• View in gallery

    Map of the state of Acre, northwestern Brazil, showing the study site. Ramal do Granada is part of the Pedro Peixoto Agricultural Settlement (shaded area in the inset), located 30–45 km northwest of the town of Acrelândia. The location of BR-364, the only paved highway connecting the capital of Acre (Rio Branco) to the rest of the country, is also indicated.

  • View in gallery

    Number of cases of dengue fever recorded in the state of Acre (population, 669,737) between 1999 and 2006. Only three dengue fever episodes were reported in 1999. During the dengue outbreak of 2004, 5,892 cases (84.1% of the 7008 cases in Acre) were diagnosed in residents of the state capital, Rio Branco (population, 314,127). The vast majority of dengue fever episodes recorded in Acre during 2004 (6,250 of 7,008 cases or 89.2%) occurred between October and December (unpublished data of the Ministry of Health of Brazil; available at http://portal.saude.gov.br/portal/svs/area.cfm?id_area=451).

  • 1

    Tauil PL, 2006. Perspectivas de controle de doenças transmitidas por vetores no Brasil. Rev Soc Bras Med Trop 39 :275–277.

  • 2

    Siqueira JB, Martelli CMT, Coelho GE, Simplício ACR, Hatch DL, 2005. Dengue and dengue hemorrhagic fever, Brazil, 1981–2002. Emerg Infect Dis 11 :48–53.

    • Search Google Scholar
    • Export Citation
  • 3

    Torres JR, Castro J, 2007. The health and economic impact of dengue in Latin America. Cad Saude Publica 23 (Suppl. 1):S23–S31.

  • 4

    Chareonsook O, Foy HM, Teeraratkul A, Silarug N, 1999. Changing epidemiology of dengue hemorrhagic fever in Thailand. Epidemiol Infect 122 :161–166.

    • Search Google Scholar
    • Export Citation
  • 5

    Strickman D, Sithiprasasna R, Kittayapong P, Innis BL, 2000. Distribution of dengue and Japanese encephalitis virus among children in rural and suburban Thai villages. Am J Trop Med Hyg 63 :27–35.

    • Search Google Scholar
    • Export Citation
  • 6

    Lian CW, Seng CM, Chai WY, 2006. Spatial, environmental and entomological risk factor analysis on a rural dengue outbreak in Lundu District in Sarawak, Malaysia. Trop Biomed 23 :85–96.

    • Search Google Scholar
    • Export Citation
  • 7

    Kumar A, Sharma SK, Padbidri VS, Thakare JP, Jain DC, Datta KK, 2001. An outbreak of dengue fever in rural areas of northern India. J Commun Dis 33 :274–281.

    • Search Google Scholar
    • Export Citation
  • 8

    Tewari SC, Thenmozhi V, Katholi CR, Manavalan R, Munirathinam A, Gajanana A, 2004. Dengue vector prevalence and virus infection in a rural area in south India. Trop Med Int Health 9 :499–507.

    • Search Google Scholar
    • Export Citation
  • 9

    Hayes CG, Phillips IA, Callahan JD, Griebenow WF, Hyams KC, Wu SJ, Watts DM, 1996. The epidemiology of dengue virus infection among urban, jungle, and rural populations in the Amazon Region of Peru. Am J Trop Med Hyg 55 :459–463.

    • Search Google Scholar
    • Export Citation
  • 10

    Reiskind MH, Baisley KJ, Calampa C, Sharp TW, Watts DM, Wilson ML, 2001. Epidemiological and ecological characteristics of past dengue virus infection in Santa Clara, Peru. Trop Med Int Health 6 :212–218.

    • Search Google Scholar
    • Export Citation
  • 11

    Silva-Nunes M, Malafronte Rdos S, Luz Bde A, Souza EA, Martins LC, Rodrigues SG, Chiang JO, Vasconcelos PF, Muniz PT, Ferreira MU, 2006. The Acre project: the epidemiology of malaria and arthropod-borne virus infections in a rural Amazonian population. Cad Saude Publica 22 :1325–1334.

    • Search Google Scholar
    • Export Citation
  • 12

    Filmer D, Pritchett LH, 2001. Estimating wealth effects without expenditure data-or tear: an application to educational enrolments in states of India. Demography 38 :115–132.

    • Search Google Scholar
    • Export Citation
  • 13

    Trape JF, 1985. Rapid evaluation of malaria parasite density and standardization of thick smear examination for epidemiological investigations. Trans R Soc Trop Med Hyg 79 :181–184.

    • Search Google Scholar
    • Export Citation
  • 14

    de Souza VAF, Fernandes S, Araújo ES, Tateno AF, Oliveira OMNPF, Oliveira RR, Pannuti CS, 2004. Use of an immunoglobulin G avidity test to discriminate between primary and secondary dengue virus infections. J Clin Microbiol 42 :1782–1784.

    • Search Google Scholar
    • Export Citation
  • 15

    Stefano I, Sato HK, Pannuti CS, Omoto TM, Mann G, Freire MS, Yamamura AM, Vasconcelos PF, Oselka GW, Weckx LW, Salgado MF, Noale LF, Souza VA, 1999. Recent immunization against measles does not interfere with the seroresponse to yellow fever vaccine. Vaccine 17 :1042–1046.

    • Search Google Scholar
    • Export Citation
  • 16

    Freire MS, Mann GF, Marchevsky RS, Yamamura AMY, Almeida LFC, Jabor AV, Malachias JMN, Coutinho ESF, Galler R, 2005. Production of yellow fever 17DD vaccine in primary culture of chicken embryo fibroblasts: yields, thermo and genetic stability, attenuation and immunogenicity. Vaccine 23 :250–252.

    • Search Google Scholar
    • Export Citation
  • 17

    Chomczynski P, Sacchi P, 1987. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162 :156–159.

    • Search Google Scholar
    • Export Citation
  • 18

    Lanciotti RS, Calisher CH, Gubler DJ, Chang GJ, Vorndam AV, 1992. Rapid detection and typing of dengue viruses from clinical samples by using reverse transcriptase-polymerase chain reaction. J Clin Microbiol 30 :545–551.

    • Search Google Scholar
    • Export Citation
  • 19

    de Morais Bronzoni RV, Baleotti FG, Ribeiro Nogueira RM, Nunes M, Moraes Figueiredo LT, 2005. Duplex reverse transcription-PCR followed by nested PCR assays for detection and identification of Brazilian alphaviruses and flaviviruses. J Clin Microbiol 43 :696–702.

    • Search Google Scholar
    • Export Citation
  • 20

    Tamura K, Dudley J, Nei M, Kumar S, 2007. MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol Biol Evol 24 :1596–1599.

    • Search Google Scholar
    • Export Citation
  • 21

    Regato M, Recarey R, Moratorio G, de Mora D, Garcia-Aguirre L, Gónzalez M, Mosquera C, Alava A, Fajardo A, Alvarez M, D’Andrea L, Dubra A, Martínez M, Khan B, Cristina J, 2008. Phylogenetic analysis of the NS5 gene of dengue viruses isolated in Ecuador. Virus Res 132 :197–200.

    • Search Google Scholar
    • Export Citation
  • 22

    Baleotti FG, Moreli ML, Figueiredo LT, 2003. Brazilian Flavivirus phylogeny based on NS5. Mem Inst Oswaldo Cruz 98 :379–382.

  • 23

    Gubler DJ, Kuno G, Sather GE, Velez M, Oliver A, 1984. Use of mosquito cell cultures and specific monoclonal antibodies in surveillance for dengue virus. Am J Trop Med Hyg 33 :158–165.

    • Search Google Scholar
    • Export Citation
  • 24

    Vanwambeke SO, van Benthem BHB, Khantikul N, Burgoorn-Mass C, Panart K, Oskam L, Lambin EF, Somboon P, 2006. Multilevel analysis of spatial and temporal determinants for dengue infection. Int J Health Geogr 5 :5.

    • Search Google Scholar
    • Export Citation
  • 25

    Kulldorff M, Nagarwalla N, 1995. Spatial disease clusters: detection and inference. Stat Med 14 :799–819.

  • 26

    Shope RE, Sather GE, 1979. Arboviruses. Lennette EH, Schmidt NJ, eds. Diagnostic Procedures for Viral, Rickettsial and Chlamydial Infections. Fifth edition. Washington, DC: American Public Health Association, 767–814.

  • 27

    Tavares-Neto J, Freitas-Carvalho J, Nunes MRT, Rocha G, Rodrigues SG, Damasceno E, Darub R, Viana S, Vasconcelos PFC, 2004. Pesquisa de anticorpos contra arbovírus e o v rus í vacinal da febre amarela em uma amostra da população de Rio Branco, antes e três meses após a vacina 17D. Rev Soc Bras Med Trop 37 :1–6.

    • Search Google Scholar
    • Export Citation
  • 28

    Galler R, Marchevsky RS, Caride E, Almeida LFC, Yamamura AMY, Jabor AV, Motta MCA, Bonaldo MC, Coutinho ESF, Freire MS, 2005. Attenuation and immunogenicity of recombinant yellow fever 17D-dengue type 2 virus for rhesus monkeys. Braz J Med Biol Res 38 :1835–1846.

    • Search Google Scholar
    • Export Citation
  • 29

    Aquino VH, Anatriello E, Gonçalves PF, da Silva EV, Vasconcelos PFC, Vieira DS, Batista WC, Bobadilla MI, Vazquez C, Morán M, Figueiredo LTM, 2006. Molecular epidemiology of dengue type 3 virus in Brazil and Paraguay, 2002–2004. Am J Trop Med Hyg 75 :710–715.

    • Search Google Scholar
    • Export Citation
  • 30

    Ministry of Health of Brazil, 2002. Programa Nacional de Controle da Dengue. Brasília: Ministério da Saúde.

  • 31

    Endy TP, Nisalak A, Chunsuttiwat S, Libraty DH, Green S, Rothman AL, Vaughn DW, Ennis FA, 2002. Spatial and temporal circulation of dengue virus serotypes: a prospective study of primary school children in Kamphaeng Phet, Thailand. Am J Epidemiol 156 :52–59.

    • Search Google Scholar
    • Export Citation
  • 32

    Thai KTD, Binh TQ, Giao PT, Phuong HL, Hung LQ, Nam NV, Nga TT, Groen J, Nagelkerke N, de Vries PJ, 2005. Seroprevalence of dengue antibodies, annual incidence and risk factors among children in southern Vietnam. Trop Med Int Health 10 :379–386.

    • Search Google Scholar
    • Export Citation
  • 33

    Balmasaeda A, Hammond SN, Tellez Y, Imhoff L, Rodriguez Y, Saborío SI, Mercado JC, Perez L, Videa E, Almanza E, Kuan G, Reyes M, Saenz L, Amador JJ, Harris E, 2006. High seroprevalence of antibodies against dengue virus in a prospective study of schoolchildren in Managua, Nicaragua. Trop Med Int Health 11 :935–942.

    • Search Google Scholar
    • Export Citation
  • 34

    Endy TP, Chunsuttiwat S, Nisalak A, Libraty DH, Green AS, Rothman AL, Vaughn DW, Ennis FA, 2002. Epidemiology of innaparent and symptomatic acute dengue virus infection: a prospective study of primary school children in Kamphaeng Phet, Thailand. Am J Epidemiol 156 :40–51.

    • Search Google Scholar
    • Export Citation
  • 35

    Guha-Sapir D, Schimmer B, 2005. Dengue fever: new paradigms for a changing epidemiology. Emerg Themes Epidemiol 2 :1.

  • 36

    Kaplan JE, Eliason DA, Moore M, Sather GE, Schonberger LB, Cabrera-Coello L, Fernandez de Castro J, 1983. Epidemiological investigations of dengue infection in Mexico, 1980. Am J Epidemiol 117 :335–343.

    • Search Google Scholar
    • Export Citation
  • 37

    Halsted SB, Nimmannitya S, Cohen SN, 1970. Observations related to pathogenesis of dengue hemorrhagic fever. IV Relation of disease severity to antibody response and virus recovered. Yale J Biol Med 42 :311–328.

    • Search Google Scholar
    • Export Citation
  • 38

    Vasconcelos PFC, Lima JWO, Raposo ML, Rodrigues SG, Travassos da Rosa JFS, Amorim SMC, Travassos da Rosa ES, Moura CMP, Fonseca N, Travassos da Rosa APA, 1999. Inquérito soroepidemiológico na Ilha de São Luís durante epidemia de dengue no Maranhão. Rev Soc Bras Med Trop 32 :171–179.

    • Search Google Scholar
    • Export Citation
  • 39

    Pan American Health Organization, 1994. Dengue and Dengue Hemorrhagic Fever in the Americas: Guidelines for Prevention and Control. Washington, DC: Pan American Health Organization.

  • 40

    Nunes-Araújo FRF, Ferreira MS, Nishioka SA, 2003. Dengue fever in Brazilian adults and children: assessment of clinical findings and their validity for diagnosis. Ann Trop Med Parasitol 97 :415–419.

    • Search Google Scholar
    • Export Citation
  • 41

    Rodrigues MBP, Freire HBM, Corrêa PRL, Mendonça ML, Silva MRI, França EB, 2005. Is it possible to identify dengue in children on the basis of Ministry of Health criteria for suspected dengue cases? J Pediatr (Rio J) 81 :209–215.

    • Search Google Scholar
    • Export Citation
  • 42

    Dietz VJ, Gubler DJ, Rigau-Pérez JG, Pinheiro F, Schatzmayr HG, Bailey R, Gunn RA, 1990. Epidemic dengue 1 in Brazil, 1986: evaluation of a clinically based dengue surveillance system. Am J Epidemiol 131 :693–701.

    • Search Google Scholar
    • Export Citation
  • 43

    Cunha RV, Schatzmayr HG, Miagostovich MP, Barbosa AMA, Paiva FP, Miranda RMO, Ramos CCF, Coelho JCO, Santos FB, Nogueira RMR, 1999. Dengue epidemic in the state of Rio Grande do Norte, Brazil, in 1997. Trans R Soc Trop Med Hyg 93 :247–249.

    • Search Google Scholar
    • Export Citation
 
 
 

 

 

 

 

 

 

Risk Factors for Dengue Virus Infection in Rural Amazonia: Population-based Cross-sectional Surveys

View More View Less
  • 1 Department of Parasitology, Institute of Biomedical Sciences, University of São Paulo, São Paulo, Brazil; Laboratory of Virology, Institute of Tropical Medicine of São Paulo, São Paulo, Brazil; Laboratory of Molecular Biology, Marília Medical School, Marília, Brazil; Laboratory of Virology, Faculty of Medicine of São José do Rio Preto, São José do Rio Preto, Brazil; Institute of Technology in Immunobiologicals, Oswaldo Cruz Foundation, Riuo de Janeiro, Brazil; Laboratory of Protozoology, Institute of Tropical Medicine of São Paulo, São Paulo, Brazil; Department of Health Sciences, Federal University of Acre, Rio Branco, Brazil; Laboratory of Arbovirology and Hemorrhagic Fevers, Evandro Chagas Institute, Belém, Brazil

A comparison of dengue virus (DENV) antibody levels in paired serum samples collected from predominantly DENV-naive residents in an agricultural settlement in Brazilian Amazonia (baseline seroprevalence, 18.3%) showed a seroconversion rate of 3.67 episodes/100 person-years at risk during 12 months of follow-up. Multivariate analysis identified male sex, poverty, and migration from extra-Amazonian states as significant predictors of baseline DENV seropositivity, whereas male sex, a history of clinical diagnosis of dengue fever, and travel to an urban area predicted subsequent seroconversion. The laboratory surveillance of acute febrile illnesses implemented at the study site and in a nearby town between 2004 and 2006 confirmed 11 DENV infections among 102 episodes studied with DENV IgM detection, reverse transcriptase-polymerase chain reaction, and virus isolation; DENV-3 was isolated. Because DENV exposure is associated with migration or travel, personal protection measures when visiting high-risk urban areas may reduce the incidence of DENV infection in this rural population.

INTRODUCTION

Dengue is the most important arboviral disease of humans, with 2.5 billion people at risk worldwide and 50 million new infections each year, mostly in Southeast Asia, Western Pacific, and the Americas (World Health Organization website; http://www.who.int/ctd/dengue/burdens.html). After the successful eradication of Aedes aegypti in the mid-1950s, Brazil remained free of known dengue vectors until 1976.1 Dengue virus (DENV) was reintroduced into northern Brazil in 1981 and has been spreading throughout the country since 1986.2 Over the past 10 years, Brazil has contributed 70% cases of dengue fever reported in the Americas,3 with 345,922 cases and 67 deaths recorded in 2006 (Ministry of Health of Brazil, unpublished data).

Urban populations are the main targets of dengue control programs worldwide, but DENV is also spreading to rural areas in several Asian countries, such as Thailand,4,5 Malaysia,6 and India.7,8 However, it remains uncertain what favors DENV transmission in these settings.

The spread of DENV transmission to rural areas in the Americas might represent a formidable challenge to dengue control strategies. Serologic evidence of DENV exposure has been documented in rural communities in the Amazon Basin of Peru in the 1990s,9,10 but no data are available for other rural Amazonian populations. Here we describe the epidemiology of DENV infection in one of the largest agricultural settlements in the Amazon Basin of Brazil: the Pedro Peixoto settlement in the state of Acre. We analyze individual and household-level risk factors for the presence of DENV antibodies at baseline and for seroconversion after 6–12 months of follow-up. We discuss the prospects for dengue control in this and other similar rural settings.

MATERIALS AND METHODS

Study area

The state of Acre is located in the Western Amazon Basin of Brazil, bordering with Peru, Bolivia, and the Brazilian states of Amazonas and Rondônia (Figure 1). The study site, Ramal do Granada (9°41′ S–9°49′ S, 67°05′ W–67°07′ W), was formerly a sparsely populated rubber tapper settlement in the eastern corner of Acre that became part of the Pedro Peixoto Agricultural Settlement Project (Figure 1) in 1982. The area is characterized by a humid equatorial climate and receives most rainfall (annual average, 2,198.5 mm) between December and March. The mean annual temperature is 24.5°C. Most inhabitants are migrants from southeast and south Brazil. Subsistence agriculture and cattle raising are currently the main economic activities, with coffee, banana, and rice as the main cash crops.

Baseline seroprevalence study

Recruitment strategies have been described elsewhere.11 Briefly, all households, enumerated during a census performed by our field team in Ramal do Granada, were visited between March and April 2004, and 466 dwellers < 1–90 years of age (98.5% of the 473 area’s permanent residents) were enrolled. An additional 43 individuals (mostly newcomers to the area) were enrolled between September and October 2004. The 425 study participants older than 5 years of age who were enrolled either in March-April or September-October 2004 were invited to contribute a 5-mL venous blood sample for serum separation. Of these, 405 subjects (95.3% of the eligible; age range, 5–90 years), living in 122 households, had their sera tested for DENV IgG antibodies and constituted the population sample analyzed in the baseline seroprevalence survey. The location of all households was determined using a hand-held, 12-channel global positioning system receiver (eTrex Personal Navigator; Garmin, Olathe, KS), which gives a positional accuracy within 15 m. A baseline questionnaire was applied to study participants to obtain demographic, clinical, and socioeconomic information. Data on housing conditions, land tenure, past dengue episodes, yellow fever vaccination status, use of bed nets, and ownership of 13 household assets were recorded.

Information on household assets was used to derive a wealth index, as described by Filmer and Pritchett.12 Principal component analysis was used to define the household asset weights. The first principal component explained 36.0% of the variability and gave the greatest weights to ownership of a refrigerator (0.831), a television set (0.773), and a parabolic antenna (0.765). Principal component analysis was carried out using the XLSTAT software, version 7.5.2 (Addinsoft, New York, NY). After standardization of these weighted asset variables,12 the highest scores were given to the ownership of a videotape player/recorder (1.310), a sofa set (0.770), and a motor vehicle (0.760). The lowest scores were assigned to households without a refrigerator (−1.300), a gas stove (−1.020), or a blender (−0.960). The asset scores were summed to a wealth index for each household (range, −8.850 to 7.520).

Seroconversion study

All households were revisited in February–March 2005. Of 405 subjects enrolled at baseline, 380 (93.8%) still lived in the area and were eligible for a second blood sample draw. The 310 study participants with paired serum samples tested for DENV antibodies with ELISA (76.5% of the original study population; age range, 5–79 years) comprised the population sample of the seroconversion study.

Surveillance of acute febrile illnesses

The clinical and laboratory surveillance of acute febrile illnesses started in Ramal do Granada in March 2004. Because both Plasmodium falciparum and P. vivax are locally endemic,11 all febrile patients were screened for malaria parasites by standardized thick-smear microscopy.13 Further laboratory investigations included detection of IgM antibodies to DENV and West Nile virus (WNV), amplification of DENV RNA by reverse transcription-polymerase chain reaction (RT-PCR), and virus isolation using the C6/36 cell system, in acute-phase serum or plasma samples.

Because the number of cases of dengue fever reported in the state of Acre increased dramatically throughout 2004 (Figure 2), we extended the laboratory surveillance of episodes of febrile illness of non-malarial origin to also include neighboring rural communities and the nearest town, Acrelândia. To recruit these additional subjects, we made periodic visits to malaria diagnosis outposts and enrolled febrile patients with negative microscopy for malaria parasites. Between March 2004 and October 2006, we studied 102 febrile episodes occurring in 90 subjects 6–60 years of age (mean, 28.1 years); 37 (41.1%) subjects lived in Ramal do Granada, 30 (33.3%) inhabited neighboring rural areas, and 23 (25.5%) lived in the town of Acrelandia. The samples for RT-PCR and virus isolation were stored in liquid nitrogen in the field and later shipped on dry ice; those for serologic analysis were stored at −20°C.

Antibody detection by ELISA

The serum samples collected at the study baseline and the follow-up visit were tested for DENV IgG antibodies by enzyme-linked immunosorbent assay (ELISA) at a 1:100 dilution. Antigens were prepared with Ae. albopictus C6/36 cells infected with DENV Types 1, 2, 3, and 4 and disrupted by sonication; uninfected C6/36 cells were processed in the same way for use as control antigens. Polysorb 96-well microplates (Nalge Nunc International, Rochester, NY) were coated with antigens diluted in phosphate-buffered saline (PBS) at 4°C overnight and subsequently blocked with a solution containing 5% skim milk in PBS for 1 hour at 37°C. To minimize background reactivity, sera were diluted in a blocking solution containing 10% (wt/vol) of an extract of uninfected C6/36 cells. After a 1-hour incubation, serum samples were removed, and horseradish peroxidase–conjugated goat anti-human IgG (Dako North America, Carpinteria, CA) was added (30 minutes at 37°C), followed by the addition of 3, 3′, 5′ tetramethyl benzidine chromogen solution (DADE Behring, Marburg, Germany). Net absorbance values were calculated by subtracting the absorbance readings at 450 nm of control antigen wells from those of DENV antigen wells. A cut-off absorbance value was defined as the mean net absorbance reading for 30 negative control sera plus 3 SD. Antibody levels were expressed in ELISA units, interpolated from a standard curve prepared with an in-house reference serum, as described elsewhere.5 Briefly, a serum sample with known levels of anti-DENV IgG antibodies14 was serially diluted and assayed on each of the microplates on which test sera were analyzed. Using log-linear regression analysis, net absorbance values of test sera, at 1:100 dilution, were compared with those of the standard curve to give antibody levels in arbitrary ELISA units. The cut-off absorbance value was set at 100 ELISA units; greater values were considered positive.

Paired serum samples were tested side-by-side on the same microplate. When the comparison of paired samples showed seroconversion or an increase > 4-fold in ELISA units, we determined the avidity of specific IgG antibodies by incorporating a 10-minute incubation with 7 M urea14 into the original test protocol of the DENV IgG DxSelect kit (Focus Diagnostics, Cypress, CA). The avidity index was calculated as the percentage of decrease in absorbance readings in urea-coated wells compared with non-urea controls.14 All ELISA-based IgG antibody assays were performed at the Laboratory of Virology of the Institute of Tropical Medicine of São Paulo.

Ninety-one acute-phase serum samples were tested for IgM antibodies using the DENV IgM Capture DxSelect kit (Focus Diagnostics), following the manufacturer’s instructions. Although WNV has never been found to infect humans in Brazil, we also tested 88 acute-phase sera for specific IgM antibodies with the West Nile Virus IgM Capture DxSelect ELISA kit (Focus Diagnostics). A single WNV-positive result was obtained with an acute-phase serum collected in August 2005 that was negative for DENV IgM. Using a IgG DxSelect ELISA kit (Focus Diagnostics), however, we found quite similar levels of WNV IgG in additional samples collected from the same subject before (April 2004 and February 2005) and after (October 2006) the febrile illness, thus arguing against the hypothesis of a primary exposure to WNV. The first sample (April 2004) had no antibodies to yellow fever virus (vaccinal 17D strain) or to DENV Types 1, 2, 3, and 4 detectable by hemagglutination inhibition assay (HIA).11

Antibody detection by plaque reduction neutralization tests

When recent exposure to DENV was suspected based on ELISA results (seroconversion or an increase > 4-fold in ELISA units in paired serum samples), the paired serum samples were further tested using neutralization assays for DENV Types 1, 2, and 3 and also for yellow fever virus (YFV).15 Plaque reduction neutralization tests (PRNTs) for antibodies to DENV were performed in 24-well tissue culture plates with serial 2-fold dilutions of inactivated serum samples (final volume, 50 μL), starting at either 1:10 (DENV-3) or 1:11 (DENV-1 and -2) dilution. A 150-μL virus suspension with 30 plaque-forming units (PFU)/well was incubated with diluted test sera for 1 hour at 37°C in 5% CO2; the negative control consisted of medium without serum. Previously prepared monolayers of Vero cells (0.2 × 105 cells/mL) were inoculated with 200 μL of each virus-serum mixture. Both virus and serum samples were diluted in 199 medium containing Earle salts, 5% fetal calf serum, 0.22% sodium bicarbonate, and antibiotics. After a 1-hour incubation at 37°C, the supernatant of each well was discarded and replaced with medium containing 3% carboxymethyl cellulose (final volume, 3 mL). The cultures were incubated for 7 days at 37°C in 5% CO2. The monolayers were fixed with formalin and stained with crystal violet, and plaques were counted. The serum dilution that reduces the plaque numbers by 50%, relative to the virus control, was determined by log-linear regression; neutralizing antibody titers were expressed as the reciprocal serum dilution giving 50% plaque reduction. PRNTs for antibodies to YFV were carried out in 96-well tissue culture plates with serial 2-fold dilutions of inactivated serum samples, starting at 1:5 dilution (final volume, 50 μL).15 YFV (25 PFU) in 50 μL was dispensed into wells; dilutions of both virus and serum samples were performed in 199 medium containing 2.5% 1 mol/L HEPES. A positive control (monkey serum with yellow fever antibody concentration calibrated against a WHO International Reference Preparation) was included in each test. After incubation for 1 hour at room temperature, 50 μL of a Vero cell suspension in 199 medium (1.6 × 105 cells/well) was added to the wells, with a further incubation for 3 hours at 37°C. The medium was discarded and replaced with 199 medium containing 3% carboxymethyl cellulose (final volume, 100 μL). After incubation for 7 days at 37°C in 5% CO2, the monolayers were fixed with formalin and stained with crystal violet, and the plaques were counted. Log-linear regression analysis was used to estimate the serum dilution leading to a 50% reduction in plaque numbers relative to the virus control. Antibody levels were expressed in mIU/mL, using the reference serum preparation in a calibration curve.16 All PRNTs were performed at the Oswaldo Cruz Foundation, Rio de Janeiro, Brazil.

RNA extraction, reverse transcriptase-polymerase chain reaction, and complementary DNA sequencing

Molecular diagnosis of DENV was carried out independently, in two laboratories, on 69 acute-phase serum or plasma samples. At the Laboratory of Molecular Biology of Marília Medical School, viral RNA was isolated from 300-μL aliquots of serum or plasma as described,17 and reverse transcriptase-polymerase chain reaction (RT-PCR) for DENV was performed with the method proposed by Lanciotti and others.18 At the Laboratory of Virology of the Faculty of Medicine of São José do Rio Preto, samples were analyzed with novel multiplex-nested-PCR (M-N-PCR) and nested-PCR (N-PCR) assays that detect several flaviviruses (DENV-1 to -4, YFV, WNV, St. Louis encephalitis, Rocio, Bussuquara, and Iguape) and alphaviruses.19 Briefly, RNA was extracted from 140-μL aliquots of serum or plasma using the QiAampViral RNA Mini Kit (Qiagen, Hilden, Germany), and the first RT-PCR was performed using generic oligonucleotide primers for flaviviruses and alphaviruses. After the second PCR, with the species-specific primers described elsewhere,19 the amplicons were loaded onto 1% agarose gels and visualized under UV illumination. Standard precautions to avoid contamination were followed, and both positive and negative controls were used in all reactions. Samples were coded and examined in a blinded fashion at both laboratories; results obtained independently by each laboratory were only compared at the end of the study.

DENV amplicons corresponding to a fragment of the NS5 gene were purified and sequenced using BigDye v3.1 terminator chemistry (Applied Biosystems, Foster City, CA) on an ABI377 automatic DNA sequencer (Applied Biosystems). Nucleotide sequences were analyzed using the DS Gene 2.0 software (Accelrys, San Diego, CA) and deposited into the GenBank database (accession numbers: EU672811–EU672815). The partial NS5 gene nucleotide sequences were aligned with homologous sequences from 136 DENV-3 isolates collected worldwide to build a linearized neighbor-joining phylogeny using MEGA 4.0 software.20 Bootstrap support values were obtained with 1,000 pseudoreplicates. The Appendix available online provides the complete list of sequences (with GenBank accession numbers, country, and date of isolate collection), the complete sequence alignment, and the neighbor-joining phylogeny based on 407 bp of the NS5 gene sequence in a total of 141 DENV-3 isolates. Partial NS5 gene sequences have been previously used in phylogenetic analyses of DENV-3 isolates from South America, with results consistent with those based on more extensive DNA sequencing.21,22

Dengue virus isolation

Virus isolation was performed for 59 acute-phase samples by inoculating 30 μL of clinical specimens onto confluent monolayers of Ae. albopictus C6/36 cells in 25-cm2 tissue culture flasks. The virus isolates were typed by indirect fluorescent antibody test with serotype-specific monoclonal DENV antibodies.23 Virus isolation and identification were performed, following identical protocols, at the Laboratory of Arbovirology and Hemorrhagic Fevers, Evandro Chagas Institute, Belém, and at the Laboratory of Molecular Biology of Marília Medical School, Marília.

Definitions

We used two combinations of ELISA results to select samples with a possible evidence of DENV infection during the follow-up (seroconversion study): 1) the first (baseline) sample was IgG-negative but the second sample (February–March 2005) was IgG-positive, irrespective of the levels of specific antibodies, or 2) DENV IgG levels (measured in ELISA units) increased by > 4-fold when comparing paired serum samples. To confirm that these changes in ELISA antibody units resulted from recent exposure to DENV, the subset of paired samples selected as described above was further examined with PRNTs for DENV (Types 1, 2, and 3) and YFV. The final criteria for defining recent exposure to DENV in this subset of samples were 1) the baseline sample was negative but the second sample was PRNT positive for at least one DENV type, or 2) there was a > 3-fold increase in PRNT antibody titers to at least one DENV type in paired samples. For convenience, both situations will be termed “seroconversion” throughout the article. Acute DENV infection was defined as 1) DENV isolation in acute-phase serum or plasma, 2) amplification of DENV RNA by RT-PCR (with either protocol) in acute-phase serum, 3) detection of DENV IgM by ELISA in acute-phase serum, or 4) PRNT antibody seroconversion (as defined above).

Data analysis

A database was created with SPSS 13.0 (SPSS, Chicago, IL). The incidence of dengue was estimated as the number of seroconverters per 100 person-years at risk, and its exact Poisson 95% confidence interval (CI) was calculated, with time at risk defined as time interval between blood draws. Multiple logistic regression models with stepwise backward deletion were built to describe independent associations between potential risk factors (independent variables) and two outcomes: 1) positive DENV serology at the study baseline and 2) seroconversion (defined by PRNT results) during the study. Variables associated with P < 0.20 in unadjusted analysis were included into the logistic regression models. Because the data have a nested structure, where individuals are nested within households, the assumption of independence of observations underlying standard logistic regression analysis was violated. We therefore used multilevel logistic models with individual and household-level risk factors.24 The HML software package (version 6.03; Scientific Software International, Lincolnwood, IL) was used for multilevel analysis. Only variables associated with statistical significance at the 5% level were maintained in the final models.

The Kulldorff spatial scan statistics was used to test whether DENV infections were randomly distributed within the study area and to identify significant spatial clusters, if present.25 Analysis was made using the Bernoulli model implemented in version 5.1 of the SaTScan software (available at http://www.satscan.org), which creates and moves circular windows systematically throughout the geographic space to identify significant clusters of infections. The windows are centered on each household; the largest possible cluster was set to encompass 50% of the households. For each location and size of the scanning window, SaTScan performs a likelihood ratio test to evaluate whether infections are more prevalent within that specific circular window compared with the outside. Separate analyses were made for 1) DENV IgG seropositivity at the study baseline and 2) PRNT-confirmed seroconversion during the study. P values were determined by 10,000 Monte Carlo replications of the data set, and a level of significance of 5% was adopted.

Ethical considerations

Approval of the study protocol was obtained from the Ethical Review Board of the Institute of Biomedical Sciences of the University of São Paulo, Brazil (538/2004). Written informed consent was obtained from all study participants or their parents/guardians.

RESULTS

Prevalence of antibodies at baseline and associated risk factors

DENV IgG antibodies were detected in 74 subjects 5–77 years of age (mean, 33.8 years) examined at the study baseline (seroprevalence rate, 18.3%; 95% CI, 14.6–22.4%). Both individual (age, sex, and migration history) and household-level (wealth, land tenure, and housing conditions) characteristics were significantly associated with the presence of DENV IgG antibodies in unadjusted analysis (Table 1). Many household characteristics that might facilitate the transmission of mosquito-borne infections were fairly homogeneous. For example, all houses had windows that could be closed, but none of them had screens. Cross-reactivity among flaviviruses could affect DENV antibody measurements in populations exposed to, or immunized against, yellow fever. In fact, subjects with antibodies to the vaccinal strain of YFV (17DD) used in Brazil, detected by HIA,26 were considerably more likely to have DENV IgG antibodies detected by ELISA (OR = 18.74; 95% CI, 8.89–39.80; P < 0.0001; 355 sera analyzed; data from Silva-Nunes and others11). However, although 87.2% of the study subjects reported yellow fever vaccination over the past 10 years, DENV IgG antibodies detected by ELISA were similarly prevalent in recently vaccinated and non-vaccinated subjects (Table 1). This finding suggests that recent yellow fever vaccination has not substantially affected DENV antibody measurements in our population. The same conclusion was reached in a recent survey comparing HIA antibody titers to DENV and 17D yellow fever virus before and after the mass vaccination campaign against yellow fever carried out in Acre in 2000; 3 months after vaccination, 92.1% of the vaccinees had HIA antibodies to 17D yellow fever virus (compared with 15.5% before vaccination), but only 2.1% of them seroconverted to DENV.27 However, a more conclusive analysis of a possible interference of preexisting antibodies to YFV with DENV serodiagnosis would require the use of more specific neutralization assays.

Multilevel logistic regression analysis showed that male sex, low wealth index, and history of migration from extra-Amazonian states were significant independent predictors of the presence of DENV IgG antibodies at baseline (Table 2), whereas self-reported past DENV infection was a predictor of borderline statistical significance. The association between age and DENV seropositivity was no longer significant after controlling for migration history and other covariates. The effect of poor housing conditions could not be properly assessed in multiple logistic regression analysis, because of the zero seroprevalence in unexposed individuals (Table 1).

DENV infection during the follow-up and associated risk factors

Of the 310 subjects who contributed paired serum samples, four had > 4-fold increases in DENV IgG levels and 16 had a seroconversion detected by ELISA. DENV IgG antibody avidity was assessed for 16 of these subjects, and only 2 had antibodies of low avidity (< 30%) in the second sample, which are suggestive of a very recent primary DENV infection.14 Avidity could not be assessed properly in four samples with low levels of DENV IgG antibodies detected by ELISA. To confirm whether the observed changes in ELISA antibody units in paired samples resulted from recent exposure to DENV, the samples were further tested by PRNT. The comparison of paired titers of antibodies to DENV and YFV (the only other flavivirus known to circulate in the study area) in paired samples, shown in Table 3, confirmed 10 of 20 instances of recent exposure to DENV suggested by ELISA. Three DENV seroconverters (Subjects 17, 172, and 229) also had > 3-fold increases in the PRNT titers of antibodies to YFV (Table 3), suggesting that they might have been recently exposed to either wild-type YFV or yellow fever vaccine. However, given the very little (if any) cross-reactivity observed between the antibodies to DENV and YFV in PRNT,28 we consider the large increases in DENV-specific PRNT antibody titers in these three subjects to be clearly diagnostic of recent exposure to the virus. The PRNT-confirmed seroconversion rate was therefore estimated as 3.67 episodes/100 person-years at risk (95% CI, 2.24–5.67 episodes/100 person-years). Two findings suggest that the clinical diagnosis of dengue fever is neither sensitive nor specific in this malaria-exposed population: 1) only 4 of 24 subjects reporting a clinically diagnosed dengue fever episode during the follow-up actually seroconverted and 2) most seroconverters (6 of 10) reported no dengue fever episode diagnosed on clinical grounds during the follow-up.

The small number of confirmed seroconversion events reduces the statistical power of risk factor analysis, but a self-reported history of clinical diagnosis of dengue fever and of travel to Rio Branco (the capital of Acre) during the follow-up emerged as strong predictors of seroconversion in unadjusted analysis (data not shown). Participants in the seroconversion study (N = 310) were older (mean age, 28.3 versus 23.5 years; P = 0.021, Mann-Whitney test) and wealthier (mean wealth index, 0.26 versus −3.52; P < 0.001, Mann-Whitney test) than the subjects who failed to provide a second serum sample (N = 95). However, univariate analysis showed no significant difference in predictors of seroconversion between the two groups. Multilevel logistic regression analysis showed male sex, in addition to the predictors shown by unadjusted analysis, to be significantly associated with seroconversion during our study (Table 4).

Spatial analysis

The Kulldorf spatial scan statistic showed no significant spatial clustering of households inhabited by the 74 subjects with DENV IgG antibodies detected at baseline. A similar analysis involving the 10 documented episodes of seroconversion confirmed by PRNT also failed to show any significant spatial clustering.

Laboratory investigation of acute febrile illnesses

Only 11 (10.8%) of 102 non-malarial febrile episodes studied between 2004 and 2006 had any laboratory evidence of acute DENV infection: 1) RT-PCR and virus isolation were both positive for DENV-3 in two episodes, 2) RT-PCR alone was positive for DENV-3 in three episodes; 3) DENV IgM together with PRNT seroconversion were detected in two episodes, 4) DENV IgM was detected in acute-phase sera from three episodes for which seroconversion analysis could not be made, and 5) PRNT seroconversion was detected in one subject with IgM-negative acute-phase sera. Sequential serum samples for seroconversion analysis were available for only 28 subjects who contributed acute-phase serum.

Both RT-PCR assays were positive for DENV-3 in two samples, whereas only the method described by Bronzoni and others19 was able to detect DENV-3 RNA in three samples. No acute-phase sample was positive for YFV or any other flavivirus tested, arguing against a major impact of antibodies elicited by exposure to other flaviviruses on the observed serologic patterns.

We noticed several discrepancies when comparing different laboratory methods. For example, three IgM-positive acute-phase sera yielded negative RT-PCR, and one was also DENV isolation negative; the remaining IgM-positive samples were not tested by either method. In addition, the acute-phase serum from one seroconverter tested negative for IgM antibodies and two acute-phase sera from seroconverters were negative by both RT-PCR and virus isolation. Both acute-phase samples with DENV detected by virus isolation were IgM negative. These inconsistencies are expected in field studies, because the timing of acute-phase sample draws may be appropriate for some but not all diagnostic methods used. For instance, RT-PCR and cell culture isolation are more effective during early infection, whereas MAC-ELISA becomes more sensitive later on. In this study, the low viral loads at the time of blood collection may have reduced the diagnostic sensitivity of virus isolation and RT-PCR in some subjects, whereas high levels of preexisting antibodies to DENV may have impaired the detection of seroconversion events in other subjects.

The acute-phase samples from which DENV-3 was isolated were collected from two residents in the town of Acrelândia presenting with a clinical diagnosis of dengue fever in November 2004 and February 2006, respectively. In both samples, virus identification was confirmed by sequencing 407 bp of the NS5 gene fragment amplified as described by Bronzoni and others.19 We also sequenced the DENV-3–specific RT-PCR products amplified from three acute-phase samples (derived from two residents in the town of Acrelândia and one resident in the rural study site) that were negative for virus isolation (collected in December 2004, February 2005, and February 2006). The five partial NS5 gene sequences from Acre were aligned with 136 GenBank-available homologous sequences from DENV-3 isolates collected worldwide. The neighbor-joining phylogeny built with these sequence data grouped all isolates from Acre together with seven DENV-3 isolates from Brazil and six from Martinique in a clade with 85% bootstrap support (Appendix available online at www.ajtmh.org). These data give further support further support to the claim that most DENV-3 isolates circulating in Brazil have Caribbean origin.29

The DENV-3 sample found in November 2004 corresponds the first isolation of this DENV type in Acre, although DENV-3 had already been found to circulate in the neighboring states of Amazonas and Rondônia since 2002.30 Because Acrelândia is located close to the only highway (BR-364) connecting Acre to the rest of the country (Figure 1), this town is the most likely port of entry of DENV-3 into Acre, leading to the 2004 outbreak that mostly affected Rio Branco, situated about 120 km west of this town (Figure 2). Unfortunately, no further DENV-3 isolates from the outbreak are available to sequence analysis to confirm the putative route of entry of DENV-3.

DISCUSSION

The first population-based study of DENV infection in a rural Amazonian population of Brazil showed a low baseline DENV seropositivity rate (18.3%) with a sizable seroconversion rate (3.67 episodes/100 person-years at risk) over the next 12 months. These findings reflect the relatively recent introduction of DENV into Acre, with the occurrence of a dengue outbreak during the study period (Figure 2). Acre was the last state in the Amazon Basin of Brazil to report autochthonous infections with DENV-1 and -2.30 Not surprisingly, the overall DENV seroprevalence in the city of Rio Branco was estimated to be only 4.4% in August 1999, with only antibodies to DENV-1 and -2 detected by HIA.27

The incidence rate of DENV infection during the follow-up is lower than that recently estimated for areas with stable transmission in Southeast Asia, such as northern Thailand31 (8.5 episodes episodes/100 person-years) and southern Vietnam32 (11.7 episodes/100 person-years), and in the Americas, such as Nicaragua33 (6.0–12.0 episodes/100 person-years). Because of the long period between blood draws, we may have missed some DENV infections occurring soon after the baseline survey, thus underestimating the incidence rate. Sequential HIAs performed at 6-month intervals detected 97% of the incident asymptomatic DENV infections in Thailand,31,34 but no comparable analyses are available for ELISA- or PRNT-based prospective studies. Because 89.2% of DENV infections recorded in Acre in 2004 occurred after September (Ministry of Health of Brazil, unpublished data available at http://portal.saude.gov.br/portal/svs/area.cfm?id_area=451), the vast majority of incident infections in our study population probably occurred up to 6 months before the second sample draw in February–March 2005.

Several lines of evidence suggest that inhabitants of Ramal do Granada have acquired most DENV infections in other sites. A history of migration from extra-Amazonian states, where DENV has been circulating for nearly two decades,2 was significantly associated with seropositivity at baseline (Tables 1 and 2), whereas a history of travel to Rio Branco, where a major dengue fever outbreak occurred in 2004 (Figure 2), predicted seroconversion during follow-up (Table 4). Although 2 of 10 seroconverters reported no travel to Rio Branco during the follow-up, short visits to the town of Acrelândia (30–45 km away) and other urban centers could not be ruled out. Significantly, four of five DENV-3 infections confirmed by RT-PCR during our study were diagnosed in residents in the town of Acrelândia, confirming that this virus circulated in this urban area. The lack of significant spatial clustering of DENV infections in Ramal do Granada further supports the hypothesis of little autochthonous transmission in the rural settlement. No larvae or adults of known DENV vector species have been found in Ramal do Granada over the past 5 years, but sampling may have been biased because most of the local entomologic research focuses on malaria vectors.11 As pointed out in a similar study carried out in the Peruvian Amazonia, vector control is inappropriate for DENV control in rural areas where little autochthonous transmission occurs.10 Personal protection measures when visiting high-risk urban centers might be more effective to prevent DENV infection in this and other similar rural populations. Learning the reasons why inhabitants in rural areas often visit urban centers may also aid in designing more effective strategies to minimize the risk of DENV introduction into their communities.

Significant sex differences in DENV infection rates have been described in several studies.35 Hospital-based studies in Asia have suggested that infections are more frequent in men, but these data may simply reflect sex-related differences in healthcare-seeking behavior.35 In contrast, the only population-based study comparing DENV infection rates according to sex in the Americas was carried out in Mexico and found an increased risk among women.36 The risk of past DENV infection at baseline and that of subsequent infections during the follow-up remained significantly higher among men, in our study, after controlling for migration patterns and travel history, suggesting that sex-related differences in exposure are unlikely to account for these findings. The biological bases for male-female differences in DENV infection rates remain undetermined. Sex differences in immune responses elicited by DENV in men and women have been put forth as an explanation for male predominate among patients with mild disease, whereas women predominate in more severe cases in Southeast Asia.37 However, this hypothesis has yet to be further explored.

The association between low socioeconomic status and DENV infection rates has been described in both Southeast Asia and the Americas,35 although discordant results have been reported in urban Brazil.38 Because baseline DENV seropositivity remained significantly associated with poverty after controlling for migration history (Table 2), poverty-related differences in migration patterns are unlikely to account for this finding in our population. If most DENV infections are not locally acquired, differences in housing conditions32,35 are unlikely to account for this association, either. Poverty, however, did not predict the risk of seroconversion during the follow-up.

Although standardized diagnostic criteria39 are widely used, the clinical diagnosis of DENV infection in Brazil remains notoriously inaccurate,4042 because several locally prevalent febrile illnesses may be misdiagnosed as dengue fever. Here we show that 1) dengue accounted for only 11 of 102 (10.8%) febrile episodes of non-malarial origin occurring in patients from rural and urban areas of eastern Acre, 2) even though a self-reported history of dengue diagnosis was a strong predictor of seroconversion during the follow-up (Table 4), only 4 of 24 subjects self-reporting a dengue fever episode during the follow-up actually seroconverted, and 3) most (6 of 10) of those with PRNT-confirmed DENV seroconversion reported no clinically diagnosed dengue fever episode during the follow-up. Most DENV infections in these subjects may have been asymptomatic. These findings underscore the need for laboratory confirmation of DENV infections for outbreak investigation and disease surveillance.43 Accordingly, our laboratory surveillance showed the circulation of DENV-3 in the urban area of Acrelândia since November 2004, suggesting that the introduction of this serotype led to the dengue fever outbreak recorded in Acre in the second semester of 2004. Nine cases of dengue hemorrhagic fever, four leading to death, were laboratory-confirmed in Acre in 2004 (Ministry of Health of Brazil, unpublished data available at http://portal.saude.gov.br/portal/svs/area.cfm?id_area=451). If the patterns of DENV circulation in the most likely ports of entry to Acre (including Acrelândia) were known earlier, classic preventive measures, such as vector control, could have been timely implemented in the urban areas that were mostly affected, thus reducing the morbidity and mortality associated with the outbreak and minimizing the risk of infection among nearby rural populations who often visit urban areas.

Table 1

Baseline prevalence of IgG antibodies to dengue virus, according to individual and household-level risk factors, Ramal do Granada, Brazil, March–April 2004

VariableNo. of subjects*Prevalence of IgG antibodiesOdds ratio (95% CI)P
* Number of individuals differ for some variables, because of missing values.
P values of χ2 tests for linear trend; all other P values are for standard χ2 or Fisher exact tests.
Age (years)
    5–141238.1%1.000.0004†
    15–3013017.7%2.43 (1.04–5.76)
    31–6012825.8%3.93 (1.75–9.02)
    > 602433.3%5.65 (1.72–18.65)
Sex
    Male21224.5%2.53 (1.42–4.51)0.001
    Female19311.4%1.00
Time of residence in Amazonia (quartiles)
    1 (shortest)9418.1%1.000.093†
    210014.0%0.74 (0.32–1.70)
    311113.5%0.71 (0.31–1.62)
    4 (longest)10028.0%1.76 (0.85–3.73)
Migrant from extra-Amazonian states
    Yes24721.9%1.93 (1.08–3.56)0.027
    No15812.7%1.00
Wealth index (quartiles)
    1 (poorest)9330.1%4.05 (1.74–9.63)0.0005†
    210517.1%1.94 (0.80–4.82)
    310317.5%1.99 (0.81–4.93)
    4 (least poor)1049.1%1.00
Land tenure
    Yes36416.8%1.000.033
    No4131.7%2.31 (1.03–4.90)
House walls
    Wood38619.2%Undefined0.032
    Brick190.0%1.00
Wall gaps
    Yes35118.5Undefined0.014
    No330.0%1.00
Water source
    Well38318.3%1.001.000
    River2218.2%0.99 (0.24–3.15)
Bed net use
    Never27517.4%0.93 (0.39–2.46)0.975
    Occasionally5416.7%0.88 (0.27–2.91)0.984
    Always4318.6%1.00
Self-reported past diagnosis of dengue
    Yes1136.4%2.64 (0.55–10.71)0.112
    No39417.8%1.00
Yellow fever vaccination in the past 10 years
    Yes35318.1%0.93 (0.43–2.19)1.000
    No5219.3%1.00
Table 2

Results of the final multilevel logistic regression model including variables significantly associated with baseline presence of IgG antibodies to dengue virus, Ramal do Granada, Brazil, March–April 2004

VariableOdds ratio95% CIP
Gender (male vs. female)2.601.55–4.340.001
Wealth (continuous variable)0.880.83–0.950.001
Past dengue episode (yes vs. no)2.960.99–8.840.051
Migrant (yes vs. no)2.261.18–4.330.014
Table 3

Levels of neutralizing antibodies to DENV Types 1, 2, and 3 and YFV, measured with PRNT, in paired serum samples collected at intervals of 6–12 months from 20 subjects with suspected seroconversion to DENV based on ELISA results, Ramal do Granada, Brazil, March 2004–March 2005

SubjectSampleDENV-1 (titer)DENV-2 (titer)DENV-3 (titer)YFV (mU/mL)
* Subjects with seroconversion to DENV confirmed by PRNT.
10*Baseline< 11138910,275
Follow-up178620> 1,0008,010
13Baseline531,019832,661
Follow-up55> 1,100936,178
14Baseline< 11< 1146> 16,125
Follow-up22< 118812,579
17*Baseline< 11< 11< 10940
Follow-up95897501,996
53*Baseline< 11< 11< 10< 126
Follow-up104401> 1,000214
84Baseline< 11< 11< 10836
Follow-up< 11< 11< 10746
96Baseline< 1112487,580
Follow-up< 11295513,950
130*Baseline< 11< 11< 106,290
Follow-up62681936,290
172*Baseline< 1139291,798
Follow-up7046828797,442
210Baseline< 11< 11< 11> 16,125
Follow-up< 11< 11< 1012,838
229*Baseline< 1152771,622
Follow-up459580> 1,0006,975
275Baseline< 11< 11< 101,628
Follow-up< 11< 11301,716
293Baseline< 11< 11< 104,774
Follow-up< 11< 11< 103,008
303Baseline< 11< 11< 10751
Follow-up< 1114< 10251
341Baseline404791995
Follow-up3685776,975
343Baseline< 11< 11< 10642
Follow-up< 112841684
357*Baseline< 1115< 101,240
Follow-up1076258082,000
411*Baseline6173253,118
Follow-up> 1,1008297503,179
446*Baseline18655515,480
Follow-up70226787,338
477*Baseline11102< 105,835
Follow-up> 1,100> 1,100> 1,1009,300
Table 4

Results of the final multilevel logistic regression model including variables significantly associated with seroconversion to DENV, confirmed by PRNT, after 6–12 months of follow-up, Ramal do Granada, Brazil, March 2004–March 2005

VariableOdds ratio95% CIP
Sex (male vs. female)4.161.17–14.850.028
Clinical dengue episode (yes vs. no)10.872.6–43.850.001
Travel to Rio Branco (yes vs. no)11.623.39–39.92< 0.0001
Figure 1.
Figure 1.

Map of the state of Acre, northwestern Brazil, showing the study site. Ramal do Granada is part of the Pedro Peixoto Agricultural Settlement (shaded area in the inset), located 30–45 km northwest of the town of Acrelândia. The location of BR-364, the only paved highway connecting the capital of Acre (Rio Branco) to the rest of the country, is also indicated.

Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 79, 4; 10.4269/ajtmh.2008.79.485

Figure 2.
Figure 2.

Number of cases of dengue fever recorded in the state of Acre (population, 669,737) between 1999 and 2006. Only three dengue fever episodes were reported in 1999. During the dengue outbreak of 2004, 5,892 cases (84.1% of the 7008 cases in Acre) were diagnosed in residents of the state capital, Rio Branco (population, 314,127). The vast majority of dengue fever episodes recorded in Acre during 2004 (6,250 of 7,008 cases or 89.2%) occurred between October and December (unpublished data of the Ministry of Health of Brazil; available at http://portal.saude.gov.br/portal/svs/area.cfm?id_area=451).

Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 79, 4; 10.4269/ajtmh.2008.79.485

*

Address correspondence to Mônica da Silva-Nunes, Department of Parasitology, Institute of Biomedical Sciences, University of São Paulo, Av. Prof. Lineu Prestes 1374, Cidade Universitária, 05508-900 São Paulo (SP), Brazil. E-mail: msnunes1@yahoo.com.br

Authors’ addresses: Mônica da Silva-Nunes, Natal Santos da Silva, and Marcelo U. Ferreira, Department of Parasitology, Institute of Biomedical Sciences, University of São Paulo, Av. Prof. Lineu Prestes 1374, 05508-900 São Paulo (SP), Brazil, Tel: 55-11-30917746, Fax: 55-11-30917417, E-mails: msnunes1@yahoo.com.br, natalss@gmail.com, and muferrei@usp.br. Vanda A. F. de Souza and Cláudio S. Pannuti, Laboratory of Virology, Institute of Tropical Medicine of São Paulo, Av. Dr. Enéas de Carvalho Aguiar 470, Cerqueira César, 05403-000 São Paulo (SP), Brazil, Tel: 55-11-30622645, Fax: 55-11-30667012, E-mails: vaueda@usp.br and cpannuti@usp.br. Márcia A. Sperança, Laboratory of Molecular Biology, Marília Medical School, Av Monte Carmelo, 650, Fragata, 17519-030 Marilia (SP), Brazil, Tel: 55-14-34331235, Fax: 55-14-34330148, E-mail: speranca@famema.br. Ana Carolina B. Terzian and Maurício L. Nogueira, Laboratory of Virology, Faculty of Medicine of São José do Rio Preto, Av Briga-deiro Faria Lima, 5416, 15090-000 São José do Rio Preto (SP), Brazil, Tel: 55-17-2105872, E-mails: anacarolinaterzian@gmail.com and mnogueira@famerp.br. Anna M. Y. Yamamura and Marcos S. Freire, Institute of Technology in Immunobiologicals, Oswaldo Cruz Foundation, Av: Brasil 4365, Manguinhos, 21040-900 Rio de Janeiro (RJ), Brazil, Tel: 55-21-38829317 ext. 9317, Fax: 55-21-22604727, E-mails: anna@bio.fiocruz.br and freire@bio.fiocruz.br. Rosely S. Malafronte, Laboratory of Protozoology, Institute of Tropical Medicine of São Paulo, Av. Dr. Enéas de Carvalho Aguiar 470, Cerqueira César, 05403-000 São Paulo (SP), Brazil, Tel: 55-11-30617017, Fax: 55-11-30885237, E-mail: rmalafronte@usp.br. Pascoal T. Muniz, Department of Health Sciences, Federal University of Acre, BR-364 km 4, Campus Universitário, 69915-900 Rio Branco (AC), Brazil, Tel: 55-68-39012648, Fax: 55-68-3901-2648, E-mail: pascoal@ufac.br. Helena B. Vasconcelos, Eliana V. P. da Silva, and Pedro F. C. Vasconcelos, Evandro Chagas Institute, Av. Almirante Barroso 492, 66093-020 Belém (PA), Brazil, Tel: 55-91-2114409, Fax: 55-91-2265262, E-mail: pedrovasconcelos@iec.pa.gov.br.

Note: Supplemental material (Appendix) appears online at www.ajtmh.org.

Acknowledgments: The authors thank the inhabitants of Ramal do Granada for enthusiastic participation in the study; Sebastião Bocalom Rodrigues (Mayor of Acrelândia), Damaris de Oliveira, and Nésio M. Carvalho (Municipal Government of Acrelândia) for logistic support; Adamílson L. de Souza, Camila Juncansen, Carlos E. Cavasini, and Kézia K. G. Scopel for help with fieldwork; Estéfano A. de Souza and Bruna A. Luz for data management; Cassiano P. Nunes for artwork; and Tatiana Havryliuk for reviewing the manuscript.

Financial support: This study was supported by grants from the Ministry of Health of Brazil (50148920037) and the Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP, 04/00373-2). M.d.S.-N. is supported by a PhD scholarship from FAPESP. C.S.P., N.S.d.S., P.F.C.V., and M.U.F. receive scholarships from the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Brazil. Focus Diagnostics (Cypress, CA) kindly donated ELISA kits for DENV and WNV antibody detection used in this study.

REFERENCES

  • 1

    Tauil PL, 2006. Perspectivas de controle de doenças transmitidas por vetores no Brasil. Rev Soc Bras Med Trop 39 :275–277.

  • 2

    Siqueira JB, Martelli CMT, Coelho GE, Simplício ACR, Hatch DL, 2005. Dengue and dengue hemorrhagic fever, Brazil, 1981–2002. Emerg Infect Dis 11 :48–53.

    • Search Google Scholar
    • Export Citation
  • 3

    Torres JR, Castro J, 2007. The health and economic impact of dengue in Latin America. Cad Saude Publica 23 (Suppl. 1):S23–S31.

  • 4

    Chareonsook O, Foy HM, Teeraratkul A, Silarug N, 1999. Changing epidemiology of dengue hemorrhagic fever in Thailand. Epidemiol Infect 122 :161–166.

    • Search Google Scholar
    • Export Citation
  • 5

    Strickman D, Sithiprasasna R, Kittayapong P, Innis BL, 2000. Distribution of dengue and Japanese encephalitis virus among children in rural and suburban Thai villages. Am J Trop Med Hyg 63 :27–35.

    • Search Google Scholar
    • Export Citation
  • 6

    Lian CW, Seng CM, Chai WY, 2006. Spatial, environmental and entomological risk factor analysis on a rural dengue outbreak in Lundu District in Sarawak, Malaysia. Trop Biomed 23 :85–96.

    • Search Google Scholar
    • Export Citation
  • 7

    Kumar A, Sharma SK, Padbidri VS, Thakare JP, Jain DC, Datta KK, 2001. An outbreak of dengue fever in rural areas of northern India. J Commun Dis 33 :274–281.

    • Search Google Scholar
    • Export Citation
  • 8

    Tewari SC, Thenmozhi V, Katholi CR, Manavalan R, Munirathinam A, Gajanana A, 2004. Dengue vector prevalence and virus infection in a rural area in south India. Trop Med Int Health 9 :499–507.

    • Search Google Scholar
    • Export Citation
  • 9

    Hayes CG, Phillips IA, Callahan JD, Griebenow WF, Hyams KC, Wu SJ, Watts DM, 1996. The epidemiology of dengue virus infection among urban, jungle, and rural populations in the Amazon Region of Peru. Am J Trop Med Hyg 55 :459–463.

    • Search Google Scholar
    • Export Citation
  • 10

    Reiskind MH, Baisley KJ, Calampa C, Sharp TW, Watts DM, Wilson ML, 2001. Epidemiological and ecological characteristics of past dengue virus infection in Santa Clara, Peru. Trop Med Int Health 6 :212–218.

    • Search Google Scholar
    • Export Citation
  • 11

    Silva-Nunes M, Malafronte Rdos S, Luz Bde A, Souza EA, Martins LC, Rodrigues SG, Chiang JO, Vasconcelos PF, Muniz PT, Ferreira MU, 2006. The Acre project: the epidemiology of malaria and arthropod-borne virus infections in a rural Amazonian population. Cad Saude Publica 22 :1325–1334.

    • Search Google Scholar
    • Export Citation
  • 12

    Filmer D, Pritchett LH, 2001. Estimating wealth effects without expenditure data-or tear: an application to educational enrolments in states of India. Demography 38 :115–132.

    • Search Google Scholar
    • Export Citation
  • 13

    Trape JF, 1985. Rapid evaluation of malaria parasite density and standardization of thick smear examination for epidemiological investigations. Trans R Soc Trop Med Hyg 79 :181–184.

    • Search Google Scholar
    • Export Citation
  • 14

    de Souza VAF, Fernandes S, Araújo ES, Tateno AF, Oliveira OMNPF, Oliveira RR, Pannuti CS, 2004. Use of an immunoglobulin G avidity test to discriminate between primary and secondary dengue virus infections. J Clin Microbiol 42 :1782–1784.

    • Search Google Scholar
    • Export Citation
  • 15

    Stefano I, Sato HK, Pannuti CS, Omoto TM, Mann G, Freire MS, Yamamura AM, Vasconcelos PF, Oselka GW, Weckx LW, Salgado MF, Noale LF, Souza VA, 1999. Recent immunization against measles does not interfere with the seroresponse to yellow fever vaccine. Vaccine 17 :1042–1046.

    • Search Google Scholar
    • Export Citation
  • 16

    Freire MS, Mann GF, Marchevsky RS, Yamamura AMY, Almeida LFC, Jabor AV, Malachias JMN, Coutinho ESF, Galler R, 2005. Production of yellow fever 17DD vaccine in primary culture of chicken embryo fibroblasts: yields, thermo and genetic stability, attenuation and immunogenicity. Vaccine 23 :250–252.

    • Search Google Scholar
    • Export Citation
  • 17

    Chomczynski P, Sacchi P, 1987. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162 :156–159.

    • Search Google Scholar
    • Export Citation
  • 18

    Lanciotti RS, Calisher CH, Gubler DJ, Chang GJ, Vorndam AV, 1992. Rapid detection and typing of dengue viruses from clinical samples by using reverse transcriptase-polymerase chain reaction. J Clin Microbiol 30 :545–551.

    • Search Google Scholar
    • Export Citation
  • 19

    de Morais Bronzoni RV, Baleotti FG, Ribeiro Nogueira RM, Nunes M, Moraes Figueiredo LT, 2005. Duplex reverse transcription-PCR followed by nested PCR assays for detection and identification of Brazilian alphaviruses and flaviviruses. J Clin Microbiol 43 :696–702.

    • Search Google Scholar
    • Export Citation
  • 20

    Tamura K, Dudley J, Nei M, Kumar S, 2007. MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol Biol Evol 24 :1596–1599.

    • Search Google Scholar
    • Export Citation
  • 21

    Regato M, Recarey R, Moratorio G, de Mora D, Garcia-Aguirre L, Gónzalez M, Mosquera C, Alava A, Fajardo A, Alvarez M, D’Andrea L, Dubra A, Martínez M, Khan B, Cristina J, 2008. Phylogenetic analysis of the NS5 gene of dengue viruses isolated in Ecuador. Virus Res 132 :197–200.

    • Search Google Scholar
    • Export Citation
  • 22

    Baleotti FG, Moreli ML, Figueiredo LT, 2003. Brazilian Flavivirus phylogeny based on NS5. Mem Inst Oswaldo Cruz 98 :379–382.

  • 23

    Gubler DJ, Kuno G, Sather GE, Velez M, Oliver A, 1984. Use of mosquito cell cultures and specific monoclonal antibodies in surveillance for dengue virus. Am J Trop Med Hyg 33 :158–165.

    • Search Google Scholar
    • Export Citation
  • 24

    Vanwambeke SO, van Benthem BHB, Khantikul N, Burgoorn-Mass C, Panart K, Oskam L, Lambin EF, Somboon P, 2006. Multilevel analysis of spatial and temporal determinants for dengue infection. Int J Health Geogr 5 :5.

    • Search Google Scholar
    • Export Citation
  • 25

    Kulldorff M, Nagarwalla N, 1995. Spatial disease clusters: detection and inference. Stat Med 14 :799–819.

  • 26

    Shope RE, Sather GE, 1979. Arboviruses. Lennette EH, Schmidt NJ, eds. Diagnostic Procedures for Viral, Rickettsial and Chlamydial Infections. Fifth edition. Washington, DC: American Public Health Association, 767–814.

  • 27

    Tavares-Neto J, Freitas-Carvalho J, Nunes MRT, Rocha G, Rodrigues SG, Damasceno E, Darub R, Viana S, Vasconcelos PFC, 2004. Pesquisa de anticorpos contra arbovírus e o v rus í vacinal da febre amarela em uma amostra da população de Rio Branco, antes e três meses após a vacina 17D. Rev Soc Bras Med Trop 37 :1–6.

    • Search Google Scholar
    • Export Citation
  • 28

    Galler R, Marchevsky RS, Caride E, Almeida LFC, Yamamura AMY, Jabor AV, Motta MCA, Bonaldo MC, Coutinho ESF, Freire MS, 2005. Attenuation and immunogenicity of recombinant yellow fever 17D-dengue type 2 virus for rhesus monkeys. Braz J Med Biol Res 38 :1835–1846.

    • Search Google Scholar
    • Export Citation
  • 29

    Aquino VH, Anatriello E, Gonçalves PF, da Silva EV, Vasconcelos PFC, Vieira DS, Batista WC, Bobadilla MI, Vazquez C, Morán M, Figueiredo LTM, 2006. Molecular epidemiology of dengue type 3 virus in Brazil and Paraguay, 2002–2004. Am J Trop Med Hyg 75 :710–715.

    • Search Google Scholar
    • Export Citation
  • 30

    Ministry of Health of Brazil, 2002. Programa Nacional de Controle da Dengue. Brasília: Ministério da Saúde.

  • 31

    Endy TP, Nisalak A, Chunsuttiwat S, Libraty DH, Green S, Rothman AL, Vaughn DW, Ennis FA, 2002. Spatial and temporal circulation of dengue virus serotypes: a prospective study of primary school children in Kamphaeng Phet, Thailand. Am J Epidemiol 156 :52–59.

    • Search Google Scholar
    • Export Citation
  • 32

    Thai KTD, Binh TQ, Giao PT, Phuong HL, Hung LQ, Nam NV, Nga TT, Groen J, Nagelkerke N, de Vries PJ, 2005. Seroprevalence of dengue antibodies, annual incidence and risk factors among children in southern Vietnam. Trop Med Int Health 10 :379–386.

    • Search Google Scholar
    • Export Citation
  • 33

    Balmasaeda A, Hammond SN, Tellez Y, Imhoff L, Rodriguez Y, Saborío SI, Mercado JC, Perez L, Videa E, Almanza E, Kuan G, Reyes M, Saenz L, Amador JJ, Harris E, 2006. High seroprevalence of antibodies against dengue virus in a prospective study of schoolchildren in Managua, Nicaragua. Trop Med Int Health 11 :935–942.

    • Search Google Scholar
    • Export Citation
  • 34

    Endy TP, Chunsuttiwat S, Nisalak A, Libraty DH, Green AS, Rothman AL, Vaughn DW, Ennis FA, 2002. Epidemiology of innaparent and symptomatic acute dengue virus infection: a prospective study of primary school children in Kamphaeng Phet, Thailand. Am J Epidemiol 156 :40–51.

    • Search Google Scholar
    • Export Citation
  • 35

    Guha-Sapir D, Schimmer B, 2005. Dengue fever: new paradigms for a changing epidemiology. Emerg Themes Epidemiol 2 :1.

  • 36

    Kaplan JE, Eliason DA, Moore M, Sather GE, Schonberger LB, Cabrera-Coello L, Fernandez de Castro J, 1983. Epidemiological investigations of dengue infection in Mexico, 1980. Am J Epidemiol 117 :335–343.

    • Search Google Scholar
    • Export Citation
  • 37

    Halsted SB, Nimmannitya S, Cohen SN, 1970. Observations related to pathogenesis of dengue hemorrhagic fever. IV Relation of disease severity to antibody response and virus recovered. Yale J Biol Med 42 :311–328.

    • Search Google Scholar
    • Export Citation
  • 38

    Vasconcelos PFC, Lima JWO, Raposo ML, Rodrigues SG, Travassos da Rosa JFS, Amorim SMC, Travassos da Rosa ES, Moura CMP, Fonseca N, Travassos da Rosa APA, 1999. Inquérito soroepidemiológico na Ilha de São Luís durante epidemia de dengue no Maranhão. Rev Soc Bras Med Trop 32 :171–179.

    • Search Google Scholar
    • Export Citation
  • 39

    Pan American Health Organization, 1994. Dengue and Dengue Hemorrhagic Fever in the Americas: Guidelines for Prevention and Control. Washington, DC: Pan American Health Organization.

  • 40

    Nunes-Araújo FRF, Ferreira MS, Nishioka SA, 2003. Dengue fever in Brazilian adults and children: assessment of clinical findings and their validity for diagnosis. Ann Trop Med Parasitol 97 :415–419.

    • Search Google Scholar
    • Export Citation
  • 41

    Rodrigues MBP, Freire HBM, Corrêa PRL, Mendonça ML, Silva MRI, França EB, 2005. Is it possible to identify dengue in children on the basis of Ministry of Health criteria for suspected dengue cases? J Pediatr (Rio J) 81 :209–215.

    • Search Google Scholar
    • Export Citation
  • 42

    Dietz VJ, Gubler DJ, Rigau-Pérez JG, Pinheiro F, Schatzmayr HG, Bailey R, Gunn RA, 1990. Epidemic dengue 1 in Brazil, 1986: evaluation of a clinically based dengue surveillance system. Am J Epidemiol 131 :693–701.

    • Search Google Scholar
    • Export Citation
  • 43

    Cunha RV, Schatzmayr HG, Miagostovich MP, Barbosa AMA, Paiva FP, Miranda RMO, Ramos CCF, Coelho JCO, Santos FB, Nogueira RMR, 1999. Dengue epidemic in the state of Rio Grande do Norte, Brazil, in 1997. Trans R Soc Trop Med Hyg 93 :247–249.

    • Search Google Scholar
    • Export Citation
Save