The Epidemiology of Malaria and Arthropod-Borne Virus Infections in a Rural Amazonian Population

ARTIGO ARTICLE 1325

The Acre Project: the epidemiology of malaria and arthropod-borne virus infections in a rural Amazonian population

Projeto Acre: epidemiologia da malária e das arboviroses em uma população rural amazônica Mônica da Silva-Nunes 1,2 Rosely dos Santos Malafronte 2,3 Bruna de Almeida Luz 1 Estéfano Alves de Souza 1 Lívia Carício Martins 4 Sueli Guerreiro Rodrigues 4 Jannifer Oliveira Chiang 4 Pedro Fernando da Costa Vasconcelos 4 Pascoal Torres Muniz 2,5 Marcelo Urbano Ferreira 1,2

Abstract Introduction

1 Instituto de Ciências The authors describe the baseline malaria preva- Since early 2003, the county (municipality) of Biomédicas, Universidade de lence and arbovirus seroprevalence among 467 Acrelândia in the easternmost area of the State São Paulo, São Paulo, Brasil. 2 Centro de Saúde de subjects in an ongoing cohort study in rural of Acre, Brazil, has been the site of a collabora- Acrelândia, Acrelândia, Brasil. Amazonia. Most subjects (72.2%) reported one tive research project involving the Universida- 3 Laboratório de or more previous episodes of malaria, and 15.6% de Federal do Acre (Federal University of Acre – Protozoologia, Instituto de Medicina Tropical de had been hospitalized for malaria, but only 3.6% UFAC), the Instituto Evandro Chagas (Evandro São Paulo, São Paulo, Brasil. of individuals five years or older had malaria Chagas Institute – IEC), and the Universidade 4 Seção de Arbovirologia parasites detected by microscopy (10 with Plas- de São Paulo (University of São Paulo – USP), e Febres Hemorrágicas, Instituto Evandro Chagas, modium vivax and 4 with P. falciparum). Anti- aimed at investigating the population’s health Belém, Brasil. bodies to Alphavirus, Orthobunyavirus, and/or conditions and proposing possible targets for 5 Departamento de Ciências Flavivirus were detected by hemagglutination public health interventions. Acrelândia is lo- da Saúde, Universidade Federal do Acre, inhibition (HI) in 42.6% of subjects aged five cated in an area with high malaria transmis- Rio Branco, Brasil. years or older, with a higher seropositivity rate sion, with an annual parasite index (number of

Correspondence among males (49.2%) than females (36.2%). malaria cases per year per thousand inhabi- M. U. Ferreira Since 98.9% of subjects had been immunized for tants) of 104.3 in 2002 1. It is also one of the pri- Departamento de yellow fever, the presence of cross-reactive anti- ority counties in the Amazon Basin for dengue Parasitologia, Instituto 2 de Ciências Biomédicas, bodies to dengue and other Flaviviruses cannot control activities . The follow-up of a cohort of Universidade de São Paulo. be ruled out, but at least 12 subjects (3.3%) with 467 individuals began in March 2004 (includ- Av. Prof. Lineu Prestes 1374, IgM antibodies to dengue virus detected by ing all age brackets) in an agricultural settle- São Paulo, SP 05508-900, Brasil. ELISA had a putative recent exposure to this virus. ment in rural Acrelândia. The principal objec- [email protected] tive of this prospective study is to identify risk Malaria; Dengue; Arboviruses; Arbovirus Infec- and protective factors associated with acute tions; Amazonian Ecosystem febrile diseases, particularly malaria and arbovirus infections. The current article de- scribes the baseline malaria prevalence and arbovirus seroprevalence in March-April 2004.

Cad. Saúde Pública, Rio de Janeiro, 22(6):1325-1334, jun, 2006 1326 Da Silva-Nunes M et al.

Methodology (with the aid of maps provided by the local teams from the Coordinating Body for the Con- Study area and population trol of Endemics under SESACRE) and visited by our team. The household census identified Acrelândia, a county (municipality) founded in 473 inhabitants (226 males, 247 females), rang- January 1993 through repartitioning of the coun- ing in age from one day to 90 years (mean 23.5 ties of Plácido de Castro and Senador Guio- years), distributed in 114 households. All mem- mard, has a population of 8,697 (estimated for bers of these households were considered eli- 2003 by the IBGE, Instituto Brasileiro de Geo- gible for entry into our cohort and were invited grafia e Estatística, or National Census Bureau, to participate in the study. 2000 Census; unpublished data). The county has a territory of 1,607.5km2 and is located be- Field study tween the Abunã and Iquiri Rivers (the latter also known as the Ituxi) in the Acre River Valley During the cross-sectional survey from March (Figure 1). The estimated human development 16 to April 29, 2004, all households in the study index (HDI) in 2000 for the total population area were visited by the field team (two physi- was 0.680 3, with an infant mortality rate of cians, a biologist, and a research assistant). The 70.75 per thousand live births (IBGE estimate project objectives were explained to each fami- for 1998; unpublished data). The overall illiter- ly in order to obtain written informed consent. acy rate is 26.7%, reaching 31.9% in the rural The following procedures were also performed: area 3, which includes 57% of the population. (a) application of an extensive questionnaire to Acrelândia is located 112km east of Rio Branco, obtain demographic, socioeconomic, and pri- capital of the State of Acre, and borders on the or disease history data and a detailed descrip- counties of Senador Guiomard and Plácido de tion of the households, (b) clinical examination Castro in Acre as well as the Brazilian States of of all household members by a field team physi- Amazonas and Rondônia and the country Bo- cian, (c) collection of venous blood samples livia (Figure 1). from all household members aged five years or The study area encompasses rural localities older, and (d) determination of the household’s known as Linha 14 Gleba Q and Reserva da Lin- exact location using portable navigation equip- ha 14 (9o41’-9o49’S, 67o05’-67o07’W), demar- ment (eTrex®, Garmin International, Olathe, cated by the Coordinating Body for the Control USA), with an accuracy of 15 meters. Data ob- of Endemics under the Acre State Secretariat of tained during this cross-sectional survey pro- Health and Sanitation (SESACRE). Linha 14 vide the baseline for the prospective cohort, main- Gleba Q includes both sides of the last 16 kilo- tained since then under clinical and epidemio- meters of Linha 14, an unpaved road that be- logical surveillance for acute febrile diseases. gins at the highway connecting Rio Branco to Porto Velho (BR-364); Reserva da Linha 14 in- Microscopic diagnosis of malaria cludes the edges of an unpaved road that be- gins at the end (km 30) of Linha 14 and is per- Microscopic diagnosis of malaria used thick pendicular to it, running parallel to the Iquiri blood smears stained with Giemsa according to River. In this article, these sites will be referred the Walker technique and thin smears fixed with to collectively as Ramal do Granada, the name methanol and stained with Giemsa 4. At least used by the local population to refer to Linha 14. 200 high-magnification (700x) microscopic fields Ramal do Granada is located in the largest were examined by two experienced microscopy farming settlement in the State of Acre, the Pe- technicians before defining the result for each dro Peixoto Organized Agricultural Settlement slide 5. Individuals with a malaria diagnosis re- (Figure 1), implemented by the National Insti- ceived free medication provided by the SESACRE tute for Colonization and Agrarian Reform in based on Ministry of Health treatment protocols the mid-1970s. The main economic activities 4. The drugs used were: (a) chloroquine and pri- are coffee and banana growing and beef and maquine for Plasmodium vivax and (b) meflo- dairy cattle-raising. There are three malaria di- quine or a quinine-doxycycline combination (in agnostic outposts maintained by the SESSACRE addition to primaquine in cases with patent ga- at kilometers 16, 24, and 30 of Ramal do Grana- metocytemia) for P.falciparum. da; the largest, at km 16, also has nursing technicians trained in primary health care. Investigation of arbovirus antibodies During the census conducted by our fieldwork team, all buildings (both inhabited and All serological assays were performed at the Di- uninhabited) in the study area were located vision of Arbovirology and Hemorrhagic Fevers

Cad. Saúde Pública, Rio de Janeiro, 22(6):1325-1334, jun, 2006 EPIDEMIOLOGY OF MALARIA AND ARTHROPOD-BORNE VIRUS INFECTIONS 1327

Figure 1

Location of the County (Municipality) of Acrelândia, Acre State, Brazil, and the Pedro Peixoto Organized Settlement Project where the study area is located.

AMAZONAS STATE

ACRE STATE SOUTH AMERICA

Rio RONDÔNIA PERU Branco STATE

AMAZONIA BOLIVIA

Brasília

BRAZIL

ve i

uxi R t I 64 PAD - 3 Pedro BR Rio Branco Peixoto Acrelândia ã River Abun

BR Senador - 36 Guiomard 4

Plácido de Castro

at IEC in Belém, Pará. Serum samples were ob- samples were also tested by ELISA for IgM anti- tained by centrifuging the venous blood in the bodies to the dengue and yellow fever viruses. field laboratory and storing at -20oC in dry ice until shipping by airplane to Belém. The samples Statistical analysis were initially submitted to microplate hemagglutination inhibition (HI) 6 against standard- Two or more proportions were compared using ized antigens from 19 types of arbovirus, in- the χ2 or linear trend χ2 tests. Odds ratios and cluding four from genus Alphavirus (Eastern respective 95% confidence intervals (95%CI) equine encephalomyelitis [EEE], Western equine were calculated for the target associations us- encephalomyelitis [WEE], Mayaro, and Mucam- ing 2x2 tables. The 95% confidence intervals bo), nine from genus Flavivirus (two samples were obtained for proportions. These analyses of yellow fever virus [wild and vaccinal], Ilhéus, used the open-access software Epidat (http:// São Luís, Rocio, dengue [DEN] 1, DEN-2, DEN-3, www.paho.org/spanish/sha/epidat.htm; Pan- and DEN-4), and six from genus Orthobunya- American Health Organization). virus (Caraparu, Catu, Guaroa, Maguari, Oro- Median exposure time until the first episode pouche, and Tacaiuma). Positive sera accord- of malaria and median age at the first episode, ing to HI (titer ≥ 1:20) against the Mayaro virus along with their respective 95%CI, were esti- were submitted to a confirmatory enzyme-linked mated by weighted linear regression models, immunosorbent assay (ELISA) for IgM antibod- commonly applied to dose-response type situ- ies 7. Regardless of the HI results, the serum ations in which the proportions of responses

Cad. Saúde Pública, Rio de Janeiro, 22(6):1325-1334, jun, 2006 1328 Da Silva-Nunes M et al.

are transformed into probits 8. In this analysis with a history at least one episode of malaria the sample population was stratified by age or correlates positively with age (linear trend χ2 = time of residence in Amazonia, and an expo- 72.70, 1 degree of freedom, p < 0.00001) and time sure “dose” was attributed to each stratum, cor- of residence in Amazonia (linear trend χ2 = responding to mean age or time of residence in 88.64, 1 degree of freedom, p < 0.00001). Amazonia, in years. The “doses” were log-trans- Based on probit models adjusted to the da- formed to adjust the probit models. In the pro- ta presented in Table 3, median residence time bit model, the proportion of individuals in in Amazonia since the first episode of malaria each stratum who reported at least one prior was estimated at 3.3 years (95%CI: 1.9-4.7 years). episode of malaria corresponds to the propor- In other words, after living 39 months in Ama- tion of “responses”. The same analytical proce- zonia, half of the sample population had al- dure was applied separately to the proportion ready experienced at least one symptomatic of individuals who reported at least one prior malaria infection, confirmed by microscopy episode of P.falciparum and P.vivax malaria. and treated. Probit models were also adjusted Calculations used a program written in BASIC 8. to the proportions of those reporting at least The study adopted a 5% significance level. one episode of malaria attributed to P.falciparum and P.vivax, resulting in median resi- Ethical aspects dence times in Amazonia of 11.9 years (95%CI: 8.3-17.5 years) and 4.8 years (95%CI: 1.8-7.9 The research protocol was analyzed and ap- years), respectively, for the first documented proved by the Research Ethics Committee for episode of malaria attributed to each Plasmod- Experimentation in Human Beings at the Insti- ium species. tute of Biomedical Sciences, USP (reviews 318/ The probit model adjusted to data in Table CEP, July 19, 2002, and 538/ CEP, January 7, 2004). 2 estimated a median age of 4.8 years (95%CI: 3.2-6.4 years) for individuals presenting their first episode of malaria. The probit models were Results also applied separately to data on history of P. falciparum and P.vivax infection, resulting in Malaria: prior exposure and prevalence medians of 20.2 years of age (95%CI: 16.2-25.4 of current infection years) and 6.4 years (95%CI: 3.7-9.1 years), respectively, at the time of primary infection by The study participants included 98.7% of the these Plasmodium species. 473 eligible individuals identified during the Of the 399 individuals five years of age or household census; four males and two females older, eligible for drawing venous blood sam- refused to participate or were not found in the ples, 387 (97.0%; mean age 27.4 years) had at residences after at least three attempts. The least one thick and thin blood smears analyzed sample population consisted of 467 individuals by microscopy for malaria diagnosis. Addition- ranging in age from 1 day to 90 years (mean, al capillary blood samples were drawn by finger 23.3 years) distributed in 113 households, with prick, at the request of the parents or guardians, 42.4% under 15 years of age. The slight pre- in 23 children under five years of age, totaling dominance of males in the sample population 410 samples submitted to microscopic exami- (1.08 males for every female) is similar to that nation. Of these, 16 were positive for P.vivax of the overall population in Acrelândia (1.15:1), and four for P.falciparum (4.9% positivity rate according to estimates by the IBGE for 2003 among the samples analyzed), with no case of (unpublished data). Table 1 shows the sample co-infection by the two species. No individual population distribution by gender and age with infection diagnosed during the cross-sec- bracket. tional survey, through active search, was classi- The impact of malaria as a cause of prior fied as asymptomatic. However, three present- morbidity in the sample population can be in- ed no signs or symptoms of malaria at the mo- ferred from the data, obtained through individ- ment of blood sampling: (a) one adult who pre- ual interviews and shown in Tables 2 and 3. The sented symptoms during the two weeks fol- vast majority (72.2%) reported at least one pre- lowing the home visit and appeared at the vious episode of malaria diagnosed by mi- SESACRE malaria diagnostic outpost for labo- croscopy and treated (mean of 9 episodes, ratory confirmation; (b) one child who had among the 337 who reported prior malaria), gone for medical care during the two weeks and 73 (15.6%) of these reported at least one prior to the home visit due to an ill-defined hospitalization due to malaria. As shown in Ta- fever; and (c) a pregnant woman on chloro- bles 2 and 3, the proportions of individuals quine chemoprophylaxis at the time of blood

Cad. Saúde Pública, Rio de Janeiro, 22(6):1325-1334, jun, 2006 EPIDEMIOLOGY OF MALARIA AND ARTHROPOD-BORNE VIRUS INFECTIONS 1329

Table 1

Sample population distribution (n = 467) by gender and age bracket. Ramal do Granada, Acre State, Brazil, 2004.

Age bracket (years) Female Male Total n% n% n%

< 1 5 2.2 6 2.5 11 2.4 1-4 27 12.1 30 12.3 57 12.2 5-14 67 29.9 63 25.9 130 27.8 15-30 63 28.1 58 23.9 121 25.9 31-60 53 53.7 71 29.2 124 26.6 > 60 9 4.0 15 6.2 24 5.1 Total 224 100.0 243 100.0 467 100.0

The percentages in each column refer to the proportions of individuals from each gender in each age bracket.

Table 2

Proportions of sample population (n = 467) by age bracket with a history of one or more episodes of malaria. Ramal do Granada, Acre State, Brazil, 2004.

Age bracket (years) Plasmodium vivax Plasmodium falciparum Either species n% n% n%

< 1 1 9.1 0 0.0 1 9.1 1-4 20 35.1 6 10.5 21 36.8 5-14 79 60.8 45 34.6 84 64.6 15-30 82 67.8 67 55.4 102 84.3 31-60 93 75.0 79 63.7 107 86.3 > 60 19 79.2 17 70.8 22 91.7 Total 294 63.0 214 45.8 337 72.2

The percentages in each line refer to the proportion of each age bracket reporting at least one prior episode of malaria diagnosed by microscopy and treated. The numbers of individuals in each age bracket are shown in Table 1.

Table 3

Proportions of sample population (n = 467) according to time of residence in Amazonia, with a history of at least one prior episode of malaria. Ramal do Granada, Acre State, Brazil, 2004.

Time of residence Plasmodium vivax Plasmodium falciparum Either species in Amazonia (years) n% n% n%

< 5 [n = 89] 30 33.7 12 13.5 32 36.0 5-9 [n = 106] 64 60.4 39 36.8 70 66.0 10-19 [n = 175] 126 72.0 93 53.1 143 81.7 20 + [n = 97] 74 76.3 70 72.2 92 94.8 Total [n = 467] 294 63.0 214 45.8 337 72.2

The percentages on each line refer to the proportion of individuals in each stratum reporting at least one prior episode of malaria diagnosed by microscopy and treated.

Cad. Saúde Pública, Rio de Janeiro, 22(6):1325-1334, jun, 2006 1330 Da Silva-Nunes M et al.

sampling and who turned symptomatic as soon others had only resided previously in other as prophylaxis was suspended. States belonging to the Legal Amazonia (Ron- Among the 387 individuals five years of age dônia and Mato Grosso). The others (ranging or older who had blood samples taken and ex- from 31 to 52 years of age) were originally from amined, 14 (3.6%; 95%CI: 1.6-5.6%) were posi- the South and Southeast of Brazil (States of Es- tive on microscopic examination; ten had P. v i - pírito Santo, Paraná, and Santa Catarina), but vax infection and four had P.falciparum. Since had lived between 13 and 28 years in Amazo- this age group was sampled systematically (re- nia. Six of the seropositive individuals work fre- gardless of symptoms), this positivity rate is quently in timber cutting, an activity that in- considered a more adequate estimate of malar- volves frequent contact with the forest envi- ia prevalence in the sample population. No sta- ronment. However, of the 15 individuals with tistically significant difference was observed (p = antibodies to the Mayaro virus and submitted 0.55, χ2 with Yates correction) in overall malar- to confirmation for detection of IgM antibod- ia prevalence between males (4.4%; 95%CI: 1.3- ies, only one was positive, suggesting recent in- 7.5%) and females (2.8%; 95%CI: 0.9-5.4%). fection by this virus (i.e., in the three previous years). Prevalence of arbovirus antibodies Due to probable exposure of the local population to some Flavivirus species (dengue and Of the 399 individuals aged five years of age and wild and vaccinal yellow fever), the heterotypi- older and eligible for venous blood sampling, cal reactions observed in the HI assay with the 357 (89.5%; mean age, 27.5 years) were tested nine Flavivirus tested may be due largely to an- for antibodies to the various types of arbovirus tibody cross-reactions to one or more of these using HI, and 358 (89.7%; mean age 27.5 years) viruses. Although 98.9% of individuals five years were tested for IgM antibodies to dengue and or older reported previous yellow fever immu- yellow fever using ELISA. Of these, 15 individu- nization, only 66 (18.5%) of the 357 assessed als with antibodies to the Mayaro virus, detect- with serology had antibodies (detected by HI) ed by HI, were tested by ELISA for IgM antibod- to the 17D vaccinal virus. Only four individuals ies to this virus. showed IgM antibodies to the yellow fever A total of 152 samples (42.6%; 95%CI: 37.4- virus detected by ELISA; in all cases, the histo- 47.9%) were positive according to HI. Of these, ry of vaccination may explain the occurrence 20 (5.6%) presented monotypical reactions (i.e., of these antibodies. antibodies to a single type of virus from this Fourteen individuals (3.9%; 95%CI: 2.2-6.6%), genus) to Alphavirus, 15 (4.2%) had antibodies seven males and seven females (from 17 to 79 to Orthobunyavirus (of which 11 were mono- years), had IgM dengue virus antibodies detect- typical and four heterotypical), and 65 (18.2%), ed by ELISA. One had a history of yellow fever all of which heterotypical, to Flavivirus (in- vaccination 30 days previously; another had cluding the four dengue serotypes and the wild IgM antibodies to the yellow fever virus, a situ- yellow fever virus). A statistically significant ation in which a cross-reaction cannot be ruled difference was observed (p = 0.02, χ2 test with out. In the remaining 12 (3.3% of the tested sam- Yates correction) in the overall prevalence of ple), the presence of specific IgM antibodies antibodies to arboviruses comparing males was interpreted as indicative of recent dengue (49.2%; 95%CI: 41.4-56.4%) and females (36.2%; infection. Of these 12, nine are relatives, and 95%CI: 29.2-43.7%), with an odds ratio of 1.69 their households are located within a 2km ra- for males (95%CI: 1.08-2.64). Table 4 shows the dius in the study area. Two individuals with gender and age distribution of individuals with dengue IgM antibodies reported previous febrile antibodies to one or more of the 19 arboviruses episodes with clinical diagnosis of dengue (but tested; the overall arbovirus seroprevalence, without serological confirmation), 11 months estimated by HI, showed a positive correlation and two years prior to the study, respectively. to age (linear trend χ2 = 24.62, 1 degree of free- Of the 467 interviewees, 11 (2.3%; six males dom, p < 0.00001). and five females from 10 to 70 years of age) re- Table 5 shows the number and proportion ported some lifetime diagnosis of dengue. In of individuals with antibodies detected by HI all of these cases the diagnosis was clinical, to each type of Alphavirus and Orthobunya- without laboratory confirmation. Of these in- virus tested. Of these viruses, based on sero- dividuals, 10 had serum samples tested by HI. prevalance, Mayaro is the only one that appears All the results were Flavivirus antibody-nega- to circulate widely in the study area. Half of the tive, although two individuals (mentioned above) 18 seropositive individuals from 11 to 49 years had dengue IgM antibodies detected by ELISA. of age were natives of the State of Acre and two

Cad. Saúde Pública, Rio de Janeiro, 22(6):1325-1334, jun, 2006 EPIDEMIOLOGY OF MALARIA AND ARTHROPOD-BORNE VIRUS INFECTIONS 1331

Discussion Table 4

Beginning in the 1970s, large-scale projects im- Positive hemagglutination inhibition for one or more arbovirus types, plemented by the National Institute for Colo- by age bracket. Ramal do Granada, Acre State, Brazil, 2004. nization and Agrarian Reform (INCRA) attract- ed migrants from the South and Southeast of Age bracket (years) Individuals Positive Positive Brazil to Amazonia. The projects were of two tested (n) individuals (n) individuals (%) types, the so-called Organized Settlement Pro- 5-14 111 26 23.4 jects (PAD), aimed at farmers with better tech- 15-30 103 48 46.6 nical training and financial resources, and the 31-60 116 64 55.2 Integrated Colonization Projects (PIC), target- > 60 27 14 51.9 ing low-income groups, specifically landless Total 357 152 42.6 rural workers 9. The impact of tropical diseases, especially malaria, on recently arrived migrants The percentages on each line refer to the proportion of individuals in each age bracket with antibodies to one or more types of arbovirus; only individuals five has been accurately described in Machadinho years or older were tested. D’Oeste, Rondônia 10, but there are only limited data on the living and health conditions in settlements that have been consolidated over the course of several decades, and which now include the majority of the population in farm- Table 5 ing areas in Amazonia. The current article pre- sents data on a sample population from a large- Number and proportion of individuals with antibodies detected by scale agricultural settlement, collected more hemagglutination inhibition to each type of Alphavirus and Orthobunyavirus. than two decades after it was originally occu- Ramal do Granada, Acre State, Brazil, 2004. pied. Virus Positive Malaria n%

Alphavirus Malaria is still one of the main parasitic en- Eastern equine encephalomyelitis 1 0.3 demics in Brazil. From 1970 to the mid-1990s, Western equine encephalomyelitis 0 0.0 the annual malaria incidence in Brazil increased Mayaro 18 5.0 ten-fold, then leveled off at around 500 thou- Mucambo 1 0.3 sand cases a year, of which 99% were acquired in Amazonia. However, in 1999 there was a new Orthobunyavirus peak in incidence, reaching a record of 630 Caraparu 4 1.1 11 thousand reported cases . In 2000, this situa- Catu 3 0.8 tion prompted the Plan for Intensification of Guaroa 5 1.4 Malaria Control Activities in the Amazon Re- Maguari 0 0.0 gion, aimed at a major reduction in the num- Oropouche 6 1.7 ber of cases in Amazonia in the following two Tacaiuma 1 0.3 years. This plan led to important decreases in malaria incidence in most of the States, with Only individuals five years or older (n = 357) were tested. The same individual can have antibodies to more than one virus. the important exception of Rondônia. The most recent available data are for 2003, with 350 thousand cases 12. Despite the large drop in malaria incidence since 2001, Acrelândia is still among the counties at greatest risk in the State riverine communities (16.9%) 17 in Amazonia. of Acre 1. With less than 2% of the State’s popu- However, the use of more sensitive diagnostic lation, Acrelândia has nearly 20% of its report- methods than conventional microscopy could ed malaria cases. show a higher proportion of malaria-infected Malaria prevalence (3.6%) in Ramal do Gra- individuals, with or without symptoms 17,18. nada in the late rainy season in 2004 is similar The impact of malaria on the study popula- to that described in peri-urban (1.6%) 13 and tion in terms of prior morbidity can be assessed rural populations (0.5-4.2%) 14 in the neighbor- by the proportion of individuals reporting at ing State of Rondônia, consisting predominant- least one episode of diagnosed and treated ly of migrants who settled there at least a decade malaria (72.2% of the sample) and at least one previously. The prevalence is still far lower than hospitalization due to malaria (15.6%). The his- in mining areas (17.3-35.0%) 15,16 and isolated torical data confirm the trend in recent years,

Cad. Saúde Pública, Rio de Janeiro, 22(6):1325-1334, jun, 2006 1332 Da Silva-Nunes M et al.

in which P.vivax predominated over P.falci- esis of local transmission is suggested by the parum in most of the Brazilian Amazonia 11; fact that nine of the 12 individuals with serolo- more individuals in the current sample report- gy compatible with recent dengue virus infec- ed previous exposure to P.vivax as compared tion reside in a geographically limited area to P. falciparum (Tables 2 and 3). The mean close to km 16 of Ramal do Granada. Aedes ae- time between initial exposure to malaria and gypti and Ae. albopictus have never been found primary infection in Ramal do Granada is com- in entomological surveys in the rural area of parable to the estimate for native Amazonian Acrelândia (SESACRE, unpublished data); how- populations; based on different modeling strate- ever, these surveys were done somewhat hap- gies, and the median age at primary infection hazardly and were focused primarily on the among natives in two different areas in the winged and aquatic forms of malaria vectors, Brazilian Amazon during the last decade was without prioritizing probable breeding sites for estimated at 4 to 6 years 8,19. Aedes. Theoretically, the increasing population density in Ramal do Granada could allow domi- Dengue and other arbovirus infections ciliary infestation with Ae. aegypti, a hypothesis whose proof depends on systematic ento- Epidemiological data on dengue transmission mological surveys, currently in the planning in Brazil come mostly from large cities in the phase. Southeast, Northeast, and more recently the In Ramal do Granada, as well as in Rio Bran- North of the country 20,21,22,23. The disease has co 21, the prevalence of antibodies to the vacci- received little epidemiological, clinical, and nal yellow fever virus (17D) was surprisingly laboratory research in small cities 24 or rural ar- low given the frequency of reported previous eas, and no data are available on dengue trans- immunization. For example, in Rio Branco on- mission dynamics and its main determinants ly 30% of individuals reporting prior yellow in typical Amazonian counties, even though fever vaccination showed antibodies to the 17D they represent nearly 10% of what are consid- virus 21. However, we feel that the proportion ered priority Brazilian counties under the Min- of the sample population effectively vaccinat- istry of Health’s National Dengue Control Pro- ed against yellow fever is higher than indicated gram 2. However, sero-epidemiological studies by the serological data. For example, the State in Loreto, in the Peruvian Amazon, show equiv- Vaccination Campaign in 1999 vaccinated alent prevalence rates for dengue IgG antibod- 416,267 individuals, more than 80% of the pop- ies in the urban population in Iquitos (66%) ulation of Acre (SESACRE, unpublished data); and in adjacent forest-dwelling populations since then, various vaccination drives have (32-67%) 25, indicating intense circulation of been held in Ramal do Granada, covering indi- the dengue virus in rural areas. In Southeast viduals who had still not been born or had not Asian countries, recent years have witnessed a arrived in the area or were not eligible for vac- predominance of hemorrhagic dengue (a dis- cination in 1999. It is important to highlight ease until recently limited to large cities) in that hemagglutination inhibition is not the test rural populations 26. of choice for evaluating the immunity conferred Although the re-introduction of the dengue by the yellow fever vaccine. The best technique virus into Brazil in 1981 started from Roraima 27, for this evaluation is the plaque reduction neu- it was only in 1999 that the dengue epidemic tralization test (PRNT), which was not used in spread throughout Amazonia 2,28. Acre was the this study. In the future it would be important last Amazonian State to report autochthonous to use PRNT to investigate the immunity in- dengue cases (DEN-1 and DEN-2), with a ma- duced by the 17D vaccine, whether in Acrelân- jor increase (40%) in incidence from 2002 and dia or in other counties in the State of Acre. 2003 (FUNASA, unpublished data). The only Current land occupation patterns in the available data on prevalence of antibodies to Amazon Basin favor the emergence of various the dengue virus and other arboviruses in pop- typically rural arbovirus diseases, including ulations in Acre refer to a sample in the city of those produced by the Oropouche virus, asso- Rio Branco 21. ciated with an epidemic in Xapuri, Acre (in Joint analysis of serological data and indi- 1996) 29 and the Mayaro virus 30. However, in vidual history suggest that at least 12 individu- Rio Branco 21 and Ramal do Granada (Table 5), als in our sample (3.3% of the tested popula- the prevalence of antibodies to the types of Al- tion) had recent exposure to the dengue virus, phavirus and Orthobunyavirus that were inves- although it is impossible to determine the pre- tigated is far inferior to that usually described cise site of infection (urban versus rural, in in rural populations of Eastern Amazonia 31. Acrelândia versus other counties). The hypoth- Although Mayaro has never been isolated in in-

Cad. Saúde Pública, Rio de Janeiro, 22(6):1325-1334, jun, 2006 EPIDEMIOLOGY OF MALARIA AND ARTHROPOD-BORNE VIRUS INFECTIONS 1333

dividuals from Acre, it appears to be the only Alphavirus circulating with a certain frequency in this State, judging by the available seroprevalence data (Table 5) 21. This hypothesis is supported by the finding that half of the individuals in our sample with antibodies to the Mayaro virus are native Acreans.

Resumo Acknowledgments

Descrevem-se aqui dados de prevalência de malária e The authors wish to thank Sebastião Bocalom Rodri- soroprevalência de arboviroses na linha de base de um gues, Damaris de Oliveira, and Nésio Mendes de Car- estudo de coorte, envolvendo 467 indivíduos, atual- valho (Acrelândia Municipal Government) and Rai- mente em curso na Amazônia rural. A maioria (72,2%) mundo A. Costa, Manoel Costa and other members dos indivíduos relatou um ou mais episódios malári- of the local malaria control team from the Acre State cos prévios, tendo 15,6% deles tido pelo menos uma in- Secretariat of Health and Sanitation for their excel- ternação hospitalar motivada pela malária, mas so- lent logistic support, Adamílson Luís de Souza for his mente 3,6% dos indivíduos com idade igual ou supe- assistance during the fieldwork, and Luís Antônio rior a cinco anos tinham plasmódios (Plasmodium vi- Vieira Garcia for assisting in the laboratory diagnos- vax em dez e P. falciparum em quatro) diagnosticados tic activities. The study received funding from the São por microscopia. Detectaram-se anticorpos contra Al- Paulo State Research Foundation (FAPESP, grants phavirus, Orthobunyavirus e Flavivirus, pelo teste de 03/09719-6 and 04/00373-2), the Brazilian National inibição de hemaglutinação, em 42,6% dos indivíduos Research Council (CNPq, grants 504332/2004-0 and com idade igual ou superior a cinco anos, com uma 302770/2002-0), and the Department of Science and taxa de soropositividade maior entre indivíduos do Technology (DECIT) of the Secretariat for Science, sexo masculino (49,2%) do que os do feminino (36,2%). Technology, and Strategic Inputs under the Brazilian Como 98,9% dos indivíduos entrevistados tinham sido Ministry of Health (grant 50.1489/2003-7). M. da Sil- previamente vacinados contra a febre amarela, a preva-Nunes and B. A. Luz are FAPESP research fellows; sença de reações cruzadas com o vírus do dengue e ou- E. A. Souza, P. F. C. Vasconcelos, and M. U. Ferreira are tros Flavivirus não pode ser descartada, mas a expo- CNPq research fellows. sição recente ao vírus do dengue foi evidenciada pela detecção de anticorpos IgM por ELISA em 12 indiví- duos (3,3%). References Malária; Dengue; Arbovírus; Infecções por Arbovirus; Ecossistema Amazônico 1. Costa RA, Leal O, Leal EAS. Evolução da infecção malárica no Município de Acrelândia, Estado do Acre (1999-2002). Rev Soc Bras Med Trop 2004; 37 Contributors Suppl I:269. 2. Fundação Nacional de Saúde, Ministério da Saú- de. Programa nacional de controle da dengue. M. da Silva-Nunes, P. F. C. Vasconcelos, and M. U. Brasília: Ministério da Saúde; 2002. Ferreira designed the field survey, which was con- 3. Instituto Nacional de Estudos e Pesquisas Educa- ducted by M. da Silva-Nunes, R. S. Malaffronte, P. T. cionais Anísio Teixeira, Ministério da Educação. Muniz, and M. U. Ferreira. L. C. Martins, S. G. Rodri- Mapa do analfabetismo no Brasil. Brasília: Minis- gues, J. O. Chiang, and P. F. C. Vasconcelos performed tério da Educação; 2004. arbovirus serology, and M. da Silva-Nunes and R. S. 4. Fundação Nacional de Saúde, Ministério da Saú- Malafronte supervised the malaria microscopy and de. Diagnóstico e tratamento da malária. Brasília: reviewed all the slides. B. A. Luz and E. A. Souza ana- Ministério da Saúde; 1995. lyzed the data, under the supervision of M. U. Ferrei- 5. Trape J-F. Rapid evaluation of malaria parasite ra. M. da Silva-Nunes, P. F. C. Vasconcelos, and M. U. density and standardization of thick smear exam- Ferreira wrote the paper, which was read and ap- ination for epidemiological investigations. Trans proved by all the other authors. R Soc Trop Med Hyg 1985; 79:181-4.

Cad. Saúde Pública, Rio de Janeiro, 22(6):1325-1334, jun, 2006 1334 Da Silva-Nunes M et al.

6. Shope RE. The use of a microhemagglutination- 19. Burattini MN, Massad E, Coutinho FAB. Malaria inhibition test to follow response after arthro- transmission rates estimated from serological da- pod-borne virus infection in a community of for- ta. Epidemiol Infect 1993; 111:503-23. est animals. An Microbiol (Rio de J) 1963; 11 (Parte 20. Schatzmayr HG, Nogueira RMR, Travassos-da- A):167-71. Rosa APA. An outbreak of dengue virus at Rio de 7. Kuno G, Gomez I, Gubler DJ. Detecting artificial Janeiro – 1986. Mem Inst Oswaldo Cruz 1986; anti-dengue IgM immune complexes using an 81:245-6. enzyme-linked immunosorbent assay. Am J Trop 21. Tavares-Neto J, Freitas-Carvalho J, Nunes MR, Med Hyg 1987; 36:153-9. Rocha G, Rodrigues SG, Damasceno E, et al. Pes- 8. Ferreira MU, Camargo LMA, Ferreira CS. The me- quisa de anticorpos contra arbovírus e o vírus va- dian age of primary malaria infection in an Ama- cinal da febre amarela em uma amostra da popu- zonian community: probit analysis of cross-sec- lação de Rio Branco, antes e três meses após a tional data. Acta Trop 1994; 58:73-8. vacina 17D. Rev Soc Bras Med Trop 2004; 37:1-6. 9. Neves AM, Lopes AMT. Os projetos de coloniza- 22. Teixeira MG, Barreto ML, Costa MCN, Ferreira ção. In: Instituto Brasileiro de Geografia e Estatís- LD, Vasconcelos PFC, Caincross S. Dynamics of tica, organizador. A organização do espaço na fai- dengue virus circulation: a silent epidemic in a xa da Transamazônica. v. 1. Rio de Janeiro: Insti- complex urban area. Trop Med Int Health 2002; tuto Brasileiro de Geografia e Estatística; 1979. p. 7:757-62. 80-9. 23. Travassos-da-Rosa APA, Vasconcelos PFC, Travas- 10. Castilla RE, Sawyer DO. Malaria rates and fate: a sos-da-Rosa ES, Rodrigues SG, Mondet B, Cruz socioeconomic study of malaria in Brazil. Soc Sci ACR, et al. Dengue epidemic in Belém, Pará, Brazil, Med 1993; 37:1137-45. 1996-97. Emerg Infect Dis 2000; 6:298-301. 11. Fundação Nacional de Saúde, Ministério da Saú- 24. Vasconcelos PFC, Mota K, Straatmann A, Santos- de. Plano de intensificação das ações de controle Torres S, Travassos-da-Rosa APA, Tavares-Neto J. da malária na Amazônia Legal. Brasília: Ministé- Epidemia de dengue em Ipupiara e Prado, Bahia. rio da Saúde; 2000. Inquérito soro-epidemiológico. Rev Soc Bras Med 12. Braz RM, Pereira MPL. Situação da malária na Trop 2000; 33:61-7. Região Amazônica, Brasil, 2003. Rev Soc Bras Med 25. Hayes CG, Phillips IA, Callahan JD, Griebenow Trop 2004; 37 Suppl I:53. WF, Hyams KC, Wu SJ, et al. The epidemiology of 13. Camargo LMA, Dal Colletto GMD, Ferreira MU, dengue virus infection among urban, jungle, and Gurgel SM, Escobar AL, Marques A, et al. Hypoen- rural populations in the Amazon region of Peru. demic malaria in Rondonia (Brazil, western Ama- Am J Trop Med Hyg 1996; 55:459-63. zon region): seasonal variation and risk groups in 26. Chareonsook O, Foy HM, Teeraratkul A, Silarug N. an urban locality. Am J Trop Med Hyg 1996; 55:32-8. Changing epidemiology of dengue hemorrhagic 14. Camargo LMA, Ferreira MU, Krieger H, Camargo fever in Thailand. Epidemiol Infect 1999; 122:161-6. EP, Silva LHP. Unstable hypoendemic malaria in 27. Osanai CH, Travassos-da-Rosa APA, Tang AT, Rondonia (western Amazon region, Brazil): epi- Amaral RS, Passos AC, Tauil PL. Surto de dengue demic outbreaks and work-associated incidence em Boa Vista, Roraima. Rev Inst Med Trop São in an agro-industrial rural settlement. Am J Trop Paulo 1983; 25:53-4. Med Hyg 1994; 51:16-25. 28. Siqueira JB, Martelli CMT, Coelho GE, Simplício 15. Fontes CJF. Epidemiologia da malária e fatores ACR, Hatch DL. Dengue and dengue hemorrhag- associados à infecção assintomática por Plasmo- ic fever, Brazil, 1981-2002. Emerg Infect Dis 2004; dium em uma população de garimpeiros da Ama- 11:48-53. zônia Brasileira (Mato Grosso, 1996) [Tese de Dou- 29. Vasconcelos PFC, Travassos-da-Rosa APA, Ro- torado]. Belo Horizonte: Faculdade de Medicina, drigues SG, Travassos-da-Rosa ES, Dégallier N, Universidade Federal de Minas Gerais; 2001. Travassos-da-Rosa JFS. Inadequate management 16. Santos EO, Loureiro EC, Jesus IM, Brabo E, Silva of natural ecosystems in the Brazilian Amazon re- RSU, Soares MCP, et al. Diagnóstico das condi- gion results in the emergence and reemergence of ções de saúde de uma comunidade garimpeira da arboviruses. Cad Saúde Pública 2001; 17 Suppl: região do Rio Tapajós, Itaituba, Pará, Brasil, 1992. 155-64. Cad Saúde Pública 1995; 11:212-25. 30. Tesh RB, Watts DM, Russell KL, Damodaran C, 17. Alves FP, Durlacher RR, Menezes MJ, Krieger H, Calampa C, Cabezas C, et al. Mayaro virus disease: Silva LHP, Camargo EP. High prevalence of asymp- an emerging mosquito-borne zoonosis in tropi- tomatic Plasmodium vivax and Plasmodium fal- cal South America. Clin Infect Dis 1999; 28:67-73. ciparum infections in native Amazonian popula- 31. Travassos-da-Rosa APA, Travassos-da-Rosa JFS, tions. Am J Trop Med Hyg 2002; 66:641-8. Pinheiro FP, Vasconcelos PFC. Arboviroses. In: Leão 18. Scopel KK, Fontes CJF, Nunes AC, Horta MF, Bra- RNQ, organizador. Doenças infecciosas e parasi- ga EM. High prevalence of Plasmodium malariae tárias: enfoque amazônico. Belém: Editora Cejup/ infections in a Brazilian Amazon endemic area Universidade do Estado do Pará; 1997. p. 208-25. (Apiacás – Mato Grosso State) as detected by poly- merase chain reaction. Acta Trop 2004; 90:61-4. Submitted on 10/Feb/2005 Final version resubmitted on 19/Aug/2005 Approved on 02/Sep/2005

Cad. Saúde Pública, Rio de Janeiro, 22(6):1325-1334, jun, 2006