Table 2.

Association between 1-year mean transportation noise levels, NO₂ and incident diabetes

	Lden road	Lden aircraft	Lden railway	NO2
	RR (95% CI)	RR (95% CI)	RR (95% CI)	RR (95% CI)
Single exposure
Model 1	1.29 (1.00, 1.67)	1.83 (1.03, 3.28)	0.96 (0.71, 1.28)	0.94 (0.68, 1.30)
Model 2	1.21 (0.94, 1.57)	1.80 (0.97, 3.35)	0.90 (0.68, 1.19)	0.88 (0.65, 1.20)
Model 3	1.20 (0.92, 1.56)	1.75 (0.96, 3.19)	0.94 (0.72, 1.22)	0.92 (0.67, 1.26)
Model 4	1.20 (0.93, 1.55)	1.86 (1.01, 3.40)	0.95 (0.73, 1.24)	0.89 (0.64, 1.23)
Model 5	1.17 (0.88, 1.53)	1.92 (1.04, 3.55)	0.94 (0.72, 1.22)	0.87 (0.62, 1.21)
Multi-exposure
Model 1	1.41 (1.07, 1.87)	1.86 (1.00, 3.45)	0.94 (0.71, 1.26)	0.84 (0.59, 1.20)
Model 2	1.38 (1.03, 1.83)	1.82 (0.93, 3.56)	0.90 (0.68, 1.20)	0.81 (0.56, 1.15)
Model 3	1.35 (1.02, 1.78)	1.86 (0.96, 3.59)	0.94 (0.71, 1.24)	0.86 (0.61, 1.22)
Model 4	1.40 (1.05, 1.86)	1.95 (1.01, 3.77)	0.98 (0.74, 1.29)	0.79 (0.55, 1.15)
Model 5	1.31 (0.98, 1.75)	1.96 (1.00, 3.81)	0.94 (0.72, 1.24)	0.86 (0.60, 1.22)
Model 6	1.35 (1.02, 1.78)	1.87 (0.96, 3.62)	0.94 (0.71, 1.24)	0.87 (0.60, 1.22)

All RRs are per IQR of respective noise metric (road: 10 dB, aircraft: 12 dB, rail: 11d B, NO₂: 15μg/m³). Single exposure models included one exposure at a time whereas multi-exposure models considered all exposure metrics at the same time. Random intercepts were applied at the level of the study areas in all models.

Model 1: unadjusted model. Unadjusted model for multi-exposure models include road, aircraft and railway traffic noise, noise truncation indicator and NO₂.

Model 2: Model 1 + age, sex, educational level, neighbourhood socioeconomic index, smoking status and pack-years, consumption of alcohol, fruits and vegetables.

Model 3: Model 2 + physical activity, body mass index and change in body mass index.

Model 4: Model 3 + noise intermittency.

Model 5: Model 3 + traffic noise annoyance.

Model 6: Model 3 + green space.