Abstract
Background
Non-alcoholic fatty liver disease (NAFLD) is a major contributor to liver diseases globally, yet there are limited studies investigating the impact of diet and lifestyle factors on its development. This study aimed to examine the association between the prevalence of NAFLD and predicted pro-inflammatory high-sensitivity C-reactive protein (hs-CRP) score.
Methods
We included 1,076 Korean adults who underwent a medical examination at the Seoul National University Hospital Gangnam Healthcare Center in Korea between May and December 2011 and updated in 2021. The predicted pro-inflammatory hs-CRP score was derived from pro-inflammatory demographic, lifestyle, dietary, and anthropometric factors, and NAFLD was diagnosed using liver ultrasound. Multivariable-adjusted odds ratios (ORs) and 95% confidence intervals (CIs) of NAFLD odds according to predicted pro-inflammatory hs-CRP score were estimated using logistic regression at a two-sided P < 0.05.
Results
Among the 1,076 participants, 320 had NAFLD. The multivariable-adjusted ORs and 95% CIs for NAFLD by tertiles of predicted pro-inflammatory hs-CRP score were 1.00, 3.30 (2.06, 5.30), 18.25 (10.47, 31.81; P < 0.0001) in men and women combined, 1.00, 1.77 (1.10, 2.84), and 3.26 (2.02, 5.28; P < 0.0001) among men only, and 1.00, 3.03 (1.39, 6.62), and 16.71 (7.05, 39.63; P < 0.0001) among women only.
Conclusions
Predicted pro-inflammatory hs-CRP score was associated with higher odds of NAFLD. Adopting dietary and lifestyle changes related to lower inflammation might be a valuable strategy for preventing NAFLD.
Keywords: non-alcoholic fatty liver disease, fatty liver, high-sensitivity C-reactive protein, pro-inflammation
Introduction
Non-alcoholic fatty liver disease (NAFLD) is associated with metabolic co-morbidities such as obesity, diabetes mellitus, and dyslipidemia, and is often characterized by a constellation of liver damage [1, 2]. It is a major contributor to liver diseases worldwide, with severe anomalies that impose a considerable cost of care [3, 4]. Current reports suggest that NAFLD has a worldwide prevalence of ∼25% and a pooled incidence of ∼28 and ∼52 per 1,000 person-years among Westerners and Asians, respectively [1, 5].
A constellation of genetic, metabolic, lifestyle, and environmental factors accounts for the burden of NAFLD [1–5]. Also, designing cost-effective interventions for the early identification of populations at risk of NAFLD cannot be underestimated in managing the burden of NAFLD. High-sensitivity C-reactive protein (hs-CRP) has been reported as a non-invasive marker for predicting NAFLD [6]. The applicability of a pro-inflammation prediction approach (derived from typical lifestyle and dietary exposures) may be necessary for the early identification of populations at risk of disease events for public health intervention. It is a cost-effective, non-invasive tool for personalized nutrition for chronic disease prevention and management. Besides, obesity [7], physical activity [8], socioeconomic status [9], and dietary factors [10, 11] have been linked to long-term prognosis of chronic inflammation. However, whether these factors can be deployed to discriminate inflammation and its relationship with NAFLD has yet to be understood.
A sturdy pro-inflammation prediction index in discerning NAFLD events is overdue in light of the mounting burden of NAFLD. Some reports have demonstrated the viability of a diet-based pro-inflammatory index on the risk of non-communicable diseases [14, 15]. A few studies have also reported that pro-inflammatory diet-based indexes are associated with higher odds of a fatty liver index and NAFLD-fatty liver score [12, 13], with limited evidence on the true significance of inflammation in actual NAFLD events. We have previously developed and validated a sex-specific pro-inflammatory predicted score for hs-CRP based on a combination of pro-inflammatory demographic, lifestyle, dietary, and anthropometric factors to predict hs-CRP levels accurately and reflect inflammation in the Korean population [14]. Whether the predicted pro-inflammatory hs-CRP score is associated with NAFLD has yet to be understood. The study aimed to assess the association of predicted pro-inflammatory hs-CRP score with ultrasound-diagnosed NAFLD among Koreans.
Patients and methods
Study approval
This study was approved by the institutional review board of the Seoul National University Hospital (No. 1909–034–1062). Also, the study was conducted following the Ethical Principles for Medical Research Involving Human Subjects as defined in the Declaration of Helsinki.
Study population
This study included participants who underwent a medical examination at the Seoul National University Hospital Gangnam Healthcare Center in Seoul, Korea, from May to December 2011 and updated in 2021. Among 2,086 participants who underwent a liver ultrasound procedure, participants were exempted for the following reasons: missing dietary data (n = 38), implausible energy intake (energy intake more or less than three standard deviations of the mean log-transformed energy intake; n = 32), missing data on hepatitis (n = 238), chronic liver disease including those positive for hepatitis B virus (n = 66) and hepatitis C virus (n = 20), or excessive alcohol consumption (>20 g/d for males and >10 g/d for females [15]; n = 616). In all, 1,076 participants were included in the final analysis of this current study (Supplementary Figure 1).
Ascertainment of non-alcoholic fatty liver disease
NAFLD status was defined in this study according to the American Association for the Study of Liver Diseases guidelines [15]. Qualified radiologists conducted a blind review of vital clinicopathologic data and liver ultrasound (Acusion, Sequoia 512, Siemens, Mountain View, CA, USA) of participants using standard operating principles [16]. NAFLD was diagnosed as the presence of fatty liver disease in the absence of any of the following conditions: (i) seropositivity for hepatitis B surface antigens or antibodies to hepatitis C virus, (ii) excessive alcohol intake, (iii) other causes of liver disease, and (iv) medications known to produce fatty liver disease.
Assessment of lifestyle and clinical factors
Dietary intake was assessed using a validated food frequency questionnaire [17] administered by a dietitian during the medical examination before the liver ultrasound. Participants reported usual food consumption and typical portion sizes in the last 12 months preceding the study. Each food and drink item had nine options ranging from ‘never’ or ‘less than once/month’ to ‘three times/day’ and three portion size options; ‘one-half of a standard serving’, ‘one standard serving,’ and ‘one and half of standard serving’. The average daily consumption of food (in g/d) and nutrients was estimated by multiplying the frequency of consumption by the reported amount. The food frequency questionnaire assessment and food intake estimation have been reported previously [14, 18]. The summation of all ethanol weight (including the multiplication of quantities and frequencies of types of liquors/alcohol) was used to determine alcohol intake in g/d.
By completing questionnaires on sociodemographic, lifestyle, and medical history, participants provided information on sex (male or female), menopausal status for females (premenopausal or postmenopausal), age (in years), marital status (single, married, or separated/divorced/widowed), highest education completed (elementary to high school or at least a college education), and current smoker (no or yes, defined as having a smoked cigarette at least once in the last 12 months). Physical activity in metabolic equivalent tasks (MET-minutes/week) was estimated based on the average number of minutes and the number of days spent on moderate to intense levels of physical activity or walking [19] and classified as ‘none’, ‘<840 MET-minutes/week,’ and ‘≥840 MET-minutes/week’, respectively. Body mass index (BMI) was derived from the weight (kg) divided by the square of the height (m). Systolic and diastolic blood pressures were measured twice using the InBody blood pressure monitor (BPBIO320; InBody Co., Ltd, Korea) and the mean values were estimated to define hypertension. Hypertension was defined as a systolic blood pressure of ≥140 mmHg, diastolic blood pressure of ≥90 mmHg, or the use of antihypertensive medication according to the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure [20]. Fasting blood samples were collected after a 12-h overnight fast and all laboratory tests were carried out, including the serum levels of fasting glucose and high-sensitivity C-reactive protein (hs-CRP). Fasting blood glucose and hs-CRP were assessed using a clinical chemical analyser (ARCHITECT c16000; Abbott Laboratories, Abbott Park, IL, USA). The coefficient of variation of these biomarkers was <2%. Diabetes mellitus was defined as a fasting blood glucose level of ≥126 mg/dL, a glycated hemoglobin level of ≥6.5%, or a history of using glucose-lowering medications [21].
Computation of the predicted pro-inflammatory hs-CRP score
The predicted pro-inflammatory hs-CRP score model used in this study was developed by using the Health Examinees study group in Korea [22] and validated with hs-CRP levels in this population [14]. Details of how the predicted pro-inflammatory hs-CRP model was developed, tested, and validated have been reported previously [14]. In brief, the predicted pro-inflammatory hs-CRP score model was derived from a constellation of pro-inflammation factors, including food groups, nutrients, alcohol intakes, BMI, smoking status, physical activities, educational levels, and menopausal status of women, using a stepwise linear regression model. The predicted pro-inflammatory hs-CRP score was computed as the summation of scores arising from the multiplication of the beta coefficients (reported in sex-combined and sex-specific predicted pro-inflammatory hs-CRP models derived from empirical models) with individual responses on demographic and lifestyle factors and food intake (g/d). A detailed description of the computation is shown in Supplementary Table 1. The predicted pro-inflammatory hs-CRP score was categorized into tertiles to include a reasonable number of respondents in each category for statistical comparison.
Statistical analysis
Characteristics of participants are presented across tertile distribution of the sex-combined and sex-specific predicted pro-inflammatory hs-CRP score. Continuous and categorical variables are presented as mean ± standard deviation (SD) and frequencies (percentages), respectively. Logistic regression models were applied to estimate the odds ratio (ORs) and 95% confidence intervals (CIs) for NAFLD prevalence by tertiles of the sex-combined and -specific predicted pro-inflammatory hs-CRP score. In the sex-combined and male-specific logistic regression models, we adjusted for age (in years, continuous), current alcohol drinking (no or yes), current smoking (no or yes), physical activity (<840 MET-minutes/week or ≥840 MET-minutes/week), and education (elementary to high school or at least a college education). Furthermore, the median value of the tertile distribution of the predicted pro-inflammatory hs-CRP score was assigned in a continuous model to assess the test for linear trends in the relationship between the predicted pro-inflammatory hs-CRP score and NAFLD. The predicted pro-inflammatory hs-CRP scores were dichotomized by median values for subgroup analyses by overweight/obesity status (<23 or ≥23 kg/m2), diabetes mellitus status (no or yes), hypertension status (no or yes), and menopausal status (premenopausal or postmenopausal) for females only. Effect modification was tested by including interaction term(s) in the subgroup analyses. The likelihood ratio test was used to compare nested models that included cross-product terms with the original models that did not include the term. Also, the receiver-operating characteristic (ROC) curve was plotted in the sex-combined and -specific regression models by treating the predicted pro-inflammatory hs-CRP score in a continuous model. The area under the curve (AUC), sensitivity, and specificity were estimated to assess the predictive ability of the predicted pro-inflammatory hs-CRP score in discriminating NAFLD. All statistical analyses were conducted using SAS 9.4 (SAS Institute Inc., Cary, NC, USA) at a two-sided P-value of <0.05.
Results
Characteristics of participants by tertiles of the predicted pro-inflammatory hs-CRP score
Overall, 320 (29.7%) of the 1,076 participants in this sample had NAFLD (Supplementary Figure 1). The general characteristics of participants by tertiles of the predicted pro-inflammatory hs-CRP score are presented in Table 1. Participants with high predicted pro-inflammatory hs-CRP scores were more likely to be male, older, current smokers, overweight/obese (≥23 kg/m2), hypertensive, or diabetic compared with those with low scores. We found similar observations in the sex-specific models of the predicted pro-inflammatory hs-CRP score (Supplementary Table 2).
Table 1.
Characteristics of participants by tertiles of the predicted pro-inflammatory hs-CRP score
| Characteristic | Tertiles of the predicted pro-inflammatory hs-CRP score |
||
|---|---|---|---|
| Tertile 1 | Tertile 2 | Tertile 3 | |
| (n = 358) | (n = 359) | (n = 359) | |
| Predicted pro-inflammatory hs-CRP score, mean ± SD | –3.1 ± 0.1 | –2.7 ± 0.1 | –2.4 ± 0.1 |
| NAFLD cases, n (%) | 34 (9.5) | 81 (22.6) | 205 (57.1) |
| Females, n (%) | 299 (83.5) | 211 (58.8) | 99 (27.6) |
| Postmenopausal women, n (%) | 208 (69.6) | 81 (38.4) | 17 (17.2) |
| Age (years), mean ± SD | 45.7 ± 7.2 | 52.4 ± 7.6 | 56.7 ± 9.6 |
| Highest education completed | |||
| College graduate or above, n (%) | 310 (86.6) | 268 (74.7) | 264 (73.5) |
| Current smoking, n (%) | 9 (2.5) | 27 (7.5) | 94 (26.2) |
| Alcohol intake (g/d), mean ± SD | 2.2 ± 3.3 | 3.8 ± 4.9 | 4.3 ± 5.9 |
| Physical activity (MET-minutes/week), mean ± SD | 1,271.8 ± 2,920.7 | 1,191.5 ± 2,073.3 | 1,267.5 ± 2,654.1 |
| ≥840 MET-minutes/week, n (%) | 130 (36.3) | 131 (36.5) | 127 (35.4) |
| BMI (kg/m2), mean ± SD | 20.3 ± 1.7 | 22.9 ± 1.5 | 25.8 ± 2.5 |
| BMI ≥ 23 kg/m2, n (%) | 21 (0.3) | 172 (9.8) | 323 (60.5) |
| hs-CRP (mg/dL), mean ± SD | 0.1 ± 0.2 | 0.1 ± 0.3 | 0.2 ± 0.3 |
| Waist circumference (cm), mean ± SD | 76.0 ± 5.3 | 82.9 ± 4.7 | 90.8 ± 6.6 |
| Systolic blood pressure (mmHg), mean ± SD | 109.3 ± 11.8 | 114.7 ± 12.9 | 120.1 ± 13.2 |
| Diastolic blood pressure (mmHg), mean ± SD | 69.2 ± 9.2 | 74.0 ± 9.8 | 76.9 ± 10.1 |
| Hypertensivesa, n (%) | 22 (6.2) | 68 (18.9) | 134 (37.3) |
| Fasting plasma glucose (mg/dL), mean ± SD | 86.2 ± 13.3 | 93.7 ± 14.6 | 100.9 ± 19.5 |
| Diabetes mellitusb, n (%) | 7 (2.0) | 26 (7.2) | 66 (18.4) |
hs-CRP: high-sensitivity C-reactive protein.
Continuous variables are presented as mean ± SD and categorical variables are presented as n (%); SD: standard deviation; NAFLD: non-alcoholic fatty liver disease; BMI: body mass index; hs-CRP: high-sensitivity C-reactive protein; MET: the metabolic equivalent of task.
Hypertension was defined as systolic blood pressure of ≥140 mmHg, diastolic blood pressure of ≥90 mmHg, or the use of antihypertensive medication.
Diabetes was defined as fasting blood glucose value of ≥126 mg/dL, a glycated hemoglobin level of ≥6.5%, or a history of using glucose-lowering medications.
Predicted pro-inflammatory hs-CRP score and odds of NAFLD
We found that a higher predicted pro-inflammatory hs-CRP score was associated with the odds of NAFLD (Table 2). In the sex-combined model, the ORs and 95% CIs for NAFLD in the first, second, and third tertiles of predicted pro-inflammatory hs-CRP score were 1.00, 3.53 (2.25, 5.55), and 19.25 (12.02, 30.82; P < 0.0001), respectively, after adjusting for age only. The ORs were slightly attenuated but remained statistically significant after adjusting for age, sex, alcohol drinking, smoking, physical activity, and highest education completed; 1.00, 3.30 (2.06, 5.30), and 18.25 (10.47, 31.81; P < 0.0001) for first, second, and third tertiles, respectively. A similar trend was observed among males and females in the sex-combined model. In the male-specific model, ORs (95% CIs) of NAFLD by tertile of the predicted pro-inflammatory hs-CRP score were 1.00, 1.77 (1.10, 2.84), and 3.26 (2.02, 5.28; P < 0.0001) for the first, second, and third tertiles, respectively, after adjusting for similar covariates in the sex-combined model. Similarly, higher predicted pro-inflammatory hs-CRP score was associated with high odds of NAFLD in the female-specific model; ORs (95% CIs): 1.00, 3.03 (1.39, 6.62), and 16.71 (7.05, 39.63; P < 0.0001) for the first, second, and third tertiles, respectively, after adjusting for similar covariates in the sex-combined model and menopausal status. These findings trended similarly in the dichotomized hs-CRP score independently of sex differentials (Supplementary Table 3).
Table 2.
Odds ratios and 95% confidence intervals for NAFLD according to the tertiles of the predicted pro-inflammatory hs-CRP score for the sex-combined model, male-specific model, and female-specific model
| Regression model | Tertiles of the predicted pro-inflammatory hs-CRP score |
|||
|---|---|---|---|---|
| Tertile 1 | Tertile 2 | Tertile 3 | P trend | |
| Sex-combined modela | ||||
| NAFLD cases/total | 34/358 | 81/359 | 205/359 | |
| Age-adjusted | 1.00 | 3.53 (2.25, 5.55) | 19.25 (12.02, 30.82) | <0.0001 |
| Model 1b | 3.30 (2.06, 5.30) | 18.25 (10.47, 31.81) | <0.0001 | |
| Males in sex-combined model | ||||
| NAFLD cases/total | 11/59 | 45/148 | 154/260 | |
| Age-adjusted | 1.00 | 2.47 (1.15, 5.30) | 10.30 (4.78, 22.21) | <0.0001 |
| Model 2c | 3.08 (1.39, 6.82) | 16.64 (6.99, 39.65) | <0.0001 | |
| Females in sex-combined model | ||||
| NAFLD cases/total | 23/299 | 36/211 | 51/99 | |
| Age-adjusted | 1.00 | 2.79 (1.53, 5.12) | 15.85 (7.73, 32.48) | <0.0001 |
| Model 3d | 2.81 (1.52, 5.21) | 16.70 (7.80, 35.74) | <0.0001 | |
| Male-specific modele | ||||
| NAFLD cases/total | 48/155 | 69/156 | 93/156 | |
| Age-adjusted | 1.00 | 1.77 (1.11, 2.82) | 3.29 (2.06, 5.24) | <0.0001 |
| Model 2 | 1.77 (1.10, 2.84) | 3.26 (2.02, 5.28) | <0.0001 | |
| Female-specific modelf | ||||
| NAFLD cases/total | 11/203 | 25/203 | 74/203 | |
| Age-adjusted | 1.00 | 3.00 (1.38, 6.53) | 14.83 (6.48, 33.93) | <0.0001 |
| Model 3 | 3.03 (1.39, 6.62) | 16.71 (7.05, 39.63) | <0.0001 | |
NAFLD: non-alcoholic fatty liver disease; hs-CRP: high-sensitivity C-reactive protein.
Tertiles of the predicted pro-inflammatory hs-CRP scores were generated for men and women combined in a sex-combined model.
Model 1 was adjusted for age (continuous, years), sex (males or females), current alcohol drinking (no or yes), smoking (no, past or current), exercise (none, <840 MET-minutes/week or ≥840 MET-minutes/week), and highest education completed (high school and less, or college education and above).
Model 2 was adjusted for age (continuous, years), current alcohol drinking (no or yes), smoking (no, past or current), exercise (none, <840 MET-minutes/week or ≥840 MET-minutes/week), and highest education completed (high school and less, or college education and above).
Model 3 was adjusted for age (continuous, years), current alcohol drinking (no or yes), smoking (no, past or current), exercise (none, <840 MET-minutes/week or ≥840 MET-minutes/week), highest education completed (high school and less or college education and above), and menopausal status (premenopausal or postmenopausal).
Tertiles of the predicted pro-inflammatory hs-CRP scores were generated among men only in a male-specific model.
Tertiles of the predicted pro-inflammatory hs-CRP scores were generated for women only in a female-specific model.
In the dichotomized predicted pro-inflammatory hs-CRP score, we tested whether overweight/obesity, diabetes mellitus, hypertension, and menopausal status modified the association of predicted pro-inflammatory hs-CRP score with NAFLD (Table 3). We found that the associations were slightly higher among participants with overweight/obesity, diabetes mellitus, or hypertension but did not reach statistically significant interactions in the sex-combined group. In the male-specific model, the predicted pro-inflammatory hs-CRP score was associated with NAFLD among non-diabetics and normotensives, with no evidence of statistically significant interaction. A more pronounced association was observed among premenopausal women than postmenopausal women, but the interaction was not statistically significant.
Table 3.
Odds ratios and 95% confidence intervals for NAFLD by subgroup analyses of obesity, diabetes mellitus, hypertension, and menopausal status according to the dichotomous category of the predicted pro-inflammatory hs-CRP score for the sex-combined, male-specific, and female-specific models
| Predicted pro-inflammatory hs-CRP score |
||||||
|---|---|---|---|---|---|---|
| Subgroup analysis | < Median |
≥ Median |
P trend | P for interaction | ||
| Cases/total | Reference | Cases/total | OR (95% CI) | |||
| Sex-combined modela | ||||||
| Overweight/obesity | ||||||
| BMI < 23 kg/m2 | 41/459 | 1.00 | 26/101 | 3.89 (1.69, 8.94) | 0.0014 | 0.77 |
| BMI ≥ 23 kg/m2 | 19/79 | 1.00 | 234/437 | 3.61 (1.86, 7.00) | 0.0001 | |
| Diabetes mellitus status | ||||||
| No | 53/522 | 1.00 | 202/455 | 7.72 (5.01, 11.89) | <0.0001 | 0.15 |
| Yes | 7/16 | 1.00 | 58/83 | 8.61 (1.95, 38.04) | 0.005 | |
| Hypertension status | ||||||
| No | 54/488 | 1.00 | 166/364 | 7.58 (4.80, 11.98) | <0.0001 | 0.56 |
| Yes | 6/50 | 1.00 | 94/174 | 8.45 (3.09, 23.12) | <0.0001 | |
| Male-specific modelb | ||||||
| Overweight/obesity | ||||||
| BMI < 23 kg/m2 | 22/111 | 1.00 | 9/41 | 1.18 (0.47, 2.91) | 0.73 | 0.95 |
| BMI ≥ 23 kg/m2 | 66/122 | 1.00 | 113/193 | 1.14 (0.72, 1.83) | 0.58 | |
| Diabetes mellitus status | ||||||
| No | 61/190 | 1.00 | 103/208 | 2.01 (1.32, 3.05) | 0.001 | 0.67 |
| Yes | 27/43 | 1.00 | 19/26 | 1.37 (0.40, 4.72) | 0.62 | |
| Hypertension status | ||||||
| No | 53/166 | 1.00 | 88/173 | 2.12 (1.35, 3.33) | 0.001 | 0.10 |
| Yes | 35/67 | 1.00 | 34/61 | 1.15 (0.55, 2.42) | 0.71 | |
| Female-specific modelb,c | ||||||
| Menopausal status | ||||||
| Premenopausal | 2/66 | 1.00 | 68/237 | 11.98 (2.74, 52.53) | <0.0001 | 0.35 |
| Postmenopausal | 18/238 | 1.00 | 22/68 | 5.83 (2.66, 12.77) | <0.0001 | |
hs-CRP: high-sensitivity C-reactive protein.
Model was adjusted for age (continuous, years), sex (males or females), current alcohol drinking (no or yes), smoking (no, past or current), exercise (none, <840 MET-minutes/week or ≥840 MET-minutes/week), and highest education completed (high school or less, college education and above).
Model was adjusted for age (continuous, years), current alcohol drinking (no or yes), smoking (no, past or current), exercise (none, <840 MET-minutes/week or ≥840 MET-minutes/week), highest education completed (high school and less or college education and above), and menopausal status (premenopausal or postmenopausal) among women only.
Because of fewer cases (one, two, and three for BMI ≥ 23 kg/m2, diabetes yes, and hypertension yes, respectively); we did not present the results for interaction by BMI, diabetes, or hypertension.
Area under the ROC curve of the predicted pro-inflammatory hs-CRP score
Furthermore, when the predicted pro-inflammatory hs-CRP score was included as a continuous variable in the regression models to assess its predictive ability in discriminating against NAFLD, AUCs and 95% CIs were 0.81 (0.78, 0.84) in the sex-combined model, 0.63 (0.58, 0.68) in the male-specific model, and 0.79 (0.75, 0.84) in the female-specific model (Supplementary Figure 2). These findings suggest that the predicted pro-inflammatory hs-CRP score potentially exercised good ability in distinguishing NAFLD cases in this sample, suggesting the feasibility of the predicted pro-inflammatory hs-CRP score for population-level risk assessment of NAFLD for early detection and timely intervention.
Discussion
In this study, we tested the association of predicted pro-inflammatory hs-CRP score (empirically derived from pro-inflammatory demographic, lifestyle, dietary, and anthropometric factors) with ultrasound-measured NAFLD. We observed that a higher predicted pro-inflammatory hs-CRP score was associated with higher odds of NAFLD among Koreans. The associations were generally more prominent among women than men. Reasons for the higher odds of NAFLD among women than men in this study are multifactorial. It might not be far-fetched from the complex difference in the homeostasis and control of vascular function among men and women [23]. Inflammation-related alteration in endothelial function has been linked to a higher risk of cardiometabolic diseases among women [24]. Also, women-specific risk factors (such as menopause and pregnancy complications, among others) that manipulate sex hormones to trigger intra-molecular changes in the endocrine systems have been linked to poor cardiovascular health [25].
Plasma hs-CRP is a major acute-phase protein and a valuable measure of systemic inflammation linked to NAFLD [26–29], but limited studies have utilized sociodemographic, lifestyle, and dietary factors to predict hs-CRP and investigated its association with NAFLD. As against the objectively measured bloodstream-derived hs-CRP profiles, the predicted pro-inflammatory hs-CRP score was derived from pro-inflammatory demographic, lifestyle, dietary, and anthropometric factors. The predicted pro-inflammatory hs-CRP score significantly correlated with circulating hs-CRP profiles [14]. Therefore, using the predicted pro-inflammatory hs-CRP score may be necessary and potentially viable to discern the inflammatory status and discriminate populations at risk of chronic inflammation for early intervention and management. Our findings suggest a novel paradigm for personalized nutrition and lifestyle approaches in identifying populations at risk of NAFLD early. This approach enables timely intervention to improve quality of life and prevent complications associated with NAFLD.
Earlier reports on this subject utilized either pro-inflammation dietary [12, 30–33] or metabolic/clinical/genetic [34–37] factors, but our model employed broadly established pro-inflammatory demographic, lifestyle, and dietary factors to derive a sturdy and validated model in predicting hs-CRP scores. In tandem with our findings, diet-derived pro-inflammatory scores have been associated with higher odds of fatty liver [12], poorer hepatic health [12, 13], and obesity-related hypertriglyceridemia [33].
The precise mechanism by which predicted pro-inflammatory hs-CRP score was associated with NAFLD events is still evolving but can be hypothesized in the following ways. First, excessive consumption of lipid-dense foods may promote the excessive release of misfolded proteins and overloaded adipocytes in the endoplasmic reticulum [38], which can lead to the amplification of membrane lipid production via the inositol-requiring enzyme 1/X-box-binding protein 1 and unfolded protein response complexes, and promote apoptotic stress (and excessive release of adipokines in the adipose tissue, such as hs-CRP, vascular endothelial growth factor, and tumor necrosis factor-α) [39] that promotes chronic low-grade inflammation and causes NAFLD [29, 40]. Second, diet plays a role in gut composition [41, 42] and disturbances in gut composition could compromise gut integrity (via a series of complex pathways) [43], particularly in an obesogenic state, which could lead to the release of pro-inflammatory factors to promote chronic diseases [44, 45]. In summary, proclivities to pro-inflammatory dietary and lifestyle factors may be related to NAFLD development by activating inflammatory pathways and releasing cytokines and adipokines, particularly in obesogenic conditions [42, 46].
In tandem with our study, some reports have alluded to a null modification by obesity status in the association between predicted inflammatory index/score and chronic diseases [12, 28, 31, 37]. However, other studies have reported the contrary [14, 29, 30, 36]. The reason(s) for these differences is largely unknown. It is worth noting that some observational [47, 48] and intervention [49] studies have linked pro-inflammatory dietary factors with obesity. Obesity is linked to a constellation of chronic disease events and chronic low-grade inflammation [38, 42], which, under most circumstances, are primary drivers of inflammation prior to the onset of chronic diseases. Although there was no evidence of significant interaction by obesity status in the association of predicted pro-inflammatory hs-CRP score with NAFLD in our study, this warrants further investigation.
Our study lends credence to the applicability of non-invasive strategies in providing interventions for promoting healthy dietary and lifestyle choices, improving quality of life, and engaging personalized medicine approaches for NAFLD prevention and management. The strength of this study includes using independently validated sex-specific pro-inflammatory models for predicting hs-CRP score, multiple sensitivity analyses to test the robustness of our findings, and ultrasound assessment of NAFLD. To enhance the accuracy of the predicted pro-inflammatory hs-CRP score, lifestyle factors, long-term diet, and BMI were considered during its development. Additionally, participants in this study presented with plasma hs-CRP profiles of <10 mg/L, thereby ruling out the potential influence of recent infections in our findings. This was important to ensure that the predicted score specifically captured the influence of chronic factors rather than acute inflammatory responses [14].
This study has some limitations. A cross-sectional study design may not be optimal for identifying the temporal relationship. However, participants provided information on habitual dietary intake prior to undergoing the liver ultrasound procedure, suggesting a potential to emphasize the role of pro-inflammatory dietary and lifestyle factors in the development of NAFLD. The presence of measurement error inherent in questionnaire assessment cannot be ruled out. Also, the predicted pro-inflammatory hs-CRP score was validated exclusively in a Korean population and validation in other populations with different dietary and lifestyle habits might be necessary. Future studies should consider testing the validity of the predicted pro-inflammatory hs-CRP score in other hepatocellular-related complications.
Conclusions
We found that predicted pro-inflammatory hs-CRP score (derived from pro-inflammatory demographic, lifestyle, dietary, and anthropometric factors) was associated with higher odds of ultrasound-measured NAFLD in a Korean population.
Supplementary Material
Acknowledgements
The authors are indebted to the staff and volunteers of this study for their commitment and dedication. The study was carried out according to the Helsinki Declaration. Ethical approval for this study was approved by the institutional review board of the Seoul National University Hospital (No. 2111–066–1271). All participants provided written informed consent.
Contributor Information
Akinkunmi Paul Okekunle, Department of Food and Nutrition, College of Human Ecology, Seoul National University, Seoul, Korea; Research Institute of Human Ecology, Seoul National University, Seoul, Korea.
Jiyoung Youn, Department of Food and Nutrition, College of Human Ecology, Seoul National University, Seoul, Korea.
Sihan Song, Department of Food and Nutrition, College of Human Ecology, Seoul National University, Seoul, Korea.
Goh Eun Chung, Healthcare System Gangnam Center, Healthcare Research Institute, Seoul National University Hospital, Seoul, Korea.
Sun Young Yang, Healthcare System Gangnam Center, Healthcare Research Institute, Seoul National University Hospital, Seoul, Korea.
Young Sun Kim, Healthcare System Gangnam Center, Healthcare Research Institute, Seoul National University Hospital, Seoul, Korea.
Jung Eun Lee, Department of Food and Nutrition, College of Human Ecology, Seoul National University, Seoul, Korea; Healthcare System Gangnam Center, Healthcare Research Institute, Seoul National University Hospital, Seoul, Korea.
Supplementary data
Supplementary data is available at Gastroenterology Report online.
Authors’ contributions
A.P.O. and J.E.L. designed the study. A.P.O. and J.Y. performed all statistical analyses. A.P.O. wrote the first draft. Y.S.K. and J.E.L. revised the manuscript for important intellectual content. J.Y., S.S., G.E.C., and S.Y.Y. contributed to the draft editing and revision. All authors read and approved the final published version of the manuscript.
Funding
The Brain Pool Program supported A.P.O. through the National Research Foundation of Korea, funded by the Ministry of Science and ICT [2020H1D3A1A04081265]. The funders had no role in the study design, data collection and analysis, publication decision, or manuscript preparation.
Conflict of Interest
None declared.
Data availability
The data presented in this study are available on request from the corresponding author.
References
- 1. Younossi ZM, Koenig AB, Abdelatif D et al. Global epidemiology of nonalcoholic fatty liver disease—meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology 2016;64:73–84. [DOI] [PubMed] [Google Scholar]
- 2. Friedman SL, Neuschwander-Tetri BA, Rinella M et al. Mechanisms of NAFLD development and therapeutic strategies. Nat Med 2018;24:908–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Younossi Z, Anstee QM, Marietti M et al. Global burden of NAFLD and NASH: trends, predictions, risk factors and prevention. Nat Rev Gastroenterol Hepatol 2018;15:11–20. [DOI] [PubMed] [Google Scholar]
- 4. Younossi ZM, Henry L. The global burden of non-alcoholic steatohepatitis. Arch cuba gastroenterol 2020;1(1). [Google Scholar]
- 5. Younossi ZM. Non-alcoholic fatty liver diseaseA global public health perspective. J Hepatol 2019;70:531–44. [DOI] [PubMed] [Google Scholar]
- 6. Lee J, Yoon K, Ryu S et al. High-normal levels of hs-CRP predict the development of non-alcoholic fatty liver in healthy men. PLoS One 2017;12:e0172666. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Lu F-B, Hu E-D, Xu L-M et al. The relationship between obesity and the severity of non-alcoholic fatty liver disease: systematic review and meta-analysis. Expert Rev Gastroenterol Hepatol 2018;12:491–502. [DOI] [PubMed] [Google Scholar]
- 8. Fedewa MV, Hathaway ED, Ward-Ritacco CL. Effect of exercise training on C reactive protein: a systematic review and meta-analysis of randomised and non-randomised controlled trials. Br J Sports Med 2017;51:670–6. [DOI] [PubMed] [Google Scholar]
- 9. Muscatell KA, Brosso SN, Humphreys KL. Socioeconomic status and inflammation: a meta-analysis. Mol Psychiatry 2020;25:2189–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Chrysohoou C, Panagiotakos DB, Pitsavos C et al. Adherence to the Mediterranean diet attenuates inflammation and coagulation process in healthy adults: The Attica study. J Am Coll Cardiol 2004;44:152–8. [DOI] [PubMed] [Google Scholar]
- 11. Barbaresko J, Koch M, Schulze MB et al. Dietary pattern analysis and biomarkers of low-grade inflammation: a systematic literature review. Nutr Rev 2013;71:511–27. [DOI] [PubMed] [Google Scholar]
- 12. Cantero I, Abete I, Babio N et al. Dietary Inflammatory Index and liver status in subjects with different adiposity levels within the PREDIMED trial. Clin Nutr 2018;37:1736–43. [DOI] [PubMed] [Google Scholar]
- 13. Ramírez-Vélez R, García-Hermoso A, Izquierdo M et al. The Dietary Inflammatory Index and hepatic health in the US adult population. J Hum Nutr Diet 2022;35:968–79. [DOI] [PubMed] [Google Scholar]
- 14. Kim S, Song S, Kim YS et al. The association between predicted inflammatory status and colorectal adenoma. Sci Rep 2020;10:2433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Chalasani N, Younossi Z, Lavine JE et al. The diagnosis and management of nonalcoholic fatty liver disease: Practice guidance from the American Association for the Study of Liver Diseases. Hepatology 2018;67:328–57. [DOI] [PubMed] [Google Scholar]
- 16. Saadeh S, Younossi ZM, Remer EM et al. The utility of radiological imaging in nonalcoholic fatty liver disease. Gastroenterology 2002;123:745–50. [DOI] [PubMed] [Google Scholar]
- 17. Ahn Y, Kwon E, Shim JE et al. Validation and reproducibility of food frequency questionnaire for Korean genome epidemiologic study. Eur J Clin Nutr 2007;61:1435–41. [DOI] [PubMed] [Google Scholar]
- 18. Chung GE, Youn J, Kim YS et al. Dietary patterns are associated with the prevalence of nonalcoholic fatty liver disease in Korean adults. Nutrition 2019;62:32–8. [DOI] [PubMed] [Google Scholar]
- 19. Ainsworth BE, Haskell WL, Herrmann SD et al. 2011 compendium of physical activities: a second update of codes and MET values. Med Sci Sports Exerc 2011;43:1575–81. [DOI] [PubMed] [Google Scholar]
- 20. Chobanian AV, Bakris GL, Black HR et al. ; National High Blood Pressure Education Program Coordinating Committee. Seventh report of the joint national committee on prevention, detection, evaluation, and treatment of high blood pressure. Hypertension 2003;42:1206–52. [DOI] [PubMed] [Google Scholar]
- 21. Alberti KG, Zimmet PZ; WHO Consultation. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med 1998;15:539–53. [DOI] [PubMed] [Google Scholar]
- 22. Health Examinees Study Group . The Health Examinees (HEXA) Study: rationale, study design and baseline characteristics. Asian Pac J Cancer Prev 2015;16:1591–7. [DOI] [PubMed] [Google Scholar]
- 23. Vaccarino V, Badimon L, Corti R et al. ; Working Group on Coronary Pathophysiology and Microcirculation. Ischaemic heart disease in women: are there sex differences in pathophysiology and risk factors?: Position Paper from the Working Group on Coronary Pathophysiology and Microcirculation of the European Society of Cardiology. Cardiovasc Res 2011;90:9–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Donahue RP, Rejman K, Rafalson LB et al. Sex differences in endothelial function markers before conversion to pre-diabetes: does the clock start ticking earlier among women?: the Western New York Study. Diabetes Care 2007;30:354–9. [DOI] [PubMed] [Google Scholar]
- 25. Gao Z, Chen Z, Sun A et al. Gender differences in cardiovascular disease. Med Novel Technol Dev 2019;4:100025. [Google Scholar]
- 26. Yoneda M, Mawatari H, Fujita K et al. High-sensitivity C-reactive protein is an independent clinical feature of nonalcoholic steatohepatitis (NASH) and also of the severity of fibrosis in NASH. J Gastroenterol 2007;42:573–82. [DOI] [PubMed] [Google Scholar]
- 27. Yu E, Hsu H-Y, Huang C-Y et al. Inflammatory biomarkers and risk of atherosclerotic cardiovascular disease. Open Med (Wars) 2018;13:208–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Kumar R, Porwal YC, Dev N et al. Association of high-sensitivity C-reactive protein (hs-CRP) with non-alcoholic fatty liver disease (NAFLD) in Asian Indians: a cross-sectional study. J Family Med Prim Care 2020;9:390–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Fricker ZP, Pedley A, Massaro JM et al. Liver fat is associated with markers of inflammation and oxidative stress in analysis of data from the Framingham Heart Study. Clin Gastroenterol Hepatol 2019;17:1157–64.e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Okada E, Shirakawa T, Shivappa N et al. Dietary inflammatory index is associated with risk of all-cause and cardiovascular disease mortality but not with cancer mortality in middle-aged and older Japanese Adults. J Nutr 2019;149:1451–9. [DOI] [PubMed] [Google Scholar]
- 31. Wang Q, Sun Y, Xu Q et al. Higher dietary inflammation potential and certain dietary patterns are associated with polycystic ovary syndrome risk in China: a case-control study. Nutr Res 2022;100:1–18. [DOI] [PubMed] [Google Scholar]
- 32. Ye C, Huang X, Wang R et al. Dietary Inflammatory Index and the Risk of Hyperuricemia: A Cross-Sectional Study in Chinese Adult Residents. Nutrients 2021;13:4504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Tavakoli A, Mirzababaei A, Moosavi H et al. Dietary inflammatory index (DII) may be associated with hypertriglyceridemia waist circumference phenotype in overweight and obese Iranian women: a cross sectional study. BMC Res Notes 2021;14:312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Klisic A, Kavaric N, Ninic A et al. Oxidative stress and cardiometabolic biomarkers in patients with non-alcoholic fatty liver disease. Sci Rep 2021;11:18455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Chang Y, Ryu S, Sung K-C et al. Alcoholic and non-alcoholic fatty liver disease and associations with coronary artery calcification: evidence from the Kangbuk Samsung Health Study. Gut 2019;68:1667–75. [DOI] [PubMed] [Google Scholar]
- 36. Koo BK, Joo SK, Kim D et al. Development and validation of a scoring system, based on genetic and clinical factors, to determine risk of steatohepatitis in Asian patients with nonalcoholic fatty liver disease. Clin Gastroenterol Hepatol 2020;18:2592–9.e10. [DOI] [PubMed] [Google Scholar]
- 37. Chen Q, Li Q, Li D et al. Association between liver fibrosis scores and the risk of mortality among patients with coronary artery disease. Atherosclerosis 2020;299:45–52. [DOI] [PubMed] [Google Scholar]
- 38. Lee H, Lee IS, Choue R. Obesity, inflammation and diet. Pediatr Gastroenterol Hepatol Nutr 2013;16:143–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Jacquemyn J, Cascalho A, Goodchild RE. The ins and outs of endoplasmic reticulum-controlled lipid biosynthesis. EMBO Rep 2017;18:1905–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Ahima RS, Flier JS. Adipose tissue as an endocrine organ. Trends Endocrinol Metab 2000;11:327–32. [DOI] [PubMed] [Google Scholar]
- 41. Singh RK, Chang H-W, Yan D et al. Influence of diet on the gut microbiome and implications for human health. J Transl Med 2017;15:73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Hariharan R, Odjidja EN, Scott D et al. The dietary inflammatory index, obesity, type 2 diabetes, and cardiovascular risk factors and diseases. Obes Rev 2022;23:e13349. [DOI] [PubMed] [Google Scholar]
- 43. Topping DL, Clifton PM. Short-chain fatty acids and human colonic function: roles of resistant starch and nonstarch polysaccharides. Physiol Rev 2001;81:1031–64. [DOI] [PubMed] [Google Scholar]
- 44. Gurung M, Li Z, You H et al. Role of gut microbiota in type 2 diabetes pathophysiology. EBioMedicine 2020;51:102590. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Naderpoor N, Mousa A, Gomez-Arango LF et al. Faecal microbiota are related to insulin sensitivity and secretion in overweight or obese adults. J Clin Med 2019;8:452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Turnbaugh PJ. Microbes and diet-induced obesity: fast, cheap, and out of control. Cell Host Microbe 2017;21:278–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Hodge AM, Karim MN, Hébert JR et al. Diet scores and prediction of general and abdominal obesity in the Melbourne collaborative cohort study. Public Health Nutr 2021;24:6157–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Khan I, Kwon M, Shivappa N et al. Proinflammatory dietary intake is associated with increased risk of metabolic syndrome and its components: results from the population-based prospective study. Nutrients 2020;12:1196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Andrade PA, Hermsdorff HHM, Leite JIA et al. Baseline pro-inflammatory diet is inversely associated with change in weight and body fat 6 months following-up to bariatric surgery. Obes Surg 2019;29:457–63. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data presented in this study are available on request from the corresponding author.