Using remote sensing and modeling techniques to investigate the annual parasite incidence of malaria in Loreto, Peru
Introduction
Malaria continues to be one of the most severe public health problems worldwide. According to the World Health Organization (WHO, 2015), 1.2 billion people are at a high risk of being infected with malaria and developing the disease and 214 million cases were reported in 2015. The majority of the cases occurred in Africa and South-East Asia, but transmission continues in several parts of South America as well (WHO, 2015). In 2012, approximately 25% of the malaria burden in South America was experienced by 12 municipalities in Peru, Brazil, and Venezuela (Zaitchik et al., 2012).
Peru is making progress towards controlling malaria but has not been able to completely eliminate the disease, thus making it the country with the second highest number of malaria cases in South America (Bautista et al., 2006, World Health Organization 2015). In 2015, Peru had an estimated population of 30,973,148, of which 12,165,089 had at least some risk of contracting malaria (WHO, 2015). During the 1990s, there was a 7-fold increase in malaria incidence in Peru, rising from 13 per 10,000 inhabitants in 1990 to a peak of 88 per 10,000 in 1996 (Roper et al., 2000). Specifically, over 60% of all malaria cases occurred in the Loreto Department of Peru (Zaitchik et al., 2012). As a result, the Loreto has been the major focus of the malaria control. In 1990, there were only 641 cases in Loreto, but the number rose to 121,268 cases by 1997 (Roper et al., 2000). Peru saw an overall decline in malaria cases from 2001–2010, but the number of cases has increased since then, especially in Loreto. In 2015, there were nearly 3 times as many malaria cases as were reported in 2011.
Peru has implemented a number of initiatives in an effort to control malaria. The Peruvian Malaria Program provides free antimalarial drugs under a Directly Observed Therapy (DOT) protocol (Chuquiyauri et al., 2012). In addition, regional efforts to improve malaria surveillance, early detection, prompt treatment, and vector management have been employed since 2000 (Herrera et al., 2012). From 2006 to 2011, Peru participated in the PAMAFRO project, a malaria control program in which long lasting insecticide-impregnated nets (LLIN) were delivered to remote communities in Loreto. Based on the most recent project report (PAMAFRO, 2010), most of the LLINs were distributed during the first years of the project (prior to 2009). Despite these efforts and increased funding for malaria control in the region, there are still gaps in understanding how different factors impact malaria transmission and elimination (Herrera et al., 2012). This brings into question the role of climate and environmental factors.
In Peru, the two main Anopheles species responsible for malaria transmission are the An. darlingi (along the Amazon basin) and An. pseudopunctipennis (along the Peruvian north coast) (Sinka et al., 2012) The seasonality patterns of the mosquito are closely related to the rainfall cycle, mainly due to rainfall increasing the availability of breeding sites leading to peak abundances of An. darlingi reported in the rainy season (Reinbold-Wasson et al., 2012). The larvae also require stable conditions in the breeding sites and prefer large water bodies such as rivers. Additionally, the mosquitoes prefer certain amount of vegetation coverage and temperatures ranging from 20 to 28°C (Hiwat and Bretas, 2011). The changing temperature trends can impact the time needed for parasite development, mosquito abundance, gonotrophic cycle, and larval development (Patz and Olson, 2006). Past studies clearly indicate that global climate variability already has and will continue to have an impact on malaria transmission. Specifically, climatic variations and extreme weather events have been shown to have a profound impact on infectious agents and their associated vector organisms (Parham and Michael, 2010, Patz et al., 2005). These two studies also showed that vectors such as mosquitoes are devoid of thermostatic mechanisms, so their reproduction and survival rates are strongly impacted by fluctuations in temperature. Parham and Michaels (2010) showed that environmental variables such as temperature, humidity, rainfall, and wind speed can affect the incidence of malaria by impacting the changes in the duration of the parasite's life cycle and parasite behavior.
Githeko and Ndegwa (2001) focused on the East African Highlands and argued that the underlying cause of the malaria epidemic is due to the changing climatic conditions in this normally cool area. An increase in temperature has been shown to accelerate the rate of mosquito larval development and the frequency of bites on humans, as well as impacting the time it takes for the malaria parasite to mature into the mosquito stage. Increases in rainfall can create additional habitats for mosquitoes to breed, thus increasing vector populations. Githeko and Ndegwa (2001) concluded that in the past decade there has been an increase in the anomalies of mean monthly temperatures, which has a strong relationship with the number of malaria cases.
Few past studies have attempted to identify the relationship between climate variables and malaria risk in Peru. Jones et al., (2004) proposed that environmental factors are responsible for changes in the mosquito population over time. This study focused on a region in Loreto, where a higher overall mosquito population was observed from October 1996 through March 1997, which corresponds to the rainy season (Jones et al., 2004). Aramburú et al. (1999) found a positive correlation between malaria transmission periods and rainfall and higher temperatures near the Amazon River. Additionally, Aramburú et al. (1999) showed that the two precipitation peaks in 1997 occurred three months and one month before the malaria cases reached their highest levels in Loreto. These studies focused on the Loreto region but failed to take into account a long time series of climate and environmental data, which is critical to observe temporal trends.
This research goes one step further, by investigating how remote sensing and modeling products can be used to analyze trends in the annual parasite incidence by expanding on these past studies to include a longer time series, a larger study area (i.e., the whole Loreto Department), by examining a more complete set of environmental variables, and by quantifying the relationship between malaria and climate/environmental conditions. Field observations in the region are limited, as the Loreto department comprises nearly one-fourth of the landmass of Peru and has a low population density (Aramburú et al., 1999), making it difficult to conduct field collections of environmental data. Hence, remote sensing and modeling techniques are extremely valuable to obtain the necessary information of the current environmental and climate conditions of the region and to investigate the impacts of those factors on malaria transmission. In this study, we focus on analyzing the association between the annual parasite incidence at 315 health centers located in Loreto and environmental and atmospheric variables, such as temperature, humidity, soil moisture, vegetation coverage, and elevation. All of the variables are entered into a Multivariate Poisson Regression Model to study the dependence of the annual parasite incidence on these environmental conditions and identify which regions of the department are suitable for malaria transmission. Results from this study can be applied to surveillance efforts and to direct elimination strategies in higher risk regions.
Section snippets
Study area
Loreto is one of the 25 departments in Peru, located in the Northeast region of the country (Fig. 1a). Loreto comprises one fourth of Peru's land area and has a total area of approximately 348,177 km2 (Griffing et al., 2013, Vittor et al., 2006). The region lies in the Amazon rainforest basin and has ecological characteristics of the Amazon lowlands (Aramburú et al., 1999). The region is characterized by two distinct wet and dry seasons, with the wet season going from November to May, although
Mixed effects poisson regression model
A multilevel mixed-effects Poisson regression model is used to study the relationship between Annual Parasite Incidence and atmospheric/environmental variables. This type of model is a multilevel Poisson regression that contains both fixed effects and random effects. There are many advantages of using a mixed-effect model for this type of analysis. First, mixed-effect models can be applied to continuous and non-normally distributed outcomes (e.g., Poisson distribution). Second, this family of
Sample characteristics
The total number of malaria cases for Loreto decreased from 2009 to 2010, then began increasing afterward, reaching 43,737 cases in 2013. The yearly number of malaria cases, population, and annual parasite incidence for the entire department are presented in Table 1 for 2009–2013.
Table 2 shows the mean annual parasite incidence for the health centers in each of the eight redes, as well as the number of health centers in each. In 2013, Maynas Ciudad, Datem del Maranon, and Maynas Periferia
Conclusion
In this study, we analyzed annual parasite incidence data from health centers in the Loreto Department of Peru, located in the Amazon basin, to assess and quantify the association between malaria and various environmental and atmospheric factors. This region has a low population density and a large land area, making it difficult to directly collect high-resolution environmental data. Remote sensing and modeling techniques are particularly useful in remote areas like Loreto, providing temporally
Acknowledgments
The authors would also like to thank the Loreto Ministry of Health for providing the malaria dataset, NASA Precipitation Processing System (PPS) for the TMPA 3B42 data, the Global Modeling and Assimilation Office (GMAO) and the GES DISC for the dissemination of MERRA.
References (57)
- et al.
Anemia and malaria at different altitudes in the western highlands of Kenya
Acta Trop.
(2004) - et al.
Altitude and the risk of bites from mosquitoes infected with malaria and filariasis among the Mianmin people of Papua New Guinea
Trans. R. Soc. Trop. Med. Hygiene
(1997) - et al.
Socio-demographics and the development of malaria elimination strategies in the low transmission setting
Acta Trop.
(2012) - et al.
Optical–biophysical relationships of vegetation spectra without background contamination
Remote Sens. Environ.
(2000) - et al.
Evaluation of TMPA satellite-based research and real-time rainfall estimates during six tropical-related heavy rainfall events over Louisiana, USA
Atmos. Res.
(2009) - et al.
Prospects for malaria elimination in non-Amazonian regions of Latin America
Acta Trop.
(2012) - et al.
Overview of the radiometric and biophysical performance of the MODIS vegetation indices
Remote Sens. Environ.
(2002) - et al.
A review of Peruvian flood plain forests: ecosystems, inhabitants and resource use
For. Ecol. Manage.
(2001) - et al.
Validation of TRMM multi-satellite precipitation analysis (TMPA) products in the Peruvian Andes
Atmos. Res.
(2015) Geographical field note an anthropogenic meander cutoff along the Ucayali river, Peruvian Amazon
Geog. Rev.
(2005)
Towards malaria risk prediction in Afghanistan using remote sensing
Malar. J.
Penalty, Shrinkage and Pretest Strategies
Malaria reemergence in the Peruvian Amazon region
Emerging Infect. Dis.s
Assessment of current malaria status in light of the ongoing control interventions, socio-demographic and environmental variables in jiga area, northwest ethiopia
PLoS ONE
Prevalence and risk factors of malaria in Ethiopia
Malar. J.
Epidemiology and spatial analysis of malaria in the Northern Peruvian Amazon
Am. J. Trop. Med. Hyg.
Relationship between altitude and intensity of malaria transmission in the Usambara Mountains, Tanzania
J. Med. Entomol.
Demography and management of the invasive plant species Hypericum perforatum. I. Using multi-level mixed-effects models for characterizing growth, survival and fecundity in a long-term data set
J. Applied Ecology
The drying out of soil moisture following rainfall in a numerical weather prediction model and implications for malaria prediction in West Africa
Weather Forecasting
Altitude-dependent and -independent variations in plasmodium falciparum prevalence in Northeastern Tanzania
J. Infect. Dis.
Regression analyses of counts and rates: poisson, overdispersed poisson, and negative binomial models
Psychol. Bull.
Advances in Analysis of Longitudinal Data
The history of 20th century malaria control in Peru
Malar. J.
Predicting malaria epidemics in the Kenyan highlands using climate data: a tool for decision makers
Global Change Hum. Health
Status of the TRMM 2A12 land precipitation algorithm
J. Atmos. Oceanic Technol.
Mapa de la hidrografía principal y secundaria, departamento de Loreto
Vegetation NDVI linked to temperature and precipitation in the upper catchments of yellow river
Environ. Model. Assess.
Cited by (8)
Predicting the location of larval habitats of Anopheles mosquitoes using remote sensing and soil type data
2022, International Journal of Applied Earth Observation and GeoinformationCitation Excerpt :In the first approach, researchers have studied the presence of malaria disease in study areas based on environmental factors. In these researches, malaria risk maps had been created using the Normalized Difference Vegetation Index (NDVI), classification of satellite images with medium and low spatial resolution with emphasis on two classes of water bodies and vegetation, as well as using meteorological station data (Jeganathan et al., 2001; Kleinschmidt et al., 2000; Mohan and Naumova, 2014; Mousam et al., 2017; Sewe et al., 2017; Adeola et al., 2019; Mazher et al., 2018; Abdelsattar and Hassan, 2021). Some of these studies have used a regression model using malaria frequency statistics in the study area to further investigate the impact of each of these environmental parameters (Beck et al., 1997; Kleinschmidt et al., 2000; Mousam et al., 2017; Sewe et al., 2017).
Errors and uncertainties associated with quasiglobal satellite precipitation products
2021, Precipitation Science: Measurement, Remote Sensing, Microphysics and ModelingHydrology, water resources and the epidemiology of water-related diseases
2017, Advances in Water ResourcesMalaria seroepidemiology in very low transmission settings in the Peruvian Amazon
2024, Scientific ReportsDescription of the UCAR/CU soil moisture product
2020, Remote Sensing