Skip to main content
. Author manuscript; available in PMC: 2014 Aug 25.
Published in final edited form as: Epidemiology. 2013 May;24(3):421–429. doi: 10.1097/EDE.0b013e31828b0631

Table 3.

Hazard ratios (95% confidence interval) for incident cardiovascular endpoints by urine cadmium concentrations (n=3,348)

Cadmium Quartiles (μg/g) 80th vs 20th
≤ 0.61a HR (95% CI) 0.62 – 0.92 HR (95% CI) 0.93 – 1.45 HR (95% CI) > 1.45 HR (95% CI) percentile HR (95% CI) Test for trendb
Cardiovascular Disease
No. Cases / Non Cases: (249 / 613) (263 / 554) (269 / 568) (303 / 529) (1084 / 2264)
Model 1c 1.00 1.17 (0.98–1.40) 1.21 (1.01–1.46) 1.34 (1.11–1.62) 1.19 (1.07–1.31) P= 0.001
Model 2d 1.00 1.26 (1.05–1.50) 1.38 (1.14–1.66) 1.58 (1.30–1.92) 1.28 (1.16–1.42) P <0.001
Model 3e 1.00 1.20 (1.00–1.44) 1.30 (1.07–1.58) 1.48 (1.21–1.80) 1.24 (1.11–1.38) P< 0.001
Coronary Heart Disease
No. Cases / Non Cases (193 / 669) (182 / 635) (198 / 639) (193 / 639) (766 / 2582)
Model 1c 1.00 1.09 (0.89–1.34) 1.19 (0.96–1.47) 1.18 (0.95–1.47) 1.15 (1.02–1.30) P =0.02
Model 2d 1.00 1.18 (0.96–1.45) 1.38 (1.11–1.71) 1.41 (1.12–1.78) 1.26 (1.12–1.42) P <0.001
Model 3e 1.00 1.13 (0.92–1.40) 1.31 (1.05–1.63) 1.33 (1.05–1.68) 1.22 (1.08–1.38) P =0.002
Stroke
No. Cases / Non Cases (51 / 811) (51 / 766) (53 / 784) (89 / 743) (244 / 3104)
Model 1c 1.00 1.14 (0.78–1.68) 1.14 (0.77–1.68) 1.91 (1.30–2.80) 1.73 (1.22–2.45) P =0.006
Model 2d 1.00 1.24 (0.84–1.84) 1.22 (0.81–1.82) 2.03 (1.35–3.06) 1.82 (1.25–2.65) P =0.006
Model 3e 1.00 1.21 (0.81–1.80) 1.11 (0.73–1.69) 1.87 (1.22–2.86) 1.75 (1.17–2.59) P =0.03
Heart Failure
No. Cases / Non Cases (60 / 802) (81 / 736) (87 / 750) (100 / 732) (328 / 3020)
Model 1c 1.00 1.35 (0.96–1.89) 1.38 (0.98–1.95) 1.41 (0.98–2.03) 1.20 (0.88–1.63) P =0.03
Model 2d 1.00 1.49 (1.06–2.11) 1.65 (1.15–2.35) 1.78 (1.22–2.59) 1.46 (1.06–2.02) P =0.002
Model 3e 1.00 1.40 (0.99–1.99) 1.52 (1.06–2.18) 1.61 (1.10–2.36) 1.39 (1.01–1.94) P =0.01
a

Reference category.

b

Models comparing the 80th (1.62 μg/g) to the 20th (0.55 μg/g) percentiles of urine cadmium distributions and associated test for trend were obtained from Cox proportional hazards models with log-transformed cadmium as a continuous variable, except for stroke and heart failure where cadmium was introduced as restricted quadratic splines with knots at percentiles 10th, 50th and 90th because of the non-linear relationships.

c

Model 1 adjusted for sex

d

Model 2 further adjusted for post-menopausal status for women (yes, no), education (< high school, ≥ high school), body mass index (kg/m2), total cholesterol (mg/dL), estimated LDL- cholesterol (mg/dL), hypertension (yes, no), diabetes (yes, no), and estimated glomerular filtration rate (ml/min/1.73m2).

e

Model 3 further adjusted for smoking status (never, former, current) and cumulative smoking dose (pack-years modeled as restricted cubic splines with knots at 10, 20 and 30 pack-years).