Traffic-related air pollution and QT interval: modification by diabetes, obesity, and oxidative stress gene polymorphisms in the normative aging study

Abstract

Background: Acute exposure to ambient air pollution has been associated with acute changes in cardiac outcomes, often within hours of exposure.

Objectives: We examined the effects of air pollutants on heart-rate-corrected QT interval (QTc), an electrocardiographic marker of ventricular repolarization, and whether these associations were modified by participant characteristics and genetic polymorphisms related to oxidative stress.

Methods: We studied repeated measurements of QTc on 580 men from the Veterans Affairs Normative Aging Study (NAS) using mixed-effects models with random intercepts. We fitted a quadratic constrained distributed lag model to estimate the cumulative effect on QTc of ambient air pollutants including fine particulate matter <or= 2.5 microm in aerodynamic diameter (PM2.5), ozone (O3), black carbon (BC), nitrogen dioxide (NO2), carbon monoxide (CO), and sulfur dioxide (SO2) concentrations during the 10 hr before the visit. We genotyped polymorphisms related to oxidative stress and analyzed pollution-susceptibility score interactions using the genetic susceptibility score (GSS) method.

Results: Ambient traffic pollutant concentrations were related to longer QTc. An interquartile range (IQR) change in BC cumulative during the 10 hr before the visit was associated with increased QTc [1.89 msec change; 95% confidence interval (CI), -0.16 to 3.93]. We found a similar association with QTc for an IQR change in 1-hr BC that occurred 4 hr before the visit (2.54 msec change; 95% CI, 0.28-4.80). We found increased QTc for IQR changes in NO2 and CO, but the change was statistically insignificant. In contrast, we found no association between QTc and PM2.5, SO2, and O3. The association between QTc and BC was stronger among participants who were obese, who had diabetes, who were nonsmokers, or who had higher GSSs.

Conclusions: Traffic-related pollutants may increase QTc among persons with diabetes, persons who are obese, and nonsmoking elderly individuals; the number of genetic variants related to oxidative stress increases this effect.

Figure 1

Effect estimates of change in mean QTc per IQR change in cumulative exposure to pollutant during the 10 hr before ECG measurement: single-pollutant model with random intercept using QCDL. All models were adjusted for age; BMI; MAP; cholesterol; diabetic status (doctor’s diagnosis of disease or FBG > 126 mg/dL vs. no diagnosis or FBG ≤ 126 mg/dL); alcohol intake (≥ 2 drinks/day, < 2 drinks/day); cigarette smoker (never, ever); use of alpha blockers, angiotensin-converting enzyme inhibitors, angiotensin receptor antagonist/blockers, beta blockers, or calcium channel blockers; day of the week (Monday through Sunday); temperature; and a natural spline for long-term time trend (date). Error bars indicate 95% CI. *p < 0.10.

Figure 2

Effect estimates of change in mean QTc per IQR change of traffic-related pollutant. Error bars and hatch-marked regions indicate 95% CI. (A–C) Single-pollutant model of cumulative exposure for a 10-hr time window with random intercept using QCDL with hourly lags: BC (A), NO₂ (B), and CO (C). (D–F) Single-pollutant model of hourly exposure with random intercept using hourly lags: BC (D), NO₂ (E), and CO (F).

Figure 3

Adjusted change in mean QTc per IQR change of traffic-related pollutant: hourly exposure with a 4-hr lag (A) or cumulative exposure during 10 hr before ECG measurement (B), by participant characteristics [diabetic status (doctor’s diagnosis of diabetes or FBG > 126 mg/dL, vs. no diagnosis or FBG ≤ 126 mg/dL), smoking (never, ever), and obesity (BMI ≥ 30, < 30)]. Error bars indicate 95% CI. *p-Value interaction < 0.01; **p-Value interaction < 0.05.

Figure 4

Adjusted change in mean QTc per IQR change of traffic-related pollutant: hourly exposure with a 4-hr lag (A) or cumulative exposure during 10 hr before ECG measurement (B), by GSS (low GSS, high GSS). Error bars indicate 95% CI. *p-Value interaction < 0.01; **p-value interaction < 0.10.