The associations between different types of infections and circadian preference and shift work

ABSTRACT Disturbed sleep and circadian disruption are reported to increase the risk of infections. People with an evening circadian preference and night workers typically report insufficient sleep, and the aims of the present study were to investigate possible associations between various types of infections and circadian preference and shift work status. Data were collected from an online cross-sectional survey of 1023 participants recruited from the Norwegian practice-based research network in general practice – PraksisNett. The participants completed questions about circadian preference (morning type, intermediate type, evening type), work schedule (day work, shift work without nights, shift work with night shifts), and whether they had experienced infections during the last three months (common cold, throat infection, ear infection, sinusitis, pneumonia/bronchitis, COVID-19, influenza-like illness, skin infection, gastrointestinal infection, urinary infection, venereal disease, eye infection). Data were analyzed with chi-square tests and logistic regression analyses with adjustment for relevant confounders (gender, age, marital status, country of birth, children living at home, and educational level). Results showed that evening types more often reported venereal disease compared to morning types (OR = 4.01, confidence interval (CI) = 1.08–14.84). None of the other infections were significantly associated with circadian preference. Shift work including nights was associated with higher odds of influenza-like illness (OR = 1.97, CI = 1.10–3.55), but none of the other infections. In conclusion, neither circadian preference nor shift work seemed to be strongly associated with risk of infections, except for venereal disease (more common in evening types) and influenza-like illness (more common in night workers). Longitudinal studies are needed for causal inferences.


Introduction
There is an interplay between sleep, circadian rhythm, and the immune system (Besedovsky et al. 2019;Irwin et al. 2016).In line with this, studies suggest that sleep of adequate duration and quality may reduce the risk of infectious diseases (Besedovsky et al. 2019; Lee and Glickman 2021;Robinson et al. 2021;Salehinejad et al. 2022).We recently reported that short sleep duration, sleep debt, and insomnia symptoms were associated with higher odds of reporting infections in a sample of Norwegian adults (Bjorvatn et al. 2023).Interestingly, for sleep debt and insomnia severity, these associations were dose dependent.The reported associations with sleep parameters were evident for many different types of infections, e.g., throat infection, ear infection, sinusitis, pneumonia/bronchitis, influenza-like illness, skin infection, and gastrointestinal infection (Bjorvatn et al. 2023).
Circadian preference is often differentiated into morning types, intermediate types, and evening types (Adan et al. 2012;Kivela et al. 2018).Morning types typically have no problem waking up early in the morning and they become sleepy relatively early in the evening, whereas evening types typically struggle with waking up early in the morning and function at their best relatively late in the day (Adan et al. 2012;Rumble et al. 2018;Zhang et al. 2015).Several studies show that evening types report more sleep problems and psychopathology than both morning and intermediate types (Kivela et al. 2018;Melo et al. 2017;Merikanto et al. 2012Merikanto et al. , 2022;;Rumble et al. 2018;Zhang et al. 2015).In a recent study conducted during the first wave of the COVID-19 pandemic, evening types reported shorter nighttime sleep duration, poorer sleep quality, more sleep onset problems, more excessive sleepiness, more fatigue, and more insomnia symptoms than morning types (Merikanto et al. 2022).Whether evening types also have higher risk of infections has not received much attention.In the mentioned study during the pandemic, evening types did not report to have had COVID-19 more often than morning types (Merikanto et al. 2022).
Shift work is associated with negative health consequences, including an increased risk of infections (Kecklund and Axelsson 2016;Rizza et al. 2021;Salehinejad et al. 2022).It is assumed that shift work, and especially night work, negatively impact the immune system, possibly through impaired sleep and circadian disruption (Besedovsky et al. 2019;Bjorvatn et al. 2020;Kecklund and Axelsson 2016;Liu et al. 2021).However, not all studies report higher prevalence of infections among shift workers.For instance, shift workers were not at increased risk of COVID-19 compared to day workers (Bjorvatn et al. 2023).However, that latter study showed that shift workers experienced more severe disease and need for hospital care, when infected by SARS-CoV-2 (Bjorvatn et al. 2023).
With this backdrop, the aims of the present study were to investigate possible associations between various infections and circadian preference (categorized as morning, intermediate, and evening types) and shift work status (regular day work, shift work without night shifts, shift work including night shifts).With the possible exception for COVID-19, our hypotheses were that evening types and shift workers would report a higher rate of infections than morning types and day workers, respectively.

Study design and participants
This study used the Norwegian practice-based research network in general practice -PraksisNett (www.praksisnett.no) to identify individuals aged 25 to 70 years to a study focusing on sleep and infections.Details are given in an earlier publication (Bjorvatn et al. 2023).In brief, general practitioners (GPs) were digitally provided with lists of eligible individuals (see below) by PraksisNett, and a total of 29 GPs sent out invitations to individuals for participation in the study.Only the GP knew the identity of the individuals.The individuals received the invitation from the GPs digitally, mostly through www.helsenorge.no(including a link to an online survey).Helsenorge.no is the official website for information about and access to health services for residents of Norway.The invited individuals responded to the invitation by clicking on the link and thereby entering the online survey (provided by Surveyxact by Ramboll).After consenting, the participants received the survey questions.Each GP was instructed to randomly invite about 20 individuals from a list generated by PraksisNett who had been diagnosed with a sleep problem during the last year (P06 in the International Classification of Primary Care, 2 nd edition -ICPC-2) and about 40 individuals from a list generated by PraksisNett without a diagnosed sleep problem during the last year.This was done to ensure that many participants actually had a sleep problem.To increase number of participants, 12 of the GPs invited individuals on two separate occasions.The invitations were sent out between March 2022 and January 2023.

Survey items
The following socio-demographic information was collected: gender (male; female; do not want to say/other), age (25-70 years, categorized as 25-40; 41-55; 56-70), marital status (single; married/cohabiting; divorced/ separated; widow/widower), country of birth (Norway; other European country; Asia; Africa; America; Oceania), children living at home (no; yes), educational level (primary school; secondary school; college/university).Furthermore, the participants responded to a question about duration of sleep problems with four response alternatives (not having a sleep problem; less than 3 months; 3 months to 1 year; more than 1 year).Use of prescribed sleep medication was reported with five response alternatives (no; sometimes; 1-2 days per week; 3-6 days per week; daily).Sick leave during the last three months due to own illness was reported as "no" or "yes." Circadian preference (chronotype) was self-reported on a five-point scale (definitively a morning type; more a morning than an evening type; neither a morning nor an evening type; more an evening than a morning type; definitely an evening type).This variable was collapsed into three categories (morning type (first two categories); neither a morning nor an evening type (middle category); evening type (last two categories)).Information about shift work was reported with three response alternatives (no; yes, but not night work; yes, including night work).
The participants reported whether they had experienced the following infections during the last three months (no; yes): common cold, throat infection, ear infection, sinusitis, pneumonia/bronchitis, COVID-19, influenza-like illness, skin infection (erysipelas, herpes labialis, etc.), gastrointestinal infection with vomit and/ or diarrhea, urinary infection (cystitis, pyelonephritis), venereal disease (chlamydia, genital herpes), and eye infection.They also reported whether the infection had resulted in a doctor's visit and whether the infection had caused sickness absence.

Ethics
The study was approved by the Regional Committee for Medical and Health Related Research Ethics (REK sør-øst, application number 268 606).Only participants who consented to participate received the survey questions.The study adhered to the ethical standards and methods outlined by Portaluppi et al. (Portaluppi et al. 2010).

Statistics
Data analyses were conducted with SPSS, version 28 (IBM SPSS Statistics).The associations between circadian preference (categorized as morning type; intermediate type; evening type) and shift work status (categorized as no; yes, but not night work; yes, including night work) and different types of infections were explored using Pearson chi-square statistics.Furthermore, logistic regression analyses, adjusted for gender, age, marital status, country of birth, children living at home and educational level, were conducted with the different infections as dependent variables (no = 0; yes = 1) and circadian preference (morning type as reference) and shift work status (not shift work as reference) as independent variables.Similar statistics were conducted for infections with a doctor's visit and for infections with sickness absence.Significance level was set to .05.

Results
Of a total of 2492 invitations sent out digitally by the GPs, 1095 individuals entered the website and responded "yes" or "no" to participation.In all, 1023 individuals consented to participate, leaving a response rate of 41.1% (1023/2492).Only 72 individuals declined to participate on the website.About 60% of the participants were female, and mean age was 48.7 (SD = 12.0) years.1).Sick leave was not associated with neither circadian preference nor shift work.
Table 2 presents the associations between the three circadian preference types and infections.Evening types reported the highest prevalence of most infections, but a statistically significant association with circadian preference was only present for influenza-like illness and venereal disease (Table 2).Restricting the statistical analyses to "infections with a doctor's visit" or "infections with sickness absence" did not lead to more significant associations with circadian preference (data not shown).
Table 3 presents the associations between shift work status and the different types of infections.Shift work, both with and without night work, was associated with increased risk of reporting influenza-like illness, but not any of the other infections (Table 3).Restricting the statistical analyses to "infections with a doctor's visit" or "infections with sickness absence" did not lead to more significant associations with shift work status (data not shown).
Table 4 shows adjusted logistic regression analyses with the different types of infection as the outcome variable.Evening types had higher odds ratio (OR) of reporting venereal disease than morning types (OR = 4.01, confidence interval (CI) = 1.08-14.84).None of the other infections, including influenza-like illness, were significantly associated with circadian preference in these adjusted logistic regression analyses.Shift work including night shifts was associated with higher odds of influenza-like illness (OR = 1.97,CI = 1.10-3.55),but none of the other infections (Table 4).

Discussion
Evening types reported higher odds for venereal disease, whereas night workers reported higher odds for influenza-like illness, compared to morning types and day workers, respectively.None of the other infections were significantly associated with circadian preference or shift work in the adjusted logistic regression analyses.Thus, our hypotheses of associations between various kinds of infections and circadian preference and shift work were not confirmed for most infections.
It is assumed that sleep of sufficient duration and quality will reduce the risk of infections (Besedovsky et al. 2019;Lee and Glickman 2021;Robinson et al. 2021), but studies are few.Since evening types and shift workers often complain of sleep problems, such as short sleep duration, poor sleep quality, and circadian misalignment (Makarem et al. 2020;Merikanto et al. 2012Merikanto et al. , 2022)), we expected an association with infections.In the same sample of Norwegian adults, we recently showed that participants with short sleep duration (<6 hours), sleep debt, and insomnia were all associated with higher odds of reporting various kinds of infections (Bjorvatn et al. 2023).The associations between infections and sleep debt and insomnia severity were dose dependent.Thus, the present results may suggest that evening types and shift workers get sufficient sleep, and thereby are not at higher risk of infectious diseases.However, our data showed that evening types did report more sleep problems and more sleep medication use than morning types, rendering such an interpretation uncertain.The present findings are novel, as no previous study has investigated the association between these common types of infections and circadian preference and shift work.
The only infection that was more common in evening types compared with morning types was venereal disease.We do not think this is related to insufficient sleep or circadian disruption per se.It is more likely related to evening types having a lifestyle where exposure to venereal diseases is higher, including being more prone to risk-taking behavior and impulsivity (Gowen et al. 2019;Kang et al. 2015;Wang and Chartrand 2015).Additionally, it is important to note that venereal disease was reported by very few participants, also among evening types, so this result should be treated with caution.The reason why none of the other infections were more common among evening types, despite evening types reporting more sleep problems, is unclear.One interpretation may be that the sleep problems experienced by evening types were not severe enough to cause an increased risk of infections.
Night workers reported to have had influenza-like illness significantly more often than day workers (45% versus 29%).Since night work is commonly associated with insufficient sleep and circadian disruption, which is shown to increase the risk of infection (Kecklund and Axelsson 2016), this was an expected finding.However, none of the other infections were more common among the night workers, questioning this interpretation.Furthermore, in our study, shift workers with or without night work did not report more sleep problems or sleep medication use than day workers.This may suggest that the shift workers in our study coped well with the work schedule, possibly explaining the lack of increase in infections.Why influenza-like illness was more common in night workers is thereby more difficult to explain.Influenza-like illness may be more sensitive to insufficient sleep and circadian disruption than other types of infections.
We did not find any associations between circadian preference, shift work status and COVID-19.This is in line with two previous studies that also reported no increased risk for COVID-19 among evening types (Merikanto et al. 2022) and shift workers (Bjorvatn et al. 2023).This may be explained by the fact that SARS-CoV-2 is very contagious, and people get infected when exposed, even if there is no sleep problem or circadian disruption (Bjorvatn et al. 2023).In the earlier study, shift/night workers suffered from more severe COVID-19 when infected (Bjorvatn et al. 2023).This is not readily supported by the findings in the present study, since restricting the statistical analyses to infections with a doctor's visit or infections with sickness absence did not change the associations with circadian preference or shift work status.
The present study had several strengths and limitations.The response rate was 41% based on how many invitations the GPs sent.However, it is uncertain if all digital invitations reached the participants.Among the 1095 participants who entered the website, only 72 declined to participate, rendering a much higher response rate.We therefore believe that the present results may be generalizable to the general adult population.We adjusted for relevant sociodemographic variables (gender, age, marital status, country of birth, children living at home, and educational level) in the regression analyses, since these variables may be associated with the risk of infections and also with circadian preference and shift work status.A major limitation was that all data were self-reported, and not supported by medical documentation.Thus, infections were not objectively assessed.Furthermore, circadian preference was based on a single question and not a validated multi-item instrument, such as the Morningness-Eveningness Questionnaire.However, using a single question for morningness/eveningness is common in epidemiological studies, and seems to estimate circadian preference fairly accurate (Bjorvatn et al. 2023;Merikanto et al. 2022).Another limitation was that we lacked detailed information about work status, such as workload (e.g., number of night shifts, hours of work per day/week).Furthermore, some of the statistical analyses were based on few individuals, calling for caution in the interpretation.Studies with larger samples are needed for replication.Recall bias (Talari and Goyal 2020), social desirability bias (Krumpal 2013), and the common method bias (Podsakoff et al. 2003) are other limitations.Importantly, the study was cross-sectional, and no causal inferences are possible.
In conclusion, this study showed that neither circadian preference nor shift work were strongly associated with infections, except for venereal disease (more common in evening types) and influenza-like illness (more common in night workers).These findings contrast with studies showing that insufficient sleep and sleep problems are associated with risk of these common types of infections, suggesting that most evening types and shift workers may get sufficient sleep in terms of infection risk.
Table 1 presents the characteristics of the study sample split by circadian preference and shift work status.Being a morning type was significantly less common in the youngest age group compared to the other age groups.Furthermore, circadian preference differed in relation to marital status, children living at home, reporting a sleep problem, and the use of sleep medication (Table 1).Morning types less commonly reported a sleep problem and use of sleep medication.With regards to shift work status, shift work (both without and with night shifts) was more common among young compared to older participants.However, shift work was not significantly associated with reporting a sleep problem or use of sleep medication (Table

Table 1 .
Characteristics of the study sample split by circadian preference types and by shift work status.
1Chi-square tests.Significant results are indicated in bold.

Table 2 .
Association between circadian preference and different types of infections.

Table 3 .
Association between work schedule and different types of infections.
χ2: Pearson chi-square; df: degrees of freedom.Significant results are indicated in bold.OR: odds ratio; CI: confidence interval; Ref: reference.Significant results are indicated in bold.