Urinary Metal Levels and Coronary Artery Calcification: Longitudinal Evidence in the Multi-Ethnic Study of Atherosclerosis (MESA)

Abstract

Objective: Growing evidence indicates that exposure to metals are risk factors for cardiovascular disease (CVD). We hypothesized that higher urinary levels of metals with prior evidence of an association with CVD, including non-essential (cadmium , tungsten, and uranium) and essential (cobalt, copper, and zinc) metals are associated with baseline and rate of change of coronary artery calcium (CAC) progression, a subclinical marker of atherosclerotic CVD.

Methods: We analyzed data from 6,418 participants in the Multi-Ethnic Study of Atherosclerosis (MESA) with spot urinary metal levels at baseline (2000-2002) and 1-4 repeated measures of spatially weighted coronary calcium score (SWCS) over a ten-year period. SWCS is a unitless measure of CAC highly correlated to the Agatston score but with numerical values assigned to individuals with Agatston score=0. We used linear mixed effect models to assess the association of baseline urinary metal levels with baseline SWCS, annual change in SWCS, and SWCS over ten years of follow-up. Urinary metals (adjusted to μg/g creatinine) and SWCS were log transformed. Models were progressively adjusted for baseline sociodemographic factors, estimated glomerular filtration rate, lifestyle factors, and clinical factors.

Results: At baseline, the median and interquartile range (25^th, 75^th) of SWCS was 6.3 (0.7, 58.2). For urinary cadmium, the fully adjusted geometric mean ratio (GMR) (95%Cl) of SWCS comparing the highest to the lowest quartile was 1.51 (1.32, 1.74) at baseline and 1.75 (1.47, 2.07) at ten years of follow-up. For urinary tungsten, uranium, and cobalt the corresponding GMRs at ten years of follow-up were 1.45 (1.23, 1.71), 1.39 (1.17, 1.64), and 1.47 (1.25, 1.74), respectively. For copper and zinc, the association was attenuated with adjustment for clinical risk factors; GMRs at ten years of follow-up before and after adjustment for clinical risk factors were 1.55 (1.30, 1.84) and 1.33 (1.12, 1.58), respectively, for copper and 1.85 (1.56, 2.19) and 1.57 (1.33, 1.85) for zinc.

Conclusion: Higher levels of cadmium, tungsten, uranium, cobalt, copper, and zinc, as measured in urine, were associated with subclinical CVD at baseline and at follow-up. These findings support the hypothesis that metals are pro-atherogenic factors.

Keywords: Metals; cadmium; cardiovascular disease; cobalt; copper; coronary artery calcification; longitudinal; mixed models; prospective; tungsten; uranium; zinc.

Figure 1.

Median and interquartile ranges [25^th, 75^th] of urine metal levels (μg/g creatinine) by participant characteristic for non-essential metals cadmium (Cd), tungsten (W), and uranium (U), and essential metals cobalt (Co), copper (Cu), and zinc (Zn). Points represent the median urine metal level and lines correspond to the interquartile range overall and for each subgroup at baseline. The n for each group is on the y-axis. The dotted line represents the overall median urine metal level.

Figure 2.

Geometric mean ratio (GMR) (95% confidence interval) of spatially weighted calcium scores (SWCS) at baseline (blue lines and shaded areas) and at 10-years of follow-up (orange lines and shaded areas) by urinary metal levels (μg/g creatinine) modeled as restricted cubic splines. Lines (shaded areas) represent the GMR (95%CI) of SWCS by metals modeled as restricted cubic splines for log transformed metal distributions with knots at 10^th, 50^th, and 90^th percentiles. The reference value was set at the 10^th percentile. Models were adjusted for age, sex, race and ethnicity, study site, education, eGFR, smoking status, pack-years, physical activity, BMI, systolic blood pressure, antihypertensive medication, LDL-cholesterol, HDL-cholesterol, lipid lowering medications, and diabetes status. The histograms in the background represent the distribution of each metal (μg/g creatinine) at baseline.