Background It has been reported that removing all anthropogenic particulate matter air pollution would increase UK life expectancy. Air pollution has been shown to be associated with cardio-respiratory mortality. Nitrogen dioxide (NO₂), an air pollutant, has been shown to be strongly positively correlated with particulate matter. Many studies rely on comparisons between cities using single exposure measures for each city; greater data availability for NO₂ makes it possible to examine the relationship with varying exposures across small areas. We investigate changes in NO₂ over two time periods, 2006–08 and 2009–11, and the relationship with cardio-respiratory mortality before and after adjustment for area deprivation.

Methods NO₂ was recorded at monitoring sites in 141 datazones (mean population 905) across seven local authorities in west central Scotland in 2007 and 2010. We used geocoded mortality and population data for 2006–08 and 2009–11 by age and sex for cardio-respiratory deaths. The Scottish Index of Multiple Deprivation (SIMD) was used to assess deprivation. Multilevel Poisson regression modelled the effect of NO₂ on death, adjusted for age and sex, before and after adjustment for deprivation. Analyses were performed separately for under 65 years and 65+.

Results There were 1848 and 1714 cardio-respiratory deaths during 2006–08 (2007 population 124,603) and 2009–11 (2010 population 130,496) respectively. Annual mean NO₂ across all datazones significantly increased (p = 0.001) from 32µgm^-3 (SD 11.5) in 2007 to 34µgm^-3 (SD 9.9) in 2010. The increase in NO₂ did not differ significantly by SIMD (p > 0.05); however, mean NO₂ in both periods was significantly associated with SIMD (p_trend <0.001). The relationship between air pollution and death did not differ significantly in the two periods in either age group (p > 0.05) thus analyses were not further stratified by period. The association between NO₂ and mortality was only significant in the under 65s; relative risk for a 10 µgm^-3 increase in NO_2, (RR)=1.14 (95% CI=1.02–1.28) for under 65s and 1.02 (0.96–1.08) for 65+. The significant association in the under 65s disappeared after adjustment for deprivation RR=1.06 (0.98–1.16).

Conclusion Annual mean NO₂ significantly increased over time across small areas in west central Scotland; however, there was less variation in NO₂ exposure. NO₂ was associated with increased risk of cardio-respiratory death in under 65s, but this effect was attenuated by deprivation. Higher levels of NO₂ in the most deprived areas highlight the risk that deprivation may confound the relationship between air pollution and mortality. Such issues should be considered in developing public health responses to premature mortality.