Sage Journals: Discover world-class research

Abstract

Background:

Although growing evidence suggests that fructose intake contributes to the development of non-alcoholic fatty liver disease (NAFLD), fructose intake in NAFLD patients has not been documented locally.

Objectives:

The objectives of this study were to compare fructose intake between NAFLD patients and controls with chronic hepatitis B, and to ascertain whether fructose intake was associated with the presence of NAFLD.

Methods:

This was a cross-sectional, case–control study. Thirty-four patients diagnosed with NAFLD and 34 controls with chronic hepatitis B participated in this study between 2012 and 2014 in the Singapore General Hospital outpatient setting. Fructose, energy and nutrient intake were assessed by using food frequency questionnaires.

Results:

NAFLD patients had higher body mass index (28.6±4.0 vs. 22.5±3.9 kg/m², p<0.001) and waist circumference (100.0±7.9 vs. 80.2±11.7 cm, p<0.001) than controls. Cases reported higher intakes of energy (2378±708 vs. 1796±398 kcal, p<0.001), protein (109±37 vs. 84±25 g, p<0.01), fat (87±33 vs. 62±19 g, p<0.001), total carbohydrate (294±83 vs. 232±63 g, p<0.001) and fructose (42±17 vs. 31±15 g, p<0.05). Logistic regression analysis showed waist circumference (odds ratio: 1.25; 95% confidence interval: 1.11–1.41; p<0.001) and energy intake (odds ratio: 1.002; 95% confidence interval: 1.001–1.004; p<0.05) were significant risk factors for NAFLD.

Conclusions:

Total calorie, macronutrient and fructose intake in NAFLD patients were significantly higher than in controls. Waist circumference and energy intake were significantly associated with NAFLD. Reducing total calorie intake and adopting healthy eating habits should be emphasized to NAFLD patients to manage their clinical conditions.

Keywords

Fructose dietary intake non-alcoholic fatty liver disease

Introduction

Non-alcoholic fatty liver disease (NAFLD) is a leading cause of chronic liver disease worldwide, with reported prevalence between 8.7% and 51%.¹ It involves a wide spectrum of clinical phenotypes including steatosis, non-alcoholic steatohepatitis and cirrhosis.²

NAFLD is a heterogeneous disease, with both genetic and environmental factors implicated in its pathogenesis. The most commonly reported risk factors for NAFLD are obesity and metabolic syndrome,³ with recent interest focusing mainly on the role of dietary fructose in the pathogenesis of NAFLD.

Fructose is a natural sugar found in fruit and honey. It is also present in common sweeteners such as sucrose (containing 50% fructose and 50% glucose) and high-fructose corn syrup (HFCS), with its fructose content typically at 55%. The mechanism of fructose in inducing NAFLD was proposed from animal data where fructose induced the activities of major transcriptional regulators of hepatic lipogenesis and subsequently promoted de novo lipogenesis in the liver.⁴ A few other feeding studies on rats established that fructose indeed increased lipogenic enzyme activities and blood triglyceride levels.^5–7 Observational human studies have also demonstrated the association of fructose intake with the presence of NAFLD.^8–10 Ouyang et al. observed a two- to three-fold increase in fructose consumption among patients with NAFLD when compared with controls.⁸ Zelber-Sagi et al. reported that compared with healthy controls, NAFLD patients consumed twice the amount of soft drinks, with fructose as the main sweetener.⁹ In another study, Abdelmalek et al. established that consuming seven servings and above per week of fructose consumption was associated with higher fibrosis stage in NAFLD patients.¹⁰ On the other hand, data from controlled trials have reported conflicting results. Two recent systematic reviews and meta-analyses of controlled feeding trials failed to support fructose-induced NAFLD changes when it was isocalorically exchanged for other carbohydrates in participants.^11,12 Serum markers of NAFLD only elevated when diets were supplemented with fructose, providing excessive calories. One postulation was that this effect was likely to be confounded by excessive energy intake, rather than fructose per se.^11,12

Studies on the association between fructose and NAFLD were primarily from western countries. There is a paucity of data about NAFLD and dietary habits, in particular, fructose intake in the Asian context. Notably, dietary patterns in Asia may vary widely from western diets. As such, we conducted this study to compare the fructose intake between adult patients with NAFLD and controls with chronic hepatitis B, and to ascertain whether fructose intake was associated with NAFLD.

Methods

This was a case–control study conducted from July 2012 to April 2014. Cases were patients diagnosed with NAFLD, while controls were patients with inactive chronic hepatitis B (CHB). Patients aged 21 years and above without other concomitant chronic liver diseases were recruited consecutively at the gastroenterology outpatient clinics at the Singapore General Hospital.

The inclusion criteria for NAFLD patients were: most recent liver enzymes with at least 1.2 times of upper limit of normal value (alanine transaminase (ALT) ⩾43 U/L or aspartate transaminase (AST) ⩾39 U/L) in the past six months and ultrasound evidence of hepatic steatosis. The normal ranges of liver enzymes (ALT: 7–36 U/L; AST: 15–33 U/L) were based on the hospital laboratory references. Patients with other contributory causes of liver disease, such as significant alcohol consumption (exceeding 20 g alcohol/day in women and exceeding 30 g alcohol/day in men),¹³ hepatotoxic drug history, chronic viral hepatitis B or C, autoimmune hepatitis and Wilson’s disease were excluded. Similarly, patients with other medical conditions requiring sugar intake restriction (e.g. diabetes or pre-diabetes) and patients who had made major changes to their dietary habits after the diagnosis of NAFLD were also excluded.

Controls were otherwise healthy inactive CHB patients with normal liver function tests. As there are no established dietary recommendations or dietary restrictions for CHB management, the dietary habits in CHB patients were assumed to be representative of healthy people. Patients were excluded if they had chronic liver diseases other than CHB (e.g. evidence of hepatic steatosis on ultrasonography, autoimmune hepatitis, hepatitis C or Wilson’s disease) or alcohol consumption exceeding 20 g/day in women and 30 g/day in men.¹³ Patients with other medical conditions requiring restriction of sugar intake (e.g. diabetes or pre-diabetes) or who had made major dietary changes in the past six months were also excluded.

Data collection included demographic, anthropometric and nutrition intake data in all patients.

Demographic data including age, gender and ethnicity were obtained from patients’ medical records. Body weight (without shoes and in light clothing) and height were taken to the nearest 0.1 kg and 0.1 cm, respectively. Waist and hip circumference were based on the measurement guidelines from the World Health Organization.¹⁴ Body mass index (BMI) and waist–hip ratio (WHR) were calculated based on the above measurements: BMI (kg/m²)=weight/height², WHR=waist circumference/hip circumference.

A food frequency questionnaire (FFQ) was administered to each patient face-to-face by a single dietitian, who was blinded to patients’ case–control status and clinical conditions. The FFQ was adapted from the validated questionnaire used in the Singapore National Health Survey 2004,¹⁵ and contained 370 commonly consumed food items and beverages. A number of dietary aids were used during the face-to-face interview to facilitate the quantification of foods and beverages consumed by the patients. Household measures such as bowls, cups, glasses, spoons, et cetera and food models of actual portion sizes were used to help patients estimate the amounts of foods and beverages consumed. The frequency of consumption for each food or beverage item was reported on a daily, weekly or monthly basis, where applicable, during the past 12 months.

For each FFQ item, the energy and nutrient values were obtained from the online database of Energy & Nutrient Composition of Food from Singapore Health Promotion Board¹⁶ and incorporated into a spreadsheet. Daily intakes of total energy, protein, fat, total carbohydrate, sugar, added sugar and fructose were computed based on the frequency of each food or beverage intake indicated in the FFQ. ‘Added sugar’ was defined as sugar that was added to drinks or foods during manufacturing, cooking or at the table. ‘Sugar’ included added sugar as well as natural sugar from fruit, vegetables, natural juice and honey. ‘Fructose’ encompassed sources from both natural and processed foods. As the fructose content values for fruit and fruit juice were not available from the Singapore Health Promotion Board, these were obtained from Food Standards Australia and New Zealand (online database), where free fructose and sucrose values per 100 g of fruit and fruit juice were available.¹⁶ The total fructose content per 100 g of fruit or juice was the sum of free fructose and half of the sucrose values per 100 g. As sucrose was the most commonly used sweetener for local sweetened beverages, sweets or desserts, fructose content for those food or beverages was estimated as 50% of the sucrose level. All dietary data were entered and analyzed by the same dietitian.

Statistical analysis was performed using IBM SPSS Statistics (Version 21.0, 2012, IBM Corp., Armonk, NY, USA). Continuous variables were expressed as means and standard deviations (SDs) while categorical variables were expressed as frequencies and percentages (%). Baseline characteristics and differences between NAFLD patients and controls were explored, with continuous variables analyzed using independent sample t-test and categorical variables analyzed using Chi-square test. Logistic regression analysis was performed to determine independent associations between variables and the presence NAFLD. A value of p<0.05 was considered statistically significant for all analyses.

Sample sizes of 34 patients with NAFLD and 34 with CHB were required to achieve an 80% power to detect a difference of 10 g of fructose intake between the two groups, with estimated standard deviation of 10 g in each group and with a significance level, alpha of 0.05 using a two-sided two-sample t-test.

The study protocol was approved by the Centralised Institutional Review Board, Singhealth Office of Research. All patients provided written informed consent forms.

Results

Thirty-four patients in each arm of NAFLD and CHB patients who met the inclusion criteria were included for analysis.

There were no statistically significant differences in age and ethnicity distribution between the two groups (Table 1). However, the proportion of men in the NAFLD group was remarkably higher than the CHB group. NAFLD patients also had noticeably greater mean BMI and waist circumference when compared with CHB patients, with nearly all patients (97%) in the NAFLD group being centrally obese.

Table 1.

Demographics and anthropometric measurements in patients with NAFLD and CHB.

Characteristics	NAFLDn=34	CHBn=34	p-value^a
Age, years, mean±SD	49±10	48±11	0.86
Men, n (%)	28 (82)	10 (29)	<0.001
Ethnicity, n (%)
Chinese	30 (88)	30 (88)	1.00
Others	4 (12)	4 (12)
Height, cm, mean±SD	167.9±9.5	162.3±7.7	0.009
Weight, kg, mean±SD	80.7±12	59.3±12	<0.001
BMI, kg/m², mean±SD	28.6±4.0	22.5±3.9	<0.001
Waist circumference, cm, mean±SD	100±8	80±12	<0.001
Hip circumference, cm, mean±SD	105±8	98±8	<0.001
Waist-hip ratio, mean±SD	0.95±0.08	0.82±0.08	<0.001
Obesity^b, n (%)	20 (59)	3 (9)	<0.001
Abdominal obesity^c, n (%)	33 (97)	13 (38)	<0.001
ALT, U/L	83±46	21±9	<0.0001

p-value calculated using independent sample t-test for continuous variables and Fisher’s exact test (2 by 2) for categorical variables.

Obesity defined as BMI ⩾27.5 kg/m².

Abdominal obesity defined as waist circumference ⩾80 cm for women and ⩾90cm for men.

NAFLD: non-alcoholic fatty liver disease; CHB: chronic hepatitis B; SD: standard deviation; BMI: body mass index; ALT: alanine aminotransferase.

NAFLD patients reported significantly higher intakes of energy, macronutrients (protein, fat and carbohydrate), added sugar and fructose (Table 2). Compared with controls, while there were no significant differences in terms of the energy contribution (%) from protein, fat, carbohydrate and fructose to total energy intake, added sugar contributed significantly more to total energy intake in NAFLD patients (Table 2).

Table 2.

Comparison of energy, macronutrients, added sugar and fructose intake between NAFLD and CHB patients.

Characteristics	NAFLDn=34	CHBn=34	p-value^a
Energy intake, kcal/day	2378±708	1796±398	<0.001
Energy requirement, kcal/day	1961±257	1695±202	<0.001
Energy intake/requirement, %	121±31	107±26	0.048
Protein intake, g/day	109±37	84±25	0.002
Fat intake, g/day	87±33	62±19	<0.001
Carbohydrate intake, g/day	294±83	232±63	0.001
Added sugar intake, g/day	55±25	36±19	0.001
Fructose intake, g/day	42±17	31±15	0.009
% Energy intake from:
Protein, %	18±3	19±4	0.68
Fat, %	32±4	31±5	0.19
Carbohydrate, %	50±7	52±8	0.36
Added sugar, %	9.6±4.3	7.6±3.4	0.037
Fructose, %	7.2±2.8	6.7±2.7	0.514

Values are expressed as mean±SD.

p-value calculated using independent sample t-test.

NAFLD: non-alcoholic fatty liver disease; CHB: chronic hepatitis B; SD: standard deviation.

The logistic regression analysis using backward stepwise method (entered variables included age, gender, waist circumference, energy intake and fructose) showed only waist circumference (odds ratio (OR): 1.25; 95% confidence interval (CI): 1.11–1.41; p<0.001) and energy intake (OR: 1.002; 95% CI: 1.001–1.004; p<0.05) were significantly associated with NAFLD (Table 3). Fructose consumption had no significant association with NAFLD (OR: 0.99; 95% CI: 0.93–1.05; p=0.708) when age, gender, waist circumference and energy intake were adjusted for.

Table 3.

Results of backward stepwise logistic regression analysis of factors associated with non-alcoholic fatty liver disease.

Variables		Risk estimates	p-value
Waist circumference, cm	Crude OR	1.253 (95% CI: 1.120–1.401)	<0.001
	Adj. OR^a	1.253 (95% CI: 1.111–1.414)	0.001
Daily energy intake, kcal	Crude OR	1.002 (95% CI: 1.001–1.004)	0.001
	Adj. OR^a	1.002 (95% CI: 1.001–1.004)	0.01
Fructose intake, g	Crude OR	1.042 (95% CI: 1.009–1.076)	0.013
	Adj. OR^a	0.988 (95% CI: 0.927–1.053)	0.708

Adjusted for age, gender. Additionally adjusted for energy intake, fructose intake and waist circumference as appropriate.

OR: odds ratio; CI: confidence interval; Adj.: adjusted.

Discussion

This is the first in-depth study investigating dietary patterns among NAFLD patients in Singapore. It provides an important insight into the fructose intake in NAFLD patients in our local Asian context. This study offers valuable information to tailor effective nutrition intervention programmes for NAFLD patients. Our study showed NAFLD patients had a significantly higher consumption of calories, macronutrients, added sugar and fructose than CHB controls. However, there was no independent relationship between fructose intake and NAFLD, after adjusting for age, gender, waist circumference and energy intake.

The results of our study contrasted with the data from several observational cross-sectional western studies, which found a positive association between fructose intake and NAFLD.^8–10 This could be in part attributed to the relatively low intake of fructose in the local population. According to the National Health and Nutrition Examination Survey performed in US residents from 2007 to 2008, adults aged 35 to 54 years consumed approximately 80.7 g added sugar daily and the mean intake (14.3% of total energy intake) surpassed the recommended limit of less than 10%.¹⁷ In contrast, the added sugar intake in our local NAFLD and CHB patients was only 55 g/day (9.6% of total energy) and 36 g/day (7.6% of total energy) respectively, values which were within the national recommendation of less than 10% in Singapore. In addition, the average intake of added sugar in our study population was 43.6% less than the American counterparts.¹⁷ Another US study showed fructose ingestion from HFCS or sweetened drinks in NAFLD patients was two- to three-fold higher than healthy controls (91 vs. 43 g/day, p<0.05).⁸ Again, the fructose intake from sweetened beverages in our NAFLD patients was much lower (only 15 g/day) and the intake from all food sources was approximately 41 g/day. Hence, such a low consumption of fructose may not be significant enough to even pose a risk for NAFLD.

Nonetheless, cross-sectional and retrospective case–control studies do not provide definite information on causal relationships. In fructose-feeding animal studies, the effect of fructose in inducing NAFLD was well-established even in the 1970s.^5–7,18 The study from Bruckdorfer et al. reported that diets with 68% energy from fructose raised hepatic fatty acid synthetase activity and plasma triglyceride concentration to the greatest extent compared with other forms of carbohydrates of the same amount (starch, maltose, glucose, sucrose) after 30 days.⁶ Livers from rats infused with fructose continuously increased the secretion of very-low-density lipoprotein and reduced oxidation of free fatty acids, but such effects were not observed in glucose infusion.¹⁸ Although promising data existed in animals, it remains prudent to be cautious in extrapolating this data to humans, as the metabolic pathway of fructose in animals and humans may be different. Furthermore, it is increasingly being recognized that animal data have limitations in making reliable predictions in human studies.¹⁹ The high level evidence as represented by the two recent systematic reviews demonstrated no independent association between fructose consumption and NAFLD in humans, which was consistent with our finding.^11,12 The review from Chiu et al. included seven isocaloric trials with a follow-up duration of three to eight weeks, in which fructose (median: 182 g/day) was administered in beverage form exchanging for equal amounts of other carbohydrates.¹¹ These isocaloric trials included studies not only with neutral energy balance for study subjects, but also with positive energy balance (from overfeeding with fructose), as long as the carbohydrate comparators were matched for the excessive calories leading to the same positive energy balance. On the other hand, the hypercaloric trials used diet supplemented with excessive energy (median: +25%) from fructose (median: 193 g/day) compared with the same diet without the excess energy for a period of more than a week to four weeks. Both trials contained similar macronutrient compositions (50–55% carbohydrates, 30–35% fat and 13–15% protein) and subjects were young and tended to be healthy. The results from isocaloric trials showed no effects of fructose on intrahepatic cellular lipids (IHCLs) and ALT. In contrast, fructose raised both IHCL and ALT in hypercaloric trials. Another systematic review showed similar findings.¹² However, both reviews could not exclude whether the excess energy intake had confounded the effect of fructose in the hypercaloric trials; hence the effect may be due to the excessive energy rather than fructose itself. Similar findings relating fructose to other cardiometabolic risk factors such as body weight, lipid profile and glucose control were reported in a few other systematic reviews, where fructose had adverse effects on these risk factors when it was added to diets and providing excessive calories, but not when it was isocalorically exchanged for other carbohydrates.^20–22 In our study, energy intake was found to be significantly associated with NAFLD.

We also found waist circumference a significant NAFLD risk factor, with an adjusted OR of 1.25 (95% CI: 1.11–1.41; p=0.001). Similarly, a systematic review on the prevalence of and risk factors for NAFLD in Asia by Cheng also showed that waist circumference (abdominal obesity) was an independent and significant risk factor for NAFLD – adjusted OR of 1.09–3.6.¹ This finding was expected because abdominal obesity is strongly associated with insulin resistance, which is a well-known risk factor for NAFLD.³

This study had several strengths. First of all, all dietary-related information was collected by a single dietitian. Hence, inter-observer bias was excluded. Furthermore, the dietitian was blinded to the clinical diagnosis of the patient, thus avoiding performance bias. In addition, the adapted FFQ was expanded to include a wider range of sweetened foods and beverages than the original version, and a series of standard food models and measuring spoons and cups were used to assist patients in quantifying food portions more accurately. All these measures improved the accuracy of food portion size estimation, which is vital in ascertaining the nutrition intake of the study population. Last but not least, this study excluded patients who had altered their dietary habits since the diagnoses of their medical conditions. This criterion is of great importance. If a patient is advised by his/her primary physician to adopt a healthier diet after the diagnosis of NAFLD or other medical conditions such as diabetes (which requires limiting the intake of sugar), his/her sugar intake may have been reduced even before the dietary interview. As a result, the actual fructose intake might have been under-estimated. Hence, any study without this exclusion criterion will underestimate the fructose intake in the subjects.

There were a few limitations in this study. First, a case–control study has its own inherent limitations. The exposure measurement (i.e. dietary intake assessed by FFQ) is subject to recall and reporting bias, as it relies heavily on the self-reporting of food intake in the past. Second, obese subjects are likely to under-report their calorie intake therefore the energy intake may be an under-estimated value.²³ Third, the subjects in the control group were not drawn from healthy individuals. While we excluded patients with cirrhosis from our study, we did not have any additional information about stage of fibrosis in our cohort, which may also be a limitation. Despite our control patients being not healthy subjects, there are no established dietary recommendations or dietary restrictions for CHB management. Hence, the dietary habits in CHB patients are assumed to be representative of healthy subjects. This assumption is further supported by the comparable energy intake of 1796 kcal/day in our CHB patients (with majority of female) to the national recommended energy intake of 1720 kcal/day for female Singaporeans aged between 30 and 59 years.²⁴ The intake of protein, carbohydrate and fat in our CHB patients contributed 19%, 52% and 31% to total energy intake, respectively, which more or less aligned with general healthy eating guidelines. Last, gender was not matched for this study. Although we have done some adjustments to address this limitation in the regression analysis, it would inevitably reduce the statistical power of this study.

Conclusion

In conclusion, our study found that patients with NAFLD tended to consume more energy, macronutrients, added sugar and fructose than the controls. However, fructose intake was not associated with NAFLD, after adjustment for age, gender, waist circumference and energy intake. We observed that excessive calorie intake and abdominal obesity were significant risk factors for NAFLD. Although hypercaloric controlled trials on diet supplemented with fructose supported the effect of fructose on raising liver enzymes, whether this effect is independent of excessive energy intake still remains an outstanding question. The relationship between fructose consumption and NAFLD remains controversial and warrants further clarification. High-quality randomized controlled trials with larger sample size and longer follow-up duration addressing these issues are required to better understand the effect of fructose on histopathological changes of NAFLD in humans. Last, in the nutrition management of NAFLD or NAFLD prevention, our study suggested that obese individuals, particularly those with abdominal obesity, should alert health care professionals for timely and effective nutrition intervention. Diet and lifestyle modifications still remain the first-line approach for NAFLD management.

Footnotes

Author contribution

XYL participated in study design, face-to-face dietary interview for each subject, data entry and statistical analysis and was responsible for writing the manuscript. GGB participated in conception and design of the study, subject recruitment, data collection and review of the manuscript. SHH contributed to subject recruitment, data collection and entry. SF performed the statistical analysis. YHL contributed to conception, design and supervision of the study and review of the manuscript. WCC initiated the study and participated in conception, design, and supervision of the study and review of the manuscript.

Availability of data and materials

The datasets generated and/or analyzed during the current study are available from the corresponding author.

Declaration of conflicting interests

The authors declare that there is no conflict of interest.

Ethical approval

Ethical approval to report these cases was obtained from Centralised Institutional Review Board, Singhealth Office of Research. Approval number ID: 2012/449/E

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Informed consent

Written informed consent was obtained from the patients for their anonymized information to be published in this article.

ORCID iD

Xiao Ying Li

References

Cheng

LF.

Non-alcoholic fatty liver disease in Asia: A systematic review. Master’s thesis, The University of Hong Kong, Pokfulam, Hong Kong, 2010.

Farrell

Larter

CZ.

Nonalcoholic fatty liver disease: From steatosis to cirrhosis. Hepatology 2006; 43: S99-S112.

Souza

Diniz

Medeiros-Filho

, et al. Metabolic syndrome and risk factors for non-alcoholic fatty liver disease. Arq Gastroenterol 2012; 49: 89–96.

Miyazaki

Dobrzyn

Man

, et al. Stearoyl-CoA desaturase 1 gene expression is necessary for fructose-mediated induction of lipogenic gene expression by sterol regulatory element-binding protein-1c-dependent and -independent mechanisms. J Biol Chem 2004; 279: 25164–25171.

Naismith

DJ.

Differences in the metabolism of dietary carbohydrates studied in the rat. Proc Nutr Soc 1971; 30: 259–265.

Bruckdorfer

Khan

Yudkin

Fatty acid synthetase activity in the liver and adipose tissue of rats fed with various carbohydrates. Biochem J 1972; 129: 439–446.

Waddell

Fallon

HJ.

The effect of high-carbohydrate diets on liver triglyceride formation in the rat. J Clin Invest 1973; 52: 2725–2731.

Ouyang

Cirillo

Sautin

, et al. Fructose consumption as a risk factor for non-alcoholic fatty liver disease. J Hepatol 2008; 48: 993–999.

Zelber-Sagi

Nitzan-Kaluski

Goldsmith

, et al. Long term nutritional intake and the risk for non-alcoholic fatty liver disease (NAFLD): A population based study. J Hepatol 2007; 47: 711–717.

10.

Abdelmalek

Suzuki

Guy

, et al. Increased fructose consumption is associated with fibrosis severity in patients with nonalcoholic fatty liver disease. Hepatology 2010; 51: 1961–1971.

11.

Chiu

Sievenpiper

de Souza

, et al. Effect of fructose on markers of non-alcoholic fatty liver disease (NAFLD): A systematic review and meta-analysis of controlled feeding trials. Eur J Clin Nutr 2014; 68: 416–423.

12.

Chung

Patel

, et al. Fructose, high-fructose corn syrup, sucrose, and nonalcoholic fatty liver disease or indexes of liver health: A systematic review and meta-analysis. Am J Clin Nutr 2014; 100: 833–849.

13.

Chitturi

Farrell

Hashimoto

, et al. Non-alcoholic fatty liver disease in the Asia-Pacific region: definitions and overview of proposed guidelines. J Gastroenterol Hepatol 2007; 22: 778–787.

14.

World Health Organization. Waist circumference and waist–hip ratio, http://www.who.int/nutrition/publications/obesity/WHO_report_waistcircumference_and_waisthip_ratio/en (2011, accessed 29 September 2014).

15.

Health Promotion Board. Report of the National Nutrition Survey, http://www.hpb.gov.sg/data/hpb.home/files/whp/health_fac/resources/nnsreport_final(merged).pdf (2004, accessed 21 September 2014).

16.

Food Standards Australia New Zealand. Australian food composition database, http://www.foodstandards.gov.au/science/monitoringnutrients/nutrientables/nuttab/Pages/default.aspx, (2014, Accessed 10 September 2014).

17.

Welsh

Sharma

Grellinger

, et al. Consumption of added sugars is decreasing in the United States. Am J Clin Nutr 2011; 94: 726–734.

18.

Topping

Mayes

PA.

Comparative effects of fructose and glucose on the lipid and carbohydrate metabolism of perfused rat liver. Br J Nutr 1976; 36:113–126.

19.

Perel

Roberts

Sena

, et al. Comparison of treatment effects between animal experiments and clinical trials: Systematic review. BMJ 2007; 334: 197.

20.

Chiavaroli

de Souza

, et al. Effect of fructose on established lipid targets: A systematic review and meta-analysis of controlled feeding trials. J Am Heart Assoc 2015; 4: e001700.

21.

Cozma

Sievenpiper

de Souza

, et al. Effect of fructose on glycemic control in diabetes: A systematic review and meta-analysis of controlled feeding trials. Diabetes Care 2012; 35: 1611–1620.

22.

Sievenpiper

JL1

de Souza

Mirrahimi

, et al. Effect of fructose on body weight in controlled feeding trials: A systematic review and meta-analysis. Ann Intern Med 2012; 156: 291–304.

23.

Johnson

RK.

Dietary intake—how do we measure what people are really eating?

Obes Res 2002; 10: 63S–68S.

24.

Ministry of Health Singapore. Recommended daily allowance, https://www.healthhub.sg/live-healthy/192/recommended_dietary_allowances (2019, accessed 2 March 2019).

Fructose intake in patients with non-alcoholic fatty liver disease

Abstract

Background:

Objectives:

Methods:

Results:

Conclusions:

Keywords

Introduction

Methods

Results

Discussion

Conclusion

Footnotes

Author contribution

Availability of data and materials

Declaration of conflicting interests

Ethical approval

Funding

Informed consent

ORCID iD

References