2011 Progress Report: Are Diabetics and the Neurologically Impaired at Increased Risk from Air Pollutant Exposures? A National AnalysisEPA Grant Number: R834900
Title: Are Diabetics and the Neurologically Impaired at Increased Risk from Air Pollutant Exposures? A National Analysis
Investigators: Zanobetti, Antonella , Dominici, Francesca , Schwartz, Joel , Koutrakis, Petros , Wang, Yun
Institution: Harvard T.H. Chan School of Public Health , The Johns Hopkins University
Current Institution: Harvard T.H. Chan School of Public Health
EPA Project Officer: Chung, Serena
Project Period: April 1, 2011 through March 31, 2014 (Extended to March 31, 2015)
Project Period Covered by this Report: April 1, 2011 through March 31,2012
Project Amount: $299,903
RFA: Exploring New Air Pollution Health Effects Links in Existing Datasets (2010) RFA Text | Recipients Lists
Research Category: Air Quality and Air Toxics , Health Effects , Air
We focused this project on two susceptible populations: individuals with neurological disorder, and individuals with diabetes. Using Medicare data to select the susceptible populations, we will estimate county specific mortality risks associated with both short- and long-term exposure to individual pollutants on a national scale. Then we will identify factors that could explain the heterogeneity of these air pollution mortality risks. The specific aims of this proposal are:
Aim 1: To estimate the chronic effects on mortality of long-term exposure to individual pollutants in several US counties in two susceptible populations defined as individuals with neurological disorders or diabetes.
Aim 2: To estimate the acute effects on mortality of short term effects of individual pollutants in a potentially susceptible population.
Aim 3: To investigate whether markers of susceptibility and vulnerability differentially influence the previously established relationships between individual pollutants and mortality, allowing us to identify subpopulations at increased risk for harmful effects of air pollution. Moreover we will examine effect modification due to the composition of multi-pollutant mixtures and to PM composition.
First we worked on preparing the data (see points 1 and 2). We then started to analyze the data, by first examining the acute effects on mortality of short term effects of PM2.5 in individuals with neurological disorder and diabetes. (Aim 2)
1) Medicare data
We used the Medicare beneficiary denominator file from CMS to identify beneficiaries who were enrolled in the Medicare FFS plan between January 2000 and December 2008. The denominator file contains information on beneficiaries’ eligibility and enrollment in Medicare and the date of death that was ascertained through the corresponding vital status file from CMS. The initial denominator file included near 400 million beneficiaries across the study period, of which we excluded beneficiaries with age<65 years, those enrolled in managed-care programs over an entire year, and those who resided outside of the targeted counties. We calculated person-years for each beneficiary to account for new enrollment, disenrollment, or death during an index year. We then linked the person-years beneficiary data with the Medicare Provider Analysis and Review (MEDPAR) inpatient data to identify all Medicare FFS patients who were hospitalized for the following targeted medical conditions between January 1, 2000 and December 31, 2008:
1. diabetes ICD-9: 250;
2. Acute myocardial infarction (AMI, ICD-9: 410);
3. Dementia ICD-9: 290
4. Neurological disorders, ICD-9: 320-359.
5. Inflammatory diseases of the central nervous system (CNS) ICD-9: 320-326;
6. Hereditary and degenerative diseases of the central nervous system ICD-9: 330-337;
7. Alzheimer's disease ICD-9: 331.0;
8. Parkinson's disease ICD-9: 332;
9. Other disorders of the central nervous system ICD-9: 340-349; d
10. Disorders of the peripheral nervous system (PNS) ICD-9: 350-359.
The MEDPAR inpatient data includes information on patient demographics (age, sex, race), dates of admission and discharge, date of death and verification code for death, admission sources and types, discharge dispositions, principal and secondary diagnosis codes, and procedure codes, defined by the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). We excluded patients who could not be merged with the Medicare denominator file. One reason for the unsuccessful merges is the incorrect MEDPAR beneficiary identification code or sex code. Patients’ states of residence information were obtained from the Medicare beneficiary denominator file.
2) Air pollution, weather, and Census data
We obtained daily air pollution (PM2.5, NO2, CO, and Ozone) from the U.S. EPA Air Quality System website, and weather from the NOAA website. Key variables in the pollution data include monitor ID, date, FIPS, and sample value. Key variables in the weather data include daily temperature (mean, min, max), daily dew point, date, and weather station and WBAN IDs.
Preliminary results: we submitted the following abstract to the ISEE entitled “PM2.5 and mortality risk: Modification by diabetes and the neurological impairment: A National analysis.” I am adding here the abstract which describe our preliminary results.
Background: Short-term exposure to ambient air pollution is associated with mortality, and diabetes, Parkinson’s disease, dementia, and Alzheimer's disease are also a growing burden.
Objectives: PM2.5 increase pollution-associated mortality risk among elderly and effect modification by neurological disorders or diabetes.
Methods: We examined 8 million deaths among all Medicare enrollees from 1999-2008 and traced previous cause of specific hospital admissions. We conducted a case-crossover study to estimate the mortality risks associated with short-term exposures to PM2.5 across all Medicare deaths. We then examined effect modification by specific cause of prior admission.
Results: We found a 0.68 % increase (95% CI: 0.48- 0.88) in mortality rate for each 10 µg/m3 increase in the 2 days average of PM2.5 among all Medicare enrollees. We found significantly higher effects in subjects with diabetes with a 1.05 % increase (95% CI: 0.57- 1.53), and in subjects with multiple sclerosis (5.34 %, 95% CI: 1.03- 9.84); we also found higher effects (0.80 % increase (95% CI: 0.10- 1.51)) (P-value for interaction <0.09) in subjects with a previous admission for Alzheimer’s Disease; Parkinson’s Disease (1.62%, 95% CI: 0.56- 2.69); with diseases of the peripheral nervous system (1.49 %, 95% CI: 0.27- 2.71); and with dementia ( 0.97 % ,95% CI: 0.09- 0.87).
CONCLUSIONS: In this multi-city study we found particles increased the risk of mortality, and we identified neurological disorders, multiple sclerosis and diabetes as susceptible subpopulations.
We will complete dataset preparation and analyses and other remaining activities outlined in the grant proposal.
We will first update the dataset with Medicare data up to 2009. We will then update our previous analysis, finalizing our main results for Aim 2. We will also examine vulnerability and susceptibility (Aim 3) in relation to our results of Aim 2 by using FIPS and zipcode level area characteristics.
We will then focus on our Aim 1, by examining the long term effects of air pollution, and other activities as outlined in the grant proposal.
We do not expect any changes to the project or to the key personnel.