A Daily Diary Panel Study of Congestive Heart Failure: Associations Between Health Outcomes and Air Pollution and Weather

Abstract

P-019 Introduction: Time series studies of persons with congestive heart failure (CHF) suggest that their health may be affected greatly from exposure to ambient air pollution. The present daily diary panel study was conducted to determine whether there were associations between intermediate health outcomes (e.g., oxygen saturation) and exposure to daily levels of ambient air pollution and weather. Methods: Thirty-one subjects, over the age of 45 years (87% male), with a confirmed diagnosis of CHF were recruited from the Heart Failure Clinic of the Royal Victoria Hospital, Montreal, between 2002 and 2003. Over a two month period, subjects were asked every morning to record values of their oxygen saturation, pulse rate, body temperature, health status, and other factors that may affect their health. We had on hand daily averages of essential weather indicators and concentrations of criteria air pollutants as measured at fixed-site monitoring stations. Mixed linear regression using an AR1 error structure were used to model the various outcome variables according to daily personal covariates and average levels of air pollution and weather, after accounting for temporal trends (random effects). Results: From adjusted multilevel models, we found numerous effects in many of the health outcome variables under consideration. For example, for mean changes in oxygen saturation (assessed across interquartile ranges (IQ)): +0.098% (95%CI:0.032–0.163%) for an increase in barometric pressure from the previous day of 1.09kPa; -0.084% (95%CI:−0.165 – −0.004%) for increases in ozone of 23.73 μg/m3; and −0.259% (95%CI:−0.364 – −0.155%) for an increase in body temperature of 0.8° C. In other single pollutant, but otherwise adjusted models, associations with oxygen saturation were also found for SO2 (−0.104% (95%CI: −0.173 – −0.034%), IQ=8.59 μg/m3) and for PM2.5 (−0.066% (95%CI:−0.122 – −0.010%), IQ=7.28 μg/m3). Discussion and Conclusions: Our findings for body temperature and decreased barometric pressure are biologically coherent. The findings for air pollution are consistent with the results of time series studies that show that air pollution and weather affect mortality and hospitalizations. The effects found in this study are small, especially in comparison to what would be found from between-person comparisons, but reflect the expected smaller within-person temporal changes which were found to correlate with daily fluctuations. The mechanisms responsible for these effects may be related to some of the current theories of general and specific inflammation of air pollution as well as vasoconstriction that may be due to increases in endothelins.