47 / 80
Seasonal Rhinitis and Environmental Factors in Madrid
J Investig Allergol Clin Immunol 2019; Vol. 29(5): 371-377
© 2019 Esmon Publicidad
doi: 10.18176/jiaci.0368
Figure 2.
CatPCA 1996.
pollen counts are much lower, although the sign continues to
be positive (R=0.28).
Also noteworthy is the weak association between the
pollution variables ozone and PM10, which, therefore, occupy
distant, almost perpendicular positions on the graph. In addtion,
this relationship has a negative sign (high levels of PM10
reduce ozone levels).
The 2009 CatPCA analysis (Figure 3) explains 70.5% of
the variance. The relationship between symptoms and pollen
becomes stronger (R=0.81) and remains positive, as does
that of ozone, albeit to a lesser extent (R=0.35), with both
coefficients higher than in 1996.
The difference between the analyses for 2009 and 1996 is
in the association between the allergy variables and the rest of
the variables (as can be clearly seen in the positions occupied
by the variables on the graph). In contrast to the 1996 period,
the 2009 variables (symptoms and grass pollen counts) are
associated mainly with ozone (R=0.35, R=0.26). Therefore,
In the 1996 CatPCA analysis, symptoms were related
mainly to grass pollen (R=0.55) and, to a lesser extent, to
temperature (R=0.38) and ozone (R=0.28); all relationships
had a positive sign (Figure 2). With PM10, the correlation
coefficient was much lower (R=0.18). Temperature is related
to pollen counts, symptoms, PM10, and ozone (the correlation
is particularly strong for ozone, R=0.63). Given the position
and proximity of the lines on the graph, the variables with
the closest association are symptoms and grass pollen counts
(R=0.55) and temperature and ozone (R=0.63). In both cases,
the relationships are positive, meaning that high values in one
parameter correspond to high values in the other, ie, higher
pollen levels, higher symptoms, higher temperatures, and
higher ozone levels. Temperature is also clearly related to
symptoms and PM10 (R=0.34 and R=0.33); the sign for this
relationship is also positive. Consequently, high temperature
levels are related to high values for symptoms and PM10.
Values for the association between temperature and grass
374
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
Dimension 2
Dimension 1
Variable Principal Normalization
0.0
0.4
0.8
0.2
0.6
1.0
Grass pollen counts
Symptoms
Temperature
PM10
O
3
Component Loadings
Figure 3.
CatPCA 2009.
0.75
0.50
0.25
0.00
-0.25
-0.50
Dimension 2
Dimension 1
Variable Principal Normalization
0.0
0.4
0.8
0.2
0.6
1.0
Grass pollen counts
Symptoms
Temperature
PM10 O
3
Component Loadings
CatPCA analysis explains 70.5% of the variance
Correlations Transformed Variables
Symptoms Grass
PM10 Ozone Tempe-
pollen counts
rature
Symptoms 1.000
0.807 0.118 0.353 0.048
Grass pollen
counts
0.807
1.000 0.186 0.259 0.100
PM10
0.118
0.186 1.000 0.056 0.359
Ozone
0.353
0.259 0.056 1.000 0.527
Temperature 0.048
0.100 0.359 0.527 1.000
CatPCA analysis explains 66.4% of the variance
Correlations Transformed Variables
Symptoms Grass
PM10 Ozone Tempe-
pollen counts
rature
Symptoms 1.000
0.547 0.179 0.278 0.337
Grass pollen
counts
0.547
1.000 0.146 0.222 0.281
PM10
0.179
0.146 1.000 -0.097 0.334
Ozone
0.278
0.222 -0.097 1.000 0.626
Temperature 0.337
0.281 0.334 0.626 1.000