Long-Term Air Pollution Exposure Linked to Incident CKD
Key Takeaways
- Long-term residential exposure to PM2.5, PM10, NO2, and black carbon was associated with modestly higher incident CKD in the main adjusted analysis.
- The NO2 association was stronger among never-smokers, and NO2 and black carbon remained positive in two-pollutant models while PM2.5 and PM10 attenuated.
- Alternative CKD-definition and competing-risk analyses were broadly similar, while the female-only cohort, hospital-based CKD ascertainment, and residential exposure modeling limit interpretation.
In the Danish Nurse Cohort analysis, 24,581 female nurses recruited in 1993 or 1999 were followed until December 31, 2018. Mean follow-up was 21.2 years. Incident CKD was defined as the first hospital contact with a primary or secondary CKD diagnosis. Annual mean PM2.5, PM10, NO2, and black carbon were estimated at residential addresses with the DEHM/UBM/AirGIS system, and 14-year moving averages were analyzed with Cox regression models.
In the fully adjusted analysis, hazard ratios per interquartile-range increase were 1.18 for PM2.5, 1.14 for PM10, 1.13 for NO2, and 1.09 for black carbon. The corresponding interquartile ranges were 2.86 micrograms per cubic meter for PM2.5, 3.33 for PM10, 8.09 for NO2, and 0.34 for black carbon. Baseline mean exposures were 13.3 micrograms per cubic meter for PM2.5, 19.7 for PM10, 19.5 for NO2, and 0.9 for black carbon. Overall, the estimates indicated modest positive associations across the pollutant set.
The NO2 association was greater among never-smokers than among ever-smokers, with an interaction P value of 0.034. The investigators interpreted that subgroup pattern cautiously, as exploratory rather than definitive. NO2 and black carbon remained positively associated in two-pollutant models, whereas PM2.5 and PM10 attenuated after adjustment for combustion-related co-pollutants. Results were broadly similar with an alternative CKD definition, in Fine-Gray competing-risk analyses, and across 1-, 5-, and 10-year exposure windows, although estimates for shorter windows were less precise and in some cases included the null; the most consistent associations were seen for NO2 and black carbon.
This observational cohort included only female nurses, narrowing generalizability beyond that occupational population. Hospital registry definitions may miss earlier-stage CKD, and residential-address exposure estimates did not capture commuting, occupational, or indoor exposures. The authors also acknowledged residual confounding, selection bias, limited case numbers, and wide confidence intervals, especially for PM2.5 and NO2.
They concluded that long-term ambient air pollution exposure was positively associated with CKD incidence and that black carbon may mark a relevant combustion-related pollutant mixture.