Advertisement

Dengue serotype-specific seroprevalence among 5- to 10-year-old children in India: a community-based cross-sectional study

Open AccessPublished:November 04, 2016DOI:https://doi.org/10.1016/j.ijid.2016.10.030

      Highlights

      • Dengue is increasing in India, but empirical data documenting burden are very limited.
      • A multi-site study found that 59.6% of children aged 5–10 years had been infected at least once.
      • Nearly half had antibody profiles suggestive of more than one previous dengue infection.
      • Antibodies against multiple dengue serotypes were identified at all eight sites.
      • These data provide strong evidence for hyper-endemic dengue transmission across India.

      Summary

      Background

      Dengue surveillance data in India are limited and probably substantially underestimate the burden of disease. A community-based study was undertaken to assess the prevalence of dengue-specific immunoglobulin G (IgG) antibodies in children across India and to examine historical dengue exposure rates. Potential associations between socio-economic factors and dengue seroprevalence were also assessed (registered at ctri.nic.in: CTRI/2011/12/002243).

      Methods

      A convenience sample of 2609 healthy children aged 5–10 years was enrolled; these children were registered at or were living in the vicinity of eight centres located at six geographically distinct sites across India. Blood samples were drawn to test for the presence of dengue IgG antibodies using ELISA. Serotype-specific neutralizing antibody titres were measured in dengue IgG-positive children using dengue plaque reduction neutralization tests. Socio-demographic and household information was collected using a questionnaire.

      Results

      Overall, 2558/2609 children had viable samples with laboratory results for dengue IgG. Dengue IgG seroprevalence across all sites was 59.6% (95% confidence interval 57.7–61.5%): the lowest (23.2%) was in Kalyani, West Bengal, and the highest (80.1%) was in Mumbai. Seroprevalence increased with age. Multivariate analysis suggested associations with household water storage/supply and type of housing. Half of the subjects with positive IgG results presented a multitypic profile, indicating previous exposure to more than one serotype.

      Conclusions

      The overall dengue seroprevalence suggests that dengue endemicity in India is comparable to that in highly endemic countries of Southeast Asia. Additional prospective studies are required to fully quantify the disease burden, in order to support evidence-based policies for dengue prevention and control in India.

      Keywords

      1. Introduction

      Dengue is caused by a mosquito-borne flavivirus that is endemic in tropical and subtropical countries, including India.
      • World Health Organization
      Global strategy for dengue prevention and control: 2012-2020.
      Sporadic outbreaks have been reported in India for over 200 years. The scale and severity of two major epidemics in the 1990s prompted the implementation of a number of strategies to aid the control and surveillance of dengue. In particular, a passive surveillance programme and publication of guidelines for dengue prevention and control was launched as an initiative of the National Vector Borne Disease Control Programme,
      • Chakravarti A.
      • Arora R.
      • Luxemburger C.
      Fifty years of dengue in India.
      in collaboration with the existing Integrated Disease Surveillance Programme. Due to non-specific and often mild symptoms, dengue is significantly under-reported in nearly all countries. This is exacerbated in India, where dengue surveillance data are collected from only approximately 500 sentinel hospitals.
      • Shepard D.S.
      • Halasa Y.A.
      • Tyagi B.K.
      • Adhish S.V.
      • Nandan D.
      • Karthiga K.S.
      • et al.
      Economic and disease burden of dengue illness in India.
      Studies using global or extrapolated data have quantified this under-reporting, and suggest that the dengue disease burden in India is likely to be the highest in the world.
      • Shepard D.S.
      • Halasa Y.A.
      • Tyagi B.K.
      • Adhish S.V.
      • Nandan D.
      • Karthiga K.S.
      • et al.
      Economic and disease burden of dengue illness in India.
      • Bhatt S.
      • W Gething P.
      • Brady O.J.
      • Messina J.P.
      • Farlow A.W.
      • Moyes C.L.
      • et al.
      The global distribution and burden of dengue.
      Dengue has spread from urban to rural areas of India in recent decades.
      • Chakravarti A.
      • Arora R.
      • Luxemburger C.
      Fifty years of dengue in India.
      • National Vector Borne Disease Control Programme
      Dengue cases and deaths in the country since 2010.
      All four virus serotypes – DENV-1 to DENV-4 – have been documented in India, without a clear geographical distribution. Areas where serotypes co-circulate are increasing in number and scale.
      • Chakravarti A.
      • Arora R.
      • Luxemburger C.
      Fifty years of dengue in India.
      Specifically, DENV-1, DENV-2, and DENV-4 were isolated during an outbreak of dengue fever in Vellore in 1963 and all four serotypes were isolated during another outbreak in the same city in 1968.
      • Chakravarti A.
      • Arora R.
      • Luxemburger C.
      Fifty years of dengue in India.
      DENV-2 was the predominant serotype from the early 1970s to 2000, responsible for large epidemics in 1993 and 1996. DENV-3 was the predominant serotype in a 2003 outbreak and co-circulated with DENV-1 in 2006 in Dehli.
      • Chakravarti A.
      • Arora R.
      • Luxemburger C.
      Fifty years of dengue in India.
      Delhi became hyperendemic for dengue, with all four serotypes isolated in 2003 and 2006.
      • Chakravarti A.
      • Arora R.
      • Luxemburger C.
      Fifty years of dengue in India.
      No study to date has taken a nationally representative view of serotype distribution.
      Cross-sectional, population-based, age-stratified seroprevalence data describe historical disease transmission intensity.
      • Farrington C.P.
      • Kanaan M.N.
      • Gay N.J.
      Estimation of the basic reproduction number for infectious diseases from age-stratified serological survey data.
      • Ferguson N.M.
      • Donnelly C.A.
      • Anderson R.M.
      Transmission dynamics and epidemiology of dengue: insights from age-stratified sero-prevalence surveys.
      A seroprevalence study was undertaken to describe the prevalence of dengue-specific immunoglobulin G (IgG) antibodies and the infecting serotype profiles of positive samples, in children from eight sites across India.

      2. Methods

      2.1 Study design and centres

      This was a community-based, descriptive, cross-sectional seroprevalence study and was conducted between January 2011 and October 2012 in healthy children (registered at ctri.nic.in: CTRI/2011/12/002243). A convenience sample of eight private or government medical colleges at six geographically distinct locations was selected (1) to provide a wide geographical distribution across India, (2) to represent rural and peri-urban areas, and (3) based on the recognized ability of the site to conduct epidemiological research. Overall, two sites were selected in New Delhi and Hyderabad, and one site each in Kalyani, Wardha, Mumbai, and Bangalore.
      This study was conducted in accordance with the latest revision of the Declaration of Helsinki (Seoul, Korea, October 2008), guidelines for Good Epidemiological Practice,
      • International Epidemiological Association
      Good epidemiological practice.
      and local regulatory requirements. The study protocol was approved by ethics committees at the study centres and by the Health Ministry Steering Committee (HMSC) of the Government of India.

      2.2 Participants

      Children, 5–10 years old, who were resident at the study sites, were eligible. This is an age at which blood sample collection is relatively straightforward. Furthermore, seroconversion, and thus the demonstration of age-specific variation in seroprevalence, was considered likely in this age group. Parents or legal guardians were invited to enrol children during routine household visits by community health workers. Written informed consent was obtained from the parents or legal guardians, and children aged 8–10 years also signed an assent form. Enrolment at the two sites in Hyderabad was school-based; parent–teacher meetings were held at randomly selected schools to explain the purpose of the study, and all eligible children at those schools were invited to participate. Permission was obtained from the District Education Officer to perform study visits, complete questionnaires, and collect blood samples from study participants on the premises of each school.
      Assuming a dengue seroprevalence of 30%, a sample size of 323 participants at each site was calculated to ensure a precision of 5% for the two-sided 95% confidence interval (CI) around the seroprevalence point estimate.

      2.3 Data and sample collection

      Socio-demographic data (participant's demographic characteristics, household occupancy, water supply/storage, self-reported history of dengue or Japanese encephalitis (JE) virus infection, and education levels attained by the parents/guardians) were collected using a questionnaire. The questionnaire was administered by the health worker through interviews with the participant's parents or legal guardians during the first visit. Participants were asked to report to the affiliated centre for blood sample collection (5 ml) by a trained laboratory technician. The participant's height and weight were recorded using standard methods. Significant medical history, current or previous medical conditions, concomitant medication, recent vaccinations, and reasons for refusal of blood sampling, where relevant, were recorded.
      Blood samples were left at room temperature for 1–2 h before centrifugation. Each serum sample was divided into aliquots and stored in 3-ml cryotubes: 0.5 ml for dengue IgG antibody assessment, 1 ml for dengue plaque reduction neutralization tests (PRNT), and 0.5 ml for JE IgG antibody detection. Serum samples were kept frozen at −20 °C or below until analysis.

      2.4 Assays

      Samples were sent to the Microbiology Department of the Maulana Azad Medical College (New Delhi) for analysis. Dengue IgG antibody levels were assessed using commercially available ELISA kits. The EL1500G kit (Focus Diagnostics, California, USA) was used for samples from the first two sites (New Delhi); however, due to supply issues, the E-DEN 10G kit (Panbio Diagnostics, Brisbane, Australia) was used for the other sites. A sensitivity analysis of the two dengue IgG-specific ELISA kits performed on 30 samples confirmed 100% concordance; data from all centres were thus pooled. JE IgG antibody testing by indirect ELISA was also performed using commercially available kits (InBios, Washington, USA). Dengue IgG-positive samples were sent to the Centre for Vaccine Development (Mahidol University, Thailand) for measurement of dengue serotype-specific neutralizing antibody titres using PRNT based on 50% or greater reduction in plaque counts (PRNT50).
      • Russell P.K.
      • Nisalak A.
      • Sukhavachana P.
      • Vivona S.
      A plaque reduction test for dengue virus neutralizing antibodies.
      Seropositivity for dengue and JE were defined according to the manufacturer's instructions. Seroprevalence was the percentage of seropositive participants.
      For the interpretation of PRNT50 titres, participants were classified as follows: ‘naïve’, if antibody titres were <10 (1/dil) for the four serotypes; ‘monotypic’, if antibody titres were ≥10 (1/dil) for only one serotype or if titres were ≥10 (1/dil) for different serotypes, with a single serotype having a high titre (>80 (1/dil) titre, and ≥5 times higher than other titres); and ‘multitypic’, if antibody titres were ≥10 (1/dil) for different serotypes without a single predominant titre.
      At each site, a designated clinical research associate performed periodic visits to monitor implementation. All serum samples were checked for quantity and storage temperature by a lot quality assurance method.

      2.5 Statistical analysis

      Descriptive statistics reported baseline characteristics and immunogenicity results. Associations between all demographic–socio-economic factors and dengue serostatus were assessed by univariate analysis using the Chi-square test or t-test (for age) and multiple logistic regression with backward selection (significant if the p-value is ≤0.05). JE serostatus was not included as a covariate due to possible cross-reaction between flavivirus antibodies. Statistical significance was considered at and below a p-value of 0.05. Statistical analyses were performed using SAS software version 9.1 (SAS Institute, Cary, NC, USA) and Stata v. 12.1 (StataCorp LP, College Station, TX, USA).

      2.6 Role of the funding source

      The sponsor participated in all operational aspects of the study, including data collection, statistical analyses, and the writing of the study report. The sponsor funded medical writing support for the development of this publication. The corresponding author had full access to all of the data in the study and had final responsibility for the decision to submit for publication.

      3. Results

      3.1 Demographic characteristics of the study population

      Overall, 2609 participants from eight health centres were enrolled in the study. A total of 18 participants were excluded, due to age <5 years or >10 years (n = 8) and/or assent form not signed (n = 13). Thus, 2591 participants were included, all of whom had a blood sample drawn; 1364 (52.6%) were female, and the mean age of all participants was 7.8 (standard deviation (SD) 1.6) years.

      3.2 Socio-economic characteristics

      The mean number of people living in the participants’ households was 5.4 (SD 2.3), including 2.6 (SD 1.2) children under 15 years of age. Most (2170/2591; 83.8%) participants lived in a house, with 381 (14.7%) living in precarious lodgings; 1121 (43.3%) had an indoor piped public water supply, 2343 (90.4%) had water storage in the house, 1675 (64.6%) were connected to the public sewer, and 1339 (51.7%) had regular organized waste collection.

      3.3 Medical history

      A notable medical history was recorded for 94 (3.6%) children; 61 (2.4%) children were undergoing at least one current treatment at the time of enrolment. A history of dengue or a family history of dengue was reported by 15 (0.6%) and 48 (1.9%) participants, respectively. No participants reported a history or family history of JE infection. Only one participant reported receiving JE vaccination.

      3.4 Dengue IgG seroprevalence

      Anti-dengue IgG results were available for 2558/2591 (98.7%) participants. Serology data were missing for 33 participants; 22 samples from a single site (Mumbai) could not be analysed due to haemolysed red blood cells. Overall, 1525/2591 (59.6%) participants were dengue seropositive, with similar prevalence in males and females. Six of the eight sites had dengue seropositivity ranging from 58.2% to 69.0%. The sites in Kalyani and Mumbai had the lowest (23.2%) and highest (80.1%) seroprevalence, respectively (Figure 1). Overall, dengue IgG seroprevalence increased with age, from 40.7% (95% CI 36.0–45.5%) in children aged 5 years to 73.4% (95% CI 67.9–78.5%) in 10-year-olds (Figure 2). At the Bangalore site, seroprevalence remained relatively stable across the age strata (varying from 58.8% in 7-year-olds to 70.9% in 8-year-old children).
      Figure thumbnail gr2
      Figure 2Percentage of participants with antibody titres ≥10 (1/dil) against dengue according to age group (all study sites). Error bars show 95% confidence intervals. n, number of participants with available results per age group.

      3.5 Socio-economic characteristics associated with dengue seroprevalence status

      In univariate analyses, children seropositive for dengue were found more likely to be from homes with more than two children (p < 0.0001), more likely to have water storage (p < 0.0001) and indoor piped water from the public water supply (p < 0.0001), and less likely to be living in precarious housing (p = 0.0048) compared with dengue seronegative children (Table 1). Multiple logistic regression confirmed possible positive associations with household water storage (odds ratio (OR) 5.00, 95% CI 3.54–7.06) and indoor piped water from the public water supply (OR 1.49, 95% CI 1.19–1.85). Increasing participant age (p < 0.0001) and living in precarious lodgings compared to a house (OR 1.54, 95% CI 1.17–2.03) were also associated with dengue status. In terms of geography, Kalyani was associated with decreased exposure (OR 0.18, 95% CI 0.10–0.31), while Wardha (OR 1.49, 95% CI 1.07–2.08) and Mumbai (OR 2.49, 95% CI 1.70–3.65) had an elevated risk, in comparison with Delhi. The pseudo R2 of the final model was 0.085.
      Table 1Demographic and socio-economic characteristics associated with dengue seroprevalence status
      Dengue IgG positiveDengue IgG negativep-Value (OR
      The odds ratio (in parentheses) is provided for significant categorical multivariate results.
      )
      Demographic characteristics
      Overall, n15251033
      Age, mean (SD) years8.13 (1.56)7.39 (1.56)<0.0001
      t-test.
      ,
      p<0.05 in multivariate analysis.
      Socio-economic characteristics
      Number of children living in the household (%)<0.0001
      Chi-square test.
       ≤2 children855 (56.07)671 (64.96)
       >2 children670 (43.93)362 (35.04)
      Type of housing (%)0.0048
      Chi-square test.
      ,
      p<0.05 in multivariate analysis.
       House1290 (84.6)855 (82.8)(1.00)
       Apartment29 (1.9)7 (0.7)(2.16)
       Precarious lodgings206 (13.5)171 (16.6)(1.54)
      Water storage in the house (%)1465 (96.1)845 (81.9)<0.0001
      Chi-square test.
      ,
      p<0.05 in multivariate analysis.
      (5.00)
      Indoor piped public water supply (%)721 (47.3)380 (36.8)<0.0001
      Chi-square test.
      ,
      p<0.05 in multivariate analysis.
      (1.54)
      Connected to public sewer (%)959 (62.9)688 (66.6)0.0541
      Chi-square test.
      IgG, immunoglobulin G; SD, standard deviation.
      a The odds ratio (in parentheses) is provided for significant categorical multivariate results.
      * p < 0.05 in multivariate analysis.
      b t-test.
      c Chi-square test.

      3.6 Dengue serotype analysis

      Of 1525 IgG seropositive participants tested, 1511 had PRNT50 data available for all four serotypes. Of these, 1468 (97.2%) had antibody titres ≥10 for at least one serotype and 1205 (79.7%) had antibody titres ≥10 against all four serotypes. Nearly half (736/1511; 48.7%) had a multitypic antibody profile and 732/1511 (48.4%) had a monotypic profile. DENV-1, DENV-2, and DENV-3 were nearly equally represented among dominant serotypes in participants with monotypic profiles overall (Figure 3).
      Figure thumbnail gr3
      Figure 3Proportion of participants with PRNT50 results showing monotypic profiles, according to dominant serotype, by site and overall.

      3.7 Japanese encephalitis IgG seroprevalence

      Anti-JE IgG results were available for 2544 (98.2%) participants. Of these, 345 (13.6%; 95% CI 12.3–15.0%) participants were seropositive against JE, 327 (94.8%) of whom were also dengue seropositive (Table 2). JE seroprevalence ranged from 4.3% (95% CI 2.4–7.1%) in Kalyani to 20.5% (95% CI 16.2–25.3%) in Wardha.
      Table 2Prevalence of dengue and Japanese encephalitis IgG in participants with blood samples
      33/2591 total samples had no results available for dengue IgG; 47 were inconclusive and 405 were missing for JE IgG.
      Dengue IgG
      nPositiveNegative
      JE IgGn259115251033
      Positive34532718
      Negative1794801990
      JE, Japanese encephalitis; n, number of participants with results corresponding to the specified category.
      a 33/2591 total samples had no results available for dengue IgG; 47 were inconclusive and 405 were missing for JE IgG.

      4. Discussion

      These findings demonstrate a high intensity of dengue transmission among children in India; more than 50% of the children had been infected at least once by the age of 6 years, results which are broadly consistent with existing, limited dengue seroprevalence data for adults and children in Chennai and Hyderabad.
      • Rodriguez-Barraquer I.
      • Solomon S.S.
      • Kuganantham P.
      • Srikrishnan A.K.
      • Vasudevan C.K.
      • Iqbal S.H.
      • et al.
      The hidden burden of dengue and chikungunya in Chennai, India.
      • Oruganti G.
      • Dinaker M.
      • Tez K.S.
      • Kattula S.R.
      • Rajesh J.G.
      • Yeldandi V.V.
      • et al.
      High sero-prevalence of dengue IgG antibodies among healthy individuals in Andhra Pradesh, India.
      All four serotypes were found to circulate, varying by geographic location. Nearly half of all participants had a multitypic dengue antibody profile. Dengue IgG seroprevalence increased with age at all but one study site, consistent with age-related cumulative exposure to dengue.
      • Rodriguez-Barraquer I.
      • Solomon S.S.
      • Kuganantham P.
      • Srikrishnan A.K.
      • Vasudevan C.K.
      • Iqbal S.H.
      • et al.
      The hidden burden of dengue and chikungunya in Chennai, India.
      The exception in Bangalore could be related to epidemiological, behavioural, or environmental factors moderating exposure risk, such as the occurrence of large, infrequent outbreaks.
      The observed level of dengue exposure was comparable to that reported in other highly endemic countries of Southeast Asia and Latin America: 56.2% for 4–9-year-olds in Yogyakarta, Indonesia (1995–1996),
      • Graham R.R.
      • Juffrie M.
      • Tan R.
      • Hayes C.G.
      • Laksono I.
      • Ma'roef C.
      • et al.
      A prospective seroepidemiologic study on dengue in children four to nine years of age in Yogyakarta, Indonesia I. Studies in 1995-1996.
      65% for 11-year-olds in Rayong, Thailand (2010),
      • Rodriguez-Barraquer I.
      • Buathong R.
      • Iamsirithaworn S.
      • Nisalak A.
      • Lessler J.
      • Jarman R.G.
      • et al.
      Revisiting Rayong: shifting seroprofiles of dengue in Thailand and their implications for transmission and control.
      34.4% for under 7-year-olds to 70.5% in 14–16-year-olds from a primary health care facility in Sri Lanka (2013–2014),
      • Jeewandara C.
      • Gomes L.
      • Paranavitane S.A.
      • Tantirimudalige M.
      • Panapitiya S.S.
      • Jayewardene A.
      • et al.
      Change in dengue and Japanese encephalitis seroprevalence rates in Sri Lanka.
      53% for under 7-year-olds and 88% by the age of 13 years among primary school children in southern Vietnam,
      • Thai K.T.
      • Binh T.Q.
      • Giao P.T.
      • Phuong H.L.
      • Hung le Q.
      • Van Nam N.
      • et al.
      Seroprevalence of dengue antibodies, annual incidence and risk factors among children in southern Vietnam.
      and 35.7% and 52.2% for 5–9 and 10–14 years age groups, respectively, in two localities in Mexico in 2011.
      • Amaya-Larios I.Y.
      • Martinez-Vega R.A.
      • Mayer S.V.
      • Galeana-Hernandez M.
      • Comas-Garcia A.
      • Sepulveda-Salinas K.J.
      • et al.
      Seroprevalence of neutralizing antibodies against dengue virus in two localities in the state of Morelos, Mexico.
      A higher seroprevalence was observed in a study in Managua, Nicaragua (2001–2003), where 80% of enrolled children were seropositive by 5 years of age.
      • Balmaseda A.
      • Hammond S.N.
      • Tellez Y.
      • Imhoff L.
      • Rodriguez Y.
      • Saborío S.I.
      • et al.
      High seroprevalence of antibodies against dengue virus in a prospective study of schoolchildren in Managua, Nicaragua.
      Considering these similarities in exposure history, it might be expected that rates of symptomatic, reported dengue are similar in India and other countries. In fact, there are huge disparities: from 2007 to 2011, India reported an approximate average annual incidence of 1.4 cases/100 000 population,
      • National Vector Borne Disease Control Programme
      Dengue cases and deaths in the country since 2010.
      whereas case notifications in Cambodia, Malaysia, the Philippines, and Sri Lanka for 2011 were 119, 70.4, 134, and 135 per 100 000, respectively.
      • Arima Y.
      • Chiew M.
      • Matsui T.
      Epidemiological update on the dengue situation in the Western Pacific Region, 2012.
      • National Dengue Control Unit
      Sri Lankan Situation.
      Despite their significant and often multitypic infection history, very few participants in the present study reported a history of dengue. Similarly low reporting was observed in the recent household-based study in Chennai,
      • Rodriguez-Barraquer I.
      • Solomon S.S.
      • Kuganantham P.
      • Srikrishnan A.K.
      • Vasudevan C.K.
      • Iqbal S.H.
      • et al.
      The hidden burden of dengue and chikungunya in Chennai, India.
      in which 744/800 (93%) subjects were dengue IgG seropositive, but only 1% of participants reported a history of dengue. The present authors are unaware of virological or genetic factors that might disassociate infection history from the incidence of symptomatic disease; likely explanations include a lack of health-seeking for patients with apparent infection, lack of recognition of the disease, or misdiagnosis of dengue.
      • World Health Organization
      Global strategy for dengue prevention and control: 2012-2020.
      • Bhatt S.
      • W Gething P.
      • Brady O.J.
      • Messina J.P.
      • Farlow A.W.
      • Moyes C.L.
      • et al.
      The global distribution and burden of dengue.
      For these reasons, and because dengue surveillance reports are collected from only sentinel sites,
      • Shepard D.S.
      • Halasa Y.A.
      • Tyagi B.K.
      • Adhish S.V.
      • Nandan D.
      • Karthiga K.S.
      • et al.
      Economic and disease burden of dengue illness in India.
      it must be assumed that dengue burdens reported in the routine surveillance system represent only a fraction of symptomatic episodes.
      In the current study, serotype-specific analyses identified historical circulation of all four dengue virus serotypes at each site, with DENV-1 present in a high proportion of samples in New Delhi and DENV-3 in Kalyani. These serological findings in children are worrying: co-circulation of multiple serotypes is a population risk factor for severe dengue because it allows for sequential infection, and because secondary infection is a risk factor for severe disease.
      • Vaughn D.W.
      • Green S.
      • Kalayanarooj S.
      • Innis B.L.
      • Nimmannitya S.
      • Suntayakorn S.
      • et al.
      Dengue viremia titer, antibody response pattern, and virus serotype correlate with disease severity.
      • Burke D.S.
      • Nisalak A.
      • Johnson D.E.
      • Scott R.M.
      A prospective study of dengue infections in Bangkok.
      Multivariate risk factor analysis suggested relationships between water availability/storage practices and dengue infection risk, an association with biological plausibility due to the aquatic larval and pupal stages of the vector life-cycle. However, these results should be interpreted with caution because they were mainly driven by data from Kalyani and are thus highly susceptible to confounding with site-specific socio-demographic covariates, or other factors. After excluding Kalyani data from the multivariate analysis, only participant age remained significantly associated with dengue positivity. Furthermore, the determination coefficient (R2) of the model was lower than 0.1, confirming the limited predictive value of these variables for dengue serostatus, in agreement with inconclusive/variable socio-economic drivers of dengue serostatus identified in other studies. In Chennai, univariate logistic regression showed a negative association with household income and no associations with other household factors.
      • Rodriguez-Barraquer I.
      • Solomon S.S.
      • Kuganantham P.
      • Srikrishnan A.K.
      • Vasudevan C.K.
      • Iqbal S.H.
      • et al.
      The hidden burden of dengue and chikungunya in Chennai, India.
      Thai et al. (2005) found associations with littering in and around the home and the types of sanitary facilities in an initial univariate analysis, but these associations were not confirmed on multivariate analysis.
      • Thai K.T.
      • Binh T.Q.
      • Giao P.T.
      • Phuong H.L.
      • Hung le Q.
      • Van Nam N.
      • et al.
      Seroprevalence of dengue antibodies, annual incidence and risk factors among children in southern Vietnam.
      A community-based study of potential risk factors for dengue transmission in Venezuela found several household and socio-economic factors, including storing water and used tyres (univariate analysis), and crowding, household size, and living in a shack (multivariate analysis), to be associated with an increased risk of dengue infection.
      • Velasco-Salas Z.I.
      • Sierra G.M.
      • Guzman D.M.
      • Zambrano J.
      • Vivas D.
      • Comach G.
      • et al.
      Dengue seroprevalence and risk factors for past and recent viral transmission in Venezuela: a comprehensive community-based study.
      The seroprevalence of anti JE IgG antibodies was also measured in the participants in the present study. JE is endemic in some regions in India, particularly in the south and north-east;
      • Tiwari S.
      • Singh R.K.
      • Tiwari R.
      • Dhole T.N.
      Japanese encephalitis: a review of the Indian perspective.
      however, during the current study period, none of the study sites were considered to be within a JE endemic area and none were in an area subject to routine JE vaccination. The observed seroprevalence of JE in the current study (13.6%; 95% CI 12.3–15.0%) confirms circulation of the virus, but is lower than that reported in a number of other studies on JE seroprevalence in endemic countries,
      • Jeewandara C.
      • Gomes L.
      • Paranavitane S.A.
      • Tantirimudalige M.
      • Panapitiya S.S.
      • Jayewardene A.
      • et al.
      Change in dengue and Japanese encephalitis seroprevalence rates in Sri Lanka.
      • Conlan J.V.
      • Vongxay K.
      • Khamlome B.
      • Jarman R.G.
      • Gibbons R.V.
      • Fenwick S.G.
      • et al.
      Patterns of flavivirus seroprevalence in the human population of Northern Laos.
      • Chatterjee S.
      • Chattopadhyay D.
      • Bhattacharya M.K.
      • Mukherjee B.
      Serosurveillance for Japanese encephalitis in children in several districts of West Bengal, India.
      perhaps because these sites are located within less-endemic areas of India.
      Cross-reaction between anti-flavivirus IgG antibodies has been documented
      • Centers for Disease Control and Prevention
      Laboratory guidance and diagnostic testing.
      and cannot be excluded here from affecting the observed JEV or dengue IgG rates. The PRNT is a more specific assay and may be used to distinguish between cross-reactive and pathogen-specific responses. Encouragingly, 99% of dengue IgG-positive samples in the current study were also dengue PRNT-positive. However, the lack of JEV PRNT data and dengue PRNT in dengue-negative samples remains a limitation in conclusively addressing this risk.
      Other limitations include the use of different dengue-specific IgG ELISAs at different sites, but the sensitivity test showing 100% concordance was reassuring. In this study, a convenience sample was selected, including some low-income settings, to provide geographical spread across India. However, sites were not randomly selected, subjects were consecutively recruited, and these data cannot be considered nationally or locally representative. As with other epidemiological studies, recall bias (during the questionnaire) and selection bias (e.g., self-selection of healthy subjects) cannot be excluded. Despite these sampling limitations, these data provide a first multi-centre view on dengue seroprevalence in India. The use of a single protocol and consistent methods between the sites strengthens the validity of the data.
      In conclusion, high levels of dengue exposure were observed in Indian children, and age-stratified data describe transmission intensity at these locations. This information may inform dengue burden estimates and populate transmission models to assess the potential impact of prevention and control measures, including vaccination programmes.

      Author contributions

      SG, AC, and GF designed the study. SG, RS, NRRM, RCG, GRJ, EG, NS, MMS, and SO refined the study design, performed the study, and collected demographic and clinical data from study subjects, in the field. AC performed laboratory assays for dengue and JE virus in accordance with good laboratory practice. JN and AM analysed the data. JN provided an initial interpretation and worked to develop a manuscript outline. All authors interpreted the data with refinements and contributed to writing the manuscript. All authors critically reviewed the manuscript while in preparation and approved the final draft.

      Acknowledgements

      This study was sponsored by Sanofi Pasteur. The authors wish to thank the following investigators for their contributions to the conduct of the study: Suman Kumar, Gandhari Basu, Chitra Chatterjee, Ganesh Oruganti, Sathish Chandra MR, and Malatesh Undi. The authors also thank Shelke Pallavi who was involved in the conduct of the study and Claire Dakowski for statistical advice (Sanofi Pasteur), and Jo-Ann West for editorial assistance and manuscript coordination on behalf of Sanofi Pasteur. Editorial assistance with the preparation of the manuscript was provided by professional medical writers, Simon Lancaster and Juliette Gray of inScience Communications, Springer Healthcare, funded by Sanofi Pasteur. The wider DNG10 authorship group included Suman Kumar, Gandhari Basu, Chitra Chatterjee, Ganesh Oruganti, and Shelke Pallavi.
      Conflict of interest: JN, AM, and SO are employees of Sanofi Pasteur. AC, EG, MMS, RM, and RS have no conflicts of interest to declare.

      References

        • World Health Organization
        Global strategy for dengue prevention and control: 2012-2020.
        WHO, Geneva2012 (Available at: http://www.who.int/denguecontrol/9789241504034/en/ (accessed July 7, 2016))
        • Chakravarti A.
        • Arora R.
        • Luxemburger C.
        Fifty years of dengue in India.
        Trans R Soc Trop Med Hyg. 2012; 106: 273-282
        • Shepard D.S.
        • Halasa Y.A.
        • Tyagi B.K.
        • Adhish S.V.
        • Nandan D.
        • Karthiga K.S.
        • et al.
        Economic and disease burden of dengue illness in India.
        Am J Trop Med Hyg. 2014; 91: 1235-1242
        • Bhatt S.
        • W Gething P.
        • Brady O.J.
        • Messina J.P.
        • Farlow A.W.
        • Moyes C.L.
        • et al.
        The global distribution and burden of dengue.
        Nature. 2013; 496: 504-507
        • National Vector Borne Disease Control Programme
        Dengue cases and deaths in the country since 2010.
        National Vector Borne Disease Control Programme, Ministry of Health and Family Welfare, India2014 (Available at: http://nvbdcp.gov.in/den-cd.html (accessed May 27, 2016))
        • Farrington C.P.
        • Kanaan M.N.
        • Gay N.J.
        Estimation of the basic reproduction number for infectious diseases from age-stratified serological survey data.
        J R Stat Soc Ser C Appl Stat. 2001; 50: 251-292
        • Ferguson N.M.
        • Donnelly C.A.
        • Anderson R.M.
        Transmission dynamics and epidemiology of dengue: insights from age-stratified sero-prevalence surveys.
        Philos Trans R Soc Lond B Biol Sci. 1999; 354: 757-768
        • International Epidemiological Association
        Good epidemiological practice.
        International Epidemiological Association, London, UK2007 (Available at: http://ieaweb.org/good-epidemiological-practice-gep/ (accessed May 27, 2016))
        • Russell P.K.
        • Nisalak A.
        • Sukhavachana P.
        • Vivona S.
        A plaque reduction test for dengue virus neutralizing antibodies.
        J Immunol. 1967; 99: 285-290
        • Rodriguez-Barraquer I.
        • Solomon S.S.
        • Kuganantham P.
        • Srikrishnan A.K.
        • Vasudevan C.K.
        • Iqbal S.H.
        • et al.
        The hidden burden of dengue and chikungunya in Chennai, India.
        PLoS Negl Trop Dis. 2015; 9: e0003906
        • Oruganti G.
        • Dinaker M.
        • Tez K.S.
        • Kattula S.R.
        • Rajesh J.G.
        • Yeldandi V.V.
        • et al.
        High sero-prevalence of dengue IgG antibodies among healthy individuals in Andhra Pradesh, India.
        Indian J Public Health Res Dev. 2014; 5: 131-135
        • Graham R.R.
        • Juffrie M.
        • Tan R.
        • Hayes C.G.
        • Laksono I.
        • Ma'roef C.
        • et al.
        A prospective seroepidemiologic study on dengue in children four to nine years of age in Yogyakarta, Indonesia I. Studies in 1995-1996.
        Am J Trop Med Hyg. 1999; 61: 412-419
        • Rodriguez-Barraquer I.
        • Buathong R.
        • Iamsirithaworn S.
        • Nisalak A.
        • Lessler J.
        • Jarman R.G.
        • et al.
        Revisiting Rayong: shifting seroprofiles of dengue in Thailand and their implications for transmission and control.
        Am J Epidemiol. 2014; 179: 353-360
        • Jeewandara C.
        • Gomes L.
        • Paranavitane S.A.
        • Tantirimudalige M.
        • Panapitiya S.S.
        • Jayewardene A.
        • et al.
        Change in dengue and Japanese encephalitis seroprevalence rates in Sri Lanka.
        PLoS One. 2015; 10: e0144799
        • Thai K.T.
        • Binh T.Q.
        • Giao P.T.
        • Phuong H.L.
        • Hung le Q.
        • Van Nam N.
        • et al.
        Seroprevalence of dengue antibodies, annual incidence and risk factors among children in southern Vietnam.
        Trop Med Int Health. 2005; 10: 379-386
        • Amaya-Larios I.Y.
        • Martinez-Vega R.A.
        • Mayer S.V.
        • Galeana-Hernandez M.
        • Comas-Garcia A.
        • Sepulveda-Salinas K.J.
        • et al.
        Seroprevalence of neutralizing antibodies against dengue virus in two localities in the state of Morelos, Mexico.
        Am J Trop Med Hyg. 2014; 91: 1057-1065
        • Balmaseda A.
        • Hammond S.N.
        • Tellez Y.
        • Imhoff L.
        • Rodriguez Y.
        • Saborío S.I.
        • et al.
        High seroprevalence of antibodies against dengue virus in a prospective study of schoolchildren in Managua, Nicaragua.
        Trop Med Int Health. 2006; 11: 935-942
        • Arima Y.
        • Chiew M.
        • Matsui T.
        Epidemiological update on the dengue situation in the Western Pacific Region, 2012.
        Western Pac Surveill Response J. 2015; 6: 82-89
        • National Dengue Control Unit
        Sri Lankan Situation.
        National Dengue Control Unit, Ministry of Health, Sri Lanka2014 (Available at: http://www.dengue.health.gov.lk/index.php/information-on-dengue/sri-lankan-situation (accessed July 22, 2016))
        • Vaughn D.W.
        • Green S.
        • Kalayanarooj S.
        • Innis B.L.
        • Nimmannitya S.
        • Suntayakorn S.
        • et al.
        Dengue viremia titer, antibody response pattern, and virus serotype correlate with disease severity.
        J Infect Dis. 2000; 181: 2-9
        • Burke D.S.
        • Nisalak A.
        • Johnson D.E.
        • Scott R.M.
        A prospective study of dengue infections in Bangkok.
        Am J Trop Med Hyg. 1988; 38: 172-180
        • Velasco-Salas Z.I.
        • Sierra G.M.
        • Guzman D.M.
        • Zambrano J.
        • Vivas D.
        • Comach G.
        • et al.
        Dengue seroprevalence and risk factors for past and recent viral transmission in Venezuela: a comprehensive community-based study.
        Am J Trop Med Hyg. 2014; 91: 1039-1048
        • Tiwari S.
        • Singh R.K.
        • Tiwari R.
        • Dhole T.N.
        Japanese encephalitis: a review of the Indian perspective.
        Braz J Infect Dis. 2012; 16: 564-573
        • Conlan J.V.
        • Vongxay K.
        • Khamlome B.
        • Jarman R.G.
        • Gibbons R.V.
        • Fenwick S.G.
        • et al.
        Patterns of flavivirus seroprevalence in the human population of Northern Laos.
        Am J Trop Med Hyg. 2015; 93: 1010-1013
        • Chatterjee S.
        • Chattopadhyay D.
        • Bhattacharya M.K.
        • Mukherjee B.
        Serosurveillance for Japanese encephalitis in children in several districts of West Bengal, India.
        Acta Paediatr. 2004; 93: 390-393
        • Centers for Disease Control and Prevention
        Laboratory guidance and diagnostic testing.
        CDC, Atlanta, GA2010 (Available at: http://www.cdc.gov/dengue/clinicalLab/laboratory.html (accessed January 20, 2016))