| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
There is an enormous heterogeneity in malaria transmission intensity in affected areas of Africa, and implementation of targeted malaria control operations would require an understanding of the forces that drive transmission. The assessment of indices relating to disease transmission therefore becomes central to disease control through quantifying the potential risk of infection and elucidating the patterns of disease transmission.1 The estimation of the entomologic inoculation rate (EIR) provides a standard and relatively simple means of quantifying levels of exposure to infected mosquitoes.2 The EIR uses an index of vector infectivity, the sporozoite rate, and the human-biting rate, which expresses the degree of human-vector contact. These two parameters are largely dependent on environmental conditions and would be expected to vary seasonally. It has been observed that vector species behavior, density, and the number of infective bites a person can receive per unit time is dependent on seasonal changes in environmental variables.3 Several studies have consequently compared transmission among different ecologic zones and even among villages within the same area, and have reported significant variations in transmission intensity among villages separated only by short distances and even between rural and urban settings within the same locality.3–5 Knowledge of the magnitude of these spatial variations is critical to understanding the transmission dynamics of the disease and the evaluation of the efficacy of vector control measures.6 This becomes even more crucial for the semi-arid regions of Africa where malaria is generally unstable and is occasionally experienced in epidemic proportions.
In this study, we examined the seasonality and spatial variations of malaria transmission at eight villages representing different ecologic zones of Eritrea. As the country aims to reduce the burden of malaria based on the RBM initiative, it is important to understand the degree of malaria distribution, risk, and burden so that control options can be targeted rationally.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). The villages represented different ecologies and variable malaria endemicity based on clinical evidence. Vector control strategies implemented at the study villages include environmental management, larviciding, and use of insecticide-impregnated bed nets. Coverage of at least two bed nets per household has been achieved in the study villages.
|
Sporozoite rates and the EIR. The head and thorax of each mosquito were separated from the abdomen and tested for the presence of Plasmodium falciparum circumsporozoite (CS) antigen.9 Mosquitoes were ground in 50 µL of boiled casein containing Nonidet 40 and final volume brought to 250 µL with blocking buffer. Fifty microliters of the triturate was used in sporozoite enzyme-linked immunosorbent assays. Positive reactions were determined visually.10 The EIR, expressed as the number of infective bites per unit time, was derived as a product of the sporozoite rate and the human-biting rate.1 The human-biting rate (HBR) was derived directly from HLCs and was expressed as the number of bites per person per night (b/p/n). The annual and per month inoculation rates were derived by multiplying the daily EIR (ib/p/n) by 365 and 30 days, respectively.
Statistical analysis. Analysis of variance was used to analyze differences in biting rates, sporozoite rates, and EIR between seasons and among villages. Log-transformed values of mosquito densities and arcsine-transformed values of sporozoite rates and EIR was used for statistical analysis. The chi-square test was used to test the significance of variations between human biting rates with regard to host location (indoor or outdoor). Data analysis was performed using SPSS version 10(SPSS Institute, Chicago, IL).
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2 = 19.29, df = 7, P = 0.007). The HBR (b/p/n) varied significantly between months (F = 6.64, df = 10, 191, P < 0.001). The An arabiensis biting activity was concentrated between July and November in the study villages in the highlands and western lowlands, with only peak biting density varying among sites. The highest HBRs of 37.8 b/p/n indoors and 54.8 b/p/n outdoors were obtained in Hiletsidi in September. In Maiaini, peak indoor (34.8 b/p/n) and outdoor (46.3 b/p/n) biting rates were recorded in September. In Adibosqual, the HBR in September was 32.5 b/p/n indoors and 47 b/p/n outdoors. Biting rates were generally low in the villages on the coast (range = 0.3–3.3 b/p/n), where biting activity was concentrated between December and February.
Plasmodium falciparum sporozoite rates A total of 4,886 An. arabiensis mosquitoes from HLCs and concurrent indoor-resting collections yielded an overall P. falciparum CS antigen positivity rate of 1.25% (n = 61). The sporozoite rates for the two villages surveyed in each zone combined was highest in the Gash-Barka zone (1.62%), which is also the most malarious region in the country and is located at an elevation between 500 and 1000 meters above sea level. The sporozoite rates recorded in the Anseba and Debub zones were 0.52% and 1.23%, respectively. None of the An. arabiensis tested (n = 103) from the two villages in the Northern Red Sea zone was positive for P. falciparum CS antigens. Overall, the difference in sporozoite rates among villages was not significantly different (F = 1.503, df = 7, 34, P = 0.209). Of the positive An. arabiensis mosquitoes, 70.5% (n = 43) were collected from the two villages in the Gash-Barka zone. Hiletsidi (570 meters above sea level) alone comprised 63.9% (n = 39) of the total positive mosquitoes with a sporozoite rate of 1.6%, while in Dasse the sporozoite rate was 1.8%.
The data revealed significant seasonal variation in sporozoite rates in An. arabiensis. For instance, the highest proportion of infected mosquitoes in Hiletsidi was recorded in August, although infection rates tended to increase between July and October (range = 0.6–5.2%). In Adibosqual, Maiaini, and Dasse, peak mosquito infection rates were recorded from July to September, coinciding with the rainy season. Generally low sporozoite rates were recorded during the dry season between November and May in all the study villages.
Entomologic inoculation rate
The EIRs differed significant among the study villages (F = 2.884, df = 7, 95, P = 0.009). The highest levels of transmission were recorded in Hiletsidi in the Gash Barka zone and in Maiaini in the Debub zone, with mean annual EIR figures of 70.6 and 32.1 infective bites per person, respectively, over the two-year period of the study (Table 1). The number of infective bites a person was exposed to annually was generally low in Dasse, Adibosqual, and Hagaz (range = 1.8–8.0 ib/p/year). The number of days it would take to receive one infective bite from An. arabiensis was also variable among the study villages, further showing the extent of geographic differences in transmission intensity. This ranged from 2.8 days in Hiletsidi to 203 days in Hagaz over the 24 months of the study.
|
Seasonal variations in the EIR
Malaria transmission, based on EIR estimates, occurred between July and October in four villages (Adibosqual, Dasse, Hagaz, and Maiaini), coinciding with the rainy season experienced in these sites from June to September (Table 2). Similarly, the main transmission season in Hiletsidi occurred between July and October (range = 5.01–43.62 ib/p/month) during the rainy season in the western lowlands. However, in this village there was evidence of malaria transmission during the dry season, with each resident exposed to 1.42 and 0.94 infective bites per month in March and May, respectively. This coincided with the short rainy season that occurs between March and May. Overall, seasonal patterns of malaria transmission were generally similar among the study villages (Figure 2). Overall, there was a significant positive correlation between EIRs and monthly rainfall amounts (r = 0.262, P < 0.001).
|
|
).
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The mean sporozoite rates for the combined villages per zone were two times higher in the villages in the Gash Barka zone than in the Debub and Anseba zones, and were significantly higher during the wet season (June to September) and the period soon after the rains (October to December). It is likely that the scarcity of rainfall during the dry season and the cold temperatures associated with the high altitude areas could have influenced geographic variations in survival rates and parasite development.13 Studies conducted at the Mwea-Tebere irrigation scheme in Kenya have shown that the survival rate of An. arabiensis was two times higher during the rainy season than in the dry season.14 It is therefore likely that due to prevailing environmental conditions at each of the villages by season, some An. arabiensis did not live long enough to complete sporogony. Although the present study did not measure house-level variations in sporozoite rates, it is also likely that differences on a microecologic scale may exist and would be important in determining the variability in malaria transmission intensity.1
The annual EIR estimates at the eight sites displayed enormous geographic variations (range = 0–70.6 ib/p/year). This confirms that malaria transmission is heterogeneous, consistent with the high degree of ecologic diversity in Eritrea. Since the EIR uses estimates of mosquito biting densities and sporozoite rates, it is likely that variation in the EIR was influenced by environmental determinants of vector populations and survival associated with different ecologic scenarios.15 The annual EIR derived in the present study for Hiletsidi and Maiaini would seem high, if one considers that these sites received only a mean annual rainfall of 286.9 mm and 378.1 mm, respectively, over the two years of study. More than 90% of the total rainfall was received between June and September, and more than 95% of the transmission risk was also restricted to the same period. The relatively small differences in rainfall and soil moisture content in these dry regions could have led to pronounced changes in ecology and consequent differences in the composition and population of mosquito vector species.16 The same factors would explain the variation in the EIR during the two years of the study. For instance, in Hiletsidi, the annual EIR was 10 times higher in the second year than in the first year, while in Dasse, the inoculation rate was twice as high in the first year as in the second year. Similar year-to-year variations in inoculation rates have been reported in two villages in Senegal (Dielmo and Ndiop).3
Despite the scarce rainfall in Eritrea, the inoculation rates obtained in the present study were generally comparable with records from most entomologic surveys across Africa, where transmission is short and restricted to approximately three months.17 However, the EIRs were generally lower compared with data from The Gambia, Kenya, Senegal, and Tanzania, where malaria is highly endemic. For example, annual inoculation rates in Tanzania ranged from 93.7 to 702.6 infective bites per person, with 75% of the transmission occurring in more than six months.17 Overall, entomologic assessment of transmission risk in semi-arid to arid regions of the continent where malaria occur mostly in severe epidemic proportions is generally lacking. Although EIRs = 0 were recorded in Shekaeyamo, Ghinda, and Gahtelay, it is likely that transmission is occurring below entomologic thresholds for detection. Some degree of variation in the EIR among study villages could also be attributed to inherent bias in selection of the villages because this was based on numbers of clinical malaria in recent years.
The study further demonstrated that EIR was significantly correlated with monthly rainfall amounts. In villages where transmission was detected entomologically, transmission was restricted to 1–3 months between July and December, except for Hiletsidi where transmission occurred also between March and May. The high wet season inoculation rates could be attributed to the fact that EIR estimates were based on HLCs, which increased with the onset of rains, and also may be a consequence of increased use of insecticide-impregnated bed nets in the study villages.18,19 The excito-repellency properties of pyrethroid insecticides could divert host-seeking mosquitoes to the exposed collectors, leading to an overestimation of the HBR. Eritrea has a successful bed net implementation program with coverage of at least two bed nets per household in at least 67% of the high-risk malaria regions (Ministry of Health, unpublished data).
Our study demonstrates that malaria transmission in Eritrea shows strong spatial and seasonal variations that have to be accounted for in planning for malaria control. For instance, the timing of bed net re-impregnation and selective indoor residual spraying would effectively be planned based on the onset of transmission in respective villages or regions. Eritrea is currently undertaking the implementation of vector control based on the RBM initiative, and the estimation of the EIR in the present study provides the malaria control program with means of quantifying levels of exposure to infected mosquitoes in different regions of the country. The data also provides important means for evaluating the efficacy of vector control measures, including use of impregnated bed nets. Despite the large differences in EIRs observed, it must be emphasized that annual EIRs, even as low as 5 or less, would still be responsible for substantial malaria prevalence and a high incidence of severe life-threatening malaria.4,15,20 Therefore, malaria control programs must also consider clinical and parasitologic assessments of disease to complement entomologic component in defining and further stratifying malarious areas in a country for effective disease control. Monitoring of rainfall amounts under similar semi-arid conditions experienced in Eritrea could provide a sensitive surveillance tool for forecasting malaria transmission and epidemics. Rainfall monitoring is currently being undertaken at a number of sentinel surveillance sites in Eritrea.
Received July 7, 2003. Accepted for publication July 30, 2003.
Acknowledgments: We are grateful to the zonal staff of the National Malaria Control Program for conducting the field surveys, and thank David Sintasath, Fesahaye Seulu, and G. E. Asmelash for their valuable technical support. The generous support of Saleh Meky (Minister of Health), Berhane G Tensae (DG Health Services), Dr. Andom Ogbamariam (Director of Communicable Diseases Control), and Linda Lou Kelley (Chief, Health Strengthening Office, United States Agency for International Development [USAID]/Asmara) is acknowledged. This work was performed by the Eritrea National Malaria Control Program to improve the knowledge base for making malaria vector control decisions.
Financial support: This study was supported by the USAID through a mechanism for cooperation with the Ministry of Health, Eritrea, and with technical support from the USAID Environmental Health Project.
Authors’ addresses: Josephat Shililu, National Malaria Control Program, Division of Communicable Diseases Control, Ministry of Health, PO Box 212, Asmara, Eritrea and International Centre of Insect Physiology and Ecology, PO Box 30772, Nairobi, Kenya. Tewolde Ghebremeskel, Solomon Mengistu, Helen Fekadu, and Mehari, Zerom, National Malaria Control Program, Division of Communicable Diseases Control, Ministry of Health, PO Box 212 Asmara, Eritrea, Telephone: 291-1-125529, Fax: 291-1-125835. Charles Mbogo and John Githure, International Centre of Insect Physiology and Ecology, PO Box 30772, Nairobi, Kenya, Telephone: 254-2-861680, Fax: 254-2-860110. Robert Novak, Illinois Natural History Survey, 607 E. Peabody, Champaign, IL 61820, Telephone: 217-333-1186, Fax: 217-333-2359. Eugene Brantly, Environmental Health Project, 1611 N. Kent Street, Suite 300, Arlington, VA 22209-2111, Telephone: 703-247-8730, Fax: 703-243-9004. John C. Beier, Department of Epidemiology and Public Health, University of Miami, School of Medicine, Highland Professional Building 1801 NW 9th Avenue, Suite 300 (D-93), Miami, FL 33136, Telephone: 305-243-2977, Fax: 305-243-3324.
Reprint requests: Josephat Shililu, National Malaria Control Program, Division of Communicable Diseases Control, Ministry of Health, PO Box 212, Asmara, Eritrea, Telephone: 291-1-125-529, Fax: 291-1-122-899 or 291-1-125835; E-mail: shililu{at}gemel.com.er.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. J. Kent, S. Mharakurwa, and D. E. Norris Spatial and Temporal Genetic Structure of Anopheles arabiensis in Southern Zambia over Consecutive Wet and Drought Years Am J Trop Med Hyg, August 1, 2007; 77(2): 316 - 323. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. KENT, P. E. THUMA, S. MHARAKURWA, and D. E. NORRIS SEASONALITY, BLOOD FEEDING BEHAVIOR, AND TRANSMISSION OF PLASMODIUM FALCIPARUM BY ANOPHELES ARABIENSIS AFTER AN EXTENDED DROUGHT IN SOUTHERN ZAMBIA Am J Trop Med Hyg, February 1, 2007; 76(2): 267 - 274. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
