We conducted a serological study of pandemic A (H1N1) 2009 in Selenghe province, Mongolia. Serological studies on pandemic A (H1N1) 2009 provide a unique opportunity to understand the transmission dynamics of influenza in the community. For seasonal influenza, high antibody cross-reactivity with previous infections and immunity induced by vaccines make it difficult to interpret serological data. Understanding the transmission of pandemic influenza virus is important to be prepared for future pandemic. Data such as overall cumulative incidence of infection and age-specific cumulative incidence of infection can provide useful information in being prepared for future pandemics.
In Mongolia, there was a unique epidemiological characteristic in the outbreak of pandemic A (H1N1) 2009. In many northern hemisphere countries, there were smaller outbreaks between May 2009 and August 2009, followed by larger outbreaks between September 2009 and early 2010.15,16 However, in Mongolia, no confirmed case of pandemic A (H1N1) 2009 occurred until October 12, 2009, despite an extensive search for cases with pandemic A (H1N1) 2009.17 Mongolia has a low population density, and traveling between Mongolia and other countries and within the country occurs much less than in other Asian countries.18 Such geographical and social characteristics may have contributed to the late introduction of pandemic A (H1N1) 2009 into the country. However, once the virus was introduced, it spread rapidly, and widespread outbreaks occurred almost simultaneously in different provinces, including the Selenghe province. A very intensive outbreak occurred in mid-November in the Selenghe province as shown in Figure 1. The present study reveals that approximately 30% of the total population in the Selenghe province was infected with pandemic A (H1N1) 2009 during the first wave of this pandemic. Our data also indicate that children aged 2–4 year and 5–9 years had higher cumulative incidences of infection than other age-groups. The serological study in England showed that the cumulative incidences of infection in those aged 5–14 years were highest and much higher than those aged <5 years.4 Another study in New Zealand also indicated higher cumulative incidences of infection in school-aged children (5–9 years) compared with infants (1–4 years).7 The fact that in the present study cumulative incidences of infection were high even in small children aged 2–4 years can be explained by different social mixing patterns in Mongolia. In countries such as the United States and the United Kingdom, schools have been implicated as a major source for community transmission of influenza.19,20 In particular, with regard to pandemic A (H1N1) 2009, older children aged 10–19 years were the main target for infections in the early stages of the epidemic in 2009.8 In Mongolia, most children aged 2 years and over go to kindergarten. Our previous analysis in Mongolia on seasonal influenza (2008–2009 season) and pandemic A (H1N1) (2009–2010 season) showed that ILI incidence was higher in children aged 1–4 years compared with those aged 5–9 years.17 In countries like Mongolia, where in most households both parents work and the preschool system is well established, kindergarten or nursery schools may have a more important role in spreading influenza viruses; thus, there are higher incidences in smaller children. Similarly, the seroprevalence among those aged 0–5 years was high in China.21 Social mixing patterns can be an important determinant for spread of infectious diseases including influenza.22,23 However, data on social mixing patters in Mongolia are not available. Another significant finding in the present study was a high cumulative incidence of infection in the elderly: 18·5% by both HI and MN for those aged 65 years or more. In other studies, estimated cumulative incidences of infection in the elderly were significantly lower than in younger age-groups. For example, the study in England showed that there was only a 0·9% difference between the baseline and after the first wave for those aged 65 years or more.4 Similarly, low cumulative incidences of infection in the elderly were seen in the United States,2 Australia,24,25 and New Zealand.7 Low cumulative incidences of infection in the elderly have been explained by preexisting cross-immunity in the elderly population before pandemic A (H1N1) 2009.2,4 In our study, there was no difference in pre-pandemic antibody prevalence and GMT between the elderly (those more than 65 years) and younger age-groups (Figure 2). Similar patterns were seen in rural farmers in Guangxi province, China, where the elderly over 60 years of age did not have any positive antibodies.12 Lack of preexisting antibodies in the elderly was also confirmed in another study in China, which tested samples from larger geographic areas including urban areas.21 There were also no preexisting antibodies in those aged between 42 and 82 years in Singapore.26 It is still unclear why many elderly had preexisting antibodies in North America and some European countries, while the preexisting antibody prevalence in this age-group was generally low in Asian countries.5 However, it should be noted that all our study subjects were <80 years of age except one aged 85 years. The study in Japan indicated that only people born in 1918 had preexisting neutralizing antibodies.27 In addition, a serological study in Finland showed that only those aged more than 80 years had significantly higher level of HI titers.3 We might have found a higher prevalence of preexisting immunity if we had tested more samples from those over the age of 80. Further studies should be conducted to compare preexisting cross-reactive antibody prevalence in different parts of the world.
We also analyzed the effects of preexisting antibodies on cumulative incidences of infection (Table 3). Those with a higher HI antibody titer (≥1:40) had significantly lower cumulative incidences of infection than those with a lower HI antibody titer (<1:40). Our results are comparable with the cohort study conducted in adults in Singapore, which also showed a protective effect of baseline HI titers of 1:40 or more.10 However, such significant protective effects were not observed for MN titers. Those with baseline MN titers ≥1:40 did not have significantly lower cumulative incidences of infection than those with MN titers <40. In general, MN titers are considered to be better indicators for assessing the protection level of influenza immunity.28,29 It was also shown that MN results are a better indicator than HI titers in assessing cumulative incidences of infection in laboratory-confirmed cases of pandemic (H1N1) 2009.30 It is not clear why our MN results were not associated with cumulative incidences of infection. Although both HI and MN tests used the same reference virus (A/California/07/2009pdm), the sensitivity of our MN test might have been low, especially in detecting low levels of cross-reacting antibodies.
There are some limitations to our study. The number of subjects in each group was relatively small, and this makes it difficult to analyze statistical differences between age-groups. We also included samples from only one province of Mongolia, namely Selenghe province, which may not represent the pattern of whole country. Samples collected from other provinces of Mongolia are now being analyzed to compare the data between provinces. In addition, the study participants took part in the study on a voluntary basis. This might be a potential source of bias, and the data from the study participants may not represent the whole population. Despite these limitations, our study is one of a few studies that analyzed paired serum samples and assessed actual seroconversion in each individual. It was also one of a few studies that was conducted in less developed countries.
In conclusion, the community-based serological cohort study based on paired sera revealed that approximately 30% of the population in Selenghe province of Mongolia had been infected during the first wave of pandemic A(H1N1) 2009. This overall incidence is comparable to previous studies in other countries. The cumulative incidence was higher in age-groups of 2–4 years and 5–9 years. The baseline antibody levels were similar in all age-groups including those >65 years.