Stress-induced headache in the general working population is moderated by the NRCAM rs2300043 genotype

Background

Previous findings show that work related psychosocial stressors increase the risk of psychological distress [1], [2], [3] and a variety of health complaints [4], [5], [6], [7], [8], [9]. Pain complaints such as headache has previously been associated with psychosocial stressors [10], [11], [12]. Despite no previous studies on its relationship with headache, abusive supervision has been established as a highly detrimental psychosocial stressor. The term abusive supervision is defined as the employees’ perceptions of the supervisor’s hostile verbal and nonverbal behaviours, not including physical contact [1, 13], with prevalence rates ranging from 10 to 16% [14].

Earlier data suggest that between 50 and 90% of adults have experienced headache within the last year [15, 16]. Moreover, headache is third largest reason for years lived with disability, globally [17]. The most common primary headaches are tension-type headache (60%) and migraine (15%) [15, 18, 19], which often co-occur [20]. In addition, cervicogenic headache (17–50%) [21, 22] and other types of headache such as chronic-, medication over-use-and cluster headache are frequently reported (4%, 1–2%, and 0.2–0.3%, respectively) [15].

Many parts of the brain may be involved in the mechanisms underlying stress-induced health complaints. For example, social stress activates the hypothalamic-pituitary-adrenal (HPA) axis [23, 24], which in turn regulates circulating glucocorticoid levels [25]. Evidence exists that stress-induced anger and anxiety may be associated with activation of the amygdala, the limbic system [24], the periaqueductal gray (PAG) and the hypothalamus [26], which are all involved in behavioural responses to threatening stimuli and pain [27]. The same brain regions are also important for the influence of past emotional experience and cognitive-emotional pain modulation [26], and thereby also perceptions of aversive event such as abusive supervision.

Stress-induced processes may differ by both gender [5] and genetic factors [6]. One genetic factor, which might also affect the pituitary gland [28], could be the Neuronal Cell Adhesion Molecule (NRCAM) gene. Located at 7q31, the NRCAM encodes a protein expressed in neurons and glial cells in many brain regions. Interestingly, previous data show that the rs2300043 C allele of the NRCAM gene is overrepresented in subjects with autism [29]. Encoding a member of the immunoglobulin superfamily protein, NRCAM might be involved in organising the neural network during neural development [30, 31]. Further, the NRCAM and its protein is proposed to have a role in neuronal plasticity [32], myelination of Schwann cells and the Na⁺ channels of the nodes of Ranvier [33].

Recent data suggest that NRCAM also may be linked to emotional aspects such as negative affect [34] and somatic complaints e.g., neuropathic pain [29]. However, studies addressing the association between individual stress exposure, the NRCAM genotype and common pain states such as headache, are lacking. Hence, based on the possible link between social stress, NRCAM and central or peripheral neuronal processes, we hypothesized that the effect of abusive supervision on headache may be modified by the genetic variability in NRCAM; i.e., the rs2300043 C allele.

In the present study we demonstrate that headache in the general working population is associated with an abusive leadership and NRCAM genotype interaction.

Methods

Results

A total of 1,185, i.e., 567 (48%) men and 618 (52%) women, were successfully genotyped regarding NRCAM rs2300043 (see Tables 1 and 2). The data showed that the distribution within the male cohort was 68 (12.0%) with CC, 265 (47.0%) with CG and 234 (41.0%) with the GG. For the female cohort the distribution was 82 (13.2%), 302 (48.8%) and 234 (37.8%), respectively. The mean experienced abusive supervision scores were 0.15 (SD=0.34) for males with the CC variant, 0.17 (SD=0.38) for those with the CG variant, and 0.21 (SD=0.46) for those with the GG variant. For women the mean experienced abusive supervision scores were 0.26 (SD=0.47), 0.15 (SD=0.41) and 0.14 (SD=0.38), respectively. Males with the CC variant showed a distribution amongst the 4 categories regarding headache as follows: 44.1% ‘not bothered’, 39.7% ‘a little bothered’, 16.7% ‘considerably bothered’ and 0% ‘seriously bothered’. For males with the CG and GG variant the distribution was 51.6, 35.8, 10.1, and 48.2, 35.0, 13.6 and 2.10%, respectively. The distribution in the female cohort with the CC variant was 29.2, 52.4, 15.9 and 2.43%. The CG and GG variant showed a distribution of 33.1, 38.4, 20.5, 7.96, 41.4, 19.2 and 4.17%, respectively. No deviation from the Hardy-Weinberg equilibrium (chi²=1.0475, p=0.32) was observed.

Regarding SNP rs2300043, the results from the regression analyses of the linear associations are shown in Table 3. In men (Table 3 left, step 1) the analysis without the interaction term only revealed a trend towards a positive covariation between abusive supervision and headache (OR 1.41, CI 0.94–2.10). In men (Table 3 left, step 2), when the Abusive supervision x rs2300043 interaction was entered, the analyses showed a much stronger correlation between abusive supervision and headache (OR 6.43, CI 1.88–21.9, effects given CC and zero abusive supervision). Also, a significant Abusive supervision x rs2300043 interaction was observed, indicating a weaker relationship between abusive supervision and headache for G carriers. The analysis of the female cohort revealed a positive significant association between abusive supervision and headache (OR 1.71, CI 1.23–2.37) (Table 3 right, step 1), but no significant interactions were observed (see Figure 1 for rs2300043 data summary).

Figure 1:

The association between abusive supervision and headache.

(a) Headache in male subjects with NRCAM CC, CG and GG exposed to abusive supervision. Adjusted for age. (b) Headache in female subjects with NRCAM CC, CG and GG exposed to abusive supervision. Adjusted for age.

The three-way interaction analysis revealed that the association between abusive supervision and headache was weaker for women vs. men with CC (Table 4).

Discussion

The data of the present study supported our hypothesis that the effect of abusive supervision on headache may be moderated by the genetic variability in the gene encoding NRCAM. In particular, a very high odds of headache among men with rs2300043 CC exposed to abusive supervision was demonstrated. This suggests that the rs2300043 C allele in men may reduce tolerance of social stress in the workplace. Hence, our data support the idea that the NRCAM and its protein may be important for maintenance of adaptive neuronal processes and or responses during stress exposure.

Due to its presence in both the hypothalamus, the pituitary gland and a vast number of brain areas [28], NRCAM might be involved in the many processes related to the stress-induced neuronal activity that ultimately leads to changed sensory or emotional neuronal activity. For example, stress may activate the HPA-axis, increase circulating glucocorticoids (CORTs) [37, 38], which in turn affects the brain e.g., hippocampus [39, 40], amygdala [41] and prefrontal cortex [38, 42]. These areas are important for memory, emotions, and pain. Thus, it is tempting to speculate that exposure to repeated social stress, i.e., abusive supervision, might be associated with neural maladaptive changes in the brain.

Although the study was designed to examine the possible role of the NRCAM during exposure to social stress, the mechanism of possible maladaptive processes in different brain regions is beyond the scope of the present study. However, based on the theoretical link between NRCAM and neuronal development, plasticity, interactions with numerous other intra- and extracellular molecules [43], it seems plausible that social stress, through NRCAM could be associated with supraspinal pain processing. In addition, previous data suggest that deletion of NRCAM results in a delay in peripheral and central Na⁺channels [33].

Sodium channels are highly important in the conduction of action potentials in pain-conducting pathways, and are highly expressed in both free nerve endings and unmyelinated C-fibres [44]. If NRCAM affects conduction in the peripheral nervous system, it may indirectly affect pain processing in many brain regions. Also, with the high density and its role in the commissural fibres in the corpus callosum, the NRCAM molecule has a crucial role in connecting the two hemispheres. This in turn, may affect several processes of the higher brain centres.

Previous data suggest that rs2300043 C may be linked to psychiatric disorders in males [45]. Although the NRCAM is located on a non-sex chromosome, our data showed clear gender differences regarding rs2300043. Hence, the present findings support the idea that men and women may be different with regard to how they respond to psychosocial stressors [5, 6, 46, 47] and how they experience pain [48], [49], [50]. Recent observations from our group show that men exposed to bullying at the workplace report more pain than women in the same situation [5, 6].

Since NRCAM is expressed throughout our whole life, it is clear that this gene has an important role in both the young and adult brain [51]. Moreover, earlier observations indicate that NRCAM also in adults may affect synaptic plasticity and structural rewiring [43, 52]. Hence, the recently discovered relationship between NRCAM and psychiatric disorders [53], addiction [43], neurodegeneration [54] and neuropathic pain [29] – which also may be linked to or triggered by environmental factors – is only the beginning of our understanding of the NRCAM gene. Conclusive evidence of NRCAMs role in sensory processes and to what extent it affects the pain matrix in the brain remains to be investigated.

Conclusions

This study supports previous findings [55] on abusive supervision as a detrimental social stressor in contemporary working life and highlights the importance of organizational efforts to reduce the occurrence of such leadership behaviour. Extending previous research, the present results indicated that the association between abusive supervision and headache in men with the NRCAM rs2300043 C allele was stronger than in other men, and in women with the C allele. This suggests that carrying one or two copies of this allele may, in males, reduce the tolerance to social stress such as exposure to repeated negative acts from a superior. In contrast, the NRCAM rs2300043 C allele was not associated with increased report of headache during exposure to social stress in women; actually, a trend towards the opposite pattern was observed. We conclude that more knowledge about the NRCAM gene may be important for a more nuanced understanding of stress-induced pain such as headache.