Abstract
Background:
The prevalence of non-alcoholic fatty liver disease (NAFLD) is increasing in the U.S. and strongly linked to obesity in many, but not all, racial/ethnic groups. It is conceivable that the lack of correspondence is related to differential fat distribution. The study objective was to examine which fat distribution measures best predicted NAFLD by sex within racial/ethnic groups.
Methods:
The analysis included 1,404 participants from the 2017-2018 National Health and Nutrition Examination Survey (NHANES). Area under the receiver operating characteristic curve (AUC) analyses compared the ability of dual energy x-ray absorptiometry measured percent total fat and abdominal fat, and measured BMI, waist circumference and waist-to-height ratio (WHR) to predict ultrasound transient elastography assessed NAFLD in each sex and racial/ethnic group.
Results:
AUC analysis found the best predictors of NAFLD among men were waist circumference and total abdominal fat area [AUC: 84.1%], and among women was visceral fat [AUC: 85.2]. NAFLD prediction by body fat measures, however, was similar between racial/ethnic groups
Conclusion:
The best predictors of NAFLD, using body fat distribution measures, vary by sex but not by racial/ethnic group. While the prediction of NAFLD using body fat measures is good, other racial/ethnic predictors of NAFLD warrant further consideration.
Keywords: Body fat, fatty liver disease, anthropometry, sex, race/ethnicity
Introduction
Non-alcoholic fatty liver disease (NAFLD), a growing health problem worldwide, is associated with increased morbidity and mortality due to its association with diabetes, cancer, liver disease and cardiovascular disease (1–3). Two recent studies using data from the National Health and Nutrition Examination Survey (NHANES) found that 48.0% of U.S. adults had NAFLD, with Hispanic persons having the highest prevalence (56.3%) and non-Hispanic Black persons having the lowest prevalence (40.0%) (4, 5). As the prevalence of NAFLD has increased greatly in a short period, efforts to identify risk factors and slow the increase in prevalence have become a research priority (6).
One of the major risk factors for NAFLD is obesity, and as obesity rates rise, rates for NAFLD are expected to rise even higher (6). In 2017-2018, the obesity prevalence in the U.S. was 42.4%, up 11.9% since 1999-2000 (7). Obesity was highest among non-Hispanic Black persons, the group with the lowest NAFLD rate (7). Conversely, obesity was lowest (17.4%) among the group with the second highest NAFLD rate, non-Hispanic Asian persons (7). The reasons for the lack of correspondence between racial/ethnic rates of obesity and NAFLD are unclear. It is possible, however, that differences in body fat distribution could explain why obesity and NAFLD are not well correlated across racial/ethnic groups. Prior research has found that body fat distribution, including percent total fat mass (8) varies by race/ethnicity. Thus, the aim of the current study was to examine whether differential body fat distribution or anthropometric body measures were better able to predict NAFLD across racial/ethnic groups in the U.S.
Methods
NHANES is representative of the noninstitutionalized United States population (9). The 2017-2018 cycle included 5,533 persons aged 18 years and older who completed both the household interview and had a physical examination at a mobile exam center (MEC). The MEC exam included measured height, weight, and waist circumference. Supplemental Figure 1 displays the exclusion criteria for the current analysis. After all exclusions/eligibility criteria the analytic population included 1,404 persons. For the analysis of android percent fat, gynoid percent fat, visceral fat area and subcutaneous fat area, an additional 110 participants were excluded due to invalid measures (N=1,294). Abbreviations used in the manuscript are: “NHB” for Non-Hispanic Black persons, “NHW” for Non-Hispanic White persons, “Hispanic” for Hispanic/Mexican persons, and “Asian” for Non-Hispanic Asian persons.
Non-alcoholic Fatty liver Disease
The transient elastography (TE) exam used the Fibroscan model 502 V2 Touch system with medium (M), and extra-large (XL) wands to obtain measurements. The controlled attenuation parameter (CAP) ranges from 100-400 dB/m (10). NAFLD was defined as TE-CAP ≥ 263 dB/m (11).
Anthropometric measures
Body mass index (kg/m2), waist circumference (cm), and standing height (cm) were extracted from the body measures dataset. Waist-to-height ratio (WHR) was calculated by dividing waist circumference by height.
Dual energy x-ray Absorptiometry (DXA)
DXA was used to assess body fat distribution. The analysis used the whole body DXA scan dataset to obtain measures for total percent body fat and trunk percent fat(12). The abdominal body fat measures (android, gynoid, android/gynoid ratio, visceral fat area, subcutaneous fat area, total abdominal fat area) were abstracted from the NHANES DXA android/gynoid measurements dataset (13). Android area was defined as the lower trunk area bounded by the pelvic horizontal cut line and the pelvic line. The upper gynoid line was 1.5 times the height of the android region below the pelvic line and the lower gynoid line such that the distance between the two gynoid lines was 2 times the height of the android region. Visceral and subcutaneous adipose tissue were measured at the interspace between the L4 and L5 vertebrae (13).
Statistical Analysis
Medians and interquartile ranges (IQR) [75th-25th percentile] were calculated for age, anthropometric and body fat measures. Proportions were used for categories of education, marital status, income, physical activity, and NAFLD status. Area under the receiver operating characteristic curve (AUC) was used to estimate the accuracy of correctly predicting NAFLD status. AUC’s were computed from sex and race/ethnic stratified logistic regressions with body fat measures as well as age, education, and physical activity as covariates to predict NAFLD status. All statistical analyses were weighted using the MEC survey weights and accounted for other aspects of the NHANES complex sample design (9). Analyses were conducted in STATA SE 16 (College Station, TX). All 95% confidence intervals were two-sided.
Results
Characteristics of the analytic population of 1,404 persons are shown in Supplemental Table 1. As shown in Table 1, 49.2% of the men had NAFLD. Hispanic men had the highest proportion of NAFLD (58.6%), followed by Asian (54.3%), NHW (49.2%) and NHB (34.9%) men. The waist-to-height ratio was greatest among Hispanic men, while BMI and waist circumference were highest among Hispanic and NHW men. Total percent fat and trunk percent fat were similar between Asian and Hispanic men. Asian men had the highest percent android fat. Gynoid fat percentage and android/gynoid ratio was similar in Asian, Hispanic and NHW men. Visceral fat area was highest in NHW and Hispanic men. Hispanic men had the largest trunk/total fat ratio, subcutaneous fat area and total abdominal fat area. NHB men had the lowest levels of all body fat measures, with notably lower levels of visceral fat area, subcutaneous fat area and abdominal fat area.
Table 1.
NAFLD and body measures among men overall and by racial/ethnic group
| All | NHW | NHB | Hispanic | Asian | |
|---|---|---|---|---|---|
|
|
|||||
| Number of participants | 744 | 229 | 190 | 160 | 165 |
| NAFLD | 49.2% | 49.2% | 34.9% | 58.6% | 54.3% |
| Anthropometric measures (median, IQR) | |||||
| Body mass Index | 28.6 (7.4) | 29.1 (7.5) | 27.2 (8.5) | 29.0 (6.8) | 26.3 (4.6) |
| Waist Circumference | 98.9 (21.0) | 100.6 (19.9) | 94.1 (29.1) | 99.2 (16.0) | 94.2 (12.1) |
| Waist-to-Height ratio | 56.5 (12.0) | 57.0 (11.3) | 53.2 (15.3) | 58.6 (9.3) | 54.6 (7.6) |
| Body fat measures (median, IQR) | |||||
| Total percent fat | 27.3 (7.2) | 27.1 (7.1) | 25.0 (11.8) | 28.2 (7.2) | 28.2 (5.8) |
| Trunk percent fat | 28.4 (9.2) | 28.2 (8.8) | 25.1 (13.5) | 30.2 (7.6) | 30.3 (6.0) |
| Trunk/total fat ratio | 49.8 (9.3) | 50.3 (9.3) | 43.1 (8.1) | 51.2 (6.7) | 50.4 (5.6) |
| Android percent fat | 32.7 (10.1) | 32.9 (9.8) | 27.8 (15.0) | 33.6 (9.6) | 34.2 (7.0) |
| Gynoid percent fat | 29.0 (7.5) | 29.1 (7.5) | 27.2 (9.9) | 29.2 (6.8) | 29.7 (6.1) |
| Android/gynoid ratio | 1.1 (0.2) | 1.1 (0.3) | 1.0 (0.2) | 1.1 (0.2) | 1.1 (0.2) |
| Visceral fat area | 98.3 (80.1) | 104.0 (89.1) | 58.0 (61.4) | 100.4 (67.6) | 97.4 (48.5) |
| Subcutaneous fat area | 265.8 (162.0) | 265.9 (162.7) | 213.3 (245.5) | 282.5 (181.4) | 261.8 (105.2) |
| Total abdominal fat area | 375.8 (217.7) | 380.6 (228.8) | 278.4 (295.0) | 402.2 (198.5) | 355.5 (122.4) |
Proportions weighted using 2-year mobile exam sample weights. Estimated weighted proportions may not add up to 100% due to rounding.
Abbreviations: NHW- Non-Hispanic White, NHB- Non-Hispanic Black, H.S.- High school education, IQR – Interquartile range
As shown in Table 2, the overall NAFLD prevalence among women was 40.8%. Hispanic women had the highest prevalence (45.8%), followed by NHW (40.7%), NHB (37.3%) and Asian (36.9%) women. An examination of the anthropometric measures found that NHB women had the highest BMI, waist circumference and waist-to-height ratio. Total percent fat, trunk percent fat, android and gynoid percent fat were similar between NHW, NHB and Hispanic women, and the median android/gynoid ratio was similar among all women. Hispanic women had the largest measured visceral fat area and trunk/total fat ratio, while NHB women had the largest subcutaneous fat and total abdominal fat area.
Table 2.
NAFLD and body measures among women overall and by racial/ethnic group
| All | NHW | NHB | Hispanic | Asian | ||
|---|---|---|---|---|---|---|
|
|
||||||
| Number of participants | 660 | 173 | 157 | 177 | 153 | |
| NAFLD | 40.8% | 40.7% | 37.3% | 45.8% | 36.9% | |
| Anthropometric measures (median, IQR) | ||||||
| Body mass Index | 28.3 (11.6) | 28.3 (13.1) | 31.7 (13.7) | 28.2 (8.7) | 24.5 (7.4) | |
| Waist Circumference | 93.9 (26.4) | 96.0 (30.7) | 99.6 (29.0) | 93.2 (21.1) | 84.5 (17.4) | |
| Waist-to-Height ratio | 58.6 (17.0) | 58.8 (20.5) | 61.8 (17.5) | 59.0 (12.9) | 54.0 (11.6) | |
| Body fat measures (median, IQR) | ||||||
| Total percent fat | 39.9 (9.3) | 39.9 (10.5) | 40.3 (8.5) | 40.5 (7.4) | 37.9 (8.1) | |
| Trunk percent fat | 38.7 (10.6) | 39.3 (10.9) | 39.0 (10.8) | 39.2 (8.5) | 36.5 (9.4) | |
| Trunk/total fat ratio | 45.5 (7.9) | 45.8 (7.36) | 42.4 (8.5) | 46.4 (6.8) | 45.0 (7.8) | |
| Android percent fat | 39.9 (11.9) | 39.9 (12.2) | 40.2 (13.0) | 40.6 (10.0) | 37.9 (9.1) | |
| Gynoid percent fat | 42.5 (7.1) | 42.2 (7.3) | 42.6 (7.9) | 43.1 (6.0) | 41.4 (7.7) | |
| Android/gynoid ratio | 0.9 (0.2) | 0.9 (0.2) | 0.9 (0.2) | 0.9 (0.2) | 0.9 (0.2) | |
| Visceral fat area | 82.2 (72.1) | 83.4 (73.3)) | 69.0 (68.7) | 92.4 (59.5) | 75.6 (59.0) | |
| Subcutaneous fat area | 373.9 (238.7) | 383.6 (241.6) | 412.9 (319.9) | 377.6 (187.8) | 296.8 (141.7) | |
| Total abdominal fat area | 460.1 (306.3) | 468.8 (326.6) | 487.4 (373.8) | 473.4 (261.4) | 373.3 (199.9) | |
*Proportions weighted using 2-year mobile exam sample weights. Estimated weighted proportions may not add up to 100% due to rounding.
Abbreviations: NHW- Non-Hispanic White, NHB- Non-Hispanic Black, H.S.- High school education, IQR – Interquartile ra
Table 3 displays the ability of the body fat measures to predict NAFLD by sex and race/ethnicity. Among men, AUC analysis found that, overall, two models performed equally well; the model that included total abdominal fat area and the model that included waist circumference both had AUCs of 84.1. Abdominal fat area was the best predictor of NAFLD among NHW, NHB and Asian men. In contrast, among Hispanic men, waist-to-height ratio performed as well as abdominal fat area in predicting NAFLD. Among women the best predictor of NAFLD overall and in each racial/ethnic group was visceral fat area (AUC: 85.2).
Table 3.
Area under the curve (%) analysis of the classification of NAFLD by anthropometric and body fat measures by sex and race/ethnicity groups.
| All* | NHW | NHB | Hispanic | Asian | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| AUC | 95% CI | AUC | 95% CI | AUC | 95% CI | AUC | 95% CI | AUC | 95%CI | ||
|
|
|
|
|
||||||||
| Men | |||||||||||
| Model 1 | 67.2 | (63.2, 71.2) | 63.3 | (59.2, 67.5) | 65.1 | (61.0, 69.2) | 62.2 | (58.0, 66.3) | 62.3 | (58.1, 66.5) | |
| Anthropometric measures | |||||||||||
| Model 2 | 83.6 | (80.6, 86.6) | 82.1 | (78.9, 85.2) | 82.4 | (79.3, 85.5) | 80.5 | (77.3, 83.7) | 82.5 | (79.4, 85.6) | |
| Model 3 | 84.1 | (81.2, 87.1) | 82.8 | (79.7, 85.8) | 83.2 | (80.2, 86.2) | 81.4 | (78.2, 84.5) | 83.3 | (80.3, 86.3) | |
| Model 4 | 83.9 | (80.9, 86.7) | 82.8 | (79.7, 85.8) | 83.4 | (80.4, 86.4) | 82.4 | (79.3, 85.4) | 83.5 | (80.5, 86.5) | |
| Body fat measures | |||||||||||
| Model 5 | 80.9 | (77.7, 84.1) | 79.7 | (76.4, 83.0) | 80.5 | (77.2, 83.7) | 79.5 | (76.2, 82.8) | 80.6 | (77.3, 83.8) | |
| Model 6 | 73.9 | (70.2, 77.5) | 71.6 | (67.8, 75.4) | 73.1 | (69.4, 76.8) | 71.7 | (67.9, 75.5) | 73.1 | (69.3, 76.8) | |
| Model 7 | 76.8 | (73.1, 80.4) | 74.2 | (70.5, 77.9) | 77.3 | (73.7, 80.8) | 76.8 | (73.2, 80.4) | 77.1 | (73.5, 80.7) | |
| Model 8 | 81.9 | (78.8, 85.1) | 81.0 | (77.8, 84.2) | 82.3 | (79.2, 85.3) | 81.2 | (78.1, 84.4) | 82.2 | (79.2, 85.3) | |
| Model 9 | 83.3 | (80.3, 86.3) | 81.9 | (78.8, 85.1) | 82.6 | (79.5, 85.7) | 81.1 | (77.9, 84.3) | 82.8 | (79.7, 85.9) | |
| Model 10 | 84.1 | (81.2, 87.1) | 83.4 | (80.4, 86.4) | 83.8 | (80.8, 86.7) | 82.4 | (79.4, 85.5) | 83.9 | (80.9, 86.8) | |
| Women | |||||||||||
| Model 1 | 64.5 | (59.8, 69.2) | 63.5 | (58.7, 68.3) | 64.2 | (59.5, 68.9) | 60.9 | (56.1, 65.7) | 62.0 | (57.2, 66.8) | |
| Anthropometric measures | |||||||||||
| Model 2 | 83.9 | (80.6, 87.2) | 81.1 | (77.6, 84.7) | 81.6 | (78.0, 85.1) | 82.2 | (78.8, 85.7) | 82.2 | (78.8, 85.6) | |
| Model 3 | 83.5 | (80.2, 86.8) | 80.5 | (76.9, 84.1) | 81.1 | (77.5, 84.7) | 81.7 | (78.1, 85.2) | 81.0 | (77.5, 84.5) | |
| Model 4 | 84.1 | (80.8, 87.4) | 81.5 | (78.0, 85.0) | 82.3 | (78.8, 85.8) | 82.6 | (79.2, 86.0) | 82.3 | (78.9, 85.8) | |
| Body fat measures | |||||||||||
| Model 5 | 78.0 | (74.2, 81.8) | 77.7 | (73.8, 81.5) | 77.2 | (73.3, 81.1) | 77.3 | (73.4, 81.1) | 76.2 | (72.3, 80.1) | |
| Model 6 | 65.5 | (60.9, 70.1) | 64.8 | (60.1, 69.4) | 64.5 | (59.9, 69.2) | 62.8 | (58.1, 67.5) | 63.1 | (58.3, 67.8) | |
| Model 7 | 79.9 | (76.2, 83.6) | 78.9 | (75.1, 82.7) | 79.3 | (75.5, 83.0) | 78.9 | (75.2, 82.7) | 79.2 | (75.5, 82.9) | |
| Model 8 | 85.2 | (82.0, 88.3) | 84.4 | (81.2, 87.7) | 84.2 | (80.9, 87.6) | 84.1 | (80.7, 87.4) | 84.5 | (81.3, 87.7) | |
| Model 9 | 80.3 | (76.7, 83.9) | 78.7 | (75.0, 82.5) | 78.0 | (74.1, 81.9) | 79.2 | (75.4, 82.9) | 78.4 | (74.6, 82.1) | |
| Model 10 | 82.7 | (79.3, 86.1) | 81.1 | (77.6, 84.7) | 80.7 | (77.0, 84.3) | 81.5 | (77.9, 85.0) | 81.1 | (77.5, 84.6) | |
Model 1- physical activity + age + education
Model 2 -Model 1 + BMI
Model 3- Model 1 + waist circumference
Model 4- Model 1 + WHR
Model 5-Model 1 + android measurement
Model 6- Model 1 + gynoid measurement
Model 7- Model 1 + Android /Gynoid ratio
Model 8- Model 1 + Visceral Fat area
Model 9- Model 1 + Subcutaneous area
Model 10- Model 1 + Total abdominal fat area
Model was also adjusted for race/ethnicity.
The analysis of all persons who had a whole body DXA scan is shown in Supplemental Table 2. The models that included total body fat and trunk percent fat did not predict NAFLD as well the models shown in Table 3.
Discussion
The current study found that body fat and anthropometry are good predictors of NAFLD, however the best predictors were measures of abdominal size (total abdominal fat, waist circumference, waist-to-hip ratio) among men, and visceral fat among women. The predictions did not vary greatly by race/ethnicity.
Other studies have examined the association of body fat distribution with NAFLD by sex. The Rotterdam study found that android fat mass was strongly associated with NAFLD, with the association strongest among women (14). Trunk fat mass was the second strongest predictor among both sexes (14). The current study also found that android measures were strong predictors of NAFLD, however total abdominal fat and visceral fat were better predictors. An Iranian study found that total fat percent was the best predictor of NAFLD among both sexes (15). The differences in study results could be due to differences in body fat distribution among populations. To our knowledge no prior study has examined prediction of NAFLD using body fat distribution measures by race/ethnicity.
There are a variety of other body composition metrics that could explain the racial/ethnic disparities in NAFLD prevalence. Measures such as bone density and/or skeletal muscle differ by race/ethnicity and thus may contribute to differences in NAFLD (8, 16). In addition, environmental and social factors could be related to disparities in NAFLD prevalence. For example, arsenic is an environmental pollutant that has been reported to be associated with NAFLD, particularly among Mexican-Americans (17). Genetic susceptibility may also explain some of the disparity. Studies have found that genetic polymorphisms in loci such as PNPLA3 are related to NAFLD and have differential distribution across racial/ethnic groups(18).
A strength of the current study is that it used a nationally representative survey of the U.S population. The study had multiple different DXA measures of body fat distribution and used TE measurements, a highly accurate measure of steatosis, to estimate NAFLD (19). A limitation of the study is that the NHANES 2017-2018 cycle restricted DXA evaluation to persons <60 years old. Studies have shown that NAFLD prevalence increases with age (20). In addition, there was no external validation of the computed AUCs as this was first time that TE measures were collected in a national sample. Future studies should examine whether the current results are reproducible in other populations.
Supplementary Material
Supplementary Figure 1. Flowchart of National Health and Nutrition Examination Survey 2017-2018 participants. Description of inclusion criteria for the analytic population. DXA, dual energy x-ray absorptiometry; NHW, Non-Hispanic white; NHB, Non-Hispanic Black; NHANES, National Health and Nutrition Examination Survey.
Study Importance.
What is already known about this subject?
Fatty liver disease is a growing health concern in the United States but is not equally distributed by sex and race/ethnicity.
Obesity is the strongest risk factor for fatty liver disease.
What are the new findings in your manuscript?
In a nationally representative survey, waist circumference and total abdominal fat measures were the strongest predictors of NAFLD in men, and visceral fat was the strongest predictor in women.
Compared to measured body fat, anthropometric measures are strong predictors of NAFLD.
Models for predicting NAFLD were consistent across race/ethnicity.
How might your results change the direction of research or the focus of clinical practice?
Non-invasive models using DXA based measures of body fat or anthropometric measures could have clinical utility in NAFLD predictions and may have similar accuracy across racial/ethnic groups.
Financial Support:
National Cancer Institute Intramural Research Program
Footnotes
Conflict of Interest: No conflicts to disclose
References
- 1.Ray K NAFLD-the next global epidemic. Nat Rev Gastroenterol Hepatol. 2013;10(11):621. [DOI] [PubMed] [Google Scholar]
- 2.Perumpail BJ, Khan MA, Yoo ER, Cholankeril G, Kim D, Ahmed A. Clinical epidemiology and disease burden of nonalcoholic fatty liver disease. World J Gastroenterol. 2017;23(47):8263–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Alvarez CS, Graubard BI, Thistle JE, Petrick JL, McGlynn KA. Attributable Fractions of Nonalcoholic Fatty Liver Disease for Mortality in the United States: Results From the Third National Health and Nutrition Examination Survey With 27 Years of Follow-up. Hepatology. 2020;72(2):430–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Jones GS, Alvarez CS, Graubard BI, McGlynn KA. Agreement Between the Prevalence of Nonalcoholic Fatty Liver Disease Determined by Transient Elastography and Fatty Liver Indices. Clin Gastroenterol Hepatol. 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kim D, Cholankeril G, Loomba R, Ahmed A. Prevalence of Fatty Liver Disease and Fibrosis Detected by FibroScan in Adults in the United States, 2017-2018. Clin Gastroenterol Hepatol. 2020. [DOI] [PubMed] [Google Scholar]
- 6.Mitra S, De A, Chowdhury A. Epidemiology of non-alcoholic and alcoholic fatty liver diseases. Transl Gastroenterol Hepatol. 2020;5:16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hales CM, Carroll MD, Fryar CD, Ogden CL. Prevalence of Obesity and Severe Obesity Among Adults: United States, 2017-2018. NCHS Data Brief. 2020(360):1–8. [PubMed] [Google Scholar]
- 8.Heymsfield SB, Peterson CM, Thomas DM, Heo M, Schuna JM Jr. Why are there race/ethnic differences in adult body mass index-adiposity relationships? A quantitative critical review. Obes Rev. 2016;17(3):262–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.National Health and Nutrition Examination Survey 2017-2018. Sample design and estimation procedures manual 2015-2018. Hyattsville, MD.: Centers for Disease Control and Prevention (CDC).National Center for Health Statistics (NCHS). Accessed date: November 1, 2021. Available at [https://wwwn.cdc.gov/nchs/nhanes/analyticguidelines.aspx] . [Google Scholar]
- 10.National Health and Nutrition Examination Survey Data 2017-2018. Liver Ultrasound Transient Elastography. Hyattsville, MD: Centers for Disease Control and Prevention (CDC). National Center for Health Statistics (NCHS). Accessed date: November 1, 2021. Available at [https://wwwn.cdc.gov/Nchs/Nhanes/2017-2018/LUX_J.htm]. [Google Scholar]
- 11.Siddiqui MS, Vuppalanchi R, Van Natta ML, Hallinan E, Kowdley KV, Abdelmalek M, et al. Vibration-Controlled Transient Elastography to Assess Fibrosis and Steatosis in Patients With Nonalcoholic Fatty Liver Disease. Clin Gastroenterol Hepatol. 2019;17(1):156–63 e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.National Health and Nutrition Examination Survey Data 2017-2018. Dual-Energy X-Ray Absorptiometry-Whole Body. Hyattsville, MD: Centers for Disease Control and Prevention (CDC). National Center for Health Statistics (NCHS; ). Accessed date: November 1, 2021. Available at [https://wwwn.cdc.gov/Nchs/Nhanes/2017-2018/DXX_J.htm]. [Google Scholar]
- 13.National Health and Nutrition Examination Survey Data 2017-2018. Dual-Energy X-Ray Absorptiometry-Android/Gynoid measurements. Hyattsville, MD: Centers for Disease Control and Prevention (CDC). National Center for Health Statistics (NCHS). Accessed date: November 1, 2021. Available at [https://wwwn.cdc.gov/Nchs/Nhanes/2017-2018/DXXAG_J.htm]. [Google Scholar]
- 14.Alferink LJM, Trajanoska K, Erler NS, Schoufour JD, de Knegt RJ, Ikram MA, et al. Nonalcoholic Fatty Liver Disease in The Rotterdam Study: About Muscle Mass, Sarcopenia, Fat Mass, and Fat Distribution. J Bone Miner Res. 2019;34(7):1254–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ariya M, Koohpayeh F, Ghaemi A, Osati S, Davoodi SH, Razzaz JM, et al. Assessment of the association between body composition and risk of non-alcoholic fatty liver. PLoS One. 2021;16(4):e0249223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Poggiogalle E, Donini LM, Lenzi A, Chiesa C, Pacifico L. Non-alcoholic fatty liver disease connections with fat-free tissues: A focus on bone and skeletal muscle. World J Gastroenterol. 2017;23(10):1747–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Frediani JK, Naioti EA, Vos MB, Figueroa J, Marsit CJ, Welsh JA. Arsenic exposure and risk of nonalcoholic fatty liver disease (NAFLD) among U.S. adolescents and adults: an association modified by race/ethnicity, NHANES 2005-2014. Environ Health. 2018;17(1):6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hernaez R, McLean J, Lazo M, Brancati FL, Hirschhorn JN, Borecki IB, et al. Association between variants in or near PNPLA3, GCKR, and PPP1R3B with ultrasound-defined steatosis based on data from the third National Health and Nutrition Examination Survey. Clin Gastroenterol Hepatol. 2013;11(9):1183–90 e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Eddowes PJ, Sasso M, Allison M, Tsochatzis E, Anstee QM, Sheridan D, 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. 2019;156(6):1717–30. [DOI] [PubMed] [Google Scholar]
- 20.Makri E, Goulas A, Polyzos SA. Epidemiology, Pathogenesis, Diagnosis and Emerging Treatment of Nonalcoholic Fatty Liver Disease. Arch Med Res. 2021;52(1):25–37. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Figure 1. Flowchart of National Health and Nutrition Examination Survey 2017-2018 participants. Description of inclusion criteria for the analytic population. DXA, dual energy x-ray absorptiometry; NHW, Non-Hispanic white; NHB, Non-Hispanic Black; NHANES, National Health and Nutrition Examination Survey.