Amazing results challenge guidelines in new study

  • A new study challenges the dietary guidelines for heart health
  • MUFAs and omega 6 PUFAs not effective at reducing atherosclerosis risk
  • Omega 3 fish oils reverse triglycerides and weight gain in an animal model of insulin resistance, despite increased calories

Read time: 9 minutes (1400 words)

Guidlines for prevention of heart disease have shifted in recent years away from a simplistic ‘reduce total fat’ message towards a more nuanced emphasis on the type of fat. The current American Heart Association (AHA) recommendation is to ‘replace saturated fats with monounsaturated (MUFA) and polyunsaturated (PUFA) fats.’

For the purpose of this post, I am going to put aside my objection to the demonising of saturated fats and instead focus on the MUFA / PUFA alternatives recommended by the AHA. Similarly, I am not going to challenge the cholesterol hypothesis nor debate the merits or otherwise of lowering LDL cholesterol here. Instead I am going to look at a recent paper that studied at the effects of MUFAs and PUFAs on atherosclerosis risk.

Note the terms n-3, n6- and n-9 in the title of this paper; these are just shorthand for omega-3 polyunsaturated fatty acids (ω3 PUFAs), omega-6 polyunsaturated fatty acids (ω6 PUFAs) and omega-9 monounsaturated fatty acids (ω9 MUFAs). I’ve made a quick reference guide for these below, showing typical foods high in each kind of fat:

The role of triglycerides in cardiovascular disease

Cardiovascular disease is often preceded by insulin resistance during which changes to the cholesterol delivery system increase the risk of atheroma formation in the artery wall. One of the hallmarks of insulin resistance is an increased production in very low density lipoproteins (VLDL) which are high in triglycerides.

The following is a quick explanation about triglycerides which you can skip if it’s a bit too much or you are already familiar with the subject…

The relevance of triglyceride-rich VLDL

VLDL is a carrier protein, produced by the liver, which transports triglycerides from the liver to the adipose (fat) tissue. It is important to note that these triglycerides are primarily endogenous i.e. manufactured by the body, rather than coming directly from dietary sources.

VLDL and triglycerides are raised principally by (1) Excess caloric intake from any source and (2) Carbohydrates (and especially fructose). In the former case the formation of VLDL can be seen as the body packing away and storing excess calories which are transported to the adipose tissue where they can be stored for a rainy day as fat.

In the case of carbohydrates, triglyceride rich VLDL is manufactured as a response to overwhelming surges in blood glucose that cannot be dealt with sufficiently by insulin induced uptake by the organs, especially muscle and liver glycogen stores. Indeed some have said that raised blood triglycerides are a reliable marker of carbohydrate consumption.

If cells become insulin resistant they stop taking up glucose effectively, increases the need for diversion of calories into the endogenous VLDL pathway. Hence insulin resistance is characterised by raised triglycerides.

Fructose has some unique metabolic problems as it does not trigger insulin, so cannot be taken up by cells as quickly as glucose. Instead it has to be processed by the liver, which can easily be overwhelmed, turning the excess into triglyceride rich VLDL.

It should not be surprising then, that the rats used in the study I am reviewing here, were made insulin resistant by feeding with sugar water (30% sucrose water) for 12 weeks. Sucrose is equal parts glucose and fructose, so their water contained 15% of each of these sugars.

So the aim of this study was to see how different dietary fats affect the dislipidemia associated with insulin resistance, especially the triglycerides. To do this the rats in this study were split into 5 groups: A control group on standard diet, whilst the other four were all made insulin resistant by feeding 30% sucrose water for 12 weeks. Of the four insulin resistant groups one was supplemented with n3 PUFAs, one with n6 PUFAs and one with n9 MUFAs.

It’s a pretty obvious experiment to undertake, and at this point you might rightly be asking why? Hasn’t this has all been done before? Surely the science on such a basic question is settled? With 60 years of American Heart Association advice you would think they had the science to back up their assertions and advice, wouldn’t you?

Well, shockingly, you would be wrong. As the authors note:

To our knowledge no studies have addressed the impact of dietary n-3, n-6 and n-9 fatty acids on VLDL composition and size in the [insulin resistant rodent] model.

There are several reasons why this basic question has not been answered before:

  1. Many previous studies looking at the effects of PUFAs on insulin resistance have applied n6 and n3 together, with only a few addressing n6 alone
  2. Previous studies evaluating n3 fish oils have used cod liver oil, which contains high levels of vitamin A, D and cholesterol, which could affect the findings.
  3. Assessment of MUFAs (n9) have usually used olive oil, which contains a broad range of phytochemicals which may be responsible for the beneficial heart effects observed in those studies, rather than the actual monounsaturated fats it contains.

To get round these problems ithe researchers used the following oils:

  • n3: fish oils from pressing whole fish, hence low in vitamin A and D.
  • n6: linoleic acid rich sunflower oil, low in phytonutrients
  • n9: high oleic sunflower oil

Rodent diets contained 15% w/w of each oil, which represents about 35% by calories i.e. similar to a standard western diet.


What they found was striking and deserves some careful reflection.

To make the findings a little easier to appreciate I have made a graph of some of the key results, but tables with all the study data are provided at the end of the post.

Data is expressed as mean percentage differences compared to the standard rodent chow diet (Reference). All four insulin resistant diets (IR) were sucrose rich; the three intervention diets consisted of supplementing with 15% w/w with n-3: deepwater fish oil; n-6: sunflower oil; n-9: high oleic sunflower oil.

Triglycerides ()  and Liver fat ()

The effects of the high sucrose feeding, as expected was a jump in triglycerides, which can be seen between the Reference and IR results above. Dramatically, supplementation with n3 fish oils almost completely reversed this dyslipidemia, returning VLDL particles to normal. Whilst remarkable, this is in line with previous epidemiological, human and animal studies that have shown n-3 PUFA have positive physiological effects on IR and lipid metabolism.

n6 and n9 oils, however, only weakly attenuated these harmful changes, failing to reverse the atherogenic state of the VLDL particles. This casts doubt on the validity of the American Heart Association recommendations.

In the case of the MUFA (n9), previous studies using olive oil have shown greater improvements in insulin resistance parameters, but as already noted, olive oil contains a broad range of bioactive phytochemicals (e.g. sterols and polyphenols). By using high oleic sunflower oil this study has been able to show that MUFAs do not of themselves produce these beneficial effects.

In relation to MUFAs this is particularly important as many processed food uses high MUFA oils that are low in phytonutrients as these can impart undesirable flavours.

Accumulated liver fat follows a similar pattern to plasma triglycerides. Again, n3 oils produce the best reductions in damage caused by insulin resistance.

Weight gain () and Calorie intake ()

Some of the most surprising results were seen in relation to caloric intake and weight gain. All groups of rats could eat ad libitum, yet in the n3 fish oils and n9 MUFA groups caloric intake was considerably raised. Extraordinarily, despite this those fed the n3 fish oils had no weight gain during this trial, whilst those fed the n9 high oleic oil had the most weight gain.

Rats fed the n6 sunflower oil supplemented diet had lower caloric intake than the n3 and n9 groups, but still gained more weight than the n3 group.


The authors of this study conclude:

In insulin resistance, while n-3 PUFA showed expected favorable effects, supplementation with n-6 PUFA and n-9 MUFA did not prevent atherogenic alterations of VLDL. Thus, the recommendations of supplementation with these fatty acids in general diet should be revised.

The authors seem somewhat nonchalant about the n3 fish oils, but it is worth reflecting for a moment just what those fish oils did: the rats were drinking insane quantities of sugar, eating a hyper caloric diet, yet avoided most of the effects of insulin resistance and weight gain. That’s a pretty impressive feat as far as I can see!

Implications for diet

This study looked at the effects of fat choice in the context of insulin resistant animal models. The results support and extend previous research in humans and epidemiological studies. Taken together these point to certain food choices: fish, seafood and olive oil are good choices based on these results; Omega 6 vegetable oils such as sunflower, safflower, corn and soya oils are best avoided, as are low-polyphenol MUFAs like high oleic oil and possibly filtered rapeseed (Canola) oil. A high quality fish oil supplement seems prudent too.

Based on the ideas suggested by this study cold pressed rapeseed oil is potentially interesting as like extra-virgin olive oil, it contains high levels of phytonutrients, but unlike olive oil it has significant levels of alpha linolenic acid (ALA), a short chain n3 fatty acid. Based on the results above I would expect it to have beneficial metabolic effects possibly similar to or slightly better than Olive oil. Well that’s my prediction. So lets see…

I searched Pubmed for “rapeseed cardiovascular”. Sure enough, in one of the first studies I found [Baxheinrich et al, 2012] patients with metabolic syndrome (which is just one step down from full blown insulin resistance), were placed on a low calorie diet enriched either with olive oil (high n9 MUFA, low n3 ALA) or cold pressed rapeseed oil (high n9 MUFA high n3 ALA) for six months. Although both groups improved similarly on many metabolic measures (body weight, systolic blood pressure, cholesterol, and insulin levels), the cold pressed rapeseed group had significantly lower triglycerides. That said rapeseed oil is probably less suitable than olive oil for high temperature cooking as the ALA it contains is very heat sensitive. Still, its a good choice for salad dressings and mayonnaise!

Additional data

For those of you who like to dig into the data here are some key tables from the study:

(1) Composition of diets; (2) Intakes and body weight; (3) Effects on adipose tissue/liver and serum parameters


Gluten Update April 2017

The effects of gluten are not confined to just the 1% of the population that suffer with Coeliac Disease and the estimated 6% who suffer with Non-Coeliac Gluten Sensitivity, but can have detrimental effects in everyone. In these updates we share some of the latest research.

  • Effects of gluten on gluten-tolerant mice
  • Relatives of coeliacs often have gluten related disorders
  • NCGS persists even after 8 years on gluten free diet
  • Gluten free diet prevents progression of potential coeliacs
  • Coeliac disease may be triggered by a common virus
  • Coeliac disease and joint and bone problems

Read time: 11 minutes (2000 words)

Effects of gluten in otherwise gluten-tolerant mice

Dietary gluten causes severe disorders like celiac disease in gluten-intolerant humans. However, currently understanding of its impact in tolerant individuals is limited.

A ground breaking study has undertaken the first detailed investigation into the effects of gluten on metabolism and microbiome in gluten tolerant mice. The purpose of the study was to identify the effects of an obesogenic diet with or without gliadin (a key, immunogenic component of gluten).

The team from The National Food Institute, Denmark (Li Zhang et al, 2017) analysed a huge array of parameters, including insulin resistance, histology of liver and adipose tissue, intestinal microbiota in three gut compartments, gut barrier function, gene expression, urinary metabolites and immune profiles in intestinal, lymphoid, liver and adipose tissues.

Using tightly controlled diets the researchers found that a relatively small change (exchanging casein for 4% gliadin) resulted in a considerable impact on the host response in the mouse model. The levels of gliadin used were comparable to that found in bread. The remainder of the diet was, however, higher in fat than a typical human meal (35g fat/100g).

The headline findings include:

  1. Gliadin Intake Affected Glucose and Lipid Metabolic Homeostasis
  2. Gliadin Intake Altered Gut Microbial Composition and Activity
  3. Gliadin Intake Caused Lower Expression of Gut Barrier Function Related Genes in Ileum
  4. Gliadin Intake Changed the Metabolic Signature of Urine
  5. Gliadin Did Not Affect Systemic Inflammatory Markers but Altered Immune Cell Composition in Liver and Inflammatory Phenotype of Visceral Adipose Tissue

The changes were predominately negative, but there is a huge amount of detail to unpack in this research, cetainly more than I want to go into in this post. That said, I will look a little more into the changes to the microbiome as they are particularly fascinating. If you are interested in going into the other areas more deeply then check out the full free text.

Changes to the intestinal microbiome

After nine weeks on the 4% gliadin diet levels of lactobacillus had fallen by more than 90%. (These are generally considered beneficial bacteria). Conversely, bacteria associated with detrimental health changes including Clostridium XI, Dorea and Coriobacteriaceae increased in abundance by more than ten fold.

Strains belonging to Clostridium XI, including also the opportunistic pathogen C. difficile, are associated with compromised health. Dorea spp. are found to be overrepresented in irritable bowel syndrome patients, and patients with non-alcoholic fatty liver disease. Coriobacteriaceae spp. have repeatedly been shown to be involved in host lipid metabolism, and many bacteria within this group are considered as opportunistic pathogens

One apparently positive change observed in the gliadin consuming mice was an increase in Akkermansia in the colon. This species is usually associated with beneficial effects on metabolic health and inflammation. However, it feeds primarily on mucin secreted from the gut wall, and the researchers suggest its proliferation may be due to increased turnover of the small intestine mucosa due to disrupted gut barrier function.

Another recent paper (The gut–kidney axis in IgA nephropathy: role of microbiota and diet on genetic predisposition, Coppo, April 2017) underscores the importance of gluten-microbiome interaction, in this case to the development of IgA nephropathy (Berger’s disease).

The importance of this research is that it is a major piece of work exploring the effects of gluten relevant to the ‘normal’, apparently gluten-tolerant population. It has shown a large range of measurable effects which will need to be investigated further, particularly to establish their relevance in humans. In terms of the gluten iceberg these researchers are undertaking the deep sea diving necessary to establish the extent of the sub surface portion of the gluten phenomena. I am sure we will see more research like this in the coming years.

First degree relatives of Coeliacs often have gluten related disorders despite testing negative for blood markers 

Here at the clinic we see a disproportionate number of patients who at some point in the consultation declare that one of their parents or siblings has coeliac disease, often as if this is some minor point that is hardly worth mentioning. Nothing could be further from the truth: The biggest risk factors for coeliac disease is being a first degree relative of a coeliac, with nearly four times the incidence than the general population.

The first-degree relatives (FDRs) of patients with coeliac disease are the main risk group for disease development.

– Vaquero et al, 2017

Parents, children or siblings of coeliacs should therefore have blood tests regularly, and if positive should immediately adopt a gluten free diet. A study by Nicola Imperatore et al, which I cover later in this post, has demonstrated the importance of early adoption of a gluten-free diet for anyone with positive coeliac blood tests, even when they have no  symptoms, to avoid progression of intestinal damage and immunological problems.

Some patients, however, claim that they have been screened for coeliac disease and have been told ‘they don’t have it’ – i.e. that their blood markers were negative. In many cases, however, I can tell from their symptoms that they are gluten sensitive despite the test results. Many go on to make remarkable progress once they adopt a properly gluten-free diet, but sometimes it can be an uphill struggle convincing them to give it a try. Such is the power of the medical establishment when it makes its decrees.

These clinical observations have recently received confirmation from a study published in the Journal of Gastroenterology and Hepatology (Vaquero et al, 2017), which investigated the value of a gluten free diet in first degree relatives who were negative for coeliac blood markers (anti-endomysial antibodies and raised anti-tissue transglutaminase ). In this study they invited 205 first degree relatives of known Coeliacs who were negative on blood tests to undergo genetic screening (for the HLA-DQ2/8 genes) and duodenal biopsy to check for intestinal damage. Symptoms were established by questionnaire at the start of the study, after which participants followed a three phase diet: (1) baseline diet (gluten containing) (2) gluten free diet (4 weeks) then (3) gluten ‘overload’ diet

Of 139 who completed the study HLA-DQ2/8 was positive in 78.4% of the participants (homozygous, 15.1%; heterozygous, 63.3%). Alterations to the intestinal mucosa were noted in 37.1% of participants who underwent duodenal biopsy (Marsh I, 32.7%; Marsh IIIa, 4.4%). At baseline more than half of the participants had gastrointestinal symptoms (57.6%), mainly associated with bloating (16.5%), constipation (15.1%), diarrhea (14.4%), and abdominal pain (5.8%). During the gluten-free phase this fell to just one quarter and increased agin during the gluten overload phase.

Symptom improvement during the gluten free diet was twice as common among women as men and three times as common among people with an existing autoimmune disorder. Both of these observations fit clinical patterns at Rosemary Cottage Clinic and are similar to risk factors identified in studies of Non Coeliac Gluten Sensitivity.

In conclusion: First degree relatives of coeliacs have a heightened risk of developing coeliac disease. Of those without coeliac blood markers many have intestinal damage. In addition gluten related symptoms are commonly present independent of intestinal damage. Such people would have to be classified as Non Coeliac Gluten Sensitive as they do not fit the definition of Coeliac.

Non Coeliac Gluten Sensistivity persists even after 8 years of a wheat-free diet

In a paper recently published in Gastroenterology (Carroccio et al, 2017) Italian researchers followed up 200 patients that had been diagnosed with NCGS many years earlier. 88% had experienced improved symptoms following their original diagnosis.

8 years later 148 of these individuals were still on a strict wheat-free diet, with virtually all of them (98%) reporting a continuation of symptom reduction. Among those that had not maintained a strict WFD only 58% had symptoms that were improved compared to the time of diagnosis.

The researchers repeated the double-blind placebo-controlled challenge with 22 patients, and found that 20 reacted to wheat.

This study demonstrates that NCGS is persistent, suggesting that it should be treated as a life long condition and a wheat-free diet adhered to.

A gluten-free diet prevents progression in Potential Coeliac Disease even when asymptomatic

Coeliac disease is diagnosed where patients have positive  blood tests (Anti-endomysial antibodies and raised anti-tissue transglutaminase) as well as evidence of villous atrophy on duodenal biopsy. In some cases, however, blood tests are positive, but there is little or no intestinal damage. Such situations are labelled Potential Coeliac Disease (PCD).

Until now it has not been clear whether such patients simply have a mild form of gluten intolerance, or whether they have early stages of Coeliac disease.  Furthermore, for the subset of PCD patients that have no gastrointestinal symptoms.

To answer these questions recent study an Italian team (Nicola Imperatore et al, Mar 2017) followed patients with PCD either on a gluten free diet or a gluten containing diet for a period of six years.

In short, those on the the gluten containing diet had increased intestinal damage and immune related disorders compared to those on the gluten-free diet. In addition, the asymptomatic patients who continued to consume gluten 69% developed coeliac-related symptoms, 46% developed villous atrophy and 61% immune mediated disorders.

This study underlines the importance of starting a gluten free diet as soon as possible following positive blood tests, regardless of symptoms and presence of villous atrophy.

Routine screening for Coeliac blood markers is not currently undertaken as many people that such tests would identify would prove to have no symptoms nor villous atrophy. It has been assumed that there would be no point in starting them on a gluten free diet as it was assumed that they would not go on to develop coeliac disease. This study challenges this position and demonstrates that many of these ‘false positives’ would indeed benefit from starting a gluten-free diet because they are at risk of developing villous atrophy and autoimmune diseases in the long term..

Coeliac Disease may be triggered by a common virus

It has long been known that the risk of coeliac disease is increased after various infections such as campylobacter, rotavirus and gastroenteritis. One hypothesis is that the presence
of both the pathogen and gluten presented simultaneously to the immune system  is enough to trigger coeliac disease.

Now a team from the University of Chicago, has found that exposing mice to a common reovirus called T1L can induce gluten intolerance. This virus was first identified in humans in the 1950s but has not been associated with any disease.

The team found that when they fed gluten to mice, those that were also infected with the virus produced two to three times as many antibodies to gluten as those that were virus free.

One of the Authors of the study Bana Jabri explained “Instead of mounting a tolerant, non-aggressive response, the immune system in the presence of the reovirus views gluten as being dangerous, promoting a destructive inflammatory response,”

The discovery has led the authors to speculate that a vaccine might be possible to prevent people from developing coeliac disease. Personally I think that’s unlikely, for a number of reasons. 1) gluten seems to cause disease by many pathways not just an aggressive immune response, 2) the researchers did not show that the mice went on to develop coeliac disease they just had a heightened immune response to gluten, 3) there is no evidence that all coeliacs had previously been infected with reovirus, and 4) physiological stressors other than pathogens appear to be able to trigger coeliac disease.

Source: New Scientist, Gluten allergy in coeliac disease may be provoked by virus (04/06/2017)

Coeliac Disease and Joint/Bone problems

In a recent letter to the journal Joint Bone Spine, Coline Daron et al report on their analysis finding 20‐30% cumulative incidence of arthralgia and arthritis among coeliac patients. They also identified an increased risk of osteoporosis:

Out of 11 case‐control studies featuring 1,008 patients with celiac disease and 13,706 controls, we noted a 2.73 [1.86‐3.99] higher risk of osteoporosis at any site in the celiac group. This increased risk was significant for femoral osteoporosis (OR=2.03 [95% CI: 1.11‐3.71]), and most of all for spinal osteoporosis (OR=7.2 [95% CI: 3.42‐15.18]. No increased risk of arthritis was noted in celiac patients compared to controls (OR=0.76 [95% CI: 0.16‐3.66]).

The authors call for physicians to undertake coeliac screening in cases of osteoporosis, arthralgia and arthritis where no obvious cause can be identified.


Salt and cardio-vascular disease: Policy and Science clash

The recent video we posted of Dr SalimYusuf’s PURE study had a section on sodium intake, where he showed that the lowest risk of cardiovascular events, cardiovascular deaths and all-cause mortality was associated with an intake between 3000 and 6000 mg of sodium per day (equivalent to 7 to 15g salt per day). The current US average sodium intake is 3800 mg placing the general population nicely within this sweet spot, although towards the lower end.

Current US and UK dietary recommendations recommend an upper limit at 2300mg of sodium (6g of salt) whilst cardiovascular recommendations by bodies such as the American Heart Association aim to reduce sodium intake to 1500 mg per day (approx 3.75 g salt). If the PURE study is right (and it is not alone in questioning the current guidelines), then these aspirations would do more harm than good.

How did such discrepancy arise? The problem may be the use of surrogate markers. The thinking goes like this: Salt raises blood pressure. Raised blood pressure increases CVD risk, so salt increases CVD risk. This kind of thinking was evident in 2011 when the American Heart Association (AHA) called for salt targets to be reduced to 1500mg per day.At the time MedPage Today explained:

The evidence linking salt intake with blood pressure — and the major adverse outcomes of heart disease, stroke, and kidney disease — is “impressive,”…

That evidence includes more than 50 trials assessing the blood pressure effects of salt, as well as a meta-analysis showing that cutting salt intake by about 1,800 mg per day lowered blood pressure by 5 mm Hg systolic and 2.7 mm Hg diastolic.

This is a “critically important public health issue,” according to Appel and colleagues, and this AHA advisory must be considered “a call to action.”

On the basis of this ‘A leads to B leads to C, therefore A leads to C’ thinking initiatives were instigated all round the world to reduce public consumption of salt. A task force of concerned scientists even formed a lobby group to put pressure on food manufacturers, which successfully led to reductions in added salt in manufactured foods.

However, within a short time of the AHA call to action reports started coming in contradicting this advice.

Over this period it is clear that scientists were becoming more and more irritated with the dogmatic approach of the AHA and government bodies, and by the last article were publicly calling the AHA anti-scientific!

Despite all of the research questioning the validity of further salt reduction US and UK policy remains stubbornly wedded to the ‘less is best trajectory’. In their 2016 survey the UK government reported proudly that average sodium consumption fell from 3500mg in 2005 to 3200 mg in 2014.

Their report claimed “Too much salt in the diet can raise blood pressure which increases the risk of heart disease and stroke. A reduction in average salt intake from 8g to 6g per day is estimated to prevent over 8000 premature deaths each year and save the NHS over £570million annually.”

Yet contrary evidence from studies including PURE would suggest that this is not simply futile but probably harmful. You would think that with the swathe of research challenging the low salt dogma that public policy would be questioning the wisdom of further reductions. Not a bit of it. Dr Alison Tedstone, chief nutritionist at Public Health England, makes no bones about it:

Our analysis makes clear that there is a steady downward trend in salt consumption. While people are having less salt than 10 years ago, we are still eating a third more than we should.

Many manufacturers and retailers have significantly reduced the salt levels in everyday foods. However, more needs to be done, especially by restaurants, cafes and takeaways.

The intransigence of health policy makers leads researchers to exasperation and despair. As one writer put it:

…the ‘salt hypothesis’ is rather like a monster from a 1950s B movie. Every time you attack it with evidence it simply shrugs it off and grows even stronger. – Malcolm Kendrick

In an interview with MedPage today researchers who found that patients with heart failure who ate more salt did better than those who ate less made the following statement which we have published before, but is such a gem it deserves another outing:

“We have had no basis for any of our recommendations regarding sodium restriction during the past 50 years, although these recommendations have changed a great deal (for no good reason). After this report, we still have no basis for any of our recommendations regarding sodium restriction. We were ignorant before; we are not any smarter now. Did we really need this report to tell us that we lack evidence for our recommendations regarding dietary sodium in patients with heart failure?”
Milton Packer, Professor in the Division of Cardiology, UT Southwestern

Further reading:

Salt vs sodium measurements

We made a boo-boo in out recent post (Cardiologist attacks diet dogma at 2017 symposium) where we summarised the findings from the PURE study regarding salt intake. We originally stated that 3 to 6 grams of salt per day appeared optimal but this should have read 3 to 6 grams of sodium per day.

What’s the difference?

Salt is a simple compound sodium chloride, NaCl, composed of sodium and chlorine atoms in a 1:1 ratio. Sodium has an atomic mass of 23 and chlorine a mass of 35, so the the fraction by weight of sodium in salt is 23/58 = 40%; or said the other way round, 1 gram of sodium is found in 2.5g of salt. To make things even more confusing, sodium is often quoted in milligrams (mg) whilst salt is given in grams, so the conversion becomes: 1000mg sodium = 2.5g salt

Let’s put this to the test with a confusing pair of health policies: UK guidelines recommend you eat no more than 6g of salt per day, whereas US guidelines place the limit at 2,300 mg of sodium per day. How do these compare?

First, converting the US 2,300 mg of sodium to grams gives 2.3 g of sodium. Next, convert this to the equivalent amount of salt by multiplying by two and a half: 2.3g x 2.5 =  5.75g. This figure rounds to 6g. i.e. they are essentially recommending the same thing, but expressing them in different ways. (That’s the special relationship for you!)

Here is a handy table for converting between sodium, salt and teaspoons:

Salt in
Sodium in mg Approx. equivalent to                Guidelines – daily limit
0.5 200 Average pinch of salt
2.5 1000 Half a teaspoon salt
3.75 1500 ¾ a teaspoon salt Recommended (AHA)
5 2000 One teaspoon salt
6 2400 1¼ teaspoons salt Upper limit (UK / US)
10 4000 2 teaspoons salt Current average consumption
15 6000 3 teaspoons salt Upper limit (PURE study*)

*The PURE study found that the lowest risk of cardiovascular and all-cause mortality was associated with a sodium intake of 3000 to 6000 mg per day. A concern I will look at in the next post is that public health policy does not take into account the lower limit, and assumes that less salt is always better. PURE and other studies suggest otherwise!

This sodium/salt mistake crops up a little too often and leads to confusion. For example MedPage Today, a respected medical news site, reported in 2011 that the American Heart Association had called for salt intake to be limited to 1,500 Mg. (I will write more about the conflict between this figure and the findings of the PURE study in my next post)

First off, I’m sure they meant milligrams (mg), not mega grams (Mg). Autocapitalising their title, put them out by a ‘trifling’ factor of one billion, but that’s forgivable. Where they really sowed confusion was by muddling up sodium and salt like I did. The AHA were calling for sodium intake to be reduced from 2300mg per day to 1500 mg per day (for adults), but MedPage reported these figures as salt not sodium.

One bemused commenter wrote “This article suggests 1 1/2 gms of salt a day, in the UK we are told 6gms per day”

So to clarify, the UK and US recommended upper limit is 2300 mg sodium per day (= 6 g salt), and the AHA recommendation is 1500 mg sodium (=just under 4g of salt). In contrast the PURE study found the ideal range was 3000 to 6000 mg sodium (=between 7.5 and 15g of salt per day).

Why then do researchers and nutrition labels quote sodium, not salt quantities? Because in principle at least, you could get sodium from sources other than sodium chloride. In practice non-salt sources of sodium are insignificant.

Anyway, hope that clears up the confusion about measuring sodium levels.

Next up I’ll tackle how the national guidelines are pushing us in the opposite direction to that suggested by the PURE study.

Cardiologist attacks diet dogma at 2017 Symposium



Dr Salim Yusuf speaking at the Cardiology Update 2017 symposium gives preliminary findings from the PURE study which followed 140,000 people in 17 countries, designed to address causation of cardiovascular disease.

He explains that the results indicate:

  • Greater fat intake is protective
  • Carbohydrates are harmful
  • High fat dairy is beneficial
  • Saturated fat from meat is neutral
  • Fruit is beneficial, but no additional benefit over 2 portions per day
  • Legumes are beneficial
  • 3 to 6 g/day sodium* intake optimal (vs US guidelines of 1.5g)
  • Eggs, fish and vegetables were neutral

*Edit 10/03/2017
Correction: I previously wrote 3 to 6 grams of salt per day, but have corrected this to 3 to 6 grams of sodium per day. This is equivalent to 7.5 to 15 g of salt per day.

Gluten – what we learned in 2016 (part 2) – the Great Imitator

20+ conditions related to gluten – a review of some 2016 papers

gluten-related-disorders-2016-20🔍 Please feel free to share this infographic or the whole post!

In part 1 we looked at some of the key developments in understanding gluten sensitivity (coeliac and non-coeliac gluten sensitivity) that emerged in 2016. In part 2 now we take a look at some case studies, and small trials that link gluten to a wider range of conditions.

Read time: 18 minutes (2400 words)

Some of the conditions that were linked to gluten in 2016

When I did my medical training I was examined on clinical diagnostic skills. We students of herbal medicine, like other medical students, were taught that whatever symptoms a patient presented with, our list of possible diagnoses we could always include tuberculosis and syphilis. These two systemic diseases were called the great imitators due to their potential for affecting any tissue or organ. From back pain, to skin rashes and mental illness, these ancient adversaries of Man could be the cause. Misdiagnosis could be serious in that the wrong treatment might be pursued, and the effective treatment missed. What we are seeing with gluten related disorders is another great imitator at play. The conditions that follow will leave you in no doubt that gluten and/or other factor in cereal grains, can play havoc with every corner of the body and should always be considered, properly tested for,m and ruled out by the responsible and updated clinician.

In reading this post there are several points to bear in mind: (1) This is far from an exhaustive list, even of 2016 papers. (2) Gluten related disorders often go unsuspected, undiagnosed and therefore untreated for many many years and 80% of coeliacs remain undiagnosed! (3) Gluten can affect any system or tissue type of the body and should always be considered as part of a differential diagnosis even if there are no intestinal symptoms. A full gluten blood test, not just the standard NHS coeliac tests, should be used, or a minimum of a 6 week exclusion diet. (4) The true extent of the gluten iceberg’s sub-surface volume has yet to be established. These reports give some insight into the extent of the murky mass.

I have seperated these reports into those related to coeliac – where there is adaptive immunity and/or autoimmune factors involved and non coeliac gluten sensitivity (NCGS) – where the innate immune system is principally involved. The latter is particularly important as it indicates the propensity of gluten to affect potentially anyone.

1. Coeliac (autoimmune) related

Coeliac hepatitis: hepatic fibrosis, advanced steatohepatitis and cirrhosis
In a Nov 2016 case study, a 20 year old woman with coeliac disease and severe liver cirrhosis had a near complete reversal once placed on a gluten-free diet.The author called this a “hepatomiracle” [Gaur, Nov 2016]
An Iranian study published in Jul 2016 found that the prevalence of coeliac disease among patients with epilepsy was 6% – about five times higher than in the general population. [Bashiri, Jul 2016]
Chronic Hepatitis C Virus (CHV)
CHV infection can lead to autoimmune diseases and shares one of the same genetic loci as coeliac disease (HLA-DQ2). Liver disease also leads to anti-tissue transglutaminase antibodies (anti-tTG). Some CHV patients develop coeliac disease during interferon therapy. There is a complex relationship with many overlapping features between these two conditions, although it is not yet clear whether the incidence of coeliac disease is higher in CHV or not [Association between celiac disease and chronic hepatitis C, 2016].
Distal Renal Tubular Acidosis Associated with Celiac Disease and Thyroiditis
A case report of a 12 year old girl with a particular form of kidney disease, (distal renal tubular acidosis, RTA) with autoimmune diseases, which is extremely rare in children. [Indian Pediatrics, Nov 2016] “Despite resolution of acidosis on bicarbonate, she continued to have poor growth and delayed puberty. Investigations revealed autoimmune thyroiditis and celiac disease. Levothyroxine and gluten-free diet were initiated. Child gained height and weight and had onset of puberty after gluten withdrawal.”
Psychiatric Case
The BMJ recently published a case study of a girl admitted to a psychiatry ward suffering with suicidal behaviours who then developed an agitated catatonic state. She was unresponsive to antidepressants, anxiolytics, antipsychotics and electroconvulsive therapy, but improved significantly when a gluten-free diet was started [Oliveira-Maia, Dec 2016]. Another recent paper The progression of coeliac disease: its neurological and psychiatric implications, Campagna G, Dec 2016, explores the current understanding of the neurological implications of coeliac disease.
Burning Tongue
An elderly woman presented with complaints of a burning tongue for the past two years as well as occasional loose stools and fatigue. Tests revealed iron deficiency anemia, zinc deficiency and an abnormal celiac panel. Ten weeks on a gluten free diet led to complete symptom resolution [Sherman, Jun 2016].
Macrophage activation syndrome
A case report in Pediatric Rheumatology Online Journal, reports on a six year old girl who was diagnosed with Macrophage activation syndrome – an autoinflammatory or rheumatic disease involving hyper inflammation and an ineffective immune response. Serology indicated coeliac disease and symptoms stabilised with the introduction of a gluten-free diet. Authors state that “Clinicians should have a low threshold for screening children with other autoimmune diseases for coeliac disease.” [Palman, Dec 2016].
Coeliac like disease in dogs
The Veterinary Record [Lowrie, Dec 2016] reports on a case of “gluten-sensitive dyskinesia (previously termed canine epileptoid cramping syndrome) is a condition of Border terriers in which the leading manifestation is neurological… responsive to a gluten-free diet.” As such, the authors suggest that gluten sensitivity in Border terriers “may manifest as a multisystem disease in a similar manner to that seen in human beings.” There is another paper on this topic: Gluten exposure and multisystem disease in dogs [Davies M, Dec 2016], but I have been unable to access it.
Gluten free diet in pregnancy and type 1 diabetes in offspring
As we have previously discussed, gluten has a significant role in the development of type 1 diabetes. A study in the Journal of Diabetes Research [Antvorskov JC, Aug 2016] investigated the a mouse model of type 1 diabetes. Withholding gluten during pregnancy prevented the subsequent development of type 1 diabetes in offspring, even when exposed to gluten after birth. However, the effect disappeared if the mother received a gluten-free diet prior to pregnancy.
Aortic stiffness may explain increased cardiovascular risk
Many studies show an increased risk of cardiovascular disease in coeliac patients which cannot be explained by traditional risk factors. A study from Antalya, Turkey used echocardiograms to compare the aortic function of 81 coeliac patients with that of 63 healthy volunteers. They found an increased level of aortic stiffness and inflammation in coeliac patients. Whilst inflammation decreased with adherence to a gluten free diet, aortic stiffness did not, suggesting that increased cardiovascular risk may persist despite a gluten-free diet. [Bayar, Mar 2016]
Hemophagocytic lymphohistiocytosis
Hemophagocytic Lymphohistiocytosis and is a life-threatening immunodeficiency. It affects people of all ages and ethnic groups. Common symptoms are fevers, enlarged spleen, low blood counts and liver abnormalities. [ref] A recent paper [Fordham NJ Sep 2016] reports a case that did not respond to standard treatment, but following blood tests established undiagnosed coeliac disease. “She initially responded to chemoimmunotherapy specific for hemophagocytic lymphohistiocytosis but relapsed within a few months of cessation of treatment and then achieved complete remission on gluten withdrawal alone.”
Down’s Syndrome
A study in Poland identified high levels of diagnosed coeliac disease among patients with Down’s syndrome (5.4% vs 1% in general population). The authors emphasise that tests for coeliac disease should be carried out in all (Polish) patients with Down’s syndrome, regardless of the clinical picture. [Szaflarska-Popławska, 2016]
Multiple autoimmune syndrome with celiac disease
Reumatologia published a case report of a 32 year old woman who had four co-existing autoimmune diseases: autoimmune hypothyroidism, Sjögren’s syndrome, systemic lupus erythematosus (SLE) and celiac disease which leads to the final diagnosis of multiple autoimmune syndrome type 3 with celiac disease. The authors point out that patients with single autoimmune disorders are at 25% risk of developing other autoimmune disorders. The case emphasises the need for continued surveillance for the development of new autoimmune disease in predisposed patients. [Harpreet, Dec 2016]

2. Non-coeliac Gluten Sensitivity (NCGS)

NOTE: I am using NCGS as a looser category than is currently accepted to include any studies where withdrawal of gluten or wheat showed clinical benefit.

Intestinal cell damage and systemic immune activation in NCGS
A study published in the BMJ’s journal Gut at the end of 2016, examined serum from 80 individuals meeting the criteria for NCGS (although the authors use the term Non Celiac Wheat Sensitivity). They found increased levels of lipopolysaccharide binding protein – indicating that these patients had raised levels of gut bacteria products passing through a damaged gut wall, known as ‘leaky gut’. Associated with this was raised markers of systemic immune activation. The researchers went on to find raised levels of fatty acid binding protein, indicative of intestinal cell damage. As NCGS patients do not have villous atrophy (which is characteristic of celiac disease) it had been assumed that there is no intestinal damage taking place in NCGS, however, this study shows that damage is indeed taking place. The authors speculate that damage in NCGS may be taking place in the mid section of the small intestine, the jejunum, rather than the first section, the duodenum, from which biopsies are usually taken. Finally, these markers were found to improve with the initiation of a gluten free diet. [Melanie Uhde, 2016]
Psychotic Illness
Schizophrenia has previously been linked to raised kynurenine and reduced tryptophan [Chiappelli, Nature, 2016], both of which are related to the production of Vitamin B3. Researchers hypothesised that this imbalance could be caused by inflammatory immune mediators such as gluten. They found schizophrenic patients had raised levels of anti-gliadin antibodies (IgG) which correlated with kynurenine/tryptophan ratios. They concluded “Our results connect nonceliac gluten sensitivity with the KYN pathway of TRP metabolism in psychotic illness” [Okusaga, 2016]. This links to an long used treatment of schizophrenia which is goodly doses of Vitamin B3 (niacin), the production of which may be getting interfered with by gluten in these patients.
Atopic Dermatitis (Eczema)
A recent study sent 169 atopic dermatitis patients a 61-question survey asking about dietary modifications they had tried and their perceptions and outcomes of such trials. The most common foods eliminated were ‘junk foods’ (68%), dairy (49.7%), and gluten (49%). The best improvement in skin was reported when removing white flour products (37 of 69, 53.6%), gluten (37 of 72, 51.4%) and the nightshade family of vegetables, i.e. potatoes, tomatoes, aubergines, peppers, chilli and paprika (18 of 35, 51.4%). [Nosrati, 2017]
Lymphocytic colitis
(This condition is characterised by chronic watery diarrhoea yet with normal colon cells when sent to the histology lab, but with an accumulation of lymphocytes in the colonic epithelium/lining)
A study in the journal PLoS One found that 91% of lymphocytic colitis patients who were identified as NCGS responded to a double blind gluten challenge indicating a causative role for gluten in these patients. [Rosinach M, Jul 2016]
Nephrotic Syndrome
This is a rare condition in childhood that presents with proteinuria, hypoalbuminemia, and oedema. Kidney function is usually normal however. Most children (>90%) respond to an initial course of oral steroids and are designated as having steroid-sensitive nephrotic syndrome (ssNS). A study in Pediatrics placed 8 children with difficult-to-manage disease (characterized by steroid dependence or frequent relapses) on a gluten-free diet. They all had clinical improvement enabling reduction or discontinuation in steroids. The role of gluten in this condition was confirmed through relapse following re-exposure to gluten. The authors conclude “Elimination of gluten from the diet, may reduce the need for potentially toxic immunosuppressant therapies” [Lemley KV, Jul 2016]
Microscopic colitis
– A new clinical and pathological entity (“lymphocytic enterocolitis”)?
Of patients with Marsh 1 duodenal damage, but no coeliac serology (i.e. negative for anti-endomysium and anti-tissue transglutaminase), half were found to have microscopic colitis. Although this study was not specifically looking at gluten, 14 patients tried a gluten free diet for at least one month and 3 of them (21%) had improvements. [Bonagura, Nov 2016]
Postural orthostatic tachycardia syndrome (POTS)
A study published in the European Journal of Gastroenterology and Hepatology found that 4% of PoTS patients had coeliac antibodies, yet 42% reported gluten sensitivity, suggesting that many POTS patients are NCGS. [Penny HA, Dec 2016]
Eosinophilic oesophagitis
Eosinophilic oesophagitis (also called EoE for those who spell oesophagus without the ‘o’) is a chronic inflammatory oesophageal disease triggered predominantly by food antigens. Although considered a food allergy, EoE is unique in not involving the immunoglobulin IgE antibody response. A recent review identifies milk and gluten elimination as the most promising cure strategies. [Molina-Infante, Dec 2016] How is EoE caused? Until recently it was considered that the oesophagus was relatively impermeable to food antigens and allergic diseases of the oesophagus were unknown. Researchers from the Mayo Clinic wondered if food antigens were present in the oesophageal tissue of sufferers. In a recently published paper they report finding increasing levels of gliadin (gluten) in the oesophageal tissue with increased severity of disease, whilst none was present amongst controls. This suggests a direct causative role of gluten in EoE.


Comparison of total anti-gliadin staining (y-axis) in patients on a gluten free diet (GFD), control patients and patients with inactive and active EoE on gluten (x-axis, 63x). E. V. Marietta, AP&T Nov 2016

Effect of a gluten-free diet in children with autism spectrum disorders
In this randomised clinical trial [Ghalichi F, Nov 2016] 80 children with ASD were randomised to gluten-free diet [40] or regular diet [40].

“In the GFD group, the prevalence of gastrointestinal symptoms decreased significantly (P<0.05) after intake of GFD (40.57% vs. 17.10%) but increased insignificantly in the RD group (42.45% vs. 44.05%). GFD intervention resulted in a significant decrease in behavioral disorders (80.03±14.07 vs. 75.82±15.37, P<0.05) but an insignificant increase in the RD group (79.92±15.49 vs. 80.92±16.24).”

Alcohol related cerebellar degeneration
Alcohol-related cerebellar degeneration is one of the commonest acquired forms of cerebellar ataxia, however, the mechanism by which alcohol causes this damage is unknown. The cerebellum is the back part of the brain, responsible for many basic things including walking. ‘Ataxia’ is difficulty in walking. In a recent study from Royal Hallamshire Hospital, Sheffield, England, a group of 38 patients with ataxia were studied. 34% were found to have circulating antigliadin antibodies vs. 12 % in healthy controls, and 39 % were found to have antibodies to transglutaminase 6 (which are neurological antibodies) vs. 4 % of healthy controls. The authors suggest that chronic alcohol abuse increases gut permeability exposing the immune system to increased levels of gluten peptides to which these two antibodies are raised. They conclude “Alcohol induced tissue injury to the central nervous system leading to cerebellar degeneration may also involve immune mediated mechanisms, including sensitisation to gluten.” [Shanmugarajah PD, Oct 2016]


We are in the middle of an epidemic of gluten related disease that have crept up on us over the last few decades. Our European culture has lived with wheat and its associated diseases, for thousands of years, and we have named and described them, yet only now are we becoming aware of the true cost to our health. At the same time, many Central America countries (Guatemala, El Salvador, Honduras, Nicaragua, Costa Rica, and Panama), which historically have had very low rates of coeliac disease, are transitioning from a diet based on maize to one increasingly centred around wheat. Maize too has its problems but the prolamine in maize, zein, is less likely to cause immune reactions than is the prolamine in wheat, namely gluten.

A paper published in the journalNutrients  [Amado Salvador Peña* and Jakob Bart Arie Crusius, Sep 2015] provides a salutary reminder of the scope of the problem:

[these] changes permit a prediction of an increase of celiac disease and other autoimmune diseases such as type I diabetes and thyroid disease.

The aim of this review is to… alert authorities responsible for the planning of education and health, to find possibilities to avoid a rise in these disorders before the epidemics start.

– Amado Salvador Peña and Jakob Bart Arie Crusius, Central America in Transition: From Maize to Wheat. Challenges and Opportunities (Nutrients, Sep 2015)

Gluten – what we learned in 2016 (part 1)

toast-head-by-rysunek-kuczynski▲ Image: Pawel Kuczynski

“Give us this day our daily bread (…) but deliver us from evil”

—Matthew 6:11, 13

This is an update on our previous series of articles “Why No One Should Eat Grains” which were published in 2015; we recommend you read them too if you want to get your brain around this topic (before gluten makes toast of it!):


  • Introduction – “deliver us from evil”
  • Gluten related disorders on the rise – but why?
  • Amylase Trypsin Inhibitors – activate myeloid cells
  • Type 1 diabetes – gluten affects the pancreas of even healthy mice
  • Grain globulins – contain coeliac and T1 diabetes reactive proteins
  • Oats – evidence that they should be avoided in coeliac disease
  • IBS and gluten sensitivity – gluten is often the problem
  • New coeliac auto-antibodies identified – linked to autoimmune polyneuropathies
  • Neurological effects of gluten – Simiar in coeliac and NCGS
  • Gluten in Latin America – high levels of self reported gluten avoidance.

Read time: 11 minutes (2200 words)

Continue reading