Influenza in long-term Dutch travelers in the tropics: symptoms and infections

Influenza is the most frequently reported vaccine-preventable disease affecting visitors to (sub) tropical countries [1]. In 2009, an increase in the proportion of travel-related respiratory disease in Europe, mostly attributable to pandemic A(H1N1) influenza, was recorded [2]. The impact of influenza virus infection on travelers in terms of morbidity, healthcare accessed and hospitalization, is well documented [3] in short-term holiday makers, business travelers, and other traveler groups (Hadj pilgrims [4], cruise-ship passengers [5]). Travelers are also an acknowledged risk for onward transmission while in transit and at home [6]. Western adults are traveling in unprecedented numbers to emerging economies (often tropical and subtropical regions), for prolonged periods for ‘gap years’, to work, study, and tour. In this study of Dutch long-term travelers (3 to 12 months) to (sub) tropical regions, we prospectively document the attack rate (AR) of influenza infection and the incidence of influenza like illness (ILI) while traveling, and identify risk factors associated with symptomatic and asymptomatic infection from 2009 to 2011.

A prospective mono-centre study of immunocompetent, Dutch travelers aged ≥18 to 64 years was conducted at the Public Health Service travel clinic in Amsterdam from December 2008 to September 2011. All clients planning to travel to any (sub) tropical country in Africa, Central America, the Caribbean, South America or Asia for ≥12 and ≤52 weeks were invited to participate. Pre-travel, clients were interviewed by a nurse or physician and asked about travel purpose (work/study or tourism), travel duration, planned destination(s), and demographic details. No influenza vaccination history was established, but in the Netherlands, influenza vaccination is not recommended for travelers or healthy, young adults. At the time of the study, vaccination was only recommended for people ≥ 65 years and people with certain chronic diseases. Participants were given a digital thermometer (Huikeshoven Medical, Tiel, The Netherlands) and asked to take their temperature if they felt feverish while traveling. They kept a structured, weekly travel diary, recording their itinerary, symptoms experienced including respiratory, gastrointestinal and other systemic complaints and physician visits while ill. Diaries were completed on paper or digitally. Travelers received a weekly email reminder and were seen 2 to 6 weeks after return. Blood samples were taken before and after travel and paired samples were tested simultaneously post-travel for antibodies against influenza viruses using the hemagglutination-inhibition (HI) assay [7, 8]. Vaccine strains used represented the epidemic strains circulating worldwide during the study period: A(H1N1pdm09), A(H3N2brisbane), A(H3N2perth), A(H1N1brisbane), B/Florida and B/Brisbane [9]. Influenza infection was defined as a ≥4-fold rise in antibody titre post-travel from a pre-travel titre of ≥10, or if pre-travel titer was <10 titer, a post-travel titer of ≥40. Fever was defined as thermometer-confirmed body temperature >38.0 °C and influenza like illness (ILI) was defined as fever and cough or sore throat in the absence of other diagnoses. Primary regions visited while travelling were categorized into Influenza Transmission Zones [10] and then into 4 regions: South East Asia, Africa, Central and Latin America and “Asia (Other)” (Fig. 1).

Differences in the proportion of confirmed influenza infections by sex, age-group, primary influenza transmission zone visited, fever >38.0 °C and ILI were tested using Pearson’s chi-squared for each influenza strain. Positive predictive value (PPV) for seroconversion was the proportion of symptomatic cases seroconverting for infection. The association between infection and ILI was tested using logistic regression and odds ratios with 95 % confidence intervals. Incidence of fever and ILI while travelling was estimated as 1st episodes of fever or ILI per 100 person-weeks of travel with Poisson distributed 95 % confidence intervals (STATA version 13).

In this young cohort of long-term travelers to (sub) tropical countries (2009 to 2012), the attack rate for confirmed influenza virus infection was 15 %. This is higher than the 1–7 % found in other, mainly short-term traveler studies [1, 12, 13] and closely mirrors seasonal rates of infection in the (unvaccinated) general population in temperate [14] and tropical [15] regions during the same period, reflecting the prolonged travel duration. For symptomatic infection, we found attack rates of 6.3 % (fever alone) and 2 % (ILI), respectively, which was higher than in two short-term studies that found attack rates of 0.9–1.3 % for confirmed infections with fever alone [1, 12] and 0.8 % for ILI [12]. The results of symptomatic infections however, should be interpreted with care. First, some studies use different definitions than ILI for symptomatic infections and do not report fever as the only symptom of confirmed infection [13]. Second, fever is a very common symptom in travelers [1, 12]. At the same time, confirmed infections without fever were found in 32 and 83 %, respectively [1, 12]. Because virological data from nasopharyngeal sampling was not available for any of the studies, we can not be sure whether the symptoms were caused by influenza or by other infections. In our study, both the proportion of travelers with fever (35 %) and the proportion of travelers with confirmed infections without fever (58 %) were high. Therefore we did not calculate IR’s for symptomatic influenza infections.

As shown elsewhere [15], travelers in their mid-20s were more susceptible to A(H1N1pdm09) infection than other viruses. The majority of respondents were immune to at least one virus pre-travel, and 1/5th departing before the pandemic was declared were positive for A(H1N1pdm09) as early as January 2009. Forty percent of travelers were infected with >1 virus while traveling and 10 people experienced a reinfection. Travelers infected with A(H1N1pdm09) during travel were more likely to be symptomatic with ILI. Overall, the positive predictive value of ILI was 38 %. We found no evidence of seasonality or destination-specific risk, and travel was not a risk factor for severe disease.

There were some limitations to our study. The pre-travel influenza vaccination status was not confirmed. However, in the Dutch healthcare system, it is very uncommon that healthy, young travelers are vaccinated. Another possible limitation is that travelers only measured their temperature if they ‘felt feverish’ rather than daily. This may have lead to an underestimate of the proportion of travelers with fever, and an overestimate of the number of asymptomatic infections.

Vaccination for travelers against influenza has been discussed [16]. In the USA, influenza vaccination is recommended for all US residents aged ≥6 months [17]. The USA guidelines for international travel (Yellow Book, 2016) say ‘any traveler who wants to reduce the risk for influenza infection should consider influenza vaccination ≥2 weeks before departure if they plan to travel to the tropics’. According to WHO, influenza vaccination should be part of the routine immunization program for international travelers belonging to a risk group, in particular during influenza seasons [18].

In the Netherlands, only people at risk for complications are adviced influenza vaccination. Because there is no evidence that travel is an additional risk factor for influenza, the Dutch guidelines for travelers health advice do not recommend influenza vaccination for travelers other than those at high risk of complications.

We found that the attack rate of influenza in long-term travelers closely mirrors seasonal rates in the general population in temperate and tropical regions and travel was not a risk factor for severe disease. Influenza vaccination pre-travel is therefore most important for people at risk of medical complications due to influenza.