This study utilized publicly available data obtained from the National Health and Nutrition Examination Survey (NHANES), a comprehensive cross-sectional survey administered by physicians and highly trained medical personnel. The survey encompasses questionnaires, physical examinations, and laboratory data. NHANES aims to ascertain the prevalence and identify risk factors associated with major diseases in the U.S. population. Released biannually, the survey provides data collected from participants across the United States, selected through a sophisticated multistage, stratified sampling method. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines for cross-sectional studies.

In present study, we utilized data from six NHANES cycles (2005–2006, 2007–2008, 2009–2010, 2011–2012, 2013–2014, and 2015–2016) that involved a total of 60,936 participants. We excluded participants younger than 20 years, leaving 34,180 adults. Participants were excluded by the following criteria (18, 27–29): missing data for the CDAI calculation; missing data for the calculation of US fatty liver index (USFLI) score; missing data for alcohol consumption or presence of considerable alcohol consumption (>21 drinks per week for male and >14 drinks per week for female); participants with the hepatitis B surface antigen or hepatitis C antibodies; missing data for covariates. Therefore, a total of 9,746 participants was included in the present study, and the flowchart of enrollment is presented in Figure 1. Referring to previous NHANES literature (30, 31), data from the NHANES project was analyzed using a complex stratified sampling method.

Diet affects the CDAI. CDAI data were derived from the two 24-h dietary recall survey of NHANES participants. The first 24h was recorded face-to-face at a mobile examination center, and the second 24 h was recorded by telephone 3–10 days later. The University of Texas Food Intake Analysis System and the United States Department of Agriculture Survey Nutrient Database were used to assess the dietary nutrient intake (32). The nutritional estimates did not include any nutrients obtained from dietary supplements or medications. The average of the two 24-h intakes was taken as the daily dietary intake for the present study (33).

The CDAI was calculated from the mean dietary intake of vitamin A, vitamin C, vitamin E, zinc, selenium, and carotenoids obtained from two 24-h recalls using a modified version developed by Wright et al. (20). Six antioxidants were standardized by subtracting the mean and dividing by the standard deviation. Then the CDAI was calculated based on the sum of these standardized values.

In this formula, X_i represents the individual daily intake of antioxidant components; μ_i represents the mean of X_i; s_i represents the standard deviation for μ_i (33–35).

NAFLD was defined according to the USFLI which was developed using the NHANES database, which was moderately improved accuracy compared to the Fatty Liver Index (FLI) in the multiethnic US population (18). And it has been validated and used in several previous studies (36–38). USFLI was developed based on race/ethnicity, age, gamma-glutamyl transferase (GGT), waist circumference (WC), fasting insulin, and fasting glucose with the following formula:

Scores range from 0 to 100. In this study, a USFLI score ≥ 30 was considered to have NAFLD as suggested by Ruhl and Everhart (18), with an area under the receiver operating characteristic curve (AUROC) of 0.8 (sensitivity: 62%; specificity: 88%).

Based on the literature, the following covariates were selected, including: age, sex, race/ethnicity, education level, marital status, family income, body mass index (BMI), alcohol drinking status, smoking status, diabetes, hypertension, metabolic syndrome, and total energy intake (39, 40). As used by NHANES, we divided race/ethnicity into Mexican American, other Hispanic, non-Hispanic White, non-Hispanic Black and Other (including multiracial). Education level was divided into two groups (high school or below and greater than high school). The marital status was classified as married, never married, living with a partner, and others (including divorced, widowed, and separated). Family income was categorized into 3 levels (< 1.3, 1.3–3.5, and ≥3.5) based on the family poverty income ratio (PIR). BMI was divided into 3 levels (< 25, 25–30, and ≥30 kg/m²). Alcohol drinking status was determined by the following survey question, “In any 1 year, have you had at least 12 drinks of any type of alcoholic beverage?” Participants who answered “yes” were defined as alcohol drinkers. Serum cotinine concentration was utilized as a proxy for environmental tobacco exposure and categorized into active/secondhand smoker (>0.011 ng/mL) and non-smoker (≤ 0.011 ng/mL). Diabetes was defined as using antidiabetic medication or a fasting glucose level equal or >126 mg/dL. Hypertension was defined as using antihypertensive medication or average systolic blood pressure ≥140 mmHg and/or average diastolic blood pressure ≥80 mmHg. Metabolic syndrome was defined based on the Adult Treatment Panel III criteria in 2005 as having at least 3 of the following: waist circumference >102 cm in men or 88 cm in women, triglyceride level >150 mg/dL, high-density lipoprotein cholesterol < 40 mg/dL in men or < 50 mg/dL in women, systolic blood pressure at least 130 mm Hg or diastolic blood pressure at least 85 mm Hg or taking hypertension medications, or fasting plasma glucose level at least 100 mg/dL or taking diabetes medications (41, 42). Total energy intake was calculated by averaging energy intake collected during two 24-h total nutrient recall interviews

According to NHANES analytic guidelines, complex sampling design and sampling weights were considered in our analyses (43). The characteristics of participants are described as means (95% CIs) for continuous variables and percentage frequencies (95% CIs) for categorical variables. Continuous data were compared using t-tests, and categorical data were compared by the χ² test. These means and frequencies can be generalized to the US adult population. No imputation method was used due to the percentage of missing data was small for any variable.

Odds ratios (ORs) and 95% CIs were calculated to assess the association between CDAI/antioxidant components and NAFLD using weighted logistic regression models. Given that the values of antioxidant components intake were skewed, a logarithmic change was performed before statistical analysis to ensure a normal distribution. And the CDAI was generally converted into categorical variables according to quartiles, and the P-values for the trend were calculated. Three models were used in this study. Model 1 was the crude model with no covariates adjusted. Model 2 was adjusted for age, sex, and race/ethnicity. Model 3 was the fully adjusted model which including age, sex, race/ethnicity, education level, marital status, PIR, BMI, alcohol drinking status, serum cotinine, diabetes, hypertension, metabolic syndrome, and total energy intake.

In addition, interaction and subgroup analyses of association between CDAI and NAFLD were also performed according to sex (male, female), age (20–40, 40–60, ≥60 years), BMI (< 25, 25–30, and ≥30 kg/m²), diabetes (no, yes), hypertension (no, yes), and metabolic syndrome (no, yes) using logistic regression models.

To ensure the robustness of our research findings, we adopted the methods used by Ruan et al. (36) and conducted several sensitivity analyses. Initially, to assess the potential hepatotoxicity of certain pharmacological agents, we conducted a sensitivity analysis excluding participants who had been administered methotrexate, acitretin, pioglitazone, liraglutide, semaglutide, atorvastatin, or aspirin (44–53). Subsequent to this, drawing upon extant literature which posited diminished prevalence rates of NAFLD and suboptimal diagnostic precision of USFLI with the non-Hispanic Black cohort (18, 36, 54, 55), we implemented a sensitivity analysis that omitted individuals belonging to this demographic. Finally, in an endeavor to mitigate the possibility of misclassification stemming from USFLI scores, we performed an additional sensitivity analysis utilizing the hepatic steatosis index (HSI) (56). Here, we defined the NAFLD as an HSI score ≥ 36 in present study with the following formula:

Considering the calculation if HSI was based on BMI, BMI was not included as a covariate in this model to avoid the over adjustment. Ultimately, numerous previous studies have examined data from the NHANES to explore risk factors associated with various diseases, and some researchers have utilized weighted analysis methods, while others have employed unweighted approaches. Although NHANES employed complex sampling techniques to improve the representativeness and applicability of findings, conclusions derived from weighted and unweighted analyses can occasionally differ. Therefore, we conducted a sensitivity analysis using unweight data. We also further adjusted for physical activity which assessed by metabolic equivalent (MET) in a typical week to ensure the robustness of our results.

All statistical analysis was performed with R (version 4.1.3, R Project for Statistical Computing, Vienna, Austria) and EmpowerStats (version 4.1, Boston, Massachusetts). In all tests, P < 0.05 (2-sided) was considered to indicate statistical significance.

A total of 9,746 participants (4,662 males [47.6%; 95% CI, 46.6–48.7%] and 5,082 females [52.4%; 95% CI, 51.3–53.4%]; 6422 without NAFLD [67.0%; 95% CI, 65.4–68.5%] and 3,324 with NAFLD [33.0%; 95% CI, 31.5–34.6%]; mean age: 48.3 years [95% CI, 47.7–48.9]) were included. Participants were excluded if they were under 20 years of age, had missing data for CDAI or USFLI calculations, reported excessive alcohol consumption, tested positive for hepatitis B or C, or had incomplete covariate information. The weighted baseline characteristics of included participants are shown in Table 1. Age, BMI, total energy intake, and zinc intake were lower in the participants without NAFLD vs. who with NAFLD (P < 0.05). Family PIR, education level, vitamin E intake, vitamin C intake, and Selenium intake were higher in the participants without NAFLD (P < 0.05). In the terms of alcohol drinking status, moderate-drinkers were more often in participants without NAFLD compared with their counterpart (P < 0.05). And participants with NAFLD were more likely to be Mexican American (P < 0.05). Compared with participants without NAFLD, those with NAFLD had a higher prevalence of diabetes (1,026 [26.6%; 95% CI, 24.5–28.7%] vs. 542 [5.7%; 95% CI, 5.0–6.5%]), hypertension (2,112 [62.2%; 95% CI, 59.8–64.5%] vs. 2,465 [33.5%; 95% CI, 31.7–35.2%]), and metabolic syndrome (1,846 [54.1%; 95% CI, 51.4–56.8%] vs. 802 [10.8%; 95% CI, 10.0–11.7%]). In addition, no significant differences were observed for serum cotinine, vitamin A intake, and carotenoids intake.

The results from weighted multivariable regression analyses are presented in Table 2. A negative correlation between CDAI and NAFLD based on USFLI score was revealed in model 1 (crude model) (OR, 0.99; 95% CI, 0.98–1.00), model 2 (OR, 0.97; 95% CI, 0.96–0.99), and model 3 (the fully adjusted model) (OR, 0.95; 95% CI, 0.93–0.98). After transforming CDAI into quartiles, we found that participants with highest quartile CDAI were 34% less likely to have NAFLD than those with the lowest quartile (OR, 0.66; 95% CI, 0.52–0.85), and the trend test was also significant (P for trend = 0.002). Furthermore, it is worth noting that the association between CDAI related antioxidant components and NAFLD was also conducted in this study. As shown in Table 2, after adjusted for all variables, vitamin A (OR, 0.85; 95% CI, 0.77–0.95), vitamin C (OR, 0.86; 95% CI, 0.80–0.93), vitamin E (OR, 0.73; 95% CI, 0.62–0.86), and Zinc (OR, 0.79; 95% CI, 0.65–0.97) intake were independently associated with NAFLD.

The results of the subgroup and interaction analyses are shown in Figure 2. CDAI was significantly associated with a lower risk of NAFLD among participants with a BMI < 25 (OR, 0.89; 95% CI, 0.83–0.95) and without metabolic syndrome (OR, 0.93; 95% CI, 0.91–0.96). The interaction tests revealed significant differences in these subgroups (P for interaction = 0.04 for BMI and 0.003 for metabolic syndrome). While the negative association between CDAI and NAFLD remained robust across most subgroups stratified by sex, age, BMI, diabetes, and hypertension, the protective effect was strongest among participants with lower BMI and those without metabolic syndrome. In contrast, no significant association was observed among participants with metabolic syndrome (OR, 0.99; 95% CI, 0.95–1.03) or those with a BMI ≥ 30 (OR, 0.96; 95% CI, 0.94–0.99). These findings emphasize that the protective effect of CDAI on NAFLD varies across specific subgroups.

The results of sensitivity analyses are summarized in Table 3. After excluding participants treated with methotrexate, acitretin, pioglitazone, liraglutide, semaglutide, atorvastatin, and aspirin (n = 8911), CDAI showed negative correlation with NAFLD (OR, 0.95; 95% CI, 0.93–0.98). Following the exclusion of non-Hispanic Black participants (n = 7,921), a negative association was observed between CDAI and NAFLD (OR, 0.95; 95% CI, 0.93–0.98), further confirming the robustness of the findings across diverse population groups. Moreover, when NAFLD was defined using the HSI score and adjusted for various covariates including age, sex, race/ethnicity, education level, marital status, PIR, alcohol consumption, serum cotinine, diabetes, hypertension, metabolic syndrome, and total energy intake (n = 9,732), the inverse correlation between CDAI and NAFLD persisted (OR, 0.97; 95% CI, 0.95–0.99). This supports the stability of the observed association even when using an alternative definition of NAFLD. Additionally, a sensitivity analysis utilizing unweighted logistic regression (n = 9,746) confirmed the stability of the negative association between CDAI and NAFLD (OR, 0.96; 95% CI, 0.94–0.98). Finally, a sensitivity analysis incorporating adjustments for physical activity, measured by MET over a typical week (n = 7,677), reaffirmed the robustness of the negative association between CDAI and NAFLD (OR, 0.96; 95% CI, 0.93–0.98).