Table 3.
Adjusted ORs of type 2 diabetes and adjusted βs of HbA1c with each additional copy of the MTNR1B G risk allele across categories of current work schedule
| Type 2 diabetes | HbA1c (mmol/mol) | |||||
|---|---|---|---|---|---|---|
| Type 2 diabetes case/control subjects, n/n | OR (95% CI) | Pint | N | β (95% CI) | Pint | |
| Overall (n = 189,488) | 5,042/184,446 | 1.10 (1.05–1.15) | 0.15 | 175,156 | 0.26 (0.23–0.28) | 0.25 |
| Day workers | 4,047/154,792 | 1.09 (1.03–1.14) | 146,993 | 0.25 (0.22–0.28) | ||
| Shift work without nights | 475/14,863 | 1.24 (1.07–1.43) | 14,110 | 0.32 (0.22–0.41) | ||
| Sometimes night shift work | 284/8,434 | 0.99 (0.82–1.20) | 8,005 | 0.36 (0.24–0.48) | ||
| Usual night shift work | 80/2,171 | 0.85 (0.58–1.25) | 2,069 | 0.20 (−0.04 to 0.45) | ||
| Always night shift work | 156/4,186 | 1.28 (0.99–1.65) | 3,979 | 0.19 (0.02–0.37) | ||
Association results are adjusted ORs (95% CI) of type 2 diabetes per each additional copy of the MTNR1B G risk allele or adjusted βs (95% CI) describing differences in HbA1c in mmol/mol per each additional copy of the MTNR1B G risk allele across categories of current work schedule. Association analyses are adjusted for age, sex, BMI, genotyping array, and 10 principal components of ancestry. Pint is log likelihood ratio test comparing models with and without cross-product interaction terms (MTNR1B and current work schedule) including main effect terms in logistic or linear regression models adjusted for the aforementioned covariates.