Traffic-Related Air Pollution and Dementia Incidence in Northern Sweden: A Longitudinal Study

http://dx.doi.org/10.1289/ehp.1408322



Table 2. Hazard ratios (HRs) and 95% confidence intervals (CIs) for dementia in association with baseline annual NOx concentration from Cox proportional hazards models.
NOxa (μg/m3) n (person-years) Cases HR (95% CI)
aThe annual non–back-extrapolated NOx baseline concentration at the residence of each participant was used as a proxy for exposure to air pollution. bModel 1: age-adjusted. cModel 2: adjusted for baseline age, education, physical activity, smoking, sex, BMI (body mass index), WHR (waist–hip ratio), alcohol, and ApoE4. dModel 3: model 2 plus baseline medical history of diabetes, hypertension, and stroke.
Model 1b
4.8–9 433 (5,024) 55 1.0
> 9–17 476 (5,787) 65 1.10 (0.77, 1.58)
> 17–26 459 (5,182) 90 1.49 (1.07, 2.09)
> 26 438 (4,540) 92 1.57 (1.12, 2.19)
Per 10 μg/m3 1.04 (0.98, 1.11)
Model 2c
4.8–9 381 (4,553) 50 1.0
> 9–17 420 (5,286) 59 1.11 (0.76, 1.63)
> 17–26 408 (4,700) 82 1.48 (1.03, 2.11)
> 26 383 (4,037) 84 1.43 (0.998, 2.05)
Per 10 μg/m3 1.05 (0.98, 1.12)
Model 3d
4.8–9 378 (4,538) 49 1.0
> 9–17 418 (5,264) 59 1.13 (0.77, 1.66)
> 17–26 406 (4,677) 81 1.49 (1.04, 2.14)
> 26 383 (4,037) 84 1.60 (1.02, 2.10)
Per 10 μg/m3 1.05 (0.98, 1.12)