Gene‒environment interaction effect of hypothalamic‒pituitary‒adrenal axis gene polymorphisms and job stress on the risk of sleep disturbances

Introduction

Sleep is essential for humans, helping to maintain energy, promote growth and development, and improve immunity (Ramar et al., 2021). However, sleep disturbances seriously reduce people’s quality of life and have become a major public health problem affecting people’s physical and mental health (Halonen et al., 2017). The global prevalence of sleep disturbances is approximately 37.9% (Wu et al., 2021); in Canada, the prevalence is 23.8% (Chaput et al., 2018); in Japan, the prevalence is 13.3% (Miyachi et al., 2021); in the US, the prevalence is 30.5% (Kadier et al., 2023); in Europe, the prevalence is 25.73% (Linh et al., 2023); in Africa, the prevalence is 32.6% (Wang et al., 2019); and in China, the prevalence is 29.2% (Shi et al., 2020). Long-term sleep disturbances negatively affect people’s physical and mental health and are early risk factors for many diseases, such as cardiovascular and cerebrovascular diseases, neuropsychiatric disorders, accidental injuries and even death (Rajaratnam et al., 2011; Morin & Jarrin, 2022).

Job stress is a negative physical and psychological reaction that occurs when job requirements do not match workers’ abilities, coping resources and demands (Basu, Qayyum & Mason, 2017). In recent years, studies have shown that excessive job stress can lead to imbalances in physiological functions, resulting in decreased sleep quality and sleep problems such as insomnia and drowsiness (Khamisa et al., 2016; Herr et al., 2018; D’Ettorre et al., 2020). Therefore, job-related stress is a major occupational risk factor that significantly increases the risk of sleep disturbances (Juster & McEwen, 2015; Linton et al., 2015). Epidemiological research has indicated that job stress is related to an increased risk of sleep disturbances (Blom et al., 2020; Hämmig, 2020). A prospective cohort study of Japanese workers (n = 1,022) with a 2-year observation period also revealed that high job stress was strongly associated with insomnia, with an OR (95% CI) of 1.72 [1.06–2.79] (Ota et al., 2009). A prospective cohort study of workers in Denmark aligns with this conclusion (Nordentoft et al., 2020). In addition, sleep disturbances also seriously affect the efficiency of workers, leading to a decline in production efficiency and the occurrence of accidents, resulting in substantial social and economic burdens (Kucharczyk, Morgan & Hall, 2012; Uehli et al., 2014). It is essential to explore the mechanism underlying the influence of job stress on sleep disturbances among occupational groups and to take active measures to reduce the occurrence of sleep disturbances.

The HPA axis is thought to be the main pathway mediating the stress response (Hirotsu, Tufik & Andersen, 2015). More importantly, the HPA axis regulates the sleep-wake cycle: activation of the HPA axis may lead to awakening and insomnia in animals and humans (de Feijter et al., 2022). Dysfunction of the HPA axis at any molecule (such as the corticotropin-releasing hormone receptor, glucocorticoid receptor or mineralocorticoid receptor) may disturb sleep (Buckley & Schatzberg, 2005). When encountering stressors (physiological or psychological), the hypothalamus releases corticotropin-releasing hormone (CRH). CRH stimulates the anterior pituitary to release corticotropin, and corticotropin activates the adrenal cortex to upregulate the production of glucocorticoids (GCs). Its main function is to restore internal physiological balance after exposure to stress. However, Weitzman et al. (1983) showed that the release of GCs was related to the occurrence and development of sleep disturbances. Moreover, most stress-related hormones promote wakefulness, and elevated HPA activity appears to contribute to stress-induced insomnia (Nicolaides, Vgontzas & Kritikou, 2000). Exploring the genes that play a role in HPA axis regulation may be useful in determining the relationship between job stress and sleep disturbances. Gerritsen et al. (2017) suggested that the CRH gene is linked to stress and sleep disturbances. In addition, individual variation in the FK506 binding protein five (FKBP5) gene is related to an imbalance in the HPA axis; this imbalance has been identified as the key neurobiological mechanism underlying psychotic symptoms (Nold et al., 2022; Wang et al., 2023). An animal study also reported that FKBP5 may be a target gene for stress-induced mood and sleep disturbances (Albu et al., 2014). Although many studies have shown that sleep disturbances are related to HPA axis genes and job stress, their interaction and effect on sleep disturbances remain unclear.

In recent years, many researchers have assessed the effects of gene‒environment interactions on sleep disturbances (Zwicker, Denovan-Wright & Uher, 2018; Zhang, Khan & Rzhetsky, 2022). Both genetic (Federenko et al., 2004) and environmental factors have been shown to influence an individual’s cortisol response to stress through the HPA axis, even if the response is extreme enough to increase the risk of sleep disturbances (Foley & Kirschbaum, 2010; Kudielka & Wüst, 2010). Moreover, interactions between several genes (the glucocorticoid receptor (GR)) (Bakker et al., 2017), FKBP5 (Matosin, Halldorsdottir & Binder, 2018; Normann & Buttenschøn, 2020), 5-hydroxytryptamine transporter (5-HTTLPR) (Huang et al., 2014) and dopamine D2 receptor (DRD2) (Jiang et al., 2020) and exposure to job stress have repeatedly been found to play a role in the onset of sleep disturbances. For instance, Brummett et al. (2007) reported that the 5-HTTLPR gene polymorphism is related to sleep quality problems in individuals exposed to long-term stress. A previous study reported that the effects of early-life stress on mental illnesses such as sleep disturbances were more prominent for the G alleles of the GR genes rs258747 and rs41423247 (Lian et al., 2014). One of the largest Trier Social Stress Test (TSST) cohorts indicated that interactions among FKBP5, corticotrophin-releasing hormone receptor type 1 gene (CRHR1) gene polymorphisms and psychosocial stress may affect the cortisol response and cause circadian rhythm disruption (Mahon et al., 2013). However, there are still single nucleotide polymorphisms (SNPs) in the HPA axis that have not been fully investigated in these interactions. Most studies have focused on the effect of a single gene-stress interaction on sleep quality, and few have examined multiple major genes regulating the HPA axis to determine the relationships among gene polymorphisms, job stress, and their interaction with sleep disturbances.

Therefore, we chose SNPs of several major genes regulating the HPA axis to investigate the independent and interactive effects of HPA axis gene polymorphisms and job stress on sleep quality among front-line railway workers in Fuzhou city, China. Our investigation focused on the interaction effect of genetic and environmental factors on sleep disturbances to provide new insights for improving sleep health.

Materials and Methods

Results

Effect of the gene–environment interaction on sleep disturbances

When sex, age, ethnicity, marital status, smoking status and drinking status were controlled as covariates, the best gene‒environment interaction models were determined by GMDR analysis (Table 5). These models showed a significant effect of the interaction between HPA axis genes and job stress on sleep disturbances. The model had the maximum cross-validation consistency coefficient (10/10), and the accuracies of the training set and testing set were 0.68 and 0.60, respectively. ERI × rs1360780 × rs947008 × rs4713916 × rs110402 was considered the best interaction model because it contained the most SNPs among the models that met the best model criteria. This suggests that the best interaction model was the interaction between job stress and FKBP5 rs1360780, rs9470080, and rs4713916 genotypes and the CRHR1 rs110402 genotype. Furthermore, we also found that under job stress, the subjects with FKBP5 rs1368780-CT, rs4713916-GG, and rs9470080-CT genotypes and the CRHR1 rs110402-AA genotype had the greatest risk of sleep disturbances (Fig. 1). In addition, we analyzed the HPA axis gene‒gene interactions, and the results showed that rs1360780 × rs947008 × rs110402 was the best gene‒gene interaction model among the nine SNPs in the genes related to the HPA axis (Table S2 and Fig. S1).

Table 5:

Best gene-environment interaction models, as identified by GMDR.

Model	Training accuracy (%)	Testing accuracy (%)	Cross-validation consistency	P-value
ERI	0.58	0.56	8/10	0.17
ERI × rs1360780	0.62	0.61	10/10	0.01
ERI × rs1360780 × rs947008	0.64	0.60	4/10	<0.01
ERI × rs1360780 × rs947008 × rs110402	0.66	0.64	10/10	<0.01
ERI × rs1360780 × rs947008 × rs4713916 × rs110402	0.68	0.60	10/10	<0.01

DOI: 10.7717/peerj.17119/table-5

Note:

Adjusted for gender, age, ethnicity, marital status, smoking status and alcohol status. The best interaction model was selected based on the balance test error of the 1/10 test sample, the accuracy of the cross-validation and P-value and more SNPs included in the model, suggest that ERI × rs1360780 × rs947008 × rs4713916 × rs110402 is the best interaction model (Cross-Validation Consistency:10/10, P < 0.01). Statistically significant P value was denoted in bold.

Discussion

To our knowledge, this is the first study to investigate the associations among multiple HPA axis gene polymorphisms, job stress, and their interactions with sleep disturbances. Our study has three main findings. (a) After controlling for confounding factors such as sex, age and ethnicity, job stress was correlated with sleep disturbances. (b) FKBP5 rs1360780-T and rs4713916-A alleles and the CRHR1 rs110402-G allele were associated with the risk of sleep disturbances. However, the FKBP5 rs9470080-T allele was a protective factor against sleep disturbances. (c) GMDR analysis showed that in individuals under job stress, the risk of sleep disturbances was the highest for the FKBP5 rs1368780-CT, rs4713916-GG, and rs9470080-CT genotypes and the CRHR1 rs110402-AA genotype.

In this study, we found that the greater the level of job stress experienced, the worse the sleep quality. Consistent with previous studies, many studies have shown that high job stress is associated with a greater risk of insomnia (Deguchi et al., 2017; Yang et al., 2018; Wang et al., 2022a). In addition, overcommitment was also meaningfully and positively correlated with the PSQI score. This result also suggested that overcommitment and job stress may be related to sleep disturbances (Yoshioka et al., 2013; Wang et al., 2020). Lallukka et al. reached the same conclusion (Lallukka et al., 2014). Job stress is a very influential environmental factor for sleep (Gosling et al., 2014). There is evidence that the basal levels of cortisol are elevated in individuals experiencing job stress, and the HPA axis of people experiencing job stress may release the cortisol that causes sleep disturbances (Fogelman & Canli, 2018; Rotvig et al., 2019). In addition, Birch & Vanderheyden (2022) explored how job stress mediates stress-induced insomnia by regulating the glucocorticoid signaling pathway in brain astrocytes. This evidence suggests that job stress interferes with normal sleep and even increases the risk of sleep disturbances by activating the HPA axis.

Consistent with previous results, our study also revealed correlations between several major HPA axis regulatory genes and sleep disturbances. This result indicated that individuals with FKBP5 rs1360780-T and rs4713916-A alleles and the CRHR1 rs110402-G allele had a greater risk of sleep disturbance. This finding is in line with a study by White et al. (2012) that showed that the interaction between FKBP5 minor alleles (including rs1360780-T and rs4713916-A alleles) and emotional neglect may increase the risk of stress-related disorders such as sleep disturbances. In addition, previous studies have shown that participants with the CRHR1 rs110402-A allele had higher cortisol levels 15 min poststress, implying a risk of sleep disturbances in the future (Weeger et al., 2020; Nold et al., 2021). A meta-analysis showed that individuals exposed to stress and carrying the rs1360780-T allele and rs3800373-C allele had significantly shorter sleep durations and greater risks of stress-related diseases (Wang, Shelton & Dwivedi, 2018). Moreover, a study by Maguire et al. (2020) suggested that stress-related alterations in HPA axis genes in individuals with PTSD may contribute to sleep difficulties. We also found a protective effect of the FKBP5 rs9470080-TT genotype against sleep disturbances, which is different from the results of another study (Li et al., 2019). We believe that these inconsistent results may be caused by different types of stress or stressors. Previous findings have been based primarily on posttraumatic stress in earthquake survivors. Another possible explanation for the differences is the questionnaires and evaluation criteria used.

Our findings provide new insights into the effects of gene‒environment interactions on sleep disturbances. We found that the HPA axis gene × job stress interaction strongly affects sleep disturbances. More importantly, the GMDR results showed that individuals with the FKBP5 rs1360780-CC genotype, rs9470080-CC genotype and CRHR1 rs110402-AA genotype have the highest risk of sleep disturbances under job stress. Previous studies have also revealed effects of gene‒environment interactions on sleep disturbances. For example, Zimmermann et al. (2011) reported that individuals carrying risk alleles of two FKBP5 SNPs (rs3000377 and rs47139611) have the highest risk of reduced sleep quality if they have experienced adverse life events. Similar results were found for the interaction between childhood trauma and risk alleles of these SNPs (Bevilacqua et al., 2012). Similarly, He et al. (2019) investigated 712 participants in a large general hospital in Beijing, and the results suggested that when experiencing work-related stress, individuals with the CRHR1 rs110402-A allele may experience reduced sleep quality. In summary, our study provides evidence that the HPA axis gene × job stress interaction may play an important role in sleep disturbances. Furthermore, according to previous research, the gene × stress interaction can be explained by the diathesis-stress model (Belsky & Pluess, 2009). The model suggests that individuals with “risk-associated genes” are prone to stress-related diseases such as sleep disturbances when confronted with stress or adverse environments, while individuals with “resilient-associated genes” are not affected (Monroe & Simons, 1991; Shao et al., 2018). As diathesis-stress research has highlighted, the interaction of FKBP5 variants with trauma and adverse environments has been found to confer risk for several psychopathological phenotypes (Zannas et al., 2016). In this study, the FKBP5 rs1360780-CC and rs9470080-CC genotypes and the CRHR1 rs110402-AA genotype may be risk factors for susceptibility to stressful environments, supporting the diathesis-stress model. Therefore, to reduce the risk of sleep disturbances, individuals with genetic susceptibility should avoid or reduce job stress as much as possible.

This study has several strengths. This is the first study to examine the effects of multiple gene polymorphisms and job stress on sleep disturbances from the perspective of the HPA axis and to determine a haplotype that increases the power to detect genetic associations (Aziz et al., 2021). Haplotype analysis can assess the role of different genotypes of the target gene in susceptibility to sleep disturbances. Furthermore, GMDR was used to investigate the pattern of gene × environment interactions, as it recognizes interactions between multiple loci or environmental factors (Hou et al., 2019). However, this research still has some limitations that can be addressed in future studies. First, the evaluation of sleep disturbances was entirely based on the PSQI, and the evaluation of job stress was based exclusively on the ERI, which are subjective questionnaires that are prone to produce false positive results, which may have affected the accuracy of the results. Second, there are different sources of sample bias, including reaction bias (e.g., subjects with poor sleep quality may be more inclined to complete the study than those with good sleep quality) and sample-selection bias (e.g., first-line railway workers are apt to work long hours in stressful environments). Finally, a cross-sectional design was used; thus, we could not examine the causality of the HPA axis gene × job stress interaction in the development of sleep disturbances. In future research, longitudinal designs should be used to further study this causal relationship, as well as experimental methods to measure subjects’ sleep disturbances and job stress to provide experimental support for the results of the present study. Importantly, we will incorporate transcriptomic, epigenomic, or HPA axis activity data to conduct more in-depth studies. The results of this study suggest that individuals who carry risk alleles may be at increased risk for sleep disturbances if they are under job stress. This study provides a reliable basis for formulating strategies to reduce employees’ job stress and improve sleep quality. This study suggested that industries should pay attention to the occupationally stressful situations of their workers by reducing the incidence of occupational stress (reduction of working hours and tasks, work incentives and support) and thus reducing the incidence of sleep disturbances.

Conclusions

This is the first study to investigate the effect of the interaction between job stress and HPA axis gene polymorphisms on sleep disturbances in railway frontline workers. The present study revealed that 48.5% of workers experienced both job stress and sleep disturbances. As the main effect of sleep quality, job stress was found to increase the risk of sleep disturbances. The FKBP5 rs1360780-T and rs4713916-A alleles and the CRHR1 rs110402-G allele were also risk factors for sleep disturbances. More importantly, the GMDR results showed that the interactions of SNPs with job stress increased the risk of sleep disturbances, which is the core conclusion of our study. These findings provide new insight into the correlation between job stress and HPA axis gene polymorphisms and their interaction with sleep disturbances.

Supplemental Information