Abstract
Objectives.
Non-alcoholic fatty liver disease (NAFLD) is a highly prevalent liver disease, with the highest prevalence observed in the U.S. among Hispanic/Latino adults. While physical activity and dietary behaviors have established protective associations with NAFLD and its severity, these associations have not been well-characterized in Hispanic/Latino adults. The purpose of this study was to assess the association of lifestyle behaviors with NAFLD and advanced fibrosis in US Hispanic/Latino adults.
Design.
We selected all Hispanic/Latino adults from the 2017–2018 National Health and Nutrition Examination Survey (NHANES). NAFLD was defined as CAP ≥285 dB/m, and advanced fibrosis as liver stiffness measurements ≥8.6 kPa. Multivariate-adjusted logistic regression models assessed associations of physical activity and sedentary behavior (Global Physical Activity Questionnaire), as well as diet quality (Healthy Eating Index [HEI]-2015) and total energy intake (24-hour recall) with NAFLD and advanced fibrosis.
Results.
In Hispanic/Latino adults, the overall prevalence of NAFLD was 41.5%, while the prevalence of advanced fibrosis among those with NAFLD was 17.2%. We found that higher levels of physical activity and high diet quality were associated with lower risk of NAFLD. Compared to those reporting on average 0 metabolic equivalent (MET) hours/week of physical activity, participants reporting high levels of physical activity (≥32 MET hours/week) had 40% lower risk of NAFLD (Adjusted OR=0.60, 95%CI 0.38, 0.93). High diet quality (HEI-2015) was associated with a 30% lower risk of NAFLD (Adjusted OR=0.70, 95% CI 0.51, 0.97) and 72% lower risk of advanced fibrosis (Adjusted OR=0.28, 95% CI 0.12, 0.66), as compared to those with low diet quality.
Conclusions.
In this population-based study, high levels of physical activity and diet quality were associated with lower risk of NAFLD in Hispanic/Latino adults. Public health and medical professionals need to concentrate efforts on lifestyle behavior change in Hispanic/Latino adults who are at high risk for serious liver disease.
Keywords: Hispanic, Latino, non-alcoholic fatty liver disease, fatty liver, fibrosis, lifestyle behaviors, physical activity, sedentary behavior, diet, diet quality, Health Eating Index
Introduction
Although the incidence and mortality rates for many cancers have been decreasing over the past few decades, those for liver cancer, particularly hepatocellular carcinoma (HCC), continue to rise (White et al. 2017). Metabolic disorders (e.g., obesity, diabetes, metabolic syndrome) and non-alcoholic fatty liver disease (NAFLD) in particular are major contributors to the upward trend in HCC incidence and mortality (Huang, El-Serag, and Loomba 2020). NAFLD is also the most rapidly increasing indication for liver transplantation (Rinella 2015) and is associated with an array of adverse health outcomes, including cardiovascular disease and extrahepatic cancers (Adams et al. 2017). Intensive efforts to prevent and treat NAFLD are needed in the U.S. to mitigate a future health and healthcare crisis.
Hispanic/Latino adults have the highest rate of HCC in the U.S., double the rate of their non-Hispanic counterparts (El-Serag et al. 2021), and the highest prevalence of NAFLD (Kim et al. 2021; Zhang et al. 2021). Hispanic/Latino adults have been shown to be genetically predisposed to NAFLD (Romeo et al. 2008). This genetic predisposition and the alarmingly high prevalence call for concerted efforts from the public health and medical community to tackle this growing health disparity.
Excess body fat is the central mechanism underlying NAFLD development and pathogenesis (Friedman et al. 2018). Thus, weight loss, resulting from changes in diet and physical activity, is the cornerstone of therapy for NAFLD (Chalasani et al. 2018). Beyond treatment, research to date has also begun to further explore the association of specific weight-related preventive or risk behaviors associated with NAFLD and fibrosis. For example, in recent population-based studies in the U.S., high diet quality, more physical activity and/or lower levels of sedentary behavior were associated with a reduced risk of NAFLD (Yoo et al. 2020; Kim et al. 2020; Park et al. 2020; Heredia et al. 2022). Furthermore, previous studies show these associations extend to a reduced risk of fibrosis in those with NAFLD (Park et al. 2020; Kim et al. 2020). However, it is yet unclear how these associations may differ by racial/ethnic group.
Given the differences between various racial/ethnic groups in the risk of development and potentially also in the progression of NAFLD (Romeo et al. 2008), it is important to examine the association of lifestyle behaviors with NAFLD by racial/ethnic group, in particular among the group with the highest rates of NAFLD in the U.S.: Hispanic/Latinos. Although previous studies have evaluated lifestyle characteristics among persons with a reported history of liver disease or who had abnormal liver enzymes, to date, there has been a paucity of evidence in Hispanic/Latino samples linking physical activity and diet with NAFLD diagnosed with imaging. Thus, the purpose of this study was to assess the association of various lifestyle behaviors, including physical activity, sedentary behavior, diet quality, and total energy intake with NAFLD and advanced fibrosis in those with NAFLD (as assessed by transient elastography) in Hispanic/Latino adults in the U.S.
Methods
Design & Sample
We used data from the 2017–2018 National Health and Nutrition Examination Survey (NHANES), a nationally representative survey of the non-institutionalized, civilian U.S. population. It utilizes a multistage probability sampling design. Additional details are readily available on the NHANES website (Centers for Disease Control & Prevention). From the 5,265 participants who complete the survey and medical examination during the 2017–2018 cycle, we included 1038 who were Hispanic/Latino and had complete vibration-controlled transient elastography (VCTE; defined below), and excluded those with hepatitis B surface antigen positivity (n=3), hepatitis C antibody positivity (n=11), significant alcohol consumption (>21 drinks/week in men and >14 drinks/week in women on average; n=90), or missing diet and/or physical activity data (n=214). The final analytic sample included 720 Hispanic/Latino adult participants. Participants included in the present study provided written consent. The National Center for Health Statistics research ethics review board approved NHANES and this study was approved by the Institutional Review Board (IRB) of The University of Texas MD Anderson Cancer Center.
Data Collection and Measures
NHANES questionnaires are interviewer-administered and collect sociodemographic and behavioral characteristics, including age, sex, race/ethnicity, education, household income, smoking, and alcohol use (Johnson et al. 2013). After excluding significant alcohol consumers (Chalasani et al. 2018), we created a categorical average daily past year alcohol use variable. The examination component is administered by trained medical personnel, and consists of medical, dental, and physiological measurements, including measured weight and height, which we used to calculate Body Mass Index (BMI). NHANES assessments also include laboratory tests, from which we ascertained history of type 2 diabetes and presence of metabolic syndrome. Metabolic syndrome was determined by the presence of three of the five Adult Treatment Panel (ATP) III criteria (Expert Panel on Detection 2001), including abdominal obesity (waist circumference >102 cm men, >88 cm in women); high blood pressure(bp)/hypertension (systolic bp≥130, diastolic bp≥ 85); high triglycerides (≥150 mg/dl), low HDL cholesterol (<40 mg/dl men, <50 mg/dl women); and high fasting glucose (≥110 mg/dl).
NAFLD and advanced fibrosis.
NAFLD and advanced fibrosis status were determined using data from FibroScan, which uses VCTE with controlled attenuation to derive liver stiffness measurements (LSM) and the controlled attenuation parameter (CAP) value. CAP values range from 100–400dB/m, with higher values indicating higher amounts of fat in the liver. LSM range from 1.5–75kPa, with higher values indicating more advanced fibrosis. NHANES technicians completed a 2-day training program with and were certified by the equipment manufacturer after completing 3 satisfactory exams (Echosens™ North America). For all examinations, the medium (M) probe was applied first; however, the operator switched to the extra-large (XL) probe if needed based on the recommendations of the device and the manufacturer’s instructions. Additional details can be found in the NHANES Liver Ultrasound Transient Elastography Procedures Manual (Centers for Disease Control & Prevention 2020). Exams were considered complete if participants fasted ≥3 hours prior to the exam, there were ≥10 complete LSM, and liver stiffness IQR/median <30% (Centers for Disease Control & Prevention). A total of 91 Hispanic/Latino individuals were excluded due to incomplete exam. We defined NAFLD as CAP score ≥285dB/m in the absence of excessive alcohol intake and NAFLD with suspected advanced fibrosis (F3–4) as CAP score ≥285dB/m and LSM ≥ 8.6 kPa.
Lifestyle behaviors.
Physical activity and sedentary behavior were collected with the Global Physical Activity Questionnaire (Armstrong and Bull 2006). Participants provided information on sitting time and typical physical activity in a week in the various domains and by intensity. Sedentary behavior was defined as total time spent sitting, in hours/day. For physical activity, we computed the metabolic equivalent (MET) as recommended by NHANES and summed activity across domains. Participants completed a 24-hour recall of all food and drink consumed during the 24-hour period prior to the interview (midnight to midnight), as well as a second follow-up interview 3–10 days later. When both recalls were available, we averaged the two. This data was processed via the United States Department of Agriculture Food Patterns Equivalents Database. From this database, we used total energy intake and calculated the Health Eating Index (HEI)-2015 (Reedy et al. 2018), a measure of overall diet quality which includes 9 components to eat enough of (Total Fruits, Whole Fruits, Total Vegetables, Greens and Beans, Whole Grains, Dairy, Total Protein Foods, Seafood and Plant Proteins, and Fatty Acids) and four components to limit (Refined Grains, Sodium, Added Sugars, and Saturated Fats). HEI-2015 is based on alignment with the Dietary Guidelines for Americans 2015–2020 (DeSalvo, Olson, and Casavale 2016). HEI-2015 has a possible range of zero to 100, with higher scores indicating higher (healthier) diet quality.
Statistical analyses
We compared characteristics of participants with and without NAFLD and advanced fibrosis using Chi-Square test or Fisher’s exact test for categorical variables. For all lifestyle behaviors, we created tertiles based on the distribution in those without NAFLD. However, for physical activity, we created three categories such that the first category included participants with 0 MET hours/week, and the second and third category cut-off was the median MET hours/week among participants with MET>0 (0, >0-<32 and ≥32). We used unadjusted and multivariable logistic regression models to estimate odds ratios (OR) and associated 95% confidence intervals (95% CI) for associations with the presence of NAFLD or advanced fibrosis. We created preliminary models that were 1) unadjusted and 2) were adjusted for sex and age. We then added covariates significantly associated (p<0.05) with the outcomes in univariate analyses. Thus, final multivariate-adjusted logistic regression models for both physical activity and sedentary behavior models were adjusted for age, sex, metabolic syndrome, and total energy. The multivariate-adjusted logistic regression model for HEI-2015 was adjusted for age, sex, metabolic syndrome, total energy, physical activity, alcohol use and for total energy intake was adjusted for age, sex, metabolic syndrome, HEI-2015, physical activity, alcohol use. Given the overlap in definitions and thus multicollinearity of BMI, diabetes and metabolic syndrome, we controlled for only metabolic syndrome. We conducted supplementary analyses that adjusted for each component of the metabolic syndrome.
Lastly, we examined the potential joint effect of physical activity and sedentary behavior, as well as analyses of the association with physical activity stratified by sedentary behavior to examine for evidence of effect modification. We did the same for HEI-2015 and total energy intake.
Weighted analyses were used to make the sample representative of the U.S. non-institutionalized population (Chen et al. 2018). We used SAS 9.4 (SAS Institute INC, Cary, NC) for all analyses and two-sided p<0.05 indicates statistical significance.
Results
The overall prevalence of NAFLD was 41.5% (95% CI 37.1, 46.0), while the prevalence of suspected advanced fibrosis (F3–4) among those with NAFLD was 17.2% (95% CI 11.0, 23.4). As compared to those without NAFLD, those with NAFLD were more likely to be ≥40 years of age and male. As expected, those with NAFLD were more likely to be obese, and have pre-diabetes, diabetes, and metabolic syndrome than those without NAFLD. Among those with NAFLD, a significantly higher proportion of those with advanced fibrosis had Class 2 or 3 obesity, type 2 diabetes and metabolic syndrome than those without advanced fibrosis (Table 1). There were no significant differences in length of time in the U.S. or language spoken at home by NAFLD or advanced fibrosis status.
Table 1.
Hispanic/Latino characteristics for total sample and by disease status
Variables | Total n=720 | NAFLD (CAP≥ 285 dB/m) | Advanced fibrosis F3–4 (LSM≥ 8.6 kPa) | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Yes n=319 | No= 401 | Yes n=54 | No n=265 | |||||||||
n | Weighted % ± SE | n | Weighted % ± SE | n | Weighted % ± SE | n | Weighted % ± SE | n | Weighted % ± SE | |||
Age | ||||||||||||
20–39 | 255 | 50.8 ± 2.9 | 97 | 44.4 ± 3.4 | 158 | 55.3 ± 4.1 | 0.04 | 12 | 36.6 ± 10.0 | 85 | 46.0 ± 3.5 | 0.49 |
40–59 | 245 | 34.7 ± 2.3 | 112 | 37.9 ± 2.4 | 133 | 32.4 ± 3.6 | 19 | 40.0 ± 9.5 | 93 | 37.5 ± 2.1 | ||
60–89 | 220 | 14.6 ± 2.0 | 110 | 17.7 ± 3.0 | 110 | 12.3 ± 2.0 | 23 | 23.4 ± 5.9 | 87 | 16.5 ± 3.1 | ||
Sex | ||||||||||||
Male | 335 | 50.3 ± 2.1 | 172 | 58.9 ± 3.5 | 163 | 44.1 ± 2.5 | <0.001 | 32 | 64.8 ± 6.2 | 140 | 57.7 ± 3.6 | 0.27 |
Female | 385 | 49.7 ± 2.1 | 147 | 41.1 ± 3.5 | 238 | 55.9 ± 2.5 | 22 | 35.2 ± 6.2 | 125 | 42.3 ± 3.6 | ||
Education | ||||||||||||
<12th grade | 294 | 27.4 ± 2.4 | 125 | 27.0 ± 2.8 | 169 | 27.7 ± 3.1 | 0.84 | 21 | 30.6 ± 6.7 | 104 | 26.2 ± 2.8 | 0.51 |
High school + | 425 | 72.6 ± 2.4 | 193 | 73.0 ± 2.8 | 232 | 72.3 ± 3.1 | 3 | 69.4 ± 6.7 | 160 | 73.8 ± 2.8 | ||
Household income | ||||||||||||
<$55,000 | 370 | 51.4 ± 2.7 | 167 | 54.4 ± 4.3 | 203 | 49.3 ± 3.7 | 0.39 | 30 | 60.4 ± 7.7 | 137 | 53.2 ± 4.2 | 0.29 |
≥$55,000 | 273 | 48.6 ± 2.7 | 116 | 45.6 ± 4.3 | 157 | 50.7 ± 3.7 | 18 | 39.6 ± 7.7 | 98 | 46.8 ± 4.2 | ||
Marital Status | ||||||||||||
Never Married | 124 | 21.7 ± 2.2 | 52 | 19.7 ± 2.3 | 72 | 23.2 ± 3.4 | 0.54 | 6 | 12.0 ± 4.1 | 46 | 21.3 ± 2.8 | 0.16 |
Married or living with a partner | 463 | 63.4 ± 2.6 | 207 | 63.9 ± 2.8 | 256 | 63.0 ± 3.7 | 35 | 65.1 ± 6.9 | 172 | 63.6 ± 2.8 | ||
Widowed, divorced or separated | 133 | 14.9 ± 1.2 | 60 | 16.4 ± 2.2 | 73 | 13.8 ± 1.6 | 13 | 22.9 ± 5.9 | 47 | 15.1 ± 2.3 | ||
Length of time in the U.S. | ||||||||||||
Born in the U.S. | 253 | 44.3 ± 2.7 | 102 | 39.0 ± 3.9 | 151 | 48.1 ± 4.7 | 0.19 | 20 | 49.6 ± 7.4 | 82 | 36.9 ± 4.7 | 0.36 |
≥20 years in the U.S. | 249 | 28.9 ± 2.6 | 80 | 28.9 ± 3.4 | 107 | 28.8 ± 2.8 | 13 | 26.6 ± 7.5 | 67 | 29.4 ± 3.5 | ||
<20 years in the U.S. | 187 | 26.9 ± 2.1 | 127 | 32.1 ± 4.4 | 122 | 23.1 ± 3.4 | 18 | 23.9 ± 6.8 | 109 | 33.8 ± 4.7 | ||
Language spoken at home | ||||||||||||
Primarily Spanish | 385 | 47.7 ± 2.1 | 177 | 51.6 ± 3.1 | 208 | 44.9 ± 3.5 | 0.07 | 26 | 45.5 ± 6.8 | 151 | 52.9 ± 4.0 | 0.65 |
Both Spanish and English | 92 | 13.1 ± 1.6 | 45 | 14.7 ± 2.7 | 47 | 11.9 ± 1.6 | 9 | 18.9 ± 8.0 | 36 | 13.9 ± 2.7 | ||
Primarily English | 232 | 39.2 ± 2.8 | 92 | 33.7 ± 3.0 | 140 | 43.2 ± 3.4 | 18 | 35.6 ± 5.7 | 74 | 33.3 ± 3.6 | ||
Smoking | ||||||||||||
Nonsmoker | 477 | 86.0 ± 2.7 | 198 | 84.8 ± 4.6 | 279 | 86.8 ± 2.3 | 0.55 | 31 | 84.2 ± 5.6 | 167 | 85.0 ± 5.1 | 0.76 |
Former smoker | 24 | 5.5 ± 1.3 | 14 | 7.0 ± 2.8 | 10 | 4.4 ± 1.1 | 2 | 5.3 ± 3.6 | 12 | 7.4 ± 3.4 | ||
Current smoker | 49 | 8.5 ± 1.6 | 20 | 8.1 ± 2.3 | 29 | 8.8 ± 1.8 | 5 | 10.6 ± 4.9 | 15 | 7.7 ± 2.1 | ||
Alcohol Use | ||||||||||||
No past year use | 195 | 20.9 ± 1.1 | 84 | 21.1 ± 1.7 | 111 | 20.8 ± 2.1 | 0.97 | 18 | 28.4 ± 7.3 | 66 | 19.6 ± 1.1 | 0.22 |
Light-to-Moderate | 223 | 30.6 ± 2.4 | 98 | 31.0 ± 3.4 | 125 | 30.4 ± 3.0 | 18 | 34.6 ± 6.9 | 80 | 30.3 ± 3.8 | ||
Heavy Use | 302 | 48.5 ± 3.0 | 137 | 47.9 ± 3.3 | 165 | 48.9 ± 3.2 | 18 | 36.9 ± 7.5 | 119 | 50.2 ± 3.5 | ||
Body Mass Index | ||||||||||||
Underweight/Normal (<25kg/m2) | 116 | 17.7 ± 1.3 | 9 | 2.1 ± 0.6 | 107 | 28.8 ± 2.7 | <0.0001 | 1 | 1.2 ± 0.2 | 8 | 2.3 ± 0.7 | <0.0001 |
Overweight (25–29.9kg/m2) | 277 | 37.7 ± 1.7 | 100 | 29.8 ± 3.5 | 117 | 43.4 ± 2.2 | 7 | 15.4 ± 7.4 | 93 | 32.8 ± 3.0 | ||
Class 1 obesity (30–34.9 kg/m2) | 183 | 25.0 ± 1.7 | 106 | 34.5 ± 2.6 | 77 | 18.2 ± 2.5 | 14 | 20.3 ± 5.5 | 92 | 37.4 ± 2.4 | ||
Class 2 obesity (35–39.9 kg/m2) | 92 | 12.5 ± 1.8 | 64 | 19.2 ± 2.8 | 28 | 7.8 ± 2.1 | 17 | 34.6 ± 9.7 | 47 | 16.0 ± 2.2 | ||
Class 3 obesity (>40 kg/m2) | 45 | 7.1 ± 1.4 | 38 | 14.4 ± 2.6 | 7 | 1.9 ± 1.0 | 15 | 28.5 ± 7.0 | 23 | 11.5 ± 3.0 | ||
Diabetes | ||||||||||||
Normal | 400 | 67.3 ± 1.7 | 134 | 52.2 ± 3.7 | 266 | 78.1 ± 3.2 | <0.0001 | 14 | 37.0 ± 8.0 | 120 | 55.3 ± 3.4 | <0.0001 |
Pre-diabetes | 162 | 19.9 ± 1.8 | 89 | 26.9 ± 4.0 | 73 | 14.9 ± 2.5 | 9 | 20.3 ± 7.3 | 80 | 28.2 ± 4.0 | ||
Diabetes | 138 | 12.8 ± 1.3 | 89 | 20.9 ± 2.1 | 49 | 6.9 ± 1.4 | 29 | 42.7 ± 5.7 | 60 | 16.5 ± 1.9 | ||
Metabolic Syndrome | ||||||||||||
Yes | 230 | 28.6 ± 2.6 | 148 | 47.1 ± 4.3 | 82 | 15.9 ± 2.1 | <0.001 | 30 | 67.7 ± 7.9 | 118 | 42.8 ± 4.1 | <0.001 |
No | 427 | 71.4 ± 2.6 | 141 | 52.9 ± 4.3 | 286 | 84.1 ± 2.1 | 17 | 32.3 ± 7.9 | 124 | 57.2 ± 4.1 | ||
Physical activity | ||||||||||||
T1 (0 MET hours/week) | 223 | 25.4 ± 2.0 | 108 | 28.4 ± 2.4 | 115 | 23.3 ± 2.5 | 0.02 | 22 | 34.7 ± 7.8 | 114 | 37.3 ± 3.7 | 0.70 |
T2 (>0 to <32 MET hours/week) | 158 | 20.6 ± 1.9 | 74 | 22.9 ± 2.0 | 84 | 19.0 ± 2.5 | 20 | 34.2 ± 7.6 | 65 | 27.2 ± 3.1 | ||
T3 (≥32 MET hours/week) | 339 | 54.0 ± 2.6 | 137 | 48.7 ± 3.1 | 202 | 57.7 ± 2.9 | 12 | 31.1 ± 8.9 | 86 | 35.5 ± 3.5 | ||
Sedentary Behavior | ||||||||||||
T1 (≤2 hours/day) | 223 | 28.5 ± 2.2 | 84 | 24.3 ± 3.6 | 139 | 31.6 ± 2.9 | 0.20 | 12 | 22.4 ± 6.9 | 72 | 24.7 ± 3.8 | 0.72 |
T2 (>2 to <5 hours/day) | 232 | 30.6 ± 1.6 | 120 | 34.6 ± 2.1 | 112 | 27.8 ± 2.6 | 19 | 30.3 ± 8.2 | 101 | 35.6 ± 2.2 | ||
T3 (≥ 5 hours/day) | 265 | 40.8 ± 3.0 | 115 | 41.1 ± 3.7 | 150 | 40.7 ± 4.5 | 23 | 47.3 ± 9.3 | 92 | 39.8 ± 4.3 | ||
Healthy Eating Index (HEI) 2015 | ||||||||||||
T1 (≤48.3) | 200 | 36.6 ± 3.8 | 101 | 42.3 ± 5.0 | 99 | 32.6 ± 3.8 | 0.07 | 16 | 52.0 ± 11.6 | 75 | 36.1 ± 5.1 | 0.05 |
T2 (>48.3 to <61.3) | 195 | 33.5 ± 2.9 | 76 | 29.0 ± 3.6 | 119 | 36.7 ± 4.1 | 14 | 33.6 ± 10.2 | 70 | 32.0 ± 3.2 | ||
T3 (≥61.3) | 215 | 29.9 ± 2.9 | 92 | 28.7 ± 3.2 | 123 | 30.8 ± 3.6 | 13 | 14.4 ± 3.3 | 81 | 31.9 ± 3.8 | ||
Total energy intake | ||||||||||||
T1 (≤1575 kcal) | 196 | 30.8 ± 3.2 | 80 | 29.1 ± 4.3 | 116 | 31.9 ± 3.8 | 0.35 | 12 | 24.7 ± 8.6 | 57 | 24.6 ± 4.4 | 0.72 |
T2 (>1575 to <2154 kcal) | 188 | 29.4 ± 1.9 | 81 | 26.0 ± 3.1 | 107 | 31.8 ± 4.2 | 11 | 21.3 ± 9.4 | 73 | 29.1 ± 3.4 | ||
T3 (≥2154 kcal) | 226 | 39.8 ± 3.1 | 108 | 44.8 ± 4.1 | 118 | 36.3 ± 4.0 | 20 | 54.1 ± 11.9 | 96 | 46.3 ± 5.8 |
Notes:
Abbreviations: NAFLD: non-alcoholic fatty liver disease, CAP: controlled attenuation parameter, LSM: liver stiffness measurements, SE: standard error, T: tertile, kcal: kilocalorie, MET: metabolic equivalent; Models: Chi-Square test or Fisher’s exact test.
Physical Activity and Sedentary Behavior
Higher levels of physical activity were associated with lower risk of NAFLD in univariate, age- and sex-adjusted (Supplementary Table 1) and multivariable-adjusted analyses (Table 2). Specifically, even after adjusting for age, sex, metabolic syndrome, and total energy, the middle (>0 to <32 MET hours/week, Multivariable-adjusted OR=0.60, 95%CI 0.36, 0.99) and highest levels (≥32 MET hours/week, Multivariable-adjusted OR=0.60, 95%CI 0.38, 0.93) of physical activity were associated with 40% lower risk of NAFLD compared to 0 MET hours/week (Table 2). Physical activity levels were not significantly associated with advanced fibrosis among the Hispanic/Latino adults with NAFLD.
Table 2.
Multivariate Analyses for NAFLD and advanced fibrosis in Hispanic/Latino sample
NAFLD Multivariable-adjusted OR (95% CI) |
Advanced Fibrosis Multivariable-adjusted OR (95% CI) |
|
---|---|---|
Physical activity | ||
T1 (0 MET hours/week) | Ref | Ref |
T2 (>0 to <32 MET hours/week) | 0.60 (0.36, 0.99) | 0.43 (0.07, 2.65) |
T3 (≥32 MET hours/week) | 0.60 (0.38, 0.93) | 1.36 (0.45, 4.11) |
Sedentary behavior | ||
T1 (≤2 hours/day) | Ref | Ref |
T2 (>2 to <5 hours/day) | 1.87 (1.21, 2.88) | 1.09 (0.33, 3.61) |
T3 (≥ 5 hours/day) | 1.51 (0.92, 2.47) | 0.92 (0.27, 3.11) |
Health Eating Index 2015 | ||
T1 (HEI≤48.3) | Ref | Ref |
T2 (HEI>48.3 to HEI<61.3) | 0.60 (0.34, 1.06) | 1.11 (0.41, 2.96) |
T3 (HEI≥61.3) | 0.70 (0.51, 0.97) | 0.28 (0.12, 0.66) |
Total energy intake | ||
T1 (≤1575 kcal) | Ref | Ref |
T2 (>1575 to <2154 kcal) | 0.77 (0.37, 1.60) | 0.35 (0.13, 0.96) |
T3 (≥2154 kcal) | 0.93 (0.61, 1.42) | 0.91 (0.32, 2.55) |
Notes:
Abbreviations: NAFLD: non-alcoholic fatty liver disease, OR: odds ratio, CI: confidence interval, T: tertile, MET: metabolic equivalent, HEI: Healthy Eating Index 2015; kcal: kilocalorie;
Models: Multivariable logistic regression;
Values: Multivariable adjusted Odds Ratios (95% Confidence Intervals);
Physical Activity models controlled for age, sex, metabolic syndrome, and total energy;
Sedentary Behavior models controlled for age, sex, metabolic syndrome, and total energy;
HEI models controlled for age, sex, metabolic syndrome, total energy, physical activity, and alcohol use (categorical);
Total energy models controlled for age, sex, metabolic syndrome, HEI, physical activity, and alcohol use (categorical).
As compared to those who reported ≤2 hours of daily sedentary behavior, those reporting >2 but < 5 hours of daily sedentary behavior had an 87% higher risk of NAFLD (Multivariable-adjusted OR=1.87, 95% CI 1.21, 2.88). However, the highest tertile of sedentary behavior was not statistically significantly associated with NAFLD risk. Sedentary behavior was not associated with risk of advanced fibrosis (Table 2).
In models stratified by levels of sedentary behavior (Table 3), we found that high levels of physical activity, compared to no physical activity, were significantly protective of NAFLD in those with the highest levels of sedentary behavior. When we examined the joint effect of physical activity and sedentary behavior, we found that, as compared to the reference group of individuals with both the lowest levels of physical activity and the highest levels of sedentary behavior, those with the highest levels of physical activity and the lowest levels of sedentary behavior had especially low risk of NAFLD (Multivariable-adjusted OR=0.27, 95% CI 0.10, 0.69) (Table 4).
Table 3.
Association of physical activity with NAFLD and advanced fibrosis, stratified by tertiles of sedentary behavior
NAFLD | Advanced fibrosis | |||||
---|---|---|---|---|---|---|
Multivariable-adjusted OR (95% CI) | Low Sed | Mid Sed | High Sed | Low Sed | Mid Sed | High Sed |
Low PA | Ref | Ref | Ref | Ref | Ref | Ref |
Mid PA | 0.43 (0.11, 1.63) | 0.61 (0.18, 2.05) | 0.62 (0.24, 1.62) | 0.38 (0.03, 4.31) | 0.11 (0.02, 0.53) | 0.66 (0.07, 6.20) |
0.52 (0.16, 1.67) | 0.79 (0.31, 2.06) | 0.44 (0.24, 0.79) | 1.27 (0.14, 11.89) | 1.88 (0.79, 4.48) | 0.78 (0.13, 4.59) |
Notes:
Abbreviations: NAFLD: non-alcoholic fatty liver disease, OR: odds ratio, CI: confidence interval;
Low, Med, High refer to Tertiles 1, 2, and 3, respectively;
Models: Multivariable logistic regression, stratified by tertiles of sedentary behavior;
Values: Multivariable adjusted Odds Ratios (95% Confidence Intervals);
Adjusted for age, sex, metabolic syndrome and total energy.
Table 4.
Joint effects of lifestyle behaviors on NAFLD and advanced fibrosis
Multivariable adjusted OR (95% CI) | NAFLD Multivariable-adjusted OR (95% CI) |
Advanced fibrosis Multivariable-adjusted OR (95% CI) |
---|---|---|
Low PA/High Sed (Reference) | ||
Low PA/Mid Sed | 0.86 (0.31, 2.36) | 0.80 (0.20, 3.20) |
Low PA/Low Sed | 0.63 (0.23, 1.73) | 0.74 (0.10, 5.73) |
Mid PA/High Sed | 0.37 (0.19, 0.72) | 1.35 (0.21, 8.54) |
Mid PA/ Mid Sed | 0.80 (0.42, 1.49) | 0.46 (0.14, 1.55) |
Mid PA/Low Sed | 0.47 (0.25, 0.86) | 2.01 (0.39, 10.37) |
High PA/ High Sed | 0.66 (0.34, 1.28) | 0.43 (0.03, 5.29) |
High PA/ Mid Sed | 0.73 (0.36, 1.50) | 1.86 (0.40, 8.74) |
High PA/ Low Sed | 0.27 (0.10, 0.69) | 0.27 (0.07, 1.06) |
Combined HEI & total energy | ||
Low HEI/High Total Energy (Reference) | ||
Low HEI/Mid Total Energy | 0.96 (0.33, 2.89) | 0.20 (0.05, 0.93) |
Low HEI/Low Total Energy | 2.42 (1.07, 5.50) | 2.45 (0.54, 11.24) |
Mid HEI/High Total Energy | 0.87 (0.42, 1.82) | 1.88 (0.50, 7.16) |
Mid HEI/Mid Total Energy | 0.78 (0.29, 2.11) | 1.18 (0.20, 6.99) |
Mid HEI/Low Total Energy | 0.59 (0.19, 1.82) | 0.12 (0.02, 0.95) |
High HEI/High Total Energy | 1.35 (0.56, 3.28) | 0.36 (0.08, 1.74) |
High HEI/Mid Total Energy | 0.71 (0.34, 1.48) | 0.05 (0.01, 1.38) |
High HEI/Low Total Energy | 0.83 (0.49, 1.40) | 0.42 (0.13, 1.33) |
Notes:
Abbreviations: NAFLD: non-alcoholic fatty liver disease, OR: odds ratio, CI: confidence interval, PA: physical activity, Sed: sedentary behavior, HEI: Healthy Eating Index 2015;
Low, Med, High refer to Tertiles 1, 2, and 3, respectively;
Models: Multivariable logistic regression;
Values: Multivariable-adjusted Odds Ratios (95% Confidence Intervals);
Combined physical activity & sedentary behavior model adjusted for age, sex, metabolic syndrome and total energy;
Combined HEI & total energy model adjusted for age, sex, metabolic syndrome, physical activity and alcohol use.
Diet Quality and Total Energy Intake
Higher adherence to current U.S. dietary recommendations, as defined by the HEI-2015, was associated with reduced risk of NAFLD and advanced fibrosis (Table 2). In fully adjusted models, the highest vs lowest tertile of the HEI-2015 score was associated with 30% lower risk of NAFLD (Multivariable-adjusted OR=0.70, 95% CI 0.51, 0.97) and 72% lower risk of advanced fibrosis in Hispanic/Latino adults with NAFLD (Multivariable-adjusted OR=0.28, 95% CI 0.12, 0.66). Compared to the lowest tertile of total energy intake, those in the middle tertile had a reduced risk of advanced fibrosis (Multivariable-adjusted OR=0.35, 95% CI 0.13, 0.96).
In models stratified by level of total energy intake (Table 5), we found that among those with the lowest levels of energy intake, both the middle and highest tertiles of HEI-2015 (as compared to low HEI-2015) were associated with reduced risk of both NAFLD and advanced fibrosis.
Table 5.
Association of HEI with NAFLD and advanced fibrosis, stratified by tertiles of total energy intake
NAFLD | Advanced fibrosis | |||||
---|---|---|---|---|---|---|
Multivariable-adjusted OR (95% CI) | Low Total Energy | Mid Total Energy | High Total Energy | Low Total Energy | Mid Total Energy | High Total Energy |
Low HEI | Ref | Ref | Ref | Ref | Ref | Ref |
Mid HEI | 0.22 (0.06, 0.73) | 0.96 (0.42, 2.16) | 0.84 (0.40, 1.73) | 0.07 (0.01, 0.89) | 7.08 (1.26, 39.81) | 1.92 (0.58, 6.34) |
0.31 (0.13, 0.72) | 1.08 (0.47, 2.46) | 1.02 (0.41, 2.53) | 0.21 (0.05, 0.87) | 0.32 (0.03, 3.86) | 0.19 (0.03, 1.27) |
Notes:
Abbreviations: HEI: Healthy Eating Index 2015; NAFLD: non-alcoholic fatty liver disease, OR: odds ratio, CI: confidence interval;
Low, Med, High refer to Tertiles 1, 2, and 3, respectively;
Models: Multivariable logistic regression, stratified by tertiles of total energy intake;
Values: Multivariable-adjusted Odds Ratios (95% Confidence Intervals);
Adjusted for age, sex, metabolic syndrome, physical activity, and alcohol use.
In supplementary analyses controlling for each component of metabolic syndrome, results were largely similar (Supplementary Tables 2–5).
Discussion
In this study using data from a sample of Hispanic/Latino adults in the U.S., we found that greater physical activity was associated with a reduced risk of NAFLD, especially among those with low sedentary behavior. Specifically, Hispanic/Latino adults with both high levels of physical activity (defined as ≥32 MET hours/week, or the equivalent of running a little more than 3 hours/week) and low levels of sedentary behavior (≤2 hours sitting time/day), had 73% lower risk of NAFLD. These findings align with previous analyses of NHANES risk among all adults that found increased physical activity and reduced sedentary time were associated with lower NAFLD risk (Kim et al. 2020), as well as cohort data from Hispanic/Latinos specifically (Li et al. 2020). However, while the independent effects of each behavior were established in both prior studies, the joint effects were not examined. Our findings, along with the growing recent literature in this area, reinforce that need for public health and medical professionals to focus on the promotion of physical activity and the reduction of sedentary time for Hispanic/Latino adults to potentially prevent NAFLD, which is becoming exceedingly common in this community.
We also found that the highest diet quality, that is, adherence to the 2015–2020 Dietary Guidelines for Americans, was associated with a 30% reduced risk of NAFLD and a 72% reduced risk of advanced fibrosis in Hispanic/Latino adults in the U.S. These findings align with data from previous population-based and cohort studies in the U.S. (Yoo et al. 2020; Park et al. 2020). We also found that among those with low total energy intake, those with an HEI-2015 score greater than 48.3 had 69–78% reduced of risk of NAFLD and 79–93% reduced risk of advanced fibrosis. In other words, people reporting daily consumption of <1575kcal/day characterized by a high diet quality, had the lowest risk of NAFLD and advanced fibrosis. This finding indicates that the composition of the diet is potentially as important for reducing adverse liver-related outcomes as a hypocaloric diet. Additional longitudinal research is needed to further assess the combined association of total energy intake and diet quality with NAFLD and advanced fibrosis.
We observed some subtle differences in our findings when we controlled for the individual components of the metabolic syndrome in supplemental analyses versus a single composite variable in the main analyses. Specifically, while the directions of associations were similar, some point estimates were attenuated towards the null, the 95% CIs changed, and/or some associations were no longer statistically significant. While these differences may be due to uncontrolled confounding in the models containing only the composite metabolic syndrome variable, it may also be due to multicollinearity from the correlation between the individual components.
While we were initially interested in exploring differences in the association of lifestyle variables with NAFLD by acculturation, neither length of time in the U.S. nor language spoken at home acculturation was significant in univariate analyses, echoing previous research in this area (Balakrishnan et al. 2017). However, future research could explore the longitudinal impact of acculturation on NAFLD risk over the long-term.
There are several limitations that needs to be acknowledged. Foremost, the analyses in this study are based on cross-sectional data. Therefore, we cannot assess temporality. While NHANES aims to be representative of the U.S. population, it is possible that with the exclusion of participants relevant to this analysis it may have reduced the representativeness of our sample. The measures presented here are from self-reported physical activity and sedentary behavior, which may be subject to reporting bias. Moreover, there are known issues with systematic reporting error in population-level energy intake data (Schoeller et al. 2013; Archer, Hand, and Blair 2013). However, the strength of this study is the use of national data, collected using well-validated questionnaires, trained personnel and established procedures. We also examined associations of lifestyle behaviors with liver outcomes in the Hispanic/Latino population, for whom specific analyses are seldom conducted and for whom liver outcomes are becoming an increasingly important disparity.
Conclusion
Physical activity combined with limited sedentary behavior is critical to the prevention of NAFLD in Hispanic/Latino adults. A high-quality diet consumed within a restricted caloric intake may also be associated with reduced risk of both NAFLD and advanced fibrosis. While longitudinal research is still needed to bring greater precision to lifestyle recommendations for NAFLD and fibrosis prevention, our study provides evidence for public health and medical professionals to develop resources targeting Hispanic/Latinos with NAFLD to make lifestyle behavior change.
Supplementary Material
Financial disclosure:
This work was supported by funding from the Prevent Cancer Foundation and Cancer Prevention and Research Institute of Texas (RP160097, RP190513, RP200537) and the National Cancer Institute/National Institutes of Health through MD Anderson’s Cancer Center Support Grant (P30CA016672). JPH reports funding from Merck.
Contributor Information
Natalia I. Heredia, The University of Texas Health Science Center at Houston, School of Public Health, Department of Health Promotion & Behavioral Sciences, 7000 Fannin, Suite 2558, Houston, TX 77030.
Xiaotao Zhang, The University of Texas MD Anderson Cancer Center, Department of Epidemiology, Houston, Texas.
Maya Balakrishnan, Baylor College of Medicine, Department of Medicine, Section of Gastroenterology and Hepatology, Houston, Texas.
Jessica P. Hwang, The University of Texas MD Anderson Cancer Center, Department of General Internal Medicine, Houston, Texas.
Aaron P. Thrift, Baylor College of Medicine, Department of Medicine, Section of Epidemiology and Population Sciences & Dan L Duncan Comprehensive Cancer Center, Houston, Texas.
References
- Adams, Leon A, Anstee Quentin M, Herbert Tilg, and Giovanni Targher. 2017. “Non-alcoholic fatty liver disease and its relationship with cardiovascular disease and other extrahepatic diseases.” Gut 66 (6):1138–53. [DOI] [PubMed] [Google Scholar]
- Archer Edward, Hand Gregory A, and Blair Steven N. 2013. “Validity of US nutritional surveillance: National Health and Nutrition Examination Survey caloric energy intake data, 1971–2010.” PloS One 8 (10):e76632. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Armstrong Timothy, and Bull Fiona. 2006. “Development of the world health organization global physical activity questionnaire (GPAQ).” Journal of Public Health 14 (2):66–70. [Google Scholar]
- Balakrishnan Maya, Kanwal Fasiha, El-Serag Hashem B, and Thrift Aaron P. 2017. “Acculturation and nonalcoholic fatty liver disease risk among hispanics of mexican origin: findings from the National Health and Nutrition Examination Survey.” Clinical Gastroenterology and Hepatology 15 (2):310–2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Centers for Disease Control & Prevention. 2021. “National Health and Nutrition Examination Survey. 2017–2018 Data Documentation, Codebook, and Frequencies. Liver Ultrasound Transient Elastography.” Accessed March 4. https://wwwn.cdc.gov/Nchs/Nhanes/2017-2018/LUX_J.htm.
- Centers for Disease Control & Prevention. 2021. “NHANES 2017–2018. .” Accessed March 4. https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/default.aspx?BeginYear=2017.
- Centers for Disease Control & Prevention. “National Health and Nutrition Examination Survey (NHANES) Liver Ultrasound Transient Elastography Procedures Manual.” https://wwwn.cdc.gov/nchs/data/nhanes/2017-2018/manuals/2018_Liver_Ultrasound_Elastography_Procedures_Manual.pdf.
- Chalasani Naga, Younossi Zobair, Joel E Lavine Michael Charlton, Cusi Kenneth, Rinella Mary, Harrison Stephen A, Brunt Elizabeth M, and Sanyal Arun J. 2018. “The diagnosis and management of nonalcoholic fatty liver disease: practice guidance from the American Association for the Study of Liver Diseases.” Hepatology 67 (1):328–57. [DOI] [PubMed] [Google Scholar]
- Chen Te-Ching, Parker Jennifer D, Clark Jason, Shin Hee-Choon, Rammon Jennifer R, and Burt Vicki L. 2018. “National health and nutrition examination survey: estimation procedures, 2011–2014.” [PubMed]
- DeSalvo, Karen B, Richard Olson, and Casavale Kellie O. 2016. “Dietary guidelines for Americans.” JAMA 315 (5):457–8. [DOI] [PubMed] [Google Scholar]
- El-Serag, Hashem B, Sardell Rebecca, Thrift Aaron P, Kanwal Fasiha, and Miller Paige. 2021. “Texas has the highest hepatocellular carcinoma incidence rates in the USA.” Digestive Diseases and Sciences 66 (3):912–6. [DOI] [PubMed] [Google Scholar]
- Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults, . 2001. “Executive summary of the third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III).” JAMA 285 (19):2486. [DOI] [PubMed] [Google Scholar]
- Friedman, Scott L, Neuschwander-Tetri Brent A, Rinella Mary, and Sanyal Arun J. 2018. “Mechanisms of NAFLD development and therapeutic strategies.” Nature Medicine 24 (7):908–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Heredia, Natalia I, Zhang Xiaotao, Balakrishnan Maya, Daniel Carrie R, Hwang Jessica P, McNeill Lorna H, and Thrift Aaron P. 2022. “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.” Preventive Medicine 154:106903. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Huang, Daniel Q, El-Serag Hashem B, and Loomba Rohit. 2020. “Global epidemiology of NAFLD-related HCC: trends, predictions, risk factors and prevention.” Nature reviews Gastroenterology & hepatology:1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Johnson CL, Paulose-Ram R, Ogden CL, Carroll MD, Kruszon-Moran D, Dohrmann SM, and Curtin LR. 2013. “National health and nutrition examination survey: analytic guidelines, 1999–2010.” Vital Health Stat 2 (161):1–24. [PubMed] [Google Scholar]
- Kim Donghee, Cholankeril George, Loomba Rohit, and Ahmed Aijaz. 2021. “Prevalence of fatty liver disease and fibrosis detected by FibroScan in adults in the United States, 2017–2018.” Clinical Gastroenterology and Hepatology 19 (7):1499–501. [DOI] [PubMed] [Google Scholar]
- Kim Donghee, Vazquez‐Montesino Luis Miguel, Li Andrew A, Cholankeril George, and Ahmed Aijaz. 2020. “Inadequate physical activity and sedentary behavior are independent predictors of nonalcoholic fatty liver disease.” Hepatology 72 (5):1556–68. [DOI] [PubMed] [Google Scholar]
- Li Jun, Hua Simin, Chen Guo‐Chong, Strizich Garrett, Kuniholm Mark H, Shan Zhilei, Talavera Gregory A, Castañeda Sheila F, Gellman Marc D, and Cai Jianwen. 2020. “Objectively measured sedentary time, physical activity and liver enzyme elevations in US Hispanics/Latinos.” Liver International 40 (8):1883–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Park Song-Yi, Noureddin Mazen, Boushey Carol, Wilkens Lynne R, and Setiawan Veronica W. 2020. “Diet quality association with nonalcoholic fatty liver disease by cirrhosis status: The multiethnic cohort.” Current developments in nutrition 4 (3):nzaa024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Reedy Jill, Lerman Jennifer L, Krebs-Smith Susan M, Kirkpatrick Sharon I, Pannucci TusaRebecca E, Wilson Magdalena M, Subar Amy F, Kahle Lisa L, and Tooze Janet A. 2018. “Evaluation of the healthy eating index-2015.” Journal of the Academy of Nutrition and Dietetics 118 (9):1622–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rinella Mary E. 2015. “Nonalcoholic fatty liver disease: a systematic review.” JAMA 313 (22):2263–73. [DOI] [PubMed] [Google Scholar]
- Romeo Stefano, Kozlitina Julia, Xing Chao, Pertsemlidis Alexander, Cox David, Pennacchio Len A, Boerwinkle Eric, Cohen Jonathan C, and Hobbs Helen H. 2008. “Genetic variation in PNPLA3 confers susceptibility to nonalcoholic fatty liver disease.” Nature Genetics 40 (12):1461–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schoeller, Dale A, Thomas Diana, Archer Edward, Heymsfield Steven B, Blair Steven N, Goran Michael I, Hill James O, Atkinson Richard L, Corkey Barbara E, and John Foreyt. 2013. “Self-report–based estimates of energy intake offer an inadequate basis for scientific conclusions.” The American journal of clinical nutrition 97 (6):1413–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- White, Donna L, Thrift Aaron P, Fasiha Kanwal, Jessica Davila, and El-Serag Hashem B. 2017. “Incidence of hepatocellular carcinoma in all 50 United States, from 2000 through 2012.” Gastroenterology 152 (4):812–20. e5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yoo, Eric R, Donghee Kim, Vazquez‐Montesino Luis M, Escober Jessica A, Li Andrew A, Tighe Sean P, Fernandes Christopher T, George Cholankeril, and Aijaz Ahmed. 2020. “Diet quality and its association with nonalcoholic fatty liver disease and all‐cause and cause‐specific mortality.” Liver International 40 (4):815–24. [DOI] [PubMed] [Google Scholar]
- Zhang Xiaotao, Heredia Natalia I, Balakrishnan Maya, and Thrift Aaron P. 2021. “Prevalence and factors associated with NAFLD detected by vibration controlled transient elastography among US adults: Results from NHANES 2017–2018.” PloS One 16 (6):e0252164. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.