Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2026 Jul 17.
Published in final edited form as: Chronobiol Int. 2025 Apr 30;42(5):561–571. doi: 10.1080/07420528.2025.2493208

Epigenetic markers of adverse lifestyle identified among evening and night shift workers in two UK population-based studies: Generation Scotland and Understanding Society

Paige M Hulls a,b, Daniel L McCartney c, Yanchun Bao d, Rosie M Walker e, Frank de Vocht a,f,g, Richard M Martin a,b,h, Caroline L Relton a,b, Kathryn L Evans c, Meena Kumari i, Riccardo E Marioni c, Rebecca C Richmond a,b,✉
PMCID: PMC7619246  EMSID: EMS214943  PMID: 40304317

Abstract

Epigenetic changes in the form of DNA methylation (DNAm) may act as biological markers of risk factors or adverse health states. In two cohort studies, Generation Scotland (GS) (n = 7,028) and Understanding Society (UKHLS) (n = 1,175), we evaluated associations between evening or night shift work and four lifestyle factors (body mass index, smoking, alcohol, education) through linear regression using both conventionally measured phenotypes and DNA methylation-based scores proxying the phenotypes. DNA methylation-based measures of biological ageing were also generated using six established “epigenetic clocks.” Meta-analysis of GS and UKHLS results was conducted using inverse-variance weighted fixed effects. Evening/night shift work was associated with higher BMI (0.79; 95%CI 0.02, 1.56; p = 0.04) and lower education (− 0.18; − 0.30, − 0.07; p = 0.002). There was weak evidence of association between evening/night shift work and DNAm scores for smoking (0.06, − 0.03, 0.15; p = 0.18) and education (− 0.24; − 0.49, 0.01; p = 0.06) in fully adjusted models (adjusted for age, sex, methylation principal components and phenotypic measures). Two of the epigenetic age measures demonstrated higher age acceleration among evening/night shift workers (0.80; 0.42, 1.18; p < 0.001 for GrimAge and 0.46; 0.00, 0.92; p = 0.05 for PhenoAge). In over 8,000 participants from two cohort studies, evening/night shift work was associated with both phenotypic and DNA methylation-based measures of higher BMI and lower education. DNAm predictors of smoking and ageing were also related to evening/night shift work. Epigenetic measures may provide insights into the health and lifestyle profiles of night shift workers.

Keywords: Shift work, smoking, BMI, alcohol, education, biomarkers, DNA methylation, understanding society, generation Scotland

Introduction

Shift work has been referred to as “a work activity scheduled outside standard daytime hours, where there may be a handover of duty from one individual or work group to another” (Health and Safety Executive 2006). Typically, shift work has been associated with industries that require 24-hour operation, such as essential public services or for practical purposes. In recent years there has been an increase in the number of shift workers in other industries (Health and Safety Executive 2006) with approximately 19% of the working population engaged in shift work as their main job (NHS Digital 2013) in the UK.

It has been argued that the introduction of shift work within wider industries does not consider the health and wellbeing costs to the individual shift worker (Folkard and Tucker 2003). Much of the research that investigates the health impacts of shift work has previously centred around circadian disruption particularly among night shift workers, which can result in disturbed sleep and excessive sleepiness during the work shift (Wright et al. 2013). However, adverse health behaviours have been found among shift workers which also put them at higher risk of disease (Dorrian et al. 2017; Loef et al. 2018; Suwazono et al. 2008) and shift work has been associated with higher risk of diseases, for example, type 2 diabetes, cardiovascular disease, stroke and cancer (Kecklund and Axelsson 2016), than non-shift workers.

Recent studies have investigated epigenetic changes in the form of circulating DNA methylation (DNAm) as an objective measure for evaluating the potential health impact of shift work (White et al. 2019). While these studies have typically focused on assessing individual sites in the genome (cytosine-phosphate-guanine “CpG” sites), recently DNAm scores derived from methylation levels at numerous CpG sites across the epigenome have been developed which can act as proxies for lifestyle exposures and may predict health outcomes (McCartney et al. 2018). Self-reported health behaviours (especially one-off measures) are subject to measurement error and bias (Prince et al. 2020), and using more objective measures of long-term exposure could identify more robust associations with shift work.

One group of DNAm scores aims to capture the epigenetic clock. DNA methylation age (DNAm Age) has been derived to provide an accurate estimate of biological age across a range of tissues, and at different life stages (Fransquet et al. 2019). Estimated DNAm Age exceeding true chronological age is known as “age acceleration” and studies suggest that DNAm age acceleration is associated with age-related health outcomes independent of chronological age (Marioni et al. 2015). As well as the clocks developed based on age, more recent “second generation” epigenetic clocks have been developed based on lifestyle factors and biomarkers which have been found to be highly predictive of both health and lifespan (Levine et al. 2018; Lu et al. 2019). Other DNAm scores which predict modifiable health, lifestyle and socio-economic factors include scores developed for alcohol consumption, smoking status, BMI and education (McCartney et al. 2018).

This study aimed to investigate associations of evening/night shift work participation and a series of blood based DNAm predictors of ageing, BMI, smoking, alcohol and education within the Generation Scotland (GS) and Understanding Society (UKHLS) studies.

Materials and Methods

Generation Scotland (GS)

The Generation Scotland: Scottish Family Health Study is a prospective cohort study comprising participants from the general population across five regions of Scotland. The recruitment protocol and cohort characteristics are described in detail elsewhere (Smith et al. 2013) and in the Supplementary Methods.

Blood DNAm was profiled using the Infinium MethylationEPIC BeadChip (Illumina Inc.) in two sample sets from GS: Set 1 and Set 2 (Supplementary Methods).

Methylation data were available for 777,193 CpGs measured in Set 1 (n = 2,578 unrelated individuals) and 773,860 CpGs measured in Set 2 (n = 4,450 unrelated individuals). The data sets had DNAm profiled at separate time points and quality control and normalisation was carried out separately. Details on the lifestyle variables and how they were derived can be found in the Supplementary Methods.

Understanding Society (UKHLS)

The UK Household Longitudinal Study (UKHLS) (also known as Understanding Society) is a longitudinal panel survey of 40 000 UK households from England, Scotland, Wales and Northern Ireland. The recruitment protocol and cohort characteristics are described in detail elsewhere (Chen et al. 2016; Hannum et al. 2013) and in the Supplementary Methods.

Blood DNAm was profiled using the same Infinium MethylationEPIC BeadChip (Illumina Inc.) as used by GS. Methylation data was available for 837,487 CpGs in 1,175 individuals (Supplementary Methods). Details on the lifestyle variables and how they were derived can be found in the Supplementary Methods.

Evening/Night Shift Work

In both Generation Scotland and UKHLS we defined evening/night shift work based on work reported outside normal daytime working hours (typically 9:00 AM-5:00 PM).

In Generation Scotland, participants were asked at baseline (when the samples for DNAm profiling were obtained) how many hours in a typical week they would work in the evening or overnight, from which a binary variable for “evening/night shift work” was generated based on those working ≥20 hours/week compared with those reporting to work 0 hours in the evening or overnight (i.e. excluding 1–19 hours of evening/night shift work).

In UKHLS, participants were asked to describe the times of day they usually worked at multiple time points, with those who reported working evenings or at night classed as “evenings/night shift workers.” Current evening/night shift work was defined based on reported shifts at the timepoint when the samples for DNAm profiling were obtained. More details of the evening/night shift work variables and how they were derived can be found in the Supplementary Methods.

DNA Methylation Scores

We derived four DNAm scores related to BMI, smoking, alcohol consumption and education which were based on CpG sites identified in a previous study (McCartney et al. 2018). Details of the DNAm scores are shown in Table 1. For each individual, DNAm scores were calculated as the sum of methylation values at each CpG multiplied by the effect sizes obtained in a previous study (McCartney et al. 2018).

Table 1. Origins of lifestyle DNA methylation scores employed in the current analysis.

Phenotype DNA methylation scores Original publication No. of CpG sites
Alcohol Alcohol Epigenetic prediction of complex traits and death (McCartney et al. 2018) 450
Smoking Smoking Epigenetic prediction of complex traits and death (McCartney et al. 2018) 233
BMI BMI Epigenetic prediction of complex traits and death (McCartney et al. 2018) 1109
Education Education Epigenetic prediction of complex traits and death (McCartney et al. 2018) 373
Ageing AgeAccelHorvath DNA methylation age of human tissues and cell types (Horvath 2013) 353
IEAA DNA methylation age of human tissues and cell types (Horvath 2013) 353
AgeAccelHannum Genome-wide methylation profiles reveal quantitative views of human ageing rates (Hannum et al. 2013) 71
EEAA DNA methylation-based measures of biological age: meta-analysis predicting time to death (Chen et al. 2016) 71
AgeAccelPheno An epigenetic biomarker of ageing for lifespan and healthspan (Levine et al. 2018) 513
AgeAccelGrim DNA methylation GrimAge strongly predicts lifespan and healthspan (Lu et al. 2019) 1,030
Open in a new tab

CpG: cytosine-phosphate-guanine; IEAA: Intrinsic epigenetic age acceleration; EEAA: Extrinsic epigenetic age acceleration.

We also derived six epigenetic biomarkers of ageing using previously published approaches (Chen et al. 2016; Hannum et al. 2013; Horvath 2013; Levine et al. 2018). (Supplementary Methods). In each case, age acceleration was defined as the residual obtained from regressing predicted age, as estimated by the epigenetic clock, on chronological age. This measure of age acceleration is independent of chronological age.

All methylation scores were standardized (mean = 0, standard deviation = 1) in both GS and UKHLS.

Statistical Analyses

We first assessed whether lifestyle factors were associated with evening/night shift work. We performed linear regression of each phenotype (as the outcome variable) in relation to evening/night shift work (as the exposure variable). This was with the exception for education, which was treated as the exposure variable given that education precedes evening/night shift work. In GS, two models were run: Model 1 with adjustment for age and sex, and Model 2 with additional adjustment for other self-reported phenotypes (e.g. for smoking, the model was also adjusted for BMI, alcohol, and education). In UKHLS, two models were run: Model 1 with adjustment for age, sex, blood processing day and batch and Model 2 with additional adjustment for other self-reported phenotypes.

We subsequently assessed whether the lifestyle factors could be proxied with DNAm scores within GS and UKHLS. We performed linear regression of the phenotypes (exposure variable) and lifestyle DNAm scores for alcohol, smoking, education, and BMI (outcome variable). For these models, a logistic regression was performed. In GS, two models were run: Model 1 with adjustment for age, sex and 20 methylation principal components (PCs) – the latter are used to identify and remove unwanted sources of variation (e.g. batch effects). Model 2 was additionally adjusted for other self-reported phenotypes (e.g. for smoking related DNAm scores, the model was also adjusted for BMI, alcohol, and education). In UKHLS, two models were run: Model 1 with adjustment for age, sex, blood processing day and batch; and Model 2 with additional adjustment for other self-reported phenotypes.

We finally assessed whether evening/night shift work was related to the lifestyle DNAm scores. We performed linear regression of each lifestyle methylation score (outcome) in relation to evening/night shift work (exposure). This again was with the exception for education, which was treated as the exposure. In GS, three models were run: Model 1 with adjustment for age, sex and 20 methylation PCs; Model 2 had additional adjustments for other self-reported phenotypes (e.g. for smoking related DNAm scores, model adjusted for BMI, alcohol, and education); and Model 3 with further adjustment for the corresponding phenotype (e.g. for smoking-related DNAm scores, the model was also adjusted for smoking). This was performed to examine whether the DNAm score was associated with shift work independently of the corresponding phenotype.

In UKHLS, three models were run: Model 1 with adjustment for age, sex, blood processing day and batch; Model 2 with additional adjustment of other self-reported phenotypes; and Model 3 with further adjustment for the corresponding phenotype.

We assessed whether evening/night shift work was related to any of the six epigenetic ageing measures. We performed linear regression of each epigenetic age acceleration (EEA) measure (as the outcome variable) in relation to evening/night shift work.

In GS, two models were run: Model 1 with adjustment for sex and 20 methylation PCs and Model 2 with additional adjustment for smoking, BMI, education, and alcohol. In UKHLS, two models were run: Model 1 with adjustment for sex, blood processing day and batch; and Model 2 with additional adjustment for smoking, BMI, education, and alcohol.

We conducted a series of meta-analyses of GS Sets 1 and 2 and UKHLS using an inverse-variance weighted fixed effects approach. For this, we used the binary measure of evening/night shift work derived in GS and the current evening/night shift work variable derived in UKHLS. The I2 statistic was used to assess heterogeneity across the studies.

We conducted meta-analyses of: i) the self-reported or conventionally measured (in the case of BMI) lifestyle factors in relation to evening/night shift work, ii) the lifestyle DNAm scores in relation to evening/night shift work and iii) the epigenetic age acceleration measures in relation to evening/night shift work.

Results

Baseline Characteristics

Summary characteristics of the participants from GS and UKHLS are presented in Table 2. 7,028 individuals with DNAm data were included from GS (n = 2,578 in Set 1 and n = 4,450 in Set 2) and 1,175 individuals from UKHLS. Set 1 and 2 of GS had a younger mean age (50.0 ± 12.5 and 51.4 ± 13.2) compared to UKHLS (58.0 ± 15.0). There was a comparable balance of men and women across the three datasets: 38.6% males in GS Set 1, 43.7% in GS Set 2 and 41.6% in UKHLS. BMI was also broadly comparable: 27.4 ± 5.5 kg/m2, 26.8 ± 5.0 kg/m2 and 28.1 ± 6.2 kg/m2, in GS Set 1, GS Set 2 and UKHLS, respectively. GS Set 1 had more current smokers whilst UKHLS had more former smokers and GS Set 2 had more never smokers. There was a higher proportion of daily drinkers in UKHLS (16.0%) compared with GS (12.4% in Set 1 and 13.2% in Set 2). There was also a higher proportion of less than monthly drinkers in UKHLS (25.2%) compared with GS (17.2% in Set 1 and 16.2% in Set 2). Years of full-time education was comparable between Set 1 and 2 of GS (13.6 ± 3.4 and 13.8 ± 3.4, respectively) whilst in UKHLS it was slightly lower (12.3 ± 5.1). In UKHLS, 1.6% of participants (n = 18) were currently working evening/night shifts while 8.8% (n = 103) had reported working evening/night shifts over the previous 11 years. In GS Set 1, 8.1% (n = 127) of participants reported working at evening/night for >20 hours per week at the time of sampling, compared with 7.9% (n = 193) in GS Set 2.

Table 2. Baseline characteristics of the participants in Generation Scotland and Understanding Society (UKHLS).

Variables Categories Generation Scotland Understanding Society (UKHLS)
Set 1 (n = 2578) Set 2 (n = 4450) Overall (n = 1175)
Mean SD Mean SD Mean SD
Age (years) 50.0 12.5 51.4 13.2 57.9 15.0
Body mass index (kg/m2) 27.2 5.4 26.8 4.9 28.1 6.2
Education (years of full-time education) 13.6 3.4 13.8 3.4 12.3 5.1
N % N % N %
Sex Male 1583 38.6 1944 43.7 489 41.6
Female 995 61.4 2506 56.3 686 58.4
Smoking status* Current 460 18.3 675 15.5 186 15.9
Former 771 30.7 1398 32.1 487 41.6
Never 1276 50.9 2276 52.3 498 42.5
Alcohol* Daily 167 12.4 334 13.2 164 16.0
More than weekly 728 54 1402 55.6 475 43.6
More than monthly 222 16.5 386 15.2 166 15.2
Less than monthly 231 17.2 411 16.2 275 25.2
Evening/night shift work* 0 hours per week 1015 64.9 1592 65.6 Never 1072 91.2
1–19 hours per week 422 27.0 643 26.5 Ever 103 8.8
>20 hours per week 127 8.1 193 7.9 Previous 79 6.9
Current 18 1.6
Open in a new tab
*

N in categories does not total overall sample size due to missing data in phenotypes; complete case analysis was performed meaning model N varies depending on phenotype, as indicated in Supplementary Tables.

Lifestyle Factors

How Does Evening/Night Shift Work Relate to Conventionally Measured Lifestyle Factors?

In GS, evening/night shift work was inversely associated with alcohol frequency in both GS Set 1 and GS Set 2 (Model 1: Effect= − 0.056; 95% CI − 0.095, −0.017; p = 0.006 category change per hour for GS Set 1, and − 0.039; − 0.072, − 0.006; p = 0.019 for GS Set 2), although associations attenuated with adjustment for the other self-reported phenotypes in GS Set 2 (Model 2: − 0.024; − 0.057, 0.009; p = 0.151) (Table S1). Evening/night shift work was positively associated with smoking status in GS Set 1 (Model 1: 0.024; 0.006, 0.042; p = 0.009 category change per hour) but were more weakly associated in GS Set 2 (Model 2: 0.012; −0.006, 0.030; p = 0.216), and when adjusted for the other phenotypes (Table S1). There were positive associations between evening/night shift work and BMI in both GS Set 1 (Model 1: 0.15; 0.03, 0.28; p = 0.02 kg/m2 per hour) and GS Set 2 (Model 2: 0.23; 0.14, 0.32; p = 1 × 10−6), although these associations attenuated when adjusted for the other phenotypes. There was an inverse association between evening/night shift work and education in GS Set 2 (Model 1: −0.07; −0.13, −0.02; p = 0.01 hour per year) which was less apparent in GS Set 1 (Model 1: −0.05; −0.12, 0.03; p = 0.21), although there was stronger evidence of association of the binary evening/night shift work variable with education, based on in both datasets (Table S1).

In UKHLS, no strong associations were found between current, ever, or previous evening/night shift work and any of the lifestyle phenotypes, although effect estimates were typically in the same direction as in GS (Table S2).

In the GS-UKHLS combined meta-analysis there was evidence of a positive association between evening/night shift work and BMI (0.79; 0.02, 1.56; p = 0.04 kg/m2 difference between those who did and did not work evening/night shifts) and an inverse association between evening/night shift work and education (−0.18; −0.30, −0.07; p = 0.002 log odds per year of education) in the fully adjusted models (Model 2) (Figure 1). There was little evidence for an association between evening/night shift work and either alcohol intake or smoking status in a meta-analysis across the three datasets (0.00; −0.19, 0.20; p = 0.97 and 0.04; −0.19, 0.27; p = 0.73 category change between those who did and did not work evening/night shifts, respectively). There was little evidence for heterogeneity between the study estimates (I2 < 5%). Results of the minimally adjusted models (Model 1) are shown in Figure S1).

Figure 1. Associations between evening/night shift work and lifestyle phenotypes in Generation Scotland (GS) and Understanding Society (UKHLS).

Figure 1

Open in a new tab

*Model 2: adjusted for age, sex, other self-reported phenotypes (e.g. for smoking, model adjusted for alcohol, body mass index and education). For these models, education was treated as the exposure and shift work was the outcome, and effect estimates are log odds ratios.

How are DNA Methylation Biomarkers Associated with Lifestyle Factors in GS and UKHLS?

There were positive associations between each DNAm score and its respective lifestyle phenotype in both GS and UKHLS (Table S3 and S4, respectively).

Is Evening/Night Shift Work Associated with DNA Methylation Biomarkers?

In GS, there were no clear associations between evening/night shift work and the alcohol DNAm score (Table S5). Evening/night hours were positively associated with the smoking DNAm score in both GS Set 1 and GS Set 2 (Model 1: 0.04 SD; 0.02, 0.06; p = 1.28 × 10−4 and 0.02 SD; 0.00, 0.04; p = 0.01), although associations attenuated with further covariate adjustment in Models 2 and 3. Number of evening/night hours was also positively associated with the BMI DNAm score in both GS Set 1 and GS Set 2 (Model 1: 0.03 SD; 0.01, 0.05; p = 0.01 and 0.03 SD; 0.01, 0.04; p = 0.04). Associations attenuated with further covariate adjustment in Models 2 and 3 (Table S5). Number of evening/night hours was inversely associated with the education DNAm score in GS Set 1 (Model 1: −0.20 hours; −0.33, −0.08; p = 0.001) and more weakly in GS Set 2 (Model 1: −0.09 hours; −0.19, 0.01; p = 0.08). With further covariate adjustment, the association persisted in GS Set 1 but was attenuated in GS Set 2 (Table S5).

None of the evening/night shift work measures were strongly associated with the lifestyle DNAm scores when UKHLS was assessed independently (Table S6). When estimates from GS were combined in a meta-analysis with UKHLS, there was little evidence of evening/night shift work associations with the alcohol and BMI DNAm scores in the fully adjusted models (Model 3) (Figure 2). However, there was some evidence of an inverse association between evening/night shift work and the education DNAm score in all three datasets (−0.24; −0.49, 0.01; p = 0.06 log odds per year) (Figure 2). There was also weak evidence for a positive association between evening/night shift work and the smoking DNAm score (0.06 SD, −0.03, 0.15; p = 0.18). There was little evidence for heterogeneity between the study estimates in the meta-analysis (I2 = 0%). Results of the minimally adjusted models (Models 1 and 2) are shown in Figures S2 and S3.

Figure 2. Associations between evening/night shift work and lifestyle methylation scores in Generation Scotland (GS) and Understanding Society (UKHLS).

Figure 2

Open in a new tab

*Model 3: adjusted for age, sex, blood processing day, rack barcode, corresponding other self-reported phenotypes (e.g. for smoking DNAm, model adjusted for smoking, alcohol, body mass index and education). For these models, education was treated as the exposure and shift work was the outcome, and effect estimates are log odds ratios

Epigenetic Ageing

There was evidence of an association between evening/night shift work and GrimAge in GS, but not for the other epigenetic clocks. For GrimAge, the number of evening/night hours was associated with higher age acceleration in both GS Set 1 (Model 1: 0.19 years, 0.11, 0.27; p = 1.18 × 10−5) and GS Set 2 (Model 1: 0.10 years, 0.04, 0.16; p = 9.45 × 10−4) (Table S7).

Associations with GrimAge acceleration remained, although were partially attenuated on adjustment for smoking, BMI, education and alcohol (Table S7). None of the evening/night shift work measures were strongly associated with epigenetic age acceleration in UKHLS (Table S8).

In the GS-UKHLS meta-analysis, evening/night shift work was associated with a 0.80 year (0.42, 1.18; p < 0.001) increase in GrimAge acceleration (Figure 3). There was also weak evidence of association with PhenoAge acceleration (0.46 years; 0.00, 0.92; p = 0.05). The other four epigenetic clocks showed limited evidence of association. There was low heterogeneity between the study estimates in the meta-analysis (I2 < 50%).

Figure 3. Associations between evening/night shift work and epigenetic age acceleration in Generation Scotland (GS) and Understanding Society (UKHLS).

Figure 3

Open in a new tab

Model 2: Adjusted for sex, 20 methylation PCs, smoking, alcohol, body mass index and education. For these models, education was treated as the exposure and shift work was the outcome, and effect estimates are log odds ratios

Discussion

We conducted analyses to investigate associations between evening/night shift work and both phenotypic and DNAm markers in two cohorts. When we assessed phenotypic traits, we found that evening/night shift work was associated with higher BMI and lower education. When assessing DNAm predictors of the same traits, there was similar evidence of association of evening/night shift work with BMI and education DNAm scores. While the association with the BMI score was attenuated after adjusting for the corresponding phenotype, evening/night shift work was nominally associated with education and smoking in fully adjusted models. Furthermore, two of the epigenetic age measures, GrimAge and PhenoAge, demonstrated higher age acceleration among evening/night shift workers.

The observational associations of evening/night shift work with lower education and higher BMI have been previously reported with comparable effect sizes (Daghlas et al. 2021; van Drongelen et al. 2011). While we did not find evidence of a phenotypic association with reported smoking, the association between evening/night shift work and the smoking methylation score is consistent with previous studies reporting that smoking behaviour is more common with shift work comparison to day work (Bae et al. 2017; Boini et al. 2022; Bushnell et al. 2010; Knutsson and Bøggild 2010; Knutsson and Nilsson 1998; Knutsson et al. 1988). One study also found that shift workers were also more likely to start smoking in comparison to their counterpart day workers (Cho et al. 2019; van Amelsvoort et al. 2006). We also found little evidence of an association between evening/night shift work and either self-reported or methylation measures of alcohol; however, previous research has found that shift workers were more likely to drink heavily (Dorrian et al. 2017).

There is growing evidence to suggest that DNAm-based measures are useful for health and lifestyle profiling (Hillary and Marioni 2021). Furthermore, several studies have shown that methylation predictors can provide a more accurate measurement of exposure than those based on self-report (Pośpiech et al. 2024; Zhang et al. 2016). For example, previous studies have shown that smoking methylation scores may provide a more accurate measure of true exposure compared with self-reported smoking (Lu et al. 2019; Zhang et al. 2016), possibly due to erroneous self-reporting, the broad categories for reporting exposure, or because DNAm is able to capture long-term biological changes as a result of smoking as well as secondary smoking. This is supported in part by our findings that the education and smoking scores were still weakly associated with night shift work even after adjusting for the corresponding self-reported exposures.

With respect to evening/night shift work, a number of studies have reported associations with individual CpG sites through epigenome-wide association analysis (EWAS) (Wackers et al. 2023), but few have looked at the relationship between evening/night shift work and DNAm profiles which capture markers of ageing and lifestyle exposure. Some small-scale studies have also reported epigenetic age acceleration in evening/night shift workers (n < 200) (Carugno et al. 2021; Wackers et al. 2023). White et al. (2019) investigated associations between epigenetic ageing in >2000 participants from the Sister Study (females only) and found that the length of time working night shifts was associated with increased PhenoAge, although GrimAge was not investigated.

Circadian oscillators have been found to contribute to epigenetic ageing (Koncevičius et al. 2024; Oh et al. 2018) and there is emerging evidence that DNAm age estimators relate to circadian rhythm (Horvath and Raj 2018). However, is should be noted that we did not find that evening/night shift work was associated with the Hannum and Horvath clocks, which were designed to estimate chronological age (Hannum et al. 2013; Horvath 2013). This absence of association is in accordance with findings from White et al. (2019) and suggests that intrinsic circadian processes are unlikely to underlie the associations observed (Levine et al. 2018; Lu et al. 2019).

Strengths and Limitations

One of the main strengths of the study is the use of GS, an epidemiological cohort study with a large sample size with DNAm data which has also captured information on evening/night shift work. Meta-analysing associations with those from the UKHLS dataset also improved power to detect associations between evening/night shift work, lifestyle factors and DNAm predictors and indicated consistency in associations across studies. Furthermore, we have the use of both self-reported and DNAm markers for the same exposure, so were able to directly compare.

There were some limitations to the study. The DNAm scores used in this study were developed in GS Set 1, which might lead to overfitting of the models. However, by using the second set of participants in GS and the independent UKHLS datasets, we hope that this should minimise any potential overfitting issues. There was also limited evidence for heterogeneity in the associations observed between GS set 1 and the other studies.

We specifically assessed evening/night shift work rather than other forms of shift work (such as morning or rotating shift work) in relation to DNA methylation. This is because of the previous evidence suggesting that work outside of normal daytime hours is likely to be particularly disruptive to biological processes and to have implications for adverse health (Kecklund and Axelsson 2016).

We acknowledge that the definition we have used to classify shift workers in our study differs from that used by the International Agency for Research on Cancer (IARC) Monographs to classify night shift work as “work occurring during the regular sleeping hours of the general population” (Ward et al. 2019). However, we were unable to separate evening from night shift work in the Generation Scotland study based on the question asked and so to best harmonise the data sources, we opted to combined those individuals working evening or at night in UKHLS as well. While this may have obscured some associations which might be more evident among pure night shift workers (e.g. those working between 11 pm and 7am), our definition captures a more diverse group of workers, which might lead to more generalizable findings across industries with varying shift schedules.

Another limitation was the lack of data on the intensity of evening/night shift work, e.g. whether the participant works 7pm-7 am three days a week or 7pm-10 pm every day which may have a differential biological impact (Stevens et al. 2011). Whilst rotating shift work (shifts which rotate or change according to a set schedule) has also been linked with circadian disruption, we did not specifically investigate this group of workers.

We are unable to make conclusions regarding causality of the associations observed. We cannot exclude reverse causation since the analysis in GS and that based on current evening/night shift work in UKHLS was assessed cross-sectionally. Given education preceeds shift work status, the association implies that those with less education engage in occupations that include night shift work. This is in line with a previous study which found an inverse association between a genetic (rather than epigenetic) risk score for higher education and shift work participation (Daghlas et al. 2021).

To support our findings, similar analysis should be performed in larger cohorts with DNAm data. Future studies should also look at evaluating whether these biomarkers could provide insights into the potential effects of night shift work on subsequent health outcomes, e.g. cardiometabolic diseases and cancer.

Conclusions

In over 8,000 participants from two cohort studies, evening/night shift work was associated with both phenotypic and DNA methylation-based measures of higher BMI and lower education. DNAm predictors of smoking and ageing were also related to evening/night shift work. Epigenetic measures may provide insights into the health and lifestyle profiles of night shift workers.

Funding

GS received core support from the Chief Scientist Office of the Scottish Government Health Directorates (CZD/16/6) and the Scottish Funding Council (HR03006). Genotyping and DNA methylation profiling of the GS samples was carried out by the Genetics Core Laboratory at the Edinburgh Clinical Research Facility, Edinburgh, Scotland and was funded by the Medical Research Council UK and the Wellcome Trust (Wellcome Trust Strategic Award STratifying Resilience and Depression Longitudinally (STRADL; Reference 104036/Z/14/Z). The DNA methylation data assayed for Generation Scotland was partially funded by a 2018 NARSAD Young Investigator Grant from the Brain & Behavior Research Foundation (Ref: 27404; awardee: Dr David M Howard) and by a JMAS SIM fellowship from the Royal College of Physicians of Edinburgh (Awardee: Dr Heather C Whalley). For UKHLS, the epigenetics methylation data were analysed by the University of Exeter Medical School (MRC grant K013807) and further facilitated by the University of Essex School of Biological Sciences. PMH is funded by a Wellcome Trust 4-year studentship (108902/Z/15/Z). RCR is a de Pass VC research fellow at the University of Bristol. RCR, RMM and CLR are members of the MRC Integrative Epidemiology Unit at the University of Bristol funded by the Medical Research Council (MC_UU_00011/5). RMM, CLR and RCR are supported by a Cancer Research UK (C18281/A20919) programme grant (the Integrative Cancer Epidemiology Programme). RMM and CLR are supported by the National Institute for Health Research (NIHR) Bristol Biomedical Research Centre which is funded by the National Institute for Health Research (NIHR) and is a partnership between University Hospitals Bristol and Weston NHS Foundation Trust and the University of Bristol. FdV is supported by the NIHR School for Public Health Research and NIHR Applied Research Collaboration West. RCR is further supported by the National Institute for Health and Care Research (NIHR) Oxford Health Biomedical Research Centre (RCR) (NIHR203316). Understanding Society is an initiative funded by the Economic and Social Research Council (ES/N00812X/1) and various Government Departments, with scientific leadership by the Institute for Social and Economic Research, University of Essex, and survey delivery by NatCen Social Research and Kantar Public.

Footnotes

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Institutional Review Board Statement

Data governance was provided by the METADAC data access committee, funded by ESRC, Wellcome, and MRC (2015–-2018: Grant Number MR/N01104X/1 2018–2020: Grant Number ES/S008349/1). Ethical approval for Understanding Society was obtained from the National Research Service (Understanding Society – UK Household Longitudinal Study: A Biosocial Component, Oxfordshire A REC, Reference: 10/H0604/2). All our consents can be found here: https://www.understandingsociety.ac.uk/documentation/health-assessment/fieldwork-documents.

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

According to the terms of consent for GS participants, access to data must be reviewed by the GS Access Committee. Applications should be made to access@generationscotland.org. Understanding Society data are available through the UK Data Service.

References

  1. Bae M-J, Song Y-M, Shin J-Y, Choi B-Y, Keum J-H, Lee E-A. The association between shift work and health behavior: findings from the korean national health and nutrition examination survey. Korean J Fam Med. 2017;38:86–92. doi: 10.4082/kjfm.2017.38.2.86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Boini S, Bourgkard E, Ferrières J, Esquirol Y. What do we know about the effect of night-shift work on cardiovascular risk factors? An umbrella review. Front Public Health. 2022;10:1034195. doi: 10.3389/fpubh.2022.1034195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bushnell PT, Colombi A, Caruso CC, Tak S. Work schedules and health behavior outcomes at a large manufacturer. Ind Health. 2010;48:395–405. doi: 10.2486/indhealth.mssw-03. [DOI] [PubMed] [Google Scholar]
  4. Carugno M, Maggioni C, Ruggiero V, Crespi E, Monti P, Ferrari L, Pesatori AC. Can night shift work affect biological age? Hints from a cross-sectional study on hospital female nurses. Int J Environ Res Pub Health. 2021;18(20):10639. doi: 10.3390/ijerph182010639. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chen BH, Marioni RE, Colicino E, Peters MJ, Ward-Caviness CK, Tsai PC, Roetker NS, Just AC, Demerath EW, Guan W, et al. DNA methylation-based measures of biological age: meta-analysis predicting time to death. Aging (Albany NY) 2016;8:1844–1865. doi: 10.18632/aging.101020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cho Y-M, Kim H-R, Kang M-Y, Myong J-P, Koo JW. Fixed night workers and failed smoking cessation. J Occup Med Toxicol. 2019;14:23. doi: 10.1186/s12995-019-0243-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Daghlas I, Richmond RC, Lane JM, Dashti HS, Ollila HM, Schernhammer ES, Smith GD, Rutter MK, Saxena R, Vetter C. Selection into shift work is influenced by educational attainment and body mass index: a Mendelian randomization study in the UK biobank. Int J Epidemiol. 2021;50:1229–1240. doi: 10.1093/ije/dyab031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dorrian J, Heath G, Sargent C, Banks S, Coates A. Alcohol use in shiftworkers. Accid Anal Prev. 2017;99:395–400. doi: 10.1016/j.aap.2015.11.011. [DOI] [PubMed] [Google Scholar]
  9. Folkard S, Tucker P. Shift work, safety and productivity. Occup Med (Chic Ill) 2003;53:95–101. doi: 10.1093/occmed/kqg047. [DOI] [PubMed] [Google Scholar]
  10. Fransquet PD, Wrigglesworth J, Woods RL, Ernst ME, Ryan J. The epigenetic clock as a predictor of disease and mortality risk: a systematic review and meta-analysis. Clin Epigenet. 2019;11:62. doi: 10.1186/s13148-019-0656-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hannum G, Guinney J, Zhao L, Zhang L, Hughes G, Sadda S, Klotzle B, Bibikova M, Fan JB, Gao Y, et al. Genome-wide methylation profiles reveal quantitative views of human aging rates. Mol Cell. 2013;49:359–367. doi: 10.1016/j.molcel.2012.10.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Health and Safety Executive. Managing shiftwork: Health and safety guidance. London: HSE; 2006. http://www.hse.gov.uk/pubns/books/hsg256.htm . [Google Scholar]
  13. Hillary RF, Marioni RE. MethylDetectR: a software for methylation-based health profiling. Wellcome Open Res. 2021;5:283. doi: 10.12688/wellcomeopenres.16458.2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Horvath S. DNA methylation age of human tissues and cell types. Genome Biology. 2013;14:R115. doi: 10.1186/gb-2013-14-10-r115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Horvath S, Raj K. DNA methylation-based biomarkers and the epigenetic clock theory of ageing. Nat Rev Genet. 2018;19:371–384. doi: 10.1038/s41576-018-0004-3. [DOI] [PubMed] [Google Scholar]
  16. Kecklund G, Axelsson J. Health consequences of shift work and insufficient sleep. BMJ. 2016;355:i5210. doi: 10.1136/bmj.i5210. [DOI] [PubMed] [Google Scholar]
  17. Knutsson A, Akerstedt T, Jonsson BG. Prevalence of risk factors for coronary artery disease among day and shift workers. Scand J Work Environ Heal. 1988;14:317–321. doi: 10.5271/sjweh.1913. [DOI] [PubMed] [Google Scholar]
  18. Knutsson A, Bøggild H. Gastrointestinal disorders among shift workers. Scand J Work Environ Heal. 2010;36:85–95. doi: 10.5271/sjweh.2897. [DOI] [PubMed] [Google Scholar]
  19. Knutsson A, Nilsson T. Tobacco use and exposure to environmental tobacco smoke in relation to certain work characteristics. Scand J Soc Med. 1998;26:183–189. doi: 10.1177/14034948980260030801. [DOI] [PubMed] [Google Scholar]
  20. Koncevičius K, Nair A, Šveikauskaitė A, Šeštokaitė A, Kazlauskaitė A, Dulskas A, Petronis A. Epigenetic age oscillates during the day. Aging Cell. 2024;23:e14170. doi: 10.1111/acel.14170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Levine ME, Lu AT, Quach A, Chen BH, Assimes TL, Bandinelli S, Hou L, Baccarelli AA, Stewart JD, Li Y, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging-Us. 2018;10:573–591. doi: 10.18632/aging.101414. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Loef B, van der Beek AJ, Holtermann A, Hulsegge G, van Baarle D, Proper KI. Objectively measured physical activity of hospital shift workers. Scand J Work Environ Heal. 2018;44:265–273. doi: 10.5271/sjweh.3709. [DOI] [PubMed] [Google Scholar]
  23. Lu AT, Quach A, Wilson JG, Reiner AP, Aviv A, Raj K, Hou L, Baccarelli AA, Li Y, Stewart JD, et al. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging (Albany NY) 2019;11:303–327. doi: 10.18632/aging.101684. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Marioni RE, Shah S, McRae AF, Chen BH, Colicino E, Harris SE, Gibson J, Henders AK, Redmond P, Cox SR, et al. DNA methylation age of blood predicts all-cause mortality in later life. Genome Biol. 2015;16:25. doi: 10.1186/s13059-015-0584-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. McCartney DL, Hillary RF, Stevenson AJ, Ritchie SJ, Walker RM, Zhang Q, Morris SW, Bermingham ML, Campbell A, Murray AD, et al. Epigenetic prediction of complex traits and death. Genome Biol. 2018;19:136. doi: 10.1186/s13059-018-1514-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. NHS Digital. Health survey for England - 2013. 2013.
  27. Oh G, Ebrahimi S, Carlucci M, Zhang AP, Nair A, Groot DE, Labrie V, Jia PX, Oh ES, Jeremian RH, et al. Cytosine modifications exhibit circadian oscillations that are involved in epigenetic diversity and aging. Nat Commun. 2018;9(1):644. doi: 10.1038/s41467-018-03073-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Pośpiech E, Rudnicka J, Noroozi R, Pisarek-Pacek A, Wysocka B, Masny A, Boroń M, Migacz-Gruszka K, Pruszkowska-Przybylska P, Kobus M, et al. DNA methylation at AHRR as a master predictor of smoke exposure and a biomarker for sleep and exercise. Clin Epigenet. 2024;16:147. doi: 10.1186/s13148-024-01757-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Prince SA, Cardilli L, Reed JL, Saunders TJ, Kite C, Douillette K, Fournier K, Buckley JP. A comparison of self-reported and device measured sedentary behaviour in adults: a systematic review and meta-analysis. Int J Behav Nutr Phys Act. 2020;17:31. doi: 10.1186/s12966-020-00938-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Smith BH, Campbell A, Linksted P, Fitzpatrick B, Jackson C, Kerr SM, Deary IJ, MacIntyre DJ, Campbell H, McGilchrist M, et al. Cohort profile: generation Scotland: Scottish family health study (GS: SFHS). The study, its participants and their potential for genetic research on health and illness. Int J Epidemiol. 2013;42:689–700. doi: 10.1093/ije/dys084. [DOI] [PubMed] [Google Scholar]
  31. Stevens RG, Hansen J, Costa G, Haus E, Kauppinen T, Aronson KJ, Castano-Vinyals G, Davis S, Frings-Dresen MH, Fritschi L, et al. Considerations of circadian impact for defining ‘shift work’ in cancer studies: IARC working group report. Occup Environ Med. 2011;68:154–162. doi: 10.1136/oem.2009.053512. [DOI] [PubMed] [Google Scholar]
  32. Suwazono Y, Dochi M, Sakata K, Okubo Y, Oishi M, Tanaka K, Kobayashi E, Kido T, Nogawa K. A longitudinal study on the effect of shift work on weight gain in male Japanese workers. Obesity (Silver Spring) 2008;16:1887–1893. doi: 10.1038/oby.2008.298. [DOI] [PubMed] [Google Scholar]
  33. van Amelsvoort LG, Jansen NW, Kant I. Smoking among shift workers: more than a confounding factor. Chronobiol Int. 2006;23:1105–1113. doi: 10.1080/07420520601089539. [DOI] [PubMed] [Google Scholar]
  34. van Drongelen A, Boot CRL, Merkus SL, Smid T, van der Beek AJ. The effects of shift work on body weight change — a systematic review of longitudinal studies. Scand J Work Environ Heal. 2011;37:263–275. doi: 10.5271/sjweh.3143. [DOI] [PubMed] [Google Scholar]
  35. Wackers P, Dollé MET, van Oostrom CTM, van Kerkhof LWM. Exploration of genome-wide DNA methylation profiles in night shift workers. Epigenetics. 2023;18:2152637. doi: 10.1080/15592294.2022.2152637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Ward EM, Germolec D, Kogevinas M, McCormick D, Vermeulen R, Anisimov VN, Aronson KJ, Bhatti P, Cocco P, Costa G, et al. Carcinogenicity of night shift work. Lancet Oncol. 2019;20:1058–1059. doi: 10.1016/S1470-2045(19)30455-3. [DOI] [PubMed] [Google Scholar]
  37. White AJ, Kresovich JK, Xu Z, Sandler DP, Taylor JA. Shift work, DNA methylation and epigenetic age. Int J Epidemiol. 2019;48:1536–1544. doi: 10.1093/ije/dyz027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Wright KP, Jr, Bogan RK, Wyatt JK. Shift work and the assessment and management of shift work disorder (SWD. Sleep Med Rev. 2013;17:41–54. doi: 10.1016/j.smrv.2012.02.002. [DOI] [PubMed] [Google Scholar]
  39. Zhang Y, Elgizouli M, Schöttker B, Holleczek B, Nieters A, Brenner H. Smoking-associated DNA methylation markers predict lung cancer incidence. Clin Epigenet. 2016;8:127. doi: 10.1186/s13148-016-0292-4. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

According to the terms of consent for GS participants, access to data must be reviewed by the GS Access Committee. Applications should be made to access@generationscotland.org. Understanding Society data are available through the UK Data Service.

RESOURCES