Abstract
Background and Aims:
Many studies have shown a link between physical activity (PA) and nonalcoholic fatty liver disease (NAFLD). However, more research is needed to investigate the relationship between different types of PA and NAFLD. This study aimed to explore the potential link between different types of PA, hepatic steatosis, and liver fibrosis.
Study:
A cross-sectional study was conducted using the data set from the National Health and Nutrition Examination Survey (NHANES) from 2017 to 2020. A multiple linear regression model was used to examine the linear relationship between different types of PA, the controlled attenuation parameter (CAP), and liver stiffness measurement (LSM). In addition, smoothing curve fitting and threshold effect analysis were used to depict their nonlinear relationship.
Results:
This study involved 5933 adults. Multiple linear regression analysis revealed a significantly negative correlation between leisure-time PA and CAP, while the relationship between occupation-related PA, transportation-related PA, and CAP was not significant. Subgroup analysis further revealed that leisure-time PA was significantly negatively correlated with CAP in women and younger age groups (under 60 y old), while the relationship was not significant in men and older age groups. In addition, there was a significant negative correlation between leisure-time PA and liver fibrosis in men.
Conclusions:
Leisure-time PA can prevent hepatic steatosis, and women and young people benefit more. Occupation-related PA is not associated with hepatic steatosis and cannot replace leisure-time PA. In men, increasing leisure-time PA is more effective in preventing liver fibrosis.
Key Words: NAFLD, physical activity, NHANES, hepatic steatosis, liver fibrosis
Nonalcoholic fatty liver disease (NAFLD) is a widespread liver disorder characterized by a fatty liver, with a liver cell fat accumulation of more than 5%, in the absence of a history of alcohol consumption or other external factors. As obesity and metabolic syndrome become more prevalent in the world, the occurrence of NAFLD is also on the rise.1–3 Nonalcoholic steatohepatitis (NASH), the active form of NAFLD, is characterized by histologic lobular inflammation and hepatocyte ballooning, and is associated with faster fibrosis progression.4 Currently, there is no approved drug treatment for NAFLD,5 Therefore, lifestyle modifications, including physical activity (PA), dietary changes, and weight loss, are the cornerstone of optimizing the management of patients with NAFLD.6
Recently, NAFLD was renamed metabolic dysfunction-associated steatotic liver disease (MASLD) at the 2023 European Association for the Study of the Liver (EASL) conference. The new definition requires MASLD to have at least one cardiac metabolic risk factor in addition to hepatic steatosis.7 There is mounting evidence that PA has beneficial impacts on all the components of metabolic syndrome and the resulting cardiovascular risk.8 Furthermore, various studies have reported that PA affects liver metabolism and is negatively correlated with the incidence of NAFLD and NASH.9 PA can significantly improve the accumulation of liver and visceral fat, increase lipid oxidation, and enhance insulin sensitivity.10 The American College of Sports Medicine (ACSM) acknowledges the association between regular PA and a reduced risk of NAFLD, and advises all NAFLD patients to engage in at least 150 minutes of moderate-intensity PA or 75 minutes of high-intensity PA per week.11 However, most studies demonstrating the beneficial effects of PA have focused on leisure-time PA, and the relationship between other types of PA and NAFLD remains unclear.12 For a significant portion of adults, work is the primary setting for PA. Workers in numerous occupations, such as construction, cleaning, waste collection, elderly care, agriculture, and manufacturing, engage in PA for the majority of their workday.13 Indeed, recent studies have shown that occupation-related PA may not improve health outcomes.14 Actually, occupation-related PA can be detrimental. These contrasting health effects of leisure-time PA and occupation-related PA constitute the so-called PA health paradox.15 Furthermore, recent studies have revealed that NAFLD is a sexually dimorphic disease with significant gender disparities that may be linked to the secretion of sex hormones.16 The association between different types of PA and NAFLD, along with gender differences, warrants further exploration.
Instantaneous elastography is widely used in the screening of NAFLD due to its advantageous noninvasive, accurate, and repeatable properties.17 Controlled attenuation parameters (CAP) and liver stiffness measurements (LSM) are utilized to evaluate liver steatosis and fibrosis, respectively.18
Therefore, we used a massive data set from the National Health and Nutrition Examination Survey (NHANES) to explore the connections between different types of PA and both CAP and LSM and conducted subgroup analyses based on gender and age.
METHODS
Study Population
All data and guidance on analytical approaches are publicly and freely available from the US Centers for Disease Control and Prevention’s National Center for Health Statistics and can be accessed through the following link: https://www.cdc.gov/nchs/nhanes/. This cross-sectional analysis utilized data from 4 years of the 2017 to 2020 NHANES cycle. A total of 15,560 participants were recruited for the study. However, we excluded 6539 participants who lacked data for either CAP or LSM, 1361 participants with missing data on various types of PA, 979 participants who reported significant alcohol consumption (4 or more drinks per day), 196 participants who were diagnosed with hepatitis B or C during laboratory testing or questionnaire surveys, and 552 participants who did not meet any cardiovascular metabolic criteria. The study involved a total of 5933 participants, as shown in Figure 1. Participants provided written consent before enrollment. NHANES procedures were approved by the National Center for Health Statistics Review Board.
FIGURE 1.
Flowchart of participant selection. CAP indicates controlled attenuation parameter; NHANES, National Health and Nutrition Examination Survey; LSM, liver stiffness measurement.
Study Variables
The NHANES survey used the Global PA Questionnaire, which gauged the duration of typical PA during the past week. The questionnaire queried the time spent on PA in various domains and intensities, ranging from vigorous and moderate activity at work to traffic activity and leisure time.
Vigorous-intensity activity refers to strenuous activities that cause a significant increase in breathing or heart rate, while moderate-intensity activity refers to moderate-intensity activities that cause a small increase in breathing or heart rate.
In our study, PA was divided into 3 types: leisure-time PA, occupation-related PA, and transportation-related PA. Among them, leisure-time PA includes vigorous-intensity recreational physical activities such as running, basketball, and football, as well as moderate-intensity recreational activities such as brisk walking, swimming, and table tennis. Occupation-related PA includes vigorous-intensity physical activities such as continuous lifting or carrying heavy objects, excavation, or construction work, as well as moderate-intensity activities such as fast walking or light weight lifting during work. Transportation-related PA refers to physical activity such as walking or cycling on the way to school or work.
The 2018 PA Guidelines for Americans (PA Guidelines) indicate that absolute intensity is the amount of energy expended during the activity without considering a person’s cardiorespiratory fitness or aerobic capacity. Absolute intensity is expressed in metabolic equivalent of task (MET) units; 1 MET is equivalent to the resting metabolic rate or the energy expenditure while awake and sitting quietly. Moderate-intensity activities have a MET value of 3 to 5.9 METs; vigorous-intensity activities have a MET value of 6 or greater.19 In the NHANES PA questionnaire survey, the recommended MET value for moderate-intensity activities is 4 METs, while the MET value for vigorous-intensity activities is 8 METs. We calculated the MET minutes for all PA questions and defined total PA as the total MET hours per week. Using the same methodology, we also calculated leisure-time PA, occupation-related PA, and transportation-related PA.
In our analysis, we have designed different types of PA as exposure variables, while CAP and LSM are used as outcome variables to assess hepatic steatosis and liver fibrosis, respectively.
The NHANES staff, equipped with the FibroScan 502 V2 Touch model, evaluated participants for vibration-controlled transient elastography (VCTE). We defined MASLD as CAP scores of 263 dB/m or more (S1, the cutoff of sensitivity fixed at 90%) and the Jordan index was used for optimization.20
The covariates included gender, age, race, education level, smoking status, body mass index (BMI), waist-to-hip ratio (WHR), alanine transaminase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), gamma glutamyl transferase (GGT), total cholesterol, triglycerides, direct high-density lipoprotein cholesterol (HDL-C), uric acid (UA), C-reactive protein (CRP), and glycated hemoglobin (HbA1c).
Statistical Analysis
We performed all statistical analyses using R (https://www.r-project.org/) and EmpowerStats (https://www.empowerstats.net/cn/), with statistical significance set at P<0.05. All estimates were calculated using sample weights according to the analytical guidelines edited by NCHS, as the goal of NHANES is to produce data that is representative of the civilian, noninstitutionalized US population. By employing these weights, the analysis takes into account the complex survey design and nonresponse of NHANES, ensuring that the resulting data accurately reflect the target population. We constructed three multivariable linear regression models—model 1: with no covariates adjusted; model 2: adjusted for age, gender, and race; and model 3: adjusted for all the covariates presented in Table 1, including factors such as educational level, smoking, body mass index, waist-to-hip ratio, laboratory data, and other potential confounding factors. Subgroup analyses were further performed to explore potential interactions. A weighted generalized additive model and smooth curve fitting were used to address potential nonlinear relationships.
TABLE 1.
Weighted Features of the Study Population Based on Controlled Attenuation Parameters (CAP)
| Study population | Non-MASLD (CAP <263, n=2835) | MASLD (CAP ≥263, n=3098) | P |
|---|---|---|---|
| Age (y) | 46.3±17.9 | 49.8±16.6 | <0.0001 |
| Gender (%) | <0.0001 | ||
| Man | 40.5 | 50.9 | |
| Woman | 59.5 | 49.1 | |
| Race/ethnicity (%) | <0.0001 | ||
| Mexican American | 6.7 | 11.3 | |
| Other Hispanic | 8.0 | 7.5 | |
| Non-Hispanic White | 61.3 | 61.8 | |
| Non-Hispanic Black | 13.7 | 9.2 | |
| Other races | 10.4 | 10.3 | |
| Education level (%) | 0.0006 | ||
| Less than high school | 9.4 | 10.9 | |
| High school | 23.7 | 27.0 | |
| More than high school | 66.9 | 62.1 | |
| Smoked at least 100 cigarettes | 0.0020 | ||
| Yes | 34.5 | 38.3 | |
| No | 65.5 | 61.7 | |
| Total PA (MET hours/week) | 79.2±119.3 | 70.0±110.4 | 0.0019 |
| Leisure-time PA (MET hours/week) | 19.1±30.8 | 13.0±23.7 | <0.0001 |
| Occupation-related PA (MET hours/week) | 56.7±110.6 | 53.4±102.0 | 0.2277 |
| Transportation-related PA (MET hours/week) | 3.4±12.8 | 3.6±16.4 | 0.7485 |
| LSM (kPa) | 4.8±2.8 | 6.4±5.4 | <0.0001 |
| BMI (kg/m2) | 27.1±5.2 | 33.3±7.0 | <0.0001 |
| WHR | 0.9±0.1 | 1.0±0.1 | <0.0001 |
| Laboratory features | |||
| ALT (IU/L) | 19.1±12.3 | 26.4±17.9 | <0.0001 |
| AST (IU/L) | 20.2±8.6 | 22.6±12.3 | <0.0001 |
| ALP (IU/L) | 73.0±23.3 | 78.5±22.7 | <0.0001 |
| GGT (IU/L) | 22.7±26.3 | 33.1±37.0 | <0.0001 |
| Total cholesterol (mmol/L) | 4.8±1.0 | 4.9±1.0 | 0.0017 |
| Triglyceride (mmol/L) | 1.3±0.8 | 1.9±1.3 | <0.0001 |
| HDL-cholesterol (mmol/L) | 1.5±0.4 | 1.3±0.4 | <0.0001 |
| UA (µmol/L) | 296.6±78.4 | 337.0±84.9 | <0.0001 |
| CRP (mg/L) | 3.1±6.9 | 4.4±6.7 | <0.0001 |
| Glycohemoglobin (%) | 5.5±0.7 | 5.9±1.1 | <0.0001 |
Mean±SD for continuous variables: P-value was calculated by weighted linear regression model.
% for categorical variables: P-value was calculated by weighted χ2 test.
ALP indicates alkaline phosphatase; ALT, alanine transaminase; AST, aspartate aminotransferase; BMI, body mass index; CAP, controlled attenuation parameter; CRP, C-reactive protein; GGT, gamma glutamyl transferase; HDL-cholesterol, high-density lipoprotein cholesterol; LSM, liver stiffness measure; MET, metabolic equivalent; UA, uric acid; WHR, waist-to-hip ratio.
RESULTS
Baseline Characteristics
In this research, a total of 5933 participants were recruited according to the established inclusion and exclusion criteria. The average age of the participants was 49.9±18.0 years, with a slightly higher proportion of women at 54.4% compared with men at 45.6%. The mean (SD) concentrations of CAP, LSM, total PA, leisure-time PA, occupation-related PA, and transportation-related PA were 267.2 (61.0) dB/m, 5.7 (4.5) kPa, 71.4 (116.2) MET hours/week, 14.4 (27.6) MET hours/week, 52.7 (106.7) MET hours/week, and 4.2 (16.5) MET hours/week, respectively.
Table 1 lists all clinical characteristics of participants with CAP as the column-wise stratification variable. Compared with the non-MASLD group, the MASLD group is more likely to be male and elderly, with a higher proportion of Mexican Americans and high school education. They also have a higher smoking status and higher levels of LSM, BMI, WHR, ALT, AST, ALP, GGT, total cholesterol, triglycerides, UA, CRP, and glycated hemoglobin, but lower levels of total PA, leisure-time PA, and direct HDL-cholesterol. However, there is no significant difference between occupation-related PA and transportation-related PA.
Association Between Different Types of Physical Activity (PA) and Controlled Attenuation Parameter (CAP)
Table 2 presents the results of multiple regression analysis, which investigated the relationship between different types of physical activity (PA) and college academic performance (CAP). On the basis of these 3 models, there is a strong negative correlation between leisure-time PA and CAP. In models 1 and 2, there is a significant negative correlation between total PA and CAP, but this correlation becomes nonsignificant in the fully adjusted model 3. Surprisingly, the correlation between occupation-related PA, transportation-related PA, and CAP is not significant in any of the 3 models.
TABLE 2.
The Correlation Between Different Types of PA and CAP
| Variables | Model 1 β (95% CI) P |
Model 2 β (95% CI) P |
Model 3 β (95% CI) P |
|---|---|---|---|
| Total PA | −0.026 (−0.039, −0.012) 0.00019 | −0.030 (−0.043, −0.016) 0.00002 | −0.011 (−0.021, 0.000) 0.05167 |
| Leisure-time PA | −0.248 (−0.305, −0.192) <0.00001 | −0.261 (−0.318, −0.205) <0.00001 | −0.063 (−0.107, −0.018) 0.00549 |
| Occupation-related PA | −0.013 (−0.028, 0.002) 0.08392 | −0.015 (−0.030, −0.000) 0.04337 | −0.008 (−0.019, 0.004) 0.18324 |
| Transportation-related PA | −0.026 (−0.131, 0.079) 0.63232 | −0.049 (−0.152, 0.055) 0.35689 | −0.009 (−0.088, 0.071) 0.83186 |
Result variable: CAP, controlled attenuation parameter.
Exposure variable: different types of physical activity.
Model 1: no covariates were adjusted. Model 2: age, gender, and race were adjusted. Model 3: age, gender, race, educational level, smoking status, BMI, ALT, AST, ALP, GGT, total cholesterol, triglyceride, HDL-C, UA, CRP, and glycohemoglobin were adjusted.
Table 3 presents the results of multiple regression subgroup analysis stratified by gender and age, examining the relationship between leisure-time PA and CAP.
TABLE 3.
The Correlation Between Leisure-Time PA and CAP Stratified by Gender and Age.
| Leisure-time PA | Model 1 β (95% CI) P |
Model 2 β (95% CI) P |
Model 3 β (95% CI) P |
|---|---|---|---|
| Man | −0.227 (−0.301, −0.153) <0.00001 | −0.197 (−0.272, −0.122) <0.00001 | −0.038 (−0.097, 0.021) 0.20739 |
| Woman | −0.395 (−0.481, −0.310) <0.00001 | −0.342 (−0.428, −0.257) <0.00001 | −0.096 (−0.164, −0.028) 0.00543 |
| Age <60 | −0.250 (−0.316, −0.184) <0.00001 | −0.311 (−0.376, −0.246) <0.00001 | −0.093 (−0.143, −0.043) 0.00025 |
| Age ≥60 | −0.212 (−0.330, −0.095) 0.00042 | −0.237 (−0.354, −0.119) 0.00008 | −0.002 (−0.102, 0.098) 0.97112 |
Result variable: CAP, controlled attenuation parameter.
Exposure variable: leisure-time physical activity.
Model 1: no covariates were adjusted. Model 2: age, gender, and race were adjusted. Model 3: age, gender, race, educational level, smoking status, BMI, ALT, AST, ALP, GGT, total cholesterol, triglyceride, HDL-C, UA, CRP, and glycohemoglobin were adjusted.
In the subgroup analysis stratified by gender and age, the model is not adjusted for gender and age, respectively.
In all gender and age subgroups, there was a significant negative correlation between leisure-time PA and CAP, both in models 1 and 2. However, in the fully adjusted model 3, only the female subgroup and those aged under 60 showed a significant negative correlation between leisure-time PA and CAP, while there was no significant relationship between the male subgroup and those aged over 60.
We use smooth curve fitting to describe the nonlinear relationship between different types of PA and CAP (Figs. 2–4).
FIGURE 2.
The association between total PA and CAP. (A) Each black point represents a sample. (B) The solid red line represents the smooth curve fit between variables. Blue bands represent the 95% CI from the fit. CAP indicates controlled attenuation parameter; MET, metabolic equivalent of task; PA, physical activity.
FIGURE 4.
The association between occupation-related PA and CAP. (A) Each black point represents a sample. (B) The solid red line represents the smooth curve fit between variables. Blue bands represent the 95% CI from the fit. CAP indicates controlled attenuation parameter; MET, metabolic equivalent of task; PA, physical activity.
FIGURE 3.
The association between leisure-time PA and CAP. (A) Each black point represents a sample. (B) The solid red line represents the smooth curve fit between variables. Blue bands represent the 95% CI from the fit. (C) The association between leisure-time PA and CAP stratified by gender. (D) The association between leisure-time PA and CAP stratified by age group. CAP indicates controlled attenuation parameter; MET, metabolic equivalent of task; PA, physical activity.
Using a 2-stage linear regression model, we discovered a threshold saturation effect between leisure-time PA and CAP, with an inflection point at 24 MET hours per week. Upon stratifying the analysis by gender and age, we also identified a threshold saturation effect between leisure-time PA and CAP in the female subgroup and the subgroup aged <60 years, with inflection points of 26 and 24 MET hours per week, respectively. However, in the male subgroup and the subgroup aged over 60 years, there was only a negative linear relationship without a significant threshold saturation effect (Table 4).
TABLE 4.
Threshold Effect Analysis of Leisure-Time PA on CAP Using Two-Stage Linear Regression Model
| CAP (dB/m) | Adjusted β (95% CI) P |
|---|---|
| Leisure-time PA(MET hours/week) | |
| Inflection point | 24 |
| ≤24 | −0.301 (−0.439, −0.162) <0.0001 |
| >24 | 0.016 (−0.046, 0.079) 0.6068 |
| Log likelihood ratio test | <0.001 |
| Woman | |
| Inflection point | 26 |
| ≤26 | −0.196 (−0.380, −0.012) 0.0367 |
| >26 | −0.055 (−0.153, 0.043) 0.2734 |
| Log likelihood ratio test | 0.250 |
| Age <60 | |
| Inflection point | 24 |
| ≤24 | −0.432 (−0.594, −0.271) <0.0001 |
| >24 | 0.014 (−0.055, 0.084) 0.6888 |
| Log likelihood ratio test | <0.001 |
Race, educational level, smoking status, BMI, WHR, ALT, AST, ALP, GGT, total cholesterol, triglyceride, HDL-C, UA, CRP, and glycohemoglobin were adjusted.
CAP indicates controlled attenuation parameter; MET, metabolic equivalent of task; PA, physical activity.
Association Between Different Types of Physical Activity (PA) and Liver Stiffness Measurement (LSM)
Table 5 presents the results of multiple regression analysis of different types of PA and LSM. In models 1 and 2, there is a significant negative correlation between leisure-time PA and LSM, but this correlation becomes nonsignificant in the fully adjusted model 3. Furthermore, the correlations between total PA, occupation-related PA, and transportation-related PA with LSM are not significant in any of the models.
TABLE 5.
The Correlation Between Different Types of PA and LSM
| Variables | Model 1 β (95% CI) P |
Model 2 β (95% CI) P |
Model 3 β (95% CI) P |
|---|---|---|---|
| Total PA | −0.001 (−0.002, 0.000) 0.20145 | −0.001 (−0.002, 0.000) 0.05319 | −0.001 (−0.002, 0.000) 0.28150 |
| Leisure-time PA | −0.009 (−0.013, −0.005) 0.00001 | −0.011 (−0.015, −0.006) <0.00001 | −0.004 (−0.008, 0.000) 0.06947 |
| Occupation-related PA | −0.000 (−0.001, 0.001) 0.69497 | −0.000 (−0.002, 0.001) 0.37078 | −0.000 (−0.001, 0.001) 0.39264 |
| Transportation-related PA | 0.004 (−0.004, 0.012) 0.30872 | 0.003 (−0.005, 0.011) 0.44081 | 0.004 (−0.003, 0.012) 0.22965 |
Result variable: LSM, liver stiffness measure.
Exposure variable: different types of physical activity.
Model 1: no covariates were adjusted. Model 2: age, gender, and race were adjusted. Model 3: age, gender, race, educational level, smoking status, BMI, WHR, ALT, AST, ALP, GGT, total cholesterol, triglyceride, HDL-C, UA, CRP, and glycohemoglobin were adjusted.
In the subgroup analysis stratified by gender and age, the model is not adjusted for gender and age, respectively.
Table 6 presents the results of multiple regression subgroup analysis stratified by gender and age, examining the relationship between leisure-time PA and LSM.
TABLE 6.
The Correlation Between Leisure-Time of PA and LSM Stratified by Gender and Age
| Leisure-time PA | Model 1 β (95% CI) P |
Model 2 β (95% CI) P |
Model 3 β (95% CI) P |
|---|---|---|---|
| Man | −0.012 (−0.018, −0.006) 0.00013 | −0.012 (−0.018, −0.005) 0.00043 | −0.006 (−0.013, −0.000) 0.04170 |
| Woman | −0.010 (−0.015, −0.004) 0.00032 | −0.009 (−0.014, −0.004) 0.00118 | 0.000 (−0.005, 0.005) 0.99436 |
| Age <60 | −0.008 (−0.012, −0.003) 0.00143 | −0.007 (−0.012, −0.003) 0.00225 | −0.003 (−0.008, 0.001) 0.14581 |
| Age ≥60 | −0.013 (−0.022, −0.004) 0.00433 | −0.013 (−0.022, −0.004) 0.00436 | −0.005 (−0.014, 0.004) 0.23777 |
Result variable: LSM, liver stiffness measure.
Exposure variable: leisure-time physical activity.
Model 1: no covariates were adjusted. Model 2: age, gender, and race were adjusted. Model 3: age, gender, race, educational level, smoking status, BMI, ALT, AST, ALP, GGT, total cholesterol, triglyceride, HDL-C, UA, CRP, and glycohemoglobin were adjusted.
In the subgroup analysis stratified by gender and age, the model is not adjusted for gender and age, respectively.
In the male subgroup, all 3 models show a significant negative correlation between leisure-time PA and LSM. However, in the female subgroup and all age subgroups, the significant negative correlation is only observed in model 1 and model 2, but not in the fully adjusted model 3.
We use smooth curve fitting to describe the nonlinear relationship between leisure-time PA and LSM (Fig. 5).
FIGURE 5.
The association between leisure-time PA and LSM. (A) Each black point represents a sample. (B) The solid red line represents the smooth curve fit between variables. Blue bands represent the 95% CI from the fit. (C) The association between leisure-time PA and LSM stratified by gender. (D) The association between leisure-time PA and LSM stratified by age group. LSM indicates liver stiffness measurement; MET, metabolic equivalent of task; PA, physical activity.
DISCUSSION
In this population-based study utilizing a nationally representative sample of American adults, we found a significant negative association between leisure-time PA and hepatic steatosis. However, the relationships between occupation-related PA, transportation-related PA, and CAP were not significant. Furthermore, there was a threshold saturation effect between leisure-time PA and CAP, with a turning point at 24 metabolic equivalent (MET) hours/week, which is equivalent to 6 hours of moderate-intensity activity or 3 hours of vigorous-intensity activity per week. This finding aligns with others that have demonstrated the beneficial effects of physical activity during leisure time on NAFLD.12,21,22 Even without dietary intervention, physical activity has a positive impact on reducing liver fat content.23 However, occupation-related physical activity does not appear to have a beneficial effect on NAFLD.24–26 A cross-sectional study has revealed a significant negative association between leisure-time PA and transportation-related PA with NAFLD, with women being more affected than men. Conversely, the relationship between occupation-related PA and NAFLD was not statistically significant.24 The reason for this may be that improving cardiovascular health requires vigorous-intensity physical activity (>60% to 80% maximum oxygen consumption) over a short period of time. However, occupation-related PA that is of low intensity or prolonged duration may not maintain or improve cardiovascular and pulmonary health.27 The prolonged lack of recovery from PA (such as long work weeks or extreme endurance training) can lead to fatigue and exhaustion in athletes, and may even increase the risk of cerebrovascular disease.28 Inflammation has been implicated in the development and progression of NAFLD, as confirmed by numerous epidemiological studies.29,30 Inflammatory markers, such as C-reactive protein, rise during physical activity, and remain heightened until the body has recovered.31 Prolonged and continuous physical activity without sufficient recovery time can result in sustained inflammation. This study demonstrates that the impact of different types of physical activity on NAFLD varies. Leisure-time PA is beneficial for NAFLD, whereas occupation-related PA is not. Occupation-related PA cannot serve as a substitute for leisure-time PA as an indicator of increased daily physical activity for promoting health.
In further subgroup analysis, it was revealed that leisure-time PA was significantly negatively correlated with CAP in female and younger age groups, specifically those younger than 60 years. This difference may be related to sex hormones. There is significant evidence to suggest the importance of sex hormones in the occurrence of NAFLD.32 Sex hormones play a crucial role in regulating the metabolism of glucose and lipids in the liver through various mechanisms. It is well established that the liver exhibits functional gender differences. The increased risk of NAFLD in postmenopausal women can be attributed to the decline in estrogen levels, which alters the distribution of visceral fat and promotes a dyslipidemic milieu.33 Rodent and clinical studies have demonstrated that estrogen, acting through ER-α signaling, exerts a protective effect against the development of liver steatosis in both males and females. The deletion of ER-α in female mice results in an increase in liver steatosis and insulin resistance, likely due to enhanced hepatic fat generation.34 Our research indicates that women and young people are more likely to reap the benefits of leisure-time PA, which can lower the risk of NAFLD. However, the precise mechanisms involved require further exploration.
Numerous studies have demonstrated that physical activity (PA) can significantly reduce the incidence of liver fibrosis.9,12,35 One study showed that a dosage exceeding the 2018 American PA Guideline recommendations by more than double can reduce the risk of significant fibrosis or cirrhosis, making it a powerful weapon against these conditions.12 In our research, there was a significant negative correlation between leisure-time PA and liver fibrosis in men, but not in women. This suggests that increasing leisure-time PA may be more effective in preventing the progression of liver fibrosis in men. Further research is needed to explore the specific mechanisms involved.
Our study also has some limitations. First and foremost, it is a cross-sectional analysis, which means that we cannot determine the temporal relationship between variables. Furthermore, although we have adjusted for several relevant confounding factors, there may be other unmeasured factors that could have influenced the results. Therefore, the findings of our study should be interpreted with caution. Third, due to the limitations of the NHANES database, the PA in this study was surveyed using the 2018 US PA guidelines, but there is still a lot of subjectivity, so our study results may not fully reflect the true situation. Fourth, the degree of hepatic steatosis and liver fibrosis in this study was determined using transient elastography, which, while widely used and recognized, lacks relevant imaging and pathologic diagnoses.
Furthermore, ultrasound-based modalities have only moderate diagnostic accuracy for liver fat content and are suitable for screening. Among noninvasive imaging modalities, MRI-derived proton density fat fraction (MRI-PDFF) has the highest diagnostic accuracy and can be used for trial registration and evaluation of the therapeutic effect of early clinical trials for nonalcoholic steatohepatitis (NASH).36 Its accuracy has been confirmed by most studies.37,38 Of course, this technology also has its drawbacks, such as high cost and limited availability, so there is no data related to MRI-PDFF in the NHANES, which is also a limitation of this study. Due to the high accuracy of MRI-PDFF in the diagnosis of NAFLD, it is recommended to conduct further research in this area in the future if conditions permit.
CONCLUSION
Leisure-time PA is significantly negatively correlated with hepatic steatosis, and there is a threshold saturation effect, with the breaking point being 6 hours of moderate-intensity activity or 3 hours of vigorous-intensity activity per week. This relationship is more significant for women and younger members. Occupation-related PA is not associated with hepatic steatosis and cannot replace leisure-time PA. In men, increasing leisure-time PA is more effective in preventing liver fibrosis.
ACKNOWLEDGMENTS
The author thanks the members of the National Health Statistics Center and the Centers for Disease Control and Prevention for collecting data and making it available to the public. The author is also grateful to the participants who took part in the surveys for their valuable contributions.
Footnotes
Data availability statement: Publicly available data sets were analyzed in this study. This data can be found here: https://www.cdc.gov/nchs/nhanes/.
This study was reviewed and approved by the NCHS Ethics Review Board. The patients/participants provided their written informed consent to participate in this study.
B.S., Y.K., J.Z., are joint first authors.
The authors declare that they have nothing to disclose.
Contributor Information
Bo Sun, Email: sun403942916@163.com.
Ying Kang, Email: kangyingwangwen@163.com.
Junming Zhou, Email: junmingnn@hotmail.com.
Ying Feng, Email: 21818332@zju.edu.cn.
Wutao Wang, Email: wangwutaosn@163.com.
Xiaowei Wu, Email: fmed@sina.com.
Xiaohua Zhang, Email: jszhxh@sina.com.
Minli Li, Email: liminli_xh@163.com.
REFERENCES
- 1. Powell EE, Wong VW, Rinella M. Non-alcoholic fatty liver disease. Lancet. 2021;397:2212–2224. [DOI] [PubMed] [Google Scholar]
- 2. Younossi Z, Tacke F, Arrese M, et al. Global perspectives on nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. Hepatology. 2019;69:2672–2682. [DOI] [PubMed] [Google Scholar]
- 3. Huang DQ, El-Serag HB, Loomba R. Global epidemiology of NAFLD-related HCC: trends, predictions, risk factors and prevention. Nat Rev Gastroenterol Hepatol. 2021;18:223–238. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Lonardo A, Byrne CD, Caldwell SH, et al. Global epidemiology of nonalcoholic fatty liver disease: meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology. 2016;64:1388–1389. [DOI] [PubMed] [Google Scholar]
- 5. Chalasani N, Younossi Z, Lavine JE, et al. The diagnosis and management of nonalcoholic fatty liver disease: practice guidance from the American Association for the Study of Liver Diseases. Hepatology. 2018;67:328–357. [DOI] [PubMed] [Google Scholar]
- 6. Dyson JK, Anstee QM, McPherson S. Non-alcoholic fatty liver disease: a practical approach to treatment. Frontline Gastroenterol. 2014;5:277–286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Rinella ME, Lazarus JV, Ratziu V, et al. A multi-society Delphi consensus statement on new fatty liver disease nomenclature. J Hepatol. 2023;79:1542–1556. [DOI] [PubMed] [Google Scholar]
- 8. Montesi L, Moscatiello S, Malavolti M, et al. Physical activity for the prevention and treatment of metabolic disorders. Intern Emerg Med. 2013;8:655–666. [DOI] [PubMed] [Google Scholar]
- 9. Berzigotti A, Saran U, Dufour JF. Physical activity and liver diseases. Hepatology. 2016;63:1026–1040. [DOI] [PubMed] [Google Scholar]
- 10. Johnson NA, Sachinwalla T, Walton DW, et al. Aerobic exercise training reduces hepatic and visceral lipids in obese individuals without weight loss. Hepatology. 2009;50:1105–1112. [DOI] [PubMed] [Google Scholar]
- 11. Stine JG, Long MT, Corey KE, et al. American College of Sports Medicine (ACSM) International Multidisciplinary Roundtable report on physical activity and nonalcoholic fatty liver disease. Hepatol Commun. 2023;7:e0108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Kim D, Konyn P, Cholankeril G, et al. Physical activity is associated with nonalcoholic fatty liver disease and significant fibrosis measured by FibroScan. Clin Gastroenterol Hepatol. 2022;20:e1438–e1455. [DOI] [PubMed] [Google Scholar]
- 13. Holtermann A, Krause N, van der Beek AJ, et al. The physical activity paradox: six reasons why occupational physical activity (OPA) does not confer the cardiovascular health benefits that leisure time physical activity does. Br J Sports Med. 2018;52:149–150. [DOI] [PubMed] [Google Scholar]
- 14. Li J, Loerbroks A, Angerer P. Physical activity and risk of cardiovascular disease: what does the new epidemiological evidence show? Curr Opin Cardiol. 2013;28:575–583. [DOI] [PubMed] [Google Scholar]
- 15. Holtermann A, Hansen JV, Burr H, et al. The health paradox of occupational and leisure-time physical activity. Br J Sports Med. 2012;46:291–295. [DOI] [PubMed] [Google Scholar]
- 16. Ballestri S, Nascimbeni F, Baldelli E, et al. NAFLD as a sexual dimorphic disease: role of gender and reproductive status in the development and progression of nonalcoholic fatty liver disease and inherent cardiovascular risk. Adv Ther. 2017;34:1291–1326. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Castera L, Friedrich-Rust M, Loomba R. Noninvasive assessment of liver disease in patients with nonalcoholic fatty liver disease. Gastroenterology. 2019;156:1264–1281.e1264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Eddowes PJ, Sasso M, Allison M, 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;156:1717–1730. [DOI] [PubMed] [Google Scholar]
- 19. Piercy KL, Troiano RP, Ballard RM, et al. The Physical Activity Guidelines for Americans. JAMA. 2018;320:2020–2028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Siddiqui MS, Vuppalanchi R, Van Natta ML, et al. Vibration-controlled transient elastography to assess fibrosis and steatosis in patients with nonalcoholic fatty liver disease. Clin Gastroenterol Hepatol. 2019;17:156–163.e152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Trovato FM, Castrogiovanni P, Malatino L, et al. Nonalcoholic fatty liver disease (NAFLD) prevention: role of Mediterranean diet and physical activity. Hepatobiliary Surg Nutr. 2019;8:167–169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Kwak MS, Kim D. Non-alcoholic fatty liver disease and lifestyle modifications, focusing on physical activity. Korean J Intern Med. 2018;33:64–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Keating SE, Hackett DA, George J, et al. Exercise and non-alcoholic fatty liver disease: a systematic review and meta-analysis. J Hepatol. 2012;57:157–166. [DOI] [PubMed] [Google Scholar]
- 24. Kim D, Vazquez-Montesino LM, Li AA, et al. Inadequate physical activity and sedentary behavior are independent predictors of nonalcoholic fatty liver disease. Hepatology. 2020;72:1556–1568. [DOI] [PubMed] [Google Scholar]
- 25. Park Y, Sinn DH, Kim K, et al. Associations of physical activity domains and muscle strength exercise with non-alcoholic fatty liver disease: a nation-wide cohort study. Sci Rep. 2023;13:4724. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Byambasukh O, Zelle D, Corpeleijn E. Physical activity, fatty liver, and glucose metabolism over the life course: The Lifelines Cohort. Am J Gastroenterol. 2019;114:907–915. [DOI] [PubMed] [Google Scholar]
- 27. Jørgensen K. Permissible loads based on energy expenditure measurements. Ergonomics. 1985;28:365–369. [DOI] [PubMed] [Google Scholar]
- 28. Sorokin AV, Araujo CG, Zweibel S, et al. Atrial fibrillation in endurance-trained athletes. Br J Sports Med. 2011;45:185–188. [DOI] [PubMed] [Google Scholar]
- 29. Farzanegi P, Dana A, Ebrahimpoor Z, et al. Mechanisms of beneficial effects of exercise training on non-alcoholic fatty liver disease (NAFLD): roles of oxidative stress and inflammation. Eur J Sport Sci. 2019;19:994–1003. [DOI] [PubMed] [Google Scholar]
- 30. de Oliveira S, Houseright RA, Graves AL, et al. Metformin modulates innate immune-mediated inflammation and early progression of NAFLD-associated hepatocellular carcinoma in zebrafish. J Hepatol. 2019;70:710–721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Kasapis C, Thompson PD. The effects of physical activity on serum C-reactive protein and inflammatory markers: a systematic review. J Am Coll Cardiol. 2005;45:1563–1569. [DOI] [PubMed] [Google Scholar]
- 32. Jaruvongvanich V, Sanguankeo A, Riangwiwat T, et al. Testosterone, sex hormone-binding globulin and nonalcoholic fatty liver disease: a systematic review and meta-analysis. Ann Hepatol. 2017;16:382–394. [DOI] [PubMed] [Google Scholar]
- 33. Palmisano BT, Zhu L, Stafford JM. Role of estrogens in the regulation of liver lipid metabolism. Adv Exp Med Biol. 2017;1043:227–256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Grossmann M, Wierman ME, Angus P, et al. Reproductive endocrinology of nonalcoholic fatty liver disease. Endocr Rev. 2019;40:417–446. [DOI] [PubMed] [Google Scholar]
- 35. Heredia NI, Zhang X, Balakrishnan M, et al. Physical activity and diet quality in relation to non-alcoholic fatty liver disease: a cross-sectional study in a representative sample of US adults using NHANES 2017-2018. Prev Med. 2022;154:106903. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Tamaki N, Ajmera V, Loomba R. Non-invasive methods for imaging hepatic steatosis and their clinical importance in NAFLD. Nat Rev Endocrinol. 2022;18:55–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Park CC, Nguyen P, Hernandez C, et al. Magnetic resonance elastography vs transient elastography in detection of fibrosis and noninvasive measurement of steatosis in patients with biopsy-proven nonalcoholic fatty liver disease. Gastroenterology. 2017;152:598–607.e592. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Imajo K, Kessoku T, Honda Y, et al. Magnetic resonance imaging more accurately classifies steatosis and fibrosis in patients with nonalcoholic fatty liver disease than transient elastography. Gastroenterology. 2016;150:626–637.e627. [DOI] [PubMed] [Google Scholar]