suggest that mental disorders are more extreme versions of the trait (Clark, 2005). For an overview of the various models proposed for the link between neuroticism and common mental disorders, see Ormel et al. (2013). Research on personality characteristics as a risk or protective factors for developing CVD is lacking (Dahlén et al., 2022). Further, a recent study addresses the importance of incorporating neuroticism when examining the role of mental health and CVD, due to the relatively scarce research on neuroticism and CVD. Although neuroticism overlaps with symptoms of anxiety and depression, neuroticism might also have an independent association to CVD (Li et al., 2022).

While neuroticism represents a vulnerability to developing affective disorders in general, the trait extraversion is especially related to depression (Watson et al., 2015), and negatively related to common mental disorders, although with weaker association than neuroticism (Kotov et al., 2010). The core of extraversion is sociability and positive emotions (McCrae & Costa, 2010). Positive affect has been identified as protective against 10-year incident coronary heart disease, also when controlling for depression (Davidson et al., 2010). Hence, personality has relevance for CVD risk not only as a predictive factor of affective disorders, but also as a direct risk factor of CVD (Denollet et al., 1996; Friedman & Rosenman, 1959; Jokela et al., 2014).

Personality traits influence both frequency and intensity of positive and negative emotions (Komulainen et al., 2014), and while personality traits describe dimensional aspects of personality, the typology approach focuses on distinct categories, like the Type A personality typology characterised, e.g. aggressiveness (Friedman & Rosenman, 1959). Despite its popularity in popular psychology, a large-scale study applying different measures of Type A assessment, concluded that there is no evidence to support the Type A as a CVD risk factor (Šmigelskas et al., 2015). Another personality construct that has received attention is the Type D personality, defined by a combination of negative affectivity and social inhibition (Denollet & Brutsaert, 1998). Although increased mortality of CVD patients with type D personality is observed (Denollet & Brutsaert, 1998; Grande et al., 2012; Kupper & Denollet, 2018; de Voogd et al., 2012), a review of the prognostic value of Type D in cardiac samples concluded that the effect sizes probably have been overestimated (Grande et al., 2012), and the research has been criticised for lacking statistical power and dichotomising Type D personality instead of treating it as a continuous trait, given that the category represent different and distinct characteristics that are normally distributed in the population (Coyne et al., 2011; Horwood & Anglim, 2017). The Type D sub-scales of social inhibition and negative affect are strongly correlated to extraversion, neuroticism, conscientiousness, agreeableness and openness traits from the five-factor model (De Fruyt & Denollet, 2002), and these continuous subscales are better predictors of health-related variables than categorical Type D (Horwood & Anglim, 2017). Whereas one study failed to identify association between Type D and incident CHD (Larson et al., 2013), the sub-component social inhibition has been associated with coronary artery plaque in CHD-free populations (Compare et al., 2014). Although different taxonomies on personality have been given attention over the years, the majority of personality constructs likewise correspond to the factors in the five-factor model (McCrae, 2010; McCrae & John, 1992), and more recent biological approach to personality emphasis the possible integration of different models, where extraversion and neuroticism are considered as the most important traits (Markon et al., 2005).

Although neuroticism predicts all-cause mortality (O’Súilleabháin & Hughes, 2018) and is linked to immune functioning (Mengelkoch et al., 2022), and extraversion is suggested as a predictor of longevity (Chapman et al., 2011) and both physical and social activities across the lifespan (Lai & Qin, 2020), several questions remain whether neuroticism and extraversion are associated with CVD or CVD mortality (Otonari et al., 2021). Studies have observed that neuroticism is related to higher risk of death from CVD (Shipley et al., 2007), and some that neuroticism has a link to CVD independent of depression (Čukić & Bates, 2015). Other studies suggest that neuroticism has a synergistic interaction with depression, increasing CVD risk (Almas et al., 2017), further arguing for the importance of including both personality and symptoms of anxiety and depression when investigating psychological risk profiles for CVD.

Jokela et al. (2014) found that extraversion was linked to increased risk of stroke mortality, while neuroticism was linked to increased risk of CHD mortality. However, a large prospective cohort study from Japan found no association between CVD mortality and either extraversion or neuroticism on CVD (Narita et al., 2020), addressing the need for more research on the link between personality and CVD.

Sex and gender differences in associations between psychological risk factors and CVD

Research on women and psychological risk factors of CVD is scarce (Espnes et al., 2015), and studies of women have been underpowered compared to those of men (Visseren et al., 2022). The lifetime prevalence and morbidity of depressive disorders are higher in females than in males (Faravelli et al., 2013; Piccinelli & Wilkinson, 2000). Gender differences in personality traits are considered small but consistent across cultures (Costa et al., 2001); women tend to score higher on neuroticism compared to men, whereas for extraversion, women score higher on the extraversion facets of warmth, gregariousness and positive emotions. However, some argue that the gender differences in personality are substantial, and that a use of multivariate approach offers a different perspective on sex differences (Kaiser et al., 2020).

Biological sex influences CVD risk through sex-specific and unique risk factors like pregnancy or polycystic ovaries syndrome) (Cho et al., 2020). Sex hormones affect neurotransmitters like serotonin and dopamine (Barth et al., 2015), and studies have identified that the personality-psychopathology connection is moderated by sex (Neumann, 2020).

Gender, which refers to the socially constructed roles, behaviour, expressions and identities of individuals, is also an important aspect to consider when investigating sex differences in CVD (Connelly et al., 2021). Both anxiety, depression and stress are important factors to consider in CVD prevention for women, as women are overrepresented concerning prevalence of affective disorders (Cho et al., 2020). Further, large-scale studies have identified and that there is a gender-specific association between depression and CVD, where depression was identified as a stronger risk marker for women compared to men (Haukkala et al., 2009; Langvik & Hjemdal, 2015). A study found that neuroticism increased the risk of CVD mortality in women with low socio-economic status (SES) but lowered the risk for women with high SES (Hagger-Johnson et al., 2012), suggesting that societal factors play an important role. For cancer mortality, the effect of personality is opposite for men and women (Otonari et al., 2021). Further, studies have shown that associations between personality and different cardiovascular outcome differ between men and women, where extraversion is associated with increased risk for stroke for women only (Jokela et al., 2014). For incident MI, a study has shown that higher score on neuroticism was associated with an increased risk, and the risk ratio was stronger for women compared to men (Dahlén et al., 2022). As the associations between mental health and risk of CVD are stronger among women than men (Li et al., 2022), suggesting that research should be both gender and outcome specific. Research has found traditional risk factors of CVD to have different prevalence and effects for men and women (Connelly et al., 2021; Gao et al., 2019; Mauvais-Jarvis et al., 2020), and some cardiovascular risk scores underestimate the risk for women (Thurston et al., 2013). For this reason, it is important to establish whether novel risk markers also exhibit sex-specific effects or not.

Summary and research aim

The aim of this study is to examine the joint contribution of personality, anxiety and depression on specific CVD outcomes. Studies on personality and CVD that report no association between the two often use CVD as a general outcome combining both MI and stroke (Almas et al., 2017; Hagger-Johnson et al., 2012). However, research suggests that the association between personality and CVD depends on both outcome and gender (Jokela et al., 2014), and large population-based studies exploring the role of personality as a risk marker for specific CVD outcome has been requested (Dahlén et al., 2022). Given the close and complex relationship between personality and affective disorders (Kotov et al., 2010; Ormel et al., 2013), gender differences in personality (Costa et al., 2001; Kaiser et al., 2020) and affective disorders (Faravelli et al., 2013; Neumann, 2020; Piccinelli & Wilkinson, 2000), including the gender-specific association between psychological variables and CVD outcomes (Hagger-Johnson et al., 2012; Haukkala et al., 2009; Jokela et al., 2014; Li et al., 2022), it is pivotal to address the role of personality as a predictor for different CVD outcomes. Previous research on the roles of psychological risk factors of CVD has been hampered by the focus on broad, heterogeneous categorisations of CVD instead of its sub-types (Karlsen et al., 2021). For this reason, we analyse the outcomes of MI mortality and stroke mortality separately. In this study, we investigate the role of the personality traits neuroticism and extraversion, together with anxiety and depression, and their associations with stroke and MI. We also investigate interactions between the personality traits and sex. The research questions of this study were as follows:

1. Are neuroticism and extraversion associated with MI and stroke?

2. Is depression and anxiety associated with MI and stroke when including neuroticism and extraversion?

3. Are there gender differences in the associations between personality and CVD?

Methods

Participants and procedure

This study uses data collected for The Trøndelag Health Study (HUNT). All participants in the Nord-Trøndelag area of Norway were invited to participate in a longitudinal population study in 1984. Every decade since, those same participants, as well as people who have moved to that area since, have been invited to participate again. This study uses data from the third wave of data collection (2006–2008). This data were combined with data from the Norwegian Cause of Death Registry (DÅR) to identify deaths related to stroke and MI. DÅR registered deaths from the start of HUNT3 and until the end of 2017. Deaths from stroke or MI were recorded continually in this period. Participants with missing data on any of the variables included in the statistical models were excluded, as were those with pre-existing cardiovascular conditions before the start of HUNT3. This left a total sample of 32,383 participants. The average age of participants at the time of the study start was 52.25 (SD = 14.18), and the average follow-up period was 10.48 years. There were more women than men in the sample, at 57.10% (n= 18,490). The study included 337,838 person-years in total. Further information about HUNT can be found in Holmen et al. (2003) and Krokstad et al. (2013).

Measures

Cardiovascular disease

The DÅR records the cause of death for each person who either dies in Norway or is registered as living in Norway. The cause of death is determined by a physician and follows the ICD-10 system for classifying diseases and health problems. In this study, ICD-10 codes I21–I22 indicated death by MI, while ICD-10 codes I60–I69 indicated death by stroke.

Personality

Extraversion and neuroticism were measured using a questionnaire based on the Revised Eysenck Personality Questionnaire (Eysenck et al., 1985). The items were translated based on the Norwegian version of the standard Eysenck Personality Questionnaire (Eysenck & Tambs, 1990). Extraversion and neuroticism were both measured using six items, each of which had a binary option (0=no, 1=yes). This yielded a final score ranging from 0 to 6. Examples of items were: ‘Are you mostly quiet and reserved when you are around other people’? (extraversion) and ‘Do you worry that terrible things might happen’? (neuroticism). Cronbach’s alpha was 0.72 for extraversion and 0.74 for neuroticism.

Symptoms of anxiety and depression

Anxiety and depression were measured as a continuous variable, using the Hospital Anxiety and Depression Scales (Zigmond & Snaith, 1983). A Norwegian translation was made specifically for the HUNT study (HUNT, 2021). The HADS consists of 14 items (7 each for anxiety and depression), querying how often the participants do or feel each item compared over the last week, on a scale from 0 to 3. Examples of items are ‘I get sudden feelings of panic,’ scored from ‘very often indeed’ to ‘not at all’ (anxiety) and ‘I look forward with enjoyment to things’, scored from ‘as much as I ever did’ to ‘hardly at all’ (depression). The composite depression and anxiety scales have a potential range from 0 to 21. Cronbach’s alpha was 0.75 for depression and 0.81 for neuroticism.

Covariates

We added common risk factors of CVD to the statistical models: systolic blood pressure, cholesterol level, waist-hip ratio, age, sex, diabetes, smoking status and antidepressant use.

Systolic blood pressure

Systolic blood pressure was measured by trained nurses using oscillometry. The participants had been seated for 2 min before measurements were taken. Three measurements were recorded, and the mean of the second and third was used.

Waist-hip ratio

Measurements of the waist and hip were taken with a steel band while participants were standing. The hip circumference was measured at the thickest part, while the waist was measured at the height of the umbilicus.

Cholesterol

Serum cholesterol was analysed by enzymatic cholesterol esterase methodology.

Additionally, sex and age were recorded. The presence of diabetes and smoking status was self-reported via questionnaires. Diabetes was operationalised as 0 (no diabetes) or 1 (diabetes). Smoking status had four levels (0 = never, 1 = former, 2 = sometimes and 3 = current). Sex was operationalised as 0 (woman) or 1 (man).

Statistical analysis

We analysed cox proportional hazards model to analyse the survival data. We ran two analyses: one with fatal MI as the outcome, the other with fatal stroke as the outcome. For each of these conditions, we analyse the sample in several steps: Initially, the simple effects of extraversion, neuroticism, depression and anxiety were investigated. Then we ran models with combinations of predictors: Model 1 included extraversion and neuroticism. Model 2 included Model 1 as well as depression and anxiety. Model 3 included Model 2 as well as the covariates, namely diabetes, systolic blood pressure, cholesterol, waist-hip ratio, smoking status, age and sex. Model 4 included Model 3 as well as two interaction terms: sex interacting with extraversion and sex interacting with neuroticism. R version 3.6.3 with the survival package version 3.2-12 was used for the analyses. For each model, Royston and Sauerbrei’s pseudo-R² was reported (Royston & Sauerbrei, 2004), alongside the Wald test, the LR test and the Log Likelihood. Since the psychological variables were measured using different scales, the models were repeated with standardised versions of these variables to compare their relative effect on the risk estimates of the outcomes.

Results

Descriptive statistics and group differences are displayed in Table 1 and Table 2. The average time to endpoint (fatal MI or stroke, or censored) was 10.48 years, during which 142 participants died of MI while 111 died of stroke. In the total sample, those who died from MI had on average higher levels of neuroticism and depression, and lower levels of extraversion, compared to those who did not die from MI. The same pattern was found for women. For men, however, there were no significant differences in extraversion or neuroticism, but those with MI had on average a higher level of depression than those without MI. In the total sample, those who died from stroke had on average higher levels of neuroticism and depression and lower levels of extraversion. Among the women, those who died from stroke had on average a higher level of neuroticism and depression than those who did not die from stroke. The men who died from stroke had on average lower levels of extraversion and higher levels of depression compared to those who did not die from stroke.

Table 1. Descriptive statistics for those who had MI vs. those who did not, for the total sample, women and men, respectively.

	Total				Women				Men
	No MI (N= 32,241)	MI (N= 142)	p	ES	No MI (N= 18,435)	MI (N = 55)	p	ES	No MI (N = 13,806)	MI (N= 87)	p	ES
Extraversion	3.69 (1.79)	3.24 (1.68)	0.003	0.25	3.78 (1.76)	3.20 (1.54)	0.015	0.33	3.57 (1.82)	3.26 (1.77)	0.118	0.17
Neuroticism	1.63 (1.71)	1.96 (1.78)	0.022	−0.19	1.89 (1.77)	2.55 (1.83)	0.006	0.37	1.28 (1.55)	1.59 (1.65)	0.067	−0.20
Depression	3.19 (2.84)	4.54 (2.96)	<0.001	−0.48	3.02 (2.81)	4.82 (2.87)	<0.001	−0.64	3.42 (2.86)	4.37 (3.02)	0.002	−0.33
Anxiety	3.99 (3.27)	3.82 (3.16)	0.544	0.05	4.33 (3.44)	4.76 (3.39)	0.354	−0.13	3.54 (2.98)	3.23 (2.87)	0.340	0.10
SBP	130.12 (18.25)	144.68 (23.53)	<0.001	−0.80	127.38 (18.93)	146.36 (25.10)	<0.001	−1.00	133.79 (16.62)	143.62 (22.56)	< 0.001	−0.59
Cholesterol	5.55 (1.08)	5.99 (1.01)	<0.001	−0.40	5.58 (1.12)	6.16 (1.16)	<0.001	−0.52	5.52 (1.04)	5.88 (0.89)	0.001	−0.34
WHR	0.90 (0.08)	0.94 (0.07)	<0.001	−0.63	0.87 (0.07)	0.91 (0.07)	<0.001	−0.57	0.94 (0.06)	0.97 (0.06)	< 0.001	−0.50
Age	52.16 (14.78)	72.16 (10.38)	<0.001	−1.35	51.57 (15.12)	75.01 (11.12)	<0.001	−1.55	52.95 (14.28)	70.36 (9.51)	< 0.001	−1.22
Diabetes
No	31,039 (96.3%)	126 (88.7%)	<0.001	0.03	17,824 (96.7%)	49 (89.1%)	0.002	0.02	13,215 (95.7%)	77 (88.5%)	< 0.001	0.03
Yes	1202 (3.7%)	16 (11.3%)			611 (3.3%)	6 (10.9%)			591 (4.3%)	10 (11.5%)
Smoking
Never	14,169 (43.9%)	51 (35.9%)	0.062	0.02	8243 (44.7%)	27 (49.1%)	0.531	0.01	5926 (42.9%)	24 (27.6%)	0.004	0.03
Former	10,428 (32.3%)	51 (35.9%)			5583 (30.3%)	16 (29.1%)			4845 (35.1%)	35 (40.2%)
Sometimes	2295 (7.1%)	7 (4.9%)			1206 (6.5%)	1 (1.8%)			1089 (7.9%)	6 (6.9%)
Current	5349 (16.6%)	33 (23.2%)			3403 (18.5%)	11 (20.0%)			1946 (14.1%)	22 (25.3%)
Stroke
No stroke	32,130 (99.7%)	142 (100.0%)	0.484	0.00	18,367 (99.6%)	55 (100.0%)	0.652	0.00	13,763 (99.7%)	87 (100.0%)	0.602	0.00
Stroke	111 (0.3%)	0 (0.0%)			68 (0.4%)	0 (0.0%)			43 (0.3%)	0 (0.0%)
Sex
Female	18,435 (57.2%)	55 (38.7%)	<0.001	0.02
Male	13,806 (42.8%)	87 (61.3%)

Note. SBP: systolic blood pressure; WHR: waist hip ratio; MI: myocardial infarction

For continuous measures, mean and standard deviations are shown. For categorical measures, frequency and percentage are shown. Effect size is Cohen’s d for continuous variables, φ for categorical variables with only two levels, and Cramer’s V for categorical variables with more than two levels.

Table 2. Descriptive statistics for those who had stroke vs. those who did not, for the total sample, women and men, respectively.

	Total				Women				Men
	No stroke	Stroke			No stroke	Stroke			No stroke	Stroke
	(N= 32,272)	(N = 111)	p	ES	(N = 1422)	(N = 68)	p	ES	(N = 13,850)	(N= 43)	p	ES
Extraversion	3.69 (1.79)	3.17 (1.70)	0.002	0.29	3.78 (1.76)	3.50 (1.56)	0.192	0.16	3.57 (1.82)	2.65 (1.81)	<0.001	0.51
Neuroticism	1.63 (1.71)	1.97 (1.77)	0.034	−0.20	1.89 (1.77)	2.38 (1.84)	0.022	−0.28	1.28 (1.55)	1.33 (1.44)	0.859	−0.03
Depression	3.19 (2.84)	4.53 (3.21)	<0.001	−0.47	3.02 (2.81)	4.49 (3.36)	<0.001	−0.52	3.43 (2.86)	4.60 (3.01)	0.007	−0.41
Anxiety	3.99 (3.27)	4.23 (3.62)	0.432	−0.07	4.33 (3.44)	4.63 (3.91)	0.473	−0.09	3.53 (2.98)	3.60 (3.03)	0.875	−0.02
SBP	130.14 (18.27)	145.57 (22.01)	<0.001	−0.84	127.37 (18.92)	145.25 (24.89)	<0.001	−0.94	133.81 (16.66)	146.07 (16.73)	<0.001	−0.74
Cholesterol	5.55 (1.08)	5.90 (1.20)	<0.001	−0.32	5.58 (1.12)	6.13 (1.25)	<0.001	−0.49	5.52 (1.04)	5.54 (1.04)	0.927	−0.01
WHR	0.90 (0.08)	0.93 (0.08)	<0.001	−0.40	0.87 (0.07)	0.90 (0.08)	<0.001	−0.48	0.94 (0.06)	0.97 (0.06)	<0.001	−0.51
Age	52.17 (14.78)	74.67 (9.93)	<0.001	−1.52	51.55 (15.10)	76.52 (9.48)	<0.001	−1.66	53.00 (14.30)	71.75 (10.03)	<0.001	−1.31
Diabetes
No	31,063 (96.3%)	102 (91.9%)	0.016	0.01	17,810 (96.7%)	63 (92.6%)	0.065	0.01	13,253 (95.7%)	39 (90.7%)	0.108	0.01
Yes	1209 (3.7%)	9 (8.1%)			612 (3.3%)	5 (7.4%)			597 (4.3%)	4 (9.3%)
Smoking
Never	14,167 (43.9%)	53 (47.7%)	0.326	0.01	8230 (44.7%)	40 (58.8%)	0.077	0.02	5937 (42.9%)	13 (30.2%)	0.225	0.02
Former	10,442 (32.4%)	37 (33.3%)			5582 (30.3%)	17 (25.0%)			4860 (35.1%)	20 (46.5%)
Sometimes	2299 (7.1%)	3 (2.7%)			1206 (6.5%)	1 (1.5%)			1093 (7.9%)	2 (4.7%)
Current	5364 (16.6%)	18 (16.2%)			3404 (18.5%)	10 (14.7%)			1960 (14.2%)	8 (18.6%)
MI
No MI	32,130 (99.6%)	111 (100.0%)	0.484	0.00	18,367 (99.7%)	68 (100.0%)	0.652	0.00	13,763 (99.4%)	43 (100.0%)	0.602	0.00
MI	142 (0.4%)	0 (0.0%)			55 (0.3%)	0 (0.0%)			87 (0.6%)	0 (0.0%)
Sex
Female	18,422 (57.1%)	68 (61.3%)	0.375	0.00	–	–	–	–	–	–	–	–
Male	13,850 (42.9%)	43 (38.7%)

SBP: systolic blood pressure; WHR: waist hip ratio; MI: myocardial infarction

For continuous measures, mean and standard deviations are shown. For categorical measures, frequency and percentage are shown. Effect size is Cohen’s d for continuous variables, φ for categorical variables with only two levels, and Cramer’s V for categorical variables with more than two levels.

The relationship among the predictors (r) is shown in Table 3. Depression and anxiety were positively correlated with neuroticism (0.47 and 0.69, respectively) and to a lesser extent (negatively) with extraversion (−0.39 and −0.12).

Table 3. Pearson’s correlation coefficient for the relationships among predictors (N = 32,383).

	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	11.
1. Extraversion	1.00
2. Neuroticism	−0.16***	1.00
3. Depression	−0.32***	0.47***	1.00
4. Anxiety	−0.12***	0.69***	0.54***	1.00
5. Diabetes	−0.03***	0.02**	0.04***	−0.01*	1.00
6. SBP	−0.11***	−0.05***	0.06***	−0.07***	0.09***	1.00
7. Cholesterol	−0.06***	0.00	0.05***	0.00	−0.08***	0.21***	1.00
8. WHR	−0.06***	−0.02***	0.13***	−0.05***	0.15***	0.27***	0.13***	1.00
9. Smoking	0.05***	0.11***	0.09***	0.11***	0.00	−0.02***	0.06***	0.06***	1.00
10. Age	−0.22***	−0.02***	0.17***	−0.07***	0.13***	0.42***	0.32***	0.28***	0.02**	1.00
11. Sex	−0.06***	−0.18***	0.07***	−0.12***	0.03***	0.17***	−0.03***	0.45***	−0.02***	0.05***	1.00

*p < .05; **p < .01; ***p < .001.

SBP: systolic blood pressure; WHR: waist hip ratio

Baseline for smoking is never, baseline for sex is female.

The results of the cox analysis on the risk of dying from MI are displayed in Table 4. Extraversion significantly associated with a reduced risk of MI (HR = 0.88, 95% CI = [0.81, 0.97], p<.01) in Model 1. When including neuroticism, depression and anxiety symptoms in Model 2, its effect was no longer significant. Neuroticism (HR = 1.19 [1.05, 1.36], p<.01) and depression (HR = 1.20 [1.13, 1.28], p<.001) was significantly associated with an increased risk of MI, while anxiety (HR = 0.83 [0.77, 0.90], p<.001) significantly associated with a decreased risk. In Model 3 with all covariates, neuroticism (HR = 1.23 [1.08, 1.40], p<.01) and depression (HR = 1.07 [1.00, 1.14], p<.05) were significantly associated with an increased risk of MI. The interaction terms were not significant in Model 4, indicating that there were no interactions between sex and extraversion or neuroticism on the risk of MI.

Table 4. Hazard ratios of neuroticism, extraversion, depression and anxiety on risk of myocardial infarction mortality (N = 32,383).

	Simple effects	Model 1	Model 2	Model 3	Model 4
	HR (95% CI)	HR (95% CI)	HR (95% CI)	HR (95% CI)	HR (95% CI)
Extraversion	0.87** (0.80, 0.95)	0.88** (0.81, 0.97)	0.96 (0.87, 1.05)	1.07 (0.97, 1.19)	1.06 (0.90, 1.25)
Neuroticism	1.11* (1.02, 1.22)	1.09 (0.99, 1.19)	1.19** (1.05, 1.36)	1.23** (1.08, 1.40)	1.24* (1.04, 1.47)
Depression	1.15*** (1.09, 1.20)	–	1.20*** (1.13, 1.28)	1.07* (1.00, 1.14)	1.07* (1.00, 1.14)
Anxiety	0.98 (0.93, 1.04)	–	0.83*** (0.77, 0.90)	0.93 (0.86, 1.00)	0.93 (0.86, 1.00)
Diabetes	–	–	–	1.61 (0.94, 2.75)	1.61 (0.94, 2.75)
SBP	–	–	–	1.01* (1.00, 1.02)	1.01* (1.00, 1.02)
Cholesterol	–	–	–	1.25** (1.07, 1.45)	1.24** (1.07, 1.45)
WHR	–	–	–	16.63* (1.53, 181.14)	16.59* (1.52, 180.48)
Smoking
Former	–	–	–	0.95 (0.63, 1.42)	0.95 (0.63, 1.42)
Sometimes	–	–	–	1.11 (0.50, 2.47)	1.12 (0.50, 2.48)
Current	–	–	–	2.35*** (1.49, 3.71)	2.36*** (1.49, 3.72)
Age	–	–	–	1.13*** (1.11, 1.15)	1.13*** (1.11, 1.15)
Sex	–	–	–	2.30*** (1.54, 3.44)	2.17 (0.86, 5.49)
Sex × extraversion	–	–	–	–	1.03 (0.84, 1.25)
Sex × neuroticism	–	–	–	–	0.99 (0.81, 1.20)
R^2D	–	0.05	0.19	0.66	0.66
Log Likelihood	–	−1457.16	−1435.75	−1263.15	−1263.11
Wald test	–	12.68** (df = 2)	62.67** (df = 4)	295.83*** (df = 13)	295.73*** (df = 15)
LR test	–	12.34** (df = 2)	55.16*** (df = 4)	400.34*** (df = 13)	400.43*** (df = 15)
Logrank test	–	12.83** (df = 2)	62.32*** (df = 4)	343.55*** (df = 13)	344.55*** (df = 15)

*p < .05; **p < .01; ***p < .001.

SBP: systolic blood pressure; WHR: waist hip ratio; HR: Hazard ratio; CI: confidence interval; LR: likelihood ratio

Model 1: Combined effects of personality traits. Model 2: Model 1 + effects of mood disorder symptoms. Model 3: Model 2 + all risk markers. Model 4: Model 3 + interaction terms. Pseudo R² is calculated according to Royston and Sauerbrei (2004). Baseline for smoking is never, baseline for sex is female.

The results of cox analysis on the risk of dying from stroke are displayed in Table 5. In Model 1, extraversion (HR = 0.86 [0.78, 0.96], p<.001) was significantly associated with a decreased risk of stroke. This association disappeared in Model 2, in which depression (HR = 1.17 [1.09, 1.25], p<.001) was associated with an increased risk of stroke while anxiety (HR = 0.90 [0.83, 0.98], p < .05) was associated with a decreased risk. In Model 3 with all variable and covariates, only age (HR = 1.16 [1.14, 1.18], p<.001) was significantly associated with stroke. However, adding interaction terms in Model 4 resulted in a significant interaction between sex and extraversion (HR = 0.74 [0.59, 0.94], p<.05). There was a small increase in the pseudo-R² value of Model 4 (R²=0.70) compared to Model 3 (R²=0.69), supporting the interaction. VIF was calculated for all models. None reached a level of 5, indicating that there were no issues of multicollinearity. The highest VIF values were for anxiety in the full Model 4 s, at 2.20 and 2.29 for MI and stroke, respectively.

Table 5. Hazard ratios of neuroticism, extraversion, depression and anxiety on risk of stroke mortality (N = 32,383).

	Simple effects	Model 1	Model 2	Model 3	Model 4
	HR (95% CI)	HR (95% CI)	HR (95% CI)	HR (95% CI)	HR (95% CI)
Extraversion	0.85** (0.77, 0.94)	0.86** (0.78, 0.96)	0.93 (0.83, 1.03)	1.05 (0.93, 1.18)	1.19* (1.02, 1.39)
Neuroticism	1.12* (1.01, 1.24)	1.09 (0.98, 1.21)	1.10 (0.95, 1.28)	1.06 (0.91, 1.24)	1.11 (0.94, 1.32)
Depression	1.15*** (1.09, 1.21)	–	1.17*** (1.09, 1.25)	1.06 (0.99, 1.14)	1.06 (0.99, 1.15)
Anxiety	1.02 (0.97, 1.08)	–	0.90* (0.83, 0.98)	1.01 (0.93, 1.10)	1.01 (0.93, 1.09)
Diabetes	–	–	–	1.02 (0.51, 2.05)	1.03 (0.51, 2.08)
SBP	–	–	–	1.01 (1.00, 1.02)	1.01 (1.00, 1.02)
Cholesterol	–	–	–	1.03 (0.86, 1.22)	1.02 (0.86, 1.22)
WHR	–	–	–	8.84 (0.63, 124.73)	9.16 (0.65, 129.75)
Smoking
Former	–	–	–	0.90 (0.58, 1.40)	0.89 (0.57, 1.38)
Sometimes	–	–	–	0.58 (0.18, 1.87)	0.58 (0.18, 1.85)
Current	–	–	–	1.66 (0.95, 2.91)	1.62 (0.93, 2.84)
Age	–	–	–	1.16*** (1.14, 1.18)	1.16*** (1.14, 1.18)
Sex	–	–	–	0.91 (0.58, 1.44)	2.72* (1.03, 7.21)
Sex × extraversion	–	–	–	–	0.74* (0.59, 0.94)
Sex × neuroticism	–	–	–	–	0.90 (0.71, 1.15)
R^2D	–	0.06	0.15	0.69	0.70
Log likelihood	–	−1127.85	−1118.17	−957.66	−954.34
Wald test	–	12.29** (df = 2)	35.32*** (df = 4)	246.34*** (df = 13)	250.70*** (df = 15)
LR test	–	11.98** (df = 2)	31.35*** (df = 4)	352.36*** (df = 13)	359.01*** (df = 15)
Logrank test	–	12.47** (df = 2)	35.79*** (df = 4)	319.37*** (df = 13)	326.97*** (df = 15)

Note. *p < 0.05; **p < 0.01; ***p < 0.001.

SBP: systolic blood pressure; WHR: waist hip ratio; HR: hazard ratio; CI: confidence interval; LR: likelihood ratio

Model 1: Combined effects of personality traits. Model 2: Model 1 + effects of mood disorder symptoms. Model 3: Model 2 + all risk markers. Model 4: Model 3 + interaction terms. Pseudo R² is calculated according to Royston and Sauerbrei (2004). Baseline for smoking is never, baseline for sex is female.

To explore the direction of the interaction effect of sex and extraversion, we ran new analyses separately for men and women with stroke as the outcome. These are displayed in Supplementary Table S1, available online or upon request. These analyses showed that extraversion was significantly associated with increased risk of stroke for women (HR = 1.21[1.04, 1.42], p<.05), but not for men (HR = 0.86 [0.72, 1.04], p≥ 0.05). Model 1 had a pseudo-R² of 0.05 for MI and 0.06 for stroke, indicating that only a small part of the risk was accounted for by personality. Adding depression and anxiety to the models increased pseudo-R² to 0.19 and 0.15, respectively, indicating that depression and anxiety explained a larger part of the variance in risk than personality. Finally, adding the covariates to the models caused a larger increase in pseudo-R² to 0.66 and 0.69.

Table 6 presents the results from analysis with standardised measure of personality, anxiety and depression. The HR per 1-SD higher score on symptoms of depression for MI was 1.22 [1.01,1.47]. For neuroticism, the respective figure was 1.43 [1.14, 1.78].

Table 6. Cox models with standardised psychological predictors.

	MI
	Simple effects	Model 1	Model 2	Model 3	Model 4
	HR (95% CI)	HR (95% CI)	HR (95% CI)	HR (95% CI)	HR (95% CI)
Extraversion	0.78** (0.66, 0.91)	0.80** (0.68, 0.94)	0.92 (0.78, 1.09)	1.14 (0.94, 1.37)	1.10 (0.82, 1.48)
Neuroticism	1.20* (1.03, 1.39)	1.15 (0.99, 1.35)	1.35** (1.08, 1.69)	1.43** (1.14, 1.78)	1.44* (1.07, 1.93)
Depression	1.47*** (1.29, 1.68)	–	1.69*** (1.43, 2.01)	1.22* (1.01, 1.47)	1.22* (1.01, 1.47)
Anxiety	0.95 (0.80, 1.12)	–	0.55*** (0.43, 0.70)	0.79 (0.62, 1.01)	0.79 (0.62, 1.01)
Sex × extraversion		–	–	–	1.05 (0.73, 1.50)
Sex × neuroticism			–	–	0.98 (0.70, 1.37)
	Stroke
	Simple effects	Model 1	Model 2	Model 3	Model 4
	HR (95% CI)	HR (95% CI)	HR (95% CI)	HR (95% CI)	HR (95% CI)
Extraversion	0.75** (0.63, 0.90)	0.77** (0.64, 0.93)	0.87 (0.72, 1.06)	1.09 (0.89, 1.35)	1.37* (1.04, 1.79)
Neuroticism	1.21* (1.02, 1.43)	1.16 (0.97, 1.38)	1.19 (0.92, 1.52)	1.11 (0.86, 1.44)	1.20 (0.90, 1.60)
Depression	1.47*** (1.27, 1.71)	–	1.56*** (1.29, 1.90)	1.19 (0.96, 1.47)	1.20 (0.97, 1.48)
Anxiety	1.07 (0.90, 1.28)	–	0.71* (0.54, 0.93)	1.03 (0.79, 1.35)	1.02 (0.78, 1.34)
Sex × extraversion		–	–	–	0.59* (0.39, 0.89)
Sex × neuroticism		–	–	–	0.84 (0.56, 1.26)

Note. *p < .05; **p < .01; ***p < .001.

HR: hazard ratio; CI: confidence interval; LR: likelihood ratio

Model 1: Combined effects of personality traits. Model 2: Model 1 + effects of mood disorder symptoms. Model 3: Model 2 + all risk markers (not shown). Model 4: Model 3 + interaction terms. Baseline for smoking is never, baseline for sex is female. Psychological variables have been standardised so that the HR represents the change in hazard associated with 1 SD of the predictor.

Discussion

In this study, we measured extraversion and neuroticism, and symptoms of depression and anxiety during 2006–2008 and followed the participants until the end of 2017, with an average follow-up time of 10.48 years. While adjusting for all covariates, neuroticism and depression was associated with an increased risk of MI, and no interactions between sex and personality was observed. While one standard unit increase in depression was associated with 1.22 [CI 1.01,1.47] increase in risk of MI, the respective figure for neuroticism was 1.43 [1.14–1.78], indicating a stronger effect of personality. Hence, for each standard unit increase in neuroticism, there is a 43% increase in probability of dying from MI. For stroke, however, extraversion and sex demonstrated an interaction effect. Investigating women and men separately, we observed that extraversion was associated with an increased risk of stroke for women, but not for men. Looking at the explained variance attributed to each model, personality explained less variance than anxiety and depression.

Given that depression has repeatedly been shown to increase the risk of CVD and CHD (Harshfield et al., 2020; Wu & Kling, 2016), and that neuroticism is associated with increased risk of stroke (Liu et al., 2021), it was unexpected that depression and neuroticism only affected the risk of MI and not stroke. Further, neither WHR, cholesterol, nor smoking was significant predictors of stroke. As none of the traditional risk factors of stroke were significant predictors, the results should be interpreted with cation. It should be noted that the role of psychosocial factors in predicting stroke has been far less conclusive than for MI due to stroke encompassing different outcomes (e.g. large-artery atherosclerosis and cardioembolism) (Graber et al., 2019). However, the observations in this study further argue for investigating specific CVD outcomes separately. The effects of anxiety on CVD are less certain, though some studies have found that anxiety also increases the risk of CVD (Batelaan et al., 2016; Pérez-Piñar et al., 2017). Our study does not support this, but rather supports the notion that anxiety does not impact CVD risk independently of depression, a finding consistent with other studies including both anxiety and depression as predictors (see Karlsen et al., 2020 for a review on the literature).

The results are comparable to those of Jokela et al. (2014), as we found that extraversion was linked to increased risk of stroke, though only for women. Likewise, we also found that neuroticism was linked to CHD (in our case MI). Personality is more stable than the more mutable affective conditions of depression and anxiety. Personality are a set of stable characteristic patterns of thought, cognition and behaviour (McCrae & Costa, 2010), while depression and anxiety are departures from normal functioning and the target of treatment with the goal of returning to the condition pre-existing the disorders. Neuroticism and extraversion are both predictors of depression (Kotov et al., 2010; Ormel et al., 2013; Watson et al., 2015), and high stability in symptom level of anxiety and depression over long-time interval is observed (Langvik & Hjemdal, 2015). Indeed, neuroticism was found to be correlated with anxiety and depression in this study. However, symptoms of depression retained its significant association with MI even after including neuroticism and extraversion, although the effect size was smaller for depression than neuroticism, when comparing the HR based on standardised measures. Further, simple effect of extraversion being negatively associated with MI and stroke was observed, although not in the models including neuroticism, anxiety and depression. As the positive general association between neuroticism and mental illness is stronger than the more specific, negative association between extraversion and depression (Kotov et al., 2010; Watson et al., 2015), these results were not surprising. The initial negative association between extraversion and CVD are hence most like attributable to the negative association between depression and extraversion.

Few studies have examined sex and gender differences in CVD risk, and updated guidelines on CVD prevention (Visseren et al., 2022) highlight the importance of investigating the role of psychological factors independent for men and women. There was a significant group difference, i.e. lower score on extraversion in the analyses of MI for women, but not men, whereas the opposite was the case in analyses of stroke, where we observed a significant lower score on extraversion for men only. We found that extraversion was related to increased risk of stroke mortality, but only for women in the fully adjusted model. As the direction of the association between extraversion and stroke changed from the initial models to the final, the clinical relevance of extraversion appears to be marginal. The positive association between extraversion and stroke needs to be further explored, especially the sex-specific effect. The brains of extraverts react to stimuli more slowly and weakly, leading to an inclination towards a strong sensory stimulation (Eysenck, 1967). The tendency for more extraverted people to involve in risk-taking adventures, that again leads to head trauma, increasing the risk of stroke, has been suggested as a possible reason for the association between extraversion and stroke (Jokela et al., 2014). However, this would not explain the sex difference. As extraversion is associated with a higher number of friends and social activities (Stephan et al., 2014), and that women, more than men provide emotional support and take more care of older parents and friends than men do (Kahn et al., 2011), it is possible that extraverted women are more susceptible to social stress in older age, that again can explain this association. Although the results concerning extraversion is inconclusive the findings in general support the notion of treating men and women as separate populations in the investigation of psychological risk profiles for CVD, as well as addressing the importance of separate analysis for different CVD outcome like MI and stroke.

Strengths and limitations

We restricted the analyses of MI and stroke to those who died of MI or stroke. Differing effects may be found when examining CVD morbidity and mortality (Karlsen et al., 2021). Thus, these results cannot be extrapolated to MI/stroke morbidity. As the HADS excludes somatic symptoms of anxiety and depression, effects of somatic-type depression can be hidden.

Among the strengths of the study is the large size of the sample. While MI and stroke are common causes of death, the occurrence of these events is still relatively rare in a normal population, necessitating a large sample to detect effects. Although the measure of personality used in this study is validated, a more comprehensive measure of personality, preferably including other possible relevant traits, would be preferable. The focus on specific CVD outcomes separately for men and women, including both personality and symptoms of anxiety and depression represent a major strength of this study.

Conclusion

In this study, we investigated the role of the personality traits neuroticism and extraversion along with anxiety and depression as gender-specific risk factors for stroke and MI, controlling for established risk factors of CVD. Contrary to our expectations, neither neuroticism nor depression or anxiety was associated with stroke. Extraversion was not significantly associated with neither MI nor Stroke in the adjusted model, however, there was an interaction with sex and extraversion for the risk of death from stroke, where a higher score on extraversion was associated with a higher risk of stroke for women only. Anxiety was not associated with either MI or stroke in the final, adjusted models. Neuroticism appears to play a prominent role in the understanding of psychological risk profiles for MI for both men and women, whereas the role of extraversion was marginal and limited to stroke for women, and most likely without clinical importance. The results showed that both neuroticism and depression were associated with increased risk of death from MI, even when adjusting for common risk factors, supporting the role of depression as an independent risk marker for MI even when controlling for dispositional factors.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

The author(s) reported that there is no funding associated with the work featured in this article.

Data availability statement

The data used in this study are owned by separate public entities in Norway. The data are available upon request from the health registries of the Norwegian Institute of Public Health and the HUNT Study, respectively. Note that a Norwegian researcher must be involved with the project to apply to HUNT. For more information, please visit: https://www.ntnu.edu/hunt/data.