Impact of Lifetime Exposure to Air and Noise Pollution on Cardiovascular Disease and Mortality - the AIRCARD Study.

The purpose of this study is to determine the impact of individually accumulated exposure to air and noise pollution, as well as exposure fluctuations (peaks), on the incidence and prevalence of CVD. Noise pollution can be a significant confounder for the relation investigated and we will therefore include accumulated data for noise over time together with air pollution data.

The ICD or SKS disease classification code system will be used to identify clinical events for patients from the two trials (described below). CVD mortality and total mortality as well as clinical events from myocardial infarctions, percutaneous coronary interventions, coronary artery bypass grafts, heart failure, apoplexies, and bypass will be examined. The analysis will be stratified by socioeconomic status to determine if there is a social inequality in the air and noise pollution impact.

We will examine the individual air pollution exposure as well as noise exposure to understand each pollutants' contribution to the accumulated risk and demonstrate the impact of air and noise pollution on CVD morbidity and mortality. We hypothesize that:

1. the amount of accumulated air as well as noise pollution is a highly independent significant risk factor for clinical CVD, when accounting for all traditional cardiovascular risk factors.
2. certain subdivisions of the air pollution exposure carry the highest contribution for development of clinical CVD. This will provide novel and valuable information as this has never been investigated with the precision we can deliver.
3. accumulated air and noise pollution from both individual home addresses and work addresses together better predicts clinical CVD than models based solely on home addresses. No study has evaluated this before.

Furthermore, the impact on the result of social inequality will be thoroughly examined.

This study is designed as a prospective registry-based observational study using modelled air/noise pollution data. The population is predominantly males from two Danish clinical trials (DANCAVAS and VIVA trials)

DANCAVAS I and II are two similar population-based randomized, multicentered, clinically controlled studies designed to evaluate the benefits of 7-step multiple cardiovascular screening and modern vascular prophylaxis in a population of men and a small subpopulation of females, aged 60-74 years, living in the southern part of Denmark. For each participant, all relevant CVD risk factors were measured or determined.

The screening included: 1) low-dose non-contrast CT scan to detect CAC and aortic/iliac aneurysms, 2) ankle-brachial blood pressure index (ABI) to detect peripheral arterial disease (PAD) and hypertension, 3) a telemetric assessment of the heart rhythm, and 4) a measurement of the cholesterol and plasma glucose levels.

The Viborg Vascular (VIVA) screening trial (44) is a randomized, multicentered, population-based clinically controlled study designed to evaluate the benefits of vascular screening and modern vascular prophylaxis in men between 65-74 years of age living in the region of Mid Denmark (Viborg County). For each participant, abdominal ultrasound scanning of the infrarenal aorta were performed to detect abdominal aortic aneurism (AAA), and ABI were measured to detect PAD and hypertension. In all, 18,749 men were screened.

A total of 33,723 participants in the age of 60-74 are included from the two cohorts. In both trials, an AAA was defined as maximal infrarenal diameter of 30 mm or more, and PAD was defined as an ABI < 0.90 or >=1.40 using the same validated hand held Doppler-based methodology (46). We have accounted for traditional CVD confounders in all participants. This is unique for our study. Previous studies on air pollution and CVD burden have not been able to prospectively consider these confounding variables to the same extent.

In Denmark a validated and reliable air pollution model system is available. The system is named DEHM/UBM/AirGIS and consists of three coupled models; the Danish Eulerian Hemispheric Model (DEHM), the Urban Background Model (UBM) and the Operational Street Pollution Model (OSPM) and a GIS system (AirGIS) that couples the modelled concentrations with the address level of the population. The system calculates air pollution concentrations of 80 chemical species as well as air pollution levels in cities, in streets and on address level even on both side of the street. These pollution levels can be calculated back to 1979 giving retrospectively data 40 years back. The model system is validated in relation to air pollution measurements throughout Denmark back to 1990 with high correlation between model estimated values and measured values. This multi-scale model system is unique, capable of running on very high temporal (hourly) and spatial (address level) resolutions. The development of the models and the calculation of air pollution and measurements is performed at Aarhus University, Department of Environmental Science (ENVS). The model is robust; taking all necessary factors into account that could contribute to the individual life-long air pollution exposure and the model is one of the best in the world.

Noise is calculated using state-of-the-art algorithms implemented in a well-known software, the SoundPLAN. The algorithms, reflecting advanced physics and mathematics-based knowledge, consider the propagation of sound in the atmosphere as well as the sound originating from the source, e.g., road transport, railway. Here, information from various national registers, such as the national traffic database, including traffic counts, travel speeds, the building register, the address register, and the Danish surface and elevation model, to name a few. In addition, advanced weather classes reflecting all meteorological conditions in Denmark are used in noise calculations. The model output is a noise estimate at the address location or any location of interest in Denmark, which can be subsequently used to investigate the health impacts of short-term and/or long-term noise exposure.

The entire study population is monitored until December 31. 2022, in the Danish national registers.

Primarily, a multivariate Cox proportional hazards regression model will be used to examine the associations between air- and noise pollution and all-cause mortality and CVD morbidity and mortality when adjusting for inclusion date, sex, and other potential confounding factors at baseline.

Descriptive statistics of the main study variables to examine noise pollution will be presented in tables, frequencies with percentages of dichotomous/ordinal variables and medians with quartiles for numeric variables. Outcomes will be analyzed both as dichotomized variables, present CAC, severe CAC (CAC score above 400), AAA and PAD, and as the underlying continuous variables, CAC score, aortic diameter, and lowest measured ankle-brachial blood pressure index. Confounding CVD variables will be considered and adjusted for (e.g., lifestyle, medical history, QoL, weight, height, waist circumference, blood pressure, ABI, calcium score for coronary arteries, aortic dimensions, HbA1c and lipid parameters).

To investigate hypothesis about noise we will, for each dichotomized outcome, estimate adjusted odds ratios using multivariable logistic regression, with cumulated noise pollution as exposure and the Framingham risk factors as confounders. The continuous variables will be analyzed in multivariable linear regression models and the cardiovascular risk factors as confounders.

To investigate the impact of work versus home address, we will compare regression models with accumulated air pollution from both work and home addresses with models with air pollution only from home addresses. ROC curves from both models will be compared to evaluate the hypothesis.