Diet and nutrition: the folly of the reductionist approach

After almost 4 years of review, in February 2013 the National Health and Medical Research Council released the latest revision of its dietary guidelines for Australians.1 Recognising that people consume foods rather than single nutrients, the new guidelines feature food-based advice, emphasising dietary patterns that are associated with health and wellbeing and are relevant to reducing the risks of obesity and chronic disease.

While researchers use a reductionist approach in analysing the adequacy of selected nutrients or nutrient density, people do not shop for protein, “omega 3s”, “carbs”, calcium or some other nutrient, but for whole foods; and our advice needs to be about which to choose more of (fruit, vegetables, wholegrains, legumes, nuts and fish) and which to limit (sweetened drinks and processed foods high in saturated fat, added sugars or salt).

Emphasising one or more particular nutrients can lead to poor food choices. For example, on a nutrient basis, a processed breakfast cereal may contain added vitamins, but the cereal may be high in added sugars or salt. In this issue of the Journal, research by Brimblecombe and colleagues shows that Indigenous Australians in remote communities consume poor-quality processed foods fortified with some nutrients rather than nutrient-rich minimally processed foods.2

There is also accumulating evidence that the source of nutrients matters. Calcium is an important factor in bone health, but there may be large differences in health outcomes associated with calcium supplied by foods and those associated with taking calcium supplements. New, and admittedly controversial, studies examining consumption of calcium supplements report increased rates of cardiovascular events, especially when added to an adequate calcium intake from foods,3-5 whereas dietary calcium was unrelated to adverse events. Other studies have shown a slightly increased risk of kidney stones with supplemental but not dietary calcium.6 However, it has been suggested that calcium from foods does not pose the same risk because it is consumed in small quantities throughout the day rather than as a single supplementary dose.7

The complexity of fruits and vegetables provides further incentive to avoid a reductionist approach. Evidence cited in the Australian Dietary Guidelines shows that an adequate intake of fruits and vegetables reduces the risk of cardiovascular disease,1 but the reasons remain unclear. In contrast, a meta-analysis of eight randomised controlled trials using beta-carotene supplements showed a small increase in all-cause and cardiovascular mortality.8 Fruits and vegetables contain hundreds of different carotenoids, which may act synergistically with each other as well as with various vitamins, minerals and different types of dietary fibre.1 Consuming foods high in carotenoids with foods such as extra virgin olive oil, as in a typical Mediterranean diet, may increase their health-promoting effects.

Studies of fish consumption also illustrate the need for a whole food approach. A number of studies show that eating fish once or twice a week is associated with a lower incidence of cardiovascular disease1,9 and a reduced risk of stroke.10 Is this due solely to their omega-3 fatty acids and could a supplement therefore replace fish? Maybe not, according to two recent meta-analyses of studies using fish oil supplements that did not show any reduction in risk of major cardiovascular events or all-cause mortality in either healthy people or those with a history of cardiovascular disease.11,12

It is understandable that researchers want to identify protective factors in foods, but the reality is that foods cannot be reduced to single beneficial components. In practice, a reductionist approach can lead to dietary distortions by which foods with undesirable properties (eg, high sugar or salt) can be marketed as beneficial if fortified with a few selected nutrients.

The recent analysis of the Sydney Diet Heart Study also shows the folly of pushing a single nutrient while ignoring other aspects of the foods containing the nutrient. Beginning in the 1960s, participants were given safflower oil and margarine with a high linoleic acid content. Other components of the margarine, such as its high content of trans fat, and of the total diet were ignored. Leaving aside other possible problems with this study, the outcomes were increased rates of death from all causes, from coronary heart disease and from cardiovascular disease.13

The realistic — and wise — course of action is to look at diet in terms of foods and eating patterns rather than taking a reductionist approach and concentrating on a single nutrient that is almost never consumed on its own.