Personality traits, as operationalized by the Five Factor Model (FFM; also referred to as the “Big Five”), have been linked consistently to long-term health outcomes, including mortality (Roberts et al., 2007). Robust evidence indicates that higher conscientiousness – a tendency to be responsible, organized, and capable of self-control – is associated with lower risk of mortality (Graham et al., 2017; Jokela et al., 2013). Low scores of this trait, for example, are associated with an approximately 40% increased risk of mortality over an average of six years of follow-up (Jokela, 2013). In contrast, neuroticism – a tendency to experience more negative emotions such as fear and sadness – tends to be associated with elevated mortality risk (Graham et al., 2017), also across long follow-up periods in very old age (O’Súilleabháin & Hughes, 2018). However, protective effects for neuroticism and mortality risk have also been reported (Weiss & Costa, 2005). The evidence has been more mixed for the remaining personality traits within the FFM (e.g. Mroczek & Spiro, 2007; Ferguson & Bibby, 2012; O’Súilleabháin & Hughes, 2018). The consistent evidence that personality, particularly conscientiousness, is associated with mortality has led to great interest in identifying the pathways that explain this association.

Stemming from developmental psychology (Baltes & Goulet, 1970), the lifespan perspective is one that provides considerable avenues to view the possible ways that personality may impact mortality risk across the life course. As discussed by Hampson and Friedman (2008), both the critical period models (whereby exposure to risk during critical periods have longer-lasting effects than at other times) and accumulation models (impact of exposures to risk accumulates across the lifespan) have been well suited to examine personality and health associations. Most work on the possible pathways that contribute to the relation between personality and long-term mortality risk has focused on health-related behaviors (Friedman et al., 1995; Mroczek et al., 2009; Turiano et al., 2015; Graham et al, 2017; Turiano et al., 2012). Notably absent from this existing literature is the direct examination of underlying physiological mechanisms that may contribute to the predictive effect of personality on mortality risk.

Decades of research have established the importance of the complex structure of the immune system in health processes across the lifespan. Two markers, cytokine interleukin 6 (IL-6) and the acute phase protein C-reactive protein (CRP) are central to the immune system. Both of these biomarkers are historically thought of in the context of inflammation that involves the body’s defences in responding to actual or potential infections. However, recent research has highlighted the breadth of processes beyond inflammation that both IL-6 and CRP are related to which are also critically important to health across the lifespan (Del Giudice & Gangestad, 2018). Both of these biomarkers have been regularly implicated as critical to disease-specific and all-cause mortality (H. Li et al., 2017; Proctor et al., 2015; Y. Li et al., 2017; Emerging Risk Factors Collaboration, 2010). These biomarkers outperform traditional health risk prediction methods such as the Framingham Risk Score (DeFilippis et al., 2015) and are linked to the onset and development of chronic illnesses (Netea et al., 2017; Vasto et al., 2007).

Over the last decade, the idea that personality traits are associated with both IL-6 and CRP has received growing interest (Sutin et al., 2010; Luchetti et al., 2014). Evidence suggests that conscientiousness may have a protective role in inflammatory-related biomarkers (Luchetti et al., 2014; see also Allen et al., 2017; Sutin et al., 2018; Elliot et al., 2017; Turiano, Mroczek et al., 2013; Wagner et al., 2019). For example, Luchetti and colleagues (2014) found higher conscientiousness to be associated with lower CRP across three large, US samples (N >26,000): For each 1SD higher conscientiousness, the risk of exceeding the clinical threshold of CRP (≥3 mg/l) was lowered by 10–15% across the samples. These results were also supported by a meta-analysis of published studies on both CRP (7 studies) and IL-6 (6 studies) where conscientiousness was negatively associated with both of the biomarkers (estimated r was −.05 for CRP and −.08 for IL-6; Luchetti et al., 2014). In addition to conscientiousness, the meta-analysis found a negative correlation between openness and CRP and no significant associations for extraversion or agreeableness (Luchetti et al., 2014). Higher neuroticism has been associated with higher levels of IL-6 (Sutin et al., 2010), but this association did not replicate in the meta-analysis (see Luchetti et al., 2014). Later studies have mostly supported the association between conscientiousness and lower inflammation; the findings for the other traits remain mixed (see Allen et al., 2017; Wagner et al., 2019, Hengartner et al., 2016; Israel et al., 2014; Graham et al., 2018).

Much of the literature to date on the mechanisms between personality and mortality has examined behavioural pathways, with little research examining potential physiological pathways. Some literature has also examined the associations between personality and indices of biomarkers, including allostatic load, and have speculated that physiological dysregulation is one pathway that links personality to poor health outcomes (Stephan, Sutin, Luchetti, & Terracciano, 2016). The research that has considered the physiological pathways are either theoretical or solely examine the association between personality and specific biomarkers, not whether these biomarkers mediate the relation between personality and mortality. In this preregistered study, we addressed whether circulating levels of biomarkers could be an underlying mechanism through which personality is associated with mortality risk. Specifically, this research investigated if IL-6 and CRP provide a pathway linking personality and mortality risk over a period of 14 years. This study focused on IL-6 and CRP due to their role in age-related morbidity and mortality and their reported associations within the existing literature on personality and biomarkers. It was hypothesised that both biomarkers would mediate associations of both conscientiousness and neuroticism with mortality risks. For extraversion, openness, and agreeableness, we hypothesized that both IL-6 and CRP would not be a pathway linking them to mortality risk.

Several baseline variables had a direct effect on mortality (see Table 2). Consistent with previous analyses of earlier follow-up periods in MIDUS (Graham et al., 2017), Conscientiousness was associated lower mortality risk (HR = 0.74; p = 0.001; 95% CI, 0.61–0.87), such that each 1 SD increase in conscientiousness was associated with a 35% reduced risk of mortality. This effect was attenuated with the introduction of demographic controls (HR = 0.80; p = 0.028; 95% CI, 0.64–0.96) and the health-related factors (HR = 0.81; p = 0.038; 95% CI, 0.65 – 0.97). The association, however, was reduced to non-significance when adjusted for IL-6 and CRP in Model 4 (HR = 0.83; p = 0.080; 95% CI, 0.66 – 1.00). No other personality trait was associated with mortality. In the fully adjusted baseline model, IL-6 was associated with greater mortality risk (HR = 2.99; p = 0.010; 95% CI, 0.49 – 5.49). There was no significant association for CRP (HR = 0.91; p = 0.733; 95% CI, 0.42 – 1.40). Of the remaining predictors within the fully adjusted baseline model, age (HR = 1.10; p < 0.001; 95% CI, 1.08 – 1.13) and ADLs (HR = 1.33; p = 0.043; 95% CI, 0.96 – 1.70) were associated with greater mortality risk.

Each personality trait was examined for the potential that IL-6 and CRP may mediate between it and mortality because mediation does not require a direct effect from the predictor to outcome variable (Preacher et al., 2007). The combined indirect effects of both biomarkers were a significant indirect pathway between conscientiousness and mortality risk (Table 3). IL-6 emerged as a robust mediator of the association between conscientiousness and mortality in each model: Model 5 (p = .001), Model 6 (p = 0.028), and Model 7 (p = 0.032)¹. These significant mediation results across models accounted for an estimated 23.46%, 18.47%, and 17.65% of the effect of conscientiousness on mortality through IL-6, respectively (see Table 3). No significant indirect effect for CRP was observed. There was some evidence of an indirect pathway from neuroticism and agreeableness to mortality through IL-6 in Model 5, but neither indirect effect remained significant with the sociodemographic adjustments. There was no significant effect for extraversion or openness. See supplementary tables for all mediation results for extraversion, openness, and agreeableness. While it was not included in the preregistration of this manuscript, we also did an exploratory test for an interaction between neuroticism and conscientiousness on its association with IL-6 and mortality. No significant association emerged.

We found support for our preregistered hypotheses: Both IL-6 and CRP were an indirect path that partially linked conscientiousness to mortality risk. Examination of both IL-6 and CRP revealed that IL-6 was the significant contributor to this mediating effect. As such, higher conscientiousness was found to be associated with a longer life partially as a result of lower IL-6. Contrary to our preregistered hypothesis, we did not find that CRP itself was a significant mediator. Although there was some initial evidence that the biomarkers mediated both neuroticism and agreeableness and mortality, the indirect effects were not significant when controlling for the sociodemographic variables. Finally, as expected, the biomarkers did not mediate either extraversion or openness and mortality risk.

The conscientiousness findings are critically important because they directly link a biomarker as a pathway from a personality trait to mortality risk. The importance of IL-6 to physical and cognitive health processes across the lifespan is well established. IL-6 is one of the main inflammatory components associated with age-related pathologies (Franceschi, & Campisi, 2014), including central nervous system diseases (Erta et al., 2012), cardiovascular disease (Fontes, Rose, & Čiháková, 2015), and also in the allocation of energetic resources and nutrients relevant to adipose tissue (Del Giudice & Gangestad, 2018). Both genetic and epidemiological research point to dysregulation in the immune system as a common pathway involved in the leading causes of deaths. Similarly, conscientiousness has been associated with a large spectrum of health-related behaviours (Graham et al., 2020), age-related pathologies (Terracciano & Sutin, 2019), coping processes (Sesker et al., 2015), and weight trajectories across the lifespan (Sutin et al., 2011). The IL-6 mediated effect of conscientiousness on mortality is likely not just a function of a single or small number of specific health associations but may permeate immunity and associated physiological systems (e.g., nervous, cardiovascular, respiratory, and endocrine systems). Given its link to morbidity and mortality, IL-6 is likely to be a key modulator between conscientiousness and health outcomes.

Importantly, several avenues that link IL-6 to the etiology of disease across the life span have previously been linked to conscientiousness (Furman et al., 2019). For instance, conscientiousness has been linked to a wide array of health-related behaviours (Bogg & Roberts, 2004; Turiano et al., 2015; Sutin et al., 2011; Graham et al., 2018) with beneficial behavioral and clinical markers that are linked with IL-6, such as moderate alcohol use, less unhealthy eating, minimal drug use, and smoking, healthier weight, and more physical activity. Also outlined by Furman and colleagues (2019), IL-6 is known to have a critical role across physiological systems in response to stress. The role of personality within the context of physiological stress processes have been documented, not just as an averaged response to a single stress exposure (Gallagher et al., 2018), but also a pattern of responsivity across time during a stressful experience (O’Súilleabháin, Howard, & Hughes, 2018a), and changes in stress (O’Súilleabháin, Howard, & Hughes, 2018b). Research examining conscientiousness and the complexity of stress responsivity is quite limited. Of note, however, is that many of these behavioral, clinical and biological factors are modifiable and thus suggest pathways that can be intervened upon.

Given the role of negative emotionality in health (Lahey, 2009), we hypothesized that these immunity markers may also be a pathway between neuroticism and mortality. However, following full adjustment for possible confounding variables, no significant associations were found for neuroticism. Despite our interest in neuroticism, the null findings are perhaps not surprising given that previous studies have reported mixed evidence of associations between neuroticism and inflammatory markers (Luchetti et al., 2014). In addition, it is also apparent that the associations between neuroticism, health, and mortality are quite nuanced (see Weiss & Deary, 2020; Graham et al., 2020; O’Súilleabháin et al., 2019).

There are theoretical and practical implications of this research. Theoretical models of personality and health have typically focused on behavioral mechanisms (Turiano et al., 2015); this research identified and provided evidence for a biological pathway. Work on personality and mortality suggests that psychological factors may be helpful to identify who is at most risk for premature mortality and why. It will be important in future work to test the usefulness of personality-informed interventions to improve health outcomes, including reducing the risk of premature mortality. Indeed, while the present study had several strengths, such as an extensive follow-up period and a variety of data forms representing known predictors of all-cause mortality. Limitations should also be noted. While this study focused specifically on both IL-6 and CRP, further research is required to examine other potentially relevant pro- and anti- inflammatory cytokines, and importantly, how the present findings may interact with other cytokines, hormones, and other biomarkers as the mediating pathway is likely to be complex. The present sample is not a nationally representative sample, and as such, these findings require replication across a diverse range of populations. Additionally, while the personality measure available within the present study demonstrates adequate reliability, this study would have benefited from a more comprehensive personality scale that measured underlying personality facets. Facets would provide a more fine-grained examination of personality associations, which may be particularly important in the context of conscientiousness and biological health indices (Sutin et al., 2018). Given number of events is a critical component in terms of statistical power when computing a cox regression, a limitation with the present study design is that less than 12% of the sample had died. While there are challenges with inferring cause of death due to complexity with comorbidities at the end of life, it may be worthwhile to examine cause-specific mortality within larger samples to determine if specific causes are responsible for the association within the present study.

To conclude, our results indicate that individuals higher in conscientiousness live longer in part because of lower circulating levels of IL-6. These results provide a critical insight into biological mechanisms that link this personality trait to longevity. In doing so, we highlight the importance and need to identify biological pathways that bridge this link from personality to future mortality risk for future work. This study provides a crucial piece to the personality-health puzzle in suggesting that the biomarker IL-6, which is at the core of inflammatory and aging processes, provides a pathway which partly explains why conscientiousness is associated with long-term mortality risk.