Table 3.

Association of air pollutant concentration with annual eGFR slope by a linear regression model with restricted cubic splines.

	Model 1		P-value	Model 2		P-value
	Regression coefficient	95% CI		Regression coefficient	95% CI
CO, ppm
β0	15	(− 38–68)	0.57	187	(− 206–580)	0.35
Spl_β1	− 45	(− 177–87)	0.50	− 41	(− 165–84)	0.52
Spl_β2	2358	(− 2233–6948)	0.31	2037	(− 2697–6771)	0.40
Spl_β3	− 5475	(− 14,429–3480)	0.23	− 4307	(− 13,390–4776)	0.35
Spl_β4	4988	(− 1216–11,191)	0.11	3781	(− 2394–9956)	0.23
NO, ppb
β0	6	(− 28–41)	0.72	54	(− 16–125)	0.13
Spl_β1	− 4	(− 18–11)	0.63	− 6	(− 22–9)	0.42
Spl_β2	10	(− 117–138)	0.88	2	(− 134–138)	0.98
Spl_β3	4	(− 329–337)	0.98	49	(− 289–387)	0.77
Spl_β4	− 47	(− 444–349)	0.82	− 92	(− 470–287)	0.63
NO2, ppb
β0	14	(− 29–56)	0.52	62	(− 97–220)	0.44
Spl_β1	− 2	(− 5–2)	0.44	− 2	(− 5–2)	0.36
Spl_β2	2	(− 2–6)	0.31	2	(− 2–6)	0.25
Spl_β3	− 5	(− 13–4)	0.28	− 5	(− 13–3)	0.19
Spl_β4	3	(− 3–9)	0.26	5	(− 1–10)	0.12
NOx, ppb
β0	11	(− 33–55)	0.63	15	(− 118–148)	0.82
Spl_β1	− 1	(− 4–2)	0.54	− 1	(− 4–2)	0.52
Spl_β2	1	(− 2–4)	0.43	1	(− 2–4)	0.45
Spl_β3	− 3	(− 8–3)	0.40	− 3	(− 8–3)	0.36
Spl_β4	2	(− 3–7)	0.39	3	(− 2–7)	0.22
O3, ppb
β0	− 57	(− 421–307)	0.76	222	(− 135–579)	0.22
Spl_β1	2	(− 10–14)	0.77	− 1	(− 12–11)	0.92
Spl_β2	0	(− 29–29)	0.99	1	(− 28–30)	0.95
Spl_β3	− 6	(− 63–50)	0.82	− 1	(− 60–59)	0.99
Spl_β4	14	(− 25–52)	0.48	2	(− 40–45)	0.91
PM10, μg/m3
β0	82	(− 95–260)	0.36	155	(− 86–396)	0.21
Spl_β1	− 1	(− 4–2)	0.34	− 1	(− 4–2)	0.45
Spl_β2	2	(− 1–4)	0.17	1	(− 1–3)	0.26
Spl_β3	− 3	(− 8–1)	0.12	− 3	(− 7–2)	0.21
Spl_β4	3	(0–6)	0.06	2	(− 1–6)	0.12
PM2.5, μg/m3
β0	73	(− 39–186)	0.20	162	(− 74–397)	0.18
Spl_β1	− 2	(− 6–1)	0.18	− 2	(− 5–2)	0.36
Spl_β2	3	(0–7)	0.08	2	(− 1–6)	0.20
Spl_β3	− 7	(− 14–0)	0.05	− 5	(− 12–2)	0.17
Spl_β4	5	(1–10)	0.03	4	(− 1–9)	0.11
SO2, ppb
β0	− 7	(− 33–18)	0.57	81	(− 7–170)	0.07
Spl_β1	2	(− 6–10)	0.63	1	(− 7–10)	0.80
Spl_β2	− 9	(− 28–9)	0.33	− 5	(− 24–15)	0.62
Spl_β3	22	(− 16–61)	0.25	9	(− 30–48)	0.65
Spl_β4	− 21	(− 54–11)	0.20	− 4	(− 36–29)	0.83

CO, carbon monoxide; NO, nitrogen monoxide; NO₂, nitrogen dioxide; NO_x, nitrogen oxide; O₃, ozone; PM₁₀, particulate matter < 10 μm in aerodynamic diameter; PM₂₅, particulate matter < 2.5 μm in aerodynamic diameter; SO₂, sulfur dioxide; ppb, parts per billion; ppm, parts per million; μg/m³, micrograms per cubic meter; Q, quartile; CI, confidence interval.

The 5-knot (percentile: 5, 27.5, 50, 72.5, and 95) restricted cubic spline function was determined for all studied air pollutant concentrations. The knots for CO, NO, NO₂, NO_X, O₃, PM₁₀, PM_2.5, and SO₂ are (0.42, 0.44, 0.47, 0.48, and 0.49), (2.36, 2.39, 2.41, 2.43, and 2.55), (12.15, 13.19, 13.69, 13.91, and 14.21), (14.53, 15.66, 16.10, 16.28, and 16.54), (30.92, 31.31, 31.53, 31.71, and 31.89), (62.45, 63.62, 64.71, 65.22, and 65.84), (34.31, 35.15, 36.06, 36.49, and 36.99), and (3.38, 3.5, 3.58, 3.67, and 4.09), respectively. β₀ is a coefficient of intercept, and Spl_β₁- Spl_β₄ are coefficients for the restricted cubic spline functions of studied air pollutant concentrations. Model 1 is estimated by a general linear model for the annual estimated glomerular filtration rate slope after only putting one restricted cubic spline function of studied air pollutant concentration. Independent variables in Model 2 are listed as the independent variable in model 1, and age, sex, educational level, diabetes mellitus, hypertension, cerebrovascular accident, congestive heart failure, ischemic heart disease, gout, angiotensin-converting enzyme inhibitor/ angiotensin II receptor blocker, temperature, and log (urine protein creatinine ratio).