Background
The association between fruit and vegetable intake and the risk of chronic diseases including cardiovascular and cerebrovascular diseases (CVD) has been confirmed by previous epidemiological studies [
1‐
3]. In Japan, CVD are among the top contributors to the disease burden [
4], accounting for over 20% of the total medical expenditure [
5]. Over the past 50 years, mortalities due to CVD in the aging population have declined, especially for those 80 years old or older. The mortality rate due to cardiovascular diseases among individuals in their 80’s decreased from 2.24% in 1965 to 0.91% in 2017 while the mortality rate among those in their 90’s decreased from 4.79% in 1965 to 2.96% in 2017 [
6,
7]. However, the CVD burden is expected to increase among the rapidly aging population of Japan [
8].
The consumption of vegetables and fruits is lower among adults in Japan than in other countries [
9]. In Japan, a minimum of 350 g/d of vegetable (5 servings at 70 g per serving, not including potato or legumes) and 200 g/d of fruit (2 servings at 100 g per serving) per day are recommended [
9]. The UK’s recommendation is 5 portions/d (400 g/d) of fruits and vegetables (not including potato or legumes) [
10]; the US recommends 2~3 cups/d of vegetables (168 g/d ~ 252 g/d, depending on age and sex) and 1~2 cups/d of fruit (depending on age, sex, and level of physical activity) for individuals who get less than 30 min per day of moderate physical activity [
11]. Recent results of the Japanese National Health and Nutrition Survey 2016 indicate that in adults over 20 years old, the daily consumption remains as low as 269.4 g for vegetables and 98.9 g for fruits (adjusted by age) [
12] and that there has been a significant decline from 2006 (vegetables: 300.5 g/d; fruits: 107.5 g/d) [
13].
The aim of this study was to estimate and project the CVD burden at different scenarios of vegetable and fruit intake and to simulate the potential impact on CVD burden in Japan by improving vegetable and fruit consumption in a long term.
Discussion
Based on convincing evidence of an inverse association between fruit and vegetable intake and the CVD rate [
17,
22,
27], our results demonstrated a probable reduction in the CVD burden if even a modest improvement in fruit and vegetable intake could be achieved, such as an additional 1/2 serving of fruit or additional 1 serving of vegetable (the moderate increase scenario). The proportion of the CVD disease burden attributable to fruit intake was projected to decrease from 12.6% (the base scenario) to 8.5% (the moderate increase scenario) and to 4.9% (the high increase scenario), and that to vegetable intake from 8.5 to 5.4% and to 2.3%.
Our results are in accordance with previous analyses showing a potentially great impact in reducing cardiovascular disease burden by increasing fruit and vegetable intake [
35‐
37]. While detailed disease burden attributed to inadequate fruit and vegetable intake was divergent due to 1) other studies considered the total effects of fruit and vegetable intake rather than separating their impacts [
35‐
37]. We consider it inappropriate to directly sum up the disease burden, as the risk ratios used in this study were separated and possible overlapping impacts due to double counting if sum up directly; 2) different average daily consumption in heterogenous populations: for example even lower fruit and vegetable intake level in South African population (235 g/d for males and 226 g/d for females) compared to that in Japan (adjusted by age: 372 g/d for males and 400.3 g/d for females) [
12,
13], which can partly explain the higher PAF in a simulation study with a South African population (ischemic heart disease PAF: 34.6%; ischemic stroke PAF: 22.2%) [
37]; and 3) different theoretical minimum risk of ‘adequate’ intake. For example, some studies chose an intake of 600 g/d in adults as the optimal consumption [
35‐
37] while we used 350 g/d of vegetable and 200 g/d of fruit as recommended by the Japanese government [
9].
In addition, daily consumption of both vegetable and fruit remains relevantly low in Japan [
25], and the consumption pattern diversifies by age and gender. The national average vegetable and fruit intake in Japan ranked 77th and 133th among 172 countries in 2013, respectively, and was equivalent to about 50% of that in other high income countries [
38]. Similar to the tendency in other developed countries, younger people tend to be less likely to consume fruit compared to the older people [
35]. On the other hand, different to other researches [
35‐
37], average consumption in female is larger than that in male, especially that of fruit. Such a tendency may be explained by the selective preference to foods: females were more likely to be concerned about freshness and food safety while males were more likely to select foods based on their taste preferences [
39].
The projected number of patients with a CVD as well as the DALY due to low fruit and vegetable intake were estimated to increase in future decades (Fig.
3); especially, there is a steady growth shown in the disease burden of female stroke (Fig.
3b). The potential reasons might be a steadily projected increase of female stroke cases from 526,000 in 2014 to 874,797 in 2060; and the majority of the CVD disease burden (> 90%) in Japan is within the growing older population (> 60 years). [
12] According to the Patients Survey in 2015: about 94.9 and 86.8% mortality cases happened in female and male group aged above 65 years, respectively. In addition, population projection in Japan showed that the age group above 65 years has a soaring increase from 2015 to 2020, 2030 and 2060 (female: 29.7, 32.1, 34.9, 43.5%; male: 23.8, 25.9, 28.0, 36.1%, respectively) [
14].
As a forefront super-aged society, CVD burden is expected to continuously increase in the upcoming decades. The Japanese government initiated the Healthy Japan 21 Project in 2002 (2nd phase from 2013), recommending increased consumption of vegetables and fruit to prevent lifestyle-related diseases [
40]. However, vegetable consumption has actually decreased in the last decade from 300.5 g/d in 2006 to 269.4 g/d in 2016 [
12] [
13]. ,possibly due to a combination of factors: an increasing preference for a western style of diet, characterised by energy dense, processed foods; increased cost, particularly impacting the younger generation with lower purchasing power [
41]; and changes in lifestyle resulting in an increased desire/need for foods requiring less preparation time and having a longer shelf-life (from suppliers and individuals), and reduced supply [
42]. Although overall fruit consumption has remained stable from 107.5 g/d/person in 2006 to 98.9 g/d/person in 2016 [
43] the proportion of fresh fruit consumed has decreased possibly due to increasing prices, ready availability of processed foods or reduced supply and/or accessibility of fresh produce [
43]. The number of green grocers fell by half between 2003 and 2014 in tandem with a decrease in areas under cultivation and the yield of fruit trees (statistics from 1975 to 2016) [
43]. Decreased fresh fruit and vegetable production in Japan is associated with the aging population and increased urbanization, with the number of farming households decreasing from 510,000 in 2005 to 370,000 in 2015 [
41] [
43]. ,Global climate change is also likely to be impacting fruit and vegetable production in Japan. Increased CO
2 concentrations may have a predominantly positive effect on yield but negative effects on nutritional quality while the increased incidence of severe weather events often cause catastrophic damage to crops [
44].
The “Dietary Guidelines for the Japanese” was first announced in 2000. In 2005, an upside-down pyramid, the so-called, “Japanese Food Guide Spinning Top” along with the Basic ‘Shokuiku’ (dietary education’) Act was published to promote nutritional education at the community level [
45]. This program aimed to teach the Japanese what a well-balanced diet per day should contain: 5–7 servings of grain (rice, bread, noodles, and pasta), 5–6 servings of vegetables, 3-5servings of fish or meat (meat, fish, egg, and soy-beans), 2 servings of dairy (milk and milk products), and 2 servings of fruit [
46]. The lack of knowledge among the general population of how the recommended daily servings translate to actual dietary intake was evidenced in one survey of 300,000 adults in 2012, in which over a half of the respondents stated that they did not know the daily recommended intake in grams or how the daily recommended intakes translated into servings [
41,
47]. Promoting adequate fruit and vegetable consumption is a significant public health challenge.
On the other hand, although abundant evidence mainly from observational studies showed fruit and vegetable consumption has a protective effect again CVD [
1‐
3], few intervention trials were long enough to examine the effects of increased fruit and vegetable consumption on CVDs without confounders driven from other dietary patterns and lifestyle modifications [
48,
49]. Factors contributing to CVDs are complicated, not limited to vegetable and fruit consumption.
Study limitations
Despite using the best available data resources to calculate the risk ratios associated with fruit and vegetable intake based on a recently published meta-analysis, some limitations remain.
First, the risk ratio of fruit/vegetable intake to the risk of CVD was not specific to the Japanese population and may therefore not accurately reflect the different genetics, dietary habits (e.g. more fish, less meat), and physical exercise levels of the Japanese population.
Second, the RR used in this study did not take each subtype of fruit and vegetable into account. Obviously, different combinations may lead to different effects. Various studies have reported on fruit and vegetable subtypes and the risk of CVDs [
2], and not every subtype showed an inverse association with CVD risk. For example, an high intake of items like apples/pears, citrus fruits, fruit juice, green leafy vegetables, and pickled vegetables might decrease the total risk of stroke [
2]; however, an high intake of apples/pears, citrus fruits, fruit juice, green leafy vegetables, tomatoes, beta-carotene rich items, and vitamin C rich items had an inverse association with the risk of coronary heart disease [
2]. Items such as berries, citrus fruit juices, dried fruits, grapes, canned fruits, strawberries, broccoli, etc. did not show this association with CVD [
2].
Third, the average energy adjusted fruit/vegetable intake (format of g/1000Kcal) would be optimal but was not used due to lack of suitable data and for the sake of simplicity. Further, data of dietary exposure is from dietary questionnaire assessments [
13], reflecting limitation of dietary ascertainment [
50].
In addition, a greater intake of fruit and vegetables does not necessarily result in more benefits. Some fruits containing high levels of carbohydrates and sugars can increase the blood glucose level, posing a danger to patients with type-2 diabetes [
51]. An high intake of pickled vegetables containing large concentrations of N-nitroso compounds might double the risk of esophageal cancer [
52,
53]. In this study, the data source on fruit intake included figures on fresh fruit and juice but did not include fruit preserves. Vegetables were categorized as yellow-green vegetables, other vegetables, vegetable juice, and Japanese pickles [
54].
The CHD, stroke prevalence, mortality rate, and the life table of the general population used as important estimation indices in this study showed a lower disease burden than the findings of the World Health Organization – Global Burden of Disease Project (WHO-GBDP). The reason might be that, in this study, the disability weight of CHD (ischemic heart disease) only considered acute myocardial infarctions and angina pectoris based on available data; the “standard expected years of life lost” for calculating YLL in this paper used population projections for Japan based on the Japanese National Institute of Population and Social Security Research for age 0 to 105 years [
24]. The “standard expected years of life lost” in the WHO-GBDP estimates were age ranging from neonatal to 85+ at 5-year intervals, leading to an higher life expectancy and disease burden (due to greater longevity) for those aged over 85 years. When dealing with the disease data from the official website, we omitted patients with an unknown age to avoid underestimation; fortunately, however, these patients contributed few data points (1/782 in CHD and 1/1046 in stroke).
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.