The relationship between 10 anthropometric measures and metabolic dysfunction-associated fatty liver disease: among United States population: the National Health and Nutrition Examination Survey 2017–2018

doi:10.21203/rs.3.rs-3397501/v1

Download PDF

Article

The relationship between 10 anthropometric measures and metabolic dysfunction-associated fatty liver disease: among United States population: the National Health and Nutrition Examination Survey 2017–2018

https://doi.org/10.21203/rs.3.rs-3397501/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

MAFLD is a rapidly progressive chronic liver disease approximately 1/3 of adults in the United States have fatty liver. Therefore, this study investigated the relationship between 10 anthropometric measures and the risk of MAFLD using the American National Health and Nutrition Examination Surveys.

Methods

Overall 3702 participants were included in this study. MAFLD was defined as hepatic steatosis with one or more of the following: (1) overweight or obesity (body mass index ≥ 25 kg/m2), (2) type 2 diabetes, or (3) two or more other metabolic risk abnormalities. The anthropometric measures included weight, hip, WC, BMI, BRI, BAI, WTHR, WHTR, ABSI, and conicity index. The relationship between anthropometric measures and MAFLD was investigated using survey-weighted logistic regression, and stratified by sex and ages.

Result

When different indexes were used as continuous variables, the odds ratio (OR) is basically between 1.01–1.08. When different indexes were used as categorical variables (divided into trisection), WHTR per SE had the highest OR 20.166 (95%CI:12.131–33.523), followed was WC per SE with OR 19.472 (95%CI: 13.046–29.064), BRI per SE with OR 19.110 (95%CI: 11.195–32.623), while the ABSI per SE with the lowest OR 1.968 (95%CI: 1.387–2.794), WC with highest AUC (0.832), WHTR and BRI with AUC (0.831), ABSI with lowest AUC (0.609). WHTR had the highest sensitivity of 0.791(0.765,0.815).

Conclusion

The results of this study suggest that anthropometric indices can be good predictors of MAFLD. WTHR, BRI and WC were better predictors of MAFLD in adults, correctly identifying 80% of patients with the disease.

Health sciences/Risk factors

Health sciences/Biomarkers/Predictive markers

Health sciences/Diseases/Gastrointestinal diseases/Liver diseases

Anthropometric measures

metabolic dysfunction-associated fatty liver disease

dietary supplements

NHANES

cross-sectional analysis

Non-alcoholic fatty liver disease (NFALD) is defined as fatty degeneration of the liver and is the leading cause of chronic liver disease ¹. Epidemiological studies have shown that the prevalence of non-alcoholic fatty liver globally is 25% ², with a high incidence in the United States, and approximately 1/3 experience it¹. The disease increases the risk of liver-related complication and cardiovascular and all-cause mortality³. The burden of disease caused by NAFLD is expected to continue to increase in the future due to the worsening obesity epidemic and lack of approved drug treatments. NAFLD is also associated with other extrahepatic diseases, including obesity, diabetes, hypertension and high cholesterol⁴. In additional, an international panel of experts from 22 countries has proposed changes in terminology and definitions to reflect disease pathogenesis more accurately. The proposed term for metabolic (dysfunctional)-associated fatty liver (MAFLD) is defined as hepatic steatosis, in addition to evidence of overweight or obesity, diabetes, or metabolic dysfunction⁵. Notably, redefining NAFLD as MAFLD may help improve our predictors of increased cardiovascular and mortality risk ⁶.

However, because diagnosing MAFLD requires the determination of liver fat by imaging or histological detection, and the judgment of blood sugar, blood lipids, and other metabolic diseases, it is complicated to incorporate it into the routine monitoring of large-scale populations. Therefore, a noninvasive, low-cost, and easy-to-calculate measurement method was chosen in this context. Excess weight, visceral fat accumulation, is considered a major risk factor for NAFLD and is strongly associated with disease severity⁷. Several studies have demonstrated that using simple surrogate obesity and lipid-related indices to identify fatty liver, including body mass index (BMI), waist circumference (WC), waist-to-hip ratio (WTHR), waist -to-height ratio (WHTR), body shape index (ABSI), Body Roundness Index (BRI), most of which are good predictors of NAFLD^7,8, Some indexes that have not yet explored the relationship with fatty liver, such as the body adipose index (BAI), and conicity index, and these indexes are also suitable for evaluating visceral content^9,10. Notably, using these widely used and easily computed anthropometric methods seems promising for predicting MAFLD. However, many studies have found that different indices have different predictive abilities for diseases related to metabolic abnormalities. Furthermore, BMI is the most widely used as an anthropometric measure of obesity¹¹. However, it is not a reliable indicator of central obesity and WC is also generally accepted as a barometer of central obesity¹². A meta-analysis supports the use of WHTR as a better predictor of cardiovascular risk factors than BMI¹³. ABSI was superiors in predicting deaths, compared to BMI or WC¹⁴. BRI is a significantly better predictor of hypertension than ABSI¹⁵.

However, the population characteristics of MAFLD are consistent with those of NFALD. Studies have shown that MFALD has a higher BMI, hypertension, diabetes, blood lipids levels, elevated transaminases levels, and higher fibrosis scores^16,17. This may result in anthropometric measures that better predict NAFLD and may not necessarily be a good screening tool for patients with MAFLD. Simultaneously, people who are not overweight may be diagnosed with metabolic syndrome. To date, no research has been conducted on the correlations between different anthropometric measures and MAFLD. MAFLD includes fatty liver and metabolic disorders such as hypertension, obesity, and diabetes. These indexes have a good positive correlation with hypertension, diabetes, and cardiovascular diseases in previous studies, therefore, there is no difference in the predictive ability of these indexes for metabolic fatty liver. Consequently, it is necessary to identify anthropometric measures with higher correlations with MAFLD and stronger predictors.

Overall, 10 anthropometric measures were included in this study, and the relationship between the 10 measurement indexes and MAFLD was analyzed, Different sex and age groups were compared to explore which anthropometric measures are suitable for screening for MAFLD.

2.1 Data source and study participates

Data was obtained from the United States Centers for Disease Control and Prevention and retrieved from the National Health and Nutrition Examination Survey (NHANES) database (https://www.cdc.gov/nchs/nhanes/index.htm). Data were collected using a stratified, multistage, probability-cluster design, including oversampling of Mexican American, non-Hispanic white, non-Hispanic black, other Hispanics and other races¹⁸. This study used NHANES data from 2017 to 2018 was the only publicly available survey database for liver fibrosis assessment using Fibroscan® which has been used MAFLD studies¹⁹. The initial data comprised 9254 individuals from 2017 to 2018. Subsequently, we excluded participants younger than 20 years (n = 3685), pregnant women (n = 56), incomplete diagnostic indicators for MAFLD (n = 644), missing BMI (n = 44), ABSI (n = 167), WTHR (n = 11), energy intake (n = 855), and minutes sedentary activity information (n = 26). Additionally, we excluded 26 women and 38 and 38 men with implausible total energy intake (< 600 or ˃6000 kcal/day) and (< 800 or ˃8000 kcal/day), respectively²⁰. Ultimately, 3702 participants were included in this study. The inclusion and exclusion criteria for study participants are shown in Fig. 1.

2.2 Diagnosis of MAFLD

Hepatic steatosis was defined using controlled attenuation parameters (CAP) and was stratified into grades S1–3 with CAP thresholds of 302, 331, and 337 dB/m, respectively²¹. Hepatic stiffness was measured using liver stiffness measurements (LSM) and stratified into F2–4 grades with LSM thresholds of 8.2, 9.7, and 13.6 kPa, respectively. Significant steatosis and stiffness were diagnosed as grades greater than those in S1 and F2²¹. MAFLD was defined as hepatic steatosis in the presence of one or more of the following conditions: (1) overweight or obesity (BMI ≥ 25 kg/m²); (2) type 2 diabetes; or (3) two or more other metabolic risk abnormalities: 1) blood pressure ≥ 130/85 mmHg or specific drug treatment; 2) overweight or obesity (BMI ≥ 25 kg/m²); 3) plasma triglycerides ≥ 150 mg/dL or specific drug treatment; 4) plasma high-density lipoprotein cholesterol < 40 mg/dL for men and < 50 mg/dL for women or specific drug treatment; 5) prediabetes (fasting glucose 100–125 mg/dL or hemoglobin A1c (HbA1c) 5.7–6.4%; 6) homeostasis model assessment of insulin resistance score ≥ 2.5; and 7) plasma C-reactive protein level > 2 mg/L¹⁹.

2.2 Anthropometric measures

Trained health technicians obtained various anthropometric measurements, including height, weight, waist circumference, and hip circumference, in a mobile examination center (MEC). The calculation formulas for the different indices were as follows:

2.2.1 BMI

$$\mathbf{B}\mathbf{M}\mathbf{I}=\frac{\varvec{w}\varvec{e}\varvec{i}\varvec{g}\varvec{h}\varvec{t}\left(\varvec{k}\varvec{g}\right)}{{\varvec{h}\varvec{e}\varvec{i}\varvec{g}\varvec{h}\varvec{t}}^{2}\left({\varvec{m}}^{2}\right)}$$

2.2.2 BRI

BRI ¹⁵ = $364.2-365.5\times [{1-\left({\frac{\varvec{w}\varvec{a}\varvec{i}\varvec{s}\varvec{t}/2\varvec{\pi }}{\varvec{h}\varvec{e}\varvec{i}\varvec{g}\varvec{h}\varvec{t}\times 0.5})}^{2}\right]}^{0.5}$

2.2.3 BAI

BAI ²² $=\frac{\varvec{h}\varvec{i}\varvec{p}\left(\varvec{c}\varvec{m}\right)}{{\varvec{h}\varvec{e}\varvec{i}\varvec{g}\varvec{h}\varvec{t}}^{1.5}\left({\varvec{m}}^{3/2}\right)}-18$

2.2.4 WTHR

WTHR ²³ $=\varvec{w}\varvec{a}\varvec{i}\varvec{s}\varvec{t}\left(\varvec{c}\varvec{m}\right)÷\varvec{h}\varvec{e}\varvec{i}\varvec{g}\varvec{h}\varvec{t}\left(\varvec{c}\varvec{m}\right)$

2.2.5 WHTR

WHTR ²³ = $\varvec{w}\varvec{a}\varvec{i}\varvec{s}\varvec{t}\left(\varvec{c}\varvec{m}\right)÷\varvec{h}\varvec{e}\varvec{i}\varvec{g}\varvec{h}\varvec{t}\left(\varvec{c}\varvec{m}\right)$

2.2.6 ABSI

ABSI ²³ $=\mathbf{w}\mathbf{a}\mathbf{i}\mathbf{s}\mathbf{t}\left(\mathbf{m}\right)\times {\mathbf{w}\mathbf{e}\mathbf{i}\mathbf{g}\mathbf{h}\mathbf{t}}^{-\frac{2}{3}}\left(\mathbf{k}\mathbf{g}\right)\times {\mathbf{h}\mathbf{e}\mathbf{i}\mathbf{g}\mathbf{h}\mathbf{t}}^{5/6}$

2.2.7 Conicity index

Conicity index ²⁴ = $\frac{\varvec{w}\varvec{a}\varvec{i}\varvec{s}\varvec{t}\left(\varvec{m}\right)}{0.109\sqrt{\frac{\varvec{w}\varvec{e}\varvec{i}\varvec{g}\varvec{h}\varvec{t}\left(\varvec{k}\varvec{g}\right)}{\varvec{h}\varvec{e}\varvec{i}\varvec{g}\varvec{h}\varvec{t}\left(\varvec{m}\right)}}}$

2.3 Other covariates

2.3.1 Demographic characteristics and lifestyle

Self-reported demographic variables and lifestyle included age (years), sex (male/female), race/ethnicity (Mexican American, other Hispanic, non-Hispanic white, non-Hispanic black, non-Hispanic Asian, or other races, including multiracial), educational level (< 11th grade, high school graduate, some college, and college graduate or above), and smoking status (never smoker (never smoked 100 cigarettes in the lifetime), smoked some days, or smoked every day). Alcohol use information was collected from the dietary interview-total nutrient intakes and minutes of sedentary activity section of the physical activity questionnaire.

2.3.2 Energy intake

Dietary intake data were obtained from two 24-hour recall interviews using the NHANES. The first interview was conducted face-to-face at the MEC, and the second was conducted by telephone 3–10 days later. The energy intake for each food or beverage was calculated using the Food and Nutrient Database for Dietary Studie.

2.3.3 Diseases and medications

The diagnostic criteria for diabetes (DM) were as follows: the doctor stated that you have diabetes, glycohemoglobin HbA1c (%) > 6.5, fasting glucose (mmol/L) ≥ 7.0, random blood glucose (mmol/ L) ≥ 11.1, or 2-hour OGTT blood glucose (mmol/ L) ≥ 11.1, or use of diabetes medication or insulin. The diagnostic criteria for hypertension were as follows: the doctor stated that you have hypertension, systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg, or the use of hypertension medication.

The pressure/cholesterol section (variable name prefix BPQ) provided personal interview data and was informed about taking a prescription for cholesterol.

2.4 Statistical methods

The sampling weights recommended by the NHANES for the planned oversampling of specific groups were used in this study. All analyses were sample-weighted and accounted for the complex stratified, multistage, cluster sampling design of NHANES^25,26. Categorical and continuous variables are expressed as counts (survey-weight percentages) and the survey-weight mean (standard error), respectively. The basic characteristics of participants between the two groups (non-MAFLD vs. MAFLD) were compared using survey-weighted linear regression and chi-square test for continuous and categorical variables, respectively.

Owing to the inconsistency of the numerical units of different body indices, the value gap was large and could reach hundreds of times. Therefore, to eliminate the gap between the unit and the value and to better compare the relationship between different indices and MAFLD, this study calculated the index per SE, implying that the results of dividing each individual’s different index values by their respective SEs were used as independent variables. The relationship between the index per SE and MAFLD was evaluated using a survey-weighted logistic regression. Furthermore, the relationship between index per SE and MAFLD phenotypes (non-MAFLD vs. MAFLD without fibrosis vs. MAFLD with fibrosis) was assessed using survey-weighted multivariate ordinal logistic regression. Model 1 was unadjusted. Model 2 was adjusted for age, sex, and race. Moreover, Model 3 was adjusted for age, sex, race, education, smoking, alcohol consumption, minutes of sedentary activity, drugs, energy, hypertension, and diabetes. To better explore the association between the difference index and MAFLD, multivariate logistic regression was conducted to evaluate the index as a categorical variable (divided into trisection). Finally, the relationship between the different indices and MAFLD was stratified according to age and sex. To compare the predictive power of various indexes, we drew the receiver operating characteristic curve, obtained the cutoff value and the area of the curve, and calculated the sensitivity and specificity.

All analyses were performed using the R software, version 4.0.3 (R Foundation for Statistical Computing, Vienna, Austria). A two-sided P-value < 0.05 was considered significant for all analyses.

3.1 The basic characteristics of participants

As shown in Table 1, compared to non-MAFLD, the population with MAFLD was showed the highest weighted mean anthropometric measures (weight, hip, WC, BMI, BRI, BAI, WTHR, WHTR, conicity index, ABSI), minutes of sedentary activity, old age, and higher weighted prevalence of males, hypertension, DM in this study (P<0.001).

3.2 The relationship between different anthropometric measures and MAFLD

As shown in Table 2, the results of the survey-weighted linear regression of the 10 anthropometric measures were all significantly associated with MAFLD, and P values for trend were significant in all three models. When different indexes were used as continuous variables, the value of increasing the risk of MAFLD by one unit of the index was similar, and the odds ratio (OR) was basically between 1.01 and 1.08. Among them, BMI, BRI, and BAI were more suitable for predicting MAFLD with fibrosis (OR, 1.20–1.80) than MAFLD alone (OR 1.02–1.06).

When different indexes were used as categorical variables (divided into trisection), the highest tertiary (T3) had a higher MAFLD prevalence in model 3 than the lowest tertiary (T1) population of the 10 anthropometric measures. The WHTR per SE had the highest OR of 20.166 (95% confidence interval (CI): 12.131–33.523), followed by waist circumference per SE with an OR of 19.472 (95% CI: 13.046–29.064), BRI per SE with an OR of 19.110 (95% CI: 11.195–32.623), and BMI per SE with an OR of 18.099 (95% CI: 12.509–26.187), whereas the ABSI per SE had the lowest OR of 1.968 (95% CI: 1.387–2.794), and the remaining index OR was between 5 and 15.

In Model 3, the highest tertiary (T3) had a higher MAFLD with fibrosis prevalence than the lowest tertiary (T1) of the 10 anthropometric measures. WHTR per SE had the highest OR of 21.922 (95% CI: 13.874–34.389), followed by BRI per SE, which had an OR of 20.760 (95% CI: 12.901–33.507), waist circumference per SE had an OR of 20.504 (95% CI: 14.029–29.966), and BMI per SE had an OR 19.080 (95% CI: 13.757–26.463), whereas the ABSI per SE had the lowest OR of 1.987 (95% CI: 1.416–2.787), and the remaining index OR was between 5 and 15.

As shown in Table 3, the WC had the highest area under the curve (AUC) (0.801) and weighted AUC (0.832); WHTR and BRI had an AUC of 0.801 and weighted AUC of 0.831, and ABSI had the lowest AUC (0.609) and weighted AUC (0.633). In addition, the cutoff values of waist circumference, WHTR, and BRI were 104.35, 0.605, and 5.545, respectively. WHTR had the highest sensitivity of 0.791 (0.765, 0.815).

3.3 The relationship between anthropometric measures and MAFLD in different sex

As shown in Table 4, the results of the survey-weighted linear regression of the 10 anthropometric measures were all significantly associated with MAFLD in both sexes. In the male population, the first three indices with the largest OR values were WHTR (1.080, 95% CI: 1.065–1.095), WC (1.072, 95% CI: 1.060–1.085), and BRI (1.071, 95% CI: 1.056–1.087), and the last three indices with the smallest OR values were ABSI (1.018, 95% CI: 1.008–1.029), BAI (1.049, 95% CI: 1.035–1.062), and hip (1.050 95% CI: 1.039–1.062). In the female population, the first three indices with the largest OR values were WHTR (1.061, 95% CI: 1.049–1.073), WC (1.061, 95% CI: 1.050–1.073), and weight (1.057, 95% CI: 1.044–1.069), and the last three indices with the smallest OR values were ABSI (1.007, 95% CI: 1.000–1.016), BAI (1.029, 95% CI: 1.022–1.037), and hip (1.034, 95% CI: 1.026–1.042). Indices that interacted with sex included WC, BRI, WHTR, BMI, conicity index, BAI, and ABSI.

3.3 The relationship between anthropometric measures and MAFLD in different age

As shown in Table 5, the results of the survey-weighted linear regression of the 10 anthropometric measures were significantly associated with MAFLD in different age groups. Whether it was in the crowd <40 years old, 40–65 years old, or ≥65years old, the first three indexes with the largest OR were WHTR (1.064, 95% CI: 1.048–1.080), WC (1.059, 95% CI: 1.047–1.071), and BRI (1.053, 95% CI: 1.038–1.069), and the last three indexes with the smallest OR were ABSI (1.012, 95% CI: 1.000–1.026), WTHR(1.051, 95% CI: 1.038–1.065), and BAI (1.036, 95% CI: 1.026–1.047). Indices that interacted with age included the WTHR, WHHR, and ABSI.

MAFLD is a new nomenclature for NAFLD, and studies have shown that the prevalence of MAFLD is similar to that of NAFLD. In contrast, the risk of cirrhosis and liver cancer is more likely to be increased in people with MAFLD because they have higher liver enzymes and more diseases related to glucose and lipid metabolism²⁷. In this study, CAP was used to diagnose fatty liver disease. This method is highly accurate and precise²⁸. However, to use a simpler and easier-to-operate method for large-scale screening, it is necessary to assess anthropometric measures that are suitable for the risk prediction of MAFLD. This study compared 10 anthropometric measurements, each of which increased the risk of MAFLD. The three anthropometric measures (WHTR, BRI, and waist circumference) were more effective in predicting the risk of MAFLD and MAFLD with fibrosis than BMI, weight, conicity index, hip, WTHR, BAI, and ABSI. The risk of MAFLD or MAFLD with fibrosis in the third highest quartile of WTHR, BRI, and WC was approximately 20 times that of the non-MAFLD and 5–15 times for other indicators. As categorical and continuous variables, every unit increase in BRI increased the risk of MAFLD by 5.6%, and the risk of MAFLD with fibrosis increased by 68%, with an AUC of 0.80. ABSI had the lowest AUC, sensitivity, and specificity compared with the other indices.

Some formulas recommended in the literature for NAFLD prediction already use anthropometric indicators such as BMI and WC^29,30, because excess weight, mainly visceral fat accumulation, is considered the major risk factor for NAFLD. In this study, waist circumference had a better prediction in MAFLD than the other indexes. Clinical anthropometric indicators, such as WHTR, are described in the literature as the most sensitive and specific for discriminating visceral fat compared to BMI and waist circumference in NAFLD³¹. In this study, we also found that WHTR was a better predictor of MAFLD than BMI. The AUC value of WTHR for MAFLD screening is close to the waist value, while the advantage of WHTR is that it integrates height and waist and is less influenced by race, age, and sex; therefore, it is relatively stable. A systematic evaluation and meta-analysis showed that WHTR is a more favorable screening tool for metabolic risk factors than BMI and waist in adults³². Furthermore, WHTR is an indicator of abdominal obesity and has been used in several studies to assess metabolic disorders; it has significantly greater discriminatory power than BMI³³. A boundary value of WHTR = 0.5 indicates increased risk, meaning the waist should be kept lower than 1/2 hip. Studies have shown an AUC above 0.71 for the WHTR to detect NAFLD and a cutoff of 0.53–0.58^7,29. In this study, the AUC was above 0.801 for WHTR to detect MFALD and a cutoff of approximately 0.605 (sensitivity: 79%, specificity: 82%), indicating that WHTR was superior to NAFLD in the prediction of MAFLD. The increase in the cutoff value may be related to the inclusion of metabolic disorders in the diagnostic criteria of MAFLD³⁴.

The BRI is a new index proposed by Thomas et al. and is considered a predictor of body fat and visceral adipose tissue volume. The study showed that BRI has a better predictive ability for NAFLD, with differences in cutoff values between men and women: the optimal threshold for BRI was 4.00 (sensitivity = 82.7%, specificity = 70.8%) for men and 5.00 (sensitivity = 83.3%, specificity = 71.7%) for women⁷. A good risk prediction ability of BRI for MAFLD was also found in this study, with an AUC of 0.786 and a cutoff value of 5.545 (sensitivity = 79.1%, specificity = 64.7%), with an interaction between sex. Although the BRI has a higher cutoff value and lower predictive power in MAFLD than in NAFLD, several studies have identified the potential of the BRI in identifying hypertension, diabetes, and cardiovascular health conditions^35–37. Therefore, BRI is as applicable as WHTR for the risk prediction of MAFLD; notably, the predictive power of BRI for MAFLD is affected by gender differences.

Most studies have compared WHTR, BRI, and BMI with ABSI for hypertension, diabetes, and metabolic syndrome risk prediction. However, the results showed good predictive power for BRI and weak risk prediction power for ABSI ^36,38. This study also found that the ABSI had the lowest predictive power among the 10 body measurement indices. It is possible that because MAFLD is a multifactorial disease involving metabolic disorders, including hypertension, diabetes, and hyperlipidemia, ABSI alone does not fully reflect their true metabolism. However, because ABSI combines height and weight with waist circumference, ABSI has a negligible correlation with BMI and can be used as an independent factor in combination with BMI. For example, a study found that combining ABSI with BMI was a better predictor of risk for diabetes and NAFLD ³⁹. Both conicity index and ABSI are composite indices that combine height, weight, and WC, with a low correlation with BMI. Studies have found that conicity index performs significantly better in screening for diabetes, regardless of whether the individual is obese or not⁴⁰. However, WHTR was found to have the highest discriminant value for metabolic syndrome, whereas conicity index showed a significantly lower ability to predict metabolic syndrome²⁴. This study also found that conicity index had lower predictive power than WHTR for MAFLD. However, because conicity index is superior to BMI in identifying visceral obesity and is positively correlated with insulin, blood pressure, and lipids⁴¹, studies have found that its combination with adiponectin factors or other indices improves the prediction of coronary heart disease⁴². Therefore, the possibility of using ABSI or conicity index in combination with BMI or other body indices for screening MAFLD should be considered.

First, to the best of our knowledge, this is the first study to examine the relationship between 10 body indices and metabolic fatty liver disease to validate the performance of different anthropometric measures related to the amount and distribution of body fat as predictors of MAFLD. Second, the NHANES database contains a large and nationally representative sample of the adult population in the United States, and the anthropometric and laboratory data in this database are high quality. Third, we evaluated the relationship between different indices and MAFLD in various sex and age groups and differentiated MFALD with and without liver fibrosis. However, this study had some limitations. First, this was a cross-sectional study of only the United States population. Therefore, there may be limitations in generalizing over different countries and races. Second, this study is not useful to further explore the predictive ability of joint application between indices for MAFLD. Therefore, further cohort studies are needed to assess the predictive ability of indices for disease progression.

The results of this study suggest that anthropometric indices can also be good predictors of MAFLD. Based on the OR, AUC, sensitivity, and specificity, WTHR, BRI, and WC were better predictors of MAFLD in adults, correctly identifying 80% of patients with MAFLD. Although ABSI and conicity index with lower predictive power, while considering that these two index formulas include height and weight with waist circumference, ABSI and conicity index may be suitable for long-term cohort studies in combination with other indices.

MFALD Metabolic dysfunction‑associated fatty liver disease

NAFLD Non-alcoholic fatty liver disease

NHANES National Health and Nutrition Examination Surveys

DGA Dietary Guidelines for Americans

SE Standard error

BMI Body mass index

BRI Body roundness index

BAI Body adipose index

WTHR WC-to-height ratio

WHTR WC- to- height ratio

ABSI Body shape index

AUC Area under the curve

CAP Controlled attenuation parameters

LSM Liver stiffness measurements

Declaration of Competing Interest

The authors declare that they have no competing interests.

Author Contributions:

All authors contributed to the study conception and design. QYM, LHW designed the study; ZW, JLH organized and analyzed the data; QYM contributed materials/analysis tools; ZW JLH wrote the paper. All authors have read and approved the final manuscript.

Funding: This study was supported by the Incubation Fund of Science and Technology Commission of Shanghai Municipality (20DZ2254500).

Institutional Review Board Statement: The study was conducted in accordance with the Declaration of Helsinki, and the Ethics Review Board of the National Center for Health Statistics approved the NHANES protocol. The experimental protocol/s was/were approved by Ethics Review Board of the National Center for Health Statistics.

Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Consent to participate

Not applicable

Consent for publication

Not applicable

Data Availability Statement:

Data described in the manuscript, code book, and analytic code will be made publicly and freely available without restriction at https://www.cdc.gov/nchs/nhanes/index.htm.

Acknowledgments: N/A

Chalasani, N. et al. The diagnosis and management of nonalcoholic fatty liver disease: Practice guidance from the American Association for the Study of Liver Diseases. Hepatology (Baltimore, Md.) 67, 328-357, doi:10.1002/hep.29367 (2018).
Younossi, Z. et al. Global Perspectives on Nonalcoholic Fatty Liver Disease and Nonalcoholic Steatohepatitis. Hepatology (Baltimore, Md.) 69, 2672-2682, doi:10.1002/hep.30251 (2019).
Mantovani, A. et al. Complications, morbidity and mortality of nonalcoholic fatty liver disease. Metabolism: clinical and experimental 111s, 154170, doi:10.1016/j.metabol.2020.154170 (2020).
Kaya, E. & Yilmaz, Y. Metabolic-associated Fatty Liver Disease (MAFLD): A Multi-systemic Disease Beyond the Liver. Journal of clinical and translational hepatology 10, 329-338, doi:10.14218/jcth.2021.00178 (2022).
Eslam, M., Sanyal, A. J. & George, J. MAFLD: A Consensus-Driven Proposed Nomenclature for Metabolic Associated Fatty Liver Disease. Gastroenterology 158, 1999-2014.e1991, doi:10.1053/j.gastro.2019.11.312 (2020).
Kim, D. et al. Metabolic dysfunction-associated fatty liver disease is associated with increased all-cause mortality in the United States. Journal of hepatology 75, 1284-1291, doi:10.1016/j.jhep.2021.07.035 (2021).
Motamed, N. et al. Body Roundness Index and Waist-to-Height Ratio are Strongly Associated With Non-Alcoholic Fatty Liver Disease: A Population-Based Study. Hepatitis monthly 16, e39575, doi:10.5812/hepatmon.39575 (2016).
Almeida, N. S., Rocha, R., Cotrim, H. P. & Daltro, C. Anthropometric indicators of visceral adiposity as predictors of non-alcoholic fatty liver disease: A review. World journal of hepatology 10, 695-701, doi:10.4254/wjh.v10.i10.695 (2018).
Ma, Z. et al. Changing trajectories of serum uric acid and risk of non-alcoholic fatty liver disease: a prospective cohort study. Journal of translational medicine 18, 133, doi:10.1186/s12967-020-02296-x (2020).
Rincón Bello, A. et al. Fat tissue and inflammation in patients undergoing peritoneal dialysis. Clinical kidney journal 9, 374-380, doi:10.1093/ckj/sfw007 (2016).
Yusuf, S. et al. Obesity and the risk of myocardial infarction in 27,000 participants from 52 countries: a case-control study. Lancet (London, England) 366, 1640-1649, doi:10.1016/s0140-6736(05)67663-5 (2005).
Agarwal, R., Bills, J. E. & Light, R. P. Diagnosing obesity by body mass index in chronic kidney disease: an explanation for the "obesity paradox?". Hypertension (Dallas, Tex. : 1979) 56, 893-900, doi:10.1161/hypertensionaha.110.160747 (2010).
Liu, J. et al. Predictive Values of Anthropometric Measurements for Cardiometabolic Risk Factors and Cardiovascular Diseases Among 44 048 Chinese. Journal of the American Heart Association 8, e010870, doi:10.1161/jaha.118.010870 (2019).
Wei, J., Liu, X., Xue, H., Wang, Y. & Shi, Z. Comparisons of Visceral Adiposity Index, Body Shape Index, Body Mass Index and Waist Circumference and Their Associations with Diabetes Mellitus in Adults. Nutrients 11, doi:10.3390/nu11071580 (2019).
Calderón-García, J. F. et al. Effectiveness of Body Roundness Index (BRI) and a Body Shape Index (ABSI) in Predicting Hypertension: A Systematic Review and Meta-Analysis of Observational Studies. International journal of environmental research and public health 18, doi:10.3390/ijerph182111607 (2021).
Lim, G. E. H. et al. An Observational Data Meta-analysis on the Differences in Prevalence and Risk Factors Between MAFLD vs NAFLD. Clinical gastroenterology and hepatology : the official clinical practice journal of the American Gastroenterological Association, doi:10.1016/j.cgh.2021.11.038 (2021).
Yamamura, S. et al. MAFLD identifies patients with significant hepatic fibrosis better than NAFLD. Liver international : official journal of the International Association for the Study of the Liver 40, 3018-3030, doi:10.1111/liv.14675 (2020).
Ciardullo, S., Monti, T. & Perseghin, G. Prevalence of Liver Steatosis and Fibrosis Detected by Transient Elastography in Adolescents in the 2017-2018 National Health and Nutrition Examination Survey. Clinical gastroenterology and hepatology : the official clinical practice journal of the American Gastroenterological Association 19, 384-390.e381, doi:10.1016/j.cgh.2020.06.048 (2021).
Weng, Z. et al. Circadian Misalignment Rather Than Sleep Duration is Associated with MAFLD: A Population-Based Propensity Score-Matched Study. Nature and science of sleep 13, 103-111, doi:10.2147/nss.S290465 (2021).
Reis, J. P. et al. Association between duration of overall and abdominal obesity beginning in young adulthood and coronary artery calcification in middle age. Jama 310, 280-288, doi:10.1001/jama.2013.7833 (2013).
Eddowes, P. J. et al. Accuracy of FibroScan Controlled Attenuation Parameter and Liver Stiffness Measurement in Assessing Steatosis and Fibrosis in Patients With Nonalcoholic Fatty Liver Disease. Gastroenterology 156, 1717-1730, doi:10.1053/j.gastro.2019.01.042 (2019).
Bergman, R. N. et al. A better index of body adiposity. Obesity (Silver Spring, Md.) 19, 1083-1089, doi:10.1038/oby.2011.38 (2011).
Wierup, I. et al. Low anthropometric measures and mortality--results from the Malmö Diet and Cancer Study. Annals of medicine 47, 325-331, doi:10.3109/07853890.2015.1042029 (2015).
Oliveira, R. G. & Guedes, D. P. Performance of anthropometric indicators as predictors of metabolic syndrome in Brazilian adolescents. BMC pediatrics 18, 33, doi:10.1186/s12887-018-1030-1 (2018).
Survey, C. f. D. C. a. P. C. N. C. f. H. S. N. N. H. a. N. E. NHANES Demographics Data. [(accessed on 10 June 2022)].
Centers for Disease Control and Prevention (CDC) National Center for Health Statistics (NCHS) National Health and Nutrition Examination Survey: Analytic Guidelines, 2011–2014 and 2015–2016.
Lin, S. et al. Comparison of MAFLD and NAFLD diagnostic criteria in real world. Liver international : official journal of the International Association for the Study of the Liver 40, 2082-2089, doi:10.1111/liv.14548 (2020).
Papatheodoridi, M. & Cholongitas, E. Diagnosis of Non-alcoholic Fatty Liver Disease (NAFLD): Current Concepts. Current pharmaceutical design 24, 4574-4586, doi:10.2174/1381612825666190117102111 (2018).
Yoo, H. J. et al. Cutoff points of abdominal obesity indices in screening for non-alcoholic fatty liver disease in Asians. Liver international : official journal of the International Association for the Study of the Liver 30, 1189-1196, doi:10.1111/j.1478-3231.2010.02300.x (2010).
Ju, D. Y. et al. The influence of waist circumference on insulin resistance and nonalcoholic fatty liver disease in apparently healthy Korean adults. Clinical and molecular hepatology 19, 140-147, doi:10.3350/cmh.2013.19.2.140 (2013).
Zhang, X. et al. Nonalcoholic fatty liver disease prevalence in urban school-aged children and adolescents from the Yangtze River delta region: a cross-sectional study. Asia Pacific journal of clinical nutrition 24, 281-288, doi:10.6133/apjcn.2015.24.2.13 (2015).
Ashwell, M. & Gibson, S. Waist-to-height ratio as an indicator of 'early health risk': simpler and more predictive than using a 'matrix' based on BMI and waist circumference. BMJ open 6, e010159, doi:10.1136/bmjopen-2015-010159 (2016).
Ashwell, M., Gunn, P. & Gibson, S. Waist-to-height ratio is a better screening tool than waist circumference and BMI for adult cardiometabolic risk factors: systematic review and meta-analysis. Obesity reviews : an official journal of the International Association for the Study of Obesity 13, 275-286, doi:10.1111/j.1467-789X.2011.00952.x (2012).
Adams, L. A., Waters, O. R., Knuiman, M. W., Elliott, R. R. & Olynyk, J. K. NAFLD as a risk factor for the development of diabetes and the metabolic syndrome: an eleven-year follow-up study. The American journal of gastroenterology 104, 861-867, doi:10.1038/ajg.2009.67 (2009).
Chang, Y. et al. A Body Shape Index and Body Roundness Index: Two New Body Indices to Identify left Ventricular Hypertrophy among Rural Populations in Northeast China. Heart, lung & circulation 25, 358-364, doi:10.1016/j.hlc.2015.08.009 (2016).
Janghorbani, M., Aminorroaya, A. & Amini, M. Comparison of Different Obesity Indices for Predicting Incident Hypertension. High blood pressure & cardiovascular prevention : the official journal of the Italian Society of Hypertension 24, 157-166, doi:10.1007/s40292-017-0186-3 (2017).
Tian, S., Zhang, X., Xu, Y. & Dong, H. Feasibility of body roundness index for identifying a clustering of cardiometabolic abnormalities compared to BMI, waist circumference and other anthropometric indices: the China Health and Nutrition Survey, 2008 to 2009. Medicine 95, e4642, doi:10.1097/md.0000000000004642 (2016).
Chang, Y. et al. A body shape index and body roundness index: two new body indices to identify diabetes mellitus among rural populations in northeast China. BMC public health 15, 794, doi:10.1186/s12889-015-2150-2 (2015).
Kuang, M. et al. The value of combining the simple anthropometric obesity parameters, Body Mass Index (BMI) and a Body Shape Index (ABSI), to assess the risk of non-alcoholic fatty liver disease. Lipids in health and disease 21, 104, doi:10.1186/s12944-022-01717-8 (2022).
Vasques, A. C. et al. Anthropometric indicators of insulin resistance. Arquivos brasileiros de cardiologia 95, e14-23, doi:10.1590/s0066-782x2010001100025 (2010).
Mantzoros, C. S., Evagelopoulou, K., Georgiadis, E. I. & Katsilambros, N. Conicity index as a predictor of blood pressure levels, insulin and triglyceride concentrations of healthy premenopausal women. Hormone and metabolic research = Hormon- und Stoffwechselforschung = Hormones et metabolisme 28, 32-34, doi:10.1055/s-2007-979126 (1996).
Coniglio, R. I. et al. [Central obesity: relationship between conicity index and lipoprotein risk factors for coronary atherosclerosis]. Medicina 57, 21-28 (1997).

Table1 Basic Characteristics of participants by categories of MAFLD

Characteristics (weighted)	Total	Non- MAFLD	MAFLD	P value
N	3702	2622	1080
Sex (%)				< 0.001
Men	48.176 (0.019)	44.617 (1.185)	57.575 (2.088)
Women	51.824 (0.017)	55.383 (1.185)	42.425 (2.088)
Race (%)				0.075
Mexican American	7.831 (0.014)	6.959 (1.269)	10.134 (1.990)
Other Hispanic	6.437 (0.008)	6.426 (0.799)	6.468 (1.171)
Non-Hispanic white	65.049 (0.039)	64.822(2.826)	65.649 (3.417)
Non-Hispanic black	11.295 (0.015)	12.178 (1.619)	8.963 (1.489)
Other races	9.387 (0.010)	9.615 (1.300)	8.786 (1.524)
Education (%)				0.003
<11th grade	9.069 (0.008)	9.116 (0.882)	8.946 (0.843)
High school graduate	26.863 (0.019)	25.561 (1.957)	30.302 (1.564)
Some college	30.961 (0.013)	29.566 (1.727)	34.647 (1.615)
College graduate or above	33.107 (0.031)	35.757 (3.021)	26.106 (2.926)
Smoking (%)				0.719
No	83.925 (0.027)	84.052 (1.436)	83.592 (1.148)
Yes	16.075 (0.014)	15.948 (1.436)	16.408 (1.148)
Hypertension (%)				< 0.0001
No	60.443 (0.023)	68.466 (1.548)	39.253 (3.291)
Yes	39.557 (0.023)	31.534 (1.548)	60.747 (3.291)
Diabetes (%)				< 0.0001
No	85.164 (0.027)	91.842 (0.829)	67.526 (1.686)
Yes	14.836 (0.008)	8.158 (0.829)	32.474 (1.686)
Age (years)	48.409 (0.649)	46.988 (0.657)	52.163 (0.811)	< 0.0001
Energy (kcal)	2102.103 (20.211)	2081.996 (25.177)	2155.214 (44.513)	0.202
Weight(kg)	84.401 (0.876)	77.980 (0.748)	101.361 (1.252)	< 0.0001
Hip (cm)	107.902 (0.489)	104.385 (0.454)	117.191 (0.746)	< 0.0001
WC (cm)	100.973 (0.777)	95.513 (0.685)	115.394 (1.004)	< 0.0001
BMI (kg/m²)	29.760 (0.291)	27.715 (0.287)	35.164 (0.388)	< 0.0001
BRI	5.673 (0.108)	4.952 (0.100)	7.575 (0.156)	< 0.0001
BAI	31.802 (0.238)	30.413 (0.259)	35.470 (0.418)	< 0.0001
WTHR	0.933 (0.003)	0.913 (0.003)	0.986(0.004)	< 0.0001
WHTR	0.602 (0.005)	0.571 (0.004)	0.682 (0.006)	< 0.0001
conicity index	1.701 (0.006)	1.668 (0.004)	1.791 (0.007)	< 0.0001
ABSI	0.0816 (0.000)	0.0810 (0.000)	0.0831 (0.000)	< 0.0001
LSM (kPa)	5.923 (0.103)	5.162 (0.082)	7.934 (0.346)	< 0.0001
Alcohol (g)	11.601 (0.579)	11.857 (0.717)	10.925 (1.600)	0.635
Minutes sedentary activity	354.478 (7.742)	346.132 (7.217)	376.522 (14.009)	0.029

Table2 The correlation between anthropometric measures and MAFLD

	Multivariate Logistic Regression of MAFLD			Multivariate Ordinal Logistic Regression of MAFLD Phenotypes
Exposure	Model 1^a	Model 2^b	Model 3^c	Model 1^a	Model 2^b	Model 3^c
Weight per SE	1.053(1.047,1.060)<0.0001	1.063(1.056,1.069)<0.0001	1.057(1.052,1.063)<0.0001	1.056(1.051, 1.060)<0.0001	1.065 (1.061, 1.069)<0.0001	1.060 (1.056,1.064)<0.0001
T1 (Reference)	1.000	1.000	1.000	1.000	1.000	1.000
T2	2.898(2.174,3.863)<0.0001	2.987(2.245,3.973)<0.0001	2.447(1.830,3.272)<0.0001	2.904(2.39,3.766)<0.0001	2.992(2.340, 3.828)<0.0001	2.472(1.894,3.227)<0.0001
T3	13.998(9.655,20.294)<0.0001	17.965(12.494,25.832)<0.0001	13.020(9.476,17.891)<0.0001	14.401(10.426,19.890)<0.0001	18.003(13.430,24.1134)<0.0001	12.738(9.529,17.027)<0.0001
Hip per SE	1.034(1.030,1.039)<0.0001	1.043(1.037,1.049)<0.0001	1.039(1.033,1.044)<0.0001	1.074(1.066,1.082)<0.0001	1.092(1.084,1.101)<0.0001	1.083(1.075,1.092)<0.0001
T1 (Reference)	1.000	1.000	1.000	1.000	1.000	1.000
T2	3.903(2.872, 5.303)<0.0001	3.869(2.690, 5.563)<0.0001	3.567(2.511, 5.067)<0.0001	3.820(2.928,4.985)<0.0001	3.772(2.931,5.027)<0.0001	3.490(2.583,4.715)<0.0001
T3	10.628(8.414,13.423)<0.0001	14.207(9.986,20.211)<0.0001	11.481(8.385,15.720)<0.0001	11.175(9.083,13.749)<0.0001	15.015(11.443,19.704)<0.0001	11.965(9.030,15.855)<0.0001
WC per SE	1.068(1.061,1.076)<0.0001	1.070(1.062,1.079)<0.0001	1.064(1.057,1.070)<0.0001	1.092(1.085,1.099)<0.0001	1.095(1.088,1.102)<0.0001	1.087(1.080,1.094)<0.0001
T1 (Reference)	1.000	1.000	1.000	1.000	1.000	1.000
T2	6.104( 3.906, 9.540)<0.0001	5.232( 3.205, 8.540)<0.0001	4.493( 2.985, 6.761)<0.0001	5.997(4.023,8.940)<0.0001	5.184(3.449,7.791)<0.0001	4.463(3.048,6.533)<0.0001
T3	28.446(17.590,46.001)<0.0001	26.789(15.991,44.878)<0.0001	19.472(13.046,29.064)<0.0001	29.975(19.627,45.776)<0.0001	28.439(18.646,43.374)<0.0001	20.504(14.029,29.966)<0.0001
BMI per SE	1.051(1.042,1.059)<0.0001	1.057(1.048,1.066)<0.0001	1.051(1.044,1.059)<0.0001	1.188(1.164,1.211)<0.0001	1.215(1.192,1.238)<0.0001	1.196(1.175,1.217)<0.0001
T1 (Reference)	1.000	1.000	1.000	1.000	1.000	1.000
T2	5.837( 3.748, 9.089)<0.0001	5.040( 3.103, 8.186)<0.0001	4.419( 2.847, 6.858)<0.0001	5.699(3.835,8.471)<0.0001	4.910(3.310,7.284)<0.0001	4.293(2.934,6.281)<0.0001
T3	22.002(14.561,33.245)<0.0001	23.757(15.090,37.403)<0.0001	18.099(12.509,26.187)<0.0001	23.449(16.309,33.716)<0.0001	25.201(17.578,36.130)<0.0001	19.080(13.757,26.463)<0.0001
BRI per SE	1.055(1.048,1.063)<0.0001	1.062(1.052,1.072)<0.0001	1.056(1.048,1.064)<0.0001	1.656(1.568,1.750)<0.0001	1.773(1.670,1.883)<0.0001	1.684(1.588,1.786)<0.0001
T1 (Reference)	1.000	1.000	1.000	1.000	1.000	1.000
T2	6.121( 3.886, 9.641)<0.0001	5.597( 3.291, 9.518)<0.0001	4.850( 3.067, 7.669)<0.0001	6.006 (3.998, 9.022)<0.0001	5.553(3.588,8.594)<0.0001	4.775(3.132,7.281)<0.0001
T3	23.402(13.542,40.442)<0.0001	25.840(13.613,49.050)<0.0001	19.110(11.195,32.623)<0.0001	24.953(17.066,46.511)<0.0001	28.174 (17.066,46.512)<0.0001	20.760(12.901,33.407)<0.0001
BAI per SE	1.020(1.016,1.023)<0.0001	1.038(1.030,1.046)<0.0001	1.034(1.027,1.041)<0.0001	1.088(1.074,1.102)<0.0001	1.175(1.149,1.201)<0.0001	1.155(1.139,1.180)<0.0001
T1 (Reference)	1.000	1.000	1.000	1.000	1.000	1.000
T2	2.266(1.656,3.103)<0.0001	3.582(2.416, 5.313)<0.0001	3.359(2.483, 4.544)<0.0001	2.295 (1.725,3.053)<0.0001	3.726(2.700,5.142)<0.0001	3.497(2.645,4.623)<0.0001
T3	4.030(2.921,5.560)<0.0001	14.319(7.634,26.858)<0.0001	10.718(6.066,18.935)<0.0001	4.246 (3.198,5.638)<0.0001	16.151(9.912,26.317)<0.0001	12.046(7.217,20.107)<0.0001
WTHR per SE	1.045(1.039,1.051)<0.0001	1.051(1.044,1.059)<0.0001	1.044(1.038,1.050)<0.0001	1.045(1.041,1.049)<0.0001	1.052(1.046,1.057)<0.0001	1.045 (1.914,3.481)<0.0001
T1 (Reference)	1.000	1.000	1.000	1.000	1.000	1.000
T2	3.035(2.140, 4.304)<0.0001	3.073(2.074, 4.553)<0.0001	2.558(1.858,3.523)<0.0001	3.016(2.202,4.131)<0.0001	3.054(2.212,4.216)<0.0001	2.590(1.927,3.482)<0.0001
T3	8.850(6.739,11.622)<0.0001	9.509(6.715,13.465)<0.0001	6.938(5.208,9.242)<0.0001	8.776(7.017,10.976)<0.0001	9.381(7.178,12.260)<0.0001	6.833 (5.273,8.856)<0.0001
WHTR per SE	1.065(1.057,1.073)<0.0001	1.073(1.062,1.084)<0.0001	1.066(1.057,1.075)<0.0001	1.067(1.060,1.074)<0.0001	1.077(1.0689,1.085)<0.0001	1.070(1.062.1.078)<0.0001
T1 (Reference)	1.000	1.000	1.000	1.000	1.000	1.000
T2	6.281( 4.011, 9.836)<0.0001	5.723( 3.399, 9.634)<0.0001	4.950( 3.159, 7.756)<0.0001	6.166(4.126,9.213)<0.0001	5.683(3.707,8.711)<0.0001	4.888(3.233,7.391)<0.0001
T3	24.781(14.774,41.567)<0.0001	27.221(14.669,50.516)<0.0001	20.166(12.131,33.523)<0.0001	26.420(16.787,41.577)<0.0001	29.701 (18.385,47,982)<0.0001	21.922(13.874,34.389)<0.0001
ABSI per SE	1.020(1.015,1.024)<0.0001	1.016(1.011,1.022)<0.0001	1.012(1.007,1.017)<0.0001	1.019(1.014, 1.024) <0.0001	1.019(1.014,1.024)<0.0001	1.008 (1.004,1.013)<0.0001
T1 (Reference)	1.000	1.000	1.000	1.000	1.000	1.000
T2	2.351(1.779,3.107)<0.0001	2.063(1.488,2.860)<0.0001	1.790(1.326,2.415<0.0001	2.408 (1.848,3.138)<0.0001	2.129(1.601,2.821)<0.0001	1.871(1.398,2.505)<0.0001
T3	3.097(2.255,4.254)<0.0001	2.523(1.739,3.660)<0.0001	1.968(1.387,2.794)<0.0001	3.065 (2.290,4.101)<0.0001	2.541(1.848,3.492)<0.0001	1.987(1.416,2.787)<0.0001
conicity index per SE	1.053(1.046,1.060)<0.0001	1.056(1.048,1.063)<0.0001	1.048(1.042,1.054)<0.0001	1.054(1.048,1.059)<0.0001	1.057(1.052,1.063)<0.0001	1.050(1.044,1.054)<0.0001
T1 (Reference)	1.000	1.000	1.000	1.000	1.000	1.000
T2	3.505(2.489, 4.936)<0.0001	3.414(2.353, 4.952)<0.0001	2.872(2.106, 3.916)<0.0001	3.512(2.579,4.784)<0.0001	3.473(2.549,4.731)<0.0001	2.914(2.191,3.876)<0.0001
T3	11.554(7.441,17.938)<0.0001	11.831(7.211,19.410)<0.0001	8.388(5.435,12.946)<0.0001	12.503(8.029,17.572)<0.0001	12.503(8.284,18.870)<0.0001	8.912(5.901,13.459)<0.0001

Multivariate ordinal regression models (Non-MAFLD vs. MAFLD without fibrosis vs. MAFLD with fibrosis)

^aNo-adjusted model: adjusted for None
^bMinimally adjusted model: adjusted for sex, age，race

^cFully adjusted model: adjusted for sex, age, race, education, smoking, alcohol, minutes sedentary activity, drug (take prescription for cholesterol), energy, hypertension, diabetes

Abbreviation: a body shape index (ABSI), body roundness index (BRI), body adiposity index (BAI), WC-to-hip ratio (WTHR), WC-to-height ratio (WHTR), body adiposity index (BAI)

Table 3 Pooled accuracy parameters in the prediction of MFALD

Exposure	Cut				Weighted
Exposure	value	Sensitivity	Specificity	AUC	AUC
WC (cm)	104.35	0.706(0.677,0.733)	0.743(0.715,0.770)	0.801(0.775, 0.825)	0.832(0.808, 0.854)
BRI	5.545	0.791(0.765,0.815)	0.647(0.617,0.676)	0.786(0.760,0.810)	0.812(0.787,0.850)
WHTR	0.605	0.791(0.765,0.815)	0.647(0.617,0.676)	0.786(0.760,0.810)	0.812(0.787,0.850)
BMI (kg/m2)	29.65	0.744(0.716,0.770)	0.677(0.637,0.705)	0.779(0.752,0.804)	0.806(0.780,0.829)
Weight	87.45	0.645(0.615,0.674)	0.765(0.738,0.790)	0.763(0.736,0.789)	0.794(0.768,0.818)
Conicity index	1.695	0.768(0.741,0.793)	0.576(0.545,0.606)	0.736(0.708,0.762)	0.769(0.742,0.794)
Hip	103.65	0.786(0.760,0.810)	0.563(0.532,0.593)	0.733(0.704,0.760)	0.763(0.736,0.789)
WTHR	0.935	0.764(0.736,0.789)	0.568(0.537,0.598)	0.732(0.704,0.759)	0.754 (0.726,0.780)
BAI	28.65	0.792(0.766,0.816)	0.455(0.424,0.486)	0.665(0.635,0.694)	0.680 (0.650,0.708)
ABSI	0.082	0.619(0.589,0.649)	0.557(0.526,0.588)	0.609(0.578,0.639)	0.633(0.603,0.662)

Table4 The correlation between anthropometric measures and MAFLD in difference sex

Exposure	Male	Female	P for interaction
WC per SE	1.072(1.060,1.085) <0.0001	1.061(1.050,1.073) <0.0001	0.034
BRI per SE	1.071(1.056,1.087) <0.0001	1.050(1.040,1.060) <0.0001	0.003
WHTR per SE	1.080(1.065,1.095) <0.0001	1.061(1.049,1.073) <0.0001	0.004
BMI per SE	1.067(1.055,1.079) <0.0001	1.045(1.035,1.056) <0.0001	0.005
Weight per SE	1.062(1.052,1.071) <0.0001	1.057(1.044,1.069) <0.0001	0.242
Conicity index per SE	1.059(1.045,1.073) <0.0001	1.041(1.032,1.050) <0.0001	0.045
WTHR per SE	1.052(1.042,1.062) <0.0001	1.038(1.025,1.051) <0.0001	0.168
Hip per SE	1.050(1.039,1.062) <0.0001	1.034(1.026,1.042) <0.0001	0.023
BAI per SE	1.049(1.035,1.062) <0.0001	1.029(1.022,1.037) <0.0001	0.004
ABSI per SE	1.018(1.008,1.029 <0.0001	1.007(1.000,1.016) <0.0001	0.174

Adjusted for sex, age, race, education, smoking, alcohol, minutes sedentary activity, drug (take prescription for cholesterol), energy, hypertension, diabetes

Abbreviation: a body shape index (ABSI), body roundness index (BRI), body adiposity index (BAI), WC-to-hip ratio (WTHR), WC-to-height ratio (WHTR), body adiposity index (BAI)

Table 5 The correlation between anthropometric measures and MAFLD in difference age groups

Exposure	<40	40-65	≥65	P for interaction
WC per SE	1.059(1.047, 1.071) <0.0001	1.064(1.057,1.070) <0.0001	1.061(1.037,1.085) <0.0001	0.362
BRI per SE	1.053(1.038, 1.069) <0.0001	1.056(1.048,1.064) <0.0001	1.055(1.034,1.076) <0.0001	0.770
WHTR per SE	1.064(1.048, 1.080) <0.0001	1.066(1.057,1.075) <0.0001	1.064(1.040,1.089) <0.0001	0.774
BMI per SE	1.050(1.039, 1.060) <0.0001	1.051(1.044,1.059) <0.0001	1.053(1.033,1.075) <0.0001	0.275
Weight per SE	1.053(1.044, 1.062) <0.0001	1.057(1.052,1.063) <0.0001	1.058(1.035,1.081) <0.0001	0.114
Conicity index per SE	1.049(1.034, 1.064) <0.0001	1.048(1.042,1.054) <0.0001	1.032(1.015,1.050) <0.0001	0.084
Hip per SE	1.038(1.029, 1.048) <0.0001	1.039(1.033,1.044) <0.0001	1.034(1.017,1.050) <0.001	0.980
WTHR per SE	1.051(1.038, 1.065) <0.0001	1.044(1.038,1.050) <0.0001	1.031(1.013,1.049) 0.0002	0.017
BAI per SE	1.036(1.026, 1.047) <0.0001	1.034(1.027,1.041) <0.0001	1.031(1.015,1.046) <0.001	0.260
ABSI per SE	1.012(0.999, 1.026) <0.0001	1.012(1.007,1.017) <0.0001	1.004(0.994,1.014) 0.379	0.036

Adjusted for sex, age, race, education, smoking, alcohol, minutes sedentary activity, drug (take prescription for cholesterol), energy, hypertension, diabetes

Abbreviation: a body shape index (ABSI), body roundness index (BRI), body adiposity index (BAI), WC-to-hip ratio (WTHR), WC-to-height ratio (WHTR), body adiposity index (BAI)

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

The relationship between 10 anthropometric measures and metabolic dysfunction-associated fatty liver disease: among United States population: the National Health and Nutrition Examination Survey 2017–2018

Status:

Version 1

Abstract

Background

Methods

Result

Conclusion

Figures

1. Introduction

2. Methods

2.1 Data source and study participates

2.2 Diagnosis of MAFLD

2.2 Anthropometric measures

2.2.1 BMI

2.2.2 BRI

2.2.3 BAI

2.2.4 WTHR

2.2.5 WHTR

2.2.6 ABSI

2.2.7 Conicity index

2.3 Other covariates

2.3.1 Demographic characteristics and lifestyle

2.3.2 Energy intake

2.3.3 Diseases and medications

2.4 Statistical methods

3. Results

4. Discussion

5. Strengths and limitations

6. Conclusion

Abbreviations

Declarations

References

Tables

Additional Declarations

Status:

Version 1