Table 1.
Main results of systematic reviews focused on the link between shift work and diabetes.
| References, Country | Type-Shift work (n) |
Assessment of
outcomes (n) |
Confounding factors (n) | Main results shift work vs. day work (n) |
|---|---|---|---|---|
| Wang et al. (16), UK | Rotating (4) Unspecified (2) |
Type 2 diabetes (1) Diabetes mellitus (5) Diabetes death certificate (1) FPG or patient under treatment (1) MD (4) |
NA (1) Adjustments (5) |
Rotating: HR = 1.67 (0.57–4.90) (♂ 1); Rotating 2-shift RR = 1.73 (0.85–3.52)/Rotating 3-shift RR = 1.33 (0.74–2.36) (♂ 1) Shift work: Age groups (♂ 1); (30–39 yo) OR = 6.75 (1.31–56.1); (40–49 yo) OR = 1.22 (0.68–2.10); (50–59 yo) OR = 0.93 (0.53–1.55) Shift work: diabetes mortality: per year β × 10−5 = 4.14 (2.46–5.81) (♂ 1) Exposure duration: Rotating night ≥10 years adjusted for age: RR = 1.64 (1.11–2.37); multiple adjustment including BMI: RR = 0.98 (0.66–1.45) (♀ 1) |
| Esquirol et al. (12), France | Permanent Night (3) Rotating (21) Unspecified (2) Evening (1) |
Type 2 Diabetes (1) Diabetes mellitus (23) Glycaemia (11) HbA1c (8) FPG (2) OGTT (2) RPG (1) Self-reported (1) Unknown (2) |
NA(11) Adjustments (13) |
Longitudinal studies: significant effect (7): Shift work: OR = 1.56 (1.18–2.05) (1); OR = 1.35 (1.05–1.75) (1) Shift work duration ♂: age > 50 yo and 19–32 years exposure: 5% increase of risk (1) Shift work: Glycaemia>110 mg/dL = 5.1% vs. 3.8% for DW (1) Cross-Sectional studies: significant effect (4): Night shift: OR = 1.7 (0.8–3.6) (1); 2-shift: OR = 2.5 (1.1–4.3) (1) Longitudinal studies non-significant effect (13) |
| Knutsson and Kempe (13), Sweden | Rotating (5) | Type 2 diabetes (5) Type 1 diabetes (1) OGTT (1) Death certificate (1) Self-reported (1) HbA1c (2) |
NA(1) Adjustments (4) |
Rotating ♂: HR = 1.67 (0.59–4.90) (1); SRR = 1.24 (0.91–1.70)/SRR = 2.29 (0.97–5.40) (1) OR = 1.35 (1.05–1.75) (1) 2-shift ♂: RR = 1.73 (0.85–3.52) (1); 3-shift ♂: RR = 1.33 (0.74–2.36) (1) Rotating duration ♂: 10–19/20–29/≥30 years: SRR = 1.41 (0.18–11.30)/1.92 (0.5– 7.32)/2.85 (1.15–7.08) (1) Rotating duration ♀: 1–2/3–9/10–19/≥ 20 years: HR = 1.03 (0.98–1.08)/1.06 (1.01–1.11)/1.10 (1.02–1.18)/1.24 (1.13–1.37) (1) Rotating ♀: HR per 5 years = 1.05 (1.04–1.06) (1) |
| Gan et al. (18), China Meta-analysis |
Night shift (3) Rotating (3 × 8) (4) Mixed rotating (2 × 8/3 × 8) (2) Unspecified (2) |
Type 2 diabetes (4) Diabetes mellitus (7) Diabetes death certificate (1) Self-reported or medical/register report (10) FPG (2) HbA1c (2) |
NA (1) Adjustments (11) |
Shift work: RR = 1.08 (1.05–1.12), I2 40.9% (corrected for publication bias) (11) Night shift: OR = 1.09 (1.04–1.14), I2 37.6% (3) Rotating shift: OR = 1.42 (1.19–1.69), I2 13.4% (4) Subgroup analyses Shift work: cohort studies OR = 1.12 (1.06–1.19), I2 52.9%/cross-sectional studies OR = 1.06 (1.03–1.09), I2 10.9% Shift work: OR = 1.09 (1.04–1.14) for ♀, I2 54.3%/OR = 1.37 (1.20–1.56) for ♂, I2 0.0% |
| Proper et al. (14), Netherlands | Permanent Night (1) Rotating (8) |
Type 2 diabetes (2) Diabetes mellitus (7) Glycaemia (4) OGTT (1) RPG (1) FPG (1) HbA1c (4) |
NA (2) Adjutments (7) |
Shift work: High FPG: age (25–29 yo) 32.9 vs. 23.1%; NS for other age classes (1) HbA1c OR = 1.35 (1.05–1.75) (1) Shift work: ♂ Positive relation with HbA1c; ♀ NS (1) Shift work: no association with FPG (2) Rotating: OR = 1.56 (1.18–2.05) (1) Rotating 3-shift: OR = 2.62 (2.17–3.17); Rotating 2-shift: OR = 1.78 (1.49–2.14) (1) Rotating 3-shift: RR = 1.33 (0.74–2.36); Rotating 2-shift: RR = 1.73 (0.85–3.52) (1) Shift duration: OR per 10 years = 1.05 (1.01–1.09) (1) |
| Anothaisintawee et al. (17), Thailand Meta-analysis |
Rotating (5) Unspecified (6) |
Type 2 Diabetes (4) Diabetes mellitus (7) FPG (4), HbA1c (3) OGTT (2) Self-reported (4) Medical diagnosis or treatment (6) |
Adjustments (8) Unspecified (2) |
Shift work: RR = 1.40 (1.18–1.66), I2 95% (11); adjusted for BMI, other covariates RR = 1.15 (1.08–1.22) (8) Rotating: RR = 1.60 (1.20–2.14), I2 97.3% (5); adjusted for BMI, other covariates RR = 1.15 (1.06–1.25) (5) |
| Li et al. (20), China Meta-analysis |
Permanent night (4) Evening (4) Rotating (5) Unspecified (3) |
Type 2 diabetes (2) Diabetes mellitus (10) Diabetes death certificate Self-reported or medical/register report HbA1c |
Adjustment (12) | Shift work: RR = 1.12 (1.07–1.17), I2 38.9% (corrected for publication bias) (12) Night/Evening shift: RR = 1.19 (1.09–1.30), I2 52.2% (4) Rotating shift RR = 1.11 (1.06–1.16), I2 48.2% (5) Dose-response analyses: Shift work: RR = 1.07 (1.04–1.09), I2 0% per 5-year exposure (2, ♀) Subgroups analyses: Shift work: RR = 1.21 (1.08–1.17) for ♀, I2 46.9%/RR = 1.28 (1.16–1.42) for ♂ (1) Shift work: RR = 1.13 (1.07–1.19) for follow-up ≥10 years, I2 49.5%/RR = 1.17 (1.10–1.24) for follow-up < 10 years (1) |
| Rosa et al. (15), Italy | Night shift (1) | Type 2 diabetes Register (1) |
MD | Night shift: HR = 1.58 (1.25–1.99) (1) |
| Gao et al. (19), China Meta-analysis |
Night shift (10) Rotating (4) Evening (2) Unspecified (5) |
Type 2 Diabetes (21) FPG HbA1c OGTT Random plasma glucose |
NA (4) Adjustments (17) |
Shift work: RR = 1.10 (1.05–1.14), I2 37.2% (21) Night shift: RR = 1.15 (1.08–1.24), I2 60.7% (10) Rotating: RR = 1.08 (1.04–1.12), I2 0% (4) Dose-response analyses (3, ♀): RR = 1.05 (1.03–1.07) per 5-year exposure of shift work; RR = 1.17 (1.11–1.24) for 15 years of shift work |
(n), number of studies concerned; SW, shift work; FPG, Fasting plasma glucose; OGTT, Oral Glucose Tolerance Test; RPG, Random-plasma glucose; HbA1c, glycated hemoglobin MD, Missing Data; NA, non Adjusted.