Table 1

Summary of studies incorporated in integrated non-linear exposure-response modelling

Reference (year)	Design	Sample size	Exposure source	Mean or median exposure range (SD or IQR)	CKD definition	Adjustments	Exposure contrast	RR (95% CI)	Risk of bias score
Bowe et al (2018)2	Cohort	2 482 737	PM_2.5	11.8 (5.0–22.1)	eGFR <60 mL/min/1.73 m²	Age, race, sex, cancer, cardiovascular disease, chronic lung disease, diabetes mellitus, hyperlipidaemia, hypertension, baseline eGFR, BMI, smoking status, angiotensin-converting enzyme inhibitor/ angiotensin receptor blocker use, county population density, number of outpatient eGFR measurements, number of hospitalisations and county percent in poverty.	Quartile 2 versus 1	1.02 (0.97 to 1.07)	8
							Quartile 3 versus 1	1.07 (1.02 to 1.12)
							Quartile 4 versus 1	1.14 (1.09 to 1.20)
Chan et al (2018)51	Cohort	100 629	PM_2.5	27.1 (5.8–49.6)	eGFR <60 mL/min/1.73 m²	Age, sex, educational level, smoking status, alcohol consumption, BMI, systolic BP, fasting glucose, total cholesterol, self-reported heart disease or stroke and baseline eGFR.	Quintile 2 versus 1	1.05 (0.95 to 1.15)	9
							Quintile 3 versus 1	1.04 (0.94 to 1.15)
							Quintile 4 versus 1	1.11 (1.01 to 1.22)
							Quintile 5 versus 1	1.15 (1.05 to 1.26)
Yang et al (2017)52	Cross-sectional	21 656	PM_2.5	26.6 (5.0)	eGFR <60 mL/min/1.73 m²	Age, sex, fasting glucose, cholesterol, hypertension, BMI, distance to major road, smoking status, alcohol consumption and education level.	Every 5.67 µg/m³ increase	1.03 (0.97 to 1.09)	8
Chen et al (2018)50	Cross-sectional	8497	PM_2.5	24.3 (12.8–48.2)	eGFR <60 mL/min/1.73 m²	Age, sex, BMI, education level, smoking status, alcohol consumption, hypertension and diabetes.	Every 4.1 µg/m³ increase	1.01 (0.96 to 1.06)	9
Bragg-Gresham et al (2018)1	Cross-sectional	1 164 057	PM_2.5	12.2 (6.1–16.8)	eGFR <60 mL/min/1.73 m²	Age, sex, race/ethnicity, hypertension, diabetes and urban/rural status.	Quartile 2 versus 1	1.02 (0.99 to 1.04)	8
							Quartile 3 versus 1	1.01 (0.98 to 1.03)
							Quartile 4 versus 1	1.05 (1.03 to 1.07)
Weaver et al (2018)53	Cross-sectional	5090	PM_2.5	12.2 (0.6)	eGFR <60 mL/min/1.73 m²	Age, sex, BMI, education level, neighbourhood socioeconomic status, medical insurance, smoking status, physical activity, alcohol consumption, occupation, hyperlipidaemia, use of non-steroidal anti-inflammatory drugs, diuretic medication, statin medications, diabetes and hypertension and accounting for clustering by census tract.	Every 1 µg/m³ increase	1.00 (0.82 to 1.22)	9
Jhee et al (2018)*54	Cohort	1948	Passive smoking	–	eGFR <60 mL/min/1.73 m²	Age, sex, BMI, systolic BP, history of hypertension, history of diabetes, alcohol status, education levels, income levels, marital status, haemoglobin and serum albumin.	Moderate secondhand smoke	1.58 (0.94 to 2.66)	9
Jhee et al (2018)*54	Cohort	1948	Passive smoking	–	eGFR <60 mL/min/1.73 m²		Severe secondhand smoke	1.62 (1.03 to 2.63)	9
Ejerblad et al (2004)*55	Case-Control	1924	Active smoking	–	eGFR <60 mL/min/1.73 m²	Age, gender, education level, alcohol consumption, use of paracetamol and salicylates, pipe smoking, cigar smoking and snuff use.	1–10 cigarettes per day versus no smoking	0.89 (0.66 to 2.11)	7
							11–20 cigarettes per day versus no smoking	1.24 (0.96 to 1.60)
							>20 cigarettes per day versus no smoking	1.51 (1.06 to 2.15)
Hall et al (2016)*56	Cohort	3648	Active smoking	–	eGFR decline ≥30%	Age, sex, BMI, diabetes, hypertension, total cholesterol, education level, physical activity, prevalent cardiovascular disease and alcohol consumption.	1–19 cigarettes per day versus no smoking	1.75 (1.18 to 2.59)	6
Hall et al (2016)*56	Cohort	3648	Active smoking	–	eGFR decline ≥30%		>19 cigarettes per day versus no smoking	1.97 (1.17 to 3.31)	6
Hippisley-Cox and Coupland (2010)*57	Cohort	3 156 494	Active smoking	–	eGFR <45 mL/min/1.73 m²	Age, ethnicity, deprivation, smoking, BMI, systolic BP, diabetes, rheumatoid arthritis, cardiovascular disease, treated hypertension, congestive cardiac failure, peripheral vascular disease, use of non-steroidal anti-inflammatory drugs and family history of kidney disease. systemic lupus erythematosus and kidney stones were additional adjusted for models in women.	<10 cigarettes/day versus no smoking in women	1.30 (1.15 to 1.23)	7
							10–19 cigarettes/day versus no smoking in women	1.27 (1.21 to 1.34)
							>19 cigarettes/day versus no smoking in women	1.43 (1.34 to 1.52)
							<10 cigarettes/day versus no smoking in men	1.15 (1.08 to 1.22)
							10–19 cigarettes/day versus no smoking in men	1.24 (1.16 to 1.32)
							>19 cigarettes/day versus no smoking in men	1.25 (1.16 to 1.34)

*Incorporated in models when proxy exposures were included.
BMI, body mass index; BP, blood pressure; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; PM_2.5, ambient fine particulate matter.