Following the introduction of commercial assays with high sensitivity for C-reactive protein (CRP) moderately elevated levels of this compound in serum have become a well-established marker of low-grade inflammation. The finding that depression is often associated with such an increase in CRP – that has by now been corroborated in several studies 1234 – is an observation of considerable importance, as is the recent finding that a similar aberration may be observed also in subjects with panic disorder 5. If this association between CRP and psychiatric morbidity is a consequence of depression and anxiety influencing the immune system, or if low-grade inflammation may contribute to the development of depression or panic disorder, however is as yet unknown. To cast further light on why CRP is associated with depression and anxiety, and to explore the possible relationship between this marker of low-grade inflammation and other aspects of behavior, including personality traits, hence are important tasks.

The relationship between depression and panic disorder on the one hand, and elevated levels of CRP on the other, is rendered even more important given the well-established but as yet unexplained relationship between these two disorders and cardiovascular morbidity 67891011121314. The observations that enhanced CRP is an important risk factor for coronary disease 15161718192021, and that CRP promote atherosclerotic processes and endothelial cell inflammation 2223, make it tempting to suggest that low-grade inflammation may be a factor of importance for the association between depression/anxiety and cardiovascular disease.

One obvious way of enhancing our insight into the relationship between CRP and psychiatric disorders is to explore to what extent CRP is associated also with other aspects of brain function, including both other psychiatric disorders and normal personality traits. Exploring the possible relationship between personality and CRP is motivated also by the fact that not only depression and panic disorder, but also certain personality traits, have been suggested to enhance the risk for cardiovascular disease 24252627.

To this end, this study was designed to explore the possible relationship between personality traits and CRP. For the assessment of personality, the Temperament and Character Inventory (TCI) 28 was chosen, being one of the most widely used instruments for the assessment of personality in current psychiatric research. The TCI is based on a self-report questionnaire and comprises seven dimensions, two of which have been suggested to render the individual more prone to develop depression – harm avoidance (being high in subjects at risk for depression) and self-directedness (being low in these subjects) 29. The primary question prompting this study hence was to what extent there is an association between personality traits as assessed using TCI on the one hand, and CRP levels on the other, in non-depressed subjects, our à priori hypothesis being that especially those personality traits that have been shown to be associated with an enhanced risk for depression might be associated with elevated CRP. A rough estimate of ongoing or past depression was obtained by means of a brief yes/no questionnaire based on the DSM general criteria of major depression. The study provides preliminary support for the notion that high harm avoidance and low self-directedness are indeed associated with elevated CRP in middle-aged women recruited from the general population and negating ongoing depression.

The mean serum levels of CRP in the women included in the analysis of personality traits (n = 168) was 1.17 ± 1.19 mg/L (median level 0.63 mg/L). The TCI factors harm avoidance and self-directedness correlated significantly with CRP (see Table 1). After adjustment for BMI and smoking, the correlations remained significant; with respect to HA, they were strengthened. The correlations between CRP, on the one hand, and harm avoidance and self-directedness, on the other, remained significant also after exclusion of women self-reporting ongoing depression (data not shown). When comparing women confirming past but not present depression (n = 71) and women without a history of depression (n = 71), harm avoidance did not correlate with serum levels of CRP in either of the groups. Women reporting a history of depression (but denying ongoing depression) showed a significant correlation between self-directedness and CRP levels (r = -0.326, p < 0.05, corrected for multiple comparisons) whereas the CRP levels of women never being depressed did not significantly correlate with self-directedness (r = -0.209, p > 0.05). When analysing the sub-dimensions of harm avoidance, fatigability/asthenia displayed the strongest correlation with CRP. With respect to self-directedness, the sub-dimensions purposefulness and congruent second nature correlated significantly with the inflammation marker after corrections for multiple comparisons (see Table 1).

Women who reported that they suffered from an ongoing depression (n = 26) displayed significantly higher levels of CRP than women denying being depressed (n = 155): 1.69 ± 1.44 versus 1.14 ± 1.17 mg/L (p < 0.05). This difference remained significant also after exclusion of smokers (data not shown). Women confirming past but not present depression (n = 78) did not differ from women without a history of depression (previous or ongoing) (n = 77) regarding CRP levels (1.20 ± 1.13 versus 1.08 ± 1.13 mg/L); similar results were obtained also when analysing smokers and non-smokers separately.

The main finding of this paper is that the temperament trait harm avoidance was positively, and the character trait self-directedness negatively, associated with CRP levels in a population-based cohort of women. In line with this outcome, an inverse correlation between harm avoidance and self-directedness has been observed by others 39, and was confirmed by our data (not shown).

This is the first study examining CRP levels in subjects in which personality traits have been assessed using the TCI. Previous studies regarding personality traits, measured with other instruments, versus CRP levels, also are sparse. Proneness for anger and hostility has however been associated with elevated levels of CRP 4404142. Also, a study regarding fear for terror attacks revealed, in women but not in men, a positive association between fear and elevated CRP levels after adjustment for generalized anxiety, depressive symptoms and other confounding variables 43.

Although the literature is not unanimous 444546, a large number of studies suggest that ongoing depression is associated with elevated levels of CRP 12347484950. In line with this notion, women in our study reporting ongoing depression displayed elevated CRP levels. Since previous studies have shown that depression is often associated with high harm avoidance and low self-directedness state 51525354, it might be suggested that the observed association between these personality traits and CRP merely reflects an underlying depression. However, the association between personality traits and CRP was found also in women denying ongoing depression, suggesting that these personality traits per se are related to low-grade inflammation, irrespectively of an ongoing depression.

Different explanations for the relationship between high levels of CRP and depression have been put forward, including the possibility that depression influences the immune system, hence eliciting low-grade inflammation, and the alternative explanation that low-grade inflammation may cause depression. The present study, indicating that CRP may be associated with personality traits, assumed to be relatively stable throughout life, casts new light on this issue. Of interest in this context is however the fact that high harm avoidance and low self-directedness, i.e. the two traits displaying significant associations with CRP in this study, are important predictors of depression 29. One possibility would be that low-grade inflammation may be both the result of actual or perceived stress 55 – and hence associated with certain personality traits – and a cause for depression, subjects with high harm avoidance and low self-directedness being characterized by low-grade inflammation that may trigger depressive episodes.

Notably, previous studies as well as our data suggest that levels of CRP are higher during an ongoing depression than after recovery, hence suggesting CRP to be a state dependent rather than a trait dependent marker of depression 47. On the other hand, the available data do not exclude the possibility that CRP levels are slightly elevated also after recovery, and between episodes in recurrent depression 1247.

Importantly, twin studies 56 suggest the heritability of CRP levels to be considerable. This might reflect that the underlying determinants of CRP formation such as obesity or proneness to low-grade inflammation are hereditary, but may also be the result of a more direct influence of genes on CRP production. Several reports have thus shown that polymorphisms in the gene coding for CRP are associated with baseline levels of circulating CRP 5758. These polymorphisms may hence be of possible importance for the inflammatory response 59. When discussing possible explanations for the association with personality traits and depressed mood on the one hand, and CRP levels on the other, in terms of causality, the possibility that certain genes may exert a parallel influence on CRP levels and on brain development and function should not be excluded. The fact that genes are important for personality traits, including those assessed by the TCI scale 37, as well as for traits such as aggression and irritability 60, is well established.

Previous studies have suggested that the apparent association between symptoms of depression and CRP may be partly due to a confounding variable, e.g. high BMI 44. It is hence important to note that the association between CRP and the traits harm avoidance and self-directedness was not weakened by correcting for BMI or smoking. On the other hand, it should be taken into consideration that we did not control for other factors that may influence CRP levels such as diet 61, exercise 62, and socioeconomic status 63. In order to exclude individuals with an acute infection or inflammation and avoid false high levels of CRP women with CRP levels exceeding 5 mg/L were excluded. However, the acute phase response is a continuum and no true normal range for CRP levels is established. It has been suggested that levels below approximately 3 mg/L represent "normal" values and CRP levels above 10 mg/L reflect a clinically significant inflammatory state 3564. Since blood samples were obtained once only, it cannot be excluded that high levels in some subjects were the results of an ongoing infection or inflammation.

CRP being associated with certain personality traits, and, at the same time, being an important risk factor for cardiovascular disease, makes it tempting to suggest that it may serve as a mediator between these two variables. The TCI sub-dimensions that most clearly correlated with CRP, i.e. fatigability/asthenia and purposefulness, may be factors predisposing individuals to burnout, a state characterized by emotional exhaustion and feelings of inadequate control over one's job [for refs, see 65]. Burnout has been associated with microinflammation 66 as well as with cardiovascular disease 6768 also after controlling for depression and other confounding variables. Interestingly, CRP has been suggested to directly influence atherosclerosis 2223. A speculative interpretation of our study would be that women being confident, optimistic and self-sufficient, by displaying lower levels of CRP, also may exhibit a decreased risk for cardiovascular disease. This suggestion is well in line with previous studies 697071.

Limitations of this study are that the number of assessed subjects was relatively small, and that the presented associations, though significant, were modest in term of r-values. The population was recruited by means of population register, but the fact that the participants had accepted to participate may render a group somewhat non-representative for the general population. A strength of the study is that all studied subjects were of the same gender and age, hence reducing age- and sex-related sources of variability, and that all blood samples taken for analyses of CRP levels were obtained in the same (follicular) phase of the menstrual cycle 72. At the same time, the homogeneity of the group with respect to sex and age obviously limits the possibility to draw more generalised conclusions from the results, and makes further studies investigating the relationship between levels of CRP and personality traits in women at other ages and in men highly warranted. Another important limitation of the study is that depression was not assessed using a structured interview, but only by means of a few questions. Previous studies however suggest that very brief depression questionnaires, comprising only a couple of questions, such as the one used in this trial, do display an acceptable sensitivity 7374; there are hence reasons to believe that the observation of a relationship between personality traits and CRP in the non-depressed group is not largely influenced by the erroneous inclusion of depressed subjects in this analysis. Also, the self-reported rate of depression in this female population is well in line with other reports 75. This notwithstanding, the results should be regarded as preliminary until replicated in women evaluated by a more comprehensive depression questionnaire, or, preferably, by a structured interview.

In this study, the personality trait harm avoidance was positively, and the character trait self-directedness negatively, associated with serum CRP levels in women. Previously published studies 76 have suggested that type D personality, described as a synergy between negative affectivity and social inhibition, is an important risk factor for morbidity and mortality in patients suffering from cardiovascular disease. In this respect it is tempting to suggest that one link between psychological distress and cardiovascular disease may be raised levels of CRP. The underlying mechanism behind the association between brain functions and peripheral low-grade inflammation, also demonstrated in ongoing depression, however remains unknown.

The author(s) declare that they have no competing interests.

SH performed the statistical analyses and has been involved in drafting the manuscript. FB, RR and GH are performing the longitudinal study with respect to women's health in which the subjects of this study are included. ML and HA performed the personality traits investigations and psychiatric evaluations. AE carried out the immunoassay, performed the analysis and interpretation of the statistical data together with SH and drafted the manuscript. All authors read, provided comments and approved the final manuscript.