Kelly Duncan
Introduction To Geographical Information Systems
June 2009
Physical And Demographic Aspects of Dengue Fever
Occurrences and Spread in Brazil
Population Spreads
January
These maps show the spread of Dengue in Rio De
Janeiro’s Metropolitan region, through the months
with the highest number of Dengue cases (January through
April, Summer).
In January (the beginning of
summer) there are a number of
cases, especially in the areas
with a larger population, and
very few in the areas with
smaller populations.
As the season progresses, the
larger population areas show
faster growth in the
Number of Cases Population
number of cases
1 - 10
12,597 - 200,000
200,001 - 400,000 (the larger the
11 - 100
400,001 - 600,000 number of cases
101 - 1000
600,001 - 800,000 the larger the
800,001 - 6,093,472 dots), while the ar1001 - 5235
eas with less population show slower
or no growth in the number of cases.
February
March
April
0
20
40
Miles
September
0
500
Summer: January, February,
March
Winter:July, August, September
0
Fall: April, May, June
Spring:October, November,
December
1,000
Miles
2,000
Rivers
These map shows the
Major Roads
correlation between proxLakes
imity to rivers, roads, and
% Cases/Population lakes to the spread and
0.0001% - 0.01% number of cases.
0.0101% - 0.02% The months of September
0.0201% - 0.03% and October were used
0.0301% - 0.04% because September
0.0401% - 0.1297% shows the least number
No Cases
of cases in a year, hence
it is easiest to see the spread of Dengue Fever between these two months.
Looking at these two cases, it is apparent,
that the spread and occurrences are correlated to the rivers/lakes (dark to light green)
and many areas on major roads (east: green/
yellow to orange; west: light green to red).
1,000
Miles
Biomes and Climates
Biome
Number of Cases
One of the biggest factors in the spread of Dengue is the season. In
many parts of Brazil, seasons directly affect the
climate. In the summer it
is hotter and more humid,
then in winter, when it is
cooler and drier.
Because the vector mosquito is very
Number of Cases susceptible to
1 - 10
the heat and hu11 - 100
midity, during
101 - 1000
winter months
1001 - 10000
there would be
10001 - 41752
far less mosquiNo Cases
toes (in areas
where there are four seasons, such as the more
south regions) then in the
summer months. Finally,
due to the decrease in
the amount of mosquitoes there are less cases
in the winter then the
summer.
Cases Versus Population: Full Year
60,000
50,000
Amazon Rainforest
Caatinga/Dry Forest
Cerrado/Savanna
Pantanal/Water Marsh
1 - 10
11 - 100
101 - 1,000
1,001 - 10,000
10,001 - 41,752
No Cases
Biome
Tropical Decid Forst
Pampa/Grassland
Dry Forest
Savanna
Water Marsh
Grassland
Rainforest
Tropical Deciduous Forest
Number of Cases
Climate
Tropical
Equatorial
1 - 10
11 - 100
101 - 1,000
1,001 - 10,000
10,001 - 41,752
No Cases
Tropical
Semi-Arid
Climate Zone
Equatorial
Subtropical
Highland Tropical
Semi-Arid
Tropical
Highland Tropical
Subtropical
0
1,000
Miles
2,000
40,000
30,000
20,000
These maps show the correlation between the number of cases and the different biomes and climate
zones. The overlay is of the summer number of
cases.
The biome map shows there is a high correlation
between the number of cases and the areas. The
biomes Savanna, Dry Forest, and the Water Marsh
area have a higher number of occurrences (where
there is more red then yellow). While the biomes of
Tropical Deciduous Forest and Grassland have a
lower number of occurrences. The biomes with a
higher amount of occurrences are most likely have
habitats more fit to the mosquitoes (ex. more water
for breeding).
The climate map shows high correlation between all
the biomes. In the Tropical and Highland Tropical
areas, where seasons have less variation in temperature and humidity, there are a large number of
cases. In the Subtropical area, where the seasons
are more extreme, hence periods of time when the
vector mosquitoes can not survive there (hence less
mosquitoes overall), there are less cases then in
the areas where there are more moderate seasons.
0
10,000
% (Cases/Population) by Population
y = 0.0023x
10,000
October
Maps
Seasonal Spread
Seasonal: Due to seasonal changes in the temperature and humidity, there is a high
correlation between the seasons and the number of occurrences and spread of Dengue
fever. In warmer and/or more humid seasons, over all of Brazil, there are a higher number occurrences and greater spread of Dengue Fever then in colder and drier months.
Population: The correlation of the number of cases versus population cannot be completely determined by this study. In larger cities there are more cases of Dengue Fever,
but looking at the rates of cases by population, there is no clear correlation between
population and how many people are affected (see below).
Roads, Rivers, and Lakes
Methods
Using information on the number of cases in 2007 of Dengue Fever by region, from Brazil’s Ministry of Health, along with information on population, biomes and climates from
the Brazilian Institute of Geography & Statistics, to view how different types of physical
and demographical (population) features cause the occurrences and spread of Dengue.
Dengue occurrences and spreads are incredibly complex, to help simplify the correlation
factors, a breakdown of the some of the physical and demographic factors was used. After using ArcGIS to visually display and edit the Shapefile data, Microsoft Excel was
usedto create graphs to show the correlation between the different factors (seasonal,
population, proximity to roads, lakes, and river, and biome & climate areas) and the number of cases in different regions.
After examining the maps and graphs to see the correlation between the spread and
number of occurrences of Dengue and different physical and demographic properties of
Brazil, this correlation is shown below:
Seasonal Population Roads, Rivers
Biome
Climate
and Lakes
Correlation
High
High
Medium
Medium
High
% (Cases/Population)
Since the re-introduction of the Aedes aegypti mosquito to Brazil and the first case
in the 1980s, (Mondini et al., 2005), Dengue Fever has become a major public health
issue in Brazil.
Dengue is transmitted by the bite of the vector mosquito, and has a four day incubation period, before an infected person will develop a fever and rash, which he or she
will recover from in approximately five days. Dengue Fever, itself, does not have a
high mortality rate, but Dengue Hemorrhagic Fever (5% morality rate), with an extra
symptom of vascular permeability, blood loss, and Dengue Shock Syndrome (up to
40% mortality rate), where there is the loss of too much blood, resulting in hypertension, are where the majority of the Dengue deaths come from.
Due to the fact that the Aedes aegypti mosquito prefers areas with temperatures
over 68° F (20° C, Nakhapakorn, “An Information value…” 2005), areas in Tropical and
Subtropical climate regions have the highest volume of the vector mosquito. Brazil’s
location relative to the equator, makes it a “hot spot” for the vector mosquito and in effect Dengue.
The health concerns of Dengue, has caused health officials to track the disease as
well as the development of models to help track the number of occurrences and
spreads, in order to understand the disease’s patterns and facilitate the optimization of
resources in the eradication of the disease. Looking at different factors of the mosquito and human components of the disease, should be helpful in finding areas of risk
for Dengue.
Certain factors that are important to these components are, seasonality, population, transportation methods, and the different biomes and climate zones. Seasonality
is important because in many areas of Brazil the different seasons bring hotter and
colder temperatures to an region where mosquitoes live, if a temperature falls below
the Aedes aegypti desired temperature they will not be found in that region. The population of a region is important feature because where population density is higher there
are more potential targets in an infected mosquito’s life span, causing greater spread
of Dengue. Transportation methods are important for both humans and mosquitoes, for
humans, the movement of the disease from an infected region to an uninfected region,
as well as the density of mosquitoes in an area. Finally the different biome and climate
regions are important to the mosquitoes habitation and seasonal patterns.
Results
Number of Cases
Introduction
100,000
1,000,000
10,000,000
6.00%
5.00%
4.00%
3.00%
2.00%
1.00%
0.00%
10,000
100,000
100,000,000
1,000,000 10,000,00 100,000,0
0
00
Population
Population
In larger cities, Dengue spreads more rapidly then in the smaller population regions (see
above). But this is more likely due to population density then strictly the number of people. The reason the population density has much more to do with the amount of spread
is due to the rates of mosquito travel and lifespan then the total amount of people does.
Roads, Rivers, and Lakes: There is a correlation between the spread of Dengue and
the roads, rivers, and lakes. This is due to two facts, one is that roads and rivers are
used for human transportation, hence the transportation of the disease in human form.
The second being that there are more mosquitoes around rivers and lakes (mosquitoes
use water as breeding grounds), hence the transportation by mosquitoes and mosquito
density of an area is greater.
Biomes and Climates: In the biomes which are more “tree covered” then other biomes
(hence wetter), there are greater number of cases and a higher spread of Dengue then
ones which are drier.
The climate has a greater effect on the occurrences and spread (as talked about above
in the seasonal section). Climate regions with more moderate seasons have higher occurrences and spread then climate regions with more extreme seasons, due to the temperature preferences of the vector mosquito.
Overall:By putting together the different factors for the occurrences and spread of Dengue Fever, a model can be made for the number of occurrences and spread in any general area in Brazil. This model could be used in order to target and treat high risk areas
better, so there is less spread and total number of occurrences in Brazil. An un-weighted
example of a model is shown below (the Actual data is Summer 2007):
Predicted
Actual
Level of Risk
0
Very High
High
Medium
Low
Very Low
500
1,000
Actual Number of Cases
1 - 10
11 - 100
101 - 1,000
1,001 - 10,000
10,001 - 41,752
No Cases
Miles
Conclusions
The correlations between the factors and Dengue occurrence and spread, show high risk
areas for Dengue. Some of these areas though have low number of cases, this could be
for many reasons, such as immunity, isolation of groups of people, or better use of mosquito nets. By looking at these areas, public health officials could bring techniques used
these areas into areas of high risk with higher number of cases of Dengue Fever.
A more comprehensive study could also be done using smaller time periods and more
accurate, if not exact, data on the location of cases, to track the spread more accurately,
as well as getting more quantitatively accurate results.
Acknowledgments: Statistics of the Number of Dengue Cases and Map of Brazil: Anthony Stevens, Universidade Federal do Rio Grande do Sul/Ministry of Health of Brazil
Sources:
Biome & Climate Maps and Population Statistics: Brazilian Institute of Geography and Statistics, www.ibge.gov.br & mapas.ibge.gov.br; Rivers, Roads, and Lakes: GfK Geomarketing
Mondini et al., “Spatial Analysis of dengue transmission in a medium-sized city in Brazil.” 2005
Nakhapakorn et al. “An information value based analysis of physical and climatic factors affecting dengue fever and dengue hemorrhagic fever incidence.” 2005
Projection (All):
Polyconic (world)
Units (All):
Decimal Degrees