Data sources and study population
This cross-sectional study used data from the National Health and Nutrition Examination Survey (NHANES) conducted by the Centers for Disease Control and Prevention from 2003 to 2020.13The NHANES project aimed to assess the health and nutritional status of noninstitutionalized Americans through a stratified multistage probability survey.14Demographic and comprehensive health information was collected at visits, screening, and clinical examinations by mobile examination centers (MECs). Approval for NHANES was obtained from the National Center for Health Statistics Institutional Review Board, and all participants provided written informed consent before participating in the study. No additional Institutional Review Board (IRB) approval was determined to be required for this secondary analysis.15This study used the NHANES dataset accessible on the official NHANES website (accessed on October 1, 2023). Our study included children aged 3–18 years who completed the interview, but excluded those with missing asthma questionnaire and variable data.
asthma
Following previous references, the assessment of asthma was based on information collected from the questionnaire portion of the US National Health Interview Survey. NHANES uses self-administered questionnaires to collect information on asthma and related symptoms.16Participants were defined as having current asthma if they answered “yes” (or by proxy if under 16 years of age) to the following two questions: “Have you ever been told by a doctor or other health professional that you have asthma?” and “Do you currently have asthma?” Participants who did not currently have asthma and answered “no” to these questions were defined as the control group.
Composite Dietary Antioxidant Index (CDAI)
The NHANES Dietary Survey implemented a fully computerized recall system with standardized questions and response options customized for food-related questions. This method was designed to ensure thorough and accurate collection of dietary information. Accurate dietary information for each participant was obtained using the Continuing Survey of Food Intake by Individuals (CSFII) and the Automated Multipass Method (AMPM).17The NHANES Dietary Surveyor Procedures Manual was utilized as a comprehensive guide providing detailed insight into the dietary survey methodology.
In the dietary survey conducted as part of NHANES, participants were asked to recall details of their diet during the past 24 hours. Six dietary antioxidants were included in the study: vitamin A, vitamin C, vitamin E, zinc, selenium, and carotenoids. The calculation of dietary antioxidant intake excluded antioxidants found in supplements, medicines, or regular drinking water. To assess aggregate exposure from dietary antioxidant intake, a modified version of the Composite Dietary Antioxidant Index (CDAI) developed by Wright et al. was used.7,18 These values were normalized by first subtracting the mean and then dividing by the standard deviation (SD).
$${\text{CDAI}} = \mathop \sum \limits_{i = 1}^{n = 6} \left( {{\text{IndividualIntake}} – {\text{Mean}}} \right) /{\text{SD}}$$
Finally, CDAI was categorized into tertiles for analysis: tertile 1 (T1) served as the reference category and indicated a CDAI below the 33rd percentile, tertile 2 (T2) included participants with a CDAI between the 33rd and 67th percentile, and tertile 3 (T3) included participants with a CDAI above the 67th percentile.
Covariates
We aimed to minimize potential confounding bias in the analysis by selecting covariates based on previous studies and clinical validity. Selected covariates included sex, race and ethnicity, poverty income ratio (PIR), body mass index category (BMIC), familial asthma, white blood cell count (WBC), eosinophil percentage (EOPC), total energy/protein/sugar/carbohydrate/fiber/fat intake, and serum cotinine. A detailed description of the measurement procedures can be found at https://www.cdc.gov/nchs/nhanes and was accessed on October 1, 2023. In particular, race and ethnicity (Mexican American, other Hispanic, non-Hispanic white, non-Hispanic black, other race) were collected.13The PIR was determined by dividing household (or individual) income by the poverty threshold for the survey year. Based on the US government report, household income was classified into three categories: low (PIR ≤ 1.3), medium (PIR > 1.3–3.5), and high (PIR > 3.5).19BMI categories (BMIC) were underweight (BMI < 5th percentile), normal weight (BMI 5th-85th percentile), overweight (BMI 85th-95th percentile), and obese (BMI ≥ 95th percentile). Complete blood counts (CBC) were performed on blood samples from all participants using a Beckman Coulter DxH 800 instrument at the NHANES MEC. Secondhand smoke exposure was measured in participants aged 3 years and older by cotinine, a nicotine metabolite. Serum cotinine concentrations were measured using isotope dilution high-performance liquid chromatography/atmospheric pressure chemical ionization tandem mass spectrometry (ID HPLC-APCI MS/MS). To reduce bias due to missing data, samples with missing covariates (PIR, WBC, EOPC, BMIC, familial asthma, and serum cotinine) were excluded.
Statistical analysis
To account for the complex survey design, sample weights, strata, and primary sampling units were used according to NHANES analytical guidelines. Differences in baseline characteristics between the asthma and non-asthma groups were compared using independent-samples t tests and χ2 tests for continuous variables.2 Tests for categorical variables. Categorical data were expressed as unweighted numbers (weighted percentages), and continuous data were expressed as means (standard errors). [SE]NHANES provides a special set of weights (WTDRD1) for analyses using the first 24-hour dietary recall data. The WTDRD1 weights were constructed by taking the MEC sample weights (WTMEC2YR) and further adjusting for additional nonresponse and differential allocation by day of the week of the dietary recall. For more information, see https://wwwn.cdc.gov/Nchs/Nhanes/2003-2004/DR1IFF_D.htm#WTDRD1.20Multivariate weighted logistic regression analysis was performed for the association between CDAI and asthma, and odds ratios (ORs) with 95% confidence intervals (CIs) were calculated. Three adjustment models were included in the multivariate logistic regression analysis: model 1 was unadjusted, and model 2 included sex, age, race and ethnicity, PIR, BMIC, and serum cotinine. Model 3 was further adjusted for variables including WBC, EOPC, total energy intake, total protein intake, total sugar intake, total carbohydrate intake, total fiber intake, total fat intake, and cycle year.
In addition, subgroup analyses were performed using stratified logistic regression models. Interactions between subgroups were evaluated using likelihood ratio tests. To increase the reliability of our findings, we conducted sensitivity analyses, first adjusting for confounders using unweighted Logstic multifactorial analysis of the entire data series and evaluating the association between CDAI and asthma. We then added the covariate of familial asthma (only data from 2007 to 2020) to exclude missing variables, and finally included 10,520 participants in the analysis, adjusted for confounders and familial asthma, and evaluated the association between CDAI and asthma using weighted Logstic multifactorial analysis. Finally, we used multiple estimation of covariates with missing data (PIR, WBC, EOPC, BMIC, familial asthma, serum cotinine) to minimize the bias that may be introduced into the results by missing data. Specifically, the MICE algorithm with fully conditional specification (FCS) described by Van Buuren and Groothuis Oudshoorn (2011, https://doi.org/10.18637/iss.v045i03) was used. These data were finally used to assess the association between CDAI and asthma.
Sample size was determined based solely on available data, therefore no prespecified statistical power estimates were performed. Analyses were performed using the R statistical software package (http://www.R-project.org, The R Foundation) and Free Statistics software, version 1.9.2.twenty oneDescriptive statistics were performed for all participants using two-sided tests with a significance level of 2.5. P< 0.05.
Ethical approval
Ethics review and approval was not required because this study involved secondary analyses and did not require additional Institutional Review Board approval. The National Health and Nutrition Examination Survey (NHANES) was approved by the Institutional Review Board of the National Center for Health Statistics (NCHS). Written informed consent was obtained from all participants prior to their participation in the study. Statement attesting to parental or legal guardian informed consent for study participation: In accordance with ethical guidelines for research involving minors, informed consent was obtained from parents or legal guardians for all study participants. Prior to participation, parents or legal guardians provided signed consent forms, and minors aged 7–15 years provided additional signed consent forms. Copies of signed consent forms and information pamphlets were distributed to all participants and their guardians. The detailed consent process followed established procedures and complied with all federal regulations and ethical standards. For more information, please visit the NHANES website at https://www.cdc.gov/nchs/nhanes/irba98.htm and https://www.cdc.gov/nchs/data/series/sr_02/sr02_163.pdf.
Impact Statement
This study comprehensively evaluates the association between the Composite Dietary Antioxidant Index (CDAI) and asthma prevalence. Although the role of oxidative stress and inflammation in asthma pathophysiology is well documented, our findings reveal no significant association between total dietary antioxidant intake and asthma risk. This lack of association persisted after rigorous adjustment for confounding factors, including familial asthma history, and sensitivity analyses. These results suggest that simply increasing dietary antioxidant intake may not be sufficient for asthma prevention. This study highlights the importance of considering the complex interplay of dietary, genetic, and environmental factors in asthma research and may guide future dietary recommendations and public health strategies for asthma management. Further studies are needed to explore these interactions in more detail and to identify more effective dietary interventions for asthma prevention and management.
