Indexed In : Science Citation Index Expanded(SCIE), MEDLINE,
Pubmed/Pubmed Central, Elsevier Bibliographic, Google Scholar,
Databases(Scopus & Embase), KCI, KoreaMed, DOAJ
Gut and Liver is an international journal of gastroenterology, focusing on the gastrointestinal tract, liver, biliary tree, pancreas, motility, and neurogastroenterology. Gut atnd Liver delivers up-to-date, authoritative papers on both clinical and research-based topics in gastroenterology. The Journal publishes original articles, case reports, brief communications, letters to the editor and invited review articles in the field of gastroenterology. The Journal is operated by internationally renowned editorial boards and designed to provide a global opportunity to promote academic developments in the field of gastroenterology and hepatology. +MORE
Yong Chan Lee |
Professor of Medicine Director, Gastrointestinal Research Laboratory Veterans Affairs Medical Center, Univ. California San Francisco San Francisco, USA |
Jong Pil Im | Seoul National University College of Medicine, Seoul, Korea |
Robert S. Bresalier | University of Texas M. D. Anderson Cancer Center, Houston, USA |
Steven H. Itzkowitz | Mount Sinai Medical Center, NY, USA |
All papers submitted to Gut and Liver are reviewed by the editorial team before being sent out for an external peer review to rule out papers that have low priority, insufficient originality, scientific flaws, or the absence of a message of importance to the readers of the Journal. A decision about these papers will usually be made within two or three weeks.
The remaining articles are usually sent to two reviewers. It would be very helpful if you could suggest a selection of reviewers and include their contact details. We may not always use the reviewers you recommend, but suggesting reviewers will make our reviewer database much richer; in the end, everyone will benefit. We reserve the right to return manuscripts in which no reviewers are suggested.
The final responsibility for the decision to accept or reject lies with the editors. In many cases, papers may be rejected despite favorable reviews because of editorial policy or a lack of space. The editor retains the right to determine publication priorities, the style of the paper, and to request, if necessary, that the material submitted be shortened for publication.
Hong Jun Yang1 , Yeon Pyo Hong2 , Tai-Young Yoon3 , Jae-Hong Ryoo4 , Joong-Myung Choi3 , Chang-Mo Oh3
Correspondence to: Chang-Mo Oh
ORCID https://orcid.org/0000-0002-5709-9350
E-mail kachas@naver.com
Joong-Myung Choi
ORCID https://orcid.org/0000-0002-3610-6662
E-mail utang513@naver.com
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Gut Liver 2023;17(4):600-609. https://doi.org/10.5009/gnl220345
Published online March 17, 2023, Published date July 15, 2023
Copyright © Gut and Liver.
Background/Aims: This study aimed to examine the independent and synergistic association of aerobic physical activity and resistance exercise with nonalcoholic fatty liver disease (NAFLD) using a nationwide representative database.
Methods: This was a cross-sectional study using data from the Korea National Health and Nutritional Examination Survey between 2007 and 2010. Multiple logistic regression models were used to examine the independent and synergistic (additive interaction) associations of aerobic physical activity and resistance exercise with NAFLD after adjusting for multiple covariates.
Results: The prevalence of NAFLD was 26.2% for men and 17.6% for women. In the fully adjusted multiple logistic regression model to examine the independent association of aerobic physical activity or resistance exercise with NAFLD, the odds ratios for NAFLD were significantly decreased in both men (p=0.03) and women (p<0.01) who had highly active aerobic physical activity. Regarding the frequency of resistance exercise, the odds ratio for NAFLD was decreased in men who did resistance exercise ≥5 days per week (p=0.04), but not in women (p=0.19). However, when investigating the synergistic associations of aerobic physical activity and resistance exercise, the odds ratios for NAFLD significantly decreased when the frequency of both exercises increased together in both men (p for interaction <0.01) and women (p for interaction<0.01).
Conclusions: Combining aerobic physical activity and resistance exercise had a synergistic preventive association for NAFLD in Korean men and women.
Keywords: Physical activity, Resistance training, Non-alcoholic fatty liver disease
Nonalcoholic fatty liver disease (NAFLD) is as an excessive fat deposit in the liver which can be observed in a liver biopsy or imaging test such as ultrasonography, computed tomography (CT) or magnetic resonance imaging.1 The global prevalence of NAFLD was estimated as 25.2%, and the prevalence of NAFLD in Asian regions (27.4%) was higher than those in European countries (23.7%) and North America (24.1%).2 The high prevalence of NAFLD in Asia was thought to partly be due to the rapidly increasing obesity and rapid change in life styles and diet in Asian countries.3 In South Korea, the prevalence of NAFLD was estimated as 30.7% in men and 21.6% in women.4 Especially the prevalence of NAFLD in men showed a rapidly increasing trend from 2007–2009 to 2016–2018.4
NAFLD is closely associated with lifestyle factors such as sedentary lifestyle, low physical activity, high calorie diet and diet rich in processed carbohydrates.5 Regarding the effects of exercise, many previous studies reported that low to moderate intensity exercise has an effect to improve metabolic abnormalities and lower alanine aminotransferase (ALT) levels among people with NAFLD.6-8 Some studies report that resistance exercise as well as aerobic exercise was also effective to improve metabolic abnormalities among people with NAFLD such as reducing fat mass and improve insulin resistance.9-11
However, most of the previous studies have focused on the association of aerobic exercise with treatment or improvement of NAFLD-related metabolic indicators such as ALT, serum insulin level, and there were few studies examining the relationship between resistance exercise and prevention of NAFLD in a large population. Although some studies have suggested that resistance exercise can improve NAFLD with less energy than aerobic exercise, it is mostly based on studies with small number of participants, and these findings were not representative for the entire population.10 In addition, aerobic physical activity and resistance exercise are expected to have a preventive association for NAFLD, but it is still questionable whether combining aerobic physical activity and resistance exercise together can provide a synergistic preventive association for NAFLD. To our best knowledge, there were few studies on the combined protective association of aerobic physical activity and resistance exercise with the NAFLD.
Therefore, we aimed to examine independent and synergistic association of resistance exercise and aerobic physical activity with NAFLD using the Korea National Health and Nutrition Examination Survey (KNHANES) database, which is a large scale and representative of general Korean population.
This study used a nationwide database from the KNHANES from 2007 to 2010. The detailed protocols, and contents of the KNHANES have been described by previous data profile study in detail.12,13 The KNHANES includes health survey database including health behavior, disease history, and nutritional status, health examination database including blood samples, urine tests, and anthropometric measurements. Quality control assessment of the survey procedures and results was conducted by both external experts and members of the Korea Center for Disease Control and Prevention. The raw database for KNHANES could be provided freely on the homepage of KNHANES according to request procedure of Korea Center for Disease Control and Prevention. Ethics approval for the research protocol was obtained by the Institutional Review Board of Kyung Hee University (IRB number: KHSIRB-21-104(EA), Seoul, Korea).
A total 33,829 people participated in the KNHANES survey from 2007 to 2010 (Supplementary Fig. 1). Of these those aged under 20 years (n=8,954) were excluded from the analysis. An additional 4,273 people who did not respond to physical activity or resistance exercise, which are the main exposure variables, were excluded from the analysis. Therefore, 20,602 people were considered as eligible participants. Of these eligible participants, 641 were excluded due to a past history of cancer, 64 were excluded due to a past history of chronic kidney disease, 780 were excluded due to a past history of cardiovascular disease, and 687 were excluded because they had hepatitis B or hepatitis C. Individuals who did not have measured covariates were excluded (n=3,453): fasting blood glucose (n=778), body mass index (BMI) (n=62), waist circumference (n=53), alcohol drinking (n=63), smoking history (n=29), blood pressure (n=2), total cholesterol (n=4), high-density lipoprotein cholesterol (n=10), and equivalent income (n=270). Lastly, high-risk drinkers were excluded from analysis (n=2,182), because excessive alcohol drink can be associated with the increased risk of alcoholic fatty liver disease. Finally, 14,977 participants were included in the analysis set.
The level of aerobic physical activity was measured the Korean version of the International Physical Activity Questionnaire Short Form in the KNHANES.14,15 The metabolic equivalents (METs) were calculated for each item of walking, moderate intensity and vigorous physical activity. Based on MET minutes, the level of aerobic physical activity was divided into inactive, minimally active, highly active physical activity according to the guideline of IPAQ. Each criterion were categorized as follows. The criterion for highly active physical activity must meet one of the two conditions: (1) vigorous physical activity 3 or more days per week and consuming at least 1,500 MET minutes per week; or (2) one or more of walking, moderate or vigorous physical activity for 7 days and consuming at least 3,000 MET minutes per week. The minimally active criterion must meet one of the three conditions: (1) vigorous physical activity 20 minutes or more per day and at least 3 days per week; or (2) moderate physical activity 20 minutes or more per day and at least 5 days per week; or (3) one or more of walking, moderate or vigorous physical activity 7 or more days per week and consuming at least 600 MET minutes per week. The inactive criterion is neither highly active physical activity nor minimally active.15 The frequency of resistance exercise was measured by the following question: “In the last 7 days, how many days did you work out push-ups, sit-ups, dumbbells or barbells, chin-ups?” Based on the guidelines of the American College of Sports Medicine, which recommends resistance exercise at least 2 days per week,16 the raw data of resistance exercise frequency were classified into the following categories: 0 to 1 day per week, 2 to 4 days per week, ≥5 days per week.
In this study, using the NAFLD liver fat score proposed by Kotronen
The NAFLD liver fat score was calculated by –2.89+1.18 ×metabolic syndrome (yes=1/no=0)+0.45×type 2 diabetes (yes=2/no=0)+0.15×fasting serum insulin (μU/L)+0.04× aspartate aminotransferase (AST; U/L)–0.94×AST/ALT (U/L)13,17
As a diagnostic method used in many previous studies, it shows a relatively high value of area under a receiver operating characteristic curve of 0.86, with a sensitivity of 84% and a specificity of 69%.
In addition, we used the hepatic steatosis index (HSI) score, a noninvasive index of NAFLD, to examine whether the results are consistent with other noninvasive index of NAFLD. The HSI score was defined as “8×ALT/AST ratio (U/L)+BMI (kg/m2) (+2 if diabetes mellitus), (+2 if women).”18 If the HSI score was equal to higher than 36, it was considered as NAFLD.18
Venous blood samples were collected from participants after at least 8 hours of fasting. Also, serum levels of fasting blood glucose, insulin, total cholesterol, AST, and ALT, were measured using a Hitachi Automatic Analyzer 7600 (Hitachi, Tokyo, Japan). Homeostasis model assessment of insulin resistance (HOMA-IR) is an important indicator of insulin resistance. It was calculated using fasting glucose and insulin, the formula is as follows:19
HOMA-IR=fasting insulin (µU/mL)×fasting glucose (mg/dL)/405
Educational level, income level, smoking history, and monthly alcohol drinking were obtained from the health survey questionnaire of the KNHANES. Monthly alcohol drinking was defined as drinking more than once per month. High-risk drinker was defined as proportion of consuming more than seven drinks (five drinks for women) at one time and drink more than twice per week. Income level was divided into quartile groups based on overall household income. Height (m) and weight (kg) were measured by trained nurses, and BMI was calculated as weight divided by square of the height. Based on the criterion of the National Cholesterol Education Program–Adult Treatment Panel III, metabolic syndrome was defined when the participants had three or more following five conditions: (1) waist circumference of ≥90 cm for men and ≥85 cm for females; (2) triglycerides of ≥150 mg/dL for participants; (3) high-density lipoprotein cholesterol of < 40 mg/dL for men and <50 mg/dL for females; (4) systolic blood pressure of ≥130 mm Hg and diastolic blood pressure ≥85 mm Hg for participants; (5) fasting plasma glucose level ≥100 mg/dL.20
All statistical analysis were performed by gender. Continuous variables were expressed as the mean±standard error and categorical variables were presented as numbers (percentage) using the complex sampling from the KNHANES survey design. Linear regression models of continuous variables were performed to compare differences in general and clinical characteristics in participants. The Rao-Scott chi-square test for categorical variables were used to test differences in general and clinical characteristics. The interaction of the prevalence of NAFLD according to resistance exercise and aerobic physical activity level was calculated by the Cochran-Mantel-Haenszel statistics. Multivariable logistic regression analyses were used to examine the independent and synergistic association between frequency of resistance exercise and aerobic physical activity levels with NAFLD by gender, after adjusting for resistance exercise or aerobic physical activity, age, BMI, waist circumference, smoking status, alcohol intake, diabetes mellitus, hypertension, total cholesterol, high-density lipoprotein cholesterol, income level and education level.13,21 Regarding synergistic association between aerobic physical activity and resistance exercise, the difference between the likelihood of the full model including the interaction term and the likelihood of the null model without the interaction term was tested using the likelihood ratio test. A sensitivity analysis was also performed as NAFLD was defined using HSI score, instead of NAFLD liver fat score.
p-values <0.05 were considered statistically significant. All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) and Stata version 17.0 (StataCorp., College Station, TX, USA).
Of the total 14,977 participants, 5,712 (38.1%) were men and 9,265 (61.9%) were women (Table 1). The age o of study participants was 44.2±0.2 years. The age of men and women was 43.1±0.3 and 45.0±0.3 years, respectively. According to the level of aerobic physical activity, the inactive group was 37.6%, minimally active group was 32.3%, and highly active group was 30.1%. By gender, the inactive group was 33.4%, the minimally active group was 31.3%, and the highly active group was 35.3% in men, and in women, the inactive group was 41.1%, the minimally active group was 33.1%, the highly active group was 25.8%. The ratio of highly active group was higher in men than in women (p<0.01). According to the frequency of resistance exercise, 6.6% of participants were doing resistance exercise 5 or more days a week, 14.9% of participants were doing resistance exercise 2 to 4 days per week, and 78.5% of participants did resistance exercise less than once a week. According to gender, the frequency of doing resistance exercise more than 5 days per week was 10.5% for men, which was significantly higher than that (3.4%) for women (p<0.01). The prevalence of metabolic syndrome as defined by the NAFLD liver fat score was 29.7% in men, which was significantly higher than 19.0% in women (p<0.01).
Table 1 Baseline Characteristics of the Study Participants
Characteristic | Overall (n=14,977) | Gender | p-value* | |
---|---|---|---|---|
Men (n=5,712) | Women (n=9,265) | |||
Continuous variable | ||||
Age, yr | 44.2±0.2 | 43.1±0.3 | 45.0±0.3 | <0.01 |
METs, min/wk | 2,962±54 | 3,472±82 | 2,534±58 | <0.01 |
ALT, U/L | 21.2±0.2 | 26.3±0.3 | 17.0±0.1 | <0.01 |
AST, U/L | 21.2±0.1 | 23.3±0.2 | 19.5±0.1 | <0.01 |
GGT, U/L (n=4,411) | 30.1±0.8 | 41.7±1.5 | 20.4±0.5 | <0.01 |
Systolic blood pressure, mm Hg | 116.6±0.2 | 119.3±0.3 | 114.3±0.3 | <0.01 |
Diastolic blood pressure, mm Hg | 76.2±0.2 | 79.2±0.2 | 73.6±0.2 | <0.01 |
Body mass index, kg/m2 | 23.5±0.04 | 23.9±0.05 | 23.2±0.05 | <0.01 |
Waist circumference, cm | 80.4±0.1 | 83.6±0.2 | 77.7±0.2 | <0.01 |
Fasting glucose level, mg/dL | 96.2±0.2 | 97.5±0.4 | 95.0±0.3 | <0.01 |
Insulin, μU/mL | 10.1±0.1 | 10.1±0.1 | 10.1±0.1 | 0.94 |
Total cholesterol, mg/dL | 186.2±0.4 | 185.9±0.6 | 186.4±0.5 | 0.45 |
HDL cholesterol, mg/dL | 48.1±0.1 | 45.0±0.2 | 50.6±0.2 | <0.01 |
Triglyceride, mg/dL | 126.5±1.1 | 146.7±1.8 | 109.9±1.0 | <0.01 |
Alcohol intake, g/day | 3.8±0.06 | 6.3±0.10 | 1.6±0.03 | <0.01 |
Categorical variable | ||||
Physical activity | <0.01 | |||
Inactive | 5,746 (37.6) | 1,928 (33.4) | 3,818 (41.1) | |
Minimally active | 4,844 (32.3) | 1,782 (31.3) | 3,062 (33.1) | |
Highly active | 4,387 (30.1) | 2,002 (35.3) | 2,385 (25.8) | |
Resistance exercise | <0.01 | |||
0–1 day/wk | 12,125 (78.5) | 3,979 (67.9) | 8,146 (87.2) | |
2–4 day/wk | 1,928 (14.9) | 1,121 (21.6) | 807 (9.4) | |
≥5 day/wk | 924 (6.6) | 612 (10.5) | 312 (3.4) | |
Smoking history | <0.01 | |||
Never-smoker | 9,887 (61.2) | 1,414 (26.1) | 8,473 (90.2) | |
Past smoker | 2,313 (15.8) | 1,976 (29.9) | 337 (4.1) | |
Current smoker | 2,777 (23.1) | 2,322 (44.1) | 455 (5.7) | |
Hypertension (n=14,936) | <0.01 | |||
Yes | 4,239 (24.4) | 1,844 (27.6) | 2,395 (21.7) | |
No | 10,697 (75.6) | 3,852 (72.4) | 6,845 (78.3) | |
Diabetes mellitus (n=14,580) | 0.08 | |||
Yes | 1,305 (7.7) | 564 (8.2) | 741 (7.3) | |
No | 13,275 (92.3) | 4,965 (91.8) | 8,310 (92.7) | |
Metabolic syndrome | <0.01 | |||
Yes | 3,638 (22.0) | 1,518 (24.4) | 2,120 (19.9) | |
No | 11,339 (78.0) | 4,194 (75.6) | 7,145 (80.1) | |
NAFLD by NLFS score (n=14,580) | 3,550 (23.8) | 1,670 (29.7) | 1,880 (19.0) | <0.01 |
Continuous variables were expressed as means±standard errors and categorical variables were expressed as number (%) by using the sampling weights to reflect complex Korea National Health and Nutrition Examination Survey sampling design.
METs, metabolic equivalents; ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, gamma glutamyl transferase; HDL, high-density lipoprotein; NAFLD, nonalcoholic fatty liver disease; NLFS, nonalcoholic fatty liver disease liver fat score.
*Linear regressions for complex sampling survey design were used to test trend for continuous variables and Rao-Scott chi-square for complex sampling survey design were used to test differences for categorical variables.
There were differences in metabolic profiles and prevalence diabetes mellitus, metabolic syndrome and NAFLD according to the level of physical activity and frequency of resistance exercise in Korean men (Supplementary Table 1). There was no significant difference in BMI according to the level of physical activity or the frequency of resistance exercise. However, there was significantly decreasing trend in waist circumference and fasting glucose level according to the level of physical activity or the frequency of resistance exercise. HOMA-IR showed a significant decreasing trend only according to the physical activity level (p for trend <0.01). ALT showed a significant decreasing trend only according to frequency of resistance exercise (p for trend <0.01), it was not significant according to level of physical activity (p for trend=0.10). The prevalences of metabolic syndrome and NAFLD showed significant decreasing trends as the level of physical activity and frequency of resistance exercise increased (Supplementary Fig. 2).
There were differences in metabolic profiles and prevalence diabetes mellitus, metabolic syndrome and NAFLD according to the level of physical activity and frequency of resistance exercise in Korean women (Supplementary Table 2). BMI showed a significant increasing trend as the level of physical activity increased, but it was not significant for frequency of resistance exercise. There was no significant difference in waist circumference and ALT according to the level of physical activity or the frequency of resistance exercise. Fasting glucose level and HOMA-IR showed significant decreasing trends only according to frequency of resistance exercise (p for trend <0.01), but they were not significant for level of physical activity. The prevalences of metabolic syndrome and NAFLD showed significant decreasing trends as frequency of resistance exercise increased, but they were not significant for the level of physical activity (Supplementary Fig. 3).
Multiple logistic regression analysis was used to examine the relationship between level of physical activity, frequency of resistance exercise and NAFLD by gender (Table 2). In men, the odds ratio (OR) for NAFLD among physically highly active people was 0.79 (95% confidence interval [CI], 0.63 to 0.98), as compared to those who were physically inactive. Regarding frequency of resistance exercise in men, the OR for NAFLD among men did resistance exercise more than five times a week was 0.73 (95% CI, 0.54 to 0.99) compared with men did resistance exercise less than once a week. In women, the OR for NAFLD among physically highly active people was 0.74 (95% CI, 0.59 to 0.91), as compared to those who were physically inactive. In addition, OR for NAFLD showed a significantly decreasing trend as physical activity increased (Fig. 1). Regarding frequency of resistance exercise in women, the OR for NAFLD among women did resistance exercise more than five times a week was 0.73 (95% CI, 0.45 to 1.17) compared with men did resistance exercise less than once a week (p=0.19). There was a decreasing trend for NAFLD as the frequency of resistance exercise increased in women, but it was not statistically significant (Fig. 2). When sensitivity analysis was performed with HSI score for NAFLD, the OR for NAFLD was significantly decreased among physically highly active men or among men doing resistance exercise for more than 5 days, but not in women (Supplementary Table 3).
Table 2 Independent Association between Physical Activity, Resistance Exercise and NAFLD among Korean Men and Women
Variable | Unadjusted OR (95% CI) | p-value | Age and physical activity/ resistance exercise adjusted OR (95% CI)* | p-value | Multivariable adjusted OR (95% CI)† | p-value |
---|---|---|---|---|---|---|
Men | ||||||
Physical activity | ||||||
Inactive | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
Minimally active | 0.86 (0.72–1.02) | 0.08 | 0.91 (0.76–1.08) | 0.28 | 0.90 (0.71–1.14) | 0.39 |
Highly active | 0.73 (0.63–0.86) | <0.01 | 0.80 (0.68–0.94) | <0.01 | 0.79 (0.63–0.98) | 0.03 |
Resistance exercise | ||||||
0–1 day/wk | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
2–4 day/wk | 0.77 (0.64–0.93) | <0.01 | 0.84 (0.70–1.02) | 0.08 | 0.86 (0.67–1.10) | 0.22 |
≥5 day/wk | 0.68 (0.53–0.88) | <0.01 | 0.74 (0.57–0.96) | 0.02 | 0.73 (0.54–0.99) | 0.04 |
Women | ||||||
Physical activity | ||||||
Inactive | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
Minimally active | 0.92 (0.80–1.06) | 0.26 | 0.94 (0.81–1.09) | 0.41 | 0.88 (0.74–1.06) | 0.18 |
Highly active | 0.83 (0.71–0.97) | 0.02 | 0.89 (0.76–1.05) | 0.18 | 0.74 (0.59–0.91) | <0.01 |
Resistance exercise | ||||||
0–1 day/wk | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
2–4 day/wk | 0.76 (0.61–0.95) | 0.01 | 0.94 (0.75–1.17) | 0.56 | 1.02 (0.76–1.38) | 0.89 |
≥5 day/wk | 0.75 (0.53–1.07) | 0.12 | 0.75 (0.53–1.08) | 0.12 | 0.73 (0.45–1.17) | 0.19 |
Logistic regression model for complex sampling survey design were used to examine the association between physical activity, resistance exercise and nonalcoholic fatty liver disease (NAFLD).
OR, odds ratio; CI, confidence interval.
*Model 1 adjusted for age and physical activity/resistance exercise; †Model 2 adjusted for age, physical activity/resistance exercise, body mass index, waist circumference, smoking status, alcohol intake, diabetes mellitus, hypertension, total cholesterol, high-density lipoprotein cholesterol, income level and education level.
We also investigated the interaction between level of physical activity, frequency of resistance exercise on NAFLD by gender (Table 3). In men, the prevalence rates of NAFLD showed decreasing trends as the frequency of resistance exercise increased, even in those who were physically active (p for interaction <0.001) (Supplementary Table 4). Similarly, there was a decreasing trend for NAFLD as the frequency of resistance exercise increased in women in those who were highly physically active. Both men and women, there were significant multiplicative interactions between level of physical activity, frequency of resistance exercise on NAFLD (p for interaction <0.001 for men, p for interaction <0.001 for women). In other words, these findings show that when resistance exercise as well as aerobic physical activity were performed together, the observed preventive association for NAFLD was much greater than expected preventive association (Fig. 3). However, in the sensitivity analysis which was performed with HSI score for NAFLD, a synergistic effect between aerobic physical activity and resistance exercise was also observed in men, but not in women (Supplementary Table 5).
Table 3 Interaction of Odds Ratio on NAFLD between Aerobic Physical Activity and Resistance Exercise for the Multivariable Logistic Regression Model
Resistance exercise | Physical activity level, odds ratio (95% CI) | p for interaction* | ||
---|---|---|---|---|
Inactive | Minimally active | Highly active | ||
Men | <0.001 | |||
0–1 day/wk | 1.00 (reference) | 1.00 (0.77–1.30) | 0.82 (0.63–1.07) | |
2–4 day/wk | 1.09 (0.71–1.68) | 0.73 (0.47–1.14) | 0.67 (0.47–0.96) | |
≥5 day/wk | 0.97 (0.42–2.25) | 0.55 (0.35–0.88) | 0.64 (0.43–0.93) | |
Women | <0.001 | |||
0–1 day/wk | 1.00 (reference) | 0.86 (0.80–1.06) | 0.76 (0.60–0.96) | |
2–4 day/wk | 0.92 (0.52–1.64) | 0.96 (0.59–1.57) | 0.75 (0.47–1.20) | |
≥5 day/wk | 1.42 (0.55–3.63) | 0.85 (0.44–1.66) | 0.28 (0.12–0.65) |
Multiple logistic regression models for complex sampling survey design were used to examine the association between physical activity, resistance exercise and nonalcoholic fatty liver disease (NAFLD) after adjusting for age, body mass index, waist circumference, smoking status, alcohol intake, diabetes mellitus, hypertension, total cholesterol, high-density lipoprotein cholesterol, income level and education level.
CI, confidence interval.
*p for interaction was tested as a likelihood ratio test between the full model with interaction and the null model without interaction.
This study examined the independent and synergistic association of frequency of resistance exercise and aerobic physical activity with the NAFLD for a total of 14,974 Korean men and women. Our findings showed the level of aerobic exercise was associated with the decreased odds of NAFLD, and the frequency of resistance exercise was also associated with the lower odds of NAFLD independently of the aerobic exercise level especially in Korean men. In addition, our study finding also showed that aerobic physical activity and resistance exercise had an interactively preventive association for NAFLD both in men and women.
Although there were cumulating evidence that increase of physical activity could reduce the risk of NAFLD, there were still lack of evidence on whether resistance exercise itself is independently associated with a reduced risk of NAFLD in a large sample of general population.6,10,22 Most previous studies have focused on whether resistance exercise could reduce liver fat, liver enzyme levels (ALT and AST), and insulin resistance in NAFLD patients.6,10,23 However, these studies have several limitations that the number of study participants was relatively small, and the preventive or synergistic association of resistance exercise with NAFLD in the general population was not investigated. Although these studies may indirectly show that resistance exercise could prevent or improve NAFLD, previous studies did not provide interactively how resistance exercise could prevent NAFLD depending on the level of aerobic physical activity. On the contrary, our study findings showed that the probability of NAFLD has decreased as aerobic physical activity level has increased even when resistance exercise was performed more than 5 days per week, or as the frequency of resistance exercise has increased even when aerobic physical activity was sufficiently active. When subgroup analysis was performed according to age group, the preventive association of aerobic physical activity with NAFLD was reduced among the old age group (≥55 years old). In addition, the synergistic association of aerobic physical activity and resistance exercise showed a reduced synergistic association among elderly women (≥55 years old) than in elderly men. In sensitivity analysis using HSI score for NAFLD, the preventive association of aerobic physical activity and resistance exercise on NAFLD was not significant in women than in men, and the synergistic association was not clear in women. It may be because the number of women doing resistance exercise was much smaller compared to men, and the efficiency of exercise itself in women or older men is lower than that of young men.24
It has been reported that combining aerobic exercise plus resistance exercise together is more effective than aerobic exercise or resistance exercise alone in type 2 diabetes, which is characterized by insulin resistance, similar to NAFLD.25 In addition, considering that sarcopenia is a risk factor for NAFLD independent of obesity or insulin resistance, it could be reasonable inference that resistance exercise that increases muscle strength could be a preventive factor of NAFLD independently of aerobic exercise.26,27 Indeed, resistance exercise has prevented the decrease of muscle strength and improved physical function in the meta-analysis of 25 randomized controlled trials.28 Furthermore, it has been previously reported that combining resistance exercise with aerobic physical activity was more effective in improving fat mass, metabolic profiles, and inflammatory state than doing aerobic exercise alone in obese adults and children.29,30 Considering these previous studies and our study findings, it is reasonable to infer that combining resistance exercise with aerobic physical activity might be more effective for preventing NAFLD than aerobic physical activity alone.
Our study has a number of strengths. First, we used the KNHANES database, which was representative the Korean population. Therefore, our study could present reliable study findings with sufficient statistical power. Our study provides evidence that the combination of aerobic physical activity and strength training would be better for preventing NAFLD than performing only one type of exercise. It could be helpful when physicians and policy makers provide exercise recommendations for NAFLD. Of course, additional verification is necessary to whether these results are externally applicable to non-Asian population. Furthermore, we excluded high-risk drinkers and people with hepatitis B or C, which could affect the risk of severity NAFLD. These strict exclusion criteria can ensure the homogeneity of population with NAFLD, thereby increasing the internal validity of our study.
Nonetheless, the limitation of the study should be acknowledged. Although we used scoring system for NAFLD, the standard method to diagnose the NAFLD is imaging study such as CT and liver biopsy. Therefore, there may be a bias to define the NAFLD. However, our study is based on the national representative sampling survey and it is impossible to assess the participants by using abdominal ultrasound or CT. Therefore, it was impossible to use abdominal ultrasound or CT to distinguish the NAFLD among general population. Second, it is impossible to examine the temporal relationship between aerobic physical activity, resistance exercise and NAFLD, because our study design is based on the cross-sectional study. Therefore, a future follow-up study is necessary to examine whether the combining association of aerobic physical activity and resistance exercise change longitudinally with synergistic interaction. Third, we could not know the history of drug use or dietary intake, therefore we could not exclude the effects of medication or dietary supplement. Finally, although the KNHANES surveyed physical activity using standardized questionnaires by trained investigators, there may be still problem about accuracy of physical activity measurement, because the information on exercise was obtained through the questionnaire. Especially, resistance exercise was surveyed only in frequency, not intensity of exercise, so it may not reflect the total amount of resistance exercise. In addition, physical activity may not have been maintained at a constant level over a long period of time.
In conclusion, to our best knowledge, the present study is the first investigation to suggest both resistance exercise and aerobic physical activity could exert a synergistic preventive association for NAFLD in a large number of general Korean population. These synergistic associations between resistance exercise and aerobic physical activity on NAFLD were consistent in both men and women. Future cohort study or clinical trial may be necessary to confirm our study findings.
Supplementary materials can be accessed at https://doi.org/10.5009/gnl220345.
No potential conflict of interest relevant to this article was reported.
Study concept and design: J.M.C., C.M.O. Data acquisition: C.M.O. Data analysis and interpretation: H.J.Y., C.M.O., Drafting of the manuscript: H.J.Y. Critical revision of the manuscript for important intellectual content: Y.P.H., T.Y.Y., J.H.R. Statistical analysis: H.J.Y., C.M.O. Administrative, technical, or material support; study supervision: J.M.C., C.M.O. Approval of final manuscript: all authors.
Gut and Liver 2023; 17(4): 600-609
Published online July 15, 2023 https://doi.org/10.5009/gnl220345
Copyright © Gut and Liver.
Hong Jun Yang1 , Yeon Pyo Hong2 , Tai-Young Yoon3 , Jae-Hong Ryoo4 , Joong-Myung Choi3 , Chang-Mo Oh3
1Institute of Health and Environment, Graduate School of Public Health, Seoul National University, 2Department of Preventive Medicine, Chung-Ang University College of Medicine, Departments of 3Preventive Medicine and 4Occupational Medicine, School of Medicine, Kyung Hee University, Seoul, Korea
Correspondence to:Chang-Mo Oh
ORCID https://orcid.org/0000-0002-5709-9350
E-mail kachas@naver.com
Joong-Myung Choi
ORCID https://orcid.org/0000-0002-3610-6662
E-mail utang513@naver.com
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background/Aims: This study aimed to examine the independent and synergistic association of aerobic physical activity and resistance exercise with nonalcoholic fatty liver disease (NAFLD) using a nationwide representative database.
Methods: This was a cross-sectional study using data from the Korea National Health and Nutritional Examination Survey between 2007 and 2010. Multiple logistic regression models were used to examine the independent and synergistic (additive interaction) associations of aerobic physical activity and resistance exercise with NAFLD after adjusting for multiple covariates.
Results: The prevalence of NAFLD was 26.2% for men and 17.6% for women. In the fully adjusted multiple logistic regression model to examine the independent association of aerobic physical activity or resistance exercise with NAFLD, the odds ratios for NAFLD were significantly decreased in both men (p=0.03) and women (p<0.01) who had highly active aerobic physical activity. Regarding the frequency of resistance exercise, the odds ratio for NAFLD was decreased in men who did resistance exercise ≥5 days per week (p=0.04), but not in women (p=0.19). However, when investigating the synergistic associations of aerobic physical activity and resistance exercise, the odds ratios for NAFLD significantly decreased when the frequency of both exercises increased together in both men (p for interaction <0.01) and women (p for interaction<0.01).
Conclusions: Combining aerobic physical activity and resistance exercise had a synergistic preventive association for NAFLD in Korean men and women.
Keywords: Physical activity, Resistance training, Non-alcoholic fatty liver disease
Nonalcoholic fatty liver disease (NAFLD) is as an excessive fat deposit in the liver which can be observed in a liver biopsy or imaging test such as ultrasonography, computed tomography (CT) or magnetic resonance imaging.1 The global prevalence of NAFLD was estimated as 25.2%, and the prevalence of NAFLD in Asian regions (27.4%) was higher than those in European countries (23.7%) and North America (24.1%).2 The high prevalence of NAFLD in Asia was thought to partly be due to the rapidly increasing obesity and rapid change in life styles and diet in Asian countries.3 In South Korea, the prevalence of NAFLD was estimated as 30.7% in men and 21.6% in women.4 Especially the prevalence of NAFLD in men showed a rapidly increasing trend from 2007–2009 to 2016–2018.4
NAFLD is closely associated with lifestyle factors such as sedentary lifestyle, low physical activity, high calorie diet and diet rich in processed carbohydrates.5 Regarding the effects of exercise, many previous studies reported that low to moderate intensity exercise has an effect to improve metabolic abnormalities and lower alanine aminotransferase (ALT) levels among people with NAFLD.6-8 Some studies report that resistance exercise as well as aerobic exercise was also effective to improve metabolic abnormalities among people with NAFLD such as reducing fat mass and improve insulin resistance.9-11
However, most of the previous studies have focused on the association of aerobic exercise with treatment or improvement of NAFLD-related metabolic indicators such as ALT, serum insulin level, and there were few studies examining the relationship between resistance exercise and prevention of NAFLD in a large population. Although some studies have suggested that resistance exercise can improve NAFLD with less energy than aerobic exercise, it is mostly based on studies with small number of participants, and these findings were not representative for the entire population.10 In addition, aerobic physical activity and resistance exercise are expected to have a preventive association for NAFLD, but it is still questionable whether combining aerobic physical activity and resistance exercise together can provide a synergistic preventive association for NAFLD. To our best knowledge, there were few studies on the combined protective association of aerobic physical activity and resistance exercise with the NAFLD.
Therefore, we aimed to examine independent and synergistic association of resistance exercise and aerobic physical activity with NAFLD using the Korea National Health and Nutrition Examination Survey (KNHANES) database, which is a large scale and representative of general Korean population.
This study used a nationwide database from the KNHANES from 2007 to 2010. The detailed protocols, and contents of the KNHANES have been described by previous data profile study in detail.12,13 The KNHANES includes health survey database including health behavior, disease history, and nutritional status, health examination database including blood samples, urine tests, and anthropometric measurements. Quality control assessment of the survey procedures and results was conducted by both external experts and members of the Korea Center for Disease Control and Prevention. The raw database for KNHANES could be provided freely on the homepage of KNHANES according to request procedure of Korea Center for Disease Control and Prevention. Ethics approval for the research protocol was obtained by the Institutional Review Board of Kyung Hee University (IRB number: KHSIRB-21-104(EA), Seoul, Korea).
A total 33,829 people participated in the KNHANES survey from 2007 to 2010 (Supplementary Fig. 1). Of these those aged under 20 years (n=8,954) were excluded from the analysis. An additional 4,273 people who did not respond to physical activity or resistance exercise, which are the main exposure variables, were excluded from the analysis. Therefore, 20,602 people were considered as eligible participants. Of these eligible participants, 641 were excluded due to a past history of cancer, 64 were excluded due to a past history of chronic kidney disease, 780 were excluded due to a past history of cardiovascular disease, and 687 were excluded because they had hepatitis B or hepatitis C. Individuals who did not have measured covariates were excluded (n=3,453): fasting blood glucose (n=778), body mass index (BMI) (n=62), waist circumference (n=53), alcohol drinking (n=63), smoking history (n=29), blood pressure (n=2), total cholesterol (n=4), high-density lipoprotein cholesterol (n=10), and equivalent income (n=270). Lastly, high-risk drinkers were excluded from analysis (n=2,182), because excessive alcohol drink can be associated with the increased risk of alcoholic fatty liver disease. Finally, 14,977 participants were included in the analysis set.
The level of aerobic physical activity was measured the Korean version of the International Physical Activity Questionnaire Short Form in the KNHANES.14,15 The metabolic equivalents (METs) were calculated for each item of walking, moderate intensity and vigorous physical activity. Based on MET minutes, the level of aerobic physical activity was divided into inactive, minimally active, highly active physical activity according to the guideline of IPAQ. Each criterion were categorized as follows. The criterion for highly active physical activity must meet one of the two conditions: (1) vigorous physical activity 3 or more days per week and consuming at least 1,500 MET minutes per week; or (2) one or more of walking, moderate or vigorous physical activity for 7 days and consuming at least 3,000 MET minutes per week. The minimally active criterion must meet one of the three conditions: (1) vigorous physical activity 20 minutes or more per day and at least 3 days per week; or (2) moderate physical activity 20 minutes or more per day and at least 5 days per week; or (3) one or more of walking, moderate or vigorous physical activity 7 or more days per week and consuming at least 600 MET minutes per week. The inactive criterion is neither highly active physical activity nor minimally active.15 The frequency of resistance exercise was measured by the following question: “In the last 7 days, how many days did you work out push-ups, sit-ups, dumbbells or barbells, chin-ups?” Based on the guidelines of the American College of Sports Medicine, which recommends resistance exercise at least 2 days per week,16 the raw data of resistance exercise frequency were classified into the following categories: 0 to 1 day per week, 2 to 4 days per week, ≥5 days per week.
In this study, using the NAFLD liver fat score proposed by Kotronen
The NAFLD liver fat score was calculated by –2.89+1.18 ×metabolic syndrome (yes=1/no=0)+0.45×type 2 diabetes (yes=2/no=0)+0.15×fasting serum insulin (μU/L)+0.04× aspartate aminotransferase (AST; U/L)–0.94×AST/ALT (U/L)13,17
As a diagnostic method used in many previous studies, it shows a relatively high value of area under a receiver operating characteristic curve of 0.86, with a sensitivity of 84% and a specificity of 69%.
In addition, we used the hepatic steatosis index (HSI) score, a noninvasive index of NAFLD, to examine whether the results are consistent with other noninvasive index of NAFLD. The HSI score was defined as “8×ALT/AST ratio (U/L)+BMI (kg/m2) (+2 if diabetes mellitus), (+2 if women).”18 If the HSI score was equal to higher than 36, it was considered as NAFLD.18
Venous blood samples were collected from participants after at least 8 hours of fasting. Also, serum levels of fasting blood glucose, insulin, total cholesterol, AST, and ALT, were measured using a Hitachi Automatic Analyzer 7600 (Hitachi, Tokyo, Japan). Homeostasis model assessment of insulin resistance (HOMA-IR) is an important indicator of insulin resistance. It was calculated using fasting glucose and insulin, the formula is as follows:19
HOMA-IR=fasting insulin (µU/mL)×fasting glucose (mg/dL)/405
Educational level, income level, smoking history, and monthly alcohol drinking were obtained from the health survey questionnaire of the KNHANES. Monthly alcohol drinking was defined as drinking more than once per month. High-risk drinker was defined as proportion of consuming more than seven drinks (five drinks for women) at one time and drink more than twice per week. Income level was divided into quartile groups based on overall household income. Height (m) and weight (kg) were measured by trained nurses, and BMI was calculated as weight divided by square of the height. Based on the criterion of the National Cholesterol Education Program–Adult Treatment Panel III, metabolic syndrome was defined when the participants had three or more following five conditions: (1) waist circumference of ≥90 cm for men and ≥85 cm for females; (2) triglycerides of ≥150 mg/dL for participants; (3) high-density lipoprotein cholesterol of < 40 mg/dL for men and <50 mg/dL for females; (4) systolic blood pressure of ≥130 mm Hg and diastolic blood pressure ≥85 mm Hg for participants; (5) fasting plasma glucose level ≥100 mg/dL.20
All statistical analysis were performed by gender. Continuous variables were expressed as the mean±standard error and categorical variables were presented as numbers (percentage) using the complex sampling from the KNHANES survey design. Linear regression models of continuous variables were performed to compare differences in general and clinical characteristics in participants. The Rao-Scott chi-square test for categorical variables were used to test differences in general and clinical characteristics. The interaction of the prevalence of NAFLD according to resistance exercise and aerobic physical activity level was calculated by the Cochran-Mantel-Haenszel statistics. Multivariable logistic regression analyses were used to examine the independent and synergistic association between frequency of resistance exercise and aerobic physical activity levels with NAFLD by gender, after adjusting for resistance exercise or aerobic physical activity, age, BMI, waist circumference, smoking status, alcohol intake, diabetes mellitus, hypertension, total cholesterol, high-density lipoprotein cholesterol, income level and education level.13,21 Regarding synergistic association between aerobic physical activity and resistance exercise, the difference between the likelihood of the full model including the interaction term and the likelihood of the null model without the interaction term was tested using the likelihood ratio test. A sensitivity analysis was also performed as NAFLD was defined using HSI score, instead of NAFLD liver fat score.
p-values <0.05 were considered statistically significant. All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) and Stata version 17.0 (StataCorp., College Station, TX, USA).
Of the total 14,977 participants, 5,712 (38.1%) were men and 9,265 (61.9%) were women (Table 1). The age o of study participants was 44.2±0.2 years. The age of men and women was 43.1±0.3 and 45.0±0.3 years, respectively. According to the level of aerobic physical activity, the inactive group was 37.6%, minimally active group was 32.3%, and highly active group was 30.1%. By gender, the inactive group was 33.4%, the minimally active group was 31.3%, and the highly active group was 35.3% in men, and in women, the inactive group was 41.1%, the minimally active group was 33.1%, the highly active group was 25.8%. The ratio of highly active group was higher in men than in women (p<0.01). According to the frequency of resistance exercise, 6.6% of participants were doing resistance exercise 5 or more days a week, 14.9% of participants were doing resistance exercise 2 to 4 days per week, and 78.5% of participants did resistance exercise less than once a week. According to gender, the frequency of doing resistance exercise more than 5 days per week was 10.5% for men, which was significantly higher than that (3.4%) for women (p<0.01). The prevalence of metabolic syndrome as defined by the NAFLD liver fat score was 29.7% in men, which was significantly higher than 19.0% in women (p<0.01).
Table 1 . Baseline Characteristics of the Study Participants.
Characteristic | Overall (n=14,977) | Gender | p-value* | |
---|---|---|---|---|
Men (n=5,712) | Women (n=9,265) | |||
Continuous variable | ||||
Age, yr | 44.2±0.2 | 43.1±0.3 | 45.0±0.3 | <0.01 |
METs, min/wk | 2,962±54 | 3,472±82 | 2,534±58 | <0.01 |
ALT, U/L | 21.2±0.2 | 26.3±0.3 | 17.0±0.1 | <0.01 |
AST, U/L | 21.2±0.1 | 23.3±0.2 | 19.5±0.1 | <0.01 |
GGT, U/L (n=4,411) | 30.1±0.8 | 41.7±1.5 | 20.4±0.5 | <0.01 |
Systolic blood pressure, mm Hg | 116.6±0.2 | 119.3±0.3 | 114.3±0.3 | <0.01 |
Diastolic blood pressure, mm Hg | 76.2±0.2 | 79.2±0.2 | 73.6±0.2 | <0.01 |
Body mass index, kg/m2 | 23.5±0.04 | 23.9±0.05 | 23.2±0.05 | <0.01 |
Waist circumference, cm | 80.4±0.1 | 83.6±0.2 | 77.7±0.2 | <0.01 |
Fasting glucose level, mg/dL | 96.2±0.2 | 97.5±0.4 | 95.0±0.3 | <0.01 |
Insulin, μU/mL | 10.1±0.1 | 10.1±0.1 | 10.1±0.1 | 0.94 |
Total cholesterol, mg/dL | 186.2±0.4 | 185.9±0.6 | 186.4±0.5 | 0.45 |
HDL cholesterol, mg/dL | 48.1±0.1 | 45.0±0.2 | 50.6±0.2 | <0.01 |
Triglyceride, mg/dL | 126.5±1.1 | 146.7±1.8 | 109.9±1.0 | <0.01 |
Alcohol intake, g/day | 3.8±0.06 | 6.3±0.10 | 1.6±0.03 | <0.01 |
Categorical variable | ||||
Physical activity | <0.01 | |||
Inactive | 5,746 (37.6) | 1,928 (33.4) | 3,818 (41.1) | |
Minimally active | 4,844 (32.3) | 1,782 (31.3) | 3,062 (33.1) | |
Highly active | 4,387 (30.1) | 2,002 (35.3) | 2,385 (25.8) | |
Resistance exercise | <0.01 | |||
0–1 day/wk | 12,125 (78.5) | 3,979 (67.9) | 8,146 (87.2) | |
2–4 day/wk | 1,928 (14.9) | 1,121 (21.6) | 807 (9.4) | |
≥5 day/wk | 924 (6.6) | 612 (10.5) | 312 (3.4) | |
Smoking history | <0.01 | |||
Never-smoker | 9,887 (61.2) | 1,414 (26.1) | 8,473 (90.2) | |
Past smoker | 2,313 (15.8) | 1,976 (29.9) | 337 (4.1) | |
Current smoker | 2,777 (23.1) | 2,322 (44.1) | 455 (5.7) | |
Hypertension (n=14,936) | <0.01 | |||
Yes | 4,239 (24.4) | 1,844 (27.6) | 2,395 (21.7) | |
No | 10,697 (75.6) | 3,852 (72.4) | 6,845 (78.3) | |
Diabetes mellitus (n=14,580) | 0.08 | |||
Yes | 1,305 (7.7) | 564 (8.2) | 741 (7.3) | |
No | 13,275 (92.3) | 4,965 (91.8) | 8,310 (92.7) | |
Metabolic syndrome | <0.01 | |||
Yes | 3,638 (22.0) | 1,518 (24.4) | 2,120 (19.9) | |
No | 11,339 (78.0) | 4,194 (75.6) | 7,145 (80.1) | |
NAFLD by NLFS score (n=14,580) | 3,550 (23.8) | 1,670 (29.7) | 1,880 (19.0) | <0.01 |
Continuous variables were expressed as means±standard errors and categorical variables were expressed as number (%) by using the sampling weights to reflect complex Korea National Health and Nutrition Examination Survey sampling design..
METs, metabolic equivalents; ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, gamma glutamyl transferase; HDL, high-density lipoprotein; NAFLD, nonalcoholic fatty liver disease; NLFS, nonalcoholic fatty liver disease liver fat score..
*Linear regressions for complex sampling survey design were used to test trend for continuous variables and Rao-Scott chi-square for complex sampling survey design were used to test differences for categorical variables..
There were differences in metabolic profiles and prevalence diabetes mellitus, metabolic syndrome and NAFLD according to the level of physical activity and frequency of resistance exercise in Korean men (Supplementary Table 1). There was no significant difference in BMI according to the level of physical activity or the frequency of resistance exercise. However, there was significantly decreasing trend in waist circumference and fasting glucose level according to the level of physical activity or the frequency of resistance exercise. HOMA-IR showed a significant decreasing trend only according to the physical activity level (p for trend <0.01). ALT showed a significant decreasing trend only according to frequency of resistance exercise (p for trend <0.01), it was not significant according to level of physical activity (p for trend=0.10). The prevalences of metabolic syndrome and NAFLD showed significant decreasing trends as the level of physical activity and frequency of resistance exercise increased (Supplementary Fig. 2).
There were differences in metabolic profiles and prevalence diabetes mellitus, metabolic syndrome and NAFLD according to the level of physical activity and frequency of resistance exercise in Korean women (Supplementary Table 2). BMI showed a significant increasing trend as the level of physical activity increased, but it was not significant for frequency of resistance exercise. There was no significant difference in waist circumference and ALT according to the level of physical activity or the frequency of resistance exercise. Fasting glucose level and HOMA-IR showed significant decreasing trends only according to frequency of resistance exercise (p for trend <0.01), but they were not significant for level of physical activity. The prevalences of metabolic syndrome and NAFLD showed significant decreasing trends as frequency of resistance exercise increased, but they were not significant for the level of physical activity (Supplementary Fig. 3).
Multiple logistic regression analysis was used to examine the relationship between level of physical activity, frequency of resistance exercise and NAFLD by gender (Table 2). In men, the odds ratio (OR) for NAFLD among physically highly active people was 0.79 (95% confidence interval [CI], 0.63 to 0.98), as compared to those who were physically inactive. Regarding frequency of resistance exercise in men, the OR for NAFLD among men did resistance exercise more than five times a week was 0.73 (95% CI, 0.54 to 0.99) compared with men did resistance exercise less than once a week. In women, the OR for NAFLD among physically highly active people was 0.74 (95% CI, 0.59 to 0.91), as compared to those who were physically inactive. In addition, OR for NAFLD showed a significantly decreasing trend as physical activity increased (Fig. 1). Regarding frequency of resistance exercise in women, the OR for NAFLD among women did resistance exercise more than five times a week was 0.73 (95% CI, 0.45 to 1.17) compared with men did resistance exercise less than once a week (p=0.19). There was a decreasing trend for NAFLD as the frequency of resistance exercise increased in women, but it was not statistically significant (Fig. 2). When sensitivity analysis was performed with HSI score for NAFLD, the OR for NAFLD was significantly decreased among physically highly active men or among men doing resistance exercise for more than 5 days, but not in women (Supplementary Table 3).
Table 2 . Independent Association between Physical Activity, Resistance Exercise and NAFLD among Korean Men and Women.
Variable | Unadjusted OR (95% CI) | p-value | Age and physical activity/ resistance exercise adjusted OR (95% CI)* | p-value | Multivariable adjusted OR (95% CI)† | p-value |
---|---|---|---|---|---|---|
Men | ||||||
Physical activity | ||||||
Inactive | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
Minimally active | 0.86 (0.72–1.02) | 0.08 | 0.91 (0.76–1.08) | 0.28 | 0.90 (0.71–1.14) | 0.39 |
Highly active | 0.73 (0.63–0.86) | <0.01 | 0.80 (0.68–0.94) | <0.01 | 0.79 (0.63–0.98) | 0.03 |
Resistance exercise | ||||||
0–1 day/wk | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
2–4 day/wk | 0.77 (0.64–0.93) | <0.01 | 0.84 (0.70–1.02) | 0.08 | 0.86 (0.67–1.10) | 0.22 |
≥5 day/wk | 0.68 (0.53–0.88) | <0.01 | 0.74 (0.57–0.96) | 0.02 | 0.73 (0.54–0.99) | 0.04 |
Women | ||||||
Physical activity | ||||||
Inactive | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
Minimally active | 0.92 (0.80–1.06) | 0.26 | 0.94 (0.81–1.09) | 0.41 | 0.88 (0.74–1.06) | 0.18 |
Highly active | 0.83 (0.71–0.97) | 0.02 | 0.89 (0.76–1.05) | 0.18 | 0.74 (0.59–0.91) | <0.01 |
Resistance exercise | ||||||
0–1 day/wk | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
2–4 day/wk | 0.76 (0.61–0.95) | 0.01 | 0.94 (0.75–1.17) | 0.56 | 1.02 (0.76–1.38) | 0.89 |
≥5 day/wk | 0.75 (0.53–1.07) | 0.12 | 0.75 (0.53–1.08) | 0.12 | 0.73 (0.45–1.17) | 0.19 |
Logistic regression model for complex sampling survey design were used to examine the association between physical activity, resistance exercise and nonalcoholic fatty liver disease (NAFLD)..
OR, odds ratio; CI, confidence interval..
*Model 1 adjusted for age and physical activity/resistance exercise; †Model 2 adjusted for age, physical activity/resistance exercise, body mass index, waist circumference, smoking status, alcohol intake, diabetes mellitus, hypertension, total cholesterol, high-density lipoprotein cholesterol, income level and education level..
We also investigated the interaction between level of physical activity, frequency of resistance exercise on NAFLD by gender (Table 3). In men, the prevalence rates of NAFLD showed decreasing trends as the frequency of resistance exercise increased, even in those who were physically active (p for interaction <0.001) (Supplementary Table 4). Similarly, there was a decreasing trend for NAFLD as the frequency of resistance exercise increased in women in those who were highly physically active. Both men and women, there were significant multiplicative interactions between level of physical activity, frequency of resistance exercise on NAFLD (p for interaction <0.001 for men, p for interaction <0.001 for women). In other words, these findings show that when resistance exercise as well as aerobic physical activity were performed together, the observed preventive association for NAFLD was much greater than expected preventive association (Fig. 3). However, in the sensitivity analysis which was performed with HSI score for NAFLD, a synergistic effect between aerobic physical activity and resistance exercise was also observed in men, but not in women (Supplementary Table 5).
Table 3 . Interaction of Odds Ratio on NAFLD between Aerobic Physical Activity and Resistance Exercise for the Multivariable Logistic Regression Model.
Resistance exercise | Physical activity level, odds ratio (95% CI) | p for interaction* | ||
---|---|---|---|---|
Inactive | Minimally active | Highly active | ||
Men | <0.001 | |||
0–1 day/wk | 1.00 (reference) | 1.00 (0.77–1.30) | 0.82 (0.63–1.07) | |
2–4 day/wk | 1.09 (0.71–1.68) | 0.73 (0.47–1.14) | 0.67 (0.47–0.96) | |
≥5 day/wk | 0.97 (0.42–2.25) | 0.55 (0.35–0.88) | 0.64 (0.43–0.93) | |
Women | <0.001 | |||
0–1 day/wk | 1.00 (reference) | 0.86 (0.80–1.06) | 0.76 (0.60–0.96) | |
2–4 day/wk | 0.92 (0.52–1.64) | 0.96 (0.59–1.57) | 0.75 (0.47–1.20) | |
≥5 day/wk | 1.42 (0.55–3.63) | 0.85 (0.44–1.66) | 0.28 (0.12–0.65) |
Multiple logistic regression models for complex sampling survey design were used to examine the association between physical activity, resistance exercise and nonalcoholic fatty liver disease (NAFLD) after adjusting for age, body mass index, waist circumference, smoking status, alcohol intake, diabetes mellitus, hypertension, total cholesterol, high-density lipoprotein cholesterol, income level and education level..
CI, confidence interval..
*p for interaction was tested as a likelihood ratio test between the full model with interaction and the null model without interaction..
This study examined the independent and synergistic association of frequency of resistance exercise and aerobic physical activity with the NAFLD for a total of 14,974 Korean men and women. Our findings showed the level of aerobic exercise was associated with the decreased odds of NAFLD, and the frequency of resistance exercise was also associated with the lower odds of NAFLD independently of the aerobic exercise level especially in Korean men. In addition, our study finding also showed that aerobic physical activity and resistance exercise had an interactively preventive association for NAFLD both in men and women.
Although there were cumulating evidence that increase of physical activity could reduce the risk of NAFLD, there were still lack of evidence on whether resistance exercise itself is independently associated with a reduced risk of NAFLD in a large sample of general population.6,10,22 Most previous studies have focused on whether resistance exercise could reduce liver fat, liver enzyme levels (ALT and AST), and insulin resistance in NAFLD patients.6,10,23 However, these studies have several limitations that the number of study participants was relatively small, and the preventive or synergistic association of resistance exercise with NAFLD in the general population was not investigated. Although these studies may indirectly show that resistance exercise could prevent or improve NAFLD, previous studies did not provide interactively how resistance exercise could prevent NAFLD depending on the level of aerobic physical activity. On the contrary, our study findings showed that the probability of NAFLD has decreased as aerobic physical activity level has increased even when resistance exercise was performed more than 5 days per week, or as the frequency of resistance exercise has increased even when aerobic physical activity was sufficiently active. When subgroup analysis was performed according to age group, the preventive association of aerobic physical activity with NAFLD was reduced among the old age group (≥55 years old). In addition, the synergistic association of aerobic physical activity and resistance exercise showed a reduced synergistic association among elderly women (≥55 years old) than in elderly men. In sensitivity analysis using HSI score for NAFLD, the preventive association of aerobic physical activity and resistance exercise on NAFLD was not significant in women than in men, and the synergistic association was not clear in women. It may be because the number of women doing resistance exercise was much smaller compared to men, and the efficiency of exercise itself in women or older men is lower than that of young men.24
It has been reported that combining aerobic exercise plus resistance exercise together is more effective than aerobic exercise or resistance exercise alone in type 2 diabetes, which is characterized by insulin resistance, similar to NAFLD.25 In addition, considering that sarcopenia is a risk factor for NAFLD independent of obesity or insulin resistance, it could be reasonable inference that resistance exercise that increases muscle strength could be a preventive factor of NAFLD independently of aerobic exercise.26,27 Indeed, resistance exercise has prevented the decrease of muscle strength and improved physical function in the meta-analysis of 25 randomized controlled trials.28 Furthermore, it has been previously reported that combining resistance exercise with aerobic physical activity was more effective in improving fat mass, metabolic profiles, and inflammatory state than doing aerobic exercise alone in obese adults and children.29,30 Considering these previous studies and our study findings, it is reasonable to infer that combining resistance exercise with aerobic physical activity might be more effective for preventing NAFLD than aerobic physical activity alone.
Our study has a number of strengths. First, we used the KNHANES database, which was representative the Korean population. Therefore, our study could present reliable study findings with sufficient statistical power. Our study provides evidence that the combination of aerobic physical activity and strength training would be better for preventing NAFLD than performing only one type of exercise. It could be helpful when physicians and policy makers provide exercise recommendations for NAFLD. Of course, additional verification is necessary to whether these results are externally applicable to non-Asian population. Furthermore, we excluded high-risk drinkers and people with hepatitis B or C, which could affect the risk of severity NAFLD. These strict exclusion criteria can ensure the homogeneity of population with NAFLD, thereby increasing the internal validity of our study.
Nonetheless, the limitation of the study should be acknowledged. Although we used scoring system for NAFLD, the standard method to diagnose the NAFLD is imaging study such as CT and liver biopsy. Therefore, there may be a bias to define the NAFLD. However, our study is based on the national representative sampling survey and it is impossible to assess the participants by using abdominal ultrasound or CT. Therefore, it was impossible to use abdominal ultrasound or CT to distinguish the NAFLD among general population. Second, it is impossible to examine the temporal relationship between aerobic physical activity, resistance exercise and NAFLD, because our study design is based on the cross-sectional study. Therefore, a future follow-up study is necessary to examine whether the combining association of aerobic physical activity and resistance exercise change longitudinally with synergistic interaction. Third, we could not know the history of drug use or dietary intake, therefore we could not exclude the effects of medication or dietary supplement. Finally, although the KNHANES surveyed physical activity using standardized questionnaires by trained investigators, there may be still problem about accuracy of physical activity measurement, because the information on exercise was obtained through the questionnaire. Especially, resistance exercise was surveyed only in frequency, not intensity of exercise, so it may not reflect the total amount of resistance exercise. In addition, physical activity may not have been maintained at a constant level over a long period of time.
In conclusion, to our best knowledge, the present study is the first investigation to suggest both resistance exercise and aerobic physical activity could exert a synergistic preventive association for NAFLD in a large number of general Korean population. These synergistic associations between resistance exercise and aerobic physical activity on NAFLD were consistent in both men and women. Future cohort study or clinical trial may be necessary to confirm our study findings.
Supplementary materials can be accessed at https://doi.org/10.5009/gnl220345.
No potential conflict of interest relevant to this article was reported.
Study concept and design: J.M.C., C.M.O. Data acquisition: C.M.O. Data analysis and interpretation: H.J.Y., C.M.O., Drafting of the manuscript: H.J.Y. Critical revision of the manuscript for important intellectual content: Y.P.H., T.Y.Y., J.H.R. Statistical analysis: H.J.Y., C.M.O. Administrative, technical, or material support; study supervision: J.M.C., C.M.O. Approval of final manuscript: all authors.
Table 1 Baseline Characteristics of the Study Participants
Characteristic | Overall (n=14,977) | Gender | p-value* | |
---|---|---|---|---|
Men (n=5,712) | Women (n=9,265) | |||
Continuous variable | ||||
Age, yr | 44.2±0.2 | 43.1±0.3 | 45.0±0.3 | <0.01 |
METs, min/wk | 2,962±54 | 3,472±82 | 2,534±58 | <0.01 |
ALT, U/L | 21.2±0.2 | 26.3±0.3 | 17.0±0.1 | <0.01 |
AST, U/L | 21.2±0.1 | 23.3±0.2 | 19.5±0.1 | <0.01 |
GGT, U/L (n=4,411) | 30.1±0.8 | 41.7±1.5 | 20.4±0.5 | <0.01 |
Systolic blood pressure, mm Hg | 116.6±0.2 | 119.3±0.3 | 114.3±0.3 | <0.01 |
Diastolic blood pressure, mm Hg | 76.2±0.2 | 79.2±0.2 | 73.6±0.2 | <0.01 |
Body mass index, kg/m2 | 23.5±0.04 | 23.9±0.05 | 23.2±0.05 | <0.01 |
Waist circumference, cm | 80.4±0.1 | 83.6±0.2 | 77.7±0.2 | <0.01 |
Fasting glucose level, mg/dL | 96.2±0.2 | 97.5±0.4 | 95.0±0.3 | <0.01 |
Insulin, μU/mL | 10.1±0.1 | 10.1±0.1 | 10.1±0.1 | 0.94 |
Total cholesterol, mg/dL | 186.2±0.4 | 185.9±0.6 | 186.4±0.5 | 0.45 |
HDL cholesterol, mg/dL | 48.1±0.1 | 45.0±0.2 | 50.6±0.2 | <0.01 |
Triglyceride, mg/dL | 126.5±1.1 | 146.7±1.8 | 109.9±1.0 | <0.01 |
Alcohol intake, g/day | 3.8±0.06 | 6.3±0.10 | 1.6±0.03 | <0.01 |
Categorical variable | ||||
Physical activity | <0.01 | |||
Inactive | 5,746 (37.6) | 1,928 (33.4) | 3,818 (41.1) | |
Minimally active | 4,844 (32.3) | 1,782 (31.3) | 3,062 (33.1) | |
Highly active | 4,387 (30.1) | 2,002 (35.3) | 2,385 (25.8) | |
Resistance exercise | <0.01 | |||
0–1 day/wk | 12,125 (78.5) | 3,979 (67.9) | 8,146 (87.2) | |
2–4 day/wk | 1,928 (14.9) | 1,121 (21.6) | 807 (9.4) | |
≥5 day/wk | 924 (6.6) | 612 (10.5) | 312 (3.4) | |
Smoking history | <0.01 | |||
Never-smoker | 9,887 (61.2) | 1,414 (26.1) | 8,473 (90.2) | |
Past smoker | 2,313 (15.8) | 1,976 (29.9) | 337 (4.1) | |
Current smoker | 2,777 (23.1) | 2,322 (44.1) | 455 (5.7) | |
Hypertension (n=14,936) | <0.01 | |||
Yes | 4,239 (24.4) | 1,844 (27.6) | 2,395 (21.7) | |
No | 10,697 (75.6) | 3,852 (72.4) | 6,845 (78.3) | |
Diabetes mellitus (n=14,580) | 0.08 | |||
Yes | 1,305 (7.7) | 564 (8.2) | 741 (7.3) | |
No | 13,275 (92.3) | 4,965 (91.8) | 8,310 (92.7) | |
Metabolic syndrome | <0.01 | |||
Yes | 3,638 (22.0) | 1,518 (24.4) | 2,120 (19.9) | |
No | 11,339 (78.0) | 4,194 (75.6) | 7,145 (80.1) | |
NAFLD by NLFS score (n=14,580) | 3,550 (23.8) | 1,670 (29.7) | 1,880 (19.0) | <0.01 |
Continuous variables were expressed as means±standard errors and categorical variables were expressed as number (%) by using the sampling weights to reflect complex Korea National Health and Nutrition Examination Survey sampling design.
METs, metabolic equivalents; ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, gamma glutamyl transferase; HDL, high-density lipoprotein; NAFLD, nonalcoholic fatty liver disease; NLFS, nonalcoholic fatty liver disease liver fat score.
*Linear regressions for complex sampling survey design were used to test trend for continuous variables and Rao-Scott chi-square for complex sampling survey design were used to test differences for categorical variables.
Table 2 Independent Association between Physical Activity, Resistance Exercise and NAFLD among Korean Men and Women
Variable | Unadjusted OR (95% CI) | p-value | Age and physical activity/ resistance exercise adjusted OR (95% CI)* | p-value | Multivariable adjusted OR (95% CI)† | p-value |
---|---|---|---|---|---|---|
Men | ||||||
Physical activity | ||||||
Inactive | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
Minimally active | 0.86 (0.72–1.02) | 0.08 | 0.91 (0.76–1.08) | 0.28 | 0.90 (0.71–1.14) | 0.39 |
Highly active | 0.73 (0.63–0.86) | <0.01 | 0.80 (0.68–0.94) | <0.01 | 0.79 (0.63–0.98) | 0.03 |
Resistance exercise | ||||||
0–1 day/wk | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
2–4 day/wk | 0.77 (0.64–0.93) | <0.01 | 0.84 (0.70–1.02) | 0.08 | 0.86 (0.67–1.10) | 0.22 |
≥5 day/wk | 0.68 (0.53–0.88) | <0.01 | 0.74 (0.57–0.96) | 0.02 | 0.73 (0.54–0.99) | 0.04 |
Women | ||||||
Physical activity | ||||||
Inactive | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
Minimally active | 0.92 (0.80–1.06) | 0.26 | 0.94 (0.81–1.09) | 0.41 | 0.88 (0.74–1.06) | 0.18 |
Highly active | 0.83 (0.71–0.97) | 0.02 | 0.89 (0.76–1.05) | 0.18 | 0.74 (0.59–0.91) | <0.01 |
Resistance exercise | ||||||
0–1 day/wk | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
2–4 day/wk | 0.76 (0.61–0.95) | 0.01 | 0.94 (0.75–1.17) | 0.56 | 1.02 (0.76–1.38) | 0.89 |
≥5 day/wk | 0.75 (0.53–1.07) | 0.12 | 0.75 (0.53–1.08) | 0.12 | 0.73 (0.45–1.17) | 0.19 |
Logistic regression model for complex sampling survey design were used to examine the association between physical activity, resistance exercise and nonalcoholic fatty liver disease (NAFLD).
OR, odds ratio; CI, confidence interval.
*Model 1 adjusted for age and physical activity/resistance exercise; †Model 2 adjusted for age, physical activity/resistance exercise, body mass index, waist circumference, smoking status, alcohol intake, diabetes mellitus, hypertension, total cholesterol, high-density lipoprotein cholesterol, income level and education level.
Table 3 Interaction of Odds Ratio on NAFLD between Aerobic Physical Activity and Resistance Exercise for the Multivariable Logistic Regression Model
Resistance exercise | Physical activity level, odds ratio (95% CI) | p for interaction* | ||
---|---|---|---|---|
Inactive | Minimally active | Highly active | ||
Men | <0.001 | |||
0–1 day/wk | 1.00 (reference) | 1.00 (0.77–1.30) | 0.82 (0.63–1.07) | |
2–4 day/wk | 1.09 (0.71–1.68) | 0.73 (0.47–1.14) | 0.67 (0.47–0.96) | |
≥5 day/wk | 0.97 (0.42–2.25) | 0.55 (0.35–0.88) | 0.64 (0.43–0.93) | |
Women | <0.001 | |||
0–1 day/wk | 1.00 (reference) | 0.86 (0.80–1.06) | 0.76 (0.60–0.96) | |
2–4 day/wk | 0.92 (0.52–1.64) | 0.96 (0.59–1.57) | 0.75 (0.47–1.20) | |
≥5 day/wk | 1.42 (0.55–3.63) | 0.85 (0.44–1.66) | 0.28 (0.12–0.65) |
Multiple logistic regression models for complex sampling survey design were used to examine the association between physical activity, resistance exercise and nonalcoholic fatty liver disease (NAFLD) after adjusting for age, body mass index, waist circumference, smoking status, alcohol intake, diabetes mellitus, hypertension, total cholesterol, high-density lipoprotein cholesterol, income level and education level.
CI, confidence interval.
*p for interaction was tested as a likelihood ratio test between the full model with interaction and the null model without interaction.