Menu
Menu
Menu
Menu
Background: Respiratory syncytial virus (RSV) is a major cause of lower respiratory tract infection (LRTI) in children. Early-life RSV LRTI might affect long-term health but there are few data from low-income and middle-income countries. We investigated the epidemiology and effect of early-life RSV LRTI on lung health in a South African birth cohort. Methods: We conducted the Drakenstein Child Health Study (DCHS), an ongoing birth cohort longitudinal study in the Western Cape province, South Africa. We enrolled pregnant women aged 18 years or older during their second trimester of pregnancy at two public health clinics. We followed up study children from birth to 2 years. The primary outcome of the study was LRTI and RSV LRTI. LRTI and wheezing episodes were identified through active surveillance; respiratory samples were tested for RSV and other pathogens. Wheezing was longitudinally identified by caregiver report and ascertainment at health facilities. Lung function was measured from 6 weeks to 2 years. We analysed the associations between RSV LRTI and subsequent LRTI, wheezing, and lung function using generalised estimating equations and mixed-effects linear regression. Findings: We enrolled 1137 mothers between March 5, 2012, and March 31, 2015. Among their 1143 infants, accruing 2093 child-years of follow-up, there were 851 cases of LRTI (incidence 0·41 episodes per child-year, 95% CI 0·38–0·43). Admission to hospital owing to LRTI occurred in 169 (20%) cases (incidence 0·08 episodes per child-year, 0·07–0·09), with a case-fatality ratio of 0·5%. RSV was detected in 164 (21%) of 785 LRTI events with a specimen available for qPCR, an incidence of 0·08 episodes per child-year (0·07–0·09); highest at age 0–6 months (0·15 episodes per child-year, 0·12–0·19). Children with a first RSV LRTI were three times as likely to develop recurrent LRTI compared with those with non-RSV LRTI (0·32 [0·22–0·48] vs 0·10 [0·07– 0·16] episodes per child-year; p<0·0001), particularly following hospitalised RSV LRTI. RSV LRTI and hospitalisation for all-cause LRTI were independently associated with recurrent wheezing (adjusted incident rate ratio 1·41, 95% CI 1·25–1·59, for RSV LRTI and 1·48, 1·30–1·68, for hospitalisation). LRTI or recurrent LRTI was associated with impaired lung function, but a similar outcome was observed following RSV LRTI or non-RSV LRTI. All-cause LRTI was associated with an average 3% higher respiratory rate (95% CI 0·01–0·06; p=0·013) and lower compliance (–0·1, −0·18 to 0·02) at 2 years compared with no LRTI. Recurrent LRTI was associated with further increased respiratory rate (0·01, 0·001–0·02), resistance (0·77 hPa s L−1, 0·07–1·47), and lower compliance (–0·6 mL hPa−1, −0·09 to −0·02) with each additional event. Interpretation: RSV LRTI was common in young infants and associated with recurrent LRTI, particularly after hospitalised RSV. Hospitalisation for all-cause LRTI, especially for RSV-LRTI, was associated with recurrent wheezing. Impairments in lung function followed LRTI or recurrent episodes, but were not specific to RSV. New preventive strategies for RSV might have an effect on long-term lung health. Funding: Bill & Melinda Gates Foundation; South African Medical Research Council; National Research Foundation South Africa; National Institutes of Health, Human Heredity and Health in Africa; Wellcome Trust.
The DCHS, located in a peri-urban area in South Africa, enrolled pregnant women from March, 2012, to March, 2015, during their second trimester of pregnancy at two public health clinics.14 Inclusion criteria were age 18 years or older, 20–28 weeks' gestation, and resident in the area. Study visits were synchronised with health-care and immunisation visits (diphtheria, tetanus, acellular pertussis, Hib, and inactivated polio vaccine at 6, 10, 14 weeks, and 18 months, measles vaccine at 9 and 18 months, and 13-valent PCV [PCV13] at 6 weeks, 14 weeks, and 9 months), with additional study visits at 6, 12, and 24 months. Mother–infant pairs could also participate in intensive one follow-up visit every 2 weeks during the first year.14 We followed up children from birth until age 2 years for LRTI or wheezing using active surveillance systems, as described.15 Continuous surveillance was implemented at local clinics and at Paarl hospital. WHO criteria were used to define pneumonia or LRTI, as previously described.13 Children were followed up throughout LRTI hospital admission, or following an ambulatory episode. Recurrent LRTI was defined as two or more episodes. Episodes of wheezing were reported by a caregiver using questions adapted from the International Study of Asthma and Allergies in Childhood or were diagnosed on auscultation by trained study staff at a study visit or during an intercurrent illness.16 Longitudinal measurement of risk factors for LRTI or wheezing including nutrition, home environment, vaccinations, smoke exposure, and maternal factors was done at study visits and during illness. Maternal smoking or passive smoke exposure was self-reported. A composite locally validated measure from the South African Stress and Health Study17 of socioeconomic status was used encompassing current employment, education, household income, and an asset index. The study was approved by the Faculty of Health Sciences Research Ethics Committee, University of Cape Town and Western Cape Provincial Research committee. Mothers gave written informed consent at enrolment and re-consented annually. At each LRTI or wheezing episode, a nasopharyngeal swab (FLOQSwabs, Copan Diagnostics, Murrieta, CA, USA) was obtained. Nucleic acid was extracted using mechanical lysis on a Tissuelyzer LT (Qiagen, Hilden, Germany) followed by extraction with the QIAsymphony Virus/Bacteria mini kit (Qiagen). Quantitative multiplex real-time PCR (qPCR) was done using FTDResp33 (Fast Track Diagnostics, Esch-sur-Alzette, Luxembourg), identifying up to 33 organisms including RSV. RSV LRTI was defined as any episode of LRTI that was positive for RSV on qPCR. Comprehensive, validated18 lung function measurements were done using the multiple breath washout technique (measuring lung volume, functional residual capacity and the lung clearance index [LCI]); tidal breathing (measuring respiratory rate, tidal volume, and flow rates); respiratory impedance, resistance and compliance, and exhaled nitric oxide in 6-week-old unsedated sleeping infants and at 1 and 2 years, as described.18 Lung function was done when infants and children were healthy, and after 4 weeks of a LRTI. The primary outcome of the study was LRTI and RSV LRTI. Data were analysed using Stata version 14.1 and R. Socioeconomic status was evaluated in quartiles. Wilcoxon rank-sum test and χ2 or Fisher's exact were used for crude comparison, as appropriate. Child follow-up period was divided into intervals of 2 weeks that were used to calculate person-time at risk, with a not at risk (excluded person-time) period of 2 weeks after an LRTI or 4 weeks (28 days) after a wheezing episode. LRTI incidence was reported as episodes per child-year with 95% CI. Episodes of LRTI occurring immediately after birth or before discharge after delivery were regarded as congenital events and were excluded from this analysis. To examine independent predictors of LRTI we used negative binomial regression with a log link during each follow-up interval; generalised estimating equations were used to account for intraindividual clustering of intervals. We used these models to compare RSV LRTI versus non-RSV LRTI, or hospitalised LRTI versus ambulatory LRTI. Separate models included all individuals in the cohort and individuals with any episode of LRTI versus no LRTI. Multivariable models were adjusted for an a-priori set of confounding factors, identified through expert consultation with investigators and construction of a directed acyclic graph to identify the minimum covariates for models. A-priori confounders included were: sex, socioeconomic status quartile, season at birth, other children in the household, HIV exposure, maternal smoking, prematurity, low birthweight, and age at mid-interval or age at LRTI if an episode occurred in the interval. Multivariable modelling of risk factors associated with LRTI was done for any LRTI versus no LRTI (model A); RSV LRTI versus no LRTI (model B); and RSV LRTI versus non-RSV LRTI (model C). Recurrent wheezing was defined a-priori as two or more observations of wheezing after the LRTI event in individuals experiencing LRTI and two or more observations of wheezing in those who did not experience an LRTI. Similar to LRTI, wheezing models were based on generalised estimating equations with a negative binomial family and link function, adjusting for similar covariates with age included at time of reported or assessed wheeze or mid-interval of no wheeze recorded. For children with recurrent wheezing, time to first wheezing event after LRTI was used as the time to event for product-limit analyses. Additional analyses for three or more observations of wheezing was also done. For the Kaplan Meier estimates, survival proportions (cumulative incidences from survival analyses) were used. Child follow-up period was calculated as described. Longitudinal changes in lung function measures from 6 weeks to 2 years were analysed using mixed-effects linear models, with random effects for lung function outcomes, thus accounting for baseline lung function in models. These were fitted using an a-priori set of confounding factors of body-mass index for age Z score, ethnicity, sex, socioeconomic status quartile, age at time of lung function test, and gestational age at birth. Study sponsors had no role in the study design; collection, analysis, and interpretation of the data; writing of the report; or the decision to submit for publication. All authors had full access to all the data and share final responsibility for the decision to submit for publication.
