Old nutrition

Brand-Miller began by stating: The old nutrition goes like this ….

• Foods can be dissected into macronutrients
• Saturated fat is the main dietary risk factor for cardiovascular disease
• A low fat diet is best for prevention of obesity, diabetes and cardiovascular disease
• “Eat a diet that is low in fat and high in complex carbohydrate”
• “Eat plenty of cereals, breads, rice, pasta and noodles, preferably wholegrain”

And then she took old nutrition apart, highlighting how the nutrition landscape had changed over the past decade or so:

• Large, long-term, randomised controlled trials had showed low fat dietary advice did not prevent heart disease, stroke, type 2 diabetes, breast cancer or colon cancer.

• The prevalence of obesity and type 2 diabetes had continued to increase, despite the adoption of lower fat diets in Australia and the United States.

• In relation to heart disease risk, some carbohydrates were worse than saturated fat. In particular, high glycaemic index carbohydrates were linked to higher risk of cardiovascular disease, as well as higher risk for diabetes and some cancers.

• Higher fat, Mediterranean-style diets produced better weight control and cardiovascular outcomes than low fat diets.

• When compared head-to-head in high quality randomised controlled trials, energy-dense, high protein, very low carbohydrate diets were associated with faster weight loss and better cardiovascular risk factors than low fat diets.

• Low glycaemic index/low glycemic load diets were associated with improved diabetes control, prevention of weight re-gain and reduced risk of type 2 diabetes.

Ouch! That hurt!

In explaining the shortcomings of the old nutrition and the benefits of alternative approaches Brand-Miller argued that foods’ effects on postprandial glycaemia and insulinemia were central. Here was a unifying mechanism that encompassed effects on blood lipids, oxidative stress, markers of inflammation and, importantly, satiation and appetite control.

Image: source

New nutrition

Warming to her task Brand-Miller continued: So the new nutrition goes something like this …

• Low fat dietary advice has been unhelpful. It is consistently associated with weight re-gain. It does not reduce the risk of chronic disease.
• We should pay more attention to protein.
• We should pay more attention to the quality of carbohydrate sources. We need to swap high GI carbohydrate for low GI. 
• Unlike Americans, Australians don’t eat too much sugar.
• There is a range of healthy macronutrient ratios.
• High fat Mediterranean diets, higher protein diets and low GI diets are all helpful .

Brand-Miller argued that this advice offers flexibility and moderation; it accommodates cultural and ethnic differences; and it is behaviourally more sustainable. She stated that some authorities are already embracing the new nutrition, including the Institute of Medicine and the Joslin Diabetes Clinic in the United States, which have begun to recommend lower carbohydrate and higher protein intakes.

Image: source

Food labelling: new system, old nutrition

Most dietitians and nutritionists would be aware that a new front-of-pack food labelling system is to be introduced in Australia. Similar to the energy star rating system on electrical goods, the easy-to-interpret labels are intended to help the general public quickly assess the nutritional value of foods and identify healthier choices. Of course, whether the scheme actually provides meaningful guidance to the consumers will depend on the criteria used to assess the nutritional quality of foods.

Brand-Miller argued that the algorithm underpinning the new star system is based on old nutrition, highlighted by the focus on energy, fat, salt and sugar. She reeled off a string of anomalies: lentils and liquorice have the same energy density; nuts are more energy dense than French fries; dried fruit is full of sugar; and soft drinks are low in sodium.

In an aside, Brand-Miller said if breast milk were sold in the dairy compartment of the supermarket it would be docked a star because of its sugar content – human milk has the highest sugar content of any mammalian milk. Isn’t it interesting how Mother Nature knows how to ensure that infants drink their milk and thrive but humans are criticised when they use the same strategy with older children, sweetening milk to improve its palatability?

Surprisingly, in spite of embracing old nutrition, the star food labelling system virtually ignores micronutrients. Isn’t the provision of essential nutrients is the fundamental nutritional role of food? Brand-Miller argued that any star system should use Adam Drewnowski’s nutritional quality index as an essential component.

Starch: goodness by default

For Brand-Miller, a major concern about the proposed food labelling system is the healthiness afforded to starch, which is achieved by default i.e. by what it’s not – fat and sugars. The idea that starch is protective relative to fat, or even saturated fat, is prehistoric. Pooled data from 11 large cohort studies shows that carbohydrate and saturated fat pose the same risk for coronary heart disease. In 2010, a European study was published showing that high GI carbohydrates posed greater risk than saturated fat, reinforcing the concern about glycaemic load flowing from studies in the United States. Many starchy foods, even some wholegrains, have high GIs. Seriously, how can a food labelling system encourage less saturated fat and more high GI starchy foods? It’s as logical as recommending more saturated fat and less high GI foods.

Rating starch above sugars is also fraught. In a previous post I highlighted that the relative amounts of sugar and starch in Australian breakfast cereals are unrelated to nutrient density, energy density or glycaemic index. In other words, in these starch-based foods the amount of sugar present is not related to any property or mechanism that might conceivably affect health. Yet the new star food labelling system will separate breakfast cereals on the basis of sugar – healthy eating advice that won’t affect health. Sounds a bit like the old nutrition, doesn’t it?

Policy failure

We would have had more insight into these issues if the National Health & Medical Research Council had commissioned systematic literature reviews into saturated fat and glycaemic index/load for the latest dietary guidelines, but it did not. As far as I know the NHMRC has never systematically reviewed the evidence relating to GI in the 30 years that the concept has been around. With the proposed food labelling system we are now paying the price for this policy failure.

If the responsible health authorities refuse to even consider some of the scientific literature dietary advice will never change. Perhaps that’s the idea.

Image: source

The first Dietary Guidelines in Australia were released in 1980 and recommended restriction of the total amount of fat in the diet as a strategy for chronic disease prevention. In the following 20 years this piece of dietary advice took precedence over all others and dominated the nutrition landscape. However, the writing has been on the wall for the low fat diet for many years now.

The end of the low fat era

The hypothesis that the percentage of dietary energy from fat was related to chronic disease risk was developed in the 1970s and put to the test in large cohort studies that commenced in the following decade and then in several randomised controlled trials. A string of null results followed in relation to breast cancer, bowel cancer, coronary heart disease and stroke. Perhaps most surprisingly, no association between total fat and obesity was observed in prospective cohort studies, and randomised controlled trials suggested minimal benefit.

In contrast, evidence from mechanistic studies, cohort studies and controlled trials confirmed that fat type was relevant to chronic disease risk, especially in relation to coronary heart disease.

For the first time in 33 years the latest Australian Dietary Guidelines no longer highlight the need to limit or ‘moderate’ total fat intake, completing the evolution of advice in favour of fat type that has gone through several stages. The guidelines now recommend limiting saturated fat and state:

• Limit intake of foods high in saturated fat such as many biscuits, cakes, pastries, pies, processed meats, commercial burgers, pizza, fried foods, potato chips, crisps and other savoury snacks.
• Replace high fat foods which contain predominantly saturated fats such as butter, cream, cooking margarine, coconut and palm oil with foods which contain predominantly polyunsaturated and monounsaturated fats such as oils, spreads, nut butters/pastes and avocado.

These recommendations are well founded. The first point focusses attention on limiting many nutrient-poor foods that are combinations of saturated fat and poor quality carbohydrate, which is now known to confer the same or greater coronary risk than saturated fat. The second point is arguably more important as it highlights the most desirable replacement for saturated fat in the diet – poly- and monounsaturated fats. This is the area where the science is clearest.

Overall, the implication is that the Mediterranean-type diet – moderate in both total fat and carbohydrate but low in saturated fat and enriched with unsaturated fats is now the preferred model for healthy diets. This general approach is also consistent with the latest Dietary Guidelines for Americans, which no longer advocate low fat diets and encourage the replacement of saturated fat with unsaturated fats.

Image: source

Change to the food guide

There is only one significant change to the Australian Guide to Healthy Eating – the separation of unsaturated vegetable oils and spreads from the ‘extras’ or junk foods. This corrects a much criticised flaw in the previous guide and has been welcomed by the Heart Foundation. When the previous Australian Guide to Healthy Eating was released in the late 1990s the Heart Foundation contested the positioning of unsaturated vegetable oils and spreads with ‘extras’ and decided that it would not use the national food guide in any of its nutrition education materials, and it never did. The Heart Foundation’s principled stand on this issue has now been vindicated and, after a 15-year stand-off, Australia’s leading nutrition authorities now largely agree on the advice in relation to these foods.

But the new food guide isn’t perfect. The use of the words ‘choose small amounts’ in relation to unsaturated vegetable oils and spreads is a mistake. It implies restriction and reflects a lingering nervousness about dietary fat and its potential to promote obesity, despite the authorities’ own literature review showing no link.

The simple fact is that children’s intakes of unsaturated fats need to increase by 30 percent if they are to approach the ‘total diets’ modelled during the development of the new food guide and this increase needs to come from somewhere. It can’t come from meats and dairy foods as these foods are also rich in saturated fat. That just leaves nuts and seeds, which are only consumed by a small proportion of the population, and unsaturated vegetable oils and spreads. This inconsistency between the science in the Dietary Guidelines and its communication really should have been sorted out. ‘Choose unsaturated oils and spreads’ would have been an appropriate and moderate recommendation.

About 36-41 percent of calories currently come from nutrient-poor foods high in saturated fat and/or poor quality carbohydrate. These foods are the obvious targets for energy restriction for weight management, not the major source of unsaturated fats.

Energy-dense, nutrient-poor foods

The term ‘energy-dense, nutrient-poor foods’ has found its way into the Dietary Guidelines, especially in relation to obesity. Interestingly, none of the evidence statements in the obesity section actually mentions ‘energy-dense, nutrient-poor foods’. It’s just a euphemism for junk foods which are assumed to be a problem. .

I must admit I find ‘energy-dense, nutrient-poor foods’ an imprecise, clunky term. After all, macadamias are energy-dense and nutrient-poor but this doesn’t mean that they have no place in a healthy diet. On the contrary, the inclusion of these nuts in a diet has a potent effect on improving fat quality. White rice is quite energy-dense and certainly nutrient-poor but for some reason it’s positioned as a food for daily consumption rather than with the other junk foods. If terms don’t have a consistent meaning alternatives need to be considered.

The ‘energy-dense, nutrient-poor’ terminology can be expected to evolve further in the years ahead. As calculations of nutrient density are usually based on a fixed amount of calories, energy density and nutrient density are not distinct concepts – they overlap. Contrary to popular belief, energy density is not a useful measure of the healthiness of foods or diets in relation to the prevention of obesity. Nutrient density is a cleaner measure and this concept has already been embraced in the Dietary Guidelines for Americans.

But nutrient density is not sufficient by itself. Future models for classifying junk foods and healthier foods will need to combine nutrient density with measures of fat quality and carbohydrate quality.

Carbohydrate quality: an evolving concept

While fat quality is now well understood, the concept of carbohydrate quality is undeveloped and remains contentious. In the latest Dietary Guidelines carbohydrate quality remains narrowly conceived around added sugar and wholegrains, with no consideration (not even a literature review) of glycaemic index. However, a debate on this topic is now underway internationally and its conclusions will guide the further evolution of dietary advice in the years ahead.

I’ll cover the latest contributions to this debate in my next post.

Image: source

The first post on The Sceptical Nutritionist highlighted several problems with the draft Australian Dietary Guidelines and these have been discussed in detail in recent months. Encouragingly, some of the issues appear to have been addressed during the consultation period. The clues come in a draft appendix to the Australian Dietary Guidelines that was recently released for consultation, which includes the latest version of the key Dietary Guidelines statements.

More emphasis on cereal fibre

There is now greater emphasis on cereal fibre. The previous statement that recommended cereal foods should be ‘mostly wholegrains’ has been changed to ‘mostly wholegrain and/or high cereal fibre varieties’. This is consistent with the science as many of the studies used to support the wholegrains recommendation used high fibre foods, suggesting that recommending high fibre cereal foods was equally valid.

Trans fats

Another positive development was the removal of trans fats from the key Dietary Guidelines statements. As no systematic literature review on trans fats was conducted and there was hardly any discussion about trans fats in the text of the Guidelines, the rationale for emphasising trans fats was a mystery. After all, the level of trans fats in the Australian diet is low, just half the maximum level recommended by the World Health Organization. Also, as most of the trans fats in the Australian diet now come from meat and dairy fat, advice to lower saturated fat from these sources also serves to further lower intake of trans fats.

Image: source

Added sugar

The original guideline statement advocating that intake of foods and drinks containing ‘added sugars’ be limited was criticised for not being scientifically based. Compared with starch, sugar has no adverse health effects, so why limit it? And why make a distinction between naturally occurring and added sugars? Also, blanket advice to limit foods with added sugar had the potential to adversely affect diet quality as some foods with added sugar, such as flavoured milk, fruity yoghurts and breakfast cereals, are nutrient-rich. What logic is there to limiting nutrient-rich foods on the basis of their sugar content? Do nutritionists recommend limiting intake of oranges because they contain sugar?

The latest version of the guideline statement on sugar at least takes a step in the right direction, stating Limit intake of foods and drinks containing added sugars such as confectionery, sugar-sweetened soft drinks and cordials, fruit drinks, vitamin waters, energy and sports drinks. This focusses attention on limiting nutrient-poor sources of sugar. But if ‘empty calories’ is the concern, why not go all the way and make nutrient density a criterion on which the nutritional quality of foods is judged? Why bother with sugar at all?

The saturated fat debacle

Saturated fat was very poorly handled in the early stages of the development of the Dietary Guidelines.

• Amazingly, no systematic literature review was conducted and consequently the process missed a significant development in the science of how macronutrients affect the risk for coronary heart disease i.e. that carbohydrate confers the same risk as saturated fat. Recommended ‘healthy’ diets were low in saturated fat and high in carbohydrate, the authorities apparently unaware of the contradiction. Not surprisingly, the relevant chapter in the text of the draft Guidelines was out of date and poorly argued.

• There was blind adherence to the notion that a healthy diet needs to be low in fat – all of the modelled Foundation Diets were low in fat. A process that was supposed to be ‘based on the best available scientific evidence’ somehow managed to miss the four reports of the Women’s Health Initiative, the largest randomised controlled dietary trial ever conducted, which showed that a low fat diet has no effect on the risk for cardiovascular disease, breast cancer and bowel cancer, and minimal effect on body weight.

• A major concern was that unsaturated fat, which had emerged as the most protective class of macronutrients in relation to coronary heart disease, was restricted. The science is clear – unsaturated fat is the preferred replacement for saturated fat. Intake of unsaturated fats needs to increase, not decrease.

• Instead of encouraging the use of vegetable oils and spreads – the major source of unsaturated fats in the Australian diet, there was a bias against their inclusion. This bias was evident in dietary modelling. Unsaturated oils and spreads were categorised differently to the other 12 groups of foods modelled, the inclusion of oils and spreads being simply to ‘reflect current culinary behaviour’. However, the dietary modelling actually demonstrated a significant and distinctive nutritional contribution, the modelled margarine being a major source of vitamin A, vitamin D, vitamin E, linoleic acid and α-linolenic acid, as well as being instrumental in achieving the desired ratios of fatty acids.

• This bias against unsaturated vegetable oils and spreads, a reflection of a belief system rather than nutrition science, was carried through into the key Dietary Guidelines statements and the draft Australian Guide to Healthy Eating. The first guideline statement ‘Eat a variety of nutritious foods …’ included reference to all foods in the food groups modelled, except for unsaturated oils and spreads. So it was not an accurate translation of the dietary modelling into key Dietary Guidelines statements. Dogma at work. Similarly, the advice about the use of unsaturated oils and spreads in the Australian Guide to Healthy Eating was to ‘use small amounts’ i.e. to restrict the use of these foods, contrary to the science.

What a mess.

Image: source

Some progress

The good news is that there has been some progress on this issue. Apparently, the chapter on saturated fat has been re-written and is now said to be consistent with the science. And the wording of the key Dietary Guidelines statement relating to saturated fat has been changed. The original wording was ‘Limit intake of foods and drinks containing saturated … fats’ and included advice to ‘include small amounts of foods that contain unsaturated fats’.

The latest draft Guidelines statement provides much greater details on which sources of saturated fat should be limited:

Limit intake of foods containing saturated fat such as many biscuits, cakes, pastries, pies, processed meats, commercial burgers, pizza, fried foods, potato chips and crisps and other savoury snacks.

and, importantly, that foods rich in unsaturated fats should replace saturated fat in the diet:

Replace high fat foods which contain predominantly saturated fats such as butter, cream, cooking margarine, coconut and palm oil with foods which contain predominantly polyunsaturated and monounsaturated fats such as oils, spreads, nut butters/pastes and avocado.

The removal of the term ‘small amounts’ in relation to unsaturated oils and spreads is a positive move. For consistency, these words will also need to be deleted from the Australian Guide to Healthy Eating. We shall see.

As a result the Guidelines fall short scientifically in seven areas. These are reviewed below and will be considered in more depth in the weeks ahead.

Low fat diets

The Dietary Guidelines continue to advocate diets low in fat, despite there being no scientific justification for this approach. None of the systematic literature reviews conducted during the development of the Guidelines established a link between total fat intake and any condition of public health significance, including obesity. Nevertheless, the low fat approach was still embraced on the grounds that it was ‘prudent’. This is dietetic dogma at work – my term for the conservative mindset that just can’t let go of old dietary advice, even when it is no longer scientifically valid.

Controversially, the Guidelines recommend restriction of all types of fat in the diet, even unsaturated fats which are considered to be protective against coronary heart disease. In an article in The Australian newspaper in March 2012 a leading Australian nutrition scientist, the Heart Foundation and the Dietitians Association of Australia all criticised this aspect of the Guidelines. This problem arose because key literature on saturated fat, carbohydrate and heart disease was simply not reviewed.

Saturated fat, carbohydrate and heart disease

Limiting saturated fat intake has been a fixture in many Dietary Guidelines for years on the grounds that saturated fat raises the level of LDL-cholesterol in the blood, thereby increasing the risk for coronary heart disease. In Australia the preferred replacement for saturated fat has been carbohydrate, encouraging diets low in both saturated fat and total fat. However, a key study published in 2009 confirmed that carbohydrate and saturated fat confer the same risk for heart disease, so swapping one for the other has no effect on risk. As a consequence, unsaturated fats have become the recommended replacement for saturated fat and this approach has been incorporated into the Dietary Guidelines for Americans (2010).

It is hard to believe that there was no systematic literature review into the role of saturated fat in heart disease risk during the development of the latest Australian Dietary Guidelines. Instead, the literature review for the 2003 Dietary Guidelines was relied upon. So in this key area, the Dietary Guidelines are out of date, inconsistent with the dietary guidelines of other countries and inconsistent with the positions of many leading research and nutrition authorities.

The higher-than-expected risk of dietary carbohydrate has highlighted the need to eat less carbohydrate (not less fat) and to choose ‘good’ carbohydrate foods. But what do the Dietary Guidelines tell us about carbohydrate quality?

Added sugar

The Dietary Guidelines recommend that intake of foods with ‘added’ sugar be limited. But is the amount of added sugar in a food really a good measure of its nutritional quality? Certainly, sugary foods and drinks can increase the risk of tooth decay, especially in those with poor oral hygiene, but this is also a problem with natural sugars. The same goes for the starch found in rice, potato or bread which breaks down to sugars in the mouth under the action of digestive enzymes.

Another problem with targeting added sugar is the widely varying nutritional value of foods with added sugars. At one end of the scale soft drinks and some confectionery may contain virtually no nutrients other than sugar, but sugar-sweetened foods such as flavoured milk, fruity yoghurts and some breakfast cereals are nutrient-rich. Is it really wise to limit intake of these nutritious foods?

Wholegrains

(Image source)

Wholegrains are enthusiastically embraced in the draft Dietary Guidelines and this is really the only measure of carbohydrate quality considered, other than added sugar. Wholegrains are recommended as being highly nutrient-dense and also protective against heart disease and type 2 diabetes. On the surface this appears to be scientifically sound and sensible advice, but the issues are not straightforward. Firstly, although nutrient density makes sense as a measure of the quality of carbohydrate-rich foods, many wholegrains are not highly nutrient-dense, as claimed. Secondly, many of the studies used to support the wholegrains recommendation were actually conducted using high fibre foods, suggesting that recommending high fibre cereal foods would be equally valid. Why recommend one but not the other?

A third concern is the lack of a plausible mechanism explaining the apparent protection of wholegrains against heart disease observed in some epidemiological studies. A study that went looking for the mechanism found nothing. Is the idea of the healthiness of wholegrains so appealing that nutritionists now accept the evidence they like and ignore the evidence they don’t?

Finally, many wholegrain foods have a high glycaemic index, which is gaining in importance as a measure of carbohydrate quality.

Glycaemic index

The glycaemic index or GI is a measure of the blood glucose-raising potential of a carbohydrate-rich food. Early interest in the GI focussed on its potential role in the diets of people with diabetes but more recent evidence suggests that GI has a broader role. GI is related to the risk for coronary heart disease – high GI foods appear to pose greater risk than saturated fat. And GI also has a role to play in weight management. A large European trial recently showed that the GI and protein content of the diet were important factors for maintaining weight loss.

One would have expected to see systematic literature reviews on these issues as part of the development of the new Dietary Guidelines, but not a single review relating to GI was conducted.

Trans fats

Trans fats have been introduced into the Australian Dietary Guidelines for the first time but one has to ask why? Again, no systematic literature review of trans fats and health was conducted so there was no evidence base. The text of Guidelines report contained just two sentences on trans fats which noted that intake of trans fats in Australia is low – half the maximum level recommended by the World Health Organization. Furthermore, three quarters of the trans fats in the Australian diet now come from ruminant fats (meat and dairy fat). As these fats are also rich in saturated fats, any recommendation to limit saturated fats will also serve to further lower trans fat intake. The problem of trans fats has been well and truly addressed over the last 15 years, so why bother alerting the general public to the issue now?

Fruit, vegetables and cancer

(Image source)

The view that fruits and vegetables provide protection against cancer is widely held among nutritionists but is largely based on the results of old case-control studies, known for their bias. When this hypothesis was tested in large cohort studies the results were not encouraging. It is now clear that if fruits and vegetables have any effect on the risk for cancer it is very weak. In response, the World Cancer Research Fund downgraded the evidence of a link in its 2007 report. In the Dietary Guidelines for Americans, the whole section on fruit, vegetables and cancer was removed from the 2010 edition.

In the Dietary Guidelines process there may not have been any systematic literature reviews into saturated fat, trans fats or glycaemic index but there was a major evidence hunt for a link between fruit, vegetables, legumes and cancer. No less that 21 literature reviews were conducted. And the result? No ‘convincing’ or even ‘probable’ evidence of a link. But rather than follow the American lead and simply delete the issue, the Australian Dietary Guidelines doubled the amount of text devoted to this topic and painted a misleading picture suggesting that evidence for a protective effect of fruit and vegetables against cancer is building.

Is this evidence-based nutrition or something else?