Sage Journals: Discover world-class research

Abstract

Objectives

Metabolic dysfunction-associated fatty liver disease (MAFLD) poses a major global health burden due to risks of cirrhosis and liver cancer; however, the relationship between MAFLD and weight-adjusted-waist index (WWI) remains unclear. Our study aims to clarify this relationship and identify potential clinical predictors for MAFLD.

Methods

A population-based cross-sectional study was conducted using data from the National Health and Nutrition Examination Survey (NHANES) 1999–2010 and 2015–2018. Multivariable logistic regression and Restricted Cubic Splines (RCS) was employed to examine the nonlinear association between WWI and MAFLD. Following this, a subgroup analysis was performed to detect any variations and ensure the robustness of the findings.

Results

The study included 17,930 participants aged ≥ 18 years, with a mean age of 45.54 years and 48.43% males. The sample was comprised of 44.36% Non-Hispanic White, 20.69% Mexican American, 19.55% Non-Hispanic Black, 8.00% Other Hispanic, and 7.41% Other/Multi-Racial individuals. Among them, 5850 individuals were diagnosed with MAFLD. The study found a statistically significant association between WWI and MAFLD (odds ratio (OR) = 4.35, 95% confidence interval [95% CI]: 4.08–4.63), which remained significant even after adjusting for all relevant factors (OR = 2.73, 95% CI: 2.48–3.01). The RCS analysis revealed a significant nonlinear relationship between WWI and MAFLD (p < 0.05). Subgroup analysis indicated that WWI remained positively correlated with MAFLD, but no significant interaction was observed (p-interaction > 0.05)

Conclusions

WWI is a significant predictor of MAFLD, suggesting its potential as a simple, non-invasive screening tool for the occurrence of MAFLD.

Keywords

Hepatic steatosis MAFLD weight-adjusted-waist index NHANES fatty liver

Introduction

Eslam et al.¹ introduced metabolic dysfunction-associated fatty liver disease (MAFLD) as a distinct clinical entity, separate from nonalcoholic fatty liver disease (NAFLD). Conversely, MAFLD diagnosis does not necessitate ruling out other liver disease causes, such as viral hepatitis or heavy alcohol consumption.¹ The updated name was selected to better reflect the disease's etiology and to offer a more practical framework for clinical practice, given its strong association with metabolic syndrome.² Approximately 25% of the global adult population is impacted by this condition, posing significant challenges to public health and economies worldwide. Despite its prevalence, there are currently no approved pharmacological treatments for MAFLD.³

Fatty liver can further develop into cirrhosis and even hepatocellular carcinoma, significantly impacting human health.³ While tissue biopsy is the definitive method for diagnosing fatty liver, its invasive and costly nature limits its practicality in clinical settings. A commonly used noninvasive examination for detecting fatty liver disease is liver ultrasound transient elastography.⁴ However, its effectiveness is hindered by a steep learning curve.⁵ Consequently, it is necessary to identify simpler indicators related to MAFLD.

Central obesity, particularly its visceral component, is recognized as a pathogenic driver of metabolic syndrome.⁶ Additionally, substantial evidence indicates that a significant proportion of fatty liver patients maintain a normal body weight.⁷ The limitations of body mass index (BMI) in differentiating between lean body mass and fat body mass have raised concerns regarding its accuracy.^8,9 Consequently, some researchers have proposed using the weight-adjusted-waist index (WWI), a new and simple physical measurement derived from waist circumference (WC) and weight.¹⁰ This index serves as a measure of central obesity and takes into account the composition of fat and muscle mass, independent of BMI.¹¹ Numerous studies have established a relationship between WWI and NAFLD.^12,13 However, the association between WWI and MAFLD remains unclear.

The National Health and Nutrition Examination Survey (NHANES) is a crucial survey that evaluates the health and nutritional status of a representative sample of individuals across the United States. This survey uniquely combines interview data—on various topics, ranging from diet to chronic diseases, with direct physical examinations and laboratory testing. The examinations include a diverse set of physiological and functional measurements, while the laboratory analyses focus on biomarkers related to metabolism, cardiovascular risks, environmental exposures, and more. The NHANES study was approved by the National Center for Health Statistics Ethics Review Board, with all participants signing informed consent forms.¹⁴ Data collected through NHANES are considered reliable and nationally representative and are readily available online for researchers and the public.

We hypothesized that higher WWI is positively associated with MAFLD prevalence among US adults. Consequently, this study aimed to investigate the relationship between WWI and MAFLD using a representative sample of US adults from the NHANES data (1999–2010 and 2015–2018).

Materials and methods

Study population

The inclusion criteria for this study were as follows: (1) individuals participating in the NHANES between 1999 and 2010 and 2015 and 2018 and (2) age ≥18 years. Exclusion criteria included: (1) missing fasting weight data; (2) missing US-fatty liver index (US-FLI) data; (3) unable to definitively diagnose MAFLD; and (4) missing high-sensitivity C-reactive protein (hs-CRP) test data. Data from 2011 to 2014 were excluded from this study due to the absence of hs-CRP measurements, which is a key diagnostic criterion for MAFLD. Additionally, data collected from 2019 to 2023 were omitted because of the potential confounding effects of the corona virus disease 2019 (COVID-19) pandemic. During this period, population activities were restricted due to the highly infectious nature of the virus, and dietary patterns among many populations may have significantly changed, potentially impacting MAFLD outcomes.¹⁵ As a result, 81,385 individuals participated in the NHANES studies during the periods 1999–2010 and 2015–2018. Additionally, our research excluded participants younger than 18 years (n = 34,158) and those without fasting weight data (n = 28,249). Participants missing the US-FLI data (n = 979) and those unable to confirm the presence of MAFLD (n = 69) were excluded from the analysis. Ultimately, our cross-sectional retrospective study included 17,930 participants, as illustrated in Figure 1. We conducted our study in accordance with the Helsinki Declaration of 1975 as revised in 2024. All patient information has been de-identified. And the reporting of our study conforms to STROBE guidelines for cross-sectional studies.¹⁶

Figure 1.

A flowchart of showing the selection of study participants.

Outcome variable: MAFLD

The US-FLI is a quantitative measure derived from clinical parameters including age, race, WC, gamma-glutamyl transferase (GGT) activity, fasting insulin levels, and fasting blood glucose levels. It is used to evaluate the presence of hepatic steatosis. A score of 30 or higher indicates a diagnosis of hepatic steatosis.¹⁷ MAFLD diagnosis necessitates hepatic steatosis and at least one metabolic criteria¹: overweight or obesity (BMI ≥ 25 kg/m²), a history of diabetes or glycohemoglobin (HbA1c) levels ≥6.5%,¹⁸ or more than one metabolic risk abnormality: (a) WC ≥ 102 cm for males or 88 cm for females; (b) blood pressure ≥ 130/85 mmHg or specific medication therapy; (c) plasma triglycerides (TGs) ≥ 150 mg/dL or specific medication therapy; (d) plasma high-density lipoprotein (HDL) <40 mg/dL for males and <50 mg/dL for females or specific drug therapy; (e) homeostatic model assessment for insulin resistance (HOMA-IR) score ≥2.5; (f) prediabetes defined as fasting plasma glucose levels 5.6–6.9 mmol/L, or HbA1c levels 5.7–6.4%; and (g) plasma hs-CRP level >2 mg/L.^1,3 Participants in this study were categorized into groups based on the diagnostic criteria for MAFLD and non-MAFLD as outlined above.

Exposure variable: WWI

WWI is a novel indicator based on WC and weight, providing a measure of central obesity. It is calculated by dividing WC (cm) by the square root of weight (kg).¹⁹ Measurements for WC and weight were taken at a mobile examination center, and detailed procedures for these measurements can be found in the anthropometric procedures video illustrating the NHANES III (https://wwwn.cdc.gov/nchs/nhanes/nhanes3/anthropometricvideos.aspx). WWI was initially analyzed as a continuous variable and later divided into four distinct quartile Q1 (8.04–10.35), Q2 (10.36–10.95), Q3 (10.96–11.55), and Q4 (11.56–15.52), which were subsequently reclassified for further examination.

Covariates

Covariates were selected based on prior literature and established variables, which included both categorical and continuous variables. These primarily encompassed sociodemographic characteristics, lifestyle factors, and clinical factors (Supplemental Table 1).

In this study, age was classified into three groups: 18–39, 40–59, and ≥60 years.^12,13 Race/ethnicity was categorized into distinct groups: Mexican American, other Hispanic, non-Hispanic Black, non-Hispanic White, and other racial classifications. Furthermore, household income was divided into three categories according to the poverty-income ratio (PIR): <1.3, 1.3–3.5, and ≥3.5.²⁰ The educational attainment of participants was assessed through interviews and classified into three categories: less than high school, high school graduate or equivalent, and college or higher.

Based on their daily alcohol intake, participants were classified as never, moderate, heavy, or binge drinkers. The criteria for each category are as follows: never drinkers, moderate drinkers (females: 1 drink/day, males: 1–2 drinks/day), heavy drinkers (females: 2–3 drinks/day, males: 3–4 drinks/day), and binge drinkers (females: ≥4 drinks/day, males: ≥5 drinks/day). Additionally, individuals were grouped into never smokers, former smokers, and current smokers based on their self-reported smoking history. Smoking status was determined by having smoked at least 100 cigarettes.³ To evaluate the participants’ levels of physical activity (PA), total PA was calculated by multiplying the number of minutes spent on each activity per week by the corresponding metabolic equivalent of task (MET) score for that activity, and then summing the results.²¹ The activities considered in this assessment included daily activities, leisure-time activities, and sedentary activities at home. The intensity of each activity is quantified using the MET, and this information is available in the NHANES raw data set. In accordance with the US PA guidelines,²² PA is categorized into two groups: high-level PAs and low-level PAs. High-level PA is defined as engaging in activities totaling ≥600 MET·min/week, while low-level PA is defined as engaging in activities totaling <600 MET·min/week.²³

Individuals with diabetes were identified through self-reported diabetes history, glycohemoglobin levels ≥6.5%, or the use of hypoglycemic medications.^18,24 Hypertension was identified based on the average of three blood pressure measurements (systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg), hypertension history, or taking antihypertensive drugs.²⁰ Additionally, individuals with hepatitis B virus (HBV) or hepatitis C virus (HCV) were identified through positive diagnostic tests or self-reported infections.^25,26 HOMA-IR was calculated by multiplying insulin levels (mU/mL) by fasting glucose (mmol/L) divided by 22.5.²⁷ Participants were categorized into three distinct groups according to their BMI: <25, 25–30, and ≥30 kg/m². The examination data encompassed various continuous variables, including weight, height, and WC, alongside laboratory indicators such as alanine aminotransferase (ALT, U/L), aspartate aminotransferase (AST, U/L), alkaline phosphatase (ALP, U/L), GGT (U/L), albumin (ALB, g/L), globulin (GLO, g/L), total bilirubin (TBIL, μmol/L), TG (mg/dL), total cholesterol (TC, mg/dL), HDL (mg/dL), low-density lipoprotein (LDL, mg/dL), platelet (PLT, ×10⁹/L), and hs-CRP (mg/L).

Statistical analysis

The categorical variables were analyzed between groups utilizing χ², with results presented as sample size and percentage. Continuous variables were assessed using weighted linear regression to compare intergroup differences, reported as a weighted mean ± standard error. NHANES employs inferential statistical methods to analyze nationally representative samples. All analyses used fasting weights to account for the complex sampling design and the fasting subsample, following the NHANES guidelines. We conducted a collinearity assessment on all covariates and determined that none of the variables exhibited collinearity (Supplemental Table 2). The association between WWI and MAFLD was evaluated using weighted multivariate logistic regression analysis. Subsequently, WWI was converted into categorical variables, and the weighted multivariate logistic regression analysis was reapplied to evaluate the relationship between WWI and MAFLD. A trend test was also conducted to clarify this association. Multivariable regression models were constructed, incorporating three models to control for potential confounding variables that may influence the outcomes. Model 1 was unadjusted, while model 2 controlled for age, race, and gender. Model 3 included adjustments for age, gender, race/ethnicity, PIR, BMI, alcohol consumption, hypertension, diabetes, HBV infection, HCV infection, and PA. Subgroup analyses were performed to examine the effects of age, gender, and race/ethnicity, followed by interaction analyses. Additionally, restricted cubic splines (RCS) were utilized to investigate the nonlinear association between WWI and MAFLD.

EmpowerStats software (version 2.0) (X&Y Solutions, Inc., Boston, MA, USA) was utilized for data integration and cleaning, and statistical analyses were executed with STATA (version 17.0) (StataCorp, College Station, TX, USA). DecisionLinnc1.0 software (https://www.statsape.com/) was employed to perform RCS analysis for graphical visualization. Statistical significance was determined with a two-tailed p < 0.05.

Results

Baseline characteristics

A total of 81,385 individuals were initially considered for inclusion in our study cycles. After a thorough screening process, 17,930 participants were ultimately enrolled and their clinical and biochemical features were presented according to their MAFLD status in Table 1. Of the participants, 8684 (48.43%) were males, 68.64% were under the age of 60, 44.36% identified as non-Hispanic White, 58.89% reported alcohol consumption, and 50.56% were nonsmokers. The average WWI of the enrolled patients was 10.86 ± 0.84. Participants with MAFLD (n = 5850, 32.63%) exhibited higher BMI, WC, HOMA-IR, and hs-CRP levels, indicating a more severe metabolic profile. Moreover, individuals diagnosed with MAFLD exhibited a higher likelihood of advanced age, elevated WWI levels, and a slightly higher prevalence among males. Participants with diabetes, hypertension, and being overweight/obese were at a higher risk of developing MAFLD. Compared to the non-MAFLD group, patients in the MAFLD cohort exhibited a significantly higher likelihood of being over 60 years of age (41.09% vs. 26.65%), belonging to the highest WWI quartile (Q4) (46.15% vs. 15.14%), having diabetes (25.95% vs. 6.56%), and possessing a BMI of 30 or greater (66.22% vs. 19.36%).

Table 1.

General characteristics of participants.

	Total (n = 17,930)	Non-MAFLD (n = 12,080)	MAFLD (n = 5850)	p value
Age (years)	45.54 ± 17.21	43.34 ± 17.19	50.56 ± 16.15	<0.001
<40	6997 (39.02%)	5539 (45.85%)	1458 (24.92%)
≥40 to <60	5310 (29.62%)	3322 (27.50%)	1988 (33.98%)
≥60	5623 (31.36%)	3219 (26.65%)	2404 (41.09%)
WWI	10.86 ± 0.84	10.60 ± 0.77	11.44 ± 0.69	<0.001
Q1 (8.04–10.35)	4436 (24.74%)	4184 (34.64%)	252 (4.31%)
Q2 (10.36–10.95)	4500 (25.10%)	3421 (28.32%)	1079 (18.44%)
Q3 (10.96–11.55)	4465 (24.9%)	2646 (21.90%)	1819 (31.09%)
Q4 (11.56–15.52)	4529 (25.26%)	1829 (15.14%)	2700 (46.15%)
Gender				<0.001
Male	8684 (48.43%)	5377 (44.51%)	3307 (56.53%)
Female	9246 (51.57%)	6703 (55.49%)	2543 (43.47%)
Race/ethnicity				<0.001
Mexican American	3709 (20.69%)	1930 (15.98%)	1779 (30.41%)
Other Hispanic	1435 (8.00%)	900 (7.45%)	535 (9.15%)
Non-Hispanic White	7953 (44.36%)	5398 (44.69%)	2555 (43.68%)
Non-Hispanic Black	3505 (19.55%)	2868 (23.74%)	637 (10.89%)
Other race, including multiracial	1328 (7.41%)	984 (8.15%)	344 (5.88%)
Education				<0.001
Less than high school	5062 (28.23%)	2933 (24.28%)	2129 (36.39%)
High school graduate or equivalent	4061 (22.65%)	2644 (21.89%)	1417 (24.22%)
College or above	7443 (41.51%)	5331 (44.13%)	2112 (36.10%)
Not recorded	1364 (7.61%)	1172 (9.70%)	192 (3.28%)
Poverty-income ratio				<0.001
<1.3	4901 (27.33%)	3187 (26.38%)	1714 (29.30%)
≥1.3 to <3.5	6410 (35.75%)	4308 (35.66%)	2102 (35.93%)
≥3.5	5072 (28.29%)	3567 (29.53%)	1505 (25.73%)
Not recorded	1547 (8.63%)	1018 (8.43%)	529 (9.04%)
Alcohol consumption				<0.001
Never	3364 (18.76%)	2193 (18.15%)	1171 (20.02%)
Moderate	5280 (29.45%)	3578 (29.62%)	1702 (29.09%)
Heavy	3366 (18.77%)	2367 (19.59%)	999 (17.08%)
Binge	1913 (10.67%)	1220 (10.10%)	693 (11.85%)
Not recorded	4007 (22.35%)	2722 (22.53%)	1285 (21.97%)
Hypertension				<0.001
No	11,103 (61.92%)	8460 (70.03%)	2643 (45.18%)
Yes	6807 (37.96%)	3606 (29.85%)	3201 (54.72%)
Not recorded	20 (0.11%)	14 (0.12%)	6 (0.10%)
Diabetes				<0.001
No	15,378 (85.77%)	11,157 (92.36%)	4221 (72.15%)
Yes	2311 (12.89%)	793 (6.56%)	1518 (25.95%)
Not recorded	241 (1.34%)	130 (1.08%)	111 (1.90%)
Smoking status				<0.001
Never	9065 (50.56%)	6219 (51.48%)	2846 (48.65%)
Former	4300 (23.98%)	2495 (20.65%)	1805 (30.85%)
Current	3390 (18.91%)	2357 (19.51%)	1033 (17.66%)
Not recorded	1175 (6.55%)	1009 (8.35%)	166 (2.84%)
HBV infection				0.003
No	15,750 (87.84%)	10,639 (88.07%)	5111 (87.37%)
Yes	87 (0.49%)	65 (0.54%)	22 (0.38%)
Not recorded	2093 (11.67%)	1376 (11.39%)	717 (12.26%)
HCV infection				<0.001
No	17,644 (98.4%)	11,905 (98.55%)	5739 (98.10%)
Yes	185 (1.03%)	101 (0.84%)	84 (1.44%)
Not recorded	101 (0.56%)	74 (0.61%)	27 (0.46%)
Physical activity				<0.001
Low-level PA	4407 (24.58%)	2921 (24.18%)	1486 (25.40%)
High-level PA	11,317 (63.12%)	7886 (65.28%)	3431 (58.65%)
Not recorded	2206 (12.30%)	1273 (10.54%)	933 (15.95%)
BMI (kg/m²)	28.51 ± 6.59	26.10 ± 4.89	34.03 ± 6.66	<0.001
<25	5672 (31.63%)	5388 (44.6%)	284 (4.85%)
≥25 to <30	6045 (33.71%)	4353 (36.03%)	1692 (28.92%)
≥30	6213 (34.65%)	2339 (19.36%)	3874 (66.22%)
Other parameters
Weight (kg)	81.78 ± 21.04	74.31 ± 15.74	98.84 ± 21.65	<0.001
Height (cm)	169.09 ± 10.02	168.56 ± 9.85	170.28 ± 10.28	<0.001
WC (cm)	97.69 ± 16.41	90.97 ± 12.29	113.05 ± 14.15	<0.001
GHB (%)	5.53 ± 0.88	5.36 ± 0.64	5.93 ± 1.18	<0.001
FPG (mg/dL)	104.15 ± 29.30	97.75 ± 19.32	118.77 ± 40.76	<0.001
HOMA-IR	3.33 ± 4.77	1.88 ± 1.07	6.63 ± 7.51	<0.001
Insulin (μIU/mL)	12.13 ± 12.13	7.78 ± 4.06	22.07 ± 17.43	<0.001
Hs-CRP (mg/dL)	1.32 ± 4.15	0.96 ± 3.34	2.16 ± 5.49	<0.001
TC (mg/dL)	184.70 ± 57.28	181.03 ± 56.36	193.08 ± 58.47	<0.001
HDL (mg/dL)	53.69 ± 16.21	57.13 ± 16.54	45.82 ± 12.21	<0.001
TG (mg/dL)	135.56 ± 116.38	113.03 ± 76.88	187.04 ± 164.87	<0.001
LDL (mg/dL)	115.71 ± 45.45	112.02 ± 39.35	124.13 ± 56.09	<0.001
ALB (g/L)	42.50 ± 3.48	42.79 ± 3.49	41.84 ± 3.36	<0.001
ALT (U/L)	25.47 ± 26.84	21.69 ± 16.42	34.12 ± 40.54	<0.001
AST (U/L)	24.77 ± 18.08	23.35 ± 16.17	28.02 ± 21.47	<0.001
ALP (U/L)	70.19 ± 24.06	67.30 ± 21.69	76.81 ± 27.64	<0.001
GGT (U/L)	28.08 ± 38.65	20.95 ± 19.50	44.40 ± 60.47	<0.001
TBIL (μmol/L)	11.99 ± 5.36	12.13 ± 5.41	11.65 ± 5.24	<0.001
GLO (g/L)	29.35 ± 4.27	28.98 ± 4.20	30.21 ± 4.32	<0.001
PLT (×10⁹/L)	255.03 ± 66.03	254.20 ± 64.97	256.93 ± 68.36	0.011

MAFLD: metabolic dysfunction-associated fatty liver disease; WWI: weight-adjusted-waist index; HBV: hepatitis B virus; HCV: hepatitis C virus; WC: waist circumference; BMI: body mass index; GHB: glycosylated hemoglobin; FPG: fasting plasma glucose; HOMA-IR: homeostatic model assessment for insulin resistance; Hs-CRP: high-sensitivity C-reactive protein; ALT: alanine aminotransferase; AST: aspartate aminotransferase; ALP: alkaline phosphatase; ALB: albumin; GGT: gamma-glutamyl transferase; GLO: globulin; TBIL: total bilirubin; HDL: high-density lipoprotein; LDL: low-density lipoprotein; TC: total cholesterol; TG: triglyceride; PLT: platelets; PA: physical activity.

Note: Mean ± SD for continuous variables: p value was calculated by weighted linear regression model; % for categorical variables: p value was calculated by weighted chi-square test. Other races include Asian, American Indian or Alaska Native, Native Hawaiian or other Pacific Islander, and multiracial persons.

Association between WWI and MAFLD

The results of logistic regression analysis examining the association between WWI and MAFLD are detailed in Table 2. The analysis revealed a positive correlation between WWI and MAFLD prevalence in model 1 (odds ratio (OR) of 4.35; 95% confidence interval [95% CI: 4.08–4.63]. The WWI variable remained significantly associated with MAFLD (OR = 2.73, 95% CI: 2.48–3.01) even after adjusting for confounders, including age, sex, race/ethnicity, PIR, alcohol consumption, hypertension, diabetes, BMI, HBV infection, HCV infection, and PA (model 3). Additionally, in model 3, the risk of MAFLD was significantly elevated in Q2 (OR = 2.73, 95% CI: 2.21–3.36), Q3 (OR = 4.41, 95% CI: 3.54–5.48), and Q4 (OR = 7.67, 95% CI: 6.02–9.76). The association between WWI and MAFLD demonstrated a positive trend with increasing WWI (p for trend <0.001), with the highest risk observed in the Q4 group.

Table 2.

Associations between WWI and MAFLD across three models.

	Model 1 OR (95% CI)	Model 2 OR (95% CI)	Model 3 OR (95% CI)
WWI	4.35 (4.08, 4.63)	6.13 (5.65, 6.65)	2.73 (2.48, 3.01)
Quintiles of WWI
Q1 (8.04–10.35)	Reference	Reference	Reference
Q2 (10.36–10.95)	5.52 (4.61, 6.61)	5.57 (4.62, 6.73)	2.73 (2.21, 3.36)
Q3 (10.96–11.55)	12.92 (10.84, 15.41)	15.27 (12.60, 18.50)	4.41 (3.54, 5.48)
Q4 (11.56–15.52)	30.24 (25.31, 36.13)	49.74 (40.50, 61.09)	7.67 (6.02, 9.76)
p for trend	p < 0.001	p < 0.001	p < 0.001

MAFLD: metabolic dysfunction-associated fatty liver disease; WWI: weight-adjusted-waist index; OR: odds ratio; 95% CI: 95% confidence interval; BMI: body mass index.

Note: Model 1: nonadjusted model; model 2 adjusted for gender; age; and race; model 3 adjusted for gender; age; race; poverty-income ratio; BMI; diabetes; hypertension; alcohol consumption; physical activity; hepatitis B virus; and hepatitis C virus.

To elucidate the specific relationship between WWI and the prevalence of MAFLD, the RCS regression analysis was conducted (Figure 2). We identified an inverted L-shaped correlation between WWI and MAFLD prevalence, which was statistically significant (p < 0.05).

Figure 2.

Restricted cubic spline plot for analyzing the association between WWI and the incidence of MAFLD.

Subgroup analysis

To explore this relationship across diverse demographic subgroups, we performed stratified weighted multivariate regression analysis based on variables, including age, sex, and race. In the fully adjusted model, the subgroup analysis indicated that WWI maintained a positive association with the incidence of MAFLD in males (OR = 2.68, 95% CI: 2.29–3.13) and females (OR = 2.57, 95% CI: 2.27–2.90). However, no significant differences were detected between the subgroups (p for interaction = 0.98). Similarly, in other subgroup analyses, the results indicated that WWI remained positively correlated with MAFLD. However, there were no differences observed within the subgroups (Table 3).

Table 3.

Subgroup analysis for the association between WWI and MAFLD across three models.

	Model 1 OR (95% CI)	Model 2 OR (95% CI)	Model 3 OR (95% CI)	p for interaction
Gender				0.98
Male	7.00 (6.29–7.80)	8.56 (7.52–9.74)	2.68 (2.29–3.13)
Female	4.85 (4.41–5.33)	4.87 (4.39–5.40)	2.57 (2.27–2.90)
Age (years)				0.062
<40	5.03 (4.49–5.63)	8.07 (7.00–9.29)	2.83 (2.39–3.35)
≥40 to <60	5.64 (4.93–6.45)	7.71 (6.60–9.01)	3.64 (3.02–4.39)
≥60	3.13 (2.78–3.53)	3.62 (3.18–4.11)	2.08 (1.78–2.43)
Race/ethnicity				0.55
Mexican American	3.26 (2.86–3.71)	4.26 (3.63–5.01)	2.10 (1.73–2.54)
Other Hispanic	4.45 (3.54–5.60)	5.68 (4.26–7.57)	3.04 (2.16–4.29)
Non-Hispanic White	4.71 (4.32–5.14)	6.75 (6.04–7.55)	2.86 (2.49–3.28)
Non-Hispanic Black	3.68 (3.25–4.18)	4.96 (4.23–5.82)	2.63 (2.16–3.20)
Other race, including multiracial	3.42 (2.74–4.25)	5.69 (4.23–7.64)	2.69 (1.87–3.86)
BMI (kg/m²)				0.14
<25	3.53 (2.91–4.28)	3.30 (2.51–4.33)	2.76 (2.10–3.61)
≥25 to <30	2.63 (2.37–2.92)	3.14 (2.73–3.62)	2.91 (2.51–3.37)
≥30	2.83 (2.54–3.15)	3.39 (2.99–3.85)	3.20 (2.81–3.65)
Education				0.45
Less than high school	3.46 (3.08–3.88)	4.70 (4.06–5.44)	2.33 (1.95–2.77)
High school graduate or equivalent	4.1 (3.61–4.66)	5.84 (4.97–6.86)	2.47 (2.02–3.02)
College or above	4.91 (4.43–5.44)	7.10 (6.23–8.08)	3.17 (2.72–3.70)
Poverty-income ratio				0.48
<1.3	3.89 (3.47–4.36)	5.35 (4.63–6.18)	2.73 (2.28–3.26)
≥1.3 to <3.5	4.15 (3.74–4.6)	6.02 (5.25–6.91)	2.71 (2.32–3.17)
≥3.5	5.47 (4.81–6.23)	7.56 (6.44–8.87)	2.97 (2.43–3.63)
Alcohol consumption				0.69
Never	3.32 (2.92–3.78)	4.03 (3.46–4.69)	2.30 (1.91–2.77)
Moderate	4.86 (4.28–5.52)	6.99 (5.96–8.19)	2.80 (2.31–3.40)
Heavy	5.14 (4.38–6.02)	7.83 (6.35–9.65)	3.38 (2.63–4.35)
Binge	5.98 (4.83–7.4)	8.5 (6.54–11.06)	3.22 (2.30–4.51)
Hypertension				0.54
No	4.61 (4.21–5.04)	6.97 (6.21–7.82)	2.93 (2.56–3.36)
Yes	3.09 (2.81–3.4)	4.53 (4.02–5.10)	2.48 (2.15–2.85)
Diabetes				0.56
No	4.11 (3.83–4.41)	6.09 (5.56–6.67)	2.78 (2.49–3.10)
Yes	2.87 (2.4–3.43)	3.84 (3.10–4.75)	2.53 (1.99–3.23)
Smoking status				0.30
Never	3.97 (3.66–4.32)	5.56 (5.00–6.18)	2.63 (2.31–3.00)
Former	4.42 (3.85–5.06)	6.70 (5.66–7.94)	3.06 (2.51–3.73)
Current	4.88 (4.17–5.71)	7.17 (5.84–8.79)	2.70 (2.12–3.44)
Physical activity				0.46
Low-level PA	4.22 (3.71–4.79)	5.65 (4.81–6.64)	2.76 (2.26–3.37)
High-level PA	4.53 (4.17–4.92)	6.61 (5.94–7.35)	2.77 (2.44–3.15)

MAFLD: metabolic dysfunction-associated fatty liver disease; WWI: weight-adjusted-waist index; OR: odds ratio; 95% CI: 95% confidence interval; PA: physical activity; BMI: body mass index.

Note: Model 1: nonadjusted model; model 2 adjusted for gender; age; and race; model 3 adjusted for gender; age; race; poverty-income ratio; BMI; diabetes; hypertension; Alcohol consumption; physical activity; hepatitis B virus; and hepatitis C virus.

Discussion

This study evaluated the correlation between WWI and MAFLD among 17,930 participants. Our findings revealed a significant association between WWI and MAFLD that remained strong even after controlling for relevant variables. Notably, we observed an inverted L-shaped relationship between WWI and MAFLD. Subgroup analyses indicated a significant positive relationship between WWI and MAFLD; however, no differences were found within the subgroups.

To date, the association between WWI and MAFLD has rarely been investigated, although earlier research examining the relationship between WWI and NAFLD. Studies conducted by Yu et al.,²⁸ Hu et al.,¹³ Zhou et al.,²⁹ and Shen et al.¹² have demonstrated a positive correlation between WWI and NAFLD. However, our study specifically focuses on the MAFLD population, which represents a broader demographic than NAFLD. Our results indicate a positive correlation between WWI and MAFLD. Additionally, Hosseini et al.³⁰ conducted a study examining predictive indicators of MAFLD within a cohort of 7836 individuals in Iran, including 642 cases of MAFLD. Their findings revealed a positive association between WWI and MAFLD (OR = 1.747, 95% CI: 1.402–1.178). This positive correlation persisted after adjusting for some factors (OR = 1.818, 95% CI: 1.399–2.364). However, it is noteworthy that their study population was exclusively Iranian, differing from the demographic composition of our study. Our study using a larger cohort of 17,930 individuals in the United States indicated that WWI is associated with a higher risk of MAFLD, both in model 1 (OR = 4.35; 95% CI: 4.08–4.63) and in the fully adjusted model 3 (OR = 2.73, 95% CI: 2.48–3.01). Moreover, our research included additional subgroup analyses to enhance the robustness of our findings and clarified the nonlinear relationship between WWI and MAFLD using RCS. Furthermore, Tang et al.³¹ conducted a correlation analysis examining the relationship between WWI and MAFLD in a Chinese cohort of 288 individuals. However, their study primarily examined the correlation between WWI and the controlled attenuation parameter and liver stiffness value measured by a FibroScan device. Conversely, we employed a larger American population sample and utilized an RCS to further elucidate the relationship between WWI and MAFLD.

Metabolic syndrome is affected by a blend of lifestyle, environmental, and genetic factors. Research indicates that managing daily caloric intake is vital for controlling this condition.³² Consuming high-calorie foods can lead to the accumulation of subcutaneous and visceral fat, particularly in the abdominal region,³³ which is a recognized key contributor to the onset of metabolic syndrome.³⁴

While BMI is the primary measure used to assess overall obesity,³⁵ it does not account for critical factors such as visceral fat and its distribution,^36,37 making it an incomplete measure. Conversely, the novel WWI has been proposed as a potentially superior metric for evaluating obesity-related health risks.³⁸ Studies indicate that WWI is better at reflecting central obesity compared to overall obesity, with various reports demonstrating its strong association with metabolic syndrome.^6,34 Central obesity is a well-known risk factor for hepatic steatosis. In the early stages of obesity, it can trigger liver cell inflammation, elevate inflammatory cytokine levels, and disrupt hepatic metabolic function.¹⁹ Ongoing hepatic inflammation may result in cell death, which can trigger immune cell infiltration and accelerate the progression to steatohepatitis.^39,40 WWI may indicate central obesity, where values imply heightened hepatic inflammation and an increased risk of MAFLD progression.

Subgroup analysis consistently showed a positive correlation between WWI and MAFLD across all groups. However, no significant differences were found within these subgroups, indicating that the relationship between WWI and MAFLD is not influenced by factors such as age, gender, or ethnicity. Elevated WWI levels were linked to a notably higher risk of MAFLD across all study populations, irrespective of gender, age, or ethnicity.

The prevalence of MAFLD is increasing annually, impacting a significant portion of the global population. Currently, managing MAFLD presents considerable challenges, mainly due to the lack of specific pharmacological treatments.⁴¹ A pertinent question is whether early identification of high-risk groups and the development of targeted interventions can effectively reduce the prevalence of MAFLD. The monitoring of WWI may be effectively conducted by assessing height and weight during routine physical examinations or screening for related conditions such as diabetes and hypertension. This approach may help identify individuals at high risk of developing MAFLD across various populations. Subsequently, targeted interventions, including weight management, appropriate dietary modifications, and moderate PAs, can be developed and implemented to mitigate the risk of MAFLD.^42,43 The cost of measuring height and weight is minimal, making the monitoring of WWI a highly cost-effective method. Furthermore, early intervention aimed at reducing the risk of MAFLD can lessen the financial burden on individuals and mitigate potential health issues.

We conducted a large-scale data study to elucidate the connection between WWI and MAFLD. Furthermore, we performed RCS analysis for the first time, which revealed a nonlinear relationship between WWI and MAFLD. This study is particularly important for identifying high-risk groups for MAFLD within the population, which may help inform targeted interventions to reduce the prevalence of the disease. However, this study has certain limitations. First, as a cross-sectional analysis, it cannot establish causation. Consequently, we cannot infer a causal link between WWI and MAFLD based on our findings. Second, the results may not be applicable to populations outside the United States, as the research was conducted exclusively on US adults. Third, despite efforts to account for confounding factors, it is acknowledged that not all potential confounders—such as dietary patterns and genetic susceptibility—were included in the analysis, which may impact the results. Additionally, in this study, fatty liver was diagnosed using methods that do not align with the pathological gold standard. This may have led to nondifferential misclassification, biasing the observed associations toward the null hypothesis, meaning the true effect is likely stronger than we reported. Finally, despite the extensive data available in the NHANES database, significant missing data and potential bias from the Mobile Examination Centers may impact the research outcomes.

Conclusion

In summary, this research indicates a positive correlation between WWI and MAFLD in the American adult population. WWI is a low-cost, accessible screening tool for MAFLD risk. It may be possible to reduce the prevalence of MAFLD by developing targeted interventions for specific populations at risk. Additional studies are required to validate the link between WWI and MAFLD. This should include longitudinal validation studies, investigations involving populations outside of the United States, and explorations into the underlying mechanisms.

Supplemental Material

sj-docx-1-sci-10.1177_00368504261420942 - Supplemental material for The weight-adjusted-waist index is positively associated with metabolic dysfunction-associated fatty liver disease in US adults: A cross-sectional NHANES Study

Supplemental material, sj-docx-1-sci-10.1177_00368504261420942 for The weight-adjusted-waist index is positively associated with metabolic dysfunction-associated fatty liver disease in US adults: A cross-sectional NHANES Study by Pingping Liu, Wei Zhao, Daochong Qiu and Yuping Li in Science Progress

Footnotes

List of abbreviations

Acknowledgements

We express our gratitude to the NHANES program for their diligent efforts in data collection and dissemination. Additionally, we extend our appreciation to the editorial team at Home for Researchers () for their meticulous language polishing.

ORCID iD

Pingping Liu

Ethical approval and consent to participate

The NHANES protocol was approved by the NCHS Research Ethics Review Board, and informed consent was obtained from all participants.

Author contributions

Conceptualization: Pingping Liu and Wei Zhao; methodology: Pingping Liu; software: Wei Zhao; validation: Pingping Liu and Wei Zhao; data curation: Daochong Qiu and Yuping Li; writing—original draft preparation: Pingping Liu; writing—review and editing: Pingping Liu and Wei Zhao; supervision: Pingping Liu and Wei Zhao; and project administration: Pingping Liu. All authors have read and agreed to the published version of the manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Pingping Liu was supported by Supported by Ganzhou “Science and Technology and National Regional Medical Center” Joint Program (2022-YB1276).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The study conducted an analysis of publicly available datasets, which can be accessed at the following URL: .

Supplemental material

Supplemental material for this article is available online.

References

Eslam

Newsome

Sarin

, et al. A new definition for metabolic dysfunction-associated fatty liver disease: an international expert consensus statement. J Hepatol 2020; 73: 202–209.

Rinella

, et al. A multisociety Delphi consensus statement on new fatty liver disease nomenclature. Hepatology 2023; 78: 1966–1986.

Xie

Wang

, et al. Trends in prevalence and all-cause mortality of metabolic dysfunction-associated fatty liver disease among adults in the past three decades: results from the NHANES study. Eur J Intern Med 2023; 110: 62–70.

Loomba

Adams

. Advances in non-invasive assessment of hepatic fibrosis. Gut 2020; 69: 1343–1352.

Vuppalanchi

Siddiqui

Van Natta

, et al. Performance characteristics of vibration-controlled transient elastography for evaluation of nonalcoholic fatty liver disease. Hepatology 2018; 67: 134–144.

Sahakyan

Somers

Rodriguez-Escudero

, et al. Normal-weight central obesity: implications for total and cardiovascular mortality. Ann Intern Med 2015; 163: 827–835.

Younossi

Anstee

Marietti

, et al. Global burden of NAFLD and NASH: trends, predictions, risk factors and prevention. Nat Rev Gastroenterol Hepatol 2018; 15: 11–20.

Chi

Wang

, et al. Trends in insulin resistance: insights into mechanisms and therapeutic strategy. Signal Transduct Target Ther 2022; 7: 16.

Johansen

. Obesity and body composition for transplant wait-list candidacy—challenging or maintaining the BMI limits? J Ren Nutr 2013; 23: 207–209.

10.

Park

Kim

Kwon

, et al. A novel adiposity index as an integrated predictor of cardiometabolic disease morbidity and mortality. Sci Rep 2018; 8: 16753.

11.

Tutunchi

Naeini

Ebrahimi-Mameghani

, et al. Metabolically healthy and unhealthy obesity and the progression of liver fibrosis: a cross-sectional study. Clin Res Hepatol Gastroenterol 2021; 45: 101754.

12.

Shen

, et al. Association between weight-adjusted-waist index with hepatic steatosis and liver fibrosis: a nationally representative cross-sectional study from NHANES 2017 to 2020. Front Endocrinol (Lausanne) 2023; 14: 1159055.

13.

Han

Shen

, et al. Association of weight-adjusted-waist index with non-alcoholic fatty liver disease and liver fibrosis: a cross-sectional study based on NHANES. Eur J Med Res 2023; 28: 263.

14.

Xie

Tan

, et al. Association of cholecystectomy with liver fibrosis and cirrhosis among adults in the USA: a population-based propensity score-matched study. Front Med (Lausanne) 2021; 8: 787777.

15.

Fujii

Nakamura

Fukumoto

, et al. Lifestyle changes during the coronavirus disease 2019 pandemic impact metabolic dysfunction-associated fatty liver disease. Liver Int 2022; 42: 995–1004.

16.

von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007; 147: 573–577.

17.

Ruhl

Everhart

. Fatty liver indices in the multiethnic United States National Health and Nutrition Examination Survey. Aliment Pharmacol Ther 2015; 41: 65–76.

18.

2. Classification and diagnosis of diabetes: standards of medical care in diabetes-2020. Diabetes Care 2020; 43: S14–S31.

19.

Semmler

Balcar

Wernly

, et al. Insulin resistance and central obesity determine hepatic steatosis and explain cardiovascular risk in steatotic liver disease. Front Endocrinol (Lausanne) 2023; 14: 1244405.

20.

Fang

Wang

Coresh

, et al. Trends in diabetes treatment and control in U.S. adults, 1999–2018. N Engl J Med 2021; 384: 2219–2228.

21.

Huang

Zheng

, et al. The prognostic role of diet quality in patients with MAFLD and physical activity: data from NHANES. Nutr Diabetes 2024; 14: 4.

22.

Piercy

Troiano

Ballard

, et al. The physical activity guidelines for Americans. Jama 2018; 320: 2020–2028.

23.

Chu

Hong

Harasemiw

, et al. Chronic kidney disease, physical activity and cognitive function in older adults-results from the National Health and Nutrition Examination Survey (2011–2014). Nephrol Dial Transplant 2022; 37: 2180–2189.

24.

Pang

Huang

. The relationship between depressive symptoms and BMI: 2005–2018 NHANES data. J Affect Disord 2022; 313: 151–157.

25.

Testing for HCV infection: an update of guidance for clinicians and laboratorians. MMWR Morb Mortal Wkly Rep 2013; 62: 362–365.

26.

Coffin

Zhou

Terrault

. New and old biomarkers for diagnosis and management of chronic hepatitis B virus infection. Gastroenterology 2019; 156: 355–368.e3.

27.

Katsuki

Sumida

Gabazza

, et al. Homeostasis model assessment is a reliable indicator of insulin resistance during follow-up of patients with type 2 diabetes. Diabetes Care 2001; 24: 362–365.

28.

Kuang

, et al. Association between weight-adjusted waist index and non-alcoholic fatty liver disease: a population-based study. BMC Endocr Disord 2024; 24: 22.

29.

Zhou

Zhang

Xiao

. Association between weight-adjusted waist index and hepatic steatosis and fibrosis: analyses of the NHANES 2017 to 2020. Medicine (Baltimore) 2025; 104: e42708.

30.

Hosseini

Alipour

Sarvandian

, et al. Assessment of the appropriate cutoff points for anthropometric indices and their relationship with cardio-metabolic indices to predict the risk of metabolic associated fatty liver disease. BMC Endocr Disord 2024; 24: 79.

31.

Tang

Fan

, et al. Association between weight-adjusted waist circumference index and metabolic disease-associated fatty liver disease: a retrospective study. Risk Manag Healthc Policy 2025; 18: 2387–2399.

32.

Jakubowicz

Barnea

Wainstein

, et al. High caloric intake at breakfast vs. dinner differentially influences weight loss of overweight and obese women. Obesity (Silver Spring) 2013; 21: 2504–2512.

33.

Krogh-Madsen

Pedersen

Solomon

, et al. Normal physical activity obliterates the deleterious effects of a high-caloric intake. J Appl Physiol (1985) 2014; 116: 231–239.

34.

Fahed

Aoun

Bou Zerdan

, et al. Metabolic syndrome: updates on pathophysiology and management in 2021. Int J Mol Sci 2022; 23: 786.

35.

Crispo

Augustin

Luongo

, et al. Central obesity, body mass index, metabolic syndrome and mortality in Mediterranean breast cancer patients. Sci Rep 2023; 13: 21208.

36.

Flegal

Kit

Orpana

, et al. Association of all-cause mortality with overweight and obesity using standard body mass index categories: a systematic review and meta-analysis. JAMA 2013; 309: 71–82.

37.

Ogden

Carroll

Kit

, et al. Prevalence of childhood and adult obesity in the United States, 2011–2012. Jama 2014; 311: 806–814.

38.

Staiano

Katzmarzyk

. Visceral, subcutaneous, and total fat mass accumulation in a prospective cohort of adolescents. Am J Clin Nutr 2022; 116: 780–785.

39.

Bijnen

Josefs

Cuijpers

, et al. Adipose tissue macrophages induce hepatic neutrophil recruitment and macrophage accumulation in mice. Gut 2018; 67: 1317–1327.

40.

Gregor

Hotamisligil

. Inflammatory mechanisms in obesity. Annu Rev Immunol 2011; 29: 415–445.

41.

Eslam

Sanyal

George

. MAFLD: a consensus-driven proposed nomenclature for metabolic associated fatty liver disease. Gastroenterology 2020; 158: 1999–2014.e1.

42.

Tian

Zhang

Xie

, et al. Dietary quality and relationships with metabolic dysfunction-associated fatty liver disease (MAFLD) among United States adults, results from NHANES 2017–2018. Nutrients 2022; 14: 4505.

43.

Keating

Sabag

Hallsworth

, et al. Exercise in the management of metabolic-associated fatty liver disease (MAFLD) in adults: a position statement from exercise and sport science Australia. Sports Med 2023; 53: 2347–2371.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB