In the present study, we demonstrate that anti-CMV IgG in pregnant women can efficiently transfer to their fetuses and the maternal anti-CMV IgG in the majority of newborns is higher than that in their mothers. Moreover, we validate that primary CMV infection in children in China mainly occurred during 1–3.5 months of age.
Transplacental transfer of maternal IgG to the fetal bloodstream is mediated by neonatal Fc receptor in syncytiotrophoblasts of the placenta and contributes to the passive immunity of newborns to pathogens. Maternal IgG antibodies in full-term newborns are usually higher than those in their mothers
. In agreement with previous studies
[10, 11], we found that anti-CMV IgG in mothers transferred to the fetuses and anti-CMV IgG in most cord sera exceeded the maternal levels. Furthermore, in the present study, we found that the seroprevalence of anti-CMV IgG was much higher in young infants (<6 months old) as compared to older children (Table
4), which is likely to be related to the persistence of maternal antibodies in this group. Similarly, the intermediate or high avidity antibodies detected in infants <3 months of age may be explained by the transplacental transfer of maternal antibodies.
Transplacentally acquired maternal antibodies may protect infants against diseases in the early period of life. In the present study, despite maternal anti-CMV IgG in infants, primary CMV infection occurred early in childhood, indicating that maternal anti-CMV IgG can not fully protect against CMV infection. On the other hand, in spite of being infected, the children showed no symptoms related with CMV infection. Thus, maternal anti-CMV IgG in infants may provide substantial protection against symptomatic diseases or sequelae. This is similar to the prophylactic purpose of hepatitis A vaccination among children
In developing countries, the incidence of primary CMV infection has been reported to peak predominantly in the first years of life
[13, 14]. However, in developed countries, it occurs not only nearly at the end of childhood
, but also in women of child-bearing age
. In our study, based on the kinetics of anti-CMV IgG (Table
2), the seroconversion of anti-CMV IgM (Table
3) and the results of CMV IgG avidity assay in children with longitudinal sera, we concluded that primary CMV infection in children mainly occurred before 3.5 months of age, which is earlier than the infection age identified before in developing countries
[13, 17]. For the children in the second group, only cross-sectional serum samples were available. It is possible that some CMV-infected children did not develop detectable specific antibodies. Therefore, some antibody negative infants may actually have been infected. However, we consider that the proportion of CMV IgG negative infected children should be very small since the CMV IgG positive rates in children at same age were similar in the retrospective and cross-sectional groups (Table
In the present study, the seroprevalence in the children aged 0–8 years was around 80% and generally constant from 2 to 8 years old (Table
4), which indicates that there may be little CMV infection taking place between 2 and 8 years of age in this population. Also, we found that 96.4% of the pregnant women were anti-CMV IgG positive, which is in agreement with the rate of anti-CMV IgG (93–98%) in child-bearing age women in China
[17, 18]. The difference of the two rates may be due to the variation in populations with different ages since CMV infection through sexual contact is likely to increase seroprevalence by the time of adulthood. Alternatively, the prevalence in the children may reflect the positive rate in child-bearing age women in the future, since social economic situation and hygienic status in China have been greatly improved. However, with no previously reported prevalence of anti-CMV IgG in Chinese children, it is difficult to determine which is more reasonable to explain the reduced prevalence in the children. Therefore, further research is needed to clarify this issue.
Our findings that most CMV infection in children occurred before 3.5-month old also suggest that the CMV infection source in children is mainly from mothers or other family members since such young infants rarely contact others in China. The results that one child of the mother with negative anti-CMV IgG was persistently negative for anti-CMV IgG also support the assumption. Indeed, CMV in body fluids, such as breast milk may be associated with the risk and severity of postnatal CMV infection
There are several limitations in the present study. First, we defined the CMV infection in infants based on the dynamic changes of anti-CMV IgG, which was quantitatively tested only using the Dia.Pro test system. Since the dynamic ranges of different IgG ELISAs are highly different, it would have been helpful to compare the results of the Dia.Pro test system with other frequently used ELISA system. However, in our study, the sera from the mother-newborn pairs and longitudinal follow-up were tested in parallel in each assay. Additionally, the inter- and intra- assay variations were well within acceptable limits. Second, the gold standard for the diagnosis of CMV infection in early life is to detect the virus in urine or saliva by viral culture and/or PCR. As our study was retrospective, we did not have children’s urine or saliva samples and only had children’s serum samples; PCR has relatively lower sensitivity in detecting CMV DNA in serum
. Thus, we did not perform viral culture and PCR to diagnose CMV infection. Third, 13.3% of the infants at the age of 1 month (Table
3) were seroconverted to anti-CMV IgM. It is difficult to ascertain whether these infants were infected postnatally although the anti-CMV IgG levels at 1-month old were lower than in the cord blood. Nevertheless, according to the seroconversion of CMV IgM and IgG, and the dynamic changes of IgG titers (decreasing, followed by increasing) in infants
, we can get the definitive diagnosis of primary CMV infection at least in most of the infants in the present study.
In view of the most CMV infection in children occurred before 3.5-month old, Chinese children may acquire CMV too early to be prevented by vaccination. Therefore, we consider that children, at least in developing countries, should not be the target population for the future licensed vaccine against CMV infection.