The association of long-term exposure to PM2.5 on all-cause mortality in the Nurses’ Health Study and the impact of measurement-error correction

Table 2 Associations of 12-month average PM _2.5 (per 10 μg/m ³ increase) from spatio-temporal model predictions or the nearest USEPA monitoring location values with all-cause mortality, with and without adjustment for exposure measurement error

	Spatio-temporal model	Nearest USEPA monitor
Cases	8,617	8,617
Person-years	628,186	628,186
Basic HR (95%CI)¹	1.20 (1.11, 1.29)	1.14 (1.06, 1.22)
Basic Measurement Error Adjusted HR (95%CI)^1,2	1.27 (1.08, 1.48)	1.26 (1.07, 1.48)
% increase in HR³	5.8%	10.5%
% increase in 95% CIs⁴	122.2%	156.3%
Multivariable HR (95%CI)⁵	1.13 (1.05, 1.22)	1.12 (1.05, 1.21)
Multivariable Measurement Error Adjusted HR (95%CI)^2,5	1.18 (1.02, 1.36)	1.22 (1.02, 1.45)
% increase in HR³	4.4%	8.9%
% increase in 95% CIs⁴	100.0%	168.8%

¹Basic model: models stratified by age in months, adjusted for race, region, year and season.
²Additionally adjusted for exposure measurement error.
³[(HR – HR _{measurement error})/HR]*100.
⁴[((UCL–LCL)-(UCL _{measurement error} –LCL _{measurement error}))/ (UCL–LCL)]*100.
⁵Multivariable: models stratified by age in months, adjusted for race. region, year, season, smoking status, pack-yrs, family history of MI, BMI, hypercholesterolemia, median family income in census tract of residence, median house value in census tract of residence, physical activity, race, Alternate Healthy Eating Index (AHEI), individual level socioeconomic status (nurses’ education level, occupation of both parents, marital status, and husband’s education).

ISSN: 1476-069X