Fruit and Veges: How Much is Good for You?

"Eat to Live, Not Live to Eat"
John H Weisburger, 2000

The Australian National Health and Medical Research Council currently recommends a daily intake of 2+5 servings of fruit and vegetables (F&V, serving sizes defined as 150g and 75g respectively) as part of a healthy diet that, together with attention to other modifiable lifestyle factors such as smoking, alcohol consumption and physical activity, can significantly reduce the risks of non-communicable diseases particularly prevalent in later life. However the latest CSIRO dietary survey (2017) of 150,000 Australians showed that only 19% of 18-30, 17% of 31-50, 26% of 51-70 and 33% of 71+ yr old adults were meeting both F&V guidelines. So the majority are short of this public health goal with substantial differences between generations, and the latest research (International Journal of Epidemiology (2017) 1029-1056) suggests the collective shortfall is worse than thought, with further preventative effects seen in some outcomes up to levels of 800g/d F&V compared to the 300 + 375g/d recommended.

Questions remain about the strength and shape of the dose-response relationship between F&V intake and risk of cardiovascular disease, cancer and mortality, and the effects of specific types of F&V. We conducted a systematic review and meta-analysis to clarify these associations. PubMed and Embase were searched up to 29 September 2016. Prospective studies of F&V intake and cardiovascular disease, total cancer and all-cause mortality were included. Summary relative risks (RRs) were calculated using a random effects model, and the mortality burden globally was estimated; 95 studies (142 publications) were included. For F&V combined, the significant summary RR per 200 g/d was 0.92 for coronary heart disease, 0.84 for stroke, 0.92 for cardiovascular disease, 0.97 for total cancer and 0.90 for all-cause mortality. Similar associations were observed for fruits and vegetables separately. Reductions in risk were observed up to 800 g/d for all outcomes except cancer (600 g/d). Inverse associations were observed between the intake of apples and pears, citrus fruits, green leafy vegetables, cruciferous vegetables, and salads and cardiovascular disease and all-cause mortality, and between the intake of green-yellow vegetables and cruciferous vegetables and total cancer risk. An estimated 5.6 and 7.8 million premature deaths worldwide in 2013 may be attributable to F&V intake below 500 and 800 g/d, respectively, if the observed associations are causal. F&V intakes were associated with reduced risk of cardiovascular disease, cancer and all-cause mortality. These results support public health recommendations to increase F&V intake for the prevention of cardiovascular disease, cancer, and premature mortality.

This large study concerned F&V dietary risks for cardiovascular disease, cancer and all-cause mortality, the former two being the most common causes of premature death worldwide. The 95 studies included in the meta-analysis were all conducted in Europe, North America, Asia and Australia. With up to 2.1 million participants in some of the analyses, more in depth dissection and extension of previously reported trends was possible. The study had increased statistical power to establish significant health benefits and the dose response relationship at higher consumption levels where earlier studies did not have sufficient data. Although the benefits in their study showed some levelling out with higher consumption levels, the authors still recommended that future studies should be conducted to establish whether levels higher than 800g/day could have further positive effects. The RR profiles for F&V and coronary heart disease and cardiovascular disease in particular were still decreasing at intakes of 800g/d, but there was insufficient data for analysis beyond this level.

The RRs quoted above may seem small, eg 8% less risk for F&V and coronary heart disease, but it should be appreciated that these are the cumulative risks for each extra 200g/day consumed, so that the RR reduction for example with coronary heart disease with an intake of 800g/d was 24%, and for all-cause mortality it was 31%. This level of 800g F&V/d is twice that currently recommended in England (ie 400g/d) and by the World Cancer Council and the World Health Organisation. In support of this higher intake level, one of the Global Burden of Disease Studies in 2016 focussing on 84 risk factors estimated that the minimum risk exposure level for fruit was 200-300g/d excluding juices and pickled or salted fruits, and 290-430g/d for vegetables excluding legumes, salted or pickled vegetables, juices, nuts and seeds and starchy vegetables (eg potatoes and corn), giving an F&V total of 490-730g/d for those foods that qualify.

Randomised clinical trials (RCTs) provide the most reliable information in such studies, but because cardiovascular disease and cancers usually take many decades to develop, this type of study with sufficient duration is very difficult if not impossible to conduct. RCTs allow causation to be evaluated whereas other designs provide only associations between risk factors and outcomes. The next best approach to RCTs uses a prospective design, and these were the only published studies considered in the literature screening process of the present work. This design avoids or minimises many of the problems with other observational studies, namely, selection bias with the need to appropriately match controls with cases in case control studies, confounding where apparent associations between risk factors and outcomes may be influenced by other factors not considered or allowed for in the analysis, reverse causation where instead of a risk factor causing some disease, it was the diagnosis of the disease that led to a change in the risk factor, recall bias where participants are asked to recall their behaviour at some time in the past, dietary measurement error with questionnaires, participants changing their behaviour during the study period etc. However in the absence of RCT data, satisfaction of a number of well recognised criteria may provide good evidence for associations being causal, and the authors in this study discussed the likelihood of this conclusion.

The analytical methods used in many epidemiological studies assume that relations between risk factors and outcomes are directly proportional (or linear) without proper examination. This assumption was examined in the present study, where in most cases it was found that the greatest risk reduction occurred at lower intakes but still continued with lesser effects at the higher levels not well-documented in earlier studies. The findings regarding individual fruits or vegetables having significant positive effects are important but probably mainly reflect those with sufficient data eg consumption of apples and citrus is high enough to provide adequate information for testing benefits, rather than those not mentioned being necessarily ineffective. In the case of Australians, the CSIRO study illustrates that generally we don't consume enough F&V in total, so this leaves little opportunity for testing effects of the less common F&V. The important point rather than trying to identify key individual foods, is to focus on whole diets with as broad a range of healthy foods as possible, ie variety is the way to go instead of seeking so-called 'super foods'. Individual foods or supplements when studied by RCTs for desired effects on disease outcomes have largely been negative, and some even detrimental. The finding that F&V can have a greater effect on cardiovascular disease than cancers has been generally reported in earlier literature.

For we in the Rare Fruit Club, the study conclusions are 'manna from heaven' as we like to grow and eat our delicious produce, and probably already consume much more than the average adult as a result of our interest. But from a health perspective we can probably allow our frugivorous predilections to increase a few more notches without feeling embarrassingly over-indulgent or being unduly worried about an unbalanced diet. Note however that fresh fruit is generally better than processed fruit which may have unwanted negative or reduced beneficial effects, even when sugars, preservatives, thickeners, acidifiers, colouring agents etc are not added in processing, eg see the following articles on juice and tinned fruit in the nutrition series on our web site:

Images/Stars Spacer. LitchiLogo