Does Wheat Cause Coronary Heart Disease?

Introduction

Coronary heart disease (CHD) is the leading cause of deaths worldwide - killing 7 millions people every year. In the following text, we will see that wheat consumption is probably a risk factor for CHD.

Conventional Wisdom on Wheat

Most health organizations currently view wheat as a safe food except for people having celiac disease - affecting up to 1% of the population - and people having non-celiac gluten sensitivity. Also whole wheat - as part of whole-grains - is considered to be one of the healthiest food. In fact a diet rich in whole-grains is considered to be protective against CHD.

Why? Because observational studies consistently find that whole-grain consumption is associated with a decreased risk of CHD. Do these results contradict wheat consumption causing CHD?

Are Whole-Grains Protective Against CHD?

According to this study:

Whole-grain intake consistently has been associated with improved cardiovascular disease outcomes, but also with healthy lifestyles, in large observational studies. Intervention studies that assess the effects of whole-grains on biomarkers for CHD have mixed results.

Indeed many studies show that whole-grain consumption is associated with a decreased risk of CHD. But these studies are observational and can only show correlation but not causation.

In fact there is an health-conscious population bias in these studies: for example people consuming the most whole-grains also exercise more and smoke less:

Data from Majken K Jensen et al., Intakes of whole grains, bran, and germ and the risk of coronary heart disease in men, 2004

Of course researchers adjust the data with these risk factors. But it is very difficult, maybe impossible, to adjust for all risk factors. For example the two previously cited studies did not adjust for important risk factors like socioeconomic status or social support.

A classic example of an occurrence of this bias can be found in hormone replacement therapy (HRT): observational studies had found that HRT was decreasing the risk of heart disease risk while a controlled study finally found that HRT was indeed slightly increasing the risk of heart disease.

A proof that this health-conscious bias could explain the seemingly protective effect of whole-grains can be found in randomized controlled studies: many of them fail to find any beneficial effect of whole-grains compared to refined grains.

So according to these randomized controlled studies whole-grains are neutral toward CHD risks. How then can we say that wheat causes CHD?

Are All Grains Created Equal?

Many randomized controlled studies compared wheat with other grains. These trials are usually quite short. So instead of looking at the number of heart attacks, short-term studies focus on risk predictors of CHD like weight gain or markers of inflammations. Apolipoprotein B (ApoB) level is another risk factor. It represents the number of LDL particles - often called “bad cholesterol”. It is now considered to be a better predictor than LDL-C - the amount of cholesterol contained in LDL particles. The lower the level of ApoB the lower is the risk of CHD.

Here are some results of these studies:

• a study concluded that a bread diet may promote fat synthesis/accumulation compared with a rice diet
• wheat increased BMI compared to flaxseed in a 12 months study
• wheat increased ApoB level by 5.4% compared to flaxseed in a 3 weeks study
• wheat increased ApoB level by 7.5% compared to flaxseed in a 3 months study
• wheat increased ApoB level by 0.05 g/L compared to flaxseed in a 12 months study
• oat decreased ApoB level by 13.7% while wheat had no significant effect in a 21 days study
• wheat increased the number of LDL particles by 14% while oat decreased them by 5% in a 12 weeks study
• ApoA to ApoB ratio (a risk predictor similar in efficiency to ApoB alone - here the higher the better) was increased by 4.7% for oat bran and 3.9% for rice bran compared to wheat bran in a 4 weeks study

These studies show that some grains like oat improve the risk factors of CHD compared to wheat. In addition, these studies often show an absolute improvement of the CHD risk profile in groups eating oat and an absolute deterioration in groups eating wheat. Although we cannot say for sure, it would suggest that oat is protective against CHD - which is confirmed by other studies - while wheat increase the risk of CHD.

That could help explaining why people eating more whole-grains are healthier in observational studies since it looks like that they eat more grains like rice and oat and less typically wheat-made food like white bread, pasta and doughnuts:
Data from Andersson A. et al., Intakes of whole grains, bran, and germ and the risk of coronary heart disease in men, 2007

Now let’s have a look at studies linking wheat and CHD.

Observational Studies on Wheat

Some observational studies linked wheat and waist circumference gains - waist circumference being a strong predictor of CHD:

• a study showed a correlation between consumption of white bread and waist circumference gains
• a study concluded that: ”reducing white bread, but not whole-grain bread consumption, within a Mediterranean-style food pattern setting is associated with lower gains in weight and abdominal fat”
• a Chinese study found that ”vegetable-rich food pattern was associated with higher risk of obesity” but as noted by obesity researcher Stephan Guyenet the association between obesity is in fact stronger with wheat flour than with vegetables

A more pertinent result is found in the data of a large observational study in China. Researchers analysed these data and found a 0.67 correlation between wheat flour intake and CHD. They also found a 0.58 correlation between wheat intake and BMI.

From Denise Minger
But this is just a single unadjusted correlation and does not prove much. However blogger Denise Minger thoroughly analysed the data of this study and found that the association held strongly after multivariate analysis with any other variable available like latitude, BMI, smoking habits, fish consumption, etc.

Since it is an observational study it cannot prove anything but it is yet another evidence suggesting that wheat consumption causes CHD. Let’s now have a look at randomized controlled trials.

Randomized Controlled Trials on Wheat

In addition to the previous randomized controlled trials comparing wheat with other grains there are two additional studies suggesting that wheat consumption causes CHD.

The first one is a study involving rabbits. While studies involving animals are not always relevant to humans - especially studies with herbivore animals like rabbit - the results of this study are quite interesting.

The researchers fed rabbits an atherogenesis diet (i.e. promoting formation of fatty masses in arterial walls) with a supplement of cottonseed oil, hydrogenated cottonseed oil, wheat germ or sucrose. And as they concludes:

Severity of atherosclerosis after 5 months was greatest on the wheat germ-supplemented diet, whereas there were no differences among the other three groups.

The second study is the Diet And Reinfarction Trial (DART). In this 2-year randomized controlled trial, people who already had recovered from an heart attack were split into groups receiving various advices. The main result of this study was that the group advised to eat fatty fish had a reduction in mortality from CHD.

One other advice - the fibre advice - was:

to eat at least six slices of wholemeal bread per day, or an equivalent amount of cereal fibre from a mixture of wholemeal bread, high-fibre breakfast cereals and wheat bran

Seeing this advice we can guess that most of cereal fibres intake by this group was from wheat although we cannot be sure.

This advice resulted on a 22% death increase:
From Stephan Guyenet
However this result bordered on statistical significance: the 95% confidence interval being 0.99–1.65.
For people not familiar with statistics, a result is usually defined as statistically significant when there is less than 5% chance that the result is due to luck alone. Here there is a 95% probability that the relative risk is between 0.99 (1% decreased chance of dying) and 1.67 (67% increased chance of dying).

Since the probability that the fibre advice resulted in a protective or neutral effect was a little too high, this result has been quite overlooked. Had the study last a little longer, it would have raised way more suspicion toward whole-grains.

In fact, researchers found this effect to be statistically significant in a follow-up study. After adjusting for pre-existing conditions and medication use, we can see in the table 4 of this study an hazard ratio of 1.35 (95% CI 1.02, 1.80) for the 2-year period of the randomized controlled trial.

These results are quite telling: according to these researchers, a 2 year randomized controlled trial showed that advising people recovering from an heart attack to eat at least six slices of wholemeal bread per day resulted in a statistically significant 35% percent chance increase of CHD compared to people not receiving this advice.

Wheat, Vitamin D Deficiency And Heart Disease

Many studies found that vitamin D deficiency is associated with CHD.
However vitamin D deficiency does not seem to cause heart disease. For example several studies found that vitamin D supplementation did not prevent heart disease.
As this study concludes:

A lower vitamin D status was possibly associated with higher risk of cardiovascular disease. As a whole, trials showed no statistically significant effect of vitamin D supplementation on cardiometabolic outcomes.

Wheat consumption causing CHD could help explaining these results. A study found that wheat consumption depletes vitamin D reserves. That could explain why vitamin D deficiency is associated with heart disease and why it does not seem to cause it: both vitamin D deficiency and heart disease could be consequences of wheat consumption.

Of course this is not the only explanation. For example the DART study shows that fish consumption prevents CHD and fish is a food rich in vitamin D.

Not the Perfect Culprit

To be clear, if it seems likely that wheat consumption is a risk factor of CHD it is not the only one nor the primary one. There are many other factors like smoking, hypertension, lack of exercice or stress. Even among dietary factors wheat is probably not the main one. For example the DART study shows that the protective effect of fish intake is stronger than the adverse effect of wheat.

In addition, deleterious wheat effects might not affect everybody. One study showed that the ApoB level variation following wheat and oat bran intake was different depending on the genotype of the individuals. In another study whole-wheat intake worsened the lipid profile only in people having a specific genotype compared to refined wheat.

How the wheat is cooked may have a role too. Studies show that sourdough bread improve mineral bioavailability (such as magnesium, iron, and zinc) compared to yeast bread or uncooked whole-wheat. Also content in proteins with potential adverse consequences like gluten or wheat germ agglutinin differs depending of the food type.

Conclusion

There are strong evidences that wheat consumption is a risk factor for CHD. People at risk of CHD should avoid wheat as should those trying to lose weight. In all cases, stopping wheat consumption for a month for example to see how one feel without wheat is always a good idea since there is currently no available method to diagnose non-celiac wheat sensitivities and that even for celiac disease the average delay in diagnostic is 11 years in the US.

More studies looking at the links between wheat and CHD are urgently needed since CHD is the leading cause of deaths while wheat is the second most widely consumed food and whole-wheat is often advised to lower risk of CHD. Studies considering grains as a whole are bound to give inconsistent results since different grains seem to have opposite effects in the case of CHD. So as much as possible future studies should treat grains separately and consider things like type of wheat products and genetic variability.

===============================================================
Read the complete article here or here.

Niacin’s Effect on Lp(a) in AIM HIGH - Dayspring

Commentary on Niacin’s Effect on Lp(a) in AIM HIGH
    

By: Thomas Dayspring, MD, FACP, FNLA, NCMP

 

In 2013 we have already published two commentaries on niacin (Commentary on Niacin vs Ezetimibe as add on to Statin) and (Examination of the Recently Announced Preliminary Results of the HPS2-THRIVE Study), specifically extended release available as Niaspan, a seemingly potent lipid- and lipoprotein-modulating drug that dates back to the 1960’s. Initially it was used to reduce elevated cholesterol levels but eventually it was found to also raise HDL-C which for a variety of reasons was assumed to be very desirable (thought being that if low HDL-C is a strong CV risk factor, then raising it must be beneficial).  Also of interest was niacin’s ability to significantly reduce lipoprotein (a) mass [Lp(a)]. Indeed, a group entitled European Atherosclerosis Society Consensus Panel issued a statement strongly advising niacin be used for CV benefits in patients with elevated Lp(a) [1]. Interestingly that panel noted there was virtually no clinical trial support for this recommendation other than the fact that niacin does indeed reduce Lp(a) mass. Most lipidologists agreed with the belief that even if reducing Lp(a) does not matter, niacin would at least reduce apolipoprotein B (apoB) which is seemingly always desirable. NCEP ATP-III simply advocated achieving LDL-C goals in persons with risk related to Lp(a) issues. 

 

 

 

 

Recent trials [The Atherothrombosis Intervention in Metabolic Syndrome with Low HDL/High Triglycerides: Impact on Global Health Outcomes (AIM HIGH) and large Heart Protection Study 2: Treatment of HDL to Reduce the Incidence of Vascular Events  (HPS THRIVE 2)] have not shown additional event reduction in well-treated patients with stable CHD related to adding niacin to a statin or statin/ezetimibe regimen [2,3]. To say the results of those studies were a shock to the lipidology community is an understatement. AIM HIGH (all of the patients had low HDL-C at baseline) was published first and for those who believed niacin’s benefit was related to raising HDL-C, the results were a punch to the jaw. Despite a substantial (25%) HDL-C increase (remember the old well accepted but never proven caveat that for every 1% rise in HDL-C there is a 3% event reduction) there was no CV outcome improvement. The usual side effects associated with niacin were present including a questionable nonsignificant rise in ischemic stroke. Then along came the still not published HPS THRIVE 2 (baseline HDL-C was not an enrollment criteria) where again the addition of niacin to a statin or statin/ezetimibe regimen provided no additional outcome benefit. Common to both AIM HIGH and HPS THRIVE 2 was the fact that the lifestyle with statin or statin/ezetimibe had normalized LDL-C, non-HDL-C and apoB. Thus niacin was being added to patients who were at those goals (keeping in mind that there is no NCEP ATP-III goal for HDL-C). Should we really have expected niacin, whose primary mechanism of action is to lower apoB (or its lipid surrogates) to do anything to CV events in persons with normal apoB?  The answer is yes if raising HDL-C or lowering Lp(a) mass is critical to event reduction (well accepted concepts that have never ever been proven in any type of trial). Well we may have those answers now and at this point one has to reasonably conclude the evidence is strong that in patients on LDL-receptor inducing drugs (statins or statin + ezetimibe) raising HDL-C (note – niacin also raises apoA-I, but not apoA-II or total HDL-P) [4] or reducing Lp(a) mass with niacin provides no benefit in folks who are at apoB (LDL-C, non-HDL-C) goal.

 

In AIM HIGH baseline apoB and apoA-I levels were low and baseline Lp(a) was elevated at 33.8 nmol/L [using Caucasian adult data from Framingham as a comparator, Lp(a) averaged 20 nmol/L]. Nearly 30% of AIM HIGH patients had severe Lp(a) elevations > 100 nmol/L compared to 20% of Framingham cohort. The addition of niacin to statin or statin + ezetimibe raised HDL-C by 25%, apoA-I by 7% and reduced LDL-C by 12%, TG by 30% and apoB by 13%. [5]

 

Lp(a) as expected was significantly associated with CV events despite the fact that LDL-C was at goal and thus elevated Lp(a) is associated with residual risk. A one standard deviation of Lp(a) was associated with a 21% increase in CV risk. There was a 21% overall reduction [but with a 20%, 39% and incredible 64% decreases in patients at the 50^th, 75^th and 90^th percentile cut points] in Lp(a) in the niacin group compared to 6% in placebo group. So the higher the Lp(a) level, the more dramatic was niacin’s ability to lower it.  Here is the shocker: there was no difference in event rate between those on or not on niacin (remember all were statin or statin + ezetimibe) DESPITE GREATER DECREASES in Lp(a) for those using niacin. Even in those in the highest Lp(a) quartile (> 125 nmol/L) there was no reduction in events when niacin was added.

 

So where do we stand with niacin? There is no level one evidence anywhere supporting the use of niacin to reduce clinical events: The Coronary Drug Project (CDP) is often quoted as proof of niacin’s efficacy but few realize that niacin monotherapy (high dose of immediate release preparation) had no impact on the primary endpoint of the study (mortality): thus the benefit of reducing non-fatal myocardial infarction (a secondary endpoint) makes this benefit hypothesis generating [6].Of course there is the famous 15 year follow up of CDP which encompassed 6 years of the trial where niacin was used and then a subsequent 9 year period off niacin. Mortality was significantly reduced in that post hoc analysis (data derived not from examination or in person review but questionnaires sent to participants): this is the weakest data imaginable [7]. So this supposedly late benefit of niacin is in fact analysis of post hoc follow data up from a trial where niacin failed to reduce the primary endpoint. If niacin was a new drug, it would have no prayer of gaining FDA approval based on the CDP. Several subsequent trials using angiographic or CIMT endpoints showed niacin monotherapy or combination with bile acid sequestrants or statins showed imaging benefit. One small (~500 patients) open-label outcome trial (Stockholm Ischemic Heart Disease Secondary Prevention Trial), combining clofibrate and IR niacin did reduce clinical events with statistical significance [8].

 

In my opinion niacin became a major lipid drug because of its ability to raise HDL-C and to lower Lp(a) and not for what is likely its real mechanism of action, namely lowering LDL-C and apoB and LDL-P. After extended-release niacin (Niaspan) hit the market, it was also heavily promoted because of its ability to increase both HDL and LDL size. KOS made a fortune by promulgating those messages as it seemingly made so much sense. Of course over time, we have learned that influencing LDL or HDL particle size or raising HDL-C and apoA-I has no effect on outcomes. Looking at lipid/lipoprotein risk factors in 2013 the outcome evidence only supports lowering apoB (LDL-P) or perhaps raising total HDL-P. At this time unfortunately, there is no support for reducing Lp(a) with niacin: admittedly the Lp(a) data from the much larger HPS THRIVE 2 study of 25,000 patients is pending.  

 

So is niacin a dead drug? Here are my thoughts as a clinical lipidologist: We must get apoB (LDL-P) to goal in all at-risk patients. Lifestyle therapies and statins are the mainstay of therapy. However residual risk is high if apoB (LDL) remains elevated despite at-goal LDL-C, non-HDL-C), any level of HDL-C or if Lp(a) mass is elevated. So I would have no hesitancy in adding niacin to high and very high risk patients who have not achieved apoB (LDL-P) goals with whatever therapies they are using or using niacin as a monotherapy in those intolerant of other apoB lowering meds. Data from HPS THRIVE 2 (discussed in a recent commentary [ADD LINK]) suggested statin plus ezetimibe was better at event reduction than statin plus niacin [9]. In view of that and the very significant side effects reported in HPS THRIVE 2 [bleeding (GI, intracranial, other) in the niacin group: 326 (2.5%) to 238 (1.9%) and infection 1031(8%) to 853 (6.6%)] [3] makes niacin a tertiary or quaternary add-on drug (some may prefer the bile acid sequestrant colesevelam as an apoB lowering medication).

 

What about our patients with elevated Lp(a) mass or Lp(a)-P? The only therapy that has so far shown an inkling of success is LDL apheresis. For now we should all try to lower apoB (LDL-P as aggressively as possible and that often requires multiple combination therapies. What about future drugs: just published is the  data the PCSK9 monoclonal antibody AMG 145 reduces Lp(a) by 32% in patients on statins [10]. There is also promising data that the remaining CETP inhibitors also reduce Lp(a) but the reality is that until such reductions by these drugs are linked to outcome benefit they are of hypothetical interest. Hopefully in the future there will also be development of an apoprotein (a) antisense oligonucleotide inhibitor. 

---

 
References:

[1] European Atherosclerosis Society Consensus Panel. Lipoprotein(a) as a cardiovascular risk factor: current status. European Heart Journal 2010;31:2844–2853.

[2] The AIM-HIGH Investigators Niacin in Patients with Low HDL Cholesterol Levels Receiving Intensive Statin Therapy. N Engl J Med 2011;365:2255-67.
[3] Presentation by Jane Armitage on behalf of the HPS2 THRIVE group to the National Lipid Association Annual Scientific sessions, Las Vegas NV June 2013.

[4] Otvos, JD.The surprising AIM-HIGH results are not surprising when viewed through a particle lens. Journal of Clinical Lipidology (2011) 5, 368–370.

[5] John J. Albers, et al. Relationship of Apolipoproteins A-1 and B, and Lipoprotein (a) to Cardiovascular Outcomes in the AIM-HIGH Trial in press JACC 10.1016/j.jacc.2013.06.051

[6] Coronary Drug project group. Clofibrate and Niacin in Coronary Heart Disease. JAMA 1975;231:360-381.
[7] Fifteen Year Mortality in Coronary Drug Project Patients: Long Term Benefit with Niacin. JACC 1986;8:1245-55.
[8] Reduction of Mortality in the Stockholm Ischaemic Heart Disease Secondary prevention Study by Combined Treatment with Clofibrate and Nicotinic Acid. Acta Med Scand 1988;223:405-418.[9] Masana, A. Cabré, N. Plana. HPS2-THRIVE results: Bad for niacin/laropiprant, good for ezetimibe? Atherosclerosis 2013;229:449-450.

[9] Masana, A. Cabre, N. Plana. HPS2-THRIVE results: Bad for Niacin/laropiprant, good for ezetimibe? Atherosclerosis 2013; 229:449-450.

[10] Nihar R. Desai, et al. AMG145, a Monoclonal Antibody Against Proprotein Convertase Subtilisin Kexin Type 9, Significantly Reduces Lipoprotein(a) in Hypercholesterolemic Patients Receiving Statin Therapy. (Circulation. 2013;128:962-969.)

=================================================================================================

Read the complete article here.