A profound difference exists between the epidemiology of the infections caused by SARS-CoV and that of all other HCoV infections. SARS-CoV emerged in November 2002 and disappeared in April 2004 (28). During these 18 months, it was isolated in many countries, some of which were very distant from each other. However, the total of 8,098 cases of SARS diagnosed worldwide (28) is substantially fewer than the number usually found during epidemics of the most common respiratory viruses, such as respiratory syncytial virus (RSV) and influenza viruses (18).

Furthermore, seroepidemiologic studies of high-risk and low-risk residential areas have clearly shown that the prevalence of immunoglobulin G against SARS-CoV was low in children and adults (29); this result indicates that SARS-CoV not only had a restricted period of circulation but also that it had limited spread. Proportionally fewer children were involved: <5% of all cases were diagnosed in patients <18 years of age (28). The biology of SARS and its low level of transmissibility seem to be the main reasons for the low risk for contagion in children. In most of the areas in which outbreaks occurred, healthcare workers and adult patients were mainly involved and, because they were immediately hospitalized, the risk for the infection spreading to children was greatly reduced because they are not usually allowed to visit hospitals (28). This hypothesis seems to be further supported by the fact that the early detection and isolation of symptomatic patients were the most important measures in controlling the SARS epidemic.

Unlike SARS-CoV, HCoV-229E, HCoV-OC43, HCoV-NL63, and HCoV-HKU1 have been in continuous circulation since their first isolation and every year cause a large number of infections that more frequently involve children than adults (3,5–26). In particular, the data regarding the earlier-appearing HCoVs indicate that they are distributed throughout the world and mainly circulate during the winter and early spring, with outbreaks occurring every 2–4 years (3). Whenever HCoVs have been sought in studies of respiratory infections in infants and children, they have been found, although generally less frequently than other respiratory viruses such as RSV and influenza (18,30). However, the real importance of HCoV-229E and HCoV-OC43 in clinical practice has not been fully defined because the collected data are often discordant. In 1974, McIntosh et al. found that the global incidence of LRTIs due to HCoV-229 and HCoV-OC43 in hospitalized children was no higher than 3.8% (31). Later studies have shown that when all respiratory infections are considered, the etiologic prevalence of these pathogens in pediatric patients can be significantly higher, varying from ≈5% to >30% (7,15).

An overall evaluation of the available data suggests that these differences can be attributed to differences in research methods. The infections caused by these viruses are more common from November through March; more frequently affect children <5 years of age, those examined in the community, and those without underlying risk factors; and are more often identified with serologic methods (32,33). In this regard, 2 recent surveys that used serologic methods alone found previous HCoV-229E infection in 42.9%–50.0% of children 6–12 months of age (32) and in 65% of those 2.5–3.5 years of age (33). Similarly, in 75% of the cases, children 2.5–3.5 years of age had antibodies against HCoV-NL63 (33). All these factors can explain why the lowest incidences of HCoV-229E and HCoV-OC43 infections are usually found when the study population includes older children or adolescents, patients with underlying severe chronic diseases or hospitalized patients, when only highly symptomatic infections are considered, and when the study is conducted during the whole year.

Moreover, what has been clearly shown is that the original HCoVs are commonly detected in childhood and frequently isolated in the nasopharyngeal secretions of children with respiratory infection. In some cases, co-infections with other respiratory viruses, mainly RSV, influenza viruses, and human metapneumovirus, have also been found (4,7–23). However, the real incidence of HCoV-229E and HCoV-OC43 co-infections with other respiratory pathogens has not yet been defined because only a few of the published studies were planned to identify all the main respiratory viruses.

Similar conclusions can be drawn in relation to the more recently identified HCoVs. HCoV-NL63, which can be found in 1.0%–9.3% of nasopharyngeal aspirates from patients with RTIs (7–23), circulates throughout the world (predominantly during the winter in temperate regions), infects mainly younger children and subjects with underlying severe chronic diseases, and is more frequently found in nonhospitalized children (Appendix Table). Although it was not isolated until 2000, HCoV-NL63 has probably been circulating for some time because one of the first detections occurred in a sample of nasopharyngeal secretions collected from a child with pneumonia that had been kept in a freezer since January 2003 before evaluation (5).

HCoV-HKU1 was identified in 2005 and has once again been found in nasopharyngeal secretions of children and adults with respiratory infections in countries that are very distant from each other. Its incidence varies from <1% to 6% (Table) (6,20,24–26), and seroepidemiologic surveys based on antibodies reacting with the recombinant HKU1 nucleocapsid protein suggest that infection may be relatively common in humans, although generally asymptomatic (32,33). HCoV-NL63 and HCoV-HKU1 are often associated with co-infections with other respiratory viruses, mainly RSV and influenza viruses (7–26).

Superficial analysis of all of the available data concerning the effects of HCoVs in children suggests that the assessment of the importance of HCoVs made before SARS-CoV was identified can still be considered valid. In general, all of these viruses (including SARS-CoV) have been confirmed as mainly respiratory viruses with limited clinical relevance in children. They cause mainly URTIs, are not frequently isolated in hospitalized children (7–26), and, because they are rarely transmitted to other household members, have a marginal socioeconomic effects on families (18). Even SARS-CoV infection, which had a dramatic effect on adults, was mainly associated with relatively mild disease in almost all patients <12 years of age (28,34). Moreover, most children with a diagnosis of severe respiratory syndrome in whom a HCoV was isolated were co-infected by other respiratory viruses (7–26). These finding suggest that the severity of the respiratory disease at least in some of these cases was attributable to the second virus.

Because only premature infants, neonates with a low birthweight, and children with an underlying severe chronic disease are at risk of experiencing a severe respiratory problem associated with HCoV infection (21,37), we could conclude that no further studies of the role of HCoVs in children are needed because what is already known is enough to make such investigations superfluous. Furthermore, on the basis of the data regarding the natural outcome of respiratory infections, developing vaccines or specific drugs appear to be unnecessary.

However, different conclusions can be drawn when the global spectrum of the diseases caused by these viruses in animals and humans is considered. It is now well known that an enormous reservoir of CoVs exists among animals, particularly horseshoe bats, and that CoV isolates recovered from animals in China have up to 99.8% nucleotide identity with SARS-CoV (27). Because CoVs can easily mutate, this means that (as in 2003) sustained exposure to the infected animals can lead to a SARS-like CoV strain that is newly adapted to infect humans and capable of causing the reappearance of SARS. Moreover, it has been shown experimentally and in nature that all CoVs undergo a high rate of genetic mutations and can recombine when 2 different strains infect the same cells (1). This finding means that it is theoretically possible that future situations similar to those involving SARS-CoV may involve CoVs that currently infect only some animals, thus leading to novel viruses with unpredictable host ranges and pathogenicity.

Phylogenetic analyses of the genes spanning the HCoV-HUK1 genome suggest that this virus may be the result of a recombination event between related but distinct HCoVs (6) and that SARS-CoV may have originated from a unique recombination (2). In this regard, the behavior of CoVs could be quite similar to that of influenza viruses, for which genetic changes and recombinations of avian or swine strains are required to allow them to cross the species barrier and replicate in humans to cause a pandemic.

Consequently, as is usually the case with influenza, a systematic evaluation of the characteristics of CoVs should be planned. Patients with severe respiratory syndrome seem to be the best target for this kind of evaluation and, in this population, studies of children (in whom the incidence of infection is higher) may also have application to adults because the findings may lead to a reduction in the risk for the spread of particularly virulent HCoV strains.

In addition to the risk for a pandemic related to the reappearance of SARS or other new CoVs, the data regarding the possible relationship between HCoV infection and CNS diseases also suggest the need for a systematic evaluation of the circulation of CoVs. If >1 HCoVs are demonstrated to play a real role in causing some of the CNS diseases with which they have been associated, substantial changes would be required in our diagnostic, prophylactic, and therapeutic approaches to many neurologic illnesses in children.

HCoV infections can be associated with respiratory and extrarespiratory manifestations, including central nervous system involvement. The clinical and genetic characteristics of circulating HCoVs in the pediatric population should be monitored to detect the spread of particularly virulent HCoV strains in the community at an early stage and, if required, to facilitate the development of adequate preventive and therapeutic measures.