Shift work, chronotype and the risk of cardiometabolic risk factors

Shift work, chronotype and the risk of cardiometabolic risk factors Abstract Background The relation between shift work and a large variety of cardiometabolic risk factors is unclear. Also, the role of chronotype is understudied. We examined relations between shift work and cardiometabolic risk factors, and explored these relations in different chronotypes. Methods Cardiometabolic risk factors (anthropometry, blood pressure, lipids, diabetes, γ-glutamyltransferase, C-reactive protein, uric acid and estimated glomerular filtration rate) were assessed among 1334 adults in 1987–91, with repeated measurements every 5 years. Using shift work history data collected in 2013–15, we identified shift work status 1 year prior to all six waves. Linear mixed models and logistic generalized estimating equations were used to estimate the longitudinal relations between shift work and risk factors 1 year later. Results Shift work was not significantly related with cardiometabolic risk factors (P ≥ 0.05), except for overweight/body mass index. Shift workers had more often overweight (OR: 1.44, 95% CI 1.06–1.95) and a higher body mass index (BMI) (β: 0.56 kg m−2, 95% CI 0.10–1.03) than day workers. A significant difference in BMI between day and shift workers was observed among evening chronotypes (β: 0.97 kg m−2, 95% CI 0.21–1.73), but not among morning chronotypes (β: 0.04 kg m−2, 95% CI −0.85 to 0.93). No differences by frequency of night shifts and duration of shift work were observed. Conclusion Shift workers did not have an increased risk of cardiometabolic risk factors compared with day workers, but, in particular shift working evening chronotypes, had an increased risk of overweight. More research is needed to verify our results, and establish whether tailored interventions by chronotype are wanted. Introduction Shift work is nowadays an inherent part of our 24/7 society, with ∼20% of the work force in Europe and the USA working in shifts outside regular daytime working hours.1,2 The chronic disruption of the internal body clock that regulates the daily cycles of biological processes (i.e. circadian rhythms) due to shift work has been associated with negative health outcomes, such as breast cancer and cardiovascular diseases.3–6 Cardiometabolic disorders, such as obesity and type 2 diabetes, develop long before the onset of cardiovascular disease, and may mediate the relation between shift work and cardiovascular diseases.7–9 A recent review showed that shift work was associated with an increased risk of overweight and impaired glucose tolerance.7 A few studies indicated shift work to be associated with other cardiometabolic risk factors, such as triglycerides,9 hypertension,9 uric acid9 and inflammatory markers,8 but according to the review of Proper et al. strong evidence is still lacking.7 This is mainly due to the low number of studies, limited methodological quality of previous studies, including cross-sectional designs.7 It has been suggested that the relation between shift work and cardiometabolic function is influenced by chronotype, i.e. internal circadian rhythm that influences the cycle of sleep and activity.10 Several studies showed that chronotype is associated with cardiometabolic risk factors.11–14 For example, evening chronotypes have been shown to have a higher risk of type 2 diabetes and metabolic syndrome than morning chronotypes.12,14 Moreover, in one study, night shifts increased the risk of type 2 diabetes in morning chronotypes, but decreased the risk in evening chronotypes.10 The reason may be that the biological rhythm is more disturbed by night shifts in morning chronotypes than in evening chronotypes. In contrast, morning shifts may disturb the biological rhythm less strongly in morning chronotypes. In view of the gaps in the scientific literature on shift work, chronotype and cardiometabolic risk factors, we investigated relations between shift work and various cardiometabolic risk factors, and explored these potential relations in different chronotypes. We hypothesize that shift workers, in particular morning chronotypes, have an increased risk of cardiometabolic risk factors compared with day workers. Methods Study population The Doetinchem Cohort Study is an ongoing population-based longitudinal study of men and women who were at the start of the study aged 20–59 years. Participants were measured in 1987–91 (N = 7768), 1993–97 (N = 6117), 1998–2002 (N = 4918), 2003–07 (N = 4520) and 2008–12 (N = 4018). Data collection of the sixth wave is still ongoing, but data of the first three measurement years (2013–15, N = 2041) were available for the present study. Response rates were 75% or higher in waves 2–6.15 We included subjects who participated in Wave 6 since the sixth wave included questions about lifelong shift work history, which enabled us to retrospectively determine shift work status at previous waves. We excluded participants who did not have paid work or had missing data on shift work status, employment status and covariates at any wave. Informed consent was obtained from all individual participants. Shift work Shift work status at each wave was assessed retrospectively in 2013–15 using a questionnaire based on the most important aspects of shift work.16 Participants were asked to indicate whether they ever worked evening shifts (i.e. shifts ending before midnight), night shifts (i.e. shifts that continued or started after midnight), sleep shifts and rotating shifts. All subjects were also asked to the year they started and stopped shift work, and the number of years and months they worked in shifts. Using this shift work history data, we classified participants as shift workers 1 year prior to each wave if they worked in evening, night or rotating shifts (with or without night shifts); participants who previously worked in irregular shift were considered former shift workers; others were considered day workers. The frequency of night shifts was categorized into no night shifts, 1–4 and ≥5 night shifts/month. Chronotype In 2013–15, chronotype was assessed by a single question: ‘How would you describe yourself?’ Based on six response categories, participants were categorized as ‘morning type’ (‘definitely a morning type’ OR ‘rather more a morning than an evening type’), ‘evening type’ (‘definitely an evening type’ OR ‘rather more an evening than a morning type’) and ‘intermediate type’ (neither). For participants who answered ‘do not know’ (n = 55), chronotype value was set as missing. This self-awareness of chronotype has been shown to be in excellent agreement with the quantitatively assessed chronotype based on sleep times.17 Measurements of cardiometabolic risk factors Weight, height, total and HDL cholesterol, diastolic and systolic blood pressure were objectively measured at all six waves. Waist circumference and random non-fasting glucose were measured at waves 2–6. In combination with self-reported diabetes, a conservative and common cut-off point of random glucose (i.e. 11.1 mmol l−1) was used to determine presence of diabetes (table 1). Details of all measurements have been described previously and in table 1.15 In 2013–14, standardised enzymatic methods were used to retrospectively determine triglycerides, γ-glutamyltransferase (GGT), high sensitivity CRP, uric acid, cystatin C and creatinine of waves 2–5 using blood plasma that had been stored in freezers. Estimated glomerular filtration rate (eGFR) was estimated with the Chronic Disease Epidemiology Collaboration (CKD-EPI) equation.18 Table 1 Description of the measurements and cut-off values of unfavourable levels of the cardiometabolic risk factors Measured waves  Cardiometabolic risk factor  Measurement method  Cut-off value unfavourable level  1–6  Body mass index  Body weight was objectively measured to the nearest 0.1 kg on calibrated scales and body height was measured to the nearest 0.5 cm. Body mass index was calculated as weight divided by height squared (kg m−2)  Overweight: ≥25 kg m−2  2–6  Waist circumference  Twice to the nearest 0.5 cm, at the level midway between the lowest rib and the iliac crest at the end of expiration, with participants in standing position. The mean of two measures was used for analysis  Abdominal obesity: men: ≥102 cm, women: ≥88 cm  1–6  Blood pressure  Diastolic and systolic blood pressure were measured in sitting position with a random zero sphygmomanometer in waves one to three. In waves four to six a Speidel Keller meter.  Hypertension: diastolic blood pressure ≥90 mmHg, and/or systolic blood pressure ≥140 mmHg, and/or taking anti-hypertensive medication  1–6  Total cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypercholesterolaemia: ≥6.0 mmol l−1 and/or taking cholesterol-lowering medication  1–6  HDL cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypo HDL-cholesterol: <0.9 mmol l−1  2–5  Triglycerides  GPO-PAP assay, colorimetric method  Hypertriglyceridaemia; ≥2.3 mmol l−1  2–5  GGT  Colorimetric method  High GGT: >50 U l−1  2–5  High sensitivity CRP  Principle of particle-enhanced immunological agglutination (Tina-quant C-reactive protein)  High CRP: >3.0 mg l−1  2–5  Uric acid  Colorimetric method  Hyperuricaemia: men: >7.0 mg dl−1, women: >6.0 mg dl−1  2–5  Estimated glomerular filtration rate  Cystatin C was measured with the particle enhanced-turbidimetric immunoassay using reagents from Gentian. Creatinine was measured with a Creatinine Plus assay (IDMS traceable)  Low estimated glomerular filtration rate <90 ml min−1/1.73 m2  2–6  Glucose  Plasma glucose levels were determined in random non-fasting venous blood samples using the hexokinase method  Type 2 diabetes: random glucose ≥11.1 mmol l−1 and/or self-reported diabetes  Measured waves  Cardiometabolic risk factor  Measurement method  Cut-off value unfavourable level  1–6  Body mass index  Body weight was objectively measured to the nearest 0.1 kg on calibrated scales and body height was measured to the nearest 0.5 cm. Body mass index was calculated as weight divided by height squared (kg m−2)  Overweight: ≥25 kg m−2  2–6  Waist circumference  Twice to the nearest 0.5 cm, at the level midway between the lowest rib and the iliac crest at the end of expiration, with participants in standing position. The mean of two measures was used for analysis  Abdominal obesity: men: ≥102 cm, women: ≥88 cm  1–6  Blood pressure  Diastolic and systolic blood pressure were measured in sitting position with a random zero sphygmomanometer in waves one to three. In waves four to six a Speidel Keller meter.  Hypertension: diastolic blood pressure ≥90 mmHg, and/or systolic blood pressure ≥140 mmHg, and/or taking anti-hypertensive medication  1–6  Total cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypercholesterolaemia: ≥6.0 mmol l−1 and/or taking cholesterol-lowering medication  1–6  HDL cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypo HDL-cholesterol: <0.9 mmol l−1  2–5  Triglycerides  GPO-PAP assay, colorimetric method  Hypertriglyceridaemia; ≥2.3 mmol l−1  2–5  GGT  Colorimetric method  High GGT: >50 U l−1  2–5  High sensitivity CRP  Principle of particle-enhanced immunological agglutination (Tina-quant C-reactive protein)  High CRP: >3.0 mg l−1  2–5  Uric acid  Colorimetric method  Hyperuricaemia: men: >7.0 mg dl−1, women: >6.0 mg dl−1  2–5  Estimated glomerular filtration rate  Cystatin C was measured with the particle enhanced-turbidimetric immunoassay using reagents from Gentian. Creatinine was measured with a Creatinine Plus assay (IDMS traceable)  Low estimated glomerular filtration rate <90 ml min−1/1.73 m2  2–6  Glucose  Plasma glucose levels were determined in random non-fasting venous blood samples using the hexokinase method  Type 2 diabetes: random glucose ≥11.1 mmol l−1 and/or self-reported diabetes  HDL, high density lipoprotein; GGT, γ-glutamyltransferase; CRP, C-reactive protein. Table 1 Description of the measurements and cut-off values of unfavourable levels of the cardiometabolic risk factors Measured waves  Cardiometabolic risk factor  Measurement method  Cut-off value unfavourable level  1–6  Body mass index  Body weight was objectively measured to the nearest 0.1 kg on calibrated scales and body height was measured to the nearest 0.5 cm. Body mass index was calculated as weight divided by height squared (kg m−2)  Overweight: ≥25 kg m−2  2–6  Waist circumference  Twice to the nearest 0.5 cm, at the level midway between the lowest rib and the iliac crest at the end of expiration, with participants in standing position. The mean of two measures was used for analysis  Abdominal obesity: men: ≥102 cm, women: ≥88 cm  1–6  Blood pressure  Diastolic and systolic blood pressure were measured in sitting position with a random zero sphygmomanometer in waves one to three. In waves four to six a Speidel Keller meter.  Hypertension: diastolic blood pressure ≥90 mmHg, and/or systolic blood pressure ≥140 mmHg, and/or taking anti-hypertensive medication  1–6  Total cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypercholesterolaemia: ≥6.0 mmol l−1 and/or taking cholesterol-lowering medication  1–6  HDL cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypo HDL-cholesterol: <0.9 mmol l−1  2–5  Triglycerides  GPO-PAP assay, colorimetric method  Hypertriglyceridaemia; ≥2.3 mmol l−1  2–5  GGT  Colorimetric method  High GGT: >50 U l−1  2–5  High sensitivity CRP  Principle of particle-enhanced immunological agglutination (Tina-quant C-reactive protein)  High CRP: >3.0 mg l−1  2–5  Uric acid  Colorimetric method  Hyperuricaemia: men: >7.0 mg dl−1, women: >6.0 mg dl−1  2–5  Estimated glomerular filtration rate  Cystatin C was measured with the particle enhanced-turbidimetric immunoassay using reagents from Gentian. Creatinine was measured with a Creatinine Plus assay (IDMS traceable)  Low estimated glomerular filtration rate <90 ml min−1/1.73 m2  2–6  Glucose  Plasma glucose levels were determined in random non-fasting venous blood samples using the hexokinase method  Type 2 diabetes: random glucose ≥11.1 mmol l−1 and/or self-reported diabetes  Measured waves  Cardiometabolic risk factor  Measurement method  Cut-off value unfavourable level  1–6  Body mass index  Body weight was objectively measured to the nearest 0.1 kg on calibrated scales and body height was measured to the nearest 0.5 cm. Body mass index was calculated as weight divided by height squared (kg m−2)  Overweight: ≥25 kg m−2  2–6  Waist circumference  Twice to the nearest 0.5 cm, at the level midway between the lowest rib and the iliac crest at the end of expiration, with participants in standing position. The mean of two measures was used for analysis  Abdominal obesity: men: ≥102 cm, women: ≥88 cm  1–6  Blood pressure  Diastolic and systolic blood pressure were measured in sitting position with a random zero sphygmomanometer in waves one to three. In waves four to six a Speidel Keller meter.  Hypertension: diastolic blood pressure ≥90 mmHg, and/or systolic blood pressure ≥140 mmHg, and/or taking anti-hypertensive medication  1–6  Total cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypercholesterolaemia: ≥6.0 mmol l−1 and/or taking cholesterol-lowering medication  1–6  HDL cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypo HDL-cholesterol: <0.9 mmol l−1  2–5  Triglycerides  GPO-PAP assay, colorimetric method  Hypertriglyceridaemia; ≥2.3 mmol l−1  2–5  GGT  Colorimetric method  High GGT: >50 U l−1  2–5  High sensitivity CRP  Principle of particle-enhanced immunological agglutination (Tina-quant C-reactive protein)  High CRP: >3.0 mg l−1  2–5  Uric acid  Colorimetric method  Hyperuricaemia: men: >7.0 mg dl−1, women: >6.0 mg dl−1  2–5  Estimated glomerular filtration rate  Cystatin C was measured with the particle enhanced-turbidimetric immunoassay using reagents from Gentian. Creatinine was measured with a Creatinine Plus assay (IDMS traceable)  Low estimated glomerular filtration rate <90 ml min−1/1.73 m2  2–6  Glucose  Plasma glucose levels were determined in random non-fasting venous blood samples using the hexokinase method  Type 2 diabetes: random glucose ≥11.1 mmol l−1 and/or self-reported diabetes  HDL, high density lipoprotein; GGT, γ-glutamyltransferase; CRP, C-reactive protein. Covariates Data on educational attainment, lifestyle and use of anti-hypertensive medication and cholesterol-lowering medication were obtained by questionnaire as described in detail elsewhere.15 Educational attainment was categorized as low (intermediate secondary education or less), intermediate (intermediate vocational or higher secondary education), and high (higher vocational education or university). Smoking was categorized as current smoker and non-smoker, and alcohol as low (≤1 glass day−1) and high (>1 glass day−1) alcohol consumption. At baseline, leisure-time physical activity was measured by asking whether adults were at least physically active during 4 h per week. In the second to sixth wave, several items on self-reported time spent in moderate-to-vigorous intensity leisure-time physical activities were combined and the conventional cut-off point at 3.5 h per week was used.19 Occupational physical activity was categorized as low (sedentary occupation), moderate (standing and light manual work), and high (heavy manual work). Statistical analysis For the continuous cardiometabolic risk factors, linear mixed models were used to examine its relations with shift work, while for dichotomous unfavourable levels of cardiometabolic risk factors, logistic generalized estimation equations with exchangeable correlation matrix were used. For both methods, we used longitudinal time lagged models in which shift work status was determined one year prior to the outcomes. A time lag of one year was chosen because shift work has previously been associated with change in BMI and LDL/HDL-ratio over a 1-year period.20 As all outcome measures at the different time-points are analyzed together, the overall results represent the combination of within-subject and between-subject relationships.21 All analyses were adjusted for age, gender, educational level and chronotype. In analyses with diastolic and systolic blood pressure as dependent variable, additional adjustment for anti-hypertensive medication was made; similarly with total cholesterol, HDL cholesterol and triglycerides as dependent variable, adjustment for cholesterol-lowering medication was made. To investigate whether the potential relations of shift work with cardiometabolic risk factors were independent from potential intermediate lifestyle factors, analyses were additionally adjusted for smoking status, alcohol consumption, occupational and leisure-time physical activity. We performed subgroup analyses separately for morning, intermediate and evening chronotypes. In the total population, three groups of shift workers (no night shifts, 1–4 night shifts/month and ≥5 night shifts/month) were compared with day workers to study exposure–response relations regarding frequency of night shifts with cardiometabolic risk factors. We also performed sensitivity analyses in which we stratified the analyses by duration of shift work, i.e. we compared shift workers who worked for <10 and ≥10 years in irregular shifts to day workers.22–24 All analyses were performed using SAS 9.3 software and a two-sided P values of <0.05 was considered statistically significant. Results Out of the 2041 subjects who participated in wave 6, we excluded 353 individuals who were not employed at all waves, 310 with insufficient shift work data and 44 participants with missing data on cardiometabolic risk factors and covariates at all waves. This led to a study population of 808 day workers, 128 former shift workers and 125 shift workers at wave 1. The remainder of the study population (n = 273) was unemployed or employment status was unknown at wave 1, but was employed at any of the other waves. The number of participants in wave 1, 2, 3, 4, 5 and 6 were 1061, 713, 937, 868, 741 and 375, respectively. The number of shift workers in wave 2, 3, 4, 5 and 6 were 70, 104, 84, 81 and 37, respectively. Compared with excluded participants, the included participants were at baseline slightly younger (37.7 vs. 38.3 years), less often men (54 vs. 59%), more often high educated (24 vs. 19%) and had slightly lower levels of cardiometabolic risk factors (e.g. BMI 24.0 kg m−2 vs. 24.5 kg m−2 and systolic blood pressure 120 mmHg vs. 122 mmHg) (data not shown). At Wave 1, the mean age was 37.7 years (SD: 8.9), 54% were men and 24% were highly educated. Among shift workers, 85% worked at least once a month during the night and 49% worked in shifts for 10 years or more (table 2). Table 2 Characteristics of day workers, former shift workers, and shift workers in 1987–89   Never shift workers  Former shift workers  Shift workers  N = 808  N = 128  N = 125  Demographics            Age (years)  39.4 ± 9.0  35.6 ± 9.7  36.2 ± 9.2      Gender (% male)  64 (518)  63 (81)  65 (81)      High educational level (%)  29 (236)  23 (29)  14 (18)  Chronotype                Morning type (%)  36 (289)  32 (41)  28 (35)          Intermediate type (%)  32 (256)  28 (36)  32 (40)          Evening type (%)  28 (226)  36 (46)  34 (42)          Unknown type (%)  5 (37)  4 (5)  6 (8)  Occupational            Number night shifts per month  ·  ·  4.9 ± 4.1          0 night shifts/month (%)  ·  ·  15 (17)          1–4 night shifts/month (%)  ·  ·  35 (41)          ≥5 night shifts/month (%)  ·  ·  50 (58)  Years of shift work      10.9 (8.5)          <10 years (%)  ·  ·  50 (63)          10–19 years (%)  ·  ·  35 (44)          ≥20 years (%)  ·  ·  14 (18)  Lifestyle            Current smoker (%)  26 (208)  30 (39)  29 (36)      High leisure-time physical activity (%)  70 (573)  73 (93)  72 (90)      High occupational physical activity (%)  4 (36)  2 (3)  2 (3)      High alcohol intake (%)  43 (353)  34 (44)  45 (56)  Cardiometabolic risk factors            Body mass index (kg m−2)  24.2 ± 2.8  23.9 ± 3.0  24.4 ± 3.0      Waist circumference (cm)a  90 ± 10  89 ± 11  90 ± 11      Systolic blood pressure (mmHg)  121 ± 13  121 ± 13  121 ± 14      Diastolic blood pressure (mmHg)  76 ± 9  76 ± 10  74 ± 10      Total cholesterol (mmol l−1)  5.5 ± 1.0  5.1 ± 0.9  5.4 ± 1.0      HDL cholesterol (mmol l−1)  1.26 ± 0.31  1.23 ± 0.32  1.3 ± 0.3      Triglycerides (mmol l−1)a  1.25 (0.90–1.75)  1.05 (0.88–1.47)  1.19 (0.87–1.58)      γ-Glutamyltransferase (U l−1)a  21 (13–32)  18 (13–27)  21 (14–31)      C-reactive protein (mg l−1)a  0.90 (0.44–1.78)  0.72 (0.38–1.41)  0.78 (0.40–1.65)      Uric Acid (mg dl−1)a  5.0 ± 1.2  4.8 ± 1.2  5.2 ± 1.1      eGFR (ml min−1/1.73 m2)a  106 ± 13  110 ± 13  107 ± 14  Medication            Anti-hypertensive medication (%)  2.3 (19)  3.2 (4)  2.1 (3)      Cholesterol-lowering medication (%)  0 (0)  0 (0)  0.0 (0)    Never shift workers  Former shift workers  Shift workers  N = 808  N = 128  N = 125  Demographics            Age (years)  39.4 ± 9.0  35.6 ± 9.7  36.2 ± 9.2      Gender (% male)  64 (518)  63 (81)  65 (81)      High educational level (%)  29 (236)  23 (29)  14 (18)  Chronotype                Morning type (%)  36 (289)  32 (41)  28 (35)          Intermediate type (%)  32 (256)  28 (36)  32 (40)          Evening type (%)  28 (226)  36 (46)  34 (42)          Unknown type (%)  5 (37)  4 (5)  6 (8)  Occupational            Number night shifts per month  ·  ·  4.9 ± 4.1          0 night shifts/month (%)  ·  ·  15 (17)          1–4 night shifts/month (%)  ·  ·  35 (41)          ≥5 night shifts/month (%)  ·  ·  50 (58)  Years of shift work      10.9 (8.5)          <10 years (%)  ·  ·  50 (63)          10–19 years (%)  ·  ·  35 (44)          ≥20 years (%)  ·  ·  14 (18)  Lifestyle            Current smoker (%)  26 (208)  30 (39)  29 (36)      High leisure-time physical activity (%)  70 (573)  73 (93)  72 (90)      High occupational physical activity (%)  4 (36)  2 (3)  2 (3)      High alcohol intake (%)  43 (353)  34 (44)  45 (56)  Cardiometabolic risk factors            Body mass index (kg m−2)  24.2 ± 2.8  23.9 ± 3.0  24.4 ± 3.0      Waist circumference (cm)a  90 ± 10  89 ± 11  90 ± 11      Systolic blood pressure (mmHg)  121 ± 13  121 ± 13  121 ± 14      Diastolic blood pressure (mmHg)  76 ± 9  76 ± 10  74 ± 10      Total cholesterol (mmol l−1)  5.5 ± 1.0  5.1 ± 0.9  5.4 ± 1.0      HDL cholesterol (mmol l−1)  1.26 ± 0.31  1.23 ± 0.32  1.3 ± 0.3      Triglycerides (mmol l−1)a  1.25 (0.90–1.75)  1.05 (0.88–1.47)  1.19 (0.87–1.58)      γ-Glutamyltransferase (U l−1)a  21 (13–32)  18 (13–27)  21 (14–31)      C-reactive protein (mg l−1)a  0.90 (0.44–1.78)  0.72 (0.38–1.41)  0.78 (0.40–1.65)      Uric Acid (mg dl−1)a  5.0 ± 1.2  4.8 ± 1.2  5.2 ± 1.1      eGFR (ml min−1/1.73 m2)a  106 ± 13  110 ± 13  107 ± 14  Medication            Anti-hypertensive medication (%)  2.3 (19)  3.2 (4)  2.1 (3)      Cholesterol-lowering medication (%)  0 (0)  0 (0)  0.0 (0)  Note: Values represent means ± standard deviations, percentages and (numbers) or median and (interquartile ranges). a Not assessed at baseline. Therefore, values of wave two (1993–97) are shown. eGFR, estimated glomerular filtration rate; HDL, high density lipoprotein. Table 2 Characteristics of day workers, former shift workers, and shift workers in 1987–89   Never shift workers  Former shift workers  Shift workers  N = 808  N = 128  N = 125  Demographics            Age (years)  39.4 ± 9.0  35.6 ± 9.7  36.2 ± 9.2      Gender (% male)  64 (518)  63 (81)  65 (81)      High educational level (%)  29 (236)  23 (29)  14 (18)  Chronotype                Morning type (%)  36 (289)  32 (41)  28 (35)          Intermediate type (%)  32 (256)  28 (36)  32 (40)          Evening type (%)  28 (226)  36 (46)  34 (42)          Unknown type (%)  5 (37)  4 (5)  6 (8)  Occupational            Number night shifts per month  ·  ·  4.9 ± 4.1          0 night shifts/month (%)  ·  ·  15 (17)          1–4 night shifts/month (%)  ·  ·  35 (41)          ≥5 night shifts/month (%)  ·  ·  50 (58)  Years of shift work      10.9 (8.5)          <10 years (%)  ·  ·  50 (63)          10–19 years (%)  ·  ·  35 (44)          ≥20 years (%)  ·  ·  14 (18)  Lifestyle            Current smoker (%)  26 (208)  30 (39)  29 (36)      High leisure-time physical activity (%)  70 (573)  73 (93)  72 (90)      High occupational physical activity (%)  4 (36)  2 (3)  2 (3)      High alcohol intake (%)  43 (353)  34 (44)  45 (56)  Cardiometabolic risk factors            Body mass index (kg m−2)  24.2 ± 2.8  23.9 ± 3.0  24.4 ± 3.0      Waist circumference (cm)a  90 ± 10  89 ± 11  90 ± 11      Systolic blood pressure (mmHg)  121 ± 13  121 ± 13  121 ± 14      Diastolic blood pressure (mmHg)  76 ± 9  76 ± 10  74 ± 10      Total cholesterol (mmol l−1)  5.5 ± 1.0  5.1 ± 0.9  5.4 ± 1.0      HDL cholesterol (mmol l−1)  1.26 ± 0.31  1.23 ± 0.32  1.3 ± 0.3      Triglycerides (mmol l−1)a  1.25 (0.90–1.75)  1.05 (0.88–1.47)  1.19 (0.87–1.58)      γ-Glutamyltransferase (U l−1)a  21 (13–32)  18 (13–27)  21 (14–31)      C-reactive protein (mg l−1)a  0.90 (0.44–1.78)  0.72 (0.38–1.41)  0.78 (0.40–1.65)      Uric Acid (mg dl−1)a  5.0 ± 1.2  4.8 ± 1.2  5.2 ± 1.1      eGFR (ml min−1/1.73 m2)a  106 ± 13  110 ± 13  107 ± 14  Medication            Anti-hypertensive medication (%)  2.3 (19)  3.2 (4)  2.1 (3)      Cholesterol-lowering medication (%)  0 (0)  0 (0)  0.0 (0)    Never shift workers  Former shift workers  Shift workers  N = 808  N = 128  N = 125  Demographics            Age (years)  39.4 ± 9.0  35.6 ± 9.7  36.2 ± 9.2      Gender (% male)  64 (518)  63 (81)  65 (81)      High educational level (%)  29 (236)  23 (29)  14 (18)  Chronotype                Morning type (%)  36 (289)  32 (41)  28 (35)          Intermediate type (%)  32 (256)  28 (36)  32 (40)          Evening type (%)  28 (226)  36 (46)  34 (42)          Unknown type (%)  5 (37)  4 (5)  6 (8)  Occupational            Number night shifts per month  ·  ·  4.9 ± 4.1          0 night shifts/month (%)  ·  ·  15 (17)          1–4 night shifts/month (%)  ·  ·  35 (41)          ≥5 night shifts/month (%)  ·  ·  50 (58)  Years of shift work      10.9 (8.5)          <10 years (%)  ·  ·  50 (63)          10–19 years (%)  ·  ·  35 (44)          ≥20 years (%)  ·  ·  14 (18)  Lifestyle            Current smoker (%)  26 (208)  30 (39)  29 (36)      High leisure-time physical activity (%)  70 (573)  73 (93)  72 (90)      High occupational physical activity (%)  4 (36)  2 (3)  2 (3)      High alcohol intake (%)  43 (353)  34 (44)  45 (56)  Cardiometabolic risk factors            Body mass index (kg m−2)  24.2 ± 2.8  23.9 ± 3.0  24.4 ± 3.0      Waist circumference (cm)a  90 ± 10  89 ± 11  90 ± 11      Systolic blood pressure (mmHg)  121 ± 13  121 ± 13  121 ± 14      Diastolic blood pressure (mmHg)  76 ± 9  76 ± 10  74 ± 10      Total cholesterol (mmol l−1)  5.5 ± 1.0  5.1 ± 0.9  5.4 ± 1.0      HDL cholesterol (mmol l−1)  1.26 ± 0.31  1.23 ± 0.32  1.3 ± 0.3      Triglycerides (mmol l−1)a  1.25 (0.90–1.75)  1.05 (0.88–1.47)  1.19 (0.87–1.58)      γ-Glutamyltransferase (U l−1)a  21 (13–32)  18 (13–27)  21 (14–31)      C-reactive protein (mg l−1)a  0.90 (0.44–1.78)  0.72 (0.38–1.41)  0.78 (0.40–1.65)      Uric Acid (mg dl−1)a  5.0 ± 1.2  4.8 ± 1.2  5.2 ± 1.1      eGFR (ml min−1/1.73 m2)a  106 ± 13  110 ± 13  107 ± 14  Medication            Anti-hypertensive medication (%)  2.3 (19)  3.2 (4)  2.1 (3)      Cholesterol-lowering medication (%)  0 (0)  0 (0)  0.0 (0)  Note: Values represent means ± standard deviations, percentages and (numbers) or median and (interquartile ranges). a Not assessed at baseline. Therefore, values of wave two (1993–97) are shown. eGFR, estimated glomerular filtration rate; HDL, high density lipoprotein. Shift work and cardiometabolic risk factors After adjustment for age, gender, education and chronotype, shift workers had more often overweight (OR: 1.44, 95% CI 1.06–1.95) than day workers (table 3), as well as a significantly higher BMI (i.e. 0.56 kg m−2, 95% CI 0.10–1.03). Shift work was not significantly related with other cardiometabolic risk factors, and former shift workers did not have an increased risk for any of the cardiometabolic risk factors compared with day workers. Results did not change after additional adjustment for lifestyle factors (data not shown). Table 3 Multivariable-adjusted differences in cardiometabolic risk factors between day workers, former shift workers, and shift workers   Former shift workers versus day workers   Shift workers versus day workers   OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.00 (0.76 to 1.32)  0.99  1.44 (1.06 to 1.95)  0.02  Abdominal obesity (waist circumference ≥88/102)  1.24 (0.91 to 1.68)  0.17  1.26 (0.89 to 1.79)  0.18  Hypertension (≥90/140 mmHg)  1.05 (0.78 to 1.40)  0.77  0.99 (0.68 to 1.43)  0.84  Hypercholesterolaemia (≥6.0 mmol l−1)  0.84 (0.64 to 1.11)  0.84  1.00 (0.73 to 1.38)  0.99  Low HDL cholesterol (<0.9 mmol l−1)  1.07 (0.68 to 1.69)  0.76  0.86 (0.52 to 1.45)  0.57  Hypertriglyceridaemia (2.3 mmol l−1)  1.09 (0.72 to 1.64)  0.69  0.96 (0.57 to 1.61)  0.92  Type 2 diabetes (≥11.1 mmol l−1 or self-reported)  1.72 (0.71 to 4.17)  0.23  1.58 (0.58 to 4.30)  0.38  High γ-glutamyltransferase (≥50 U l−1)  1.31 (0.81 to 2.11)  0.27  0.82 (0.45 to 1.45)  0.41  High C-reactive protein (≥3.0 mg l−1)  0.82 (0.57 to 1.19)  0.29  1.34 (0.87 to 2.06)  0.24  Hyperuricaemia (≥2.3 mmol l−1)  0.78 (0.44 to 1.41)  0.41  0.73 (0.39 to 1.38)  0.32  Low eGFR (<90 ml min−1/1.73 m2)  1.00 (0.66 to 1.52)  1.00  1.16 (0.74 to 1.82)  0.57    β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.09 (−0.34 to 0.52)  0.69  0.56 (0.10 to 1.03)  0.02  Waist circumference (cm)  0.04 (−1.4 to 1.5)  0.96  1.1 (−0.5 to 2.7)  0.18  Systolic blood pressure (mmHg)b  −0.7 (−2.5 to 1.1)  0.43  −0.6 (−2.6 to 1.4)  0.53  Diastolic blood pressure (mmHg)b  −0.6 (−1.7 to 0.6)  0.33  −0.7 (−2.0 to 0.6)  0.28  Total cholesterol (mmol l−1)b  −0.14 (−0.27 to −0.01)  0.04  −0.03 (−0.18 to 0.12)  0.69  HDL cholesterol (mmol l−1)b  −0.02 (−0.06 to 0.03)  0.52  −0.01 (−0.06 to 0.04)  0.78  Triglycerides (log mmol l−1)b  −0.06 (−0.12 to 0.01)  0.11  0.01 (−0.07 to 0.09)  0.84  γ-Glutamyltransferase (log U l−1)  0.001 (−0.08 to 0.08)  0.99  −0.02 (−0.11 to 0.08)  0.71  C-reactive protein (log mg l−1)  −0.09 (−0.23 to 0.04)  0.19  0.01 (−0.14 to 0.17)  0.87  Uric acid (mg dl−1)  0.01 (−0.14 to 0.16)  0.89  0.11 (−0.06 to 0.28)  0.21  eGFR (ml min−1/1.73 m2)  −0.5 (−2.2 to 1.2)  0.57  −1.3 (−3.2 to 0.6)  0.18    Former shift workers versus day workers   Shift workers versus day workers   OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.00 (0.76 to 1.32)  0.99  1.44 (1.06 to 1.95)  0.02  Abdominal obesity (waist circumference ≥88/102)  1.24 (0.91 to 1.68)  0.17  1.26 (0.89 to 1.79)  0.18  Hypertension (≥90/140 mmHg)  1.05 (0.78 to 1.40)  0.77  0.99 (0.68 to 1.43)  0.84  Hypercholesterolaemia (≥6.0 mmol l−1)  0.84 (0.64 to 1.11)  0.84  1.00 (0.73 to 1.38)  0.99  Low HDL cholesterol (<0.9 mmol l−1)  1.07 (0.68 to 1.69)  0.76  0.86 (0.52 to 1.45)  0.57  Hypertriglyceridaemia (2.3 mmol l−1)  1.09 (0.72 to 1.64)  0.69  0.96 (0.57 to 1.61)  0.92  Type 2 diabetes (≥11.1 mmol l−1 or self-reported)  1.72 (0.71 to 4.17)  0.23  1.58 (0.58 to 4.30)  0.38  High γ-glutamyltransferase (≥50 U l−1)  1.31 (0.81 to 2.11)  0.27  0.82 (0.45 to 1.45)  0.41  High C-reactive protein (≥3.0 mg l−1)  0.82 (0.57 to 1.19)  0.29  1.34 (0.87 to 2.06)  0.24  Hyperuricaemia (≥2.3 mmol l−1)  0.78 (0.44 to 1.41)  0.41  0.73 (0.39 to 1.38)  0.32  Low eGFR (<90 ml min−1/1.73 m2)  1.00 (0.66 to 1.52)  1.00  1.16 (0.74 to 1.82)  0.57    β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.09 (−0.34 to 0.52)  0.69  0.56 (0.10 to 1.03)  0.02  Waist circumference (cm)  0.04 (−1.4 to 1.5)  0.96  1.1 (−0.5 to 2.7)  0.18  Systolic blood pressure (mmHg)b  −0.7 (−2.5 to 1.1)  0.43  −0.6 (−2.6 to 1.4)  0.53  Diastolic blood pressure (mmHg)b  −0.6 (−1.7 to 0.6)  0.33  −0.7 (−2.0 to 0.6)  0.28  Total cholesterol (mmol l−1)b  −0.14 (−0.27 to −0.01)  0.04  −0.03 (−0.18 to 0.12)  0.69  HDL cholesterol (mmol l−1)b  −0.02 (−0.06 to 0.03)  0.52  −0.01 (−0.06 to 0.04)  0.78  Triglycerides (log mmol l−1)b  −0.06 (−0.12 to 0.01)  0.11  0.01 (−0.07 to 0.09)  0.84  γ-Glutamyltransferase (log U l−1)  0.001 (−0.08 to 0.08)  0.99  −0.02 (−0.11 to 0.08)  0.71  C-reactive protein (log mg l−1)  −0.09 (−0.23 to 0.04)  0.19  0.01 (−0.14 to 0.17)  0.87  Uric acid (mg dl−1)  0.01 (−0.14 to 0.16)  0.89  0.11 (−0.06 to 0.28)  0.21  eGFR (ml min−1/1.73 m2)  −0.5 (−2.2 to 1.2)  0.57  −1.3 (−3.2 to 0.6)  0.18  a Analyses were adjusted for age, gender, educational level, and chronotype. b Systolic and diastolic blood pressure were additionally adjusted for anti-hypertensive medication, total cholesterol, HDL cholesterol and triglycerides for cholesterol-lowering medication. eGFR, estimated glomerular filtration rate, OR, odds ratio; CI, confidence interval; β, regression coefficient. Boldface indicates statistical significance (P < 0.05). Table 3 Multivariable-adjusted differences in cardiometabolic risk factors between day workers, former shift workers, and shift workers   Former shift workers versus day workers   Shift workers versus day workers   OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.00 (0.76 to 1.32)  0.99  1.44 (1.06 to 1.95)  0.02  Abdominal obesity (waist circumference ≥88/102)  1.24 (0.91 to 1.68)  0.17  1.26 (0.89 to 1.79)  0.18  Hypertension (≥90/140 mmHg)  1.05 (0.78 to 1.40)  0.77  0.99 (0.68 to 1.43)  0.84  Hypercholesterolaemia (≥6.0 mmol l−1)  0.84 (0.64 to 1.11)  0.84  1.00 (0.73 to 1.38)  0.99  Low HDL cholesterol (<0.9 mmol l−1)  1.07 (0.68 to 1.69)  0.76  0.86 (0.52 to 1.45)  0.57  Hypertriglyceridaemia (2.3 mmol l−1)  1.09 (0.72 to 1.64)  0.69  0.96 (0.57 to 1.61)  0.92  Type 2 diabetes (≥11.1 mmol l−1 or self-reported)  1.72 (0.71 to 4.17)  0.23  1.58 (0.58 to 4.30)  0.38  High γ-glutamyltransferase (≥50 U l−1)  1.31 (0.81 to 2.11)  0.27  0.82 (0.45 to 1.45)  0.41  High C-reactive protein (≥3.0 mg l−1)  0.82 (0.57 to 1.19)  0.29  1.34 (0.87 to 2.06)  0.24  Hyperuricaemia (≥2.3 mmol l−1)  0.78 (0.44 to 1.41)  0.41  0.73 (0.39 to 1.38)  0.32  Low eGFR (<90 ml min−1/1.73 m2)  1.00 (0.66 to 1.52)  1.00  1.16 (0.74 to 1.82)  0.57    β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.09 (−0.34 to 0.52)  0.69  0.56 (0.10 to 1.03)  0.02  Waist circumference (cm)  0.04 (−1.4 to 1.5)  0.96  1.1 (−0.5 to 2.7)  0.18  Systolic blood pressure (mmHg)b  −0.7 (−2.5 to 1.1)  0.43  −0.6 (−2.6 to 1.4)  0.53  Diastolic blood pressure (mmHg)b  −0.6 (−1.7 to 0.6)  0.33  −0.7 (−2.0 to 0.6)  0.28  Total cholesterol (mmol l−1)b  −0.14 (−0.27 to −0.01)  0.04  −0.03 (−0.18 to 0.12)  0.69  HDL cholesterol (mmol l−1)b  −0.02 (−0.06 to 0.03)  0.52  −0.01 (−0.06 to 0.04)  0.78  Triglycerides (log mmol l−1)b  −0.06 (−0.12 to 0.01)  0.11  0.01 (−0.07 to 0.09)  0.84  γ-Glutamyltransferase (log U l−1)  0.001 (−0.08 to 0.08)  0.99  −0.02 (−0.11 to 0.08)  0.71  C-reactive protein (log mg l−1)  −0.09 (−0.23 to 0.04)  0.19  0.01 (−0.14 to 0.17)  0.87  Uric acid (mg dl−1)  0.01 (−0.14 to 0.16)  0.89  0.11 (−0.06 to 0.28)  0.21  eGFR (ml min−1/1.73 m2)  −0.5 (−2.2 to 1.2)  0.57  −1.3 (−3.2 to 0.6)  0.18    Former shift workers versus day workers   Shift workers versus day workers   OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.00 (0.76 to 1.32)  0.99  1.44 (1.06 to 1.95)  0.02  Abdominal obesity (waist circumference ≥88/102)  1.24 (0.91 to 1.68)  0.17  1.26 (0.89 to 1.79)  0.18  Hypertension (≥90/140 mmHg)  1.05 (0.78 to 1.40)  0.77  0.99 (0.68 to 1.43)  0.84  Hypercholesterolaemia (≥6.0 mmol l−1)  0.84 (0.64 to 1.11)  0.84  1.00 (0.73 to 1.38)  0.99  Low HDL cholesterol (<0.9 mmol l−1)  1.07 (0.68 to 1.69)  0.76  0.86 (0.52 to 1.45)  0.57  Hypertriglyceridaemia (2.3 mmol l−1)  1.09 (0.72 to 1.64)  0.69  0.96 (0.57 to 1.61)  0.92  Type 2 diabetes (≥11.1 mmol l−1 or self-reported)  1.72 (0.71 to 4.17)  0.23  1.58 (0.58 to 4.30)  0.38  High γ-glutamyltransferase (≥50 U l−1)  1.31 (0.81 to 2.11)  0.27  0.82 (0.45 to 1.45)  0.41  High C-reactive protein (≥3.0 mg l−1)  0.82 (0.57 to 1.19)  0.29  1.34 (0.87 to 2.06)  0.24  Hyperuricaemia (≥2.3 mmol l−1)  0.78 (0.44 to 1.41)  0.41  0.73 (0.39 to 1.38)  0.32  Low eGFR (<90 ml min−1/1.73 m2)  1.00 (0.66 to 1.52)  1.00  1.16 (0.74 to 1.82)  0.57    β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.09 (−0.34 to 0.52)  0.69  0.56 (0.10 to 1.03)  0.02  Waist circumference (cm)  0.04 (−1.4 to 1.5)  0.96  1.1 (−0.5 to 2.7)  0.18  Systolic blood pressure (mmHg)b  −0.7 (−2.5 to 1.1)  0.43  −0.6 (−2.6 to 1.4)  0.53  Diastolic blood pressure (mmHg)b  −0.6 (−1.7 to 0.6)  0.33  −0.7 (−2.0 to 0.6)  0.28  Total cholesterol (mmol l−1)b  −0.14 (−0.27 to −0.01)  0.04  −0.03 (−0.18 to 0.12)  0.69  HDL cholesterol (mmol l−1)b  −0.02 (−0.06 to 0.03)  0.52  −0.01 (−0.06 to 0.04)  0.78  Triglycerides (log mmol l−1)b  −0.06 (−0.12 to 0.01)  0.11  0.01 (−0.07 to 0.09)  0.84  γ-Glutamyltransferase (log U l−1)  0.001 (−0.08 to 0.08)  0.99  −0.02 (−0.11 to 0.08)  0.71  C-reactive protein (log mg l−1)  −0.09 (−0.23 to 0.04)  0.19  0.01 (−0.14 to 0.17)  0.87  Uric acid (mg dl−1)  0.01 (−0.14 to 0.16)  0.89  0.11 (−0.06 to 0.28)  0.21  eGFR (ml min−1/1.73 m2)  −0.5 (−2.2 to 1.2)  0.57  −1.3 (−3.2 to 0.6)  0.18  a Analyses were adjusted for age, gender, educational level, and chronotype. b Systolic and diastolic blood pressure were additionally adjusted for anti-hypertensive medication, total cholesterol, HDL cholesterol and triglycerides for cholesterol-lowering medication. eGFR, estimated glomerular filtration rate, OR, odds ratio; CI, confidence interval; β, regression coefficient. Boldface indicates statistical significance (P < 0.05). Subgroup analysis by chronotype Among evening chronotypes, shift work was related with higher BMI (β: 0.97 kg m−2, 95% CI 0.21–1.73), which was not the case among morning or intermediate chronotypes (β: 0.04 kg m−2, 95% CI −0.85 to 0.93; respectively, 0.72, 95% CI −0.03 to 1.48). Shift work was not significantly related with other cardiometabolic risk factors in any of the chronotype strata (table 4). Though not statistically significant, in morning chronotypes, shift workers had an odds ratio for type 2 diabetes of 5.50 (95% CI 0.86–35.11). Table 4 Differences in cardiometabolic risk factors between day workers and shift workers, stratified by chronotype   Morning chronotype   Intermediate chronotype   Evening chronotype   OR 95% CI a  P-value  OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.27 (0.72 to 2.22)  0.41  1.65 (0.96 to 2.82)  0.07  1.46 (0.88 to 2.43)  0.15  Abdominal obesity (waist circumference ≥88/102)  0.99 (0.53 to 1.87)  0.98  1.30 (0.71 to 2.37)  0.39  1.51 (0.88 to 2.61)  0.13  Hypertension (≥90/140 mmHg)  0.94 (0.47 to 1.86)  0.86  0.91 (0.47 to 1.79)  0.79  1.01 (0.57 to 1.77)  0.86  Hypercholesterolaemia (≥6.0 mmol l−1)  1.12 (0.64 to 1.97)  0.69  0.78 (0.46 to 1.32)  0.35  1.15 (0.68 to 1.96)  0.56  Low HDL cholesterol (<0.9 mmol l−1)  1.43 (0.61 to 3.36)  0.41  0.51 (0.17 to 1.51)  0.22  0.84 (0.40 to 1.79)  0.65  Hypertriglyceridaemia (2.3 mmol l−1)  1.35 (0.59 to 3.10)  0.48  0.79 (0.29 to 2.15)  0.65  0.90 (0.39 to 2.06)  0.80  Type 2 diabetes (≥11.1 mmol l−1 or self to reported)  5.50 (0.86 to 35.11)  0.07  0.97 (0.11 to 8.18)  0.98  1.20 (0.29 to 5.01)  0.65  High γ-glutamyltransferase (≥50 U l−1)  1.47 (0.55 to 3.90)  0.44  0.71 (0.23 to 2.16)  0.55  0.44 (0.16 to 1.23)  0.12  High C-reactive protein (≥3.0 mg l−1)  1.30 (0.59 to 2.83)  0.52  1.03 (0.43 to 2.42)  0.96  1.47 (0.79 to 2.73)  0.23  Hyperuricaemia (≥2.3 mmol l−1)  0.15 (0.02 to 1.21)  0.08  0.83 (0.29 to 2.42)  0.73  0.94 (0.38 to 2.33)  0.97  Low eGFR (<90 ml min−1/1.73 m2)  0.92 (0.40 to 2.11)  0.84  0.74 (0.30 to 1.80)  0.50  1.93 (0.92 to 4.03)  0.09    β 95% CIa  β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.04 (−0.85 to 0.93)  0.93  0.72 (−0.03 to 1.48)  0.06  0.97 (0.21 to 1.73)  0.03  Waist circumference (cm)  −0.5 (−3.5 to 2.5)  0.75  1.2 (−1.4 to 3.9)  0.37  2.4 (−0.3 to 5.1)  0.10  Systolic blood pressure (mmHg)b  −0.03 (−3.6 to 3.5)  0.99  −1.9 (−5.3 to 1.6)  0.28  −0.2 (−3.7 to 3.2)  0.87  Diastolic blood pressure (mmHg)b  0.02 (−2.3 to 2.4)  0.98  −1.2 (−3.4 to 1.0)  0.29  −0.8 (−3.1 to 1.4)  0.42  Total cholesterol (mmol l−1)b  0.05 (−0.22 to 0.32)  0.71  −0.06 (−0.31 to 0.19)  0.63  −0.06 (−0.31 to 0.20)  0.72  HDL cholesterol (mmol l−1)b  −0.03 (−0.12 to 0.06)  0.46  0.02 (−0.07 to 0.11)  0.69  0.01 (−0.07 to 0.08)  0.88  Triglycerides (log mmol l−1)b  0.10 (−0.03 to 0.24)  0.13  −0.07 (−0.21 to 0.07)  0.30  0.002 (−0.14 to 0.14)  0.94  GGT (log U l−1)  0.10 (−0.07 to 0.26)  0.24  −0.02 (−0.18 to 0.15)  0.83  −0.13 (−0.30 to 0.04)  0.28  CRP (log mg l−1)  0.00 (−0.27 to 0.27)  1.00  −0.06 (−0.33 to 0.22)  0.68  0.10 (−0.18 to 0.37)  0.48  Uric acid (mg dl−1)  0.02 (−0.27 to 0.30)  0.91  0.11 (−0.18 to 0.40)  0.45  0.14 (−0.16 to 0.44)  0.19  eGFR (ml min−1/1.73 m2)  0.8 (−2.5 to 4.1)  0.62  −1.9 (−5.5 to 1.6)  0.28  −2.0 (−5.1 to 1.1)  0.11    Morning chronotype   Intermediate chronotype   Evening chronotype   OR 95% CI a  P-value  OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.27 (0.72 to 2.22)  0.41  1.65 (0.96 to 2.82)  0.07  1.46 (0.88 to 2.43)  0.15  Abdominal obesity (waist circumference ≥88/102)  0.99 (0.53 to 1.87)  0.98  1.30 (0.71 to 2.37)  0.39  1.51 (0.88 to 2.61)  0.13  Hypertension (≥90/140 mmHg)  0.94 (0.47 to 1.86)  0.86  0.91 (0.47 to 1.79)  0.79  1.01 (0.57 to 1.77)  0.86  Hypercholesterolaemia (≥6.0 mmol l−1)  1.12 (0.64 to 1.97)  0.69  0.78 (0.46 to 1.32)  0.35  1.15 (0.68 to 1.96)  0.56  Low HDL cholesterol (<0.9 mmol l−1)  1.43 (0.61 to 3.36)  0.41  0.51 (0.17 to 1.51)  0.22  0.84 (0.40 to 1.79)  0.65  Hypertriglyceridaemia (2.3 mmol l−1)  1.35 (0.59 to 3.10)  0.48  0.79 (0.29 to 2.15)  0.65  0.90 (0.39 to 2.06)  0.80  Type 2 diabetes (≥11.1 mmol l−1 or self to reported)  5.50 (0.86 to 35.11)  0.07  0.97 (0.11 to 8.18)  0.98  1.20 (0.29 to 5.01)  0.65  High γ-glutamyltransferase (≥50 U l−1)  1.47 (0.55 to 3.90)  0.44  0.71 (0.23 to 2.16)  0.55  0.44 (0.16 to 1.23)  0.12  High C-reactive protein (≥3.0 mg l−1)  1.30 (0.59 to 2.83)  0.52  1.03 (0.43 to 2.42)  0.96  1.47 (0.79 to 2.73)  0.23  Hyperuricaemia (≥2.3 mmol l−1)  0.15 (0.02 to 1.21)  0.08  0.83 (0.29 to 2.42)  0.73  0.94 (0.38 to 2.33)  0.97  Low eGFR (<90 ml min−1/1.73 m2)  0.92 (0.40 to 2.11)  0.84  0.74 (0.30 to 1.80)  0.50  1.93 (0.92 to 4.03)  0.09    β 95% CIa  β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.04 (−0.85 to 0.93)  0.93  0.72 (−0.03 to 1.48)  0.06  0.97 (0.21 to 1.73)  0.03  Waist circumference (cm)  −0.5 (−3.5 to 2.5)  0.75  1.2 (−1.4 to 3.9)  0.37  2.4 (−0.3 to 5.1)  0.10  Systolic blood pressure (mmHg)b  −0.03 (−3.6 to 3.5)  0.99  −1.9 (−5.3 to 1.6)  0.28  −0.2 (−3.7 to 3.2)  0.87  Diastolic blood pressure (mmHg)b  0.02 (−2.3 to 2.4)  0.98  −1.2 (−3.4 to 1.0)  0.29  −0.8 (−3.1 to 1.4)  0.42  Total cholesterol (mmol l−1)b  0.05 (−0.22 to 0.32)  0.71  −0.06 (−0.31 to 0.19)  0.63  −0.06 (−0.31 to 0.20)  0.72  HDL cholesterol (mmol l−1)b  −0.03 (−0.12 to 0.06)  0.46  0.02 (−0.07 to 0.11)  0.69  0.01 (−0.07 to 0.08)  0.88  Triglycerides (log mmol l−1)b  0.10 (−0.03 to 0.24)  0.13  −0.07 (−0.21 to 0.07)  0.30  0.002 (−0.14 to 0.14)  0.94  GGT (log U l−1)  0.10 (−0.07 to 0.26)  0.24  −0.02 (−0.18 to 0.15)  0.83  −0.13 (−0.30 to 0.04)  0.28  CRP (log mg l−1)  0.00 (−0.27 to 0.27)  1.00  −0.06 (−0.33 to 0.22)  0.68  0.10 (−0.18 to 0.37)  0.48  Uric acid (mg dl−1)  0.02 (−0.27 to 0.30)  0.91  0.11 (−0.18 to 0.40)  0.45  0.14 (−0.16 to 0.44)  0.19  eGFR (ml min−1/1.73 m2)  0.8 (−2.5 to 4.1)  0.62  −1.9 (−5.5 to 1.6)  0.28  −2.0 (−5.1 to 1.1)  0.11  GGT, γ-glutamyltransferase; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate, OR, odds ratio; CI, confidence interval; β, regression coefficient. Boldface indicates statistical significance (P < 0.05). a Analyses were adjusted for age, gender, and educational level. b Systolic and diastolic blood pressure were additionally adjusted for anti-hypertensive medication, total cholesterol, HDL cholesterol and triglycerides for cholesterol-lowering medication. Table 4 Differences in cardiometabolic risk factors between day workers and shift workers, stratified by chronotype   Morning chronotype   Intermediate chronotype   Evening chronotype   OR 95% CI a  P-value  OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.27 (0.72 to 2.22)  0.41  1.65 (0.96 to 2.82)  0.07  1.46 (0.88 to 2.43)  0.15  Abdominal obesity (waist circumference ≥88/102)  0.99 (0.53 to 1.87)  0.98  1.30 (0.71 to 2.37)  0.39  1.51 (0.88 to 2.61)  0.13  Hypertension (≥90/140 mmHg)  0.94 (0.47 to 1.86)  0.86  0.91 (0.47 to 1.79)  0.79  1.01 (0.57 to 1.77)  0.86  Hypercholesterolaemia (≥6.0 mmol l−1)  1.12 (0.64 to 1.97)  0.69  0.78 (0.46 to 1.32)  0.35  1.15 (0.68 to 1.96)  0.56  Low HDL cholesterol (<0.9 mmol l−1)  1.43 (0.61 to 3.36)  0.41  0.51 (0.17 to 1.51)  0.22  0.84 (0.40 to 1.79)  0.65  Hypertriglyceridaemia (2.3 mmol l−1)  1.35 (0.59 to 3.10)  0.48  0.79 (0.29 to 2.15)  0.65  0.90 (0.39 to 2.06)  0.80  Type 2 diabetes (≥11.1 mmol l−1 or self to reported)  5.50 (0.86 to 35.11)  0.07  0.97 (0.11 to 8.18)  0.98  1.20 (0.29 to 5.01)  0.65  High γ-glutamyltransferase (≥50 U l−1)  1.47 (0.55 to 3.90)  0.44  0.71 (0.23 to 2.16)  0.55  0.44 (0.16 to 1.23)  0.12  High C-reactive protein (≥3.0 mg l−1)  1.30 (0.59 to 2.83)  0.52  1.03 (0.43 to 2.42)  0.96  1.47 (0.79 to 2.73)  0.23  Hyperuricaemia (≥2.3 mmol l−1)  0.15 (0.02 to 1.21)  0.08  0.83 (0.29 to 2.42)  0.73  0.94 (0.38 to 2.33)  0.97  Low eGFR (<90 ml min−1/1.73 m2)  0.92 (0.40 to 2.11)  0.84  0.74 (0.30 to 1.80)  0.50  1.93 (0.92 to 4.03)  0.09    β 95% CIa  β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.04 (−0.85 to 0.93)  0.93  0.72 (−0.03 to 1.48)  0.06  0.97 (0.21 to 1.73)  0.03  Waist circumference (cm)  −0.5 (−3.5 to 2.5)  0.75  1.2 (−1.4 to 3.9)  0.37  2.4 (−0.3 to 5.1)  0.10  Systolic blood pressure (mmHg)b  −0.03 (−3.6 to 3.5)  0.99  −1.9 (−5.3 to 1.6)  0.28  −0.2 (−3.7 to 3.2)  0.87  Diastolic blood pressure (mmHg)b  0.02 (−2.3 to 2.4)  0.98  −1.2 (−3.4 to 1.0)  0.29  −0.8 (−3.1 to 1.4)  0.42  Total cholesterol (mmol l−1)b  0.05 (−0.22 to 0.32)  0.71  −0.06 (−0.31 to 0.19)  0.63  −0.06 (−0.31 to 0.20)  0.72  HDL cholesterol (mmol l−1)b  −0.03 (−0.12 to 0.06)  0.46  0.02 (−0.07 to 0.11)  0.69  0.01 (−0.07 to 0.08)  0.88  Triglycerides (log mmol l−1)b  0.10 (−0.03 to 0.24)  0.13  −0.07 (−0.21 to 0.07)  0.30  0.002 (−0.14 to 0.14)  0.94  GGT (log U l−1)  0.10 (−0.07 to 0.26)  0.24  −0.02 (−0.18 to 0.15)  0.83  −0.13 (−0.30 to 0.04)  0.28  CRP (log mg l−1)  0.00 (−0.27 to 0.27)  1.00  −0.06 (−0.33 to 0.22)  0.68  0.10 (−0.18 to 0.37)  0.48  Uric acid (mg dl−1)  0.02 (−0.27 to 0.30)  0.91  0.11 (−0.18 to 0.40)  0.45  0.14 (−0.16 to 0.44)  0.19  eGFR (ml min−1/1.73 m2)  0.8 (−2.5 to 4.1)  0.62  −1.9 (−5.5 to 1.6)  0.28  −2.0 (−5.1 to 1.1)  0.11    Morning chronotype   Intermediate chronotype   Evening chronotype   OR 95% CI a  P-value  OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.27 (0.72 to 2.22)  0.41  1.65 (0.96 to 2.82)  0.07  1.46 (0.88 to 2.43)  0.15  Abdominal obesity (waist circumference ≥88/102)  0.99 (0.53 to 1.87)  0.98  1.30 (0.71 to 2.37)  0.39  1.51 (0.88 to 2.61)  0.13  Hypertension (≥90/140 mmHg)  0.94 (0.47 to 1.86)  0.86  0.91 (0.47 to 1.79)  0.79  1.01 (0.57 to 1.77)  0.86  Hypercholesterolaemia (≥6.0 mmol l−1)  1.12 (0.64 to 1.97)  0.69  0.78 (0.46 to 1.32)  0.35  1.15 (0.68 to 1.96)  0.56  Low HDL cholesterol (<0.9 mmol l−1)  1.43 (0.61 to 3.36)  0.41  0.51 (0.17 to 1.51)  0.22  0.84 (0.40 to 1.79)  0.65  Hypertriglyceridaemia (2.3 mmol l−1)  1.35 (0.59 to 3.10)  0.48  0.79 (0.29 to 2.15)  0.65  0.90 (0.39 to 2.06)  0.80  Type 2 diabetes (≥11.1 mmol l−1 or self to reported)  5.50 (0.86 to 35.11)  0.07  0.97 (0.11 to 8.18)  0.98  1.20 (0.29 to 5.01)  0.65  High γ-glutamyltransferase (≥50 U l−1)  1.47 (0.55 to 3.90)  0.44  0.71 (0.23 to 2.16)  0.55  0.44 (0.16 to 1.23)  0.12  High C-reactive protein (≥3.0 mg l−1)  1.30 (0.59 to 2.83)  0.52  1.03 (0.43 to 2.42)  0.96  1.47 (0.79 to 2.73)  0.23  Hyperuricaemia (≥2.3 mmol l−1)  0.15 (0.02 to 1.21)  0.08  0.83 (0.29 to 2.42)  0.73  0.94 (0.38 to 2.33)  0.97  Low eGFR (<90 ml min−1/1.73 m2)  0.92 (0.40 to 2.11)  0.84  0.74 (0.30 to 1.80)  0.50  1.93 (0.92 to 4.03)  0.09    β 95% CIa  β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.04 (−0.85 to 0.93)  0.93  0.72 (−0.03 to 1.48)  0.06  0.97 (0.21 to 1.73)  0.03  Waist circumference (cm)  −0.5 (−3.5 to 2.5)  0.75  1.2 (−1.4 to 3.9)  0.37  2.4 (−0.3 to 5.1)  0.10  Systolic blood pressure (mmHg)b  −0.03 (−3.6 to 3.5)  0.99  −1.9 (−5.3 to 1.6)  0.28  −0.2 (−3.7 to 3.2)  0.87  Diastolic blood pressure (mmHg)b  0.02 (−2.3 to 2.4)  0.98  −1.2 (−3.4 to 1.0)  0.29  −0.8 (−3.1 to 1.4)  0.42  Total cholesterol (mmol l−1)b  0.05 (−0.22 to 0.32)  0.71  −0.06 (−0.31 to 0.19)  0.63  −0.06 (−0.31 to 0.20)  0.72  HDL cholesterol (mmol l−1)b  −0.03 (−0.12 to 0.06)  0.46  0.02 (−0.07 to 0.11)  0.69  0.01 (−0.07 to 0.08)  0.88  Triglycerides (log mmol l−1)b  0.10 (−0.03 to 0.24)  0.13  −0.07 (−0.21 to 0.07)  0.30  0.002 (−0.14 to 0.14)  0.94  GGT (log U l−1)  0.10 (−0.07 to 0.26)  0.24  −0.02 (−0.18 to 0.15)  0.83  −0.13 (−0.30 to 0.04)  0.28  CRP (log mg l−1)  0.00 (−0.27 to 0.27)  1.00  −0.06 (−0.33 to 0.22)  0.68  0.10 (−0.18 to 0.37)  0.48  Uric acid (mg dl−1)  0.02 (−0.27 to 0.30)  0.91  0.11 (−0.18 to 0.40)  0.45  0.14 (−0.16 to 0.44)  0.19  eGFR (ml min−1/1.73 m2)  0.8 (−2.5 to 4.1)  0.62  −1.9 (−5.5 to 1.6)  0.28  −2.0 (−5.1 to 1.1)  0.11  GGT, γ-glutamyltransferase; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate, OR, odds ratio; CI, confidence interval; β, regression coefficient. Boldface indicates statistical significance (P < 0.05). a Analyses were adjusted for age, gender, and educational level. b Systolic and diastolic blood pressure were additionally adjusted for anti-hypertensive medication, total cholesterol, HDL cholesterol and triglycerides for cholesterol-lowering medication. Frequency of night shifts and duration of shift work No exposure–response relationships between shift work and cardiometabolic risk factors were observed by frequency of night shifts (Supplementary table S1). Results were also similar for shift workers who worked for <10 years and those who work for ≥10 years in irregular shifts (Supplementary table S2). With the exception of overweight, those who worked <10 years in irregular shifts had no increased risk of being overweight compared with day workers (OR: 1.19, 95% CI 0.80–1.78), while those who worked for ≥10 years in irregular shifts had an OR for being overweight of 1.56 (95% CI 1.08–2.24). Discussion Shift work was not related with an increased risk of cardiometabolic risk factors, except for overweight/BMI. Shift workers had a 44% higher risk of being overweight and had a 0.58 kg m−2 higher BMI than day workers. The increased risk for overweight was particularly present among shift working evening chronotypes. Our research findings do not confirm our hypothesis that shift work is related with cardiometabolic risk factors other than overweight. The few existing previous studies also showed inconsistent findings, with some finding relations between shift work and cardiometabolic risk factors while others could not confirm such findings.7–9,25 The absence of relations between shift work and cardiometabolic risk factors may partly be due to limitation in the design of our study. A large part of the study population dropped out due to the long follow-up period of up to 26 years or were excluded due to missing data. These excluded individuals were lower educated and more often men than the included individuals, but levels of cardiometabolic risk factors at the first examination wave were only slightly more unfavourable. Since shift work is more common in adults of lower socio-economic status and in men, exclusion may have mitigated the relations between shift work and cardiometabolic risk factors. In addition, we lack details about the shifts schedule, such as direction and speed of rotation. It is therefore possible that particular types of shifts are related with metabolic risk factors in our population. However, we did not find dose-response relations between duration of shift work and frequency of night shifts and cardiometabolic risk factors, which are considered to be the most important characteristics of shift work related to adverse health outcomes.16 Thus, it might also be that shift work does not increase the risk of the investigated cardiometabolic risk factors as shown in some other studies.7–9 This means that other determinants mediate the relation between shift work and CVD, such as lifestyle behaviours, psychosocial determinants and other physiological determinants including disturbed autonomic regulation of the cardiovascular system.4,26 In line with other studies, we observed that shift work was related with overweight and BMI.7 We extended those findings and showed that this may be particularly the case for evening chronotypes, and that frequency of night shifts may not influence the risk of overweight. One reason may be differences in average behaviour and timing of food intake and physical activity between day and shift workers. In our study, however, physical activity and dietary patterns did not seem to play a role as adjustment for smoking, alcohol, and physical activity did not attenuate the relations. We did not adjust the primary analyses for diet because this was only measured at Waves 2–4. In sensitivity analyses based on Waves 2–4, analysis was adjusted for energy intake and the Mediterranean diet score. This did not attenuate any of the relations (data not shown). Another potential reason is the disruption of circadian rhythms, which interacts with hormones, such as insulin, leptin, cortisol, growth hormone, and melatonin, and thereby increases the risk of overweight.27–29 Finally, other plausible explanations are more exposure to light at night, less exposure to day light and possibly sleep deprivation related to shift work.27,30,31 Reasons for the particular increased risk among evening chronotypes are unclear. As described before, there is a scarcity on research to the role of chronotype in the shift work related health effects. Of the limited research, contrasting results have been observed with a higher risk of shift work for type 2 diabetes among morning and not among evening chronotypes.10 Future research is needed to understand mechanisms for the increased risk of overweight among shift workers by chronotype. Previous studies showed contrasting findings regarding the relation between shift work and blood glucose, HbA1c and type 2 diabetes, with studies finding a positive association22,32–34 and other studies not.24,35 We did not find a significant relationship between shift work and type 2 diabetes, but the OR pointed, in particular among morning chronotypes (OR: 5.50, 95% CI 0.86–35.11), to an increased risk of type 2 diabetes. The non-significance of this OR is likely due to limited statistical power of the analyses within each chronotype. A recent study observed that among female morning chronotypes, the risk of type 2 diabetes increased with more years exposure to rotating night shifts, while the opposite was true for female evening chronotypes.10 The possible moderation of chronotype may be because the shift schedule, especially night shifts, interferes more strongly with the circadian rhythm among morning chronotypes than evening chronotypes. This may lead to unfavourable biological processes, such as secretion of hormones and sleep problems, which increases the risk of type 2 diabetes.28,29,36,37 However, more research with a larger sample is needed on the moderating effect of chronotype in the relation between shift work and type 2 diabetes before strong conclusions can be drawn. A strength of this study includes the wide variety of objectively measured cardiometabolic risk factors. We were also able to investigate exposure–response relationships with frequency of night shifts and differences across chronotypes. Besides the earlier mentioned limitations, our results are vulnerable for finding false statistically significant relations by chance because we related shift work to 11 cardiometabolic risk factors. We only found shift work to statistically significantly increase the risk of overweight/BMI. Although this could be a spurious finding, it is in line with previous studies and can be explained by several mechanisms, such as an unhealthy lifestyle and hormonal changes.7,27–31 Therefore, this relation is unlikely to be a false positive finding. In addition, it is often argued to use change scores to analyze prospective relations. However, the error of change scores are large,38,39 and may even be larger than the between-subject difference. We therefore used a model in which we compared levels of cardiometabolic risk factors at several time points, and its association with shift work status (within an individual) one year before. Furthermore, shift work status was assessed retrospectively, which makes our results susceptive to information bias. However, we believe the participants were able to reliably recall their shift work exposure as being a shift worker has a significant impact on someone’s life and is relatively easy to remember. Finally, although our single question on chronotype is in excellent agreement with the validated and more quantitatively assessed chronotype of the MCTQ,17 we may have misclassified some participants. If so, this would have probably led to a small underestimation of differences in associations between shift work and cardiometabolic risk factors by chronotype. In conclusion, shift workers had a 44% higher risk of being overweight and a 0.56 kg m−2 higher BMI than day workers, but the risk of other cardiometabolic risk factors did not differ between groups. Shift working evening chronotypes had a particular higher BMI compared with day working evening chronotypes. The increased risk of type 2 diabetes of shift work among morning chronotypes in the study of Vetter et al.10 could not be replicated in our study, although the OR pointed in the same unfavourable direction. Thus, more research is needed on the moderating effects of chronotype to establish whether tailored preventive measures by chronotype are useful for shift workers. Supplementary data Supplementary data are available at EURPUB online. Acknowledgements The authors would like to thank the epidemiologists and data managers at the Dutch National Institute of Public Health and Environment, and the fieldworkers of the Municipal Health Service in Doetinchem for their contribution to the data collection for this study. Funding This research was funded by the ReVanche Program of the EMGO+ Institute for Health and Care Research of the VU University Medical Center, Amsterdam (grant number: 9729). This work was also supported by The Netherlands Organization for Health Research (ZonMW) within the program Electromagnetic Fields and Health Research under grant numbers 85200001 and 85800001. The Doetinchem Cohort Study is funded by the National Institute for Public Health and the Environment, which works under the authority of the Ministry of Health, Welfare and Sport of the Netherlands. Conflicts of interest: None declared. Key points Shift workers have an increased risk of being overweight and higher body mass index compared with day workers. Chronotype is a moderator in the association between shift work and risk of overweight; the association is especially strong in evening chronotypes. In none of the chronotypes is shift work associated with an increased risk of other cardiometabolic risk factors. References 1 U.S. Department of Labor, Bureau of Labor Statistics. Employment and Wages, Annual Averages 2004. Washington, U.S., 2008. Available at: http://www.bls.gov/cew/cewbultn04.htm (14 March 2016, date last accessed). 2 Trost SG, Owen N, Bauman AE, et al.   Correlates of adults' participation in physical activity: review and update. Med Sci Sports Exerc  2002; 34: 1996– 2001. Google Scholar CrossRef Search ADS PubMed  3 Gan Y, Yang C, Tong X, et al.   Shift work and diabetes mellitus: a meta-analysis of observational studies. Occup Environ Med  2015; 72: 72– 8. Google Scholar CrossRef Search ADS PubMed  4 Puttonen S, Harma M, Hublin C. Shift work and cardiovascular disease - pathways from circadian stress to morbidity. Scand J Work Environ Health  2010; 36: 96– 108. Google Scholar CrossRef Search ADS PubMed  5 Vyas MV, Garg AX, Iansavichus AV, et al.   Shift work and vascular events: systematic review and meta-analysis. BMJ  (Clin Res Ed) 2012; 345: e4800. Google Scholar CrossRef Search ADS PubMed  6 Wang F, Yeung KL, Chan WC, et al.   A meta-analysis on dose-response relationship between night shift work and the risk of breast cancer. Ann Oncol  2013; 24: 2724– 32. Google Scholar CrossRef Search ADS PubMed  7 Proper KI, van de Langenberg D, Rodenburg W, et al.   The relationship between shift work and metabolic risk factors: a systematic review of longitudinal studies. Am J Prev Med  2016; 50: e147– 57. Google Scholar CrossRef Search ADS PubMed  8 Puttonen S, Viitasalo K, Harma M. Effect of shiftwork on systemic markers of inflammation. Chronobiol Int  2011; 28: 528– 35. Google Scholar CrossRef Search ADS PubMed  9 Sookoian S, Gemma C, Fernandez Gianotti T, et al.   Effects of rotating shift work on biomarkers of metabolic syndrome and inflammation. J Intern Med  2007; 261: 285– 92. Google Scholar CrossRef Search ADS PubMed  10 Vetter C, Devore EE, Ramin CA, et al.   Mismatch of sleep and work timing and risk of type 2 diabetes. Diabetes Care  2015; 38: 1707– 13. Google Scholar CrossRef Search ADS PubMed  11 Merikanto I, Lahti T, Puolijoki H, et al.   Associations of chronotype and sleep with cardiovascular diseases and type 2 diabetes. Chronobiol Int  2013; 30: 470– 7. Google Scholar CrossRef Search ADS PubMed  12 Osonoi Y, Mita T, Osonoi T, et al.   Morningness-eveningness questionnaire score and metabolic parameters in patients with type 2 diabetes mellitus. Chronobiol Int  2014; 31: 1017– 23. Google Scholar CrossRef Search ADS PubMed  13 Wong PM, Hasler BP, Kamarck TW, et al.   Social Jetlag, chronotype, and cardiometabolic risk. J Clin Endocrinol Metab  2015; 100: 4612– 20. Google Scholar CrossRef Search ADS PubMed  14 Yu JH, Yun CH, Ahn JH, et al.   Evening chronotype is associated with metabolic disorders and body composition in middle-aged adults. J Clin Endocrinol Metab  2015; 100: 1494– 502. Google Scholar CrossRef Search ADS PubMed  15 Verschuren WM, Blokstra A, Picavet HS, Smit HA. Cohort profile: the Doetinchem Cohort Study. Int J Epidemiol  2008; 37: 1236– 41. Google Scholar CrossRef Search ADS PubMed  16 Stevens RG, Hansen J, Costa G, et al.   Considerations of circadian impact for defining ‘shift work’ in cancer studies: iARC Working Group Report. Occup Environ Med  2011; 68: 154– 62. Google Scholar CrossRef Search ADS PubMed  17 Roenneberg T, Wirz-Justice A, Merrow M. Life between clocks: daily temporal patterns of human chronotypes. J Biol Rhythms  2003; 18: 80– 90. Google Scholar CrossRef Search ADS PubMed  18 Inker LA, Schmid CH, Tighiouart H, et al.   Estimating glomerular filtration rate from serum creatinine and cystatin C. New Engl J Med  2012; 367: 20– 9. Google Scholar CrossRef Search ADS PubMed  19 Kemper H, Ooijendijk W, Stiggelbout M. Consensus about the Dutch physical activity guideline. Tijdschr Soc Geneeskd  2000; 78: 180– 3. 20 van Amelsvoort LG, Schouten EG, Kok FJ. Impact of one year of shift work on cardiovascular disease risk factors. J Occup Environ Med  2004; 46: 699– 706. Google Scholar CrossRef Search ADS PubMed  21 Twisk JWR. Applied Longitudinal Data Analysis for Epidemiology. A Practical Guide. Cambridge: Cambridge University Press, 2003. 22 De Bacquer D, Van Risseghem M, Clays E, et al.   Rotating shift work and the metabolic syndrome: a prospective study. Int J Epidemiol  2009; 38: 848– 54. Google Scholar CrossRef Search ADS PubMed  23 Guo Y, Rong Y, Huang X, et al.   Shift work and the relationship with metabolic syndrome in Chinese aged workers. PLoS One  2015; 10: e0120632. Google Scholar CrossRef Search ADS PubMed  24 Pan A, Schernhammer ES, Sun Q, Hu FB. Rotating night shift work and risk of type 2 diabetes: two prospective cohort studies in women. PLoS Med  2011; 8: e1001141. Google Scholar CrossRef Search ADS PubMed  25 Morikawa Y, Nakagawa H, Miura K, et al.   Effect of shift work on body mass index and metabolic parameters. Scand J Work Environ Health  2007; 33: 45– 50. Google Scholar CrossRef Search ADS PubMed  26 Knutsson A, Boggild H. Shiftwork and cardiovascular disease: review of disease mechanisms. Rev Environ Health  2000; 15: 359– 72. Google Scholar CrossRef Search ADS PubMed  27 Antunes LC, Levandovski R, Dantas G, et al.   Obesity and shift work: chronobiological aspects. Nutr Res Rev  2010; 23: 155– 68. Google Scholar CrossRef Search ADS PubMed  28 Bass J, Takahashi JS. Circadian integration of metabolism and energetics. Science  2010; 330: 1349– 54. Google Scholar CrossRef Search ADS PubMed  29 Shi SQ, Ansari TS, McGuinness OP, et al.   Circadian disruption leads to insulin resistance and obesity. Curr Biol  2013; 23: 372– 81. Google Scholar CrossRef Search ADS PubMed  30 McFadden E, Jones ME, Schoemaker MJ, et al.   The relationship between obesity and exposure to light at night: cross-sectional analyses of over 100,000 women in the Breakthrough Generations Study. Am J Epidemiol  2014; 180: 245– 50. Google Scholar CrossRef Search ADS PubMed  31 Patel SR, Hu FB. Short sleep duration and weight gain: a systematic review. Obesity (Silver Spring)  2008; 16: 643– 53. Google Scholar CrossRef Search ADS PubMed  32 Suwazono Y, Uetani M, Oishi M, et al.   Calculation of the benchmark duration of shift work associated with the development of impaired glucose metabolism: a 14-year cohort study on 7104 male workers. Occup Environ Med  2010; 67: 532– 7. Google Scholar CrossRef Search ADS PubMed  33 Li Y, Sato Y, Yamaguchi N. Shift work and the risk of metabolic syndrome: a nested case-control study. Int J Occup Environ Health  2011; 17: 154– 60. Google Scholar CrossRef Search ADS PubMed  34 Oyama I, Kubo T, Fujino Y, et al.   Retrospective cohort study of the risk of impaired glucose tolerance among shift workers. Scand J Work Environ Health  2012; 38: 337– 42. Google Scholar CrossRef Search ADS PubMed  35 Biggi N, Consonni D, Galluzzo V, et al.   Metabolic syndrome in permanent night workers. Chronobiol Int  2008; 25: 443– 54. Google Scholar CrossRef Search ADS PubMed  36 Scheer FA, Hilton MF, Mantzoros CS, Shea SA. Adverse metabolic and cardiovascular consequences of circadian misalignment. Proc Natl Acad Sci USA  2009; 106: 4453– 8. Google Scholar CrossRef Search ADS PubMed  37 Cappuccio FP, D'Elia L, Strazzullo P, Miller MA. Quantity and quality of sleep and incidence of type 2 diabetes: a systematic review and meta-analysis. Diabetes Care  2010; 33: 414– 20. Google Scholar CrossRef Search ADS PubMed  38 Cain KC, Kronmal RA, Kosinski AS. Analysing the relationship between change in a risk factor and risk of disease. Stat Med  1992; 11: 783– 97. Google Scholar CrossRef Search ADS PubMed  39 Twisk J, Proper K. Evaluation of the results of a randomized controlled trial: how to define changes between baseline and follow-up. J Clin Epidemiol  2004; 57: 223– 8. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The European Journal of Public Health Oxford University Press

Shift work, chronotype and the risk of cardiometabolic risk factors

Loading next page...
 
/lp/ou_press/shift-work-chronotype-and-the-risk-of-cardiometabolic-risk-factors-hOUccZOddA
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved.
ISSN
1101-1262
eISSN
1464-360X
D.O.I.
10.1093/eurpub/cky092
Publisher site
See Article on Publisher Site

Abstract

Abstract Background The relation between shift work and a large variety of cardiometabolic risk factors is unclear. Also, the role of chronotype is understudied. We examined relations between shift work and cardiometabolic risk factors, and explored these relations in different chronotypes. Methods Cardiometabolic risk factors (anthropometry, blood pressure, lipids, diabetes, γ-glutamyltransferase, C-reactive protein, uric acid and estimated glomerular filtration rate) were assessed among 1334 adults in 1987–91, with repeated measurements every 5 years. Using shift work history data collected in 2013–15, we identified shift work status 1 year prior to all six waves. Linear mixed models and logistic generalized estimating equations were used to estimate the longitudinal relations between shift work and risk factors 1 year later. Results Shift work was not significantly related with cardiometabolic risk factors (P ≥ 0.05), except for overweight/body mass index. Shift workers had more often overweight (OR: 1.44, 95% CI 1.06–1.95) and a higher body mass index (BMI) (β: 0.56 kg m−2, 95% CI 0.10–1.03) than day workers. A significant difference in BMI between day and shift workers was observed among evening chronotypes (β: 0.97 kg m−2, 95% CI 0.21–1.73), but not among morning chronotypes (β: 0.04 kg m−2, 95% CI −0.85 to 0.93). No differences by frequency of night shifts and duration of shift work were observed. Conclusion Shift workers did not have an increased risk of cardiometabolic risk factors compared with day workers, but, in particular shift working evening chronotypes, had an increased risk of overweight. More research is needed to verify our results, and establish whether tailored interventions by chronotype are wanted. Introduction Shift work is nowadays an inherent part of our 24/7 society, with ∼20% of the work force in Europe and the USA working in shifts outside regular daytime working hours.1,2 The chronic disruption of the internal body clock that regulates the daily cycles of biological processes (i.e. circadian rhythms) due to shift work has been associated with negative health outcomes, such as breast cancer and cardiovascular diseases.3–6 Cardiometabolic disorders, such as obesity and type 2 diabetes, develop long before the onset of cardiovascular disease, and may mediate the relation between shift work and cardiovascular diseases.7–9 A recent review showed that shift work was associated with an increased risk of overweight and impaired glucose tolerance.7 A few studies indicated shift work to be associated with other cardiometabolic risk factors, such as triglycerides,9 hypertension,9 uric acid9 and inflammatory markers,8 but according to the review of Proper et al. strong evidence is still lacking.7 This is mainly due to the low number of studies, limited methodological quality of previous studies, including cross-sectional designs.7 It has been suggested that the relation between shift work and cardiometabolic function is influenced by chronotype, i.e. internal circadian rhythm that influences the cycle of sleep and activity.10 Several studies showed that chronotype is associated with cardiometabolic risk factors.11–14 For example, evening chronotypes have been shown to have a higher risk of type 2 diabetes and metabolic syndrome than morning chronotypes.12,14 Moreover, in one study, night shifts increased the risk of type 2 diabetes in morning chronotypes, but decreased the risk in evening chronotypes.10 The reason may be that the biological rhythm is more disturbed by night shifts in morning chronotypes than in evening chronotypes. In contrast, morning shifts may disturb the biological rhythm less strongly in morning chronotypes. In view of the gaps in the scientific literature on shift work, chronotype and cardiometabolic risk factors, we investigated relations between shift work and various cardiometabolic risk factors, and explored these potential relations in different chronotypes. We hypothesize that shift workers, in particular morning chronotypes, have an increased risk of cardiometabolic risk factors compared with day workers. Methods Study population The Doetinchem Cohort Study is an ongoing population-based longitudinal study of men and women who were at the start of the study aged 20–59 years. Participants were measured in 1987–91 (N = 7768), 1993–97 (N = 6117), 1998–2002 (N = 4918), 2003–07 (N = 4520) and 2008–12 (N = 4018). Data collection of the sixth wave is still ongoing, but data of the first three measurement years (2013–15, N = 2041) were available for the present study. Response rates were 75% or higher in waves 2–6.15 We included subjects who participated in Wave 6 since the sixth wave included questions about lifelong shift work history, which enabled us to retrospectively determine shift work status at previous waves. We excluded participants who did not have paid work or had missing data on shift work status, employment status and covariates at any wave. Informed consent was obtained from all individual participants. Shift work Shift work status at each wave was assessed retrospectively in 2013–15 using a questionnaire based on the most important aspects of shift work.16 Participants were asked to indicate whether they ever worked evening shifts (i.e. shifts ending before midnight), night shifts (i.e. shifts that continued or started after midnight), sleep shifts and rotating shifts. All subjects were also asked to the year they started and stopped shift work, and the number of years and months they worked in shifts. Using this shift work history data, we classified participants as shift workers 1 year prior to each wave if they worked in evening, night or rotating shifts (with or without night shifts); participants who previously worked in irregular shift were considered former shift workers; others were considered day workers. The frequency of night shifts was categorized into no night shifts, 1–4 and ≥5 night shifts/month. Chronotype In 2013–15, chronotype was assessed by a single question: ‘How would you describe yourself?’ Based on six response categories, participants were categorized as ‘morning type’ (‘definitely a morning type’ OR ‘rather more a morning than an evening type’), ‘evening type’ (‘definitely an evening type’ OR ‘rather more an evening than a morning type’) and ‘intermediate type’ (neither). For participants who answered ‘do not know’ (n = 55), chronotype value was set as missing. This self-awareness of chronotype has been shown to be in excellent agreement with the quantitatively assessed chronotype based on sleep times.17 Measurements of cardiometabolic risk factors Weight, height, total and HDL cholesterol, diastolic and systolic blood pressure were objectively measured at all six waves. Waist circumference and random non-fasting glucose were measured at waves 2–6. In combination with self-reported diabetes, a conservative and common cut-off point of random glucose (i.e. 11.1 mmol l−1) was used to determine presence of diabetes (table 1). Details of all measurements have been described previously and in table 1.15 In 2013–14, standardised enzymatic methods were used to retrospectively determine triglycerides, γ-glutamyltransferase (GGT), high sensitivity CRP, uric acid, cystatin C and creatinine of waves 2–5 using blood plasma that had been stored in freezers. Estimated glomerular filtration rate (eGFR) was estimated with the Chronic Disease Epidemiology Collaboration (CKD-EPI) equation.18 Table 1 Description of the measurements and cut-off values of unfavourable levels of the cardiometabolic risk factors Measured waves  Cardiometabolic risk factor  Measurement method  Cut-off value unfavourable level  1–6  Body mass index  Body weight was objectively measured to the nearest 0.1 kg on calibrated scales and body height was measured to the nearest 0.5 cm. Body mass index was calculated as weight divided by height squared (kg m−2)  Overweight: ≥25 kg m−2  2–6  Waist circumference  Twice to the nearest 0.5 cm, at the level midway between the lowest rib and the iliac crest at the end of expiration, with participants in standing position. The mean of two measures was used for analysis  Abdominal obesity: men: ≥102 cm, women: ≥88 cm  1–6  Blood pressure  Diastolic and systolic blood pressure were measured in sitting position with a random zero sphygmomanometer in waves one to three. In waves four to six a Speidel Keller meter.  Hypertension: diastolic blood pressure ≥90 mmHg, and/or systolic blood pressure ≥140 mmHg, and/or taking anti-hypertensive medication  1–6  Total cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypercholesterolaemia: ≥6.0 mmol l−1 and/or taking cholesterol-lowering medication  1–6  HDL cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypo HDL-cholesterol: <0.9 mmol l−1  2–5  Triglycerides  GPO-PAP assay, colorimetric method  Hypertriglyceridaemia; ≥2.3 mmol l−1  2–5  GGT  Colorimetric method  High GGT: >50 U l−1  2–5  High sensitivity CRP  Principle of particle-enhanced immunological agglutination (Tina-quant C-reactive protein)  High CRP: >3.0 mg l−1  2–5  Uric acid  Colorimetric method  Hyperuricaemia: men: >7.0 mg dl−1, women: >6.0 mg dl−1  2–5  Estimated glomerular filtration rate  Cystatin C was measured with the particle enhanced-turbidimetric immunoassay using reagents from Gentian. Creatinine was measured with a Creatinine Plus assay (IDMS traceable)  Low estimated glomerular filtration rate <90 ml min−1/1.73 m2  2–6  Glucose  Plasma glucose levels were determined in random non-fasting venous blood samples using the hexokinase method  Type 2 diabetes: random glucose ≥11.1 mmol l−1 and/or self-reported diabetes  Measured waves  Cardiometabolic risk factor  Measurement method  Cut-off value unfavourable level  1–6  Body mass index  Body weight was objectively measured to the nearest 0.1 kg on calibrated scales and body height was measured to the nearest 0.5 cm. Body mass index was calculated as weight divided by height squared (kg m−2)  Overweight: ≥25 kg m−2  2–6  Waist circumference  Twice to the nearest 0.5 cm, at the level midway between the lowest rib and the iliac crest at the end of expiration, with participants in standing position. The mean of two measures was used for analysis  Abdominal obesity: men: ≥102 cm, women: ≥88 cm  1–6  Blood pressure  Diastolic and systolic blood pressure were measured in sitting position with a random zero sphygmomanometer in waves one to three. In waves four to six a Speidel Keller meter.  Hypertension: diastolic blood pressure ≥90 mmHg, and/or systolic blood pressure ≥140 mmHg, and/or taking anti-hypertensive medication  1–6  Total cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypercholesterolaemia: ≥6.0 mmol l−1 and/or taking cholesterol-lowering medication  1–6  HDL cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypo HDL-cholesterol: <0.9 mmol l−1  2–5  Triglycerides  GPO-PAP assay, colorimetric method  Hypertriglyceridaemia; ≥2.3 mmol l−1  2–5  GGT  Colorimetric method  High GGT: >50 U l−1  2–5  High sensitivity CRP  Principle of particle-enhanced immunological agglutination (Tina-quant C-reactive protein)  High CRP: >3.0 mg l−1  2–5  Uric acid  Colorimetric method  Hyperuricaemia: men: >7.0 mg dl−1, women: >6.0 mg dl−1  2–5  Estimated glomerular filtration rate  Cystatin C was measured with the particle enhanced-turbidimetric immunoassay using reagents from Gentian. Creatinine was measured with a Creatinine Plus assay (IDMS traceable)  Low estimated glomerular filtration rate <90 ml min−1/1.73 m2  2–6  Glucose  Plasma glucose levels were determined in random non-fasting venous blood samples using the hexokinase method  Type 2 diabetes: random glucose ≥11.1 mmol l−1 and/or self-reported diabetes  HDL, high density lipoprotein; GGT, γ-glutamyltransferase; CRP, C-reactive protein. Table 1 Description of the measurements and cut-off values of unfavourable levels of the cardiometabolic risk factors Measured waves  Cardiometabolic risk factor  Measurement method  Cut-off value unfavourable level  1–6  Body mass index  Body weight was objectively measured to the nearest 0.1 kg on calibrated scales and body height was measured to the nearest 0.5 cm. Body mass index was calculated as weight divided by height squared (kg m−2)  Overweight: ≥25 kg m−2  2–6  Waist circumference  Twice to the nearest 0.5 cm, at the level midway between the lowest rib and the iliac crest at the end of expiration, with participants in standing position. The mean of two measures was used for analysis  Abdominal obesity: men: ≥102 cm, women: ≥88 cm  1–6  Blood pressure  Diastolic and systolic blood pressure were measured in sitting position with a random zero sphygmomanometer in waves one to three. In waves four to six a Speidel Keller meter.  Hypertension: diastolic blood pressure ≥90 mmHg, and/or systolic blood pressure ≥140 mmHg, and/or taking anti-hypertensive medication  1–6  Total cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypercholesterolaemia: ≥6.0 mmol l−1 and/or taking cholesterol-lowering medication  1–6  HDL cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypo HDL-cholesterol: <0.9 mmol l−1  2–5  Triglycerides  GPO-PAP assay, colorimetric method  Hypertriglyceridaemia; ≥2.3 mmol l−1  2–5  GGT  Colorimetric method  High GGT: >50 U l−1  2–5  High sensitivity CRP  Principle of particle-enhanced immunological agglutination (Tina-quant C-reactive protein)  High CRP: >3.0 mg l−1  2–5  Uric acid  Colorimetric method  Hyperuricaemia: men: >7.0 mg dl−1, women: >6.0 mg dl−1  2–5  Estimated glomerular filtration rate  Cystatin C was measured with the particle enhanced-turbidimetric immunoassay using reagents from Gentian. Creatinine was measured with a Creatinine Plus assay (IDMS traceable)  Low estimated glomerular filtration rate <90 ml min−1/1.73 m2  2–6  Glucose  Plasma glucose levels were determined in random non-fasting venous blood samples using the hexokinase method  Type 2 diabetes: random glucose ≥11.1 mmol l−1 and/or self-reported diabetes  Measured waves  Cardiometabolic risk factor  Measurement method  Cut-off value unfavourable level  1–6  Body mass index  Body weight was objectively measured to the nearest 0.1 kg on calibrated scales and body height was measured to the nearest 0.5 cm. Body mass index was calculated as weight divided by height squared (kg m−2)  Overweight: ≥25 kg m−2  2–6  Waist circumference  Twice to the nearest 0.5 cm, at the level midway between the lowest rib and the iliac crest at the end of expiration, with participants in standing position. The mean of two measures was used for analysis  Abdominal obesity: men: ≥102 cm, women: ≥88 cm  1–6  Blood pressure  Diastolic and systolic blood pressure were measured in sitting position with a random zero sphygmomanometer in waves one to three. In waves four to six a Speidel Keller meter.  Hypertension: diastolic blood pressure ≥90 mmHg, and/or systolic blood pressure ≥140 mmHg, and/or taking anti-hypertensive medication  1–6  Total cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypercholesterolaemia: ≥6.0 mmol l−1 and/or taking cholesterol-lowering medication  1–6  HDL cholesterol  Until 1998 in non-fasting EDTA-plasma and from 1998 onwards in serum  Hypo HDL-cholesterol: <0.9 mmol l−1  2–5  Triglycerides  GPO-PAP assay, colorimetric method  Hypertriglyceridaemia; ≥2.3 mmol l−1  2–5  GGT  Colorimetric method  High GGT: >50 U l−1  2–5  High sensitivity CRP  Principle of particle-enhanced immunological agglutination (Tina-quant C-reactive protein)  High CRP: >3.0 mg l−1  2–5  Uric acid  Colorimetric method  Hyperuricaemia: men: >7.0 mg dl−1, women: >6.0 mg dl−1  2–5  Estimated glomerular filtration rate  Cystatin C was measured with the particle enhanced-turbidimetric immunoassay using reagents from Gentian. Creatinine was measured with a Creatinine Plus assay (IDMS traceable)  Low estimated glomerular filtration rate <90 ml min−1/1.73 m2  2–6  Glucose  Plasma glucose levels were determined in random non-fasting venous blood samples using the hexokinase method  Type 2 diabetes: random glucose ≥11.1 mmol l−1 and/or self-reported diabetes  HDL, high density lipoprotein; GGT, γ-glutamyltransferase; CRP, C-reactive protein. Covariates Data on educational attainment, lifestyle and use of anti-hypertensive medication and cholesterol-lowering medication were obtained by questionnaire as described in detail elsewhere.15 Educational attainment was categorized as low (intermediate secondary education or less), intermediate (intermediate vocational or higher secondary education), and high (higher vocational education or university). Smoking was categorized as current smoker and non-smoker, and alcohol as low (≤1 glass day−1) and high (>1 glass day−1) alcohol consumption. At baseline, leisure-time physical activity was measured by asking whether adults were at least physically active during 4 h per week. In the second to sixth wave, several items on self-reported time spent in moderate-to-vigorous intensity leisure-time physical activities were combined and the conventional cut-off point at 3.5 h per week was used.19 Occupational physical activity was categorized as low (sedentary occupation), moderate (standing and light manual work), and high (heavy manual work). Statistical analysis For the continuous cardiometabolic risk factors, linear mixed models were used to examine its relations with shift work, while for dichotomous unfavourable levels of cardiometabolic risk factors, logistic generalized estimation equations with exchangeable correlation matrix were used. For both methods, we used longitudinal time lagged models in which shift work status was determined one year prior to the outcomes. A time lag of one year was chosen because shift work has previously been associated with change in BMI and LDL/HDL-ratio over a 1-year period.20 As all outcome measures at the different time-points are analyzed together, the overall results represent the combination of within-subject and between-subject relationships.21 All analyses were adjusted for age, gender, educational level and chronotype. In analyses with diastolic and systolic blood pressure as dependent variable, additional adjustment for anti-hypertensive medication was made; similarly with total cholesterol, HDL cholesterol and triglycerides as dependent variable, adjustment for cholesterol-lowering medication was made. To investigate whether the potential relations of shift work with cardiometabolic risk factors were independent from potential intermediate lifestyle factors, analyses were additionally adjusted for smoking status, alcohol consumption, occupational and leisure-time physical activity. We performed subgroup analyses separately for morning, intermediate and evening chronotypes. In the total population, three groups of shift workers (no night shifts, 1–4 night shifts/month and ≥5 night shifts/month) were compared with day workers to study exposure–response relations regarding frequency of night shifts with cardiometabolic risk factors. We also performed sensitivity analyses in which we stratified the analyses by duration of shift work, i.e. we compared shift workers who worked for <10 and ≥10 years in irregular shifts to day workers.22–24 All analyses were performed using SAS 9.3 software and a two-sided P values of <0.05 was considered statistically significant. Results Out of the 2041 subjects who participated in wave 6, we excluded 353 individuals who were not employed at all waves, 310 with insufficient shift work data and 44 participants with missing data on cardiometabolic risk factors and covariates at all waves. This led to a study population of 808 day workers, 128 former shift workers and 125 shift workers at wave 1. The remainder of the study population (n = 273) was unemployed or employment status was unknown at wave 1, but was employed at any of the other waves. The number of participants in wave 1, 2, 3, 4, 5 and 6 were 1061, 713, 937, 868, 741 and 375, respectively. The number of shift workers in wave 2, 3, 4, 5 and 6 were 70, 104, 84, 81 and 37, respectively. Compared with excluded participants, the included participants were at baseline slightly younger (37.7 vs. 38.3 years), less often men (54 vs. 59%), more often high educated (24 vs. 19%) and had slightly lower levels of cardiometabolic risk factors (e.g. BMI 24.0 kg m−2 vs. 24.5 kg m−2 and systolic blood pressure 120 mmHg vs. 122 mmHg) (data not shown). At Wave 1, the mean age was 37.7 years (SD: 8.9), 54% were men and 24% were highly educated. Among shift workers, 85% worked at least once a month during the night and 49% worked in shifts for 10 years or more (table 2). Table 2 Characteristics of day workers, former shift workers, and shift workers in 1987–89   Never shift workers  Former shift workers  Shift workers  N = 808  N = 128  N = 125  Demographics            Age (years)  39.4 ± 9.0  35.6 ± 9.7  36.2 ± 9.2      Gender (% male)  64 (518)  63 (81)  65 (81)      High educational level (%)  29 (236)  23 (29)  14 (18)  Chronotype                Morning type (%)  36 (289)  32 (41)  28 (35)          Intermediate type (%)  32 (256)  28 (36)  32 (40)          Evening type (%)  28 (226)  36 (46)  34 (42)          Unknown type (%)  5 (37)  4 (5)  6 (8)  Occupational            Number night shifts per month  ·  ·  4.9 ± 4.1          0 night shifts/month (%)  ·  ·  15 (17)          1–4 night shifts/month (%)  ·  ·  35 (41)          ≥5 night shifts/month (%)  ·  ·  50 (58)  Years of shift work      10.9 (8.5)          <10 years (%)  ·  ·  50 (63)          10–19 years (%)  ·  ·  35 (44)          ≥20 years (%)  ·  ·  14 (18)  Lifestyle            Current smoker (%)  26 (208)  30 (39)  29 (36)      High leisure-time physical activity (%)  70 (573)  73 (93)  72 (90)      High occupational physical activity (%)  4 (36)  2 (3)  2 (3)      High alcohol intake (%)  43 (353)  34 (44)  45 (56)  Cardiometabolic risk factors            Body mass index (kg m−2)  24.2 ± 2.8  23.9 ± 3.0  24.4 ± 3.0      Waist circumference (cm)a  90 ± 10  89 ± 11  90 ± 11      Systolic blood pressure (mmHg)  121 ± 13  121 ± 13  121 ± 14      Diastolic blood pressure (mmHg)  76 ± 9  76 ± 10  74 ± 10      Total cholesterol (mmol l−1)  5.5 ± 1.0  5.1 ± 0.9  5.4 ± 1.0      HDL cholesterol (mmol l−1)  1.26 ± 0.31  1.23 ± 0.32  1.3 ± 0.3      Triglycerides (mmol l−1)a  1.25 (0.90–1.75)  1.05 (0.88–1.47)  1.19 (0.87–1.58)      γ-Glutamyltransferase (U l−1)a  21 (13–32)  18 (13–27)  21 (14–31)      C-reactive protein (mg l−1)a  0.90 (0.44–1.78)  0.72 (0.38–1.41)  0.78 (0.40–1.65)      Uric Acid (mg dl−1)a  5.0 ± 1.2  4.8 ± 1.2  5.2 ± 1.1      eGFR (ml min−1/1.73 m2)a  106 ± 13  110 ± 13  107 ± 14  Medication            Anti-hypertensive medication (%)  2.3 (19)  3.2 (4)  2.1 (3)      Cholesterol-lowering medication (%)  0 (0)  0 (0)  0.0 (0)    Never shift workers  Former shift workers  Shift workers  N = 808  N = 128  N = 125  Demographics            Age (years)  39.4 ± 9.0  35.6 ± 9.7  36.2 ± 9.2      Gender (% male)  64 (518)  63 (81)  65 (81)      High educational level (%)  29 (236)  23 (29)  14 (18)  Chronotype                Morning type (%)  36 (289)  32 (41)  28 (35)          Intermediate type (%)  32 (256)  28 (36)  32 (40)          Evening type (%)  28 (226)  36 (46)  34 (42)          Unknown type (%)  5 (37)  4 (5)  6 (8)  Occupational            Number night shifts per month  ·  ·  4.9 ± 4.1          0 night shifts/month (%)  ·  ·  15 (17)          1–4 night shifts/month (%)  ·  ·  35 (41)          ≥5 night shifts/month (%)  ·  ·  50 (58)  Years of shift work      10.9 (8.5)          <10 years (%)  ·  ·  50 (63)          10–19 years (%)  ·  ·  35 (44)          ≥20 years (%)  ·  ·  14 (18)  Lifestyle            Current smoker (%)  26 (208)  30 (39)  29 (36)      High leisure-time physical activity (%)  70 (573)  73 (93)  72 (90)      High occupational physical activity (%)  4 (36)  2 (3)  2 (3)      High alcohol intake (%)  43 (353)  34 (44)  45 (56)  Cardiometabolic risk factors            Body mass index (kg m−2)  24.2 ± 2.8  23.9 ± 3.0  24.4 ± 3.0      Waist circumference (cm)a  90 ± 10  89 ± 11  90 ± 11      Systolic blood pressure (mmHg)  121 ± 13  121 ± 13  121 ± 14      Diastolic blood pressure (mmHg)  76 ± 9  76 ± 10  74 ± 10      Total cholesterol (mmol l−1)  5.5 ± 1.0  5.1 ± 0.9  5.4 ± 1.0      HDL cholesterol (mmol l−1)  1.26 ± 0.31  1.23 ± 0.32  1.3 ± 0.3      Triglycerides (mmol l−1)a  1.25 (0.90–1.75)  1.05 (0.88–1.47)  1.19 (0.87–1.58)      γ-Glutamyltransferase (U l−1)a  21 (13–32)  18 (13–27)  21 (14–31)      C-reactive protein (mg l−1)a  0.90 (0.44–1.78)  0.72 (0.38–1.41)  0.78 (0.40–1.65)      Uric Acid (mg dl−1)a  5.0 ± 1.2  4.8 ± 1.2  5.2 ± 1.1      eGFR (ml min−1/1.73 m2)a  106 ± 13  110 ± 13  107 ± 14  Medication            Anti-hypertensive medication (%)  2.3 (19)  3.2 (4)  2.1 (3)      Cholesterol-lowering medication (%)  0 (0)  0 (0)  0.0 (0)  Note: Values represent means ± standard deviations, percentages and (numbers) or median and (interquartile ranges). a Not assessed at baseline. Therefore, values of wave two (1993–97) are shown. eGFR, estimated glomerular filtration rate; HDL, high density lipoprotein. Table 2 Characteristics of day workers, former shift workers, and shift workers in 1987–89   Never shift workers  Former shift workers  Shift workers  N = 808  N = 128  N = 125  Demographics            Age (years)  39.4 ± 9.0  35.6 ± 9.7  36.2 ± 9.2      Gender (% male)  64 (518)  63 (81)  65 (81)      High educational level (%)  29 (236)  23 (29)  14 (18)  Chronotype                Morning type (%)  36 (289)  32 (41)  28 (35)          Intermediate type (%)  32 (256)  28 (36)  32 (40)          Evening type (%)  28 (226)  36 (46)  34 (42)          Unknown type (%)  5 (37)  4 (5)  6 (8)  Occupational            Number night shifts per month  ·  ·  4.9 ± 4.1          0 night shifts/month (%)  ·  ·  15 (17)          1–4 night shifts/month (%)  ·  ·  35 (41)          ≥5 night shifts/month (%)  ·  ·  50 (58)  Years of shift work      10.9 (8.5)          <10 years (%)  ·  ·  50 (63)          10–19 years (%)  ·  ·  35 (44)          ≥20 years (%)  ·  ·  14 (18)  Lifestyle            Current smoker (%)  26 (208)  30 (39)  29 (36)      High leisure-time physical activity (%)  70 (573)  73 (93)  72 (90)      High occupational physical activity (%)  4 (36)  2 (3)  2 (3)      High alcohol intake (%)  43 (353)  34 (44)  45 (56)  Cardiometabolic risk factors            Body mass index (kg m−2)  24.2 ± 2.8  23.9 ± 3.0  24.4 ± 3.0      Waist circumference (cm)a  90 ± 10  89 ± 11  90 ± 11      Systolic blood pressure (mmHg)  121 ± 13  121 ± 13  121 ± 14      Diastolic blood pressure (mmHg)  76 ± 9  76 ± 10  74 ± 10      Total cholesterol (mmol l−1)  5.5 ± 1.0  5.1 ± 0.9  5.4 ± 1.0      HDL cholesterol (mmol l−1)  1.26 ± 0.31  1.23 ± 0.32  1.3 ± 0.3      Triglycerides (mmol l−1)a  1.25 (0.90–1.75)  1.05 (0.88–1.47)  1.19 (0.87–1.58)      γ-Glutamyltransferase (U l−1)a  21 (13–32)  18 (13–27)  21 (14–31)      C-reactive protein (mg l−1)a  0.90 (0.44–1.78)  0.72 (0.38–1.41)  0.78 (0.40–1.65)      Uric Acid (mg dl−1)a  5.0 ± 1.2  4.8 ± 1.2  5.2 ± 1.1      eGFR (ml min−1/1.73 m2)a  106 ± 13  110 ± 13  107 ± 14  Medication            Anti-hypertensive medication (%)  2.3 (19)  3.2 (4)  2.1 (3)      Cholesterol-lowering medication (%)  0 (0)  0 (0)  0.0 (0)    Never shift workers  Former shift workers  Shift workers  N = 808  N = 128  N = 125  Demographics            Age (years)  39.4 ± 9.0  35.6 ± 9.7  36.2 ± 9.2      Gender (% male)  64 (518)  63 (81)  65 (81)      High educational level (%)  29 (236)  23 (29)  14 (18)  Chronotype                Morning type (%)  36 (289)  32 (41)  28 (35)          Intermediate type (%)  32 (256)  28 (36)  32 (40)          Evening type (%)  28 (226)  36 (46)  34 (42)          Unknown type (%)  5 (37)  4 (5)  6 (8)  Occupational            Number night shifts per month  ·  ·  4.9 ± 4.1          0 night shifts/month (%)  ·  ·  15 (17)          1–4 night shifts/month (%)  ·  ·  35 (41)          ≥5 night shifts/month (%)  ·  ·  50 (58)  Years of shift work      10.9 (8.5)          <10 years (%)  ·  ·  50 (63)          10–19 years (%)  ·  ·  35 (44)          ≥20 years (%)  ·  ·  14 (18)  Lifestyle            Current smoker (%)  26 (208)  30 (39)  29 (36)      High leisure-time physical activity (%)  70 (573)  73 (93)  72 (90)      High occupational physical activity (%)  4 (36)  2 (3)  2 (3)      High alcohol intake (%)  43 (353)  34 (44)  45 (56)  Cardiometabolic risk factors            Body mass index (kg m−2)  24.2 ± 2.8  23.9 ± 3.0  24.4 ± 3.0      Waist circumference (cm)a  90 ± 10  89 ± 11  90 ± 11      Systolic blood pressure (mmHg)  121 ± 13  121 ± 13  121 ± 14      Diastolic blood pressure (mmHg)  76 ± 9  76 ± 10  74 ± 10      Total cholesterol (mmol l−1)  5.5 ± 1.0  5.1 ± 0.9  5.4 ± 1.0      HDL cholesterol (mmol l−1)  1.26 ± 0.31  1.23 ± 0.32  1.3 ± 0.3      Triglycerides (mmol l−1)a  1.25 (0.90–1.75)  1.05 (0.88–1.47)  1.19 (0.87–1.58)      γ-Glutamyltransferase (U l−1)a  21 (13–32)  18 (13–27)  21 (14–31)      C-reactive protein (mg l−1)a  0.90 (0.44–1.78)  0.72 (0.38–1.41)  0.78 (0.40–1.65)      Uric Acid (mg dl−1)a  5.0 ± 1.2  4.8 ± 1.2  5.2 ± 1.1      eGFR (ml min−1/1.73 m2)a  106 ± 13  110 ± 13  107 ± 14  Medication            Anti-hypertensive medication (%)  2.3 (19)  3.2 (4)  2.1 (3)      Cholesterol-lowering medication (%)  0 (0)  0 (0)  0.0 (0)  Note: Values represent means ± standard deviations, percentages and (numbers) or median and (interquartile ranges). a Not assessed at baseline. Therefore, values of wave two (1993–97) are shown. eGFR, estimated glomerular filtration rate; HDL, high density lipoprotein. Shift work and cardiometabolic risk factors After adjustment for age, gender, education and chronotype, shift workers had more often overweight (OR: 1.44, 95% CI 1.06–1.95) than day workers (table 3), as well as a significantly higher BMI (i.e. 0.56 kg m−2, 95% CI 0.10–1.03). Shift work was not significantly related with other cardiometabolic risk factors, and former shift workers did not have an increased risk for any of the cardiometabolic risk factors compared with day workers. Results did not change after additional adjustment for lifestyle factors (data not shown). Table 3 Multivariable-adjusted differences in cardiometabolic risk factors between day workers, former shift workers, and shift workers   Former shift workers versus day workers   Shift workers versus day workers   OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.00 (0.76 to 1.32)  0.99  1.44 (1.06 to 1.95)  0.02  Abdominal obesity (waist circumference ≥88/102)  1.24 (0.91 to 1.68)  0.17  1.26 (0.89 to 1.79)  0.18  Hypertension (≥90/140 mmHg)  1.05 (0.78 to 1.40)  0.77  0.99 (0.68 to 1.43)  0.84  Hypercholesterolaemia (≥6.0 mmol l−1)  0.84 (0.64 to 1.11)  0.84  1.00 (0.73 to 1.38)  0.99  Low HDL cholesterol (<0.9 mmol l−1)  1.07 (0.68 to 1.69)  0.76  0.86 (0.52 to 1.45)  0.57  Hypertriglyceridaemia (2.3 mmol l−1)  1.09 (0.72 to 1.64)  0.69  0.96 (0.57 to 1.61)  0.92  Type 2 diabetes (≥11.1 mmol l−1 or self-reported)  1.72 (0.71 to 4.17)  0.23  1.58 (0.58 to 4.30)  0.38  High γ-glutamyltransferase (≥50 U l−1)  1.31 (0.81 to 2.11)  0.27  0.82 (0.45 to 1.45)  0.41  High C-reactive protein (≥3.0 mg l−1)  0.82 (0.57 to 1.19)  0.29  1.34 (0.87 to 2.06)  0.24  Hyperuricaemia (≥2.3 mmol l−1)  0.78 (0.44 to 1.41)  0.41  0.73 (0.39 to 1.38)  0.32  Low eGFR (<90 ml min−1/1.73 m2)  1.00 (0.66 to 1.52)  1.00  1.16 (0.74 to 1.82)  0.57    β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.09 (−0.34 to 0.52)  0.69  0.56 (0.10 to 1.03)  0.02  Waist circumference (cm)  0.04 (−1.4 to 1.5)  0.96  1.1 (−0.5 to 2.7)  0.18  Systolic blood pressure (mmHg)b  −0.7 (−2.5 to 1.1)  0.43  −0.6 (−2.6 to 1.4)  0.53  Diastolic blood pressure (mmHg)b  −0.6 (−1.7 to 0.6)  0.33  −0.7 (−2.0 to 0.6)  0.28  Total cholesterol (mmol l−1)b  −0.14 (−0.27 to −0.01)  0.04  −0.03 (−0.18 to 0.12)  0.69  HDL cholesterol (mmol l−1)b  −0.02 (−0.06 to 0.03)  0.52  −0.01 (−0.06 to 0.04)  0.78  Triglycerides (log mmol l−1)b  −0.06 (−0.12 to 0.01)  0.11  0.01 (−0.07 to 0.09)  0.84  γ-Glutamyltransferase (log U l−1)  0.001 (−0.08 to 0.08)  0.99  −0.02 (−0.11 to 0.08)  0.71  C-reactive protein (log mg l−1)  −0.09 (−0.23 to 0.04)  0.19  0.01 (−0.14 to 0.17)  0.87  Uric acid (mg dl−1)  0.01 (−0.14 to 0.16)  0.89  0.11 (−0.06 to 0.28)  0.21  eGFR (ml min−1/1.73 m2)  −0.5 (−2.2 to 1.2)  0.57  −1.3 (−3.2 to 0.6)  0.18    Former shift workers versus day workers   Shift workers versus day workers   OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.00 (0.76 to 1.32)  0.99  1.44 (1.06 to 1.95)  0.02  Abdominal obesity (waist circumference ≥88/102)  1.24 (0.91 to 1.68)  0.17  1.26 (0.89 to 1.79)  0.18  Hypertension (≥90/140 mmHg)  1.05 (0.78 to 1.40)  0.77  0.99 (0.68 to 1.43)  0.84  Hypercholesterolaemia (≥6.0 mmol l−1)  0.84 (0.64 to 1.11)  0.84  1.00 (0.73 to 1.38)  0.99  Low HDL cholesterol (<0.9 mmol l−1)  1.07 (0.68 to 1.69)  0.76  0.86 (0.52 to 1.45)  0.57  Hypertriglyceridaemia (2.3 mmol l−1)  1.09 (0.72 to 1.64)  0.69  0.96 (0.57 to 1.61)  0.92  Type 2 diabetes (≥11.1 mmol l−1 or self-reported)  1.72 (0.71 to 4.17)  0.23  1.58 (0.58 to 4.30)  0.38  High γ-glutamyltransferase (≥50 U l−1)  1.31 (0.81 to 2.11)  0.27  0.82 (0.45 to 1.45)  0.41  High C-reactive protein (≥3.0 mg l−1)  0.82 (0.57 to 1.19)  0.29  1.34 (0.87 to 2.06)  0.24  Hyperuricaemia (≥2.3 mmol l−1)  0.78 (0.44 to 1.41)  0.41  0.73 (0.39 to 1.38)  0.32  Low eGFR (<90 ml min−1/1.73 m2)  1.00 (0.66 to 1.52)  1.00  1.16 (0.74 to 1.82)  0.57    β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.09 (−0.34 to 0.52)  0.69  0.56 (0.10 to 1.03)  0.02  Waist circumference (cm)  0.04 (−1.4 to 1.5)  0.96  1.1 (−0.5 to 2.7)  0.18  Systolic blood pressure (mmHg)b  −0.7 (−2.5 to 1.1)  0.43  −0.6 (−2.6 to 1.4)  0.53  Diastolic blood pressure (mmHg)b  −0.6 (−1.7 to 0.6)  0.33  −0.7 (−2.0 to 0.6)  0.28  Total cholesterol (mmol l−1)b  −0.14 (−0.27 to −0.01)  0.04  −0.03 (−0.18 to 0.12)  0.69  HDL cholesterol (mmol l−1)b  −0.02 (−0.06 to 0.03)  0.52  −0.01 (−0.06 to 0.04)  0.78  Triglycerides (log mmol l−1)b  −0.06 (−0.12 to 0.01)  0.11  0.01 (−0.07 to 0.09)  0.84  γ-Glutamyltransferase (log U l−1)  0.001 (−0.08 to 0.08)  0.99  −0.02 (−0.11 to 0.08)  0.71  C-reactive protein (log mg l−1)  −0.09 (−0.23 to 0.04)  0.19  0.01 (−0.14 to 0.17)  0.87  Uric acid (mg dl−1)  0.01 (−0.14 to 0.16)  0.89  0.11 (−0.06 to 0.28)  0.21  eGFR (ml min−1/1.73 m2)  −0.5 (−2.2 to 1.2)  0.57  −1.3 (−3.2 to 0.6)  0.18  a Analyses were adjusted for age, gender, educational level, and chronotype. b Systolic and diastolic blood pressure were additionally adjusted for anti-hypertensive medication, total cholesterol, HDL cholesterol and triglycerides for cholesterol-lowering medication. eGFR, estimated glomerular filtration rate, OR, odds ratio; CI, confidence interval; β, regression coefficient. Boldface indicates statistical significance (P < 0.05). Table 3 Multivariable-adjusted differences in cardiometabolic risk factors between day workers, former shift workers, and shift workers   Former shift workers versus day workers   Shift workers versus day workers   OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.00 (0.76 to 1.32)  0.99  1.44 (1.06 to 1.95)  0.02  Abdominal obesity (waist circumference ≥88/102)  1.24 (0.91 to 1.68)  0.17  1.26 (0.89 to 1.79)  0.18  Hypertension (≥90/140 mmHg)  1.05 (0.78 to 1.40)  0.77  0.99 (0.68 to 1.43)  0.84  Hypercholesterolaemia (≥6.0 mmol l−1)  0.84 (0.64 to 1.11)  0.84  1.00 (0.73 to 1.38)  0.99  Low HDL cholesterol (<0.9 mmol l−1)  1.07 (0.68 to 1.69)  0.76  0.86 (0.52 to 1.45)  0.57  Hypertriglyceridaemia (2.3 mmol l−1)  1.09 (0.72 to 1.64)  0.69  0.96 (0.57 to 1.61)  0.92  Type 2 diabetes (≥11.1 mmol l−1 or self-reported)  1.72 (0.71 to 4.17)  0.23  1.58 (0.58 to 4.30)  0.38  High γ-glutamyltransferase (≥50 U l−1)  1.31 (0.81 to 2.11)  0.27  0.82 (0.45 to 1.45)  0.41  High C-reactive protein (≥3.0 mg l−1)  0.82 (0.57 to 1.19)  0.29  1.34 (0.87 to 2.06)  0.24  Hyperuricaemia (≥2.3 mmol l−1)  0.78 (0.44 to 1.41)  0.41  0.73 (0.39 to 1.38)  0.32  Low eGFR (<90 ml min−1/1.73 m2)  1.00 (0.66 to 1.52)  1.00  1.16 (0.74 to 1.82)  0.57    β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.09 (−0.34 to 0.52)  0.69  0.56 (0.10 to 1.03)  0.02  Waist circumference (cm)  0.04 (−1.4 to 1.5)  0.96  1.1 (−0.5 to 2.7)  0.18  Systolic blood pressure (mmHg)b  −0.7 (−2.5 to 1.1)  0.43  −0.6 (−2.6 to 1.4)  0.53  Diastolic blood pressure (mmHg)b  −0.6 (−1.7 to 0.6)  0.33  −0.7 (−2.0 to 0.6)  0.28  Total cholesterol (mmol l−1)b  −0.14 (−0.27 to −0.01)  0.04  −0.03 (−0.18 to 0.12)  0.69  HDL cholesterol (mmol l−1)b  −0.02 (−0.06 to 0.03)  0.52  −0.01 (−0.06 to 0.04)  0.78  Triglycerides (log mmol l−1)b  −0.06 (−0.12 to 0.01)  0.11  0.01 (−0.07 to 0.09)  0.84  γ-Glutamyltransferase (log U l−1)  0.001 (−0.08 to 0.08)  0.99  −0.02 (−0.11 to 0.08)  0.71  C-reactive protein (log mg l−1)  −0.09 (−0.23 to 0.04)  0.19  0.01 (−0.14 to 0.17)  0.87  Uric acid (mg dl−1)  0.01 (−0.14 to 0.16)  0.89  0.11 (−0.06 to 0.28)  0.21  eGFR (ml min−1/1.73 m2)  −0.5 (−2.2 to 1.2)  0.57  −1.3 (−3.2 to 0.6)  0.18    Former shift workers versus day workers   Shift workers versus day workers   OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.00 (0.76 to 1.32)  0.99  1.44 (1.06 to 1.95)  0.02  Abdominal obesity (waist circumference ≥88/102)  1.24 (0.91 to 1.68)  0.17  1.26 (0.89 to 1.79)  0.18  Hypertension (≥90/140 mmHg)  1.05 (0.78 to 1.40)  0.77  0.99 (0.68 to 1.43)  0.84  Hypercholesterolaemia (≥6.0 mmol l−1)  0.84 (0.64 to 1.11)  0.84  1.00 (0.73 to 1.38)  0.99  Low HDL cholesterol (<0.9 mmol l−1)  1.07 (0.68 to 1.69)  0.76  0.86 (0.52 to 1.45)  0.57  Hypertriglyceridaemia (2.3 mmol l−1)  1.09 (0.72 to 1.64)  0.69  0.96 (0.57 to 1.61)  0.92  Type 2 diabetes (≥11.1 mmol l−1 or self-reported)  1.72 (0.71 to 4.17)  0.23  1.58 (0.58 to 4.30)  0.38  High γ-glutamyltransferase (≥50 U l−1)  1.31 (0.81 to 2.11)  0.27  0.82 (0.45 to 1.45)  0.41  High C-reactive protein (≥3.0 mg l−1)  0.82 (0.57 to 1.19)  0.29  1.34 (0.87 to 2.06)  0.24  Hyperuricaemia (≥2.3 mmol l−1)  0.78 (0.44 to 1.41)  0.41  0.73 (0.39 to 1.38)  0.32  Low eGFR (<90 ml min−1/1.73 m2)  1.00 (0.66 to 1.52)  1.00  1.16 (0.74 to 1.82)  0.57    β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.09 (−0.34 to 0.52)  0.69  0.56 (0.10 to 1.03)  0.02  Waist circumference (cm)  0.04 (−1.4 to 1.5)  0.96  1.1 (−0.5 to 2.7)  0.18  Systolic blood pressure (mmHg)b  −0.7 (−2.5 to 1.1)  0.43  −0.6 (−2.6 to 1.4)  0.53  Diastolic blood pressure (mmHg)b  −0.6 (−1.7 to 0.6)  0.33  −0.7 (−2.0 to 0.6)  0.28  Total cholesterol (mmol l−1)b  −0.14 (−0.27 to −0.01)  0.04  −0.03 (−0.18 to 0.12)  0.69  HDL cholesterol (mmol l−1)b  −0.02 (−0.06 to 0.03)  0.52  −0.01 (−0.06 to 0.04)  0.78  Triglycerides (log mmol l−1)b  −0.06 (−0.12 to 0.01)  0.11  0.01 (−0.07 to 0.09)  0.84  γ-Glutamyltransferase (log U l−1)  0.001 (−0.08 to 0.08)  0.99  −0.02 (−0.11 to 0.08)  0.71  C-reactive protein (log mg l−1)  −0.09 (−0.23 to 0.04)  0.19  0.01 (−0.14 to 0.17)  0.87  Uric acid (mg dl−1)  0.01 (−0.14 to 0.16)  0.89  0.11 (−0.06 to 0.28)  0.21  eGFR (ml min−1/1.73 m2)  −0.5 (−2.2 to 1.2)  0.57  −1.3 (−3.2 to 0.6)  0.18  a Analyses were adjusted for age, gender, educational level, and chronotype. b Systolic and diastolic blood pressure were additionally adjusted for anti-hypertensive medication, total cholesterol, HDL cholesterol and triglycerides for cholesterol-lowering medication. eGFR, estimated glomerular filtration rate, OR, odds ratio; CI, confidence interval; β, regression coefficient. Boldface indicates statistical significance (P < 0.05). Subgroup analysis by chronotype Among evening chronotypes, shift work was related with higher BMI (β: 0.97 kg m−2, 95% CI 0.21–1.73), which was not the case among morning or intermediate chronotypes (β: 0.04 kg m−2, 95% CI −0.85 to 0.93; respectively, 0.72, 95% CI −0.03 to 1.48). Shift work was not significantly related with other cardiometabolic risk factors in any of the chronotype strata (table 4). Though not statistically significant, in morning chronotypes, shift workers had an odds ratio for type 2 diabetes of 5.50 (95% CI 0.86–35.11). Table 4 Differences in cardiometabolic risk factors between day workers and shift workers, stratified by chronotype   Morning chronotype   Intermediate chronotype   Evening chronotype   OR 95% CI a  P-value  OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.27 (0.72 to 2.22)  0.41  1.65 (0.96 to 2.82)  0.07  1.46 (0.88 to 2.43)  0.15  Abdominal obesity (waist circumference ≥88/102)  0.99 (0.53 to 1.87)  0.98  1.30 (0.71 to 2.37)  0.39  1.51 (0.88 to 2.61)  0.13  Hypertension (≥90/140 mmHg)  0.94 (0.47 to 1.86)  0.86  0.91 (0.47 to 1.79)  0.79  1.01 (0.57 to 1.77)  0.86  Hypercholesterolaemia (≥6.0 mmol l−1)  1.12 (0.64 to 1.97)  0.69  0.78 (0.46 to 1.32)  0.35  1.15 (0.68 to 1.96)  0.56  Low HDL cholesterol (<0.9 mmol l−1)  1.43 (0.61 to 3.36)  0.41  0.51 (0.17 to 1.51)  0.22  0.84 (0.40 to 1.79)  0.65  Hypertriglyceridaemia (2.3 mmol l−1)  1.35 (0.59 to 3.10)  0.48  0.79 (0.29 to 2.15)  0.65  0.90 (0.39 to 2.06)  0.80  Type 2 diabetes (≥11.1 mmol l−1 or self to reported)  5.50 (0.86 to 35.11)  0.07  0.97 (0.11 to 8.18)  0.98  1.20 (0.29 to 5.01)  0.65  High γ-glutamyltransferase (≥50 U l−1)  1.47 (0.55 to 3.90)  0.44  0.71 (0.23 to 2.16)  0.55  0.44 (0.16 to 1.23)  0.12  High C-reactive protein (≥3.0 mg l−1)  1.30 (0.59 to 2.83)  0.52  1.03 (0.43 to 2.42)  0.96  1.47 (0.79 to 2.73)  0.23  Hyperuricaemia (≥2.3 mmol l−1)  0.15 (0.02 to 1.21)  0.08  0.83 (0.29 to 2.42)  0.73  0.94 (0.38 to 2.33)  0.97  Low eGFR (<90 ml min−1/1.73 m2)  0.92 (0.40 to 2.11)  0.84  0.74 (0.30 to 1.80)  0.50  1.93 (0.92 to 4.03)  0.09    β 95% CIa  β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.04 (−0.85 to 0.93)  0.93  0.72 (−0.03 to 1.48)  0.06  0.97 (0.21 to 1.73)  0.03  Waist circumference (cm)  −0.5 (−3.5 to 2.5)  0.75  1.2 (−1.4 to 3.9)  0.37  2.4 (−0.3 to 5.1)  0.10  Systolic blood pressure (mmHg)b  −0.03 (−3.6 to 3.5)  0.99  −1.9 (−5.3 to 1.6)  0.28  −0.2 (−3.7 to 3.2)  0.87  Diastolic blood pressure (mmHg)b  0.02 (−2.3 to 2.4)  0.98  −1.2 (−3.4 to 1.0)  0.29  −0.8 (−3.1 to 1.4)  0.42  Total cholesterol (mmol l−1)b  0.05 (−0.22 to 0.32)  0.71  −0.06 (−0.31 to 0.19)  0.63  −0.06 (−0.31 to 0.20)  0.72  HDL cholesterol (mmol l−1)b  −0.03 (−0.12 to 0.06)  0.46  0.02 (−0.07 to 0.11)  0.69  0.01 (−0.07 to 0.08)  0.88  Triglycerides (log mmol l−1)b  0.10 (−0.03 to 0.24)  0.13  −0.07 (−0.21 to 0.07)  0.30  0.002 (−0.14 to 0.14)  0.94  GGT (log U l−1)  0.10 (−0.07 to 0.26)  0.24  −0.02 (−0.18 to 0.15)  0.83  −0.13 (−0.30 to 0.04)  0.28  CRP (log mg l−1)  0.00 (−0.27 to 0.27)  1.00  −0.06 (−0.33 to 0.22)  0.68  0.10 (−0.18 to 0.37)  0.48  Uric acid (mg dl−1)  0.02 (−0.27 to 0.30)  0.91  0.11 (−0.18 to 0.40)  0.45  0.14 (−0.16 to 0.44)  0.19  eGFR (ml min−1/1.73 m2)  0.8 (−2.5 to 4.1)  0.62  −1.9 (−5.5 to 1.6)  0.28  −2.0 (−5.1 to 1.1)  0.11    Morning chronotype   Intermediate chronotype   Evening chronotype   OR 95% CI a  P-value  OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.27 (0.72 to 2.22)  0.41  1.65 (0.96 to 2.82)  0.07  1.46 (0.88 to 2.43)  0.15  Abdominal obesity (waist circumference ≥88/102)  0.99 (0.53 to 1.87)  0.98  1.30 (0.71 to 2.37)  0.39  1.51 (0.88 to 2.61)  0.13  Hypertension (≥90/140 mmHg)  0.94 (0.47 to 1.86)  0.86  0.91 (0.47 to 1.79)  0.79  1.01 (0.57 to 1.77)  0.86  Hypercholesterolaemia (≥6.0 mmol l−1)  1.12 (0.64 to 1.97)  0.69  0.78 (0.46 to 1.32)  0.35  1.15 (0.68 to 1.96)  0.56  Low HDL cholesterol (<0.9 mmol l−1)  1.43 (0.61 to 3.36)  0.41  0.51 (0.17 to 1.51)  0.22  0.84 (0.40 to 1.79)  0.65  Hypertriglyceridaemia (2.3 mmol l−1)  1.35 (0.59 to 3.10)  0.48  0.79 (0.29 to 2.15)  0.65  0.90 (0.39 to 2.06)  0.80  Type 2 diabetes (≥11.1 mmol l−1 or self to reported)  5.50 (0.86 to 35.11)  0.07  0.97 (0.11 to 8.18)  0.98  1.20 (0.29 to 5.01)  0.65  High γ-glutamyltransferase (≥50 U l−1)  1.47 (0.55 to 3.90)  0.44  0.71 (0.23 to 2.16)  0.55  0.44 (0.16 to 1.23)  0.12  High C-reactive protein (≥3.0 mg l−1)  1.30 (0.59 to 2.83)  0.52  1.03 (0.43 to 2.42)  0.96  1.47 (0.79 to 2.73)  0.23  Hyperuricaemia (≥2.3 mmol l−1)  0.15 (0.02 to 1.21)  0.08  0.83 (0.29 to 2.42)  0.73  0.94 (0.38 to 2.33)  0.97  Low eGFR (<90 ml min−1/1.73 m2)  0.92 (0.40 to 2.11)  0.84  0.74 (0.30 to 1.80)  0.50  1.93 (0.92 to 4.03)  0.09    β 95% CIa  β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.04 (−0.85 to 0.93)  0.93  0.72 (−0.03 to 1.48)  0.06  0.97 (0.21 to 1.73)  0.03  Waist circumference (cm)  −0.5 (−3.5 to 2.5)  0.75  1.2 (−1.4 to 3.9)  0.37  2.4 (−0.3 to 5.1)  0.10  Systolic blood pressure (mmHg)b  −0.03 (−3.6 to 3.5)  0.99  −1.9 (−5.3 to 1.6)  0.28  −0.2 (−3.7 to 3.2)  0.87  Diastolic blood pressure (mmHg)b  0.02 (−2.3 to 2.4)  0.98  −1.2 (−3.4 to 1.0)  0.29  −0.8 (−3.1 to 1.4)  0.42  Total cholesterol (mmol l−1)b  0.05 (−0.22 to 0.32)  0.71  −0.06 (−0.31 to 0.19)  0.63  −0.06 (−0.31 to 0.20)  0.72  HDL cholesterol (mmol l−1)b  −0.03 (−0.12 to 0.06)  0.46  0.02 (−0.07 to 0.11)  0.69  0.01 (−0.07 to 0.08)  0.88  Triglycerides (log mmol l−1)b  0.10 (−0.03 to 0.24)  0.13  −0.07 (−0.21 to 0.07)  0.30  0.002 (−0.14 to 0.14)  0.94  GGT (log U l−1)  0.10 (−0.07 to 0.26)  0.24  −0.02 (−0.18 to 0.15)  0.83  −0.13 (−0.30 to 0.04)  0.28  CRP (log mg l−1)  0.00 (−0.27 to 0.27)  1.00  −0.06 (−0.33 to 0.22)  0.68  0.10 (−0.18 to 0.37)  0.48  Uric acid (mg dl−1)  0.02 (−0.27 to 0.30)  0.91  0.11 (−0.18 to 0.40)  0.45  0.14 (−0.16 to 0.44)  0.19  eGFR (ml min−1/1.73 m2)  0.8 (−2.5 to 4.1)  0.62  −1.9 (−5.5 to 1.6)  0.28  −2.0 (−5.1 to 1.1)  0.11  GGT, γ-glutamyltransferase; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate, OR, odds ratio; CI, confidence interval; β, regression coefficient. Boldface indicates statistical significance (P < 0.05). a Analyses were adjusted for age, gender, and educational level. b Systolic and diastolic blood pressure were additionally adjusted for anti-hypertensive medication, total cholesterol, HDL cholesterol and triglycerides for cholesterol-lowering medication. Table 4 Differences in cardiometabolic risk factors between day workers and shift workers, stratified by chronotype   Morning chronotype   Intermediate chronotype   Evening chronotype   OR 95% CI a  P-value  OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.27 (0.72 to 2.22)  0.41  1.65 (0.96 to 2.82)  0.07  1.46 (0.88 to 2.43)  0.15  Abdominal obesity (waist circumference ≥88/102)  0.99 (0.53 to 1.87)  0.98  1.30 (0.71 to 2.37)  0.39  1.51 (0.88 to 2.61)  0.13  Hypertension (≥90/140 mmHg)  0.94 (0.47 to 1.86)  0.86  0.91 (0.47 to 1.79)  0.79  1.01 (0.57 to 1.77)  0.86  Hypercholesterolaemia (≥6.0 mmol l−1)  1.12 (0.64 to 1.97)  0.69  0.78 (0.46 to 1.32)  0.35  1.15 (0.68 to 1.96)  0.56  Low HDL cholesterol (<0.9 mmol l−1)  1.43 (0.61 to 3.36)  0.41  0.51 (0.17 to 1.51)  0.22  0.84 (0.40 to 1.79)  0.65  Hypertriglyceridaemia (2.3 mmol l−1)  1.35 (0.59 to 3.10)  0.48  0.79 (0.29 to 2.15)  0.65  0.90 (0.39 to 2.06)  0.80  Type 2 diabetes (≥11.1 mmol l−1 or self to reported)  5.50 (0.86 to 35.11)  0.07  0.97 (0.11 to 8.18)  0.98  1.20 (0.29 to 5.01)  0.65  High γ-glutamyltransferase (≥50 U l−1)  1.47 (0.55 to 3.90)  0.44  0.71 (0.23 to 2.16)  0.55  0.44 (0.16 to 1.23)  0.12  High C-reactive protein (≥3.0 mg l−1)  1.30 (0.59 to 2.83)  0.52  1.03 (0.43 to 2.42)  0.96  1.47 (0.79 to 2.73)  0.23  Hyperuricaemia (≥2.3 mmol l−1)  0.15 (0.02 to 1.21)  0.08  0.83 (0.29 to 2.42)  0.73  0.94 (0.38 to 2.33)  0.97  Low eGFR (<90 ml min−1/1.73 m2)  0.92 (0.40 to 2.11)  0.84  0.74 (0.30 to 1.80)  0.50  1.93 (0.92 to 4.03)  0.09    β 95% CIa  β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.04 (−0.85 to 0.93)  0.93  0.72 (−0.03 to 1.48)  0.06  0.97 (0.21 to 1.73)  0.03  Waist circumference (cm)  −0.5 (−3.5 to 2.5)  0.75  1.2 (−1.4 to 3.9)  0.37  2.4 (−0.3 to 5.1)  0.10  Systolic blood pressure (mmHg)b  −0.03 (−3.6 to 3.5)  0.99  −1.9 (−5.3 to 1.6)  0.28  −0.2 (−3.7 to 3.2)  0.87  Diastolic blood pressure (mmHg)b  0.02 (−2.3 to 2.4)  0.98  −1.2 (−3.4 to 1.0)  0.29  −0.8 (−3.1 to 1.4)  0.42  Total cholesterol (mmol l−1)b  0.05 (−0.22 to 0.32)  0.71  −0.06 (−0.31 to 0.19)  0.63  −0.06 (−0.31 to 0.20)  0.72  HDL cholesterol (mmol l−1)b  −0.03 (−0.12 to 0.06)  0.46  0.02 (−0.07 to 0.11)  0.69  0.01 (−0.07 to 0.08)  0.88  Triglycerides (log mmol l−1)b  0.10 (−0.03 to 0.24)  0.13  −0.07 (−0.21 to 0.07)  0.30  0.002 (−0.14 to 0.14)  0.94  GGT (log U l−1)  0.10 (−0.07 to 0.26)  0.24  −0.02 (−0.18 to 0.15)  0.83  −0.13 (−0.30 to 0.04)  0.28  CRP (log mg l−1)  0.00 (−0.27 to 0.27)  1.00  −0.06 (−0.33 to 0.22)  0.68  0.10 (−0.18 to 0.37)  0.48  Uric acid (mg dl−1)  0.02 (−0.27 to 0.30)  0.91  0.11 (−0.18 to 0.40)  0.45  0.14 (−0.16 to 0.44)  0.19  eGFR (ml min−1/1.73 m2)  0.8 (−2.5 to 4.1)  0.62  −1.9 (−5.5 to 1.6)  0.28  −2.0 (−5.1 to 1.1)  0.11    Morning chronotype   Intermediate chronotype   Evening chronotype   OR 95% CI a  P-value  OR 95% CIa  P-value  OR 95% CIa  P-value  Overweight (≥25 kg m−2)  1.27 (0.72 to 2.22)  0.41  1.65 (0.96 to 2.82)  0.07  1.46 (0.88 to 2.43)  0.15  Abdominal obesity (waist circumference ≥88/102)  0.99 (0.53 to 1.87)  0.98  1.30 (0.71 to 2.37)  0.39  1.51 (0.88 to 2.61)  0.13  Hypertension (≥90/140 mmHg)  0.94 (0.47 to 1.86)  0.86  0.91 (0.47 to 1.79)  0.79  1.01 (0.57 to 1.77)  0.86  Hypercholesterolaemia (≥6.0 mmol l−1)  1.12 (0.64 to 1.97)  0.69  0.78 (0.46 to 1.32)  0.35  1.15 (0.68 to 1.96)  0.56  Low HDL cholesterol (<0.9 mmol l−1)  1.43 (0.61 to 3.36)  0.41  0.51 (0.17 to 1.51)  0.22  0.84 (0.40 to 1.79)  0.65  Hypertriglyceridaemia (2.3 mmol l−1)  1.35 (0.59 to 3.10)  0.48  0.79 (0.29 to 2.15)  0.65  0.90 (0.39 to 2.06)  0.80  Type 2 diabetes (≥11.1 mmol l−1 or self to reported)  5.50 (0.86 to 35.11)  0.07  0.97 (0.11 to 8.18)  0.98  1.20 (0.29 to 5.01)  0.65  High γ-glutamyltransferase (≥50 U l−1)  1.47 (0.55 to 3.90)  0.44  0.71 (0.23 to 2.16)  0.55  0.44 (0.16 to 1.23)  0.12  High C-reactive protein (≥3.0 mg l−1)  1.30 (0.59 to 2.83)  0.52  1.03 (0.43 to 2.42)  0.96  1.47 (0.79 to 2.73)  0.23  Hyperuricaemia (≥2.3 mmol l−1)  0.15 (0.02 to 1.21)  0.08  0.83 (0.29 to 2.42)  0.73  0.94 (0.38 to 2.33)  0.97  Low eGFR (<90 ml min−1/1.73 m2)  0.92 (0.40 to 2.11)  0.84  0.74 (0.30 to 1.80)  0.50  1.93 (0.92 to 4.03)  0.09    β 95% CIa  β 95% CIa  β 95% CIa  Body mass index (kg m−2)  0.04 (−0.85 to 0.93)  0.93  0.72 (−0.03 to 1.48)  0.06  0.97 (0.21 to 1.73)  0.03  Waist circumference (cm)  −0.5 (−3.5 to 2.5)  0.75  1.2 (−1.4 to 3.9)  0.37  2.4 (−0.3 to 5.1)  0.10  Systolic blood pressure (mmHg)b  −0.03 (−3.6 to 3.5)  0.99  −1.9 (−5.3 to 1.6)  0.28  −0.2 (−3.7 to 3.2)  0.87  Diastolic blood pressure (mmHg)b  0.02 (−2.3 to 2.4)  0.98  −1.2 (−3.4 to 1.0)  0.29  −0.8 (−3.1 to 1.4)  0.42  Total cholesterol (mmol l−1)b  0.05 (−0.22 to 0.32)  0.71  −0.06 (−0.31 to 0.19)  0.63  −0.06 (−0.31 to 0.20)  0.72  HDL cholesterol (mmol l−1)b  −0.03 (−0.12 to 0.06)  0.46  0.02 (−0.07 to 0.11)  0.69  0.01 (−0.07 to 0.08)  0.88  Triglycerides (log mmol l−1)b  0.10 (−0.03 to 0.24)  0.13  −0.07 (−0.21 to 0.07)  0.30  0.002 (−0.14 to 0.14)  0.94  GGT (log U l−1)  0.10 (−0.07 to 0.26)  0.24  −0.02 (−0.18 to 0.15)  0.83  −0.13 (−0.30 to 0.04)  0.28  CRP (log mg l−1)  0.00 (−0.27 to 0.27)  1.00  −0.06 (−0.33 to 0.22)  0.68  0.10 (−0.18 to 0.37)  0.48  Uric acid (mg dl−1)  0.02 (−0.27 to 0.30)  0.91  0.11 (−0.18 to 0.40)  0.45  0.14 (−0.16 to 0.44)  0.19  eGFR (ml min−1/1.73 m2)  0.8 (−2.5 to 4.1)  0.62  −1.9 (−5.5 to 1.6)  0.28  −2.0 (−5.1 to 1.1)  0.11  GGT, γ-glutamyltransferase; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate, OR, odds ratio; CI, confidence interval; β, regression coefficient. Boldface indicates statistical significance (P < 0.05). a Analyses were adjusted for age, gender, and educational level. b Systolic and diastolic blood pressure were additionally adjusted for anti-hypertensive medication, total cholesterol, HDL cholesterol and triglycerides for cholesterol-lowering medication. Frequency of night shifts and duration of shift work No exposure–response relationships between shift work and cardiometabolic risk factors were observed by frequency of night shifts (Supplementary table S1). Results were also similar for shift workers who worked for <10 years and those who work for ≥10 years in irregular shifts (Supplementary table S2). With the exception of overweight, those who worked <10 years in irregular shifts had no increased risk of being overweight compared with day workers (OR: 1.19, 95% CI 0.80–1.78), while those who worked for ≥10 years in irregular shifts had an OR for being overweight of 1.56 (95% CI 1.08–2.24). Discussion Shift work was not related with an increased risk of cardiometabolic risk factors, except for overweight/BMI. Shift workers had a 44% higher risk of being overweight and had a 0.58 kg m−2 higher BMI than day workers. The increased risk for overweight was particularly present among shift working evening chronotypes. Our research findings do not confirm our hypothesis that shift work is related with cardiometabolic risk factors other than overweight. The few existing previous studies also showed inconsistent findings, with some finding relations between shift work and cardiometabolic risk factors while others could not confirm such findings.7–9,25 The absence of relations between shift work and cardiometabolic risk factors may partly be due to limitation in the design of our study. A large part of the study population dropped out due to the long follow-up period of up to 26 years or were excluded due to missing data. These excluded individuals were lower educated and more often men than the included individuals, but levels of cardiometabolic risk factors at the first examination wave were only slightly more unfavourable. Since shift work is more common in adults of lower socio-economic status and in men, exclusion may have mitigated the relations between shift work and cardiometabolic risk factors. In addition, we lack details about the shifts schedule, such as direction and speed of rotation. It is therefore possible that particular types of shifts are related with metabolic risk factors in our population. However, we did not find dose-response relations between duration of shift work and frequency of night shifts and cardiometabolic risk factors, which are considered to be the most important characteristics of shift work related to adverse health outcomes.16 Thus, it might also be that shift work does not increase the risk of the investigated cardiometabolic risk factors as shown in some other studies.7–9 This means that other determinants mediate the relation between shift work and CVD, such as lifestyle behaviours, psychosocial determinants and other physiological determinants including disturbed autonomic regulation of the cardiovascular system.4,26 In line with other studies, we observed that shift work was related with overweight and BMI.7 We extended those findings and showed that this may be particularly the case for evening chronotypes, and that frequency of night shifts may not influence the risk of overweight. One reason may be differences in average behaviour and timing of food intake and physical activity between day and shift workers. In our study, however, physical activity and dietary patterns did not seem to play a role as adjustment for smoking, alcohol, and physical activity did not attenuate the relations. We did not adjust the primary analyses for diet because this was only measured at Waves 2–4. In sensitivity analyses based on Waves 2–4, analysis was adjusted for energy intake and the Mediterranean diet score. This did not attenuate any of the relations (data not shown). Another potential reason is the disruption of circadian rhythms, which interacts with hormones, such as insulin, leptin, cortisol, growth hormone, and melatonin, and thereby increases the risk of overweight.27–29 Finally, other plausible explanations are more exposure to light at night, less exposure to day light and possibly sleep deprivation related to shift work.27,30,31 Reasons for the particular increased risk among evening chronotypes are unclear. As described before, there is a scarcity on research to the role of chronotype in the shift work related health effects. Of the limited research, contrasting results have been observed with a higher risk of shift work for type 2 diabetes among morning and not among evening chronotypes.10 Future research is needed to understand mechanisms for the increased risk of overweight among shift workers by chronotype. Previous studies showed contrasting findings regarding the relation between shift work and blood glucose, HbA1c and type 2 diabetes, with studies finding a positive association22,32–34 and other studies not.24,35 We did not find a significant relationship between shift work and type 2 diabetes, but the OR pointed, in particular among morning chronotypes (OR: 5.50, 95% CI 0.86–35.11), to an increased risk of type 2 diabetes. The non-significance of this OR is likely due to limited statistical power of the analyses within each chronotype. A recent study observed that among female morning chronotypes, the risk of type 2 diabetes increased with more years exposure to rotating night shifts, while the opposite was true for female evening chronotypes.10 The possible moderation of chronotype may be because the shift schedule, especially night shifts, interferes more strongly with the circadian rhythm among morning chronotypes than evening chronotypes. This may lead to unfavourable biological processes, such as secretion of hormones and sleep problems, which increases the risk of type 2 diabetes.28,29,36,37 However, more research with a larger sample is needed on the moderating effect of chronotype in the relation between shift work and type 2 diabetes before strong conclusions can be drawn. A strength of this study includes the wide variety of objectively measured cardiometabolic risk factors. We were also able to investigate exposure–response relationships with frequency of night shifts and differences across chronotypes. Besides the earlier mentioned limitations, our results are vulnerable for finding false statistically significant relations by chance because we related shift work to 11 cardiometabolic risk factors. We only found shift work to statistically significantly increase the risk of overweight/BMI. Although this could be a spurious finding, it is in line with previous studies and can be explained by several mechanisms, such as an unhealthy lifestyle and hormonal changes.7,27–31 Therefore, this relation is unlikely to be a false positive finding. In addition, it is often argued to use change scores to analyze prospective relations. However, the error of change scores are large,38,39 and may even be larger than the between-subject difference. We therefore used a model in which we compared levels of cardiometabolic risk factors at several time points, and its association with shift work status (within an individual) one year before. Furthermore, shift work status was assessed retrospectively, which makes our results susceptive to information bias. However, we believe the participants were able to reliably recall their shift work exposure as being a shift worker has a significant impact on someone’s life and is relatively easy to remember. Finally, although our single question on chronotype is in excellent agreement with the validated and more quantitatively assessed chronotype of the MCTQ,17 we may have misclassified some participants. If so, this would have probably led to a small underestimation of differences in associations between shift work and cardiometabolic risk factors by chronotype. In conclusion, shift workers had a 44% higher risk of being overweight and a 0.56 kg m−2 higher BMI than day workers, but the risk of other cardiometabolic risk factors did not differ between groups. Shift working evening chronotypes had a particular higher BMI compared with day working evening chronotypes. The increased risk of type 2 diabetes of shift work among morning chronotypes in the study of Vetter et al.10 could not be replicated in our study, although the OR pointed in the same unfavourable direction. Thus, more research is needed on the moderating effects of chronotype to establish whether tailored preventive measures by chronotype are useful for shift workers. Supplementary data Supplementary data are available at EURPUB online. Acknowledgements The authors would like to thank the epidemiologists and data managers at the Dutch National Institute of Public Health and Environment, and the fieldworkers of the Municipal Health Service in Doetinchem for their contribution to the data collection for this study. Funding This research was funded by the ReVanche Program of the EMGO+ Institute for Health and Care Research of the VU University Medical Center, Amsterdam (grant number: 9729). This work was also supported by The Netherlands Organization for Health Research (ZonMW) within the program Electromagnetic Fields and Health Research under grant numbers 85200001 and 85800001. The Doetinchem Cohort Study is funded by the National Institute for Public Health and the Environment, which works under the authority of the Ministry of Health, Welfare and Sport of the Netherlands. Conflicts of interest: None declared. Key points Shift workers have an increased risk of being overweight and higher body mass index compared with day workers. Chronotype is a moderator in the association between shift work and risk of overweight; the association is especially strong in evening chronotypes. In none of the chronotypes is shift work associated with an increased risk of other cardiometabolic risk factors. References 1 U.S. Department of Labor, Bureau of Labor Statistics. Employment and Wages, Annual Averages 2004. Washington, U.S., 2008. Available at: http://www.bls.gov/cew/cewbultn04.htm (14 March 2016, date last accessed). 2 Trost SG, Owen N, Bauman AE, et al.   Correlates of adults' participation in physical activity: review and update. Med Sci Sports Exerc  2002; 34: 1996– 2001. Google Scholar CrossRef Search ADS PubMed  3 Gan Y, Yang C, Tong X, et al.   Shift work and diabetes mellitus: a meta-analysis of observational studies. Occup Environ Med  2015; 72: 72– 8. Google Scholar CrossRef Search ADS PubMed  4 Puttonen S, Harma M, Hublin C. Shift work and cardiovascular disease - pathways from circadian stress to morbidity. Scand J Work Environ Health  2010; 36: 96– 108. Google Scholar CrossRef Search ADS PubMed  5 Vyas MV, Garg AX, Iansavichus AV, et al.   Shift work and vascular events: systematic review and meta-analysis. BMJ  (Clin Res Ed) 2012; 345: e4800. Google Scholar CrossRef Search ADS PubMed  6 Wang F, Yeung KL, Chan WC, et al.   A meta-analysis on dose-response relationship between night shift work and the risk of breast cancer. Ann Oncol  2013; 24: 2724– 32. Google Scholar CrossRef Search ADS PubMed  7 Proper KI, van de Langenberg D, Rodenburg W, et al.   The relationship between shift work and metabolic risk factors: a systematic review of longitudinal studies. Am J Prev Med  2016; 50: e147– 57. Google Scholar CrossRef Search ADS PubMed  8 Puttonen S, Viitasalo K, Harma M. Effect of shiftwork on systemic markers of inflammation. Chronobiol Int  2011; 28: 528– 35. Google Scholar CrossRef Search ADS PubMed  9 Sookoian S, Gemma C, Fernandez Gianotti T, et al.   Effects of rotating shift work on biomarkers of metabolic syndrome and inflammation. J Intern Med  2007; 261: 285– 92. Google Scholar CrossRef Search ADS PubMed  10 Vetter C, Devore EE, Ramin CA, et al.   Mismatch of sleep and work timing and risk of type 2 diabetes. Diabetes Care  2015; 38: 1707– 13. Google Scholar CrossRef Search ADS PubMed  11 Merikanto I, Lahti T, Puolijoki H, et al.   Associations of chronotype and sleep with cardiovascular diseases and type 2 diabetes. Chronobiol Int  2013; 30: 470– 7. Google Scholar CrossRef Search ADS PubMed  12 Osonoi Y, Mita T, Osonoi T, et al.   Morningness-eveningness questionnaire score and metabolic parameters in patients with type 2 diabetes mellitus. Chronobiol Int  2014; 31: 1017– 23. Google Scholar CrossRef Search ADS PubMed  13 Wong PM, Hasler BP, Kamarck TW, et al.   Social Jetlag, chronotype, and cardiometabolic risk. J Clin Endocrinol Metab  2015; 100: 4612– 20. Google Scholar CrossRef Search ADS PubMed  14 Yu JH, Yun CH, Ahn JH, et al.   Evening chronotype is associated with metabolic disorders and body composition in middle-aged adults. J Clin Endocrinol Metab  2015; 100: 1494– 502. Google Scholar CrossRef Search ADS PubMed  15 Verschuren WM, Blokstra A, Picavet HS, Smit HA. Cohort profile: the Doetinchem Cohort Study. Int J Epidemiol  2008; 37: 1236– 41. Google Scholar CrossRef Search ADS PubMed  16 Stevens RG, Hansen J, Costa G, et al.   Considerations of circadian impact for defining ‘shift work’ in cancer studies: iARC Working Group Report. Occup Environ Med  2011; 68: 154– 62. Google Scholar CrossRef Search ADS PubMed  17 Roenneberg T, Wirz-Justice A, Merrow M. Life between clocks: daily temporal patterns of human chronotypes. J Biol Rhythms  2003; 18: 80– 90. Google Scholar CrossRef Search ADS PubMed  18 Inker LA, Schmid CH, Tighiouart H, et al.   Estimating glomerular filtration rate from serum creatinine and cystatin C. New Engl J Med  2012; 367: 20– 9. Google Scholar CrossRef Search ADS PubMed  19 Kemper H, Ooijendijk W, Stiggelbout M. Consensus about the Dutch physical activity guideline. Tijdschr Soc Geneeskd  2000; 78: 180– 3. 20 van Amelsvoort LG, Schouten EG, Kok FJ. Impact of one year of shift work on cardiovascular disease risk factors. J Occup Environ Med  2004; 46: 699– 706. Google Scholar CrossRef Search ADS PubMed  21 Twisk JWR. Applied Longitudinal Data Analysis for Epidemiology. A Practical Guide. Cambridge: Cambridge University Press, 2003. 22 De Bacquer D, Van Risseghem M, Clays E, et al.   Rotating shift work and the metabolic syndrome: a prospective study. Int J Epidemiol  2009; 38: 848– 54. Google Scholar CrossRef Search ADS PubMed  23 Guo Y, Rong Y, Huang X, et al.   Shift work and the relationship with metabolic syndrome in Chinese aged workers. PLoS One  2015; 10: e0120632. Google Scholar CrossRef Search ADS PubMed  24 Pan A, Schernhammer ES, Sun Q, Hu FB. Rotating night shift work and risk of type 2 diabetes: two prospective cohort studies in women. PLoS Med  2011; 8: e1001141. Google Scholar CrossRef Search ADS PubMed  25 Morikawa Y, Nakagawa H, Miura K, et al.   Effect of shift work on body mass index and metabolic parameters. Scand J Work Environ Health  2007; 33: 45– 50. Google Scholar CrossRef Search ADS PubMed  26 Knutsson A, Boggild H. Shiftwork and cardiovascular disease: review of disease mechanisms. Rev Environ Health  2000; 15: 359– 72. Google Scholar CrossRef Search ADS PubMed  27 Antunes LC, Levandovski R, Dantas G, et al.   Obesity and shift work: chronobiological aspects. Nutr Res Rev  2010; 23: 155– 68. Google Scholar CrossRef Search ADS PubMed  28 Bass J, Takahashi JS. Circadian integration of metabolism and energetics. Science  2010; 330: 1349– 54. Google Scholar CrossRef Search ADS PubMed  29 Shi SQ, Ansari TS, McGuinness OP, et al.   Circadian disruption leads to insulin resistance and obesity. Curr Biol  2013; 23: 372– 81. Google Scholar CrossRef Search ADS PubMed  30 McFadden E, Jones ME, Schoemaker MJ, et al.   The relationship between obesity and exposure to light at night: cross-sectional analyses of over 100,000 women in the Breakthrough Generations Study. Am J Epidemiol  2014; 180: 245– 50. Google Scholar CrossRef Search ADS PubMed  31 Patel SR, Hu FB. Short sleep duration and weight gain: a systematic review. Obesity (Silver Spring)  2008; 16: 643– 53. Google Scholar CrossRef Search ADS PubMed  32 Suwazono Y, Uetani M, Oishi M, et al.   Calculation of the benchmark duration of shift work associated with the development of impaired glucose metabolism: a 14-year cohort study on 7104 male workers. Occup Environ Med  2010; 67: 532– 7. Google Scholar CrossRef Search ADS PubMed  33 Li Y, Sato Y, Yamaguchi N. Shift work and the risk of metabolic syndrome: a nested case-control study. Int J Occup Environ Health  2011; 17: 154– 60. Google Scholar CrossRef Search ADS PubMed  34 Oyama I, Kubo T, Fujino Y, et al.   Retrospective cohort study of the risk of impaired glucose tolerance among shift workers. Scand J Work Environ Health  2012; 38: 337– 42. Google Scholar CrossRef Search ADS PubMed  35 Biggi N, Consonni D, Galluzzo V, et al.   Metabolic syndrome in permanent night workers. Chronobiol Int  2008; 25: 443– 54. Google Scholar CrossRef Search ADS PubMed  36 Scheer FA, Hilton MF, Mantzoros CS, Shea SA. Adverse metabolic and cardiovascular consequences of circadian misalignment. Proc Natl Acad Sci USA  2009; 106: 4453– 8. Google Scholar CrossRef Search ADS PubMed  37 Cappuccio FP, D'Elia L, Strazzullo P, Miller MA. Quantity and quality of sleep and incidence of type 2 diabetes: a systematic review and meta-analysis. Diabetes Care  2010; 33: 414– 20. Google Scholar CrossRef Search ADS PubMed  38 Cain KC, Kronmal RA, Kosinski AS. Analysing the relationship between change in a risk factor and risk of disease. Stat Med  1992; 11: 783– 97. Google Scholar CrossRef Search ADS PubMed  39 Twisk J, Proper K. Evaluation of the results of a randomized controlled trial: how to define changes between baseline and follow-up. J Clin Epidemiol  2004; 57: 223– 8. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

The European Journal of Public HealthOxford University Press

Published: May 22, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off