The current cross-sectional study was part of the Global Asthma Network (GAN), which took place in Yazd, central Iran, in early 2020. The GAN is a cross-sectional, multi-center, multi-country epidemiological study that builds on and enhances the ISAAC Phase three methodology (30). According to the GAN recommendation, at least 3,000 samples are required to estimate a good prevalence of asthma (31). In the present study, students from 48 and 36 high and elementary schools (state and private) were randomly selected from two educational districts using a cluster sampling design, respectively. Moreover, non-Iranian students were excluded from the study. All the subjects aged 13–14 and parents of 6–7 years were invited to fill out online electronic questionnaires on asthma and its symptoms and risk factors, which were designed and placed in the virtual education groups of schools. The information was collected using paper questionnaires for a group of participants aged 6–8 years. Due to the school closures during the COVID-19 pandemic, the rest of the 6-8year old students were provided with the electronic questionnaires. Out of 7214 and 3026 adolescents, and children, 5141 and 2526 completed the questionnaire, with response rates of 71.3% and 83.5%, respectively, and then demographic data that seemed unacceptable was reviewed by telephone and corrected if necessary. At the beginning of the study, due to schools' closure during the outbreak of COVID-19, an electronic questionnaire was designed and placed in the virtual education groups of schools.
The ethics committee of Shahid Sadoughi University (SSU) of Medical Sciences in Yazd, Iran, approved the GAN study on Iranian youngsters (IR.SSU.REC.1398.244). The ethics committee also gave its approval for the current study. The Yazd Education administration then granted permission to conduct the study at elementary and guidance schools. Parents gave their informed consent as well. The consent form was included at the start of the internet questionnaires so that kids and their parents may feel entirely comfortable participating in the research.
Asthma and its symptoms confirmation
The GAN questionnaire, derived from the ISAAC questionnaire (29), includes questions about the symptoms of allergic diseases and related risk factors. In this study, we used some questions about asthma symptoms, "use of asthma medication" and "asthma confirmed by a doctor", as well as the amount of dairy consumed in the diet last year. According to the protocol of this study, current asthma was defined as a history of confirmed asthma by a doctor and having had wheezing and/or use of asthma medication in the past 12 months. Once the questionnaire was translated into Persian, the reliability of the translated version was confirmed by a study conducted on 100 selected subjects using Cronbach's alpha. The alpha coefficient for asthma symptoms was estimated to be 0.862, thus exhibiting appropriate internal consistency. Finally, the questionnaire was translated back into English and sent to the GAN principals in order to be approved.
Assessment of dietary intake
In this study, the dietary intake of children's food groups in the last 12 months was assessed using one of the multiple choice questions in the GAN questionnaire (32). The frequency response section had three options (never or occasionally/once or twice a week most or all days of the week). Milk, other dairy products, and total dairy consumption were assessed. The frequency of total dairy intake was assessed by summing dietary milk (including flavored milk), cheese, and yoghurt consumption.
Assessment of other variables: The data on participants’ height, weight, ethnicity (Kord/Turk/Persian/Lor/Arab/Balooch), watching television and computer use (2–4 hours/5–8 hours/9–14 hours a day) were obtained using a self-reported online GAN questionnaire. The body mass index (BMI) was calculated by using the following formula: weight (kg) divided by height squared (m2).
Statistical methods. We used STATA version 14 (State Crop., College Station, TX) for all analyses. To compare continuous and categorical variables between individuals with and without "asthma confirmed by a doctor" and "usage of asthma medication," independent samples of student t tests and chi-square testing was used, respectively. Multivariable logistic regression analysis was used in crude and multivariable controlled models to investigate the relationship between dairy intake and the odds of asthma confirmed by a doctor, current asthma, usage of asthma medication, and wheezing in the last 12 months. In the first model, adjustments were made for age and sex. Additional adjustments were made for watching TV and computer use in the second model. In the last model (third model), an additional adjustment for BMI was made. In all models, the lowest level of dairy intake (never or only occasionally) was considered as the reference category. P-value < 0.05 was considered as statistically significant.