1. Introduction
Nonalcoholic fatty liver disease (NAFLD) is among the most common causes of chronic liver disease worldwide and is a growing medical problem in industrialized countries [
1]. A wide spectrum of histological changes has been observed in NAFLD, ranging from nonalcoholic fatty liver (NAFL), which is generally non-progressive, to nonalcoholic steatohepatitis (NASH). A proportion of patients with NASH develop cirrhosis and hepatocellular carcinoma [
2]. Approximately 30% of the general population has NAFLD, and up to 5% of this population has NASH [
3,
4]. The prevalence of NAFLD is reported to be 20–40% in Western countries and 12–30% in Asian countries [
5,
6]. Annual health checkup data show that 9–30% of Japanese adults have ultrasonography (US)-diagnosed NAFLD, and NASH is diagnosed in 10–20% or more of NAFLD cases [
5,
6,
7]. There are more than 20 million NAFLD patients in Japan, and it is feared that this number will increase in the future [
8,
9].
Lifestyle modification, including reduction of calorie intake and/or increased physical activity to achieve weight loss, should be advised for all NAFLD patients [
10,
11,
12]. Weight loss induces liver fat reduction and improves the liver pathophysiological status of NAFLD patients. NAFLD is an independent risk factor for cardiovascular disease (CVD) and cancer of the liver and other organs. Weight loss improves type 2 diabetes mellitus, dyslipidemia, and hypertension. However, whether weight reduction can improve cancer development is unknown. Loss of >=5% total body weight (TBW) improves hepatic steatosis, >=7% TBW induces histological NASH resolution, and >=10% TBW leads to liver fibrosis regression [
13,
14,
15]. In revised Japanese NAFLD guidelines, >=7% weight reduction is also recommended for obese NASH patients [
16]. However, the achievement rates for these reductions are as low as 30%, 18%, and 10%, respectively [
15]. To achieve weight loss, various kinds of medical and multidisciplinary methodology should be combined to support patient behavior change [
17]. The TBW reduction rate in non-obese NAFLD patients [body mass index (BMI) < 25] is considered to be lower than in obese NAFLD patients. Half of patients have been shown to achieve NAFLD remission (defined as an intrahepatic triglyceride content below 5.0% by proton-magnetic resonance spectroscopy) with 3–5% weight reduction within 12-months of lifestyle intervention in non-obese NAFLD patients, and the same was achieved with 7–10% TBW reduction in obese NAFLD patients [
18].
Although there is a wealth of information about NAFLD worldwide, there is no study that investigates the accurate relationships between nutritional intake and disease progression in biopsy-confirmed NAFLD patients in Japan. In this study, we conducted an observation study that elucidates the accurate nutritional intake using photographs of meals and nutrition interviews in biopsy-confirmed NAFLD patients.
2. Materials and Methods
2.1. Ethical Committee Approval
The protocol and informed consent were approved as a multicenter study by each of the following institutional review boards; Osaka University Hospital, Saga University Hospital, JA Hiroshima Kouseiren General Hospital, and Yokohama City University Hospital. Written informed consent was obtained from all subjects at the time of liver biopsy or enrollment in each institute, and the study was conducted in accordance with the Helsinki Declaration.
2.2. Study Subjects
A total of 37 patients of biopsy-confirmed NAFLD patients were enrolled in 2020 during the registration period in this study from three hepatology centers in Japan; namely, Saga University Hospital, JA Hiroshima Kouseiren General Hospital, and Yokohama City University Hospital. All biopsy-confirmed NAFLD patients in this study had undergone a percutaneous liver needle biopsy. The indication of liver biopsy was routinely performed for the diagnosis of liver histology in each center. In order to elucidate the recent nutritional intake of NAFLD patients, we set the study registration period to six months. The biopsied liver samples were embedded in paraffin blocks according to standard procedures and stained with hematoxylin and eosin and Masson’s trichrome stains. All biopsy specimens were centrally evaluated by an experienced pathologist (S.A.) who was blinded to the clinical data. Adequate liver samples were defined as > 1.5 cm long and/or having more than six portal tracts. NASH was confirmed according to Matteoni’s classification [
19]. NAFLD patients with ballooning hepatocytes (Matteoni type 3) and NAFLD patients with liver fibrosis (Matteoni type 4) were placed in the NASH cohort. Patients whose liver biopsy specimens showed simple steatosis or steatosis with non-specific inflammation were placed in the NAFL cohort. Samples were also investigated and quantified according to NAFLD activity scoring (NAS) [
20]. Steatosis (0–3), lobular inflammation (0–2), and hepatocellular ballooning (0–2) were quantified. The individual parameters of fibrosis were scored independently according to the NASH Clinical Research Network scoring system [
20]. Early fibrosis was classified as a stage 0–2 (F0–2) and advanced fibrosis was classified as a stage 3–4 (F3–4) disease. The exclusion criteria for this study included a history of other hepatic diseases, a substance abuse-induced hepatic disorder, and a history of alcohol abuse (defined as >20 g of alcohol daily). Based on the median BMI of 30.3, NAFLD patients were divided into two groups; namely, the BMI Low group (BMI <=30.3) and the BMI High group (BMI > 30.3).
2.3. Study Design
This was a cohort study that prospectively followed the nutritional intake of subjects with a confirmed diagnosis of NAFLD.
2.4. Nutritional Analysis Method
Each subject recorded dietary intake of nutrients for seven consecutive days using a dietary questionnaire and photographs of each diet. A dietician analyzed and quantified the nutritional data collected by each NAFLD patient. For the food ingredients list frequency survey sheet used (
Figure 1a,
Supplementary Figure S1), we created and used our own sheet for each meal that allows the names of commonly eaten ingredients to be checked based on the classification of the food composition table.
For the evaluation of nutrient intake, the Food Composition Database of the Ministry of Education, Culture, Sports, Science and Technology (MEXT) was used in this study. Data from photographs taken by patients of everything they ate and drank, with a ruler placed as a guide to determine size (
Figure 1b), were checked with the food ingredients list frequency survey sheet, which was checked at each meal, and nutritional calculations were made using the Ministry of Education’s food database. In order to ensure a reliable understanding of what was eaten, the photographs of each food and the food ingredients list frequency survey sheet were used together in this study.
2.5. Anthropometry and Laboratory Measurements
Anthropometric variables (height and weight) were measured in the standing position, and BMI was calculated as weight (in kg) divided by the square of height in meters (m2). Serum biochemical variables [aspartate aminotransferase (AST), alanine aminotransferase (ALT), γ-glutamyltransferase (GGT), total cholesterol (T-Cho), triglyceride (TG), total protein (TP), albumin (Alb), iron (Fe), creatinine (Cre), hemoglobin A1c (HbA1c), red blood cell count (RBC), and platelet count (Plt)] were measured with a conventional automated analyzer.
2.6. Statistical Analysis
Two analyses were used: an internal comparison within NAFLD patients and an external comparison between NAFLD patients and the Nutrition Examination Survey control population. Continuous variables are summarized as mean and standard deviation (S.D.), and categorical variables are summarized as the number and proportion. The Welch’s t-test and Pearson chi-square test or Fisher exact test were used as appropriate for internal comparison within NAFLD patients. The Welch’s t-test was used to examine differences in nutritional intake status between NAFLD patients and Nutrition Examination Survey individuals. Because individual information was not available for the nutrition survey, t-values and degrees of freedom were calculated based on aggregate information of mean, standard deviation, and number of subjects as shown in the formula below [
21,
22].
Two-sided p values of <0.05 were considered statistically significant. All data were statistically analyzed using JMP statistical software (version 16.1; SAS Institute Inc., Cary, NC, USA).
4. Discussion
The results of this study show an excess or deficiency of multiple nutrients in NAFLD patients compared with control subjects. We compared the nutritional intake data with Japanese dietary guidelines for nutrient intake (
https://www.mhlw.go.jp/stf/newpage_22536.html (accessed on 2 July 2021)). There are excesses and deficiencies compared with recommendation in nutrition intake of NAFLD patients in our study. In male NAFLD patients, lipid intake was higher than in control male subjects. In female NAFLD patients, zinc and copper intakes were lower than in control female subjects. In both male and female NAFLD patients, niacin and dietary fiber intakes were lower than in controls. In NAFLD patients with high BMI, total intake calories and SFA intake were higher than in NAFLD patients with low BMI. Interestingly, according to the nutrient intake corrected by total caloric intake, SFA intake was higher, and copper and Vit E intakes were lower in NAFLD patients with high BMI than in patients with low BMI. These findings indicate that NAFLD patients have an unbalanced nutritional intake, a fact that may contribute to the progression of NAFLD.
Underreporting is a common issue in dietary investigations [
23], and might have effect on our study. For example, total calorie intake was not significantly different between NAFLD patients and control subjects in this study. This might be due to underreporting of study subjects. Although this issue would present, we analyzed based on the obtained data in this study, and we found several nutritional intakes were lower in NAFLD patients. Deficiency of Vit A and copper intakes were lower in NAFLD patients than controls, and even lower in patients with advanced fibrosis NAFLD patients. Interestingly, copper and Vit A intakes were further decreased in advanced liver fibrosis NAFLD patients than in early fibrosis NAFLD patients. Liver is the main storage organ of Vit A in animals and human [
24,
25], and hepatic stellate cell (HSC) is the main storage site of Vit A [
26]. Quiescent HSCs store vitamin A in lipid droplets, and activated HSCs lose lipid droplets containing Vit A and produce extracellular matrix proteins, leading to liver fibrosis [
27]. Our study demonstrates that Vit A intake was lower in NAFLD patients than in control subjects and further decreased in advanced NAFLD patients. This finding indicates that lower Vit A intake would lead to liver fibrosis progression via hepatic Vit A deficiency in NAFLD patients. Our study also demonstrated Vit E intake was lower in advanced liver fibrosis NAFLD patients than in early fibrosis NAFLD patients. The effects of Vit E on NAFLD were demonstrated in the PIVENS (Pioglitazone vs Vitamin E vs Placebo for Treatment of Non-Diabetic Patients with Nonalcoholic Steatohepatitis) trial [
28]. The efficacy of vitamin E on NAFLD/NASH is mainly due to its antioxidant effects. Lower intake of Vit E in advanced fibrosis patients would enhance disease severity in NAFLD patients.
Copper is a critical part of several cellular processes, including antioxidant function [
29]. Cu/Zn superoxide dismutase (SOD1) reduces radical oxygen species by the redox cycling activity of copper in its catalytic center. In our study, not only copper, but also zinc intake was lower in NAFLD patients than in controls. In particular, both copper and zinc intake were significantly reduced in female NAFLD patients compared with in control female subjects. These findings indicate that the deficiency of copper and/or zinc could lead to NAFLD progression.
Niacin and dietary fiber intakes were also lower in NAFLD patients than in controls in this study. Gender analysis revealed that these intakes were decreased in both male and female NAFLD subjects compared with male and female control subjects. Niacin prevents hepatic steatosis progression in rodent models [
30,
31]. Niacin also inhibits fatty acid flux from adipose tissue to liver, resulting in the reduction of hepatic triglyceride synthesis and an increase hepatic lipid oxidation. A recent lifestyle intervention study using
1H-MR spectroscopy demonstrated that high niacin intake decreases liver fat in a dose-dependent manner [
32]. Dietary fiber intake could reduce NAFLD risk [
12]. Another recent study demonstrated that increasing consumption of dietary fiber might reduce the risk of NAFLD and its progression [
33]. In a Chinese study, dietary intake of total dietary fiber was shown to reduce NAFLD risk in a dose-dependent manner [
34]. In our study, dietary fiber intake, especially insoluble dietary fiber intake, was lower in NAFLD patients than in controls. This trend was stronger for females than males, and the total dietary fiber intake was also lower in female NAFLD patients.
Our analysis according to BMI demonstrates that total calorie and cholesterol intakes tended to be higher in the BMI High group than in the BMI Low group and that the intake of fat, especially SFA, was significantly higher in the BMI High group than the BMI Low group. Comparison of the ratio of each nutrient to total caloric intake, to examine biases in nutritional intake, revealed that several nutrient intake ratios were lower in the BMI High group. Interestingly, intake ratios of copper and Vit E were significantly lower in the BMI High group than the BMI Low group. These findings indicate that uneven nutritional intake could promote the progression of NAFLD. In addition, we compared the nutritional intake between NAFLD subjects with or without type 2 diabetes mellitus (T2DM). T2DM was diagnosed as HbA1c ≥ 6.5%, fasting blood sugar ≥ 126 mg/dL, or treatment with anti-diabetic drugs according to the Japanese clinical practice guideline for diabetes [
35]. Among 37 NAFLD patients, 21 subjects were diagnosed as T2DM. We compared the nutritional intake in NAFLD patients with or without T2DM. There are no significant differences between the two groups (data not shown). T2DM had no effects on the nutritional intake in our study subjects.
Our study has several limitations. First, due to the short time period of the nutritional intake survey (1 week), it is unclear whether it accurately reflects the patients’ usual nutritional intake. Second, the number of NAFLD patients in our study was relatively small. Third, we did not measure patatin-like phospholipase domain-containing protein 3 (
PNPLA3) gene polymorphism, which is more common in Asian than in Western populations [
9]. This gene polymorphism has homozygous mutations in approximately 20% of the general Japanese population [
36] and is associated with NAFLD onset and progression [
37,
38]. Forth, the registration period of our study was during COVID (coronavirus disease)-19 pandemic and the data of control subjects were collected before pandemic. Recently, we investigated the lifestyle changes between before and during COVID19 pandemic [
39]. In the study, we found new metabolic dysfunction-associated fatty liver disease (MAFLD) diagnoses and daily alcohol intake increased during pandemic. Although NAFLD patients in the present study did not increase alcohol intake, there might have some additional effects of pandemic on our study results. Fifth, our control data were not collected with the same questionnaire, nor our control subjects have not taken pictures. Since the survey method is to sequentially weigh and record dietary contents rather than having the researcher interview and recall them, we think the survey method is equivalent to taking a photograph of each meal. The National Health and Nutrition Survey (NHNS) used in this study as a control group for comparison includes not only nutritional intake but also lifestyle habits and health examinations (blood data) as used in many other studies [
40,
41]. This survey, which uses cluster sampling to select and conduct a full survey, has more external validity than using healthy volunteers as the control group because it includes a wider range of subjects, including latent conditions. Despite these limitations, a relationship between inadequate intake of some nutrients and NAFLD disease progression was noted.