Respiratory syncytial virus (RSV) is probably the single major cause of lower respiratory
infection (LRI) among infants worldwide. Its relative importance may be underestimated, as the diagnosis is based on
antigen detection and
antigen may only be detectable in the early phase of
infection. We have therefore assessed the duration of
secretory IgM and
IgA antibody responses and whether assays for these
antibodies can be used to improve the diagnosing of RSV-associated
infections. During two RSV epidemics in Guinea-Bissau, 32 RSV
antigen-positive children with LRI were followed with sequential nasopharyngeal suction on days 7, 14, 30, 60 and 120 in the first epidemic and every fortnight for 6 mo after the second epidemic to measure the duration of
secretory IgM and
IgA responses. Nearly all of the children had an
IgM response during the first month after
infection. The response ratio was highest on days 7 and 14, being 84% and 71%, respectively. After 30 d the
IgM response decreased rapidly. Among 27 age- and sex-matched controls, only 1 child was positive for
IgM. During the second epidemic, when the children were followed more intensively, half of the children were
IgM-positive after the acute phase of
infection. A secondary response may be more likely in children with low
IgM responses in the acute phase (RR = 2.08 (95% confidence interval (CI) 0.92-4.70)). The
IgA response was highest on days 28 and 42 after
antigen detection, 72% having a detectable
IgA response within the first 1.5 mo. Among 27 controls, only 2 were
IgA-positive (7%). In the second epidemic with more intensive follow-up, 62% (8/13) of the
IgA-positive children had a response that lasted 10 wk. Of the children with no persistent
IgA response, half (5/10) had a subsequent
IgA-positive response after the first 42 d. All of these children had a simultaneous
IgM-positive response. When 29 of the children were tested after an epidemic when they were 1-3-y-old, >80% again had high
IgM (24/29, 82%) and
IgA (28/29, 94%) levels. Among samples collected over a 1-y period from infants with LRI in a community morbidity surveillance conducted at the local health centre and via paediatric outpatient consultation, 17% (110/659) were
antigen-positive, 26% (171/659)
IgM-positive and 38% (248/659) either
antigen- or
IgM-positive.
IgM responses are short-lived among infants and may therefore be used as an indication of recent
RSV infection among children with LRI. Using both
antigen and
IgM detection may significantly improve our detection of
RSV infections.