1. Study population
A retrospective cohort study by case-control design was conducted in the Department of Pediatrics, XiangYa hospital, Central South University in Changsha, China from May 2017 to April 2019. We recruited 699 children under 14 years old who were diagnosed with pneumonia and set them as the case group, and 811 healthy children who registered for physical examination as the control group. Children were excluded if they had abnormal congenital development in the airway and lung parenchyma, chronic lung disease, obstruction in the airway or compression outside the tube, congenital heart disease, tuberculosis, heart, liver and kidney diseases, or a known diagnosis of immunodeficiency.
The case group consisted of children with a diagnosis of pneumonia made in accordance with the following WHO diagnosis criteria [27]. Cough or difficult breathing plus at least one of the following signs: (1) fast breathing (rate >60 breaths per minute if aged <2 months, >50 breaths per minute if aged 2-11 months, and >40 breaths per minute if aged 12-59 months); or (2) lower chest wall indrawing. In addition, either crackles or pleural rub may be present on chest auscultation. The control group included healthy children who didn’t have a history of pneumonia diagnosis based on WHO criteria at the time of recruitment.
2. Demographic data
We collected the demographic data from the electronic medical records system in the hospital, including children's gender, age, birth season, gestational weeks, parity, mode of delivery, and parental atopy. Parental atopy was defined by the presence of a history of any allergic diseases among child’s mother and/or father.
3. Exposure assessment
3.1. Exposure timing windows
Exposure timing-windows included preconceptional and prenatal periods in this study. The preconceptional exposure involved in two timing windows: one year before conception and three months before conception. The prenatal period was defined from the first day of the mother’s last menstrual period to the delivery day. The prenatal exposure was further divided into three trimesters: the first trimester (from the 1st to 12th weeks of gestation), second trimester (the 13th to 27th weeks of gestation), and third trimesters (from the 28th gestational week to the birthday of the child).
3.2. Personal exposure to diurnal temperature variation (DTV)
We obtained the data of daily temperature (including the mean, maximum, and minimum temperatures) at ten different monitoring stations in Changsha during 2003-2019 from China Meteorological Administration: Kaifu district, Yuhua district, Furong district, Liuyang county, Mapoling, Ningxiang county, Tianxin district, Wangcheng district, Yuelu district, and Changsha county. The diurnal temperature variation (DTV) was calculated as the difference between the daily maximum and minimum temperature. Children's exposure was estimated by DTV at the station where their residence was located. Then, the individual exposure to outdoor DTV for each child was calculated as follows: (1) Preconceptional exposure to DTV was calculated as the average of daily DTV during one year and three months before conception, (2) Prenatal exposure to DTV was calculated by the average of daily DTV during the period from the mother’s last menstrual period to the delivery day; (3) The exposure to DTV in the first, send, and third trimester was respectively calculated as the average of daily DTV during the periods of 1st–12th gestational weeks, 13th–27th gestational weeks, and from the 28th gestational week till the date of birth,.
3.3 Exposure to outdoor air pollution
Exposure to outdoor air pollution was considered as an important confounding variable for pneumonia risk in this study. We obtained daily 24h-averaged concentrations of three main air pollutants, including nitrogen dioxide (NO2), sulfur dioxide (SO2), and particulate matter ≤10 µm in diameter (PM10), from 7 municipal air quality monitoring stations in Changsha from 2003 to 2019. Individual exposure to ambient air pollution during different timing windows was estimated by the inverse distance weight (IDW) method described in our previous work [28].
4. Statistical analysis
Statistical analyses were performed by SPSS software (version 23.0, SPSS Inc., Chicago, USA). The relationship between exposure to outdoor DTV during preconceptional and prenatal periods and childhood pneumonia was assessed by using multiple logistic regression model, with adjusting potential variables in Table 1 and the three air pollutants (NO2, SO2, and PM10). The associations in regression analysis was calculated by odds ratio (OR) of 95% confidence interval (95% CI). In our study, OR (95% CI) was estimated by per 1 °C increase in exposure to outdoor DTV. P-value≤0.05 was considered as statistically significant.