1. Analysis of three years correlations between weather variability and
seasonal asthma episodes in Miami Dade, Florida
David Quesada
School of Science, Technology and Engineering Management,
St. Thomas University, Miami Gardens FL 33054
Climatic and environmental changes occurring since the middle of the Twentieth Century as
well as th aggravating pollution l
ll the ti ll ti levels i megacities are exacerbating asthma episodes and
l in iti b ti th i d d
the number of hospitalizations due to this disease. Since 1999, in Miami Dade County the
hospitalization rates were doubling the Healthy People 2010 objectives in every age group. A
comprehensive weather database including outdoor temperature (T), humidity (H),
barometric pressure (P), wind direction (θw) and speed (vw) as well as the values of
p ( ) ( p (
maximum and minimum and the range of all these variables has been created. As a result, a
seasonal pattern emerged, with a maximum appearing around the middle of December and a
minimum around the middle of March every year for the three years of analysis.
Second Symposium on Environment and Health, AMS 91st Annual Meeting, 23 – 27 January 2011 in Seattle, WA.

2. Content
• Why Asthma? Motivation of the study.
• Previous results within continental USA and Miami Dade.
• WeatherBug Mesonet and Asthma – Weather connection.
• Mi i l Bio-Physical model.
Minimal Bi Ph i l d l
• Conclusions.
Second Symposium on Environment and Health, AMS 91st Annual Meeting, 23 – 27 January 2011 in Seattle, WA.

3. Why to study Asthma? How far Bio-Meteorology may help with?
Asthma Statistics Worldwide
Number of people diagnosed: more than 150 M
Europe: the # of cases has doubled
USA: the number of cases has increased more
than 60%
India: between 15 and 20 M
Africa: between 11 and 18% population
Number of deaths yearly: around 180,000
Miami Dade County , Florida
7.1% Middle and HS children were reported with
asthma
The number of hospitalizations due to asthma
has doubled.
The number 1 cause of school absences and 35 %
of parents missed work
Second Symposium on Environment and Health, AMS 91st Annual Meeting, 23 – 27 January 2011 in Seattle, WA.

4. Seasonal Variations in Asthma Hospital Admissions in the United
States
Asthma admission by year
16
14
12
0,000
2000
1999
zed rate per 10
1998
10
1997
1996
1995
8
1994
1993
1992
Annualiz
6
1991
1989
1988
4
2
0
1 2 3 4 5 6 7 8 9 10 11 12
Admission month
Source:Nationwide Inpatient Sample and US Census
Aichatou Hassane UNH; Robert Woodward,
Hassane, Woodward • Asthma seasonal variations confirmed
PhD, UNH; Ross Gittell, PhD, UNH - May 27, • Larger seasonal variation associated
2004 with a decrease in age.
Second Symposium on Environment and Health, AMS 91st Annual Meeting, 23 – 27 January 2011 in Seattle, WA.

5. Seasonal Variations in Asthma Hospital Admissions in the United
States
2000 Asthma Admission by US region
S
16
14
12
Annualized rate per 10,000
10
Northeast
e
8
Midwest
South
6 West
4
2
0
1 2 3 4 5 6 7 8 9 10 11 12
Admission month
Source:Nationw ide Inpatient Sample and US Census
Regional seasonal variation exists:
• Midwest has the largest rate of Asthma - East North
Central division: Illinois and Wisconsin
• West region has the lowest rate of Asthma - Mountain
division: Arizona and Colorado

6. Miami Dade Asthma Snapshot
180
175
170
Ra per 100,000 persons
165
160
155
150
145
ate
140
135
130
2001 2002 2003 2004 2005 2006 2007 2008
Areas of major incidence
Second Symposium on Environment and Health, AMS 91st Annual Meeting, 23 – 27 January 2011 in Seattle, WA.

7. Create a database of weather parameters and environmental
triggers for asthma ( WeatherBug & WeatherBug Achieve)
Feature Range Accuracy Range Accuracy
(English)
(E li h) (English)
(E li h) (Metric)
(M t i ) (Metric)
(M t i )
Temperature -55F – 150F +/- 1F -45C – 60C +/- 0.5C
Relative Humidity 0 – 100% +/- 2% 0 – 100% +/- 2%
Wind Speed 0 – 125 mph +/- 2 mph 0 – 275 kph +/- 4 kph
Wind Direction 0 – 360 deg +/- 3 deg 0 – 360 deg +/- 3 deg
Barometric Pressure 28 – 32” Hg +/- 0.05”Hg 900 – 1100 mbar +/- 5 mbar
Rainfall Unlimited +/- 2% Unlimited +/- 2%
Light Intensity 0 – 100% N/A 0 – 100% N/A

8. Zip codes patients came from
WeatherBug Mesonet stations
NWS stations, MIA & Tamiami
Year White White Non White African
Hispanic Hispanic American
2008 490 505 820 510
2009 350 256 650 525
2010 528 495 605 657
Year
Y Total
T t l Total
T t l Total
T t l % of
f
Patients Respiratory Asthma asthma
2008 5172 2950 2222 43
2009 6981 4301 2680 38
2010 7813 4960 2853 37

9. Number of asthma ca
ases
100
150
200
250
300
350
400
450
15-Jan 500
15-Feb
15-Mar
15-Apr
15-May
15-Jun
15-Jul
15-Aug
15-Sep
15-Oct
15-Nov
15-Dec
15-Jan
15-Feb
15-Mar
15-Apr
15-May
15-Jun
15-Jul
15-Aug
15-Sep
15-Oct
15-Nov
15-Dec
15-Jan
15-Feb
15-Mar
15-Apr
15-May
15-Jun
Kendall Medical Group in Miami Dade, FL
15-Jul
15-Aug
15-Sep
15-Oct
Seasonal Variations of Asthma diagnosed cases by the
15-Nov
15-Dec

10. Seasonal Variations of Asthma diagnosed cases
in standard units Z = (N – Nave)/S
by the Kendall Medical Group in Miami Dade, FL
1.5
1
ve/St.Dev)
0.5
Number of cases in z - units (N - Nav
0
-0.5
-1
-1.5
-2

11. 100 90
90
Tmax
80
80
70
70
60
60
50 50
40
Tmin 40
30 Tmean=(Tmax+Tmin)/2
( )
30
20 1/1/2008 1/1/2009 1/1/2010
500 15
1/1/2008 1/1/2009 1/1/2010
450 dTmean/dt = T[i+1] - T[i]
10
Number of asthma cases
400
5
350
300 0
a
250
-5
200
-10
150
100 -15
15…
28…
15…
31…
15…
31…
15…
…
15…
28…
15…
31…
15…
31…
15…
…
15…
28…
15…
31…
15…
31…
15…
30…

12. 30 0.6
ΔT=Tmax-Tmin ΔT/Tmean
25 0.5
20 0.4
15 0.3
10 0.2
5 0.1
0 0
1/1/2008 1/1/2009 1/1/2010 1/1/2008 1/1/2009 1/1/2010
Second Symposium on Environment and Health, AMS 91st Annual Meeting, 23 – 27 January 2011 in Seattle, WA.

13. 30.6 30.6
30.4 Pmax Pmean
30.4
30.2
30 2
30.2
30
29.8 30
29.6 29.8
29 8
29.4 Pmin
29.6
29.2
29.4
29
1/1/2008 1/1/2009 1/1/2010
1/1/2008 1/1/2009 1/1/2010
500 0.5
450 0.4
dPmean/dt
Number of asthma cases
400 0.3
0.2
350
0.1
300
a
0
250
-0.1
200
-0.2
150 -0.3
100 -0.4
15…
28…
15…
31…
15…
31…
15…
…
15…
28…
15…
31…
15…
31…
15…
…
15…
28…
15…
31…
15…
31…
15…
30…

14. Hmax 100
100
90
80 80
70
60
60
40
50
Hmin Hmean
20 40
30
0
1/1/2008 1/1/2009 1/1/2010
1/1/2008 1/1/2009 1/1/2010
500 40
450 dHmean/dt = H[i+1] - H[i]
30
Number of asthma cases
400
20
350
10
300
a
0
250
-10
200
150 20
-20
100 -30
15…
28…
15…
31…
15…
31…
15…
…
15…
28…
15…
31…
15…
31…
15…
…
15…
28…
15…
31…
15…
31…
15…
30…

15. Pearson Correlation between the number of cases and the given
set of variables (Excel)
t f i bl (E l)
Tmax Tmin ΔT Tmean dT/dt ΔT/Tmean
# cases - 0.52 - 0.59 - 0.55 0.99 - 0.16 - 0.86
ΔP Pmean dP/dt ΔP/Pmean
# of cases - 0 11
0.11 0.28
0 28 - 0 002
0.002 0.1
01
ΔH Hmean dH/dt ΔH/Hmean
# of cases 0.08 - 0.25 - 0.1 - 0.76
Second Symposium on Environment and Health, AMS 91st Annual Meeting, 23 – 27 January 2011 in Seattle, WA.

16. Correlations between the number of cases and the given set of variables
(IBM-SPSS-19)
Tmax Tmin ΔT Tmean dT/dt ΔT/Tmean
Pearson (r) - 0.524 - 0.529 0.357 - 0.531 - 0.122 0.487
P - value 0.000 0.000 0.002 0.000 0.306 0.000
Kendall - τ - 0.325 - 0.301 0.159 - 0.311 - 0.122 0.264
P - value 0.000 0.000 0.048 0.000 0.132 0.002
Spearman - ρ - 0.485 - 0.463 0.224 - 0.475 - 0.148 0.375
P - value 0.000 0.000 0.059 0.000 0.215 0.001
ΔP Pmean dP/dt ΔP/Pmean ΔH Hmean dH/dt ΔH/Hmean
Pearson (r) 0.367 - 0.021 0.082 0.42 0.452 - 0.213 - 0.015 0.445
P - value 0.002 0.862 0.491 0.000 0.000 0.073 0.899 0.000
Kendall - τ 0.269
0 269 0.008
0 008 0.045
0 045 0.291
0 291 0.282
0 282 - 0 052
0.052 0.006
0 006 0.264
0 264
P - value 0.001 0.922 0.579 0.000 0.000 0.521 0.938 0.001
Spearman - ρ 0.388 0.001 0.063 0.415 0.402 -0.091 0.003 0.373
P - value 0.001 0.996 0.600 0.000 0.000 0.445 0.979 0.001
Second Symposium on Environment and Health, AMS 91st Annual Meeting, 23 – 27 January 2011 in Seattle, WA.

17. N = Constant + a (Tmax) + b (Tmin) + c (Tmean) + d (ΔT/Tmean) + e (ΔP) + f (ΔH) + g (ΔH/Hmean)
Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
1 .695a .483 .427 62.65654
ANOVAb
Model Sum of Squares df Mean Square F Sig.
1 Regression234902.995 7 33557.571 8.548 .000a
Residual 251253.880 64 3925.842
Total 486156.875 71
Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients
B Std.
Std Error Beta t Sig.
Sig
1 (Constant) 236.329 292.762 .807 .423
VAR00003 -69.515 20.571 -5.727 -3.379 .001
VAR00004 53.801 19.021 5.375 2.829 .006
VAR00006 15.977
15 977 16.645
16 645 1.436
1 436 .960
960 .341
341
VAR00008 3026.508 1076.097 1.902 2.812 .007
VAR00009 -431.218 480.090 -.114 -.898 .372
VAR00013 14.140 3.409 1.016 4.148 .000
VAR00016 -326 596
-326.596 130.111
130 111 -.571
- 571 -2.510
-2 510 .015
015
a. Dependent Variable: VAR00001

19. Conclusions
• African Americans and Non White Hispanics are more affected by asthma
asthma.
• Zip codes from Miami Dade with the major incidence seem to be related with
socio-economic background rather than particular microclimatic conditions.
• Among weather variables, Tmean, ΔT/Tmean, Tmin, and ΔH/Hmean appear to
correlate better with the number of asthma cases.
• The observed patterns seem to be originated in the thermoregulation response
to cold weather, rather than in allergic pathways.
• More statistical work is needed in order to establish an Asthma Index for
Bio-Meteorological applications.
applications
Acknowledgments
• Oscar Hernandez M.D. and Elizabeth Fontora, Medical Group, Miami Dade, FL
• School of Science, St. Thomas University
Second Symposium on Environment and Health, AMS 91st Annual Meeting, 23 – 27 January 2011 in Seattle, WA.