Associations between acute and long-term exposure to PM2.5 components and temperature with QT interval length in the VA Normative Aging Study

Abstract

Aims: Our study adds to the sparse literature on the effect of multiple fine particulate matter (PM2.5) components on QT interval length, an outcome with high clinical relevance in vulnerable populations. To our knowledge, this is the first study to examine the association between spatiotemporally resolved exposures to PM2.5 components and QT interval length.

Methods and results: Among 578 men living in Eastern Massachusetts between 2000 and 2011, we utilized time-varying linear mixed-effects regressions with a random intercept to examine associations between acute (0-3 days), intermediate (4-28 days), and long-term (1 year) exposure to PM2.5 components, temperature, and heart-rate corrected QT interval (QTc). Each of the PM2.5 components and temperature was geocoded to the participant's residential address using validated ensemble and hybrid exposure models and gridMET predictions. We also evaluated whether diabetic status modified the association between PM2.5 components and QTc interval. We found consistent results that higher sulfate levels and colder temperatures were associated with significant longer QTc across all moving averages except the day of exposure. The greatest effect of sulfate and temperature was detected for the 28-day moving average. In the multi-pollutant model, each 1.5 µg/m3 IQR increase in daily sulfate was associated with a 15.1 ms [95% confidence interval (CI): 10.2-20.0] increase in QTc interval and in the single-pollutant models a 15.3 ms (95% CI: 11.6-19.1) increase in QTc interval. Other secondary particles, such as nitrate and organic carbon, also prolonged QT interval, while elemental carbon decreased QT interval. We found that diabetic status did not modify the association between PM2.5 components and QTc interval.

Conclusion: Acute and long-term exposure to PM2.5 components and temperature are associated with changes in ventricular repolarization as measured by QT interval.

Keywords: Air pollution; Arrhythmias; QT interval; Temperature; Ventricular repolarization.

Figure 1

Changes in milliseconds and 95% confidence interval in QTc interval for an interquartile range increase in 0–28-day moving average and 1-year moving average of each PM_2.5 component. The results are presented in a multi-pollutant model where all the PM_2.5 components and temperature are included in the same model and the single-pollutant model, which includes each individual PM_2.5 component or temperature. The models are adjusted for age (years), race, maximum years of education, body mass index (kg/m²), total cholesterol (mg/dL), mean arterial pressure (mmHg), diabetic status, use of beta-blocker medication, alcohol intake (2> drinks/day or < 2 drinks/day as reference), smoking status (current, former, or never as reference), cumulative smoking pack-years, census tract percent of population age 25 years or older with less than a high school diploma, relative humidity (%), and seasonality (sine and cosine).

Figure 2

Change in QTc interval length and 95% confidence interval for an interquartile range increase in 0–28-day and 1-year moving averages of each PM_2.5 component and temperature, but not including PM_2.5 mass. The results are presented in a multi-pollutant model where all the PM_2.5 components and temperature are included in the same model except for the indicated PM_2.5 component. The model was adjusted for age (years), race, maximum years of education, body mass index (kg/m²), total cholesterol (mg/dL), mean arterial pressure (mmHg), diabetic status, use of beta-blocker medication, alcohol intake (2> drinks/day or < 2 drinks/day as reference), smoking status (current, former, or never as reference), cumulative smoking pack-years, census tract percent of population age 25 years or older with less than a high school diploma, relative humidity (%), and seasonality (sine and cosine).

Figure 3

Change in QTc interval length and 95% confidence interval for an interquartile range increase in 0–28-day and 1-year moving for each PM_2.5 component and temperature in the multi-pollutant model among diabetic and non-diabetic individuals. The model was adjusted for the main effects of each PM_2.5 component and temperature, the interaction term between the PM_2.5 component and the dichotomous indicator variable for diabetic status, age (years), race, maximum years of education, body mass index (kg/m²), total cholesterol (mg/dL), mean arterial pressure (mmHg), diabetic status, use of beta-blocker medication, alcohol intake (2> drinks/day or < 2 drinks/day as reference), smoking status (current, former or never as reference), cumulative smoking pack-years, census tract percent of population age 25 years or older with less than a high school diploma, relative humidity (%), and seasonality (sine and cosine).