In this study, whole-blood RNA sequencing was performed before and after the night shift in four physicians working the night shift to investigate changes in gene expression. Principal component analysis showed differences in mRNA expression patterns between the subjects. As paired tests were conducted on the same subjects before and after the night shift, and the downstream analysis was performed using only statistically significant DEGs, the influence of inter-subject differences was considered to be minimal. IPA core analysis was conducted, and canonical pathway analysis, upstream regulator analysis, and functional network analysis consistently showed that innate immune responses and inflammatory reactions were activated after the night shift. RNA bulk deconvolution also showed changes in the proportions of several immune cells. In addition, IPA analysis match revealed that gene expression patterns after the night shifts were highly correlated with several diseases, including major depressive disorder.

Relationship between shift work and inflammatory response in existing studies

Previous studies have reported increases in inflammatory markers such as CRP and IL6 after night shift work [5,6,7]. Night shifts have also been associated with changes in white blood cell counts, such as monocyte, lymphocyte, and neutrophil counts [22, 23], and these changes have been reported to be more pronounced in workers who work more frequent and consecutive night shifts [24].

Atwater et al. reported that in addition to elevated inflammatory cytokines and increased leukocytes, the level of lipopolysaccharide-binding protein was also elevated after night shifts. They concluded that shift work, sleep deprivation, stress, and circadian rhythm disturbances may increase intestinal permeability and lipopolysaccharide-binding protein by transferring lipopolysaccharide into the blood, thus contributing to an enhanced innate immune response and inflammatory response [25].

Several studies have used transcriptome analysis of peripheral blood and peripheral blood mononuclear cells in experiments in which healthy volunteers were placed in a simulated night shift environment to analyze sleep deprivation, such as that caused by shift work and jet lag. Disruption of circadian rhythms, such as through sleep deprivation and shift work, alters the expression rhythm of clock-related genes, reducing their expression amplitude and altering the expression of genes involved in innate immune responses and inflammation, such as MAL, TREM1, IL6, and STAT3 [26,27,28].

Results of the IPA core analysis and RNA bulk deconvolution

In this study, the most activated pathway in the canonical pathway analysis was the MyD88:MAL (TIRAP) cascade initiated on the plasma membrane. MyD88 is an adaptor protein that signals downstream of pattern-recognition receptor TLRs expressed on macrophages and dendritic cells and is involved in the initiation of TLR-mediated innate immune responses. Other innate immune response-related pathways, such as TLR-related pathways, are also activated.

Upstream regulator analysis also suggested that the potentially activated upstream regulators included inflammatory cytokines such as CSF2, IL2, IL6, IL1b, and IFNG; inflammation-related transcription factors such as STAT3 and HIF-1; and inflammation-related signaling factors such as p38 MAPK, indicating that upstream regulators related to the innate immune response and inflammatory response factors were activated.

Functional network analysis predicted the differentiation, maturation, and activation of phagocytes such as macrophages, dendritic cells, and antigen-presenting cells, as well as immune responses in epithelial cells, and predicted functional changes centered on innate immune responses.

In summary, IPA core analysis that included canonical pathway analysis, upstream regulator analysis, and functional network analysis consistently indicated that inflammatory responses centered on innate immune responses are elicited after the night shift.

In addition, we performed RNA bulk deconvolution to estimate the proportions of immune cells from the gene expression results and found that the proportions of immune cells changed before and after the night shift. The results of IPA core analysis and RNA bulk deconvolution were consistent with those of previous studies [22,23,24, 26,27,28].

These results showed that the changes in gene expression related to innate immune and inflammatory responses and those in immune cell proportions were observed after the night shift in actual healthcare workers working the night shift.

Relationships between shift work and other diseases according to existing studies

With respect to shift work and health risks, epidemiological studies have reported an association between shift work and metabolic diseases such as type 2 diabetes, coronary artery disease, stroke, certain types of cancer, mental disorders such as depression, and autoimmune diseases [3, 4].

A meta-analysis assessing the impact of shift work on mental health revealed that shift work was associated with a 1.28-fold increased risk of mental health deterioration in general and a 1.33-fold increased risk of depression in particular. The combined effects of circadian rhythm disturbances, sleep deprivation, and other factors are considered to be the cause [29].

Approximately one-third of depressed patients have elevated inflammatory markers even in the absence of medical illness, and patients with inflammatory diseases and those receiving cytokine therapy, such as interferon, are more likely to develop depression [30, 31]. Inflammatory mediators affect neurotransmission and neuroendocrine function and may contribute to the pathophysiology of depression [32]. These findings suggest that depression and inflammation may bidirectionally influence each other.

Results of the IPA analysis match

In this study, IPA analysis match was used to identify diseases with similar gene expression patterns to those observed after night shifts, and interestingly, major depressive disorder had the highest correlation. Other diseases showing correlations were autoimmune diseases, malignant tumors, ischemic diseases, and infectious diseases. Heatmap and hierarchical clustering analyses were performed for each disease based on z scores of the three analytical metrics. Hierarchical clustering analysis indicated that changes in gene expression after the night shift and major depressive disorder were the most closely related. The heatmaps showed similarities in the colors of the z scores related to the innate immune response and the inflammatory response. These findings suggest that gene expression changes after a night shift are correlated with major depressive disorder in terms of the innate immune response and inflammatory response.

Previous studies have shown that shift work increases the risk of depression and that depression is related to the inflammatory response. In terms of the gene expression variation in the present study, the results suggest that shift work may increase the expression of genes related to innate immune responses and inflammation and potentially increase the risk of depression.

Implications of the study findings

This study examining changes in gene expression in the blood of healthcare workers before and after the night shift work revealed that the innate immune response and inflammatory response are elicited after night shift work. In addition, gene expression changes before and after the night shift were correlated with gene expression changes in major depressive disorder and several other diseases, suggesting that shift work may affect health risks through innate immune and inflammatory responses. The findings of this study may provide a basis for future research on shift work and health risks, including major depressive disorder.

Strengths and limitations

The present study has several strengths. Relatively few studies have conducted transcriptome analyses of whole blood from shift workers. Further, all previous reports were conducted on healthy volunteers in simulated night shift or sleep deprivation environments. To the best of our knowledge, this is the first report of whole-blood transcriptome analysis and gene expression analysis conducted in actual healthcare workers working the night shift and is also the first study to analyze similarities and correlations with other diseases via IPA analysis match.

The present study also has several limitations. First, this study was exploratory in nature and included a small sample size. The small sample size and homogeneity of the participants, who were all doctors from a single institution with similar backgrounds (e.g., gender, age, race, and job descriptions), may have introduced selection bias, which potentially limits the generalizability of the findings to broader populations. The insufficient sample size also reduced the statistical power, which could lead to overinterpretation or underestimation of the results. Second, the VAS-F score used in this study to evaluate the subjects’ level of fatigue is a self-reported type of subjective evaluation, and the criteria for evaluating the degree of fatigue may differ from subject to subject. However, this study evaluated the change in each subject’s level of fatigue before and after the night shift, and the effect of differences in the way each subject felt fatigued was considered minimal. Third, this study did not assess working conditions (e.g., stress, busyness, breaks, sleeping time), so other factors may have influenced the results. Fourth, only changes in gene expression were assessed one at a time before and after the night shift, and owing to the short observation period, the long-term effects are unclear. Fifth, the analysis methods are limited. RNA sequencing and IPA were the main methods used in this study, but a more comprehensive understanding could be achieved by analyzing blood markers and combining other analysis methods. Finally, although this study described a possible association with several diseases, including major depressive disorder, on the basis of similarities in gene expression changes, no other validation has been conducted, and the actual associations are unknown. Therefore, larger, long-term studies, integrated approaches using multiple analytical methods, and intervention studies are desirable to further validate these results.