No association between fruits or vegetables and non-alcoholic fatty liver disease in middle-aged men and women

Highlights

• Fruit intake was inversely associated with prevalence of non-alcoholic fatty liver disease (NAFLD).

• Adjustment for BMI attenuated the association between fruit intake and NAFLD prevalence.

• Total vegetable intake was not associated with NAFLD prevalence.

• None of the subtypes of vegetables were clearly associated with NAFLD prevalence.

Abstract

Objective

It has been hypothesized that fruit and vegetable intake is inversely associated with non-alcoholic fatty liver (NAFLD). However, some studies have speculated that fruit intake might be positively associated with NAFLD owing to the fructose content of the fruit. This might cause consumers to hesitate consuming fruit. The aim of this study was to assess the association between fruit and vegetable consumption and NAFLD.

Methods

This was a cross-sectional study of 977 men and 1467 women, 40 to 69 y of age without current liver disease other than NAFLD and who did not report excess alcohol intake (i.e., ≥30 g/d in men and ≥20 g/d in women). Dietary intake was assessed using a validated diet history questionnaire. NAFLD was diagnosed from abdominal ultrasonography results. The association between quartiles of fruit or vegetable consumption and NAFLD prevalence was assessed using logistic regression analysis, with lowest category as reference.

Results

The prevalence of NAFLD was 34.9% in men and 11.7% in women. Adjusted for age and lifestyle factors, fruit intake was inversely associated with NAFLD in both sexes. However, these associations disappeared after further adjustment for body mass index. Consumption of total vegetables was not associated with NAFLD. In women, a linear inverse association was demonstrated between green and yellow vegetable intake and NAFLD in the final model (P_trend = 0.04), but odds ratios for any intake category did not reach significance.

Conclusions

No obesity-independent association was found between fruit or vegetable intake and NAFLD. According to the findings of this study, Japanese do not need to restrict fruit consumption to limit fructose intake as a means of preventing NAFLD.

Introduction

Non-alcoholic fatty liver disease (NAFLD) is characterized by fat accumulation in the liver not caused by excess alcohol intake. NAFLD is a common liver disease both in Western [1], [2] and Asian countries [3] and can proceed to more severe liver diseases such as liver cirrhosis or hepatocellular carcinoma [4]. Thus, it is important to clarify the lifestyle factors that relate to NAFLD. Although the pathogenesis of NAFLD has not been fully revealed, it is strongly associated with increased peripheral fatty acid flux in the insulin-resistant state and also with de novo lipogenesis [5].

It is well known that fruit and vegetable intake is inversely associated with several insulin resistance-related diseases. This has naturally led to the hypothesis that fruit and vegetable intake also is associated with a low prevalence of NAFLD. Fruits and vegetables supply potentially protective dietary factors, such as antioxidant vitamins, for NAFLD [6]. In fact, the Mediterranean diet, which is rich in antioxidants from fruits and vegetables, is considered a potential dietary regimen for NAFLD prevention, although studies regarding its efficacy are limited cross-sectional studies and intervention studies in high-risk populations [7]. On the contrary, one study showed that fruit intake was positively associated with NAFLD prevalence in Chinese women [8]. Another study in Korea showed that vegetable intake, but not fruit intake, was inversely associated with NAFLD prevalence, and the authors speculated that the lack of an inverse relation of fruits to NAFLD might be due to fructose [9]. These findings are partly supported by studies using dietary pattern analysis: NAFLD prevalence was positively associated with a “fruits” pattern [10] and inversely with a “grain and vegetables” pattern [11].

Despite the failure of epidemiologic studies to demonstrate an association of fructose with NAFLD pathogenesis independently of excess energy intake [12], fructose still has received attention as a potential risk factor for NAFLD, mainly based on biochemical knowledge [13], [14]. Some consumers might not accurately understand recommendations for regulating added sugar intake or distinguish food sources of added versus naturally occurring sugar [15]. Indeed, even patients with diabetes educated by health professionals incorrectly believe that they need to avoid foods containing any sugar [16]. Thus, misunderstanding the relationship between health and sugar is considered common and might interfere with the global health campaign recommending fruit and vegetable consumption [17]. Additional studies examining the relationship between the intake of fruits and vegetables with NAFLD may help develop dietary strategies to prevent NAFLD. To our knowledge, such studies are few and none have been conducted in a Japanese population.

In the present study, we assessed the relationship between the consumption of fruits and vegetables and the prevalence of NAFLD in middle-aged men and women. Considering that each subtype of vegetable might be differently associated with insulin resistance [18], we also assessed the relationship by vegetable subtype.