This study documents for the first time the seroprevalence of measles-specific IgG antibodies among children in the CAR. The results suggest that population immunity to measles vaccine is well below the threshold (>90%) necessary to interrupt measles transmission. Protective antibody levels were more prevalent among children eligible for vaccination and among those reported to have been vaccinated; however, only 50% of vaccination-age children had protective antibodies. Passively derived maternal protection to measles vaccine among infants aged 0-8 months was also low (9.8%) and undetectable past 3 months of age, leaving this vulnerable population susceptible to wild virus circulation. The reasons for low maternal transfer of measles-specific antibody and the lower-than-expected population protection are not clear.
Low, waning maternal antibody protection, which results in an increased period of susceptibility to natural infection before vaccination eligibility, typically at 9 months of age, has been documented in various developing countries [4–9]. In a recent study in Bangladesh, only 50% of infants aged 1-3 months had protective antibodies and infants aged >3-9 months had no protection at all, similar to our findings . Two studies conducted in the past 10 years in sub-Saharan Africa reported similar results. In Nigeria, Hartter et al. found that only 17% of 4-month-old Nigerian infants were still protected against measles , while in Mali, only 30% of 2-month-old infants had protective antibody titers . Reporting results from Libreville, Gabon, more than 20 years ago, Grendel et al. found that 90% of children consulting outpatient services were seronegative by 4 months of age . During the same period, Dabis et al. found rapid rates of passive maternal antibody loss, from 96% at 2 months of age to less than 10% at 7-8 months of age in Brazzaville, Congo .
This body of research presents the accumulated descriptive epidemiological evidence on this topic. Unfortunately, there remains a dearth of analytic research in Africa to determine the role and relative importance of specific covariates associated with this phenomenon. Furthermore, the epidemiologic context of measles transmission has changed dramatically in the past decade . Accelerated measles control activities, including improved routine immunization coverage, the provision of a second dose of measles vaccine at supplementary vaccination activities and case-based surveillance with laboratory confirmation, have reduced measles-associated morbidity and mortality to record lows. Bolstered by these achievements, the African Region is poised to adopt an elimination goal. Our results are important, because they provide a rare glimpse into the serological profile of measles protection in Africa after concerted, targeted efforts to raise population immunity and reduce morbidity and mortality.
Factors known to affect the placental transfer of antibodies, such as HIV and malaria infection, may be associated with our study results . CAR has an estimated HIV seroprevalence rate of 6.2%, one the highest in central Africa . Attempts were made to ascertain the HIV status of potential study subjects; however, serological screening was not undertaken. Falciparum malaria transmission is haloendemic in much of the country, including the capital Bangui, the site of the study. Maternal histories of malaria episodes were not collected.
UNICEF estimates that 38% of the country's children suffer from global chronic malnutrition and 10% suffer from global acute malnutrition . While the effect of malnutrition on maternal transfer of passive antibody immunity is poorly understood, the scope and severity of the problem indicate that it cannot be ignored. The source of maternal immunity, wild virus versus vaccine-derived, may also affect the amount of maternal antibody transferred transplacentally, with vaccinated mothers passing on less immunity to their newborn infants. Although routine vaccination rates increased from 35% in 2003 to 62% in 2006, before this, routine vaccination with measles antigen remained consistently below 50%. For this reason, it is believed that the majority of maternal immunity is likely derived from exposure to circulating wild virus.
Our study also suggests lower-than-expected levels of measles-specific IgG in the vaccination-eligible population, and even a possible waning of serological immunity among children aged ≥10 years. This is another result that deserves further investigation. Reduced vaccine effectiveness, due to compromised vaccine or host characteristics that negatively affect seroconversion as well as reduced exposure to circulating measles virus may all play a role.
This study has some potential limitations. The presence of protective antibodies may have been misclassified because of the diagnostic test used. The selection of study subjects may also have biased our results. Although children with recent episodes of acute illness were ineligible for the study, some were nonetheless recruited from outpatient clinic services. In addition, the quota convenience sampling method used does not allow results to be extrapolated to the wider population. Determining the vaccination status of children and the reliability of reports is always problematic in the absence of the gold standard vaccination card. Thus, the vaccination status of study subjects may have been overestimated. Lack of verification of the vaccination status of the mothers is another weakness of this study.
National immunization programs must determine the optimal age for the first dose of MCV1 by weighing the risk for primary vaccine failure at younger ages against the risk for measles virus infection before vaccination. This study describes the wide window of vulnerability, between the disappearance of passively acquired maternal antibody and the eligible age of vaccination. The premature loss of maternal antibodies, which, if present, would inhibit vaccination seroconversion, opens the door to discussions of the optimal age for vaccination in developing countries.