Prevalence of Dengue Virus Infection in US Travelers Who Have Lived in or Traveled to Dengue‐Endemic Countries

Carolina Sanchez‐Vegas MD, Davidson H. Hamer MD, Lin H. Chen MD, Mary E. Wilson MD, Christine Benoit BA, Elizabeth Hunsperger PhD, William B. MacLeod ScD, Emily S. Jentes PhD, Winnie W. Ooi MD, Adolf W. Karchmer MD, Laura Kogelman MD, Emad Yanni MD, MSc, MA, Nina Marano DVM, Elizabeth D. Barnett MD
DOI: http://dx.doi.org/10.1111/jtm.12057 352-360 First published online: 1 November 2013

Abstract

Background Dengue virus (DENV) infections may occur in travelers.

Objectives To determine prevalence of anti‐DENV IgG antibody in travelers who lived in or visited dengue‐endemic countries and to describe risk factors and characteristics associated with infection and subsequent anti‐DENV IgG antibody presence.

Methods Participants were enrolled from travel clinics of the Boston Area Travel Medicine Network from August 2008 through June 2009. Demographic information, trip duration, travel history, and a blood sample were collected. Serum samples were tested for anti‐DENV IgG antibody by indirect IgG enzyme‐linked immunosorbent assay (ELISA), and antibody‐mediated virus neutralization by plaque reduction neutralization test (PRNT) for anti‐DENV IgG antibody‐positive and selected negative samples. Participants were stratified into group 1: born in dengue‐endemic countries; group 2: born in nonendemic countries but lived continuously for ≥1 year in a dengue‐endemic country; group 3: born in nonendemic countries and traveled to a dengue‐endemic country for ≥2 weeks but <1 year.

Results Six hundred travelers were enrolled. Anti‐DENV IgG antibody was identified in 113 (19%) when tested by ELISA (51% in group 1, 40% in group 2, and 6.9% in group 3) and in 71 (12%) by PRNT (42% primary monotypic and 58% heterotypic reactive responses). Sensitivity and specificity of the ELISA based on PRNT results were 85% to 100% and 79% to 94%, assuming up to 15% misclassification of ELISA negative results. Presence of anti‐DENV IgG antibody by ELISA was associated with years lived in dengue‐endemic countries and birthplace in the Caribbean for group 1, receipt of Japanese encephalitis vaccine in group 3, and self‐reported history of dengue in all three groups.

Conclusions Nineteen percent of participants who were born, lived in, or traveled to dengue‐endemic countries had anti‐DENV IgG antibody by ELISA; 12% had antibodies by PRNT, 85% of whom had no history of dengue. Presence of DENV antibodies was associated with years lived in dengue‐endemic countries and self‐reported history of dengue.

Dengue infection, caused by an RNA virus transmitted through the bite of Aedes mosquitoes, is endemic in most tropical and subtropical areas of the world. Dengue viruses (DENVs) are estimated to cause 50 to 100 million infections, 500,000 hospitalizations, and 22,000 deaths yearly.1 Among mosquito‐borne viral diseases in the world, dengue is spreading most rapidly with a 30‐fold increase in incidence in the last 50 years and expansion of its geographic range into new countries in both urban and rural settings.2

Dengue is a common cause of acute febrile illness in travelers. It has been reported as the most frequent specific diagnosis causing febrile illness in travelers returning from Asia, South America, and the Caribbean, and is a frequent cause of hospitalization in travelers returning from the tropics.3,4,5 Autochthonous cases of DENV infections have increased in the United States, as illustrated by the recent report of >40 DENV infections in Key West, Florida.6

There are four distinct serotypes of DENV. Infection with one serotype provides lifelong immunity to only that serotype. The spectrum of potential outcomes following DENV infection includes asymptomatic infection, nonspecific viral syndrome, and severe disease with hemorrhage, shock, and death.2 Severe dengue is more common during subsequent infection with a different DENV serotype.7,8

Studies from endemic areas suggest that 14% to 87% of all DENV infections are either asymptomatic or associated with nonspecific symptoms.9,10,11 Studies of presence of DENV antibodies in travelers with clinical symptoms consistent with dengue infection have shown prevalence ranging from 5% to 65%; few studies have assessed asymptomatic DENV infection in travelers.12–18 One study of 104 long‐term (≥3 months) travelers found a DENV antibody seroconversion rate of 7%, with a ratio of asymptomatic to symptomatic infection of 1 to 1.3; a second study of 447 short‐term travelers to Asia found an asymptomatic‐to‐symptomatic ratio of 3.3 : 1.17,18 A study of 1,207 travelers from the Netherlands tested before and after travel identified 6.5% of travelers with serologic evidence in the pre‐travel sample of prior DENV infection, and a 1.2% seroconversion rate in those without evidence of dengue antibody in the pre‐travel sample, for an incidence rate of 14.6 per 1,000 person‐months.19

Neither vaccine nor drugs are currently available for prevention of dengue, and there are no antiviral agents available for treatment. Because of increasing numbers of dengue cases in travelers, there is growing interest in improving education of health professionals and travelers about dengue and its prevention. Travel medicine providers may be asked specifically to provide advice about DENV infection for travelers who have a history of dengue or who have traveled extensively to dengue‐endemic areas. The purpose of this study was to determine prevalence of anti‐DENV IgG antibody in travelers who were born, lived in or visited dengue‐endemic countries and to compare demographic and trip characteristics associated with presence of antibody.

Methods

Participants and Procedures

Participants were enrolled from travel clinics in the Boston Area Travel Medicine Network (BATMN) from August 2008 through June 2009. BATMN is a research collaboration of five travel clinics in the greater Boston area, seeing approximately 7,500 travelers yearly in urban, suburban, academic, and university‐affiliated facilities. Participation in the study was offered to travelers who had spent at least 2 weeks in a dengue‐endemic country in the past.

Demographic information, relevant travel history including purpose of travel; previous diagnosis of dengue, West Nile fever, yellow fever (YF), Japanese encephalitis (JE); or previous vaccination against YF and/or JE, and a blood sample were collected from each participant. Serum samples were stored at −70°C for future testing.

Participants provided written consent. The study was approved by the Institutional Review Boards of all participating institutions and of the Centers for Disease Control and Prevention.

DENV Antibody Testing

Serum samples were tested for anti‐DENV IgG antibody using a commercial indirect IgG enzyme‐linked immunosorbent assay (ELISA, Focus Diagnostics, Cypress, CA, USA). The assay is nonquantitative and intended for use in establishing acute or previous DENV infection or for anti‐DENV IgG seroprevalence studies. Sensitivity and specificity listed by the manufacturer for asymptomatic past infection were 39 and 93%, respectively.20 Assays were performed in duplicate according to manufacturer's instructions.

Antibody‐mediated DENV neutralization was measured by plaque reduction neutralization test (PRNT) on all anti‐DENV IgG‐positive samples (including those previously vaccinated with YF or JE vaccines) and on a subset of anti‐DENV IgG‐negative samples selected to represent a range of optical densities on the anti‐DENV IgG ELISA and all sites involved in the study. Testing was done at the Centers for Disease Control and Prevention (CDC), Dengue Branch in San Juan, Puerto Rico, according to a protocol based on World Health Organization (WHO) guidelines.21 Titers were calculated based on reciprocal of the titer in which there was a 90% reduction in plaque counts (PRNT90) compared with the virus control. The PRNT assay was standardized within each test using positive controls, back titration, and negative controls, with each test validated based on number of plaques obtained in the positive control. Validation standards were used routinely for each test and for each virus tested in the PRNT to minimize inter‐assay variability. Anti‐DENV IgG‐positive PRNT results were interpreted as a primary monotypic response to a DENV infection when the serum sample neutralized against a single serotype, and as heterotypic antibody response when the sample neutralized multiple DENV serotypes with equivalent or similar reactivity, indicating secondary dengue infection with 2, 3, or 4 additional serotypes.

Sample Size and Data Analysis

Using an estimated anti‐DENV IgG antibody sero‐prevalence in the study population of 10%, a 2.5% margin of error, and 95% confidence intervals (CIs), we calculated a sample size of 554 participants. To address potential missing values for response variables, we increased the sample size by approximately 10%, to 600 participants.

Data were entered into a password‐protected database (CS Pro, US Census Bureau, Washington, DC). Data analysis involved calculation of proportions, prevalence ratios (PRs), and 95% CI. Differences in means were calculated by using a t‐test, and significance was calculated with the Wilcoxon Two‐Sample Test.

Subjects were stratified into three groups for analysis. Group 1 included those born in dengue‐endemic countries, defined as those listed by the WHO as dengue‐endemic at the time of the study.2 Group 2 included those born in nonendemic countries and had lived continuously for ≥1 year in a dengue‐endemic country. This criterion ensured that they had lived in these areas throughout all seasons of the year, accounting for seasonal DENV transmission. Group 3 included those born in nonendemic countries and had traveled to a dengue‐endemic country for ≥2 weeks but <1 year during any time of the year.

To evaluate possible cross‐reaction of YF or JE vaccines to antibody detected by the anti‐DENV IgG ELISA, analyses were performed first for all subjects, and then by excluding participants who had been immunized previously with YF or JE vaccines. Testing the samples for YF or JE antibody by PRNT was not possible because of resource limitations.

Predictors of anti‐DENV IgG antibody were determined based on results of ELISA and PRNT testing. For this analysis, only samples positive for anti‐DENV IgG antibody by PRNT were considered positive. Sera that tested anti‐DENV IgG‐positive by ELISA but negative by PRNT, those that tested negative by ELISA and PRNT, and those that tested negative by ELISA (but were not tested by PRNT) were all considered negative. Resource limitations precluded testing all ELISA‐negative samples by PRNT; therefore, we calculated ranges for sensitivity and specificity assuming misclassification of ELISA negative samples of up to 15%.

Results

Traveler Characteristics

Six hundred travelers were enrolled (Table 1); 280 were male (46.8%); ages ranged from 11 to 86 years (median 40 years); 63% were white, 14% Asian, 10% black, 6% Hispanic, and 7.5% biracial/multiracial/other/unknown. Seven subjects were excluded because of unknown country of birth and two because dengue results were unavailable. There were 140 travelers in group 1, 30 in group 2, and 421 in group 3.

View this table:
Table 1

Demographic and trip characteristics of 600 travelers tested for anti‐DENV IgG antibody

CharacteristicNumber (%)
Male gender (%)280 (47)
Age range (years)
Median age (years)
11–86
40
Country of birth (N = 593)
Dengue‐endemic country* (group 1)140 (23.7)
Dengue nonendemic country451 (76.3)
Lived for ≥1 year in dengue‐endemic country (group 2)30
Travel to dengue‐endemic countries (group 3)421
Self‐reported history of dengue16 (2.7)
Self‐reported prior yellow fever vaccination229 (40)
Self‐reported prior Japanese encephalitis vaccination25 (4.3)
Purpose of travel (most recent trip)
Tourism250 (41.8)
Visiting friends and relatives86 (14.4)
Business83 (13.9)
Education/research46 (7.7)
Volunteer31 (5.2)
Other purposes7 (1.2)
Missing purpose95 (15.9)
  • DENV = dengue virus; WHO = World Health Organization

  • * On the basis of Centers for Disease Control and Prevention and WHO map/list of countries at risk for dengue transmission 2007 (http://www.who.int/topics/dengue/en/).

  • Group 3 included those born in nonendemic countries who had traveled to a dengue‐endemic country for ≥2 weeks but <1 year.

Sixteen participants (2.7%) reported history of dengue, 5% in group 1, 7% in group 2, and 1% in group 3. None reported a history of West Nile virus, YF, or JE infection; 229 subjects (39%) reported immunization against YF and 25 (4.3%) against JE. The most common reason for travel for participants' most recent trip was tourism (41.8%).

Anti‐DENV IgG ELISA Results

Anti‐DENV IgG antibody was identified in 113 of 591 travelers (19%); antibody was present in 72 of 140 (51.4%) of those born in endemic countries (group 1) and in 41 of 451 (9%) of those born in nonendemic countries (groups 2 and 3) (PR, 5.65; 95% CI, 4.05–7.89). Among travelers in group 2, 12 of 30 (40%) were seropositive, compared with 29 of 421 (6.9%) travelers in group 3, who had shorter exposure periods in endemic countries (PR, 5.81; 95% CI, 3.31–10.18).

For those in group 1, we found a positive relationship between median number of years lived in the dengue‐endemic birth country and presence of antibody (31.2 years for anti‐DENV IgG‐positive vs 23.3 years for anti‐DENV IgG‐negative subjects, p < 0.05; Table 2). Compared with travelers born in Latin America, those born in Asia and the Caribbean were more likely to have anti‐DENV IgG antibody (PR, 2.0; 95% CI, 1.1–3.7 and PR, 2.1; 95% CI, 1.1–4.1, respectively).

View this table:
Table 2

Presence of antibody to DENV by ELISA and characteristics of travelers (all subjects)

Characteristics of travelersDengue IgG positive/number tested (%)Prevalence ratios (95% CI)
Group 1: Born in dengue‐endemic country (n = 140)72/140 (51.4)
Region of birth*
Africa12/31 (39)1.3 (0.6–2.6)
Asia39/63 (62)2.0 (1.1–3.7)
Caribbean13/20 (64)2.2 (1.1–4.0)
Latin America8/26 (31)1.0 (reference)
History of dengue
Yes7/8 (88)1.8 (1.3–2.4)
No65/132 (49)1.0 (reference)
History of YF immunization
Yes16/40 (40)0.7 (0.5–1.1)
No56/100 (56)1.0 (reference)
History of JE immunization
Yes1/3 (33)0.6 (0.1–3.2)
No71/137 (52)1.0 (reference)
Years lived in dengue‐endemic countryMedian (range)
Dengue IgG positive31.2 (19.1–43.2)p < 0.05§
Dengue IgG negative23.3 (11.8–36.3)
Group 2: Born in dengue nonendemic country—lived in dengue‐endemic country ≥1 year (n = 30)12/30 (40)
History of dengue
Yes2/2 (100)2.8 (1.7–4.6)
No10/28 (36)1.0 (reference)
History of YF immunization
Yes7/18 (39)0.9 (0.4–2.3)
No5/12 (42)1.0 (reference)
History of JE immunization
Yes3/4 (75)2.2 (0.99–4.7)
No9/26 (34)1.0 (reference)
Years lived in dengue‐endemic countryMedian (range)
Dengue IgG positive3.0 (1.75–6.50)1.7 (−0.7–4.1)
Dengue IgG negative2.8 (2.00–4.00)
Group 3: Born in dengue nonendemic country—traveled to dengue‐endemic country (n = 421)29/421 (6.9)
History of dengue
Yes3/5 (60)9.6 (4.3–21.5)
No26/416 (6)1.0 (reference)
History of YF immunization
Yes17/168 (10)2.1 (1.1–4.4)
No12/253 (5)1.0 (reference)
History of JE immunization
Yes4/18 (22)3.6 (1.4–9.2)
No25/403 (6)1.0 (reference)
Total weeks of travel to dengue‐endemic countryMedian (range)
Dengue IgG positive12.2 (8.8–24.9)3.4 (−5.5–12.4)
Dengue IgG negative7.7 (3.8–16.0)
  • CI = confidence interval; DENV = dengue virus; ELISA = enzyme‐linked immunosorbent assay; JE = Japanese encephalitis; YF = yellow fever.

  • * Africa: Angola, Burkina Faso, Comoros, Congo, Congo (Dem. Rep.), Djibouti, Equatorial Guinea, Ethiopia, Ghana, Guinea, Guinea‐Bissau, Ivory Coast, Kenya, Madagascar, Mozambique, Nigeria, Senegal, Seychelles, Sierra Leone, Somalia, South Africa, Sudan, and Tanzania. Asia: Bangladesh, Bhutan, Brunei, Burma (Myanmar), Cambodia, China, Hong Kong, India, Indonesia, Laos, Malaysia, Maldives, Mauritius, Nepal, Pakistan, Papua New Guinea, Philippines, Singapore, Sri Lanka, Taiwan, Thailand, and Vietnam. Caribbean: Anguilla, Antigua and Barbuda, Bahamas, Barbados, Dominica, Grenada, Haiti, Jamaica, Martinique, Montserrat, St. Kitts and Nevis, St. Lucia, St. Vincent & the Grenadines, Trinidad & Tobago, Turks & Caicos, Virgin Is. (British), Virgin Is. (US), and Puerto Rico. Latin America: Argentina, Belize, Bolivia, Brazil, Columbia, Costa Rica, Cuba, Dominica Rep., Ecuador, El Salvador, French Guiana, Guatemala, Guyana, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Suriname, and Venezuela. Middle East: Saudi Arabia. Oceania: Australia, French Polynesia, Guam, Kirbati, Marshall Islands, Micronesia, Nauru, New Caledonia, New Zealand, Palau, Samoa, Solomon Islands, Tonga, and Tuvalu.

  • Statistically significant difference, p < 0.05.

  • Indicates mean differences.

  • § Statistically significant difference, p < 0.05, Wilcoxon two‐sample test.

For those in group 2, there was no relationship between years lived in endemic countries and presence of anti‐DENV IgG antibody. For group 3 travelers, presence of anti‐DENV IgG antibody was related neither to number of weeks spent in dengue‐endemic countries nor to purpose of travel.

Reported receipt of YF or JE vaccine was not associated with the presence of anti‐DENV IgG antibody in travelers in groups 1 (PR, 0.7; 95% CI, 0.5–1.1 and PR, 0.6; 95% CI, 0.1–3.2, respectively) or 2 (PR, 0.9; 95% CI, 0.4–2.3 and PR, 2.2; 95% CI, 0.99–.7, respectively). In group 3, seropositive were more likely than seronegative travelers to have received YF (PR, 2.1; 95% CI, 1.1–4.4) or JE (PR, 3.6; 95% CI, 1.4–9.2) vaccines.

Twelve of 113 subjects (10.6%) with anti‐DENV IgG antibody and 3 of 478 without antibody (0.6%) reported history of dengue. Antibody presence was significantly associated with history of dengue for all three groups (Table 2).

Anti‐DENV IgG ELISA Results for Those Without History of YF or JE Immunization

For the 351 participants (59%) who received neither YF nor JE vaccine, anti‐DENV IgG was present in 55 of 100 (55%) of those in group 1, in 4 of 10 (40%) of those in group 2, and in 10 of 241 (4%) of those in group 3 (Table 3). Among group 1 participants, presence of anti‐DENV IgG was associated with history of dengue and years lived in a dengue‐endemic country. History of dengue was associated with presence of anti‐DENV IgG for those in group 3 (PR, 14.0; 95% CI, 14.0–50.6). For those born in dengue‐endemic countries there was a trend toward association with birth in the Caribbean and Asia (PR, 2.05; 95% CI, 1–4.1 and PR, 1.71; 95% CI, 0.9–3.4, respectively) and presence of anti‐DENV IgG when compared with those born in Latin America.

View this table:
Table 3

Presence of antibody to DENV by ELISA and characteristics of travelers (those with no history of YF or JE vaccination only, n = 351)

Characteristics of travelersDengue IgG positive/total number (%)Prevalence ratio (95% CI)
Group 1: Born in dengue‐endemic country (n = 100)55/98 (56)
Region of birth
Africa4/10 (45)1.13 (0.4–3.1)
Asia32/53 (60)1.71 (0.9–3.4)
Caribbean13/18 (67)2.05 (1.0–4.1)
Latin America6/17 (35)1.0 (reference)
History of dengue
Yes5/6 (83)1.5 (1.0–2.3)*
No50/92 (54)1.0 (reference)
Years lived in dengue‐endemic country. Median (range)
Dengue IgG positive27.8 (18.5–43.2)12.4 (3.4–21.4)
Dengue IgG negative19.4 (5.8–28.3)
Group 2: Born in dengue nonendemic country—lived in dengue‐endemic country >1 year (n = 10)4/10 (40)
History of dengue
Yes0/0 (0)NA
No4/10 (40)
Years lived in dengue‐endemic country. Median (range)
Dengue IgG positive3.5 (2.5–7.5)2.0 (−2.3–6.3)*
Dengue IgG negative2.5 (2.0–4.0)
Group 3: Born in dengue nonendemic country—traveled to dengue‐endemic country (n = 241)10/241 (4)
History of dengue
Yes1/1 (100)26.7 (14.0–50.6)
No9/240 (4)1.0 (reference)
Total weeks of travel to dengue‐endemic country. Median (range)
Dengue IgG positive9.2 ( 4.3–10.0)0.4 (−16.3–17.2)*
Dengue IgG negative6.0 (3.0–12.8)
  • CI = confidence interval; DENV = dengue virus; ELISA = enzyme‐linked immunosorbent assay; JE = Japanese encephalitis; NA = not applicable; YF = yellow fever.

  • * Statistically significant difference, p < 0.05.

  • Indicates mean differences.

  • Statistically significant difference, p < 0.05, Wilcoxon two‐sample test.

Anti‐DENV IgG PRNT Results

One hundred and nine samples that tested positive and 13 that tested negative for anti‐DENV IgG antibody by ELISA were tested by PRNT; insufficient serum volume prevented testing of four additional positive samples. For ELISA‐positive samples, 71 of 109 (65%) were anti‐DENV IgG‐positive by PRNT; 29 of 71 (41%) had primary monotypic anti‐DENV response and 42 of 71 (59%) had a heterotypic response to multiple DENV‐types. No ELISA‐negative samples tested positive by PRNT. Anti‐DENV IgG prevalence by PRNT was 12%. Sensitivity and specificity of the anti‐DENV IgG ELISA based on PRNT results, and assuming up to 15% misclassification of negative ELISA results, were calculated to be 85% to 100% and 79% to 94%, respectively.

Five hundred and eighty‐seven participants were included in testing for predictors of anti‐DENV IgG antibody using PRNT results. The highest proportion of participants testing anti‐DENV IgG‐positive by PRNT came from group 1 (42%), followed by those in groups 2 (23%) and 3 (1.2%) (Table 4). For group 1, presence of anti‐DENV IgG was associated with birth in the Caribbean and history of dengue. For every 5 years of residence in a dengue‐endemic country, the number of anti‐DENV IgG‐positive participants increased significantly (PR, 1.23; 95% CI, 1.06–1.42, Figure 1). For group 2, presence of anti‐DENV IgG was associated only with history of dengue, and for group 3, with history of both dengue infection and JE immunization. For groups 1 and 2, those with cross‐reactive antibody responses to multiple DENV serotypes had lived longer in dengue‐endemic countries (mean 36 and 9 years, respectively) compared with those who had monotypic (mean 26 and 1.5 years, respectively) or negative results (24 years and 3 years, respectively).

View this table:
Table 4

Presence of antibody to DENV by PRNT and characteristics of travelers (all subjects)

Characteristics of travelersDengue IgG positive/total number (%)Prevalence ratio (95% CI)
Group 1: Born in dengue‐endemic country (n = 139)57/137 (42)
Region of birth
Africa8/31 (26)01.0 (0.4–2.3)
Asia30/60 (50)1.9 (0.8–3.7)
Caribbean12/20 (60)2.2 (1.1–4.5)*
Latin America7/26 (27)1.0 (reference)
History of dengue
Yes7/8 (89)2.3 (1.6–3.2)*
No50/129 (38)1.0 (reference)
History of YF immunization
Yes10/40 (25)0.5 (0.3–0.9)
No47/97 (49)1.0 (reference)
History of JE immunization
Yes0/3 (0)0.3 (0.1–3.9)
No57/134 (43)1.0 (reference)
Years lived in dengue‐endemic countryMedian (range)
Dengue IgG positive30.2 (19.9–43.2)7 (−0.5–14.5)
Dengue IgG negative24.6 (15.5–37.3)
Group 2: Born in dengue nonendemic country—lived in dengue‐endemic country ≥1 year (n = 30)7/30 (23)
History of dengue
Yes2/2 (100)5.6 (2.5–12.4)*
No5/28 (18)1.0 (reference)
History of YF immunization
Yes5/18 (28)1.7 (0.4–7.2)
No2/12 (17)1.0 (reference)
History of JE immunization
Yes1/4 (25)1.1 (0.2–6.8)
No6/26 (23)1.0 (reference)
Years lived in dengue‐endemic countryMedian (range)
Dengue IgG positive2.0 (1.5–11.0)1.8 (−1.0–4.6)
Dengue IgG negative3.0 (2.0–4.2)
Group 3: Born in dengue nonendemic country—traveled to dengue‐endemic country (n = 420)7/420 (1.2)
History of dengue
Yes2/5 (40)33.2 (8.3–132.3)*
No5/415 (1)1.0 (reference)
History of YF immunization
Yes4/168 (2)2.0 (0.5–8.8)
No3/252 (1)1.0 (reference)
History of JE immunization
Yes2/17 (12)9.5 (1.9–45.4)*
No5/403 (1)1.0 (reference)
Total weeks of travel to dengue‐endemic countryMedian (range)
Dengue IgG positive11.7 (10.0–20.8)−0.37 (−18.1–17.4)
Dengue IgG negative8.0 (4.0–16.4)
  • CI = confidence interval; DENV = dengue virus; JE = Japanese encephalitis; YF = yellow fever.

  • * Statistically significant difference, p < 0.05.

  • Statistically significant difference, p < 0.05, Wilcoxon two‐sample test.

  • Indicates mean differences and 95% CI.

Figure 1

Proportion of participants with anti‐dengue virus (DENV) IgG positive by plaque reduction neutralization test (PRNT) by years lived in country of birth (dengue‐endemic countries only).

Discussion

In our study population, 19% of all travelers who lived in or traveled to dengue‐endemic countries had anti‐DENV IgG antibody when tested by ELISA. Results by PRNT testing showed that 12% had DENV‐specific antibodies; 85% of those had no prior history of dengue. Risk of severe dengue for travelers with history of dengue is poorly understood, although travel medicine clinicians are not infrequently asked to advise travelers about whether they face increased risk and if they need to take any special precautions based upon history of dengue. Should new data about risk of severe dengue following an initial infection become available, these study data may help clinicians identify travelers who would benefit most from this information.7,8

Data are limited on risk of severe dengue in travelers. From 2003 to 2005, the European Network Surveillance of Imported Infectious Diseases (TropNetEurop) reported 219 DENV symptomatic infections in returning travelers.22 Secondary immune response was associated with more severe disease manifestations in these travelers. During 1999 to 2002, TropNetEurop also reported 483 cases of probable dengue, of which 2.7% had dengue hemorrhagic fever (DHF); DHF was 4.3 times more common in immigrants and in travelers visiting friends and relatives in endemic areas than in general travelers.23

Identification of anti‐DENV IgG antibody by ELISA was associated with longer residence in a dengue endemic‐country. Higher seroprevalence was also detected in those born in the Caribbean and Asia, consistent with a recent report on travel‐associated dengue surveillance during 2006 to 2008, in which the highest proportion of laboratory‐confirmed cases among US travelers with known travel histories was in persons who traveled to the Dominican Republic (20%), Mexico (9%), and India (7%).24 These regional differences should be interpreted with caution, however, as this study included participant travel histories spanning many years during which there were significant changes in global epidemiology of dengue.

Currently, there are no specific recommendations to guide health professionals who provide advice to travelers who may have had DENV infection in the past. Our data described characteristics of travelers that may be associated with increased likelihood of having anti‐DENV antibodies. For those born in dengue‐endemic countries, characteristics associated with increased risk may include duration of residence in those countries, history of dengue, and birth in the Caribbean or Asia. Travelers born in nonendemic countries, especially those who have resided in an endemic country for an extended period of time and those who report a history of dengue, may also be at increased risk of severe dengue.

Pre‐travel consultations by health care providers should include discussion of DENV infection risk and vector avoidance. Patient education should include information about daytime biting habits and preference of DENV vector mosquitoes for indoor locations, especially in dark, cool places such as closets, bathrooms, behind curtains, and under beds. Pre‐travel counseling should include information about whether the destination will be in a dengue‐endemic area, and advice about seeking health care for febrile illnesses occurring during or after travel.

This study had several limitations. Self‐reported histories of illness and immunization are subject to recall bias; we were unable to verify either diagnoses of dengue or other flavivirus illness or receipt of YF or JE vaccines. We could have underestimated the number of subjects with past symptomatic infections, especially in the group who spent their childhood in endemic countries. The duration of exposure of those who lived and worked in dengue‐endemic countries and the extensive travel histories of the relatively small number of travelers born in nonendemic countries who were positive for anti‐DENV antibodies also limited the number of comparisons that could be made and conclusions that could be reached about countries associated with anti‐DENV antibody presence.

Testing for anti‐DENV IgG antibody by ELISA is limited by cross‐reactivity with antibody from other flavivirus infections, especially West Nile, YF, JE, and St. Louis encephalitis virus, and by YF and JE antibody produced by immunization.25 Testing by PRNT, undertaken to confirm ELISA results, indicated that ELISA testing overestimated the proportion of travelers with prior DENV infection, especially for subjects in groups 2 and 3. We were unable to test for YF, JE, or West Nile antibody to evaluate possible cross‐reactivity.

Serologic diagnosis of DENV infection is challenging. Detection of antibody based on capture IgM and IgG ELISA has become the standard for detection of and differentiation of primary and secondary acute DENV infections. Many commercial kits with good sensitivity and specificity are available.26 Commercially available immunoassays for detecting DENV‐specific IgM and IgG antibodies have been evaluated.27,28 One IgG enzyme immunoassay from Focus Diagnostics (former MRL laboratories) that was used in our study was evaluated and found to have sensitivity of 100% and specificity of 88% for diagnosis of acute DENV infection.

Schwartz and colleagues measured the potential for false‐positive serologic results for DENV in travelers vaccinated with YF and/or JE vaccines and concluded that specificity of an IgG ELISA assay was lower in travelers who received YF or JE vaccines than in those who did not.29 In our study, YF immunization was associated significantly with presence of DENV antibody as measured by PRNT only for group 1 subjects and JE immunization for group 3 subjects.

Another potential study limitation relates to assumptions made about PRNT results of ELISA‐negative samples. Because we were unable to test all ELISA‐negative samples by PRNT, we assumed that some samples that were antibody negative by ELISA but not tested by PRNT may have been misclassified, and then calculated the ranges for sensitivity and specificity.

Finally, definition of endemic countries for dengue was based on 2007 WHO data. Given the retrospective nature of the study and variable presence of DENV throughout a country and over time, some travelers' exposure risk might have been misclassified.

Conclusions

Individuals born in dengue‐risk countries and travelers with self‐reported history of dengue were more likely to have IgG antibody to DENV than short‐term visitors or those without history of dengue. Presence of anti‐DENV IgG antibody was associated with increased number of years lived in dengue‐endemic countries. Dengue prevention efforts should be directed at all travelers planning to visit dengue‐endemic countries, including specific advice about prevention of mosquito bites during the day and indoors.

Declaration of Interests

L. H. C. has received royalties from Wiley‐Blackwell and honorarium from Shoreline Inc. The other authors state they have no conflicts of interest to declare.

Acknowledgments

The authors acknowledge the contributions of M. Geary, N. Soodoo, E. Gleva, D. Gannon, R. Dufur, M. Bhussar, M. del Milagro Sosa, and M. Beltran.

This study was supported by Cooperative Agreement U19CI000508‐01 from the Centers for Disease Control and Prevention (CDC) to Boston Medical Center and by a research grant from Sanofi Pasteur.

References

View Abstract