2 MATERIALS AND METHODS
2.1 Setting and patients
The clinical data of SMA children in the division of respiratory and sleep disorders of the Children’s Hospital Affiliated with the Capital Institute of Pediatrics in Beijing, China, from March 2016 to December 2021 were analyzed retrospectively. RRTIs and ARF were considered severe conditions requiring intervention, and clinical characteristics were analyzed to determine the risk factors for RRTIs and/or ARF in children with SMA who had not started treatment with disease-modifying medications or mechanical ventilation. This study was reviewed and approved by the ethics committee of the Capital Institute of Pediatrics (Identifier, SHERLL2019014). The study was exempt from informed consent.
2.2 Diagnostic criteria
SMA was diagnosed by a defect in the SMN1 gene localized to 5q11.2-q13.3 and classified into clinical groups on the basis of age of onset and maximum motor function achieved: very weak infants unable to sit unsupported (type 1), non-ambulatory patients able to sit independently (type 2), and ambulant pediatric patients (type 3).1
RRTIs were diagnosed by the following criteria: for children under two, upper respiratory infection ≥ 7 episodes/year, bronchitis ≥ 3 episodes/year, or pneumonia ≥ 2 episodes/year; two to five years old, upper respiratory infection ≥ 6 episodes/year, bronchitis ≥ 2 episodes/year, or pneumonia ≥ 2 episodes/year; over five years old, upper respiratory infection ≥ 5 episodes/year, bronchitis ≥ 2 episodes/year, or pneumonia ≥ 2 episodes/year. The number of lower respiratory infections could replace the number of upper respiratory infections, but not vice versa. The interval between onsets of respiratory infection was at least seven days.5
The diagnostic criterion of ARF caused by infectious pneumonia was severe dysfunction of ventilation with arterial partial pressure of oxygen < 8.0 kPa (60 mmHg) while breathing room air, with/without arterial partial pressure of carbon dioxide (PCO2) > 6.7 kPa (50 mmHg).6
2.3 Inclusion and exclusion criteria
At enrollment, children were under eighteen years old and genetically confirmed to have homozygous SMN1 gene alterations; all completed PSG without signs of respiratory tract infection. Exclusion criteria included the initiation of treatment with disease-modifying medications or mechanical ventilation, alone or combined with the presence of other congenital diseases or other neuromuscular disorders. All children were divided into a disease group and a control group according to whether RRTIs and/or ARF occurred within one year.
2.4 Data collection
The electronic and handwritten medical records created by doctors in the division of respiratory and sleep disorders were reviewed to collect the clinical data, including sex, age, height, weight, type of SMA, genetic results, regular use of mechanical insufflation-exsufflation (MI-E), number of occurrences of respiratory tract infection and ARF in one year, past history and medication history. Spirometry was performed in children over five years old who could complete the flow-volume loop test using MasterScreen (Jaeger, Germany) equipment. The percent predicted scores of spirometry were reported. PSG was attended by trained pediatric sleep nurse using Alice 6 LDx (Philips, USA) equipment and was scored using the American Academy of Sleep Medicine Version 2.3 pediatric criteria7 (see further details in E-appendice 1). The age-specific body mass index z score (BMIz) was obtained according to the WHO child growth standard and the 2000 American CDC growth curve standard.8,9
2.5 Statistical analysis
Continuous data conforming to a normal distribution are represented by (x ± s), and the differences between two groups were evaluated by the t test. Data not conforming to a normal distribution are represented by M (Q1, Q3), and the differences between two groups were evaluated by the Mann–Whitney U test. Categorical variables are expressed as cases (%), and comparisons between groups were performed with the chi-square test and Fisher’s exact test. The differences in clinical indices between the two groups were compared, and the variables with statistically significant differences were included in the binary multivariate logistic regression. The results showed the risk of each covariate in the model and 95% confidence interval to obtain the independent risk factors for RRTIs and/or ARF in children with SMA. Receiver operating characteristic (ROC) curves were constructed to determine the best cutoff point for positive indicators that could best predict the occurrence of RRTIs and/or ARF. The statistical significance of all verifications was based on a bilateral P value < 0.05. SPSS 25.0 software was used for statistical processing, and GraphPad Prism 9.3.1 was used to draw ROC curves.
The clinical data were screened before logistic regression on the basis of the clinical characteristics. The parameters of the spirometry were not included because the test was not performed in children under five. Because sleep stages vary greatly among children of different ages, for instance, the sleep structure of infants is very different from that of older children, such indicators with age differences were not considered. In addition, in order to minimize the possible effects of multicollinearity, the representative apnea hypopnea index (AHI) was chosen because there is overlap among the parameters of respiratory events in PSG; moreover, the AHI is the sum of apneas and hypopneas occurring per hour. In addition, the mean pulse oxygen saturation (MSpO2) with small individual differences was selected because the standard deviation of MSpO2 was less than the lowest pulse oxygen saturation (LSpO2).