SEARCH

SEARCH BY CITATION

Keywords:

  • visceral leishmaniasis;
  • Leishmania chagasi/infantum;
  • Lutzomyia pseudolongipalpis;
  • temporal and spatial distribution;
  • risk factors;
  • Venezuela
  • leishmaniose viscérale;
  • Leishmania chagasi/infantum;
  • Lutzomyia pseudolongipalpis;
  • distribution temporalle et spatiale;
  • facteurs de risque;
  • Venezuela
  • leishmaniasis visceral;
  • Leishmania chagasi/infantum;
  • Lutzomyia pseudolongipalpis;
  • distribución temporal y espacial;
  • factores de riesgo;
  • Venezuela

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Objective  To relate entomological, epidemiological and geographical data to understand the transmission dynamics of visceral leishmaniasis (VL) in a closed focus in western Venezuela.

Methods  Spatial and temporal patterns of Lutzomyia pseudolongipalpis, the most prevalent phlebotomine sand fly species (99.7%), were studied in El Brasilar, Curarigua, Lara State, Venezuela, a small rural community of 20 dwellings and 118 inhabitants. The sand fly population was monitored using Centers for Disease Control light traps monthly throughout 1 year in the domestic and sylvatic habitats and for 3 months in all inhabited houses.

Results  Temporal variation followed the yearly bimodal pattern of precipitation with the highest population densities in April and December. Infection with flagellates suggestive of Leishmania spp. was detected in 0.01% of 10 026 dissected females of L. pseudolongipalpis, which proved to be highly endophilic. Prevalence of Leishmania infection in people, as measured by the leishmanin skin test, was correlated with distance of the houses from the woodland and with sand fly abundance. A logistic regression model showed that for people who live in the village, the proximity to the woodland (linear) should be considered a risk factor for Leishmania infection (binary) (z = −2.02, P = 0.04, OR = 0.98, 95% CI = 0.97–0.99). This was consistent with the association between the proportion of VL infection and the log of sand fly abundance, which was negatively correlated with distance from the woodland.

Conclusion  We discuss strategies that might be useful in controlling VL transmission in this endemic focus.

Objectifs  Rapporter des données entomologiques, épidémiologiques et géographiques pour la compréhension de la dynamique de transmission de la leishmaniose viscérale dans foyer délimitée dans l'ouest du Venezuela.

Méthodes  Les profils spatiaux et temporaux de Lutzomyia pseudolongipalpis, l'espèce de phlébotomie la plus prévalente (99.7%) ont étéétudiés à El Brasilar à Curarigua dans l’état de Lara au Venezuela, une petite communauté rurale constituée de 20 habitations et de 118 habitants. La population de phlébotomes a été suivie mensuellement en utilisant des pièges lumineux CDC pendant une année dans les habitats domestiques et sylvatiques et pendant 3 mois dans toutes les maisons inhabitées.

Résultats  Les variations temporelles ont suivi le profil bimodal annuel des précipitations avec la population la plus dense en avril et décembre. L'infection par des flagellées suggestive de Leishmania spp a été détectée chez 0.01% des 10026 femelles disséquées de L. pseudolongipalpis qui s'est révélée fortement endophilique. La prévalence de l'infection leishmania mesurée par le test cutanée à la leishmanine corrélait avec la distance des maisons aux bois et à l'abondance des phlébotomes. Un modèle de régression logistique a montré que pour les personnes vivant dans le village, la proximité des bois devrait être considérée comme facteur de risque pour l'infection leishmania. Cette observation était consistante avec l'association entre la proportion d'infection leishmania et le log de l'abondance des phlébotomes, qui corrélait négativement avec la distance aux bois.

Conclusion  Nous discutons de stratégies qui pourraient être utiles dans le contrôle de la transmission de la leishmaniose viscérale dans ce foyer endémique.

Objetivo  Relacionar datos entomológicos, epidemiológicos y geográficos para entender la dinámica de transmisión de la leishmaniasis visceral (LV) en un foco cerrado de Venezuela.

Métodos  Se estudiaron los patrones espaciales y temporales de Lutzomyia pseudolongipalpis, la especie de flebótomo predominante (99.7%), en El Brasilar, Curarigua, estado de Lara, Venezuela, una pequeña comunidad rural de 20 domicilios y 118 habitantes. Se monitorizó la población de flebótomos, utilizando trampas de luz CDC mensuales durante 1 año en hábitats domésticos y selváticos, y durante 3 meses en todas las casas habitadas.

Resultados  La variación temporal siguió el patrón bimodal de precipitación anual, con las mayores densidades de población en Abril y Diciembre. Se detectó infección con sugestivos de Leishmania spp en 0.01% de 10,026 hembras disecadas de L. pseudolongipalpis disectadas, que resultó ser altamente endofílico. La prevalencia de infección por Leishmania, medida mediante la prueba cutánea con leishmanina, estaba correlacionada con la distancia de las casas al bosque y con la abundancia de flebótomos. Un modelo de regresión logística mostró que para las personas que viven en el poblado, la proximidad al bosque (linear) debería considerarse como un factor de riesgo de infección por Leishmania (binaria) (z = −2.02, P = 0.04, OR = 0.98, IC 95% = 0,.97–0.99). Esto era consistente con la asociación entre la proporción de infección por LV y el logaritmo de la abundancia de los flebótomos, que estaba a su vez correlacionada con la distancia al bosque.

Conclusión  Se discuten estrategias que podrían ser útiles en el control de la transmisión de LV en este foco endémico.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Visceral leishmaniasis (VL), a parasitic disease caused by the protozoan Leishmania chagasi/infantum and transmitted by the bite of phlebotomine sand flies, occurs globally with the exception of Australia and is responsible for 57 000 deaths and 1 932 000 disability-adjusted life years in 80 countries annually (Morel 2000). As in all vector-borne diseases, it is well known that the establishment and maintenance of VL is related to environmental factors that favour the presence of vectors and vertebrate hosts. It is also accepted that the understanding of these factors would be useful in guiding control programmes and land management. The domestic and peridomestic transmission and risk factors of this disease were reviewed by Desjeux (2001), who pointed out environmental conditions for VL infection that explain the emergence or resurgence of this disease. Because of the complexity of each epidemiological situation, he recommended that specific strategies must be tailored to each eco-epidemiological entity. Spatial heterogeneity together with temporal changes in abundance constitutes important elements for the understanding of the dynamics of vector populations (Liebhold et al. 1993) and is fundamental for the development of adequate prevention and control strategies. Specific research on spatial determinants of disease transmission is, therefore, needed on a small and a large scale to understand and to predict the epidemiological situation in a geographical area.

In a previous study, we examined the epidemiological features of the Leishmania chagasi/infantum infection in people living in El Brasilar, a closed focus of VL in Lara State, Venezuela (Feliciangeli et al. 2005). In this study, we deal with the spatial distribution of infection and we present data on the spatial and temporal distribution of the vector in an attempt to detect environmental risk factors associated with VL at this focus.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Study area

El Brasilar (69° 55′ W, 9° 59′ N) is a small rural community in west central Venezuela, 96 km from Barquisimeto, capital of Lara State, and about 3 km from the nearest town Curarigua. Figure 1 is a thematic mapper Landsat image of the area showing its location in a valley (between 600 and 650 m above sea level), surrounded on the southwest by a chain of hills. To the east, the Curarigua River runs through an extensively cultivated area. To the north side, there is also a chain of volcanic rocks. The entrance to the village, which communicates El Brasilar to Curarigua, is to the northeast. Primary vegetation of spiny shrubs, cacti (Opuntia spp.) and trees such as yabo (Cercidium praecox), yacure (Pitbecolobium unquiscati) and urapaquita (Castella nicholsoni) is typical of this tropical thorny forest, where the mean annual temperature is 24 °C and the annual precipitation is 250–500 mm (Ewel et al. 1976). In February 2002, when the census was carried out, the village consisted of 25 houses, five of which (3,4,8,17 and 21) were uninhabited (Figure 2). A total of 118 people were living at that time in the village. Except for one or two houses made of cement bricks, the rest of the dwellings were constructed of a mixture of mud and straw supported by a structure of sticks, with zinc roofs. Agriculture is the principal local economy.

image

Figure 1.  Landsat image of the study area in Lara State, Venezuela, showing the valley where El Brasilar is located, the relative position to Curarigua town and Curarigua River (Landsat 7 ETM+; acquisition date 20 November 2000; path = 006; row = 053).

Download figure to PowerPoint

image

Figure 2.  Diagram of El Brasilar showing the relative location of the houses in relation to the nearest distance from the woodland (The line shows the borders of the woodland).

Download figure to PowerPoint

Leishmania infection among the population

The first cases of VL were registered in El Brasilar in 1989, and since then the disease has only been reported in children under the age of 10 years. A prospective study was carried out between 1998 and 2002. In the second survey, 79 people living in 20 houses were tested; 34 people (43.0%) showed a positive reaction (a papule geqslant R: gt-or-equal, slanted5 mm) to antibodies of Leishmania spp. with the leishmanin skin test (Feliciangeli et al. 2005). Demographic variables, age, sex and occupation, were not found to be significant risk factors for Leishmania infection, leading to the conclusion that all the inhabitants were equally exposed to the risk of infection and therefore the transmission seemed to occur mainly indoors.

In this article, we give information and analyse the spatial distribution of prevalence per house in 2002. The coordinates of each house and its distance to the nearest woodland were recorded using a Trimble Geoexplorer III (Trimble Navigation Ltd, Sunnyale, CA, USA). In the diagram map drawn with these coordinates using the software package ArcView 3.2 (ERSI) (Figure 2), we show the relative position of each house, and its respective distance from the woodland is given in Table 1. We considered the woodland border the limit at which the vegetation was not disturbed by man.

Table 1.   Spatial, entomological and epidemiological data collected at El Brasilar, Lara State, Venezuela
House no.Distance from the woodland (m)No. of Lutzomyia pseudolongipalpis females/trap/night LST+ people proportion (n)
  1. †Trapping was incomplete in these houses; therefore they were excluded from the analysis.

111172.731.000 (8)
24681.271.000 (2)
5621.821.000 (2)
6650.30 (3)
791–†0.500 (2)
91025.820.333 (3)
1042–†0.555 (9)
11791.550 (1)
121419.730.750 (4)
13151.640.800 (5)
141162.730 (4)
1512010.333 (6)
161460.450.333 (3)
1819700.143 (7)
192200.270 (1)
2024900 (1)
222680.360 (5)
2328500.250 (4)
2426000 (4)
25420.450.667 (5)

Sand fly trapping

Monthly sand fly collections were conducted in El Brasilar from June 1997 to May 1998 during three to four consecutive nights per month, except in October 1997, because of logistic problems. Catches were performed indoors with Centres for Disease Control (CDC) miniature light traps (Model 512; John W. Hock Company). One trap was set up overnight (18:30–6:30 hours) in the main bedroom of two houses (1 and 2) where VL cases had occurred. At the same time, CDC light traps were placed in the woodland among shrubs. One to four traps/night, but in most cases three traps/night, were used, as on a few nights heavy rains hindered the placement of traps outdoors. Monthly precipitation data were provided during the study period by the weather station of the Ministry of the Environment located in Curarigua. Spatial sand fly distribution was studied for 3 months (March 1998 to May 1998) by placing a CDC trap in the main bedroom of the 20 inhabited houses. However, because of the repeated absence of the inhabitants in the houses 7 and 10, only results obtained from those where all the catches were actually carried out (n = 18) were included in the analysis.

Sand fly identification and processing for searching natural infection with Leishmania spp.

Male sand flies were separated from females and kept in 70% ethanol for later identification, while females were stored in vials with 10% dimethyl sulphoxide (DMSO) and cryopreserved in liquid nitrogen. Male identification was carried out after clarifying in Nesbitt's solution (Krantz 1978) and inspection of the morphology of the head and genitalia. Females were dissected in sterile phosphate-buffered saline solution, pH 7.2, under a stereoscopic microscope and the gut was observed under a phase contrast microscope at 400× for Leishmania detection. If parasites were seen, the slide was stained with Giemsa to confirm the promastigote morphology. Sand fly species identification was based on the morphology of characters in the head (ascoids, palps, pharynx, cibarial armature) and spermathecae (Young & Duncan 1994).

Statistical analysis

The counts of sand flies collected (number of sand flies per trap/night) were ln (x + 1) transformed to homogenize the variance, and the Williams geometric mean was calculated. The male:female ratios in the house and in the field were compared using the chi-square test. The effect of the number of traps on the sand fly abundance in the woodland was tested using anova. Assuming that the count of sand flies was representative of the population abundance/house at any given time, a linear regression analysis was performed to test for an association between the log number of Lutzomyia pseudolongipalpis/house (as the dependent variable) and the distance from houses to the woodland (as the independent variable). A logistic (binary) regression analysis was used to test for an association between the presence/absence of people infected with Leishmania chagasi/infantum in each house (as dependent variable) and the distance from each house to the woodland (independent variable). A generalized lineal model was also used to perform a logistic regression analysis to test for an association between the proportion of VL infection and the log sand fly abundance, where the dependent grouped numerator was the total number of people infected with Leishmania sp. and the dependent grouped denominator the total number of people surveyed. All analyses were performed using a stata 8.1 package.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Table 1 gives data on the distance of each house from the woodland, the abundance of female L. pseudolongipalpis (no. of traps/night) collected per house in the period March 1998 to May 1998 and the proportion of people infected per house.

Prevalence of Leishmania infection vs. distance to the woodland

The distance of the houses from the woodland ranged from 10–285 m; 60% of them were located within the mean distance (116.9 m) and 85.3% of people infected with Leishmania spp. inhabited in those houses (χ2 = 13.32; Yates corrected P = 0.0003). The risk of infection with Leishmania spp. for people living in the village significantly decreased, as the distance from the house to the woodland increased (OR = 0.988; 95% CI = 0.97–0.99; z = −2.02; P = 0.043).

Spatial sand fly distribution

A total of 10 075 L. pseudolongipalpis (6664 males and 3411 females) were collected in 11 nights in 18 houses of El Brasilar from March to May 1998. However, 98.16% of them were caught in only four houses (1, 2, 9 and 12), whose average distance from the woodland was 20.25 ± 14.94 m. Linear regression analysis indicated that sand fly abundance was significantly associated with the distance to the woodland (F1,16 = 12.51; P = 0.003; R2 = 0.44) with a negative intercept at c. 240 m. The prevalence of VL in each house was significantly associated with log of sand fly abundance (coeff.: 0.8; 95% CI: 0.4–1.3; P = 0.001), taking into account the total number of people at risk.

Temporal sand fly abundance

During the year study, five sand fly species were collected in El Brasilar. Four of them were collected in negligible amounts indoors (H) as well as in the woodland (W): Lutzomyia evansi (1 female H), Lutzomyia cayennensis (22 males, 12 females H; 8 males, 9 females W); Lutzomyia trinidadensis (1 male, 3 females H; 5 males, 4 females W) and Lutzomyia dubitans (3 females H; 5 males, 3 females W). Lutzomyia pseudolongipalpis was the most prevalent species. Overall, a total of 80 samples in 2 houses yielded a total of 14 674 males and 7375 females and 106 samples in the woodland yielded 9995 males and 5200 females. In this habitat, the anova analysis demonstrated that the number of traps did not have a significant effect on the mean sand fly abundance (df = 3; P > 0.05). Males were almost twice abundant as females (Table 2), and no significant difference was found in comparing these proportions domestically and sylvatically. Temporally, L. pseudolongipalpis indoors displayed a bimodal annual abundance pattern, with two peaks that lagged behind the rainfall peaks, the first peak in December and the second in April (Figure 3). In the field, sand fly density was lower than that in the houses, showing a peak only in April (Figure 4).

Table 2.   Abundance of Lutzomyia pseudolongipalpis in El Brasilar, Lara State, Venezuela
 No. of traps (no. of nights)MalesWilliams geometric mean (95% CI)FemalesWilliams geometric mean (95% CI)Sex ratio (m:f)
Indoors80 (40)14 67480.04(3.89–4.89)737544.18(3.38–4.23)(1.95:1)
Field106 (39)999555.81(3.67–4.41)520029.39(3.06–3.76)(1.92:1)
Total 24 679  12 575  (1.96:1)
image

Figure 3.  Monthly rainfall and temporal pattern of Lutzomyia pseudolongipalpis indoors.

Download figure to PowerPoint

image

Figure 4.  Monthly rainfall and temporal pattern of Lutzomyia pseudolongipalpis in the field.

Download figure to PowerPoint

Sand fly natural infection with Leishmania spp.

A total of 10 026 females of L. pseudolongipalpis randomly selected, among those collected in the field and in the houses, were dissected and 1 (0.01%) was found positive for flagellates. The Giemsa-stained slide showed that they were promastigotes, suggestive of Leishmania spp.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This is the first spatial approach in Venezuela to study the epidemiology of VL in an attempt to relate human L. chagasi/infantum infection to entomological and environmental factors. The presence of infected individuals in the houses was negatively associated with the distance of the houses from the woodland. It is apparent that the closeness of the houses to the woodland is a likely risk factor for L. chagasi/infantum infection for people living in El Brasilar. In fact, about 85% of infected people lived in houses between 10 and 117 m from the woodland. The spatial distribution of L. pseudolongipalpis showed that greater numbers were captured in houses very close to the woodland. Different anthropic behaviour in the management of household environment (e.g. clearing vegetation and cleaning organic material from the peridomestic area) might explain the lower abundance in dwellings at similar distance. However, L. pseudolongipalpis still entered houses at 268 m and one infected person was found at 285 m. The distance from bordering forest and pastures at which Werneck et al. (2002) observed clustering of VL cases in Teresina, Brazil, was 300 m. Morrison et al. (1993) reported that 49% of Lutzomyia longipalpis marked in El Callejón, Colombia, were recaptured within 0–50 m, 48% within 100–300 m and nearly 3% at geqslant R: gt-or-equal, slanted0.5 km from the release site. If, from these and our data, we assume a similar dispersal pattern for L. pseudolongipalpis, the transmission dynamics in El Brasilar might be explained in the context of the landscape, showing how people, trapped in a small enclosed area within the flight range of L. pseudolongipalpis, are all exposed to the risk of infection. However, greater closeness to the woodland means a greater number of sand flies in the house and an increased risk of being bitten and of becoming infected. This conclusion is consistent with the significant association between sand fly abundance and proximity of the houses to the woodland and with the association between sand fly abundance and proportion of infected people in such houses. Similarly, Werneck (2000) in Teresina, Brazil, showed that in 1061 recorded cases during an epidemic outbreak in 1993–1996, the vector index (a surrogate for the sand fly abundance) was associated with a threefold increased risk of VL.

The bimodal pattern of the annual L. pseudolongipalpis distribution in El Brasilar, with two peaks of high densities following the peaks of rainfall, was clearly evident in the indoor population and is consistent with observations in other sand fly species (Travi et al. 1996, Salomón et al. 2004). On the contrary, the abundance of L. pseudolongipalpis indoors might be explained by the phototropism that leads females, attracted by porch lights, to concentrate in and around houses in search of a blood meal and males seeking a mate. Because of the dispersion, and the lower light level of the CDC traps, smaller numbers of sand flies would be trapped in the woodland.

Information on the bionomics and on the temporal distribution trends of the vector species is essential for planning vector-control strategies. The use of impregnated curtains has been effective in decreasing the transmission of cutaneous leishmaniasis in Trujillo State, Venezuela (Kroeger et al. 2002) and similarly in Colombia where curtains and bed nets were also employed (Alexander et al. 1995). By comparison, in hot areas like in El Brasilar, people do not use bed nets frequently. On the Island of Margarita, residual spraying with λ-cyhalothrin was shown to diminish the population density of L. longipalpis (Feliciangeli et al. 2003). The abundance and high endophily of L. pseudolongipalpis combine to pose a high risk for indoor VL transmission in El Brasilar. Here, we would recommend that at night, small children should always be protected by bed nets, even untreated ones, and that applying two residual sprayings at the beginning of the rainfalls would reduce the L. pseudolongipalpis population and control Leishmania transmission in the village. Additionally, in order to avoid abandoning of the village, people should be encouraged to reallocate their houses far from the woodland. However, research on knowledge, aptitudes and practice is undoubtedly necessary before implementing a control programme in which communities could actively participate.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We are indebted with the inhabitants of El Brasilar for their warm hospitality and helpful collaboration during the fieldwork. We thank Prof. Omar Verde for the helpful discussions on the statistical analysis and Marco Gaiani who provided the Landsat image of the study area. This research was supported by the University of Carabobo (CDCH, Project FCS-2001–005).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Alexander B, Usma MC, Cadena H et al. (1995) Evaluation of deltamethrin-impregnated bednets and curtains against phlebotomine sandflies in Valle del Cauca, Colombia. Medical and Veterinary Entomology 9, 279283.
  • Desjeux P (2001) The increase in risk factors for leishmaniasis worldwide. Transactions of the Royal Society of Tropical Medicine and Hygiene 95, 239243.
  • Ewel J, Madriz y A & Tosi J (1976) Zonas de Vida de Venezuela, 2nd Edn. Ministerio de Agricultura y Cría, Fondo Nacional de Investigaciones Agropecuarias, Caracas, Venezuela, pp. 265.
  • Feliciangeli MD, Mazzarri MB, San Blas S & Zerpa O (2003) Control trial of Lutzomyia longipalpis s.l. in the Island of Margarita. Tropical Medicine and International Health 8, 11311136.
  • Feliciangeli MD, Delgado O, Suarez B & Chiurillo MA (2005) The burden of the Leishmania chagasi/infantum infection in a closed rural focus of visceral leishmaniasis in Lara state, west-central Venezuela. Tropical Medicine and International Health 10, 444449.
  • Krantz GW (1978) A Manual of Acarology. Oregon State University, Corvallis, Oregon.
  • Kroeger A, Avila EV & Morison L (2002) Insecticide impregnated curtains to control domestic transmission of cutaneous leishmaniasis in Venezuela: cluster randomised trial. British Medical Journal 325, 810813.
  • Liebhold AM, Rossi RE & Kemp WP (1993) Geostatistics and geographic information systems in applied insect ecology. Annual Review of Entomology 38, 303327.
  • Morel CM (2000) Reaching maturity–25 years of the TDR. Parasitology Today 16, 522526.
  • Morrison AC, Ferro C, Morales A, Tesh RB & Wilson ML (1993) Dispersal of the sand fly Lutzomyia longipalpis (Diptera: psychodidae) at an endemic focus of visceral leishmaniasis in Colombia. Journal of Medical Entomology 30, 427435.
  • Salomón OD, Wilson ML, Mustermann LE & Travi BL (2004) Spatial and temporal patterns of phlebotomine sand flies (Diptera: Psychodidae) in a cutaneous leishmaniasis focus in northern Argentina. Journal of Medical Entomology 41, 3339.
  • Travi BL, Montoya J, Gallego J, Jaramillo C, Llano R & Velez I (1996) Bionomics of Lutzomyia evansi (Diptera: Psychodidae) vector of visceral leishmaniasis in northern Colombia. Journal of Medical Entomology 33, 278285.
  • Werneck GL (2000) Spatial approaches to study the epidemiology of visceral leishmaniasis in Teresina, Brazil. Cadernos de Saúde Pública 8, 9395.
  • Werneck GL, Costa CH, Walker AM, David JR, Wand M & Maguire JH (2002) The urban spread of visceral leishmaniasis: clues from spatial analysis. Epidemiology 13, 364367.
  • Young DG & Duncan MA (1994) Guide to the identification and geographic distribution of Lutzomyia sand flies in Mexico, the West Indies, central and South America (Diptera: Psychodidae). Memoirs of the American Entomological Institute 54, 1–881.