This report is an update and extension of our original October 2011 research and report about air quality and hospital admissions for asthma in Los Angeles. Our data now cover the period 2010-2011.
Summary
Soot and smog were not correlated with emergency admissions for asthma at a large Los Angeles hospital during 2010-2011.
Introduction
The U.S. Environmental Protection Agency asserts that ambient ozone (smog or O3) and fine particulate matter (soot or PM2.5) exposures increase hospital emergency admissions for asthma.
The EPA claims, for example, that reducing the 8-hour ozone standard from the current 75 parts per billion (ppb) to 65 ppb would reduce the number of asthma exacerbations by 38,000 and related hospitalizations by 11,000 annually.
The EPA claims that its Cross-State Air Pollution Rule, which is intended to reduce O3 and PM2.5 levels in 30 states will prevent 400,000 cases of aggravated asthma and 19,000 hospital and emergency department visits.
So we undertook to compare ozone and PM2.5 levels with hospital admissions for asthma in Los Angeles, which has some of the “worst” air quality in the U.S.
Methods
Through the Freedom of Information Act, we obtained the daily tally of hospital admissions for asthma from the Veterans Administration West Los Angeles Medical Center for the period January 1 2010 to December 31, 2011.
We correlated these tallies with daily O3 and PM2.5 maximums as collected by the California Air Resources Board for the South Coast Air Basin, the air quality management district that includes the Los Angeles area.
We correlated the air quality and hospital admissions data on a same day, 1-day lag, 2-day lag and 3-day lag bases.
Results
The Pearson correlations and 95% confidence intervals between o3 and PM2.5 levels in the Los Angeles’ area and hospital admissions are presented in the table below. A Pearson’s correlation of 1.0 indicates perfect correlation, a −1.0 indicates perfect inverse correlation and 0 indicates no correlation. All the correlations presented below are essentially zero.
| Pollutant | Pearson’s correlation | 95% Confidence Interval |
| O3 (No lag) | -0.061 | (-0.132, 0.011) |
| O3 (1-day lag) | -0.033 | (-0.105, 0.039) |
| O3 (2-day lag) | -0.058 | (-0.13, 0.014) |
| O3 (3-day lag) | -0.084 | (-0.155, −0.012) |
| PM2.5 (No lag) | 0.011 | (-0.061, 0.083) |
| PM2.5 (1-day lag) | 0.090 | (0.018, 0.161) |
| PM2.5 (2-day lag) | -0.025 | (-0.097, 0.047) |
| PM2.5 (3-day lag) | 0.014 | (-0.059, 0.086) |
Discussion
These results indicate that maximum ozone and fine particulate matter levels in the Los Angeles area were not correlated with admissions for asthma at the VA West Los Angeles Medical Center for the period January 1, 2010 to December 31, 2011.
The Los Angeles metropolitan area has some of the worst air quality in America.
If ambient O3 and PM2.5 were associated with hospital admissions for asthma, one could reasonably expect to find a correlation in these data. But we did not.
You’re making the argument against EPA’s position from the other side. 19,000 hospitalizations sounds like a lot until you realize that it is in a 30 state area covering >100,000,000 people in a year period. That’s 52/day in a population of about 10,000,000 asthmatics (about 10% of general population is estimated to suffer from asthma). It’s pure conjecture on EPA’s part that these numbers are statistically relevant.
But the California Air Research Board is still printing articles for newspapers claiming the correlation.
And the Sacramento Bee will not print anything to the contrary, no matter how many letters are sent to them.
Surely the EPA would not lie to us in order to justify more regulation, funding and/or fines!
It is not just a case of ‘not finding a correlation between A and B’.
Specifically your statistics indicate that (at the 95% confidence level) the null hypothesis ‘there is no correlation’ cannot be rejected, AND that the assertion ‘there is a correlation’ cannot be accepted.
This is also a case of (positively) finding that there is NO correlation between A and B.
Given that the 95% CL is in use, it is not unexpected that one of the eight measures (PM2.5 (1-day lag)) shows a Pearson’s coefficient that exhibits an apparently significant range that does not include the value 0. The 95% CL means that *on the average* one of 20 measures may show a false correlation. Use of a 99% CL (+/- 3 sigma, instead of 2 sigma) would change this lower limit from 0.018 to -0.018, meaning that the null hypothesis cannot be rejected at the 99% CL.
How about a study on the effects of asthma-related hospital visits and deaths after the US government banned the only OTC asthma medication available in the country? Or are we too busy gloating over another victory for mommy earth to notice all the living beings slowly suffocating?
Green is just another word for death as far as I’m concerned anymore.
Torture data long enough and they’ll confess to anything.
A massive study was done on Canadian data on the asthma/hospitalization question by Koop, McKitrick and Tole, Environmental Modeling and Software. They came to the same conclusion: “…we find the here-observed health effects of air pollution are very small and insignificant …” And “variations in smoking and income likely have a significant effect on lung-related health, but variations in pollution likely do not.” The data set was large. The statistics were done very well.
We need a FOIA request for the calculation and/or model used to correlate O3 and PM2.5 with asthma.
This study is evidence that air pollution even at the most polluted end of the ambient air quality, does not have a discernible impact on a recognized measure. So is it possible that data torturing is the method used by EPA researchers?