Climate
Change and Malaria
The effects of temperature on both the vectors and parasites of malaria
are easily seen in the latitudinal and altitudinal boundaries to malaria
transmission. However, these
boundaries seem to be changing as many highland areas have experienced malaria
epidemics in the past few years. It
has been hypothesized that increasing temperatures could be part of the reason
why malaria can now survive at higher altitudes.
Many other confounding factors, however, could be causing the increase in
malaria in these areas (Patz and Lindsay, 1999).
In addition to predictions of the effects of climate change on malaria,
studies which identify the factor or factors that are most responsible for any
changes in malaria are important in order to understand the complexities of
malaria in the actual world. There
are many variables that affect malaria transmission in addition to climatic
changes, such as environmental modification (e.g. deforestation, increases in
irrigation, swamp drainage), population growth, limited access to health care
systems, and lack of or unsuccessful malaria control measures (Patz and Lindsay,
1999). Some studies have been done
on the subject, yielding differing results as to which factor or factors are
most responsible for the increase in malaria.
Most of the studies, however, do not take into account all of the factors
that are related to malaria transmission. This
makes it difficult to assess the true determinants of malaria in each area.
Time-series studies
Mean temperature, night-time temperature, temperature in combination with
rainfall, and mean November and December temperature, were found to be related
to malaria in Zimbabwe, the Debre Zeit sector of Ethiopia, Rwanda, and the
Northwest Frontier Province in Pakistan, respectively (Freeman and Bradley,
1996; Freeman, 1995; Tulu, 1996; Loevinsohn, 1994; Bouma et
al., 1996). In Pakistan,
rainfall along with humidity in December, predicted malaria rates fairly well (Bouma
et al., 1996). In a more qualitative study, climate and forest clearing were
alleged to be related to malaria rates in the Usambara Mountains of Tanzania (Matola
et al., 1987). One study in the
highlands of Kenya claimed that climate was not a factor in malaria transmission
there because average temperature and rainfall did not change during the time
that malaria rates changed. According
to the study, deforestation might have been a reason behind changes in malaria
transmission in the highlands of Kenya (Malakooti et
al., 1998).
Another study in Kenya found that soil moisture correlated with the
human-biting rate of malaria vectors with a two-week time lag which was
explained as the length of time it takes for mosquito larvae to develop.
This same study also found that soil moisture correlates with
entomological inoculation rate (which is the product of the human-biting rate
and the proportion of female mosquitoes carrying infective parasites in their
salivary glands ready to be delivered to the next host) with a six-week time
lag. Six weeks is the amount of
time necessary for the development of the infective parasites in the mosquito
plus the length of time the mosquito survives (Patz et
al., 1998).
Studies of Epidemics
Three studies of epidemics of malaria in different areas of the highlands
of Kenya found that increased rainfall was related to the epidemic, although one
of the studies also claimed that increasing drug resistance had an effect, and
another study found that increasing temperature and relative humidity were also
involved (Kigotho 1997; Khaemba et al.,
1994; Fontaine et al., 1959).
An epidemic in Ethiopia was attributed to higher temperatures, rainfall
and relative humidity than in previous years.
The study of this epidemic also noted that although the epidemic was
associated with higher rainfall, this was not always true for that area because
in 1993 there was excess rainfall but no malaria epidemic, whereas in 1984-5,
there was high malaria incidence but very little rainfall (Woube, 1997).
One study of a malaria epidemic in the highlands of Madagascar claimed
that the epidemic was caused by anthropogenic climate change, but no statistics
were shown to defend this assertion (de Zulueta, 1994). Two other studies of highland malaria epidemics in Madagascar
related these epidemics to lack of anti-malarial medications, lack of control
techniques, and low levels of immunity in the population due to little previous
exposure to malaria (Fonteneille et al.,
1991; Mouchet et al., 1997).
Marimbu et al. (1993) found
that increasing temperatures were related to malaria in a highland area of
Burundi. Near the Manyuchi dam in
Zimbabwe, high winter temperatures were found to be related to increases in
malaria rates, but many other factors could be confounding this result (Freeman,
1994). In Irian Jaya, malaria
epidemics in the highlands correlated to population movements and increased
mosquito breeding grounds while meteorological variables remained constant
(Anthony et al., 1992; Bangs et
al., 1996).
A large group of studies relate the El Niņo Southern Oscillation (ENSO)
to malaria epidemics. Many areas
have experienced periodic malaria epidemics every five to eight years, which may
have been related to the ENSO cycle. Malaria
epidemics in the former British Punjab, Pakistan, Sri Lanka, the highlands of
Uganda, Columbia, Argentina, Ecuador, Peru, and Bolivia were proposed to be
associated with ENSO cycles (Bouma et al.,
1994; Kilian et al., 1999; Bouma and
van der Kaay, 1995; Bouma et al.,
1995).
The results of all of these studies reveal that more research needs to be
done on the causes of epidemics or increases in malaria transmission, taking
into account all of the factors that could be relevant to malaria.
It is important to add that, although many of these studies found a link
between climatic variables and malaria, it is hard to extrapolate each of those
findings to an assertion that anthropogenic climate change is the cause of the
changes in meteorological variables that were found to be related to changes in
malaria rates. These changes could
be smaller time scale changes which could just be related to interannual
variability or just changes in weather. Understanding
the relationships between climate and malaria in an area may allow prediction of
when there are likely to be malaria epidemics.
Modeling
Several mathematical and computer models have attempted to link all of
the factors that affect malaria transmission in order to predict the effect of
predicted global climate change on malaria.
One such model relates variables associated with human-induced climate
change (temperature, precipitation, relative humidity, and wind), environmental
factors (drought and desertification, sea level rise, changing vegetation, and
agricultural practices), the parasite development rate, the vector population
(death rate, breeding places, density, and insecticide resistance), and the
human population (migration, spread of drug resistance, change in immune status,
and spread of pathogens into new areas) (Martens, 1998).
Vectorial capacity is often used to determine the degree of malaria risk
posed by the specific vectors in the area (Singh et
al., 1990). Martens (1998)
created an equation that calculates the vectorial capacity multiplied by the
duration of the infectious period in humans, which is called the basic
reproduction rate (Ro) of malaria in the area.
The factors that are involved in the calculation of Ro
include: the ratio of mosquitoes to
people, the number of mosquito bites per person per day, the efficiency with
which an infected mosquito infects a human, the chance that a mosquito becomes
infected during a blood meal, the incubation period, and the daily survival
probability of the mosquito. Indirect
factors that affect the ones that are listed above include:
the availability of breeding sites which is related to precipitation,
human population density, human population migration; the feeding habits of the
mosquitoes; the presence of other animals on which the mosquitoes feed; human
exposure which can be affected by the use of bednets or other interventions;
temperature; the immunological and nutritional status of the population; the
effectiveness of medical treatment; natural enemies of the mosquitoes; and
control efforts. This model is
further complicated by algorithms that predict changing genetic adaptations in
the parasite and vector that lead to resistance.
This model predicts that the number of people in developing countries
that are likely to be at risk of malaria infection will increase by 5-15%
because of climate change, depending on which the Global Circulation Model (GCM)
and climate change scenario is used. The
areas that are expected to have the most increase in malaria transmission are
ones that are at the fringes of transmission now.
These areas, because they have low levels of immunity, will likely
experience epidemics unless they are able to use control efforts to effectively
reduce the impact of the epidemics (Martens, 1995).
This model, however, still needs to be better validated (Kovats, 2000,
personal communication).
