WHO’s record on sugar

The World Health Organization (WHO) is a leading global health agency with a proud history of sound dietary advice, including advice about sugar. In a 1990 report, WHO recommended a limit on intake of ‘free sugars’ of no more than 10% of daily calories, which is about the current average intake of Australian adults. Free sugars means all sugars added to foods by the manufacturer, cook or consumer, plus sugars in honey, fruit juices and syrups.

WHO’s rationale for limiting sugar intake was to lower the risk for tooth decay. No lower limit on intake of free sugars was recommended.

Thirteen years later WHO again looked at the science of sugar and health and found ‘convincing’ evidence that both the amount of free sugars and the frequency of sugar consumption increased the risk for tooth decay. And again WHO recommended a limit of 10% of daily calories.

The 2015 WHO report

In its latest report WHO found … wait for it … that eating too much sugar causes tooth decay and that the intake of free sugars should be limited to … wait for it … less than 10% of daily calories.

Interestingly, WHO also made a ‘conditional recommendation’ that the intake of free sugars could to be lowered to below 5% of daily calories for better prevention of tooth decay. But then WHO stated that this recommendation was based on ‘very low quality evidence’.

In this day and age it is a mystery why any health organisation would make a recommendation based on ‘very low quality evidence’. If the evidence is so poor, why didn’t WHO just stick with the old advice that there was no recommended lower limit on intake of free sugars?

WHO also reviewed the evidence in relation to whether sugar intake is related to body weight. Its recommendations are cautious e.g. WHO states that the evidence ‘suggests’ an association between reduction of free sugars intake and lower body weight in adults. There was no association in children. The quality of the evidence varied between moderate to low.

Perhaps the reason for the soft recommendation was WHO’s desire to maintain advice about free sugars, whereas other organisations have tended to focus more on sugar-sweetened beverages where there is more persuasive evidence. WHO found that substituting sugar for other carbohydrates (starch) had no effect on body weight, so there is nothing inherently fattening about sugar – it all depends on how many calories you eat.

Image: source

The myths: what WHO doesn’t say

The latest WHO report is notable for what it doesn’t say about sugar. It doesn’t say sugar is addictive, toxic, uniquely fattening, or that it gives you fatty liver, heart disease or diabetes. That’s because these are all just myths peddled by attention-seeking, non-nutritionists to boost their celebrity, sell books and make money.

Too much sugar is bad for your teeth. And sugar contains calories, which cause weight gain when consumed in excess of the body’s needs.

It’s not rocket science; it’s nutrition science.

Fat in your liver

Most of the carbohydrate we eat ends up in the bloodstream as either glucose or fructose. The myth goes that glucose is the good sugar as it is used to power the brain, the muscles and most of the cells in the body. And the fructose is the bad sugar which is quickly taken up by the liver and turned into fat, giving rise to fatty liver.

Unfortunately for the myth-makers, no reputable health authority in the world agrees. Fatty liver is certainly a common problem but the experts see it as part of the metabolic syndrome – a cluster of abnormalities linked to central obesity and insulin resistance, where the cells of the body become less sensitive to insulin.

There is no recommended diet for fatty liver. Instead, health authorities encourage people with fatty liver to lose some weight and increase their physical activity, both of which improve insulin resistance.

Fat in your blood

Our liver certainly has the ability to turn both glucose and fructose into fat – it’s the perfect way to turn any excess carbohydrate calories into a form that can be stored for later use. And sooner or later this fat appears in the blood as ‘triglycerides’.

However, the idea that all the fructose we eat turns to fat pushing up the level of triglycerides in the blood is just plain wrong. If you are a healthy, normal weight person eating enough food to maintain your body weight your liver only turns a tiny fraction of fructose into fat, about 1-3%. Most of the fructose taken up by the liver is actually turned into glucose – supposedly the good sugar, so it’s much more accurate to say ‘fructose turns to glucose’ than it is to say ‘fructose turns to fat’.

It’s a different story if you overeat thereby forcing the body to turn excess sugars into fat. But the underlying problem here is not fructose; it’s overeating.

The best ways to lower the level of triglycerides in the blood are to lose some excess weight, increase physical activity and limit alcohol intake.

Image: source

Fat on your body

The silliest myth about fructose is that it is uniquely fattening. You would have thought that we would all have learned something from the fat-makes-you-fat fiasco, where the blame for the obesity epidemic was laid at the feet of just one nutrient – fat. This focussed everyone’s attention on the third of our calories that came from fat and allowed us to ignore the rest. And the nation got fatter.

Does it make any sense to target fructose, which typically provides just 10% of calories? Do the other 90% of calories not count?

None of the nutrients is uniquely fattening. People put on weight when they regularly eat more calories than their bodies need.

Blood lipids and heart disease risk

The approach to managing blood lipids to lower heart disease risk has evolved over time as our understanding of this complicated field has grown. Decades ago the focus was on lowering total cholesterol but this was later narrowed down to LDL-cholesterol – ‘bad’ cholesterol for the lay person. ‘Good’ HDL-cholesterol was considered protective and the balance between the bad and the good – the LDL/HDL ratio – came into use. More recently, the total cholesterol/HDL ratio has been considered to be a better measure as all non-HDL-cholesterol appears to increase heart disease risk.

What about triglycerides?

If you are interested in how triglycerides affect the risk for heart disease and how diet affects blood triglycerides look no further than the American Heart Association’s scientific statement Triglycerides and Cardiovascular Disease. You may need to allocate an afternoon as the paper is very comprehensive and has over 500 references.

The AHA report states that high levels of triglycerides are associated with higher risk for heart disease but then delves into the long-standing debate about whether high triglycerides actually promotes heart disease or is simply a marker of risk.

The issue here is that high triglycerides are often associated with low levels of protective HDL-cholesterol. In fact, the two are metabolically linked – if triglycerides go up, HDL-cholesterol goes down. As low levels of HDL-cholesterol strongly predict higher heart disease risk, HDL seems to be the key player and triglycerides its faithful companion.

Image: source

Factors affecting triglycerides and HDL

Several aspects of diet and lifestyle affect triglycerides and HDL-cholesterol in the blood. Weight loss of 5-10% results in a 20% fall in triglycerides and increases HDL-cholesterol. Similar effects can be achieved with increased physical activity. Long chain omega 3s or fish oils also lower triglycerides, though quite high intakes are required to achieve the effect. And high intakes of alcohol can increase triglycerides. The amounts of fat and carbohydrate in the diet are relevant too.

High carbohydrate diets: adverse effects?

Very high carbohydrate diets certainly raise triglycerides and lower HDL-cholesterol. This is one of the reasons that the US Institute of Medicine and the NHMRC recommend upper limits of carbohydrate intake of 60-65% of calories. If these upper limits appear high it’s because increasing the carbohydrate content of the diet at the expense of saturated fat also lowers total and LDL-cholesterol – beneficial effects.

So when considering the effects of high and low carbohydrate diets it’s important to consider overall effects on blood lipids. Just focussing on triglycerides will only tell part of the story.

Carbohydrate quality

Carbohydrate-rich foods vary enormously – from jelly beans to red kidney beans – and it would not be surprising to find that different carbohydrate-rich foods affect blood lipids differently. The DASH diet was rich in carbohydrate from fruits, vegetables and wholegrains and lowered blood pressure, but it did not raise triglycerides. Carbohydrate quality was high and consequently high carbohydrate intake did not adversely affect blood lipids.

Fructose has a reputation for being triglyceride-raising though a meta-analysis showed that replacing starch or sucrose with fructose had no effect on fasting triglycerides at intakes of up to 100g a day. The triglyceride-raising effect was observed at higher intakes of fructose. Another meta-analysis suggested that people with diabetes, who are prone to high triglycerides and low HDL, may be more sensitive to fructose. But intakes of over 60g a day were still needed to demonstrate an effect on triglycerides, which equates to about 12% of calories in a 2000 calorie diet. So even in this at-risk group the fructose message is about moderation rather than avoidance.

New meta-analysis: sugars and blood lipids

A new meta-analysis of studies into the effect of sugars on blood lipids was published by Te Morenga and colleagues in May of this year. As expected they reported that higher sugar intakes are associated with higher blood triglycerides but perplexingly they found that sugars raised the level of protective HDL-cholesterol. This doesn’t make much sense given the metabolic link between the two. One of the problems with this study is that it is not clear what sugars are being compared to. Is it starch, or total fat, or specific types of fat?

Sugars were found to increase LDL-cholesterol but again we don’t know what the reference is. Based on previous work it could be expected that sugars would raise LDL-cholesterol compared to polyunsaturated fat but lower it relative to saturated fat. The reference is critical to the finding. Unfortunately, effects on overall measures of lipid-related risk, such as the total cholesterol/HDL ratio, were not reported. And some of the studies included did not have strict control of energy intake, which is interesting given the effects of positive and negative energy balance on blood lipids. Quite a few of the studies were conducted in the 1960s or 70s and heterogeneity was high.

Make up your own mind on this one. It’s clear that we need some new, well designed studies to guide policy on sugars.

Image: source

Low carbohydrate diets: good or bad for blood lipids?

The effect of carbohydrate on triglycerides is a weak rationale for recommending a high or low carbohydrate diet. Overall effects on blood lipids need to be considered and in the case of low carbohydrate diets this will largely be determined by the type of fat that replaces carbohydrate. If it is unsaturated fat, as recently trialled by CSIRO in people with diabetes, a low carbohydrate diet confers benefit. However, there is no benefit if saturated fat is used to replace carbohydrate.

One of the lingering mysteries about the current enthusiasm for low carbohydrate diets is the frequent recommendation by unqualified, self-declared authorities to increase intake of saturated fat, rather than unsaturated fats.

Hmmm. Why would anyone make such a recommendation?

Glycaemic response or fructose?

One view is that the key driver of carbohydrate-related risk is glycaemic response i.e. the degree to which carbohydrates raise blood glucose and insulin levels. If glycaemic response is the critical factor then the dietary measure of interest would be glycaemic load, which takes into consideration both the amount of carbohydrate consumed and its potential to raise blood glucose.

An alternative view is that fructose drives the chronic disease risk associated with dietary carbohydrate. If so, we should expect fructose or perhaps sucrose (the major dietary source of fructose) to be associated with risk for chronic disease in large population studies. And, importantly, glycaemic load would not be linked with risk.

Let’s look at the epidemiology and see if it supports one argument or the other.

Glycaemic load and coronary heart disease risk

Four meta-analyses of prospective cohort studies of glycaemic load and risk for coronary heart disease have been published in the last couple of years (Dong 2012, Fan 2012, Ma 2012, Mirrahimi 2012). The findings are essentially the same – glycaemic load is associated with heart disease in women but not men. In two of the meta-analyses the links between glycaemic load and heart disease were stronger in overweight subjects. Fan (2012) also found an association with stroke incidence, as did Sieri (2013) in a recent cohort study. In all four meta-analyses the relative risks for heart disease in men were positive – they just failed to reach statistical significance, perhaps due to smaller numbers.

A study in a Chinese cohort published last year found glycaemic load was associated with coronary heart disease risk in both genders (Yu 2013). Compared to studies in western populations carbohydrate intake was high (68% of calories) but sugar intake was relatively low. Most of the carbohydrate was starch, 87% of which came from white rice and refined wheat products.

Image: source

Fructose or sucrose and coronary heart disease risk

There are virtually no prospective cohort studies that find a significant association between fructose intake and risk for coronary disease. The few studies that have looked at total sugars and heart disease risk generally find no association (Beulens 2007, Sieri 2013). A recent exception found an association between ‘added sugars’ and cardiovascular disease in a study in the United States (Yang 2014).

There are few data about the consumption of sugar-sweetened beverages and risk for coronary heart disease. And the findings are mixed – Fung (2009) found a positive association but this was not confirmed by Eshak (2012). Eshak (2012) and Larsson (2014) found links with stroke, albeit inconsistent between genders.

Glycaemic load and risk for type 2 diabetes

In the last three years at least four meta-analyses of cohort studies have considered whether glycaemic load affects the risk for type 2 diabetes, the latest published this year by the Harvard team (Bhupathiraju 2014). All four meta-analyses found that high glycaemic load increases risk for type 2 diabetes.

A recent meta-analysis found that consumption of white rice, a high GI food, was associated with increased risk for type 2 diabetes in Asian populations where it is by far the major contributor to glycaemic load (Hu 2012). The association was also positive in western populations but failed to reach significance.

Fructose or sucrose and risk for type 2 diabetes

No meta-analysis of studies into whether fructose or sucrose affects risk for type 2 diabetes has been published and there are just a few cohort studies to go on. The findings are mixed. In a small cohort Montonen (2007) found intake of glucose and fructose combined, but not sucrose, was associated with increased risk of type 2 diabetes whereas in the large EPIC cohort sucrose, glucose and fructose were not associated with diabetes risk in men or women.

A couple of months ago a meta-analysis of studies into sugar-sweetened beverages and risk for type 2 diabetes and found a positive association (Greenwood 2014). Interestingly, the same study also found a positive association between non-sugar beverage consumption and diabetes. This raises questions about whether these links are causal or alternatively whether soft drink consumption is a marker for other lifestyle behaviours that are associated with risk.

Image: source

Glycaemic load or fructose?

Given the public debate about fructose and sucrose the paucity of data linking these sugars with risk for type 2 diabetes and coronary heart disease risk comes as a surprise. It is clear that glycaemic load is a more consistent predictor of risk for these conditions than fructose or any individual sugar or total sugars or added sugars. This doesn’t let sugars off the hook but rather suggests that the risk for chronic disease posed by sugars should be seen in the context of their contribution to glycaemic load.

The significant contribution of sugar-sweetened beverages to glycaemic load in western diets has an interesting parallel in Asian diets – white rice, a high GI, starch-rich food. Both are also nutrient-poor. Some dietary recommendations in Australia still encourage the replacement of sugars with starch but the rationale for this is lacking if limiting glycaemic load is the most effective means of lowering carbohydrate-related risk.

Poor quality carbohydrate is not a narrow concept. It’s broad and includes some traditional foods currently recommended for healthy eating.

In 2011, Professor Jennie Brand-Miller from the University of Sydney and Dr Alan Barclay, Chief Scientific Officer at the Glycaemic Index Foundation and Head of Research at the Australian Diabetes Council published a paper on the Australian Paradox – the apparent fall in sugar consumption that occurred over a period when rates of obesity in this country increased. The paper was obviously intended to stir the pot a little.

The narrative in the United States at the time was that the increase in obesity prevalence in that country had coincided with increasing sugar intake, so perhaps sugar was a causative factor. In their paper Barclay and Brand-Miller pointed out that while that may well be the case in the United States, sugar intakes had remained fairly constant in the United Kingdom over the same period and had actually fallen by 16 per cent in Australia. However, both countries had experienced an increase in rates of obesity, hence the Australian Paradox.

Under normal circumstances this simple paper may have dissolved away into the vast ocean of scientific literature and never been heard of again. However, a member of the public took exception to the finding that sugar intakes in Australia were falling. Despite not having any qualifications in nutrition or science he had formed a view that sugar intakes in Australia were in fact rising. A social media campaign was initiated to attack the nutrition researchers and their findings and a formal complaint was lodged with the University of Sydney.

The accusation

It was alleged that Barclay and Brand-Miller deliberately included falsified data in their analysis, were reckless, caused harm or risked public health and gained personally from their conduct. For nutrition research scientists the accusations could not have been more serious. After all, the researchers were required to adhere to the University of Sydney Research Code of Conduct and the Australian Code for the Responsible Conduct of Research. If they were found to have breached these codes their respective careers in research were finished.

The investigator

The University of Sydney took the complaint seriously and decided that an internal enquiry may not have been sufficient to resolve the issue. An external, independent person of recognised high academic standing and with substantial experience in overseeing matters of ethics and integrity was sought to conduct the review. Professor Robert Clark AO, Chair of Energy Strategy and Policy at the University of New South Wales, Former Chief Defence Scientist of Australia and CEO of the Defence Science and Technology Organisation accepted the role. His 86-page report has just been released after a six-month inquiry.

The findings

The University of Sydney summed up the key findings of the report in its press release:

• A formal inquiry into allegations brought against a University of Sydney academic and her research collaborator has found no research misconduct occurred.
• There was no breach of the University’s research code of conduct or of the Australian Code for the Responsible Conduct of Research.
• Consequently, there is no basis for further investigation and the allegations have been dismissed.

Although the nutrition researchers must be relieved by the findings of the review we can only imagine what it was like to suffer the sustained attack on personal and professional integrity they had to endure. Professor Clark’s comments in the body of the report provide some insight:

It is my view that, at interview, Professor Brand-Miller and Dr Barclay presented as open, honest, and well-intentioned academics … The stress resulting from the impact of the Complainant’s allegations on their scientific integrity was apparent. They each expressed the view that over the past few years they had effectively undergone ‘trial by internet’ due to the manner in which the Complainant had aired the allegations in the public domain.

An inconvenient truth

Was all this really necessary?

It would appear that Barclay and Brand-Miller were targeted because they were generating data that undermined the sugar scare being perpetuated by various media personalities, celebrities and booksellers, who presumably were on a nice little earner. To sustain the sugar scare there had to be a crisis –a tsunami of sugar crashing across Australia wreaking havoc on the population’s health. Evidence that sugar intake was actually falling simply represented an inconvenient truth that had to be buried, along with the nutrition scientists who dared to publish it.

Science is all about a contest of ideas – debate between differing views is inherent in the process. But the place for such arguments is in scientific meetings or in peer-reviewed journals where dissent can be recorded for posterity. But fundamental to such debates is soundly based scientific evidence, civility and respect for one’s opponent.

Sadly these were missing in the attack on Brand-Miller and Barclay.

Last month Professor Jim Mann** addressed the annual conference of the European Association for the Study of Diabetes (EASD) in Barcelona on the controversial topic of carbohydrate quality. Although he was mainly talking about carbohydrates in the diets of people with diabetes, what’s good for this group is good for most of us.

Sugar

Given the current hysteria about sugar it was interesting that Professor Mann had little to say about it. He indicated that the current EASD recommendation for the general population i.e. that total free sugars be limited to 10% of energy, was appropriate and was likely to be retained when new EASD guidelines are released. He had a lot more to say about starch.

Image: source

Pity those seeking advice

Professor Mann highlighted some of the contradictory advice about starchy foods currently being offered to people with diabetes. In the United Kingdom, NHS Dietitians and Diabetes UK recommend eating plenty of starchy carbohydrate foods while the American Diabetes Association recommends the exact opposite – eating less.

Professor Mann suggested that the differing advice was the result of misinterpretations of current recommendations. Although these recommendations allow for a wide range of carbohydrate intakes (45-60% of energy) he stressed that when carbohydrate intake is at the upper end of this range it is particularly important to emphasise foods rich in dietary fibre and low in glycaemic index (GI). He cited a recent meta-analysis highlighting the benefits of fibre but also noted that many of the studies showing benefits of higher carbohydrate intakes employed legumes.

But typically most people in western countries don’t eat a lot of legumes and some frequently recommended ‘healthy’ alternatives don’t shape up as well.

Wholegrain bread: hero or villain?

Surprisingly, Professor Mann expressed concern about wholemeal bread and dietitians’ near-universal endorsement of it:

“The problem is that many dietitians around the world are telling people to have wholegrain bread when most wholegrain bread is roughly comparable to eating a bag of glucose.”

This was a reference to the high glycaemic response most people experience when they eat many wholemeal breads. Professor Mann argued that rather than actually containing whole grains, which he supported, many so-called wholegrain foods had had the structure of the grains disrupted.

He quoted “one of the most important papers that has ever been published in terms of the nature of carbohydrate” by Jarvi and colleagues. In this study the effects of two diets composed of the same foods and with identical macronutrient composition and fibre were compared, the only difference being that in one diet the structure of the starch had been disrupted, increasing the GI. Glycaemic control and blood lipids were significantly better on the diet with intact starch. Professor Mann encouraged the use of ‘genuine’ wholegrains rather than many of the products that currently pass for wholegrains.

Although he considered GI to be important, Professor Mann admitted to being a bit ‘nervous’ about it and suggested that it needed to be used ‘intelligently’. He argued that other nutritional attributes of foods had to be considered in combination with GI.

This is a common criticism of GI but it can be addressed by considering a food’s nutrient density and GI together, as Manny Noakes and I did in our carbohydrate quality model.

White rice: pure, white and deadly?

Professor Mann also took aim at white rice, citing a recent meta-analysis by Hu and colleagues showing that higher white rice consumption is associated with a significantly increased risk of type 2 diabetes, especially in Asian populations. He said it had been known for three decades that high carbohydrate, high refined starch diets were associated with a deterioration of glycaemic control.

So what was Diabetes UK thinking when they advised people with diabetes to eat plenty of starchy foods?

Image: source

Summing up

Professor Mann came to the following conclusions:

• A wide range of carbohydrate intakes is still acceptable, between 40-60% energy
• Legumes, pulses, fruits and ‘genuine’ wholegrains were optimal choices among carbohydrate-rich foods
• Dietary fibre was a good indicator of a healthy choice
• Rapidly digested starchy vegetables (such as potato), white rice and many breads should be limited, even some breads labelled wholegrain.

Comment

The thing I found interesting about Professor Mann’s talk was that he effectively put sugar and refined starch in the same category – both should be limited. This is an inescapable conclusion, yet the old starch-good, sugar-bad paradigm persists among many dietitians.

If sugar and refined starch are in the same category, it follows that the sugar content of a starchy food is a very poor way of assessing its nutritional quality. Other criteria (that have some relevance to health) are required to discriminate between such foods but none of them is perfect. It’s about shades of grey, rather than black or white. Carbohydrate quality is a complex, multi-faceted concept and efforts to simplify it inevitably give rise to misleading dietary advice.

A second thing that struck me is how little of the carbohydrate in the diets of many people is of high quality i.e. legumes, pulses, fruits and ‘genuine’ wholegrains. There are plenty more carbohydrate-rich foods that have at least some positive qualities, such as high fibre, low GI or nutrient-rich.

But there are lots more that have little going for them, including white rice which still features in many healthy eating guides. A traditional food isn’t necessarily a healthy food. Less of these foods would be better (Kodama and colleagues). Not surprisingly, in Professor Mann’s recommendations the lower boundary of the recommended carbohydrate range has edged down to 40% of energy.

Have a listen to Professor Mann’s presentation.

** Jim Mann is a Professor in Human Nutrition and Medicine from New Zealand and has a major interest in the role of nutrition in diabetes and cardiovascular disease. He is one of the leading nutritionists in the southern hemisphere and has worked on several European advisory groups including the EASD and the World Health Organisation.

Those who have been following the carbohydrate quality debate will enjoy a recent feisty review by Professor Tom Wolever from the University of Toronto published in the European Journal of Clinical Nutrition. Wolever is the co-inventor of the glycaemic index concept and is obviously a little peeved that other measures of carbohydrate quality such as wholegrain have gained acceptance in dietary guidelines whereas GI has not.

Wholegrains versus GI

Here is how Wolever weighs up the relative merits of wholegrains and GI as measures of carbohydrate quality (minimally edited).

The evidence for the health benefits of wholegrains comes from epidemiological studies showing strong associations between high intake of wholegrains and reduced risk for cardiovascular disease, diabetes and obesity. Recent large clinical trials in subjects without diabetes showed that increased wholegrain consumption reduced systolic blood pressure, but none showed any significant effect of wholegrains on blood glucose, insulin sensitivity and/or insulin secretion, inflammatory markers or body weight (Tighe 2010, Brownlee 2010, Kristensen 2012). One study found that wholegrains reduced LDL-cholesterol, but another found the opposite.

On the other hand, trials of low-GI diets of similar magnitude in subjects without diabetes showed significant beneficial effects of a low-GI diet on blood lipids,(Jebb 2010, McMillan-Price 2006) inflammatory markers (Gogebakan 2011) body weight (Larsen 2010) and possibly insulin sensitivity. When the results of the four low-GI studies are combined, there is a significant reduction in LDL-cholesterol and C-reactive protein and a strong trend towards weight reduction.

Low-GI diets have at least as many, if not more, statistically significant effects than wholegrain-enriched diets; this does not support the hypothesis that wholegrain is a better marker of carbohydrate quality than GI.

I confess I have struggled to understand how wholegrains could lower risk for diabetes and coronary heart disease for the reasons that Wolever summarises. A small effect on systolic blood pressure really isn’t good enough. There is always the possibility that wholegrains may be operating via a yet-to-be-discovered mechanism, but this appears unlikely. Wholegrains have no effect on the inflammatory marker CRP suggesting no effect on underlying disease processes.

So how can one explain nutritionists’ recent embrace of wholegrains as a measure of carbohydrate quality? Perhaps the decision makers like the message – simple and easy to understand. But then again, if the message has shortcomings should we be broadcasting it? ‘Eat less fat’ was a simplification of ‘eat less saturated fat’ and look where that took us.

Image: source

Alternatively, perhaps there is a philosophical dimension to it. Nutritionists like the idea of wholegrains because it fits nicely with our philosophy about nutritional health being a natural consequence of eating simple, whole foods. But we are not philosophers; nutritionists are meant to be scientists and to make recommendations based on the facts.

The facts about GI?

The experts are yet to agree on what the facts relating to GI actually are. If you want to understand the GI debate and the case for and against, Wolever’s review is for you. He lists the criticisms of GI made by its detractors, such as imprecise measurement, lack of reproducibility, inaccurate tables and the notion that GI does not apply to mixed meals. He then addresses the issues one by one and challenges his critics to prove him wrong.

And the politics

As you may have guessed, there is a little politics involved here. In North America several influential nutritionists took a stand against GI early on and have defended their position ever since. They now concede that glycaemic ‘response’ is relevant to health but have reservations about the relevance of the GIs of individual foods. Wolever won’t have an easy task convincing them. Sometimes the acceptance of new approaches in science proceeds slowly, funeral by funeral.

In Europe leading nutritionists have been more accepting of the GI concept, especially following the results of the Diogenes study, which showed beneficial effects on weight management. The European Union is now funding a large new trial, with an Australian arm, into whether a high protein, low GI diet is superior to a conventional healthy diet for the prevention of type 2 diabetes.

It is noteworthy that Wolever doesn’t see GI as an alternative to wholegrains but as a complement to it. They are both measures of carbohydrate quality. I would add the caveat that some wholegrain foods and some low GI foods are nutrient-poor, again highlighting that carbohydrate quality is a complex, multi-faceted concept.

Image: source

Discriminating between carbohydrate-rich foods

To illustrate the argument I used the model for discriminating between carbohydrate-rich foods developed by Professor Manny Noakes and myself. For more information on this model please refer to my previous post or the published paper.

In brief, the model is based on nutrient density and glycaemic index (GI). Nutrient density was chosen as it reflects the fundamental nutritional role of foods – the delivery of essential nutrients. And GI was chosen because it relates to the physiological effect of the carbohydrate itself. When carbohydrate-rich foods are run through the model they end up in one of four carbohydrate quality quadrants.

Dairy products – sweetened and unsweetened

Here is what happens when milks and yoghurts are run through the model.

All these products fall into the highest quality quadrant i.e. they are all nutrient dense foods and they all have low GIs. In Dr Alan Barclay’s presentation at ILSI he stated that nutrition dilution (not disease prevention) was the major nutritional negative associated with high sugar consumption and the model demonstrates this effect. Note the positions of reduced-fat milk and flavoured milk, which is essentially reduced-fat milk plus sugar. The flavoured milk is displaced to the left indicating that it has lower nutrient density – fewer nutrients for the same calories. This is the nutrient dilution effect of sugar. The same effect is apparent with natural low-fat yoghurt and sweetened low-fat yoghurt. So sugar is bad, right?

The other thing to note is that the sweetened dairy foods still fall in the highest quality quadrant – they are all nutrient-rich and all have low GIs, irrespective of their sugar content. Why then would nutritionists divide these foods into good and bad on the basis of their sugar content? Aren’t they all good? Including chocolate milk in a child’s diet does not have a nutrient diluting effect. It’s a nutrient-rich food – it has the opposite effect. A recent study showed that Australian children who drink plain or flavoured milk have higher micronutrient intakes but similar body mass index to those who do not drink milk. How can one argue that chocolate milk is an undesirable food?

Grain foods – sweetened and unsweetened

Here are the results when some refined cereal foods and breakfast cereals are run through the model.

Firstly, note that white rice, polenta, couscous, semolina and white pasta tend to cluster in the lowest quality quadrant – nutrient density is very low and GI varies from moderate to very high. Until recently these cereals were referred to as ‘core’ foods and their consumption was encouraged. I wonder why – look at white rice sitting deep in the lowest quality quadrant. Just because a food is a traditional food doesn’t necessarily mean it’s a healthy food. Although these grain foods contain no sugar their nutrient densities are so low that their inclusion in the diet has a nutrient dilution effect.

Compared to the unsweetened ‘core’ grain foods the nutrient densities of the breakfast cereals are consistently high, driven by their fortification. Some of these breakfast cereals have added sugar but in this context there is no nutrient dilution effect of sugar at all. In fact, including any of the breakfast cereals in a diet at the expense of any of the unsweetened ‘core’ cereals would enrich a diet with nutrients. Relying on the old starch-good/sugar-bad paradigm can produce perverse outcomes.

Sugar, energy density and GI

I recently surveyed Australian breakfast cereals to investigate whether their sugar content was associated with energy density or GI. Please take a look at the findings, which are now published. They show that sugar content of breakfast cereals has absolutely no association with either energy density or GI.

Think about this: In the context of breakfast cereals, sugar content has no association with nutrient density, energy density or GI. In other words, the sugar content is not related to any nutritional or physiological parameter that might actually affect health. This is why sugar content is a very poor measure of the nutritional quality of a carbohydrate-rich food.

What’s needed?

Currently, health authorities are considering front-of-pack labelling systems to guide consumer choice. Using sugar, or added sugar, as a criterion for such a system would be a mistake. It would inevitably mislead. In developing front-of-pack food labelling and in nutrition education more generally two things are needed:

1. Meaningful measures of nutritional quality. If nutrient dilution is the concern with sugar, then why not make nutrient density a criterion?

2. Criteria that can be applied universally. If a criterion can’t be applied universally, then it is obviously flawed. If sugar is to be a criterion, it has to be applied to fruits. There is no place for subjectivity.

Nutrient density and GI are meaningful measures that can be applied to all carbohydrate-rich foods, which is why we used them in our model.

Image: source

Recommended carbohydrate intakes

A couple of speakers discussed the recommended range for carbohydrate intakes, which is 45-65% of daily calories in both Australia and the United States. In the United States, the major factor that determined the upper boundary of intake was adverse effects of carbohydrate on levels of triglycerides and HDL-cholesterol in the blood. At the lower end of the scale it was argued that fibre requirements are unlikely to be met at intakes of carbohydrate below 45% of energy (in the low fibre US context). It’s interesting that dietary fibre is considered as a carbohydrate-amount issue in the United States. To my mind it is very much a carbohydrate quality issue.

Mean adult intake of carbohydrate in Australia is about 46% of daily calories, right at the lower end of recommended intake range. This was perceived as low by some speakers and a reason for focussing on glycaemic index as the preferred means of lowering the total glycaemic load of the Australian diet. Others saw the current intake as ‘moderate’, pointing out that the beneficial diets in the Diogenes study contained just 43% of calories from carbohydrate – below the lower boundary of recommended carbohydrate intake.

I introduced my talk by arguing that both the upper and lower boundaries of carbohydrate intake were too high. They were framed at a time when it was thought that a higher percentage of dietary energy from fat was detrimental to health. However, the preferred model for healthy eating is now one with more unsaturated fats and less of both saturated fat and carbohydrate – a Mediterranean-type diet.

No speaker advocated very low carbohydrate diets for the general population. Whether the adverse effect on blood vessel function of very low carbohydrate diets is due to low carbohydrate intake per se or due to the typically high saturated fat content of such diets is unclear and is being researched.

Image: source

The benefits of carbohydrate intake

Professor Manny Noakes and Dr Alan Barclay gave two excellent talks addressing the benefits and risks of carbohydrate intakes. Provision of micronutrients and dietary fibre were the obvious benefits of carbohydrate-rich foods. Dr Barclay summarised the findings of a recent review showing that a high dietary fibre intake is associated with lower risk for obesity, type 2 diabetes, cardiovascular disease and colorectal cancer at varying levels of evidence. As much dietary fibre comes from wholegrains it’s difficult to clearly attribute the benefit to one or the other. Professor Noakes mentioned ‘intriguing’ data suggesting that wholegrains may affect body composition but also that CSIRO studies were unable to find any effect of wholegrains on other risk markers or risk factors for coronary heart disease.

Dietary carbohydrate also lowers LDL-cholesterol in the blood when it replaces saturated fat, which should provide benefit. However, this exchange also raises serum triglycerides and lowers HDL-cholesterol so whether there is net benefit on blood lipid-related risk for heart disease is an interesting point for debate.

Professor Noakes acknowledged US data showing that those who successfully maintain weight loss tend to be on low carbohydrate diets. However, she said prescriptions for low carbohydrate diets often threw ‘the baby out with the bathwater’, a reference to recommendations to not only restrict intake of nutrient-poor carbohydrate foods such as soft drinks, confectionery, pastries, biscuits and pies but also to restrict nutrient-rich carbohydrate foods such as fruit, bread and breakfast cereals. The benefits of carbohydrate-rich foods in providing essential nutrients and fibre need to be balanced against any risks associated with carbohydrate intake.

Image: source

Risks: sugar or GI?

Discussions on the risks associated with carbohydrate intake revolved around sugar and glycaemic index/glycaemic load, reflecting the current international debate. Dr Barclay discussed the latest review on dietary sugars and body weight conducted for the World Health Organization (WHO) by a team from Otago University in Dunedin. They found that intake of free sugars or sugar-sweetened beverages is a determinant of body weight. Like the review for the Dietary Guidelines for Americans, this review found that the best evidence on sugars and body weight is from studies on sugar-sweetened drinks and perhaps the authors of the WHO review should have limited their findings to beverages. Importantly, their other key finding was that exchanging sugars for other carbohydrates was not associated with weight change. That is, sugar and starch have the same effect!

While you are considering the WHO review take a look at the editorial that accompanied it, written by Walter Willett and David Ludwig. Although generally supportive of the WHO review, these authors take a subtle shot at the WHO for previously disregarding evidence suggesting that starchy products have metabolic effects comparable to those of sugar. They state: Efforts to reduce sugar intake are appropriate, but they should form part of a broader effort to improve the quality of carbohydrates …

This is the nub of the current debate. Carbohydrate quality is not a simple concept that can be reduced down to ‘limit sugar’. It’s complicated and has several dimensions, including nutrient density, dietary fibre/wholegrains and glycaemic index.

Dr Barclay argued that glycaemic index was more important that sugar. He cited the Institute of Medicine’s report on sugar which recommends an upper limit on added sugar intake of 25% of daily calories. But he drew our attention to the rationale for this rather high upper limit, which is not the prevention of disease – it’s to limit nutrient dilution. In contrast, Dr Barclay argued, glycaemic index and glycaemic load are predictors of disease, including obesity, type 2 diabetes and coronary heart disease. His key message was to focus on the body’s response to carbohydrate foods, rather than on whether the carbohydrate was in the form of sugars or starch.

Take-home messages

• Very low carbohydrate diets are not recommended for the general public
• Moderate intakes of nutrient-rich carbohydrate foods are recommended
• Lower GI and high fibre/wholegrain foods are protective

Image: source

One of the most important findings from nutrition research in recent years was that carbohydrate confers similar risk for coronary heart disease to saturated fat. This turned 20 years of dietary advice on its head as it meant that the long-recommended low fat diet was not protective against heart disease. It followed that carbohydrate recommendations were too high and if lower intakes of carbohydrate were to be recommended, it would make sense to preference ‘good’ carbohydrate.

But what is good carbohydrate? Ask this question to a room full of nutritionists and the argument will go on for hours. Thirty years ago starch was considered good and sugars were bad, based on the assumption that starch was digested more slowly and raised blood glucose levels more gradually than sugars. However, studies of glycaemic index showed this assumption to be wrong. So should nutritionists forget about sugar and start recommending low GI foods? There are also arguments over whether ‘wholegrain’ or dietary fibre is the better measure of a good cereal food. In the United States, nutrition authorities have adopted nutrient density as an over-arching principle of their latest dietary guidelines. It would certainly make sense to favour nutrient-rich carbohydrate foods if total carbohydrate intake (or calories) was to be limited. So there are lots of options and opinions but no consensus on what represents good and bad carbohydrate.

Against this swirling background Professor Manny Noakes from CSIRO and I recently published a new model for assessing the nutritional quality of carbohydrate-rich foods.

Assessing nutritional quality

The new model has now been published in the journal Nutrition and Dietetics so it won’t be reviewed in detail here. Instead I’ll highlight some of the more challenging findings.

The model assesses the nutritional quality of carbohydrate-rich foods based on their nutrient density and glycaemic index. Nutrient density was chosen as it reflects the fundamental nutritional role of food – to provide the body with essential nutrients. Glycaemic index was included as it captures how the body responds to the carbohydrate per se. Both measures were thought to be objective and therefore applicable to all carbohydrate-rich foods and drinks. Subjective measures of nutritional quality were avoided e.g. whether foods were natural, fresh, traditionally considered healthy, processed or ‘junk’. The challenge was to demonstrate the nutritional quality of foods.

All results were plotted on the grid that appears in Figure 1, which is broken up into four carbohydrate quality quadrants. Quadrant 1 (high nutrient density, low GI) was considered to be the highest quality quadrant. Quadrant 4 (low nutrient density, high GI) was the lowest quality quadrant. To distinguish between the two intermediate quadrants nutrient density was prioritised over GI.

Figure 1:

The model showed considerable capacity to distinguish between groups of carbohydrate-rich foods. Of particular interest was whether the model was consistent with the dietary advice frequently offered in relation to these food groups. When legumes were run through the model they clustered in Quadrant 1– the highest quality quadrant, which was certainly consistent with the advice in most dietary guidelines (Figure 2).

Figure 2

Milks and yoghurts appeared in the top quality quadrant too. The nutrient density of these dairy foods varied considerably due to the ‘nutrient dilution’ effect of the fat and sugar in some of these foods, but they all still appeared in Quadrant 1. In contrast, soft drinks, confectionery, honey and jam appeared in the lowest quality quadrant. Again, this was very consistent with mainstream dietary advice. So far, so good.

Fruits – a surprise

The model highlighted a few issues that might surprise some nutritionists. Firstly, fruits were not positioned as a tight group in one particular quadrant but were widely dispersed, mainly across Quadrants 1 and 3. In other words, the GI of fruits was typically low but nutrient density varied widely. Although fruits are often described as being nutrient-rich foods, some fruits such as apples, pears and grapes are actually quite nutrient-poor. The nutrient density of apple juice was similar to that of orange cordial. So fruits vary widely in their ability to provide the distinctive nutrients that this group of foods has to offer, though the generally low GIs are a plus.

Cereals – a conundrum

The cereal foods were also widely dispersed, being spread over all four quadrants but they were more concentrated in Quadrants 2 and 4, highlighting major differences in nutrient density. To make sense of the cereal foods I have divided them up into four groups (Figure 3):

• Biscuits, cakes, pastries
• Refined cereal foods (not enriched) – white rice, couscous, polenta, semolina
• Wholegrains (not enriched)
• Enriched cereals foods – breads and breakfast cereals

Figure 3

The model placed biscuits, cakes and pastries exactly where you would expect to find them – clustered in Quadrant 4, the lowest quality quadrant. This is consistent with traditional healthy eating advice to limit intake of these foods. But look where the model placed white rice, couscous, polenta, semolina – in virtually the same place. [Actually, the biscuits, cakes and pastries appear marginally to the left, their sugar and fat content resulting in a little ‘nutrient dilution’. However, refined cereal foods are so low in nutrients to start with that this effect is minimal.]

Until recently, white rice, couscous, polenta and semolina would have been preferred to biscuits, cakes and pastries on the basis of the latter foods’ higher saturated fat content. However, if high GI carbohydrate confers equal (or greater) risk for heart disease than saturated fat, this rationale no longer holds.

The next surprise was the relatively low nutrient density of non-enriched wholegrain foods. They were certainly more nutrient-rich than the refined cereals, notably the wholegrain wheat and oat products, but brown rice was positioned deep in the lowest quality quadrant. Wholegrains are frequently described as being nutrient-dense, but this is simply not the case.

The breads and breakfast cereals mostly fell in Quadrant 2, their nutrient enrichment predictably resulting in higher nutrient densities. The sugar content and the wholegrain status of breakfast cereals were unrelated to nutrient density as this was determined by nutrient enrichment.

What cereal foods should be recommended?

The Dietary Guidelines for Americans (2010) recommended that wholegrain and enriched cereals be included in healthy diets. The logic is clear – the wholegrains provide the fibre in under-supply in the American diet and the enriched cereals provide micronutrients. These recommendations are certainly consistent with our new model.

There is no dispute that biscuits, cakes and pastries should be limited – these foods carry plenty of calories but are nutrient-poor and many have high GIs. In nutritional terms, they have no redeeming features and are logical targets for reduction in healthy diets. However, all of these comments can equally be applied to white rice, couscous, polenta and semolina. Why do Australian nutrition authorities continue to recommend these foods in healthy diets? Doesn’t make sense.

What are your thoughts? How much cereal food should people eat? What type? And why?