Studies on the association between dyslipidemia and asthma in children reported between January 2000 and May 2022 were searched using PubMed (Table 1). The search was performed using the terms dyslipidemia, TC, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), very low-density lipoprotein cholesterol, asthma, and children. Eligible studies had to be published in English and included randomized controlled trials and prospective followup, retrospective, and cross-sectional studies. Letters, editorials, reviews, commentaries, case reports, and personal communication were not included. The population of included studies comprised those with children or adolescents under the age of 18 years who could have asthma and control groups without asthma. In addition, studies had to assess at least one part of the lipid profile of their study population, such as TC, HDL-C, LDL-C, or triglyceride (TG). Additionally, eligible studies had to include quantitative results regarding the outcomes of interest. Candidate studies were screened using a 2-step process. First, by reviewing the titles and abstracts of each study, studies that did not meet the inclusion criteria were excluded. Second, the full texts of the remaining studies were reviewed according to inclusion and exclusion criteria. Two reviewers identified the eligible studies. A third reviewer was consulted in the case of any uncertainty regarding eligibility. Among 226 studies, 11 were included.

We extracted the following relevant data from the included studies: name of first author, publication years, study design, participation age and number, exposure, and outcome of interest, and summarized the results.

The present study used 8 hospital-based cohorts that were converted to the OMOP-CDM format using the FEEDERNET platform, which provides EHR data without patients’ personal information. The Observational Health Data Sciences and Informatics (OHDSI) organization is an international collaboration that works to create high-quality evidence by creating and using open-source data analytics solutions on a large network of health databases from different countries [6]. This allows for the systematic analysis of disparate observational databases. The concept behind this approach is to transform the data in those databases into a common format and representation (terminologies, vocabularies, coding schemes) and then use a library of standard analytical routines that have been written based on the common format to do systematic analyses. A key infrastructure requirement for large-scale distributed comparative effectiveness research is that all healthcare systems use CDM [7]. Once a database has been converted to the OMOP-CDM, evidence can be generated using standardized analytics tools. The CDM contains 18 data tables: person, drug exposure, drug era, condition occurrence, condition error, observation period, observation, procedure occurrence, visit occurrence, death, drug cost, procedure cost, location, provider, organization, care site, payment plan period, and cohort [6].

The 5 secondary or tertiary hospitals included Kangdong Sacred Heart Hospital in Seoul (KDH), Kyung Hee University Hospital at Gangdong in Seoul (KHNMC), Kangwon National University Hospital in Chuncheon (KWMC), Gyeongsang National University Hospital in Changwon (GNUH), and Deagu Catholic University Hospital in Deagu (DCMC). All hospitals signed a memorandum of understanding for research in border-free zones. The enrollment period and total number of patients were 1986 to 2018 and 1,689,604 in KDH, 2006 to 2017 and 822,183 in KHNMC, 2003 to 2018 and 519,700 in KWMC, 2009 to 2022 and 618,246 in GNUH, and 2005 to 2018 and 1,688,980 in DCMC, respectively. The total number of enrolled patients was 5,338,713 (Fig. 1). The study protocol was approved by the Institutional Review Board of Hallym University (IRB 2019-09-005) without approval from the institutional review boards of other institutions in accordance with the Memorandum of Understanding on the Research Border-Free Zone.

This was a retrospective cohort study. A flowchart of the study is shown in Fig. 1. The index date was the date when the blood was drawn for TC measurements. Children under 15 years of age who underwent blood tests for the measurement of TC were identified. Children diagnosed with asthma before the index date were excluded. The target group was the high TC group, defined as a TC level greater than 170 mg/dL [8]. The comparator group was the normal TC group, defined as a TC level of 170 mg/dL or less. In both groups, participants were censored either at the time of outcome identification or at the end of the observation period in the database. Children were excluded if they belonged to either group by performing the TC level test several times. Finally, there were 88,823 children in the normal TC group and 29,038 in the high TC group.

The primary outcome was the first diagnosis of asthma. Asthma was defined as one or more principal diagnoses of the International Classification of Diseases, Tenth Edition (ICD-10) codes for asthma (J45.X) and 2 or more prescriptions for asthma treatment drugs, such as inhaled corticosteroids (ICS), combination ICS and long-acting beta-agonists, and leukotriene modifiers [9-11].

To balance the baseline characteristics between the high TC and normal TC groups, the demographic and clinical variables were considered covariates. Age at the index date and sex were regarded as demographic characteristics. In addition, the diagnosed diseases and medications during the 365 days before the index date were regarded as clinical characteristics. Diagnosed diseases were identified using ICD-10. Medications were prescribed at the hospital visit. The covariates in each hospital are shown in Supplementary Table 1 to 5.

In all hospital datasets after PS matching, the high TC group as a target group included 29,038 children, and the normal TC group as a comparator group included 88,823 children. Table 3 shows the baseline characteristics of the matched cohort. The baseline demographic and clinical data of the unmatched and matched cohorts in each hospital are described in Supplementary Table 1 to 5. Before PS matching, age group distribution; sex ratio; medical history, such as acute respiratory disease and urinary tract infection; and medication history, such as antibiotics and anti-inflammatory drugs, differed between the high TC and normal TC groups. However, after PS matching, the age group distribution, sex ratio, medical history, and medication history were balanced between the high TC and normal TC groups. Each hospital had slightly different characteristics, but the age group of 5–9 years accounted for the largest proportion, and the male-to-female ratio was comparable in all hospitals.

Table 4 and Fig. 2 show the association between total levels and asthma in children. The asthma incidence rate (per 1,000 patient-years) of the high TC group tended to be higher than that of the normal TC group, except for KWMC. The meta-analysis showed that the high TC group was significantly associated with an increased risk of asthma (pooled HR, 1.30; 95% CI, 1.12–1.52). There was no significant heterogeneity across the databases (I²=0%, P=0.68). The survival curves for asthma in each hospital are shown in Fig. 3.