My interest in heart disease was first piqued, many moons ago, by the knowledge that the French had a very low rate of heart disease, despite having the entire raft of conventional risk factors. Over the years, I have found far more outstanding paradoxes than the French, but at that time they stuck out like a sore thumb, a mute reproach to the conventional theories about heart disease. A reproach that people have continually tried, and failed, to explain.
To be frank, I found the mainstream excuses about the protective effects of eating garlic, drinking red wine and lightly cooking their vegetables to be utter bunk. It was obvious that these factors had only emerged to keep the cholesterol hypothesis alive and sweep the French paradox under the carpet.
But what else, I thought, could explain the low rate of heart disease in France? As I’m from Scotland – where the heart disease rate at the time I became interested was the highest in the world – my mind turned to the way that the French eat, and the importance of food and eating, in their culture. In Scotland, eating is seen as somewhat akin to filling up your car with petrol. A waste of ten minutes, but it is something that has to be done before going out and getting ‘pished’ on a Saturday night.
As for cooking, my memory of a traditional Scots recipe is, as follows:
| Step one: | Place a three-pound lump of beef in a saucepan with a carrot and an onion and boil for eight hours. |
| Step two: | Eat with boiled potatoes. |
And as everyone knows, the Scots love a fry-up. Even a fried-up Mars bar:
| Step one: | Take a frozen Mars bar and cover in batter. Place in deep-fat fryer for two minutes. |
| Step two: | Eat with chips while walking home in the rain. |
When I was growing up in Scotland, there used to be a substance called zinc ointment – maybe there still is. If you have never heard of it, thank the Lord. It was used as a cream substitute in things like chocolate éclairs. The Scots felt that cream tasted far too delicious to besmirch their puritanical souls. Therefore, it should be replaced with an off-white substance of little taste, although such taste as it had was distinctly bitter and unpleasant. I suspect it was constructed entirely from E-numbers in a petrochemical plant.
A few years back, I took my wife to watch Dunfermline Athletic play football. During the match, she made the extremely rash decision to buy a mutton pie. I did warn her, but she wouldnae listen, she just wouldnae. A mutton pie washed down with Bovril, no less. However, it is the mutton pie itself that is a true specialité de la Scottish cuisine. A pastry coating that the British Army has since discovered can prevent uranium-depleted shells from piercing tank armour. And if you do manage to get through this defensive barrier without breaking your teeth, you will discover a small piece of gristle, surrounded by half a pint of grey, liquid fat. The fat usually spurts out, covering all clothing within a ten-foot radius, and it cannot be removed by any form of washing powder yet created.
As a general observation, therefore, it can be said that food and eating, are not given quite the same status in Scotland as in France. For the French, food is a central part of life. Mealtimes are a major social occasion. People spend a long time buying food, preparing meals and then eating. In Scotland they don’t, or at least they certainly didn’t. This, to me, marked a very obvious difference between the two countries.
Could this attitude to food and eating somehow be the reason for the difference in heart-disease rates between the two countries? And if so, how? Was it something to do with being relaxed while trying to digest food, rather than shovelling it down as fast as possible?
With this thought in mind, I began what turned into a 25-year journey of discovery. It has to be said that I have taken many wrong turns along the way. There were several years when I thought that heart disease did not actually have a cause, or causes at all, so often did I find myself in another blind alleyway. Eventually, everything did come together in a way that makes sense and is actually supported by the facts. The primary cause of heart disease, I finally discovered, is… stress.
Well, hey, like haven’t about ten million people been saying this for the last fifty years? Indeed they have. However, there is a major problem here, which is that the word ‘stress’ doesn’t really mean anything at all. Or perhaps it means too many things. Or perhaps it just means different things to different people.
So how can anyone say that stress causes heart disease, when there is little agreement as to what stress actually means? It’s a good question. In order to answer it, I have to attempt to define rather more clearly what I mean by stress.
WHAT IS STRESS?
I suppose many people think of stress as a form of time pressure. Busy, busy, busy, so much to do, so little time to do it. Others think of stress as a type of constant grinding worry, like money problems, or having an oppressive boss at work. Stress can also be seen as a transient state – for instance, moving house, or getting married, or even getting up on a Monday morning.
Many people believe that stress is good for us, and without it we would just lounge about doing nothing at all – certainly true in my case. This lack of any clear definition, or even agreement about fundamental principles, such as whether stress per se is healthy, or unhealthy, does make it tricky to measure it in any repeatable way. And without a measurement the medical profession tends to lose interest rapidly.
With a cholesterol test, you know exactly what you’re getting. It’s 5.9, or 6.2, or 3.8, or 2.79. Once you have your figure, you can then give drugs and watch the figure change in front of your very eyes. Then you can draw graphs, do an audit, write papers… and all sorts of things. Proper science, no less. The sort that gets published in proper journals and leads to proper promotions.
But with things like stress, no such measurements exist. We are in the world of the subjective experience, where we have to rely on personal testimony and suchlike. It is not a place where many medical researchers like to venture.
An Eastern Tale
(As told to me)
A stranger was passing through a village one day when he spotted a wise man scrabbling about in the dust. ‘What are you doing?’ asked the stranger. ‘I am looking for my key,’ replied the wise man. ‘I shall help,’ the stranger exclaimed, and immediately set about the search. He looked under leaves, he sifted in the dust, he looked everywhere. After about two hours there was still no sign of the key. ‘Are you sure you lost it out here?’ the by now very dusty and thirsty stranger asked. ‘Oh, no,’ the wise man replied. ‘I lost it in my house.’ Understandably, the stranger was somewhat irritated. ‘Then why are we looking out here?’ he demanded. The wise man smiled. ‘Because,’ he said ‘out here the light is so much better for looking.’
I have to admit that I too, enjoy looking where the light is so much better. But sometimes you have to bite the bullet and accept that this doesn’t actually happen to be where the answers lie. And with stress, and heart disease, you must search in a world where some of the answers cannot be directly seen. You can only know they are there by the effects that they have on things around them.
Just to give one example, from a study published in the BMJ in 2001 entitled ‘The Hound of the Baskervilles effect: natural experiment on the influence of psychological stress on timing of death’:
In Mandarin, Cantonese, and Japanese, the words ‘death’ and ‘four’ are pronounced nearly identically, and consequently the number 4 evokes discomfort and apprehension in many Chinese and Japanese people. Because of this, the number 4 is avoided in floor and room numbers in some Chinese and Japanese hospitals, and in some Chinese and Japanese restaurant telephone numbers. In addition, the mainland Chinese airforce avoids the number 4, but uses other numbers, to designate its military aircraft, apparently because of the superstitious association between ‘four’ and ‘death’.
The study by Phillips and his co-authors finds that cardiac deaths peak on the fourth of the month for Americans of Chinese and Japanese descent, and that this pattern is not seen among whites. The study used computerized US death certificates to examine more than 200,000 Chinese and Japanese deaths, and 47,000,000 white deaths, from 1973 to 1998.
On one hand we have a real and scientifically measurable effect, which is that Chinese and Japanese die more often on the fourth day of the month – and you can’t argue with that, it’s a fact. On the other hand, we have something much more difficult to deal with, which is that the reason for this increased risk of death appears to be that the Chinese and Japanese consider the number four to be unlucky.
But has anyone worked out how to measure the physiological effect of ‘unluckiness’? Can we invent a drug to protect against the damage cause by the number four? I tend to doubt it. Yet the deadly effect of ‘four’ exists, nonetheless.
So what do we do? Simply ignore this finding, because it is considered virtually impossible to analyse in the reductionist way so beloved of medical science? Or do we bite the bullet and accept that, for some people, the number four can cause deadly stress? Even if we must also accept that it is very tricky to get a handle on what it is about this number that creates stress? I suggest the latter.
However, the mainstream has tended to the former approach, i.e. ignoring. Just to give one example of this tendency, a huge study was carried out in 52 countries looking at 29,000 people in order to establish the similarities, or differences, between risk factors across a wide range of different populations. This was the INTERHEART study.10
As part of INTERHEART, they measured psychosocial stress – hallelujah! Annika Rosengren, Professor of Cardiology at Goteborg University, Sweden – who led the stress aspect of the research – noted that people’s psychosocial wellbeing, judged by simple measures, was significant:
‘Collectively these [measures] were responsible for about one third of the risk of the population studies,’ she said. ‘Persistent severe stress makes it two and a half times more likely that an individual will have a heart attack compared with someone who is not stressed.’ She said stress and depression together increased the risk threefold.
‘The public thinks stress is very important in their heart attack. My patients often say they think it was due to stress, but previous studies have shown contrary effects of stress. But the INTERHEART study shows definitively that stress is one of the most important factors in heart attack in all ethnic groups and in all countries.’ http://www.telegraph.co.uk/news/main.jhtml?xml=/news/2004/09/02/wstres02.xml&sSheet=/portal/2004/09/02/ixportal.html
Most interesting, and surely something worth pursuing further? However, the mainstream response to this was best encapsulated, in the same article, by Professor Sir Charles George, Medical Director of the British Heart Foundation. He did say that the results ‘suggested’ that stress might have more of a role as a cause of heart attacks than many people had previously thought. (Don’t you just love the use of that word ‘suggested’?) However, he went on to caution that the findings were the result of ‘self-reported’ stress that had not been confirmed by chemical measures – of hormones in saliva, for example. In short, you didn’t really measure it in the approved scientific manner, therefore it doesn’t really exist. In such a casually dismissive fashion is the evidence about perceived stress swept under the carpet, time and time again.
Of course, stress is not a simple concept, and measuring it is even more difficult. However, if you are willing to accept proof, and facts, that are less rigid than p < 0.005 (CI 0.63–84), then good. After all, as Albert Einstein was wont to say:
Not everything that can be measured, matters, and not everything that matters can be measured.
[One of several different versions of this saying attributed to him.]
This does not mean that I am simply going to claim that stress is the main cause of heart disease and leave it at that. Indeed, I intend to use a great deal of evidence to make the case for – or should that be against – stress. Just because you can’t accurately measure everything does not mean that you should give up, or try to use good scientific research wherever possible.
At this point, therefore, having thrown enough caveats into the air to sink a battleship, I will take the plunge and attempt to show you exactly how stress causes heart disease. Before I can do this, though, I have to dismantle stress into its component parts.
Firstly, we’ll need to separate out the things that cause stress – the ‘stressors’ – from the ‘stress response’, i.e. the physiological effects that stressors create. Of course, not all stressors will create a stress response. For example, the number four will have no effect on most people in the West-whereas the number thirteen might.
After separating cause and effect, a further split is necessary because there are two basic stress responses: healthy and unhealthy. As a further subdivision, I need to make the distinction between the two types of stressor: physical and psychological.
To try and make this a bit clearer, I have created a list of the type of stressors I am talking about, and the likely effects that they have.
1: Examples of physical stressors that create a healthy response:
• Exercise
• Competitive sport
• Massage
• Sauna
• Moderate alcohol consumption
• Singing
• Bungee jumping
• Rock climbing
• Roller-coaster rides
2: Examples of psychological stressors that create a healthy stress response:
• Your football team winning
• Passing an exam
• Clinching a successful business deal
• Organising an enjoyable social evening
• A tight sales deadline – but not too tight
• Giving a well-received lecture
• A busy shift in Accident and Emergency with no one dying
• Being Prime Minister
3: Examples of physical stressors that create an unhealthy stress response:
• Excessive, intense, forced exercise in adverse conditions, e.g. working deep below the ground in a coal mine in Russia
• Extreme environmental change/rapid alteration in temperature
• Being a fighter-jet pilot
• Rheumatoid arthritis
• Cocaine use
• Smoking
• Eating under pressure
• Major trauma/surgery
• Spinal cord injury
• Steroid use
• Disease of the hormonal system
– Cushing’s disease (too much cortisol)
– Phaechromocytoma (too much adrenaline)
– Diabetes (too much blood sugar)
– Acromegaly (too much growth hormone)
4: Examples of psychological stressors that create an unhealthy stress response:
• Bullying boss
• Suffering racism
• Being ‘dislocated’ from the surrounding population/culture
• Money worries, long-term debt
• Low status in social hierarchy
• Poor social network
• Non-supportive, unloving or abusive spouse
• Football team losing
• Getting caught in an earthquake (though this is a physical stressor too)
• Getting up on Monday morning
• Forced emigration/social dislocation
• The number four
These are not full lists, by any manner of means – and not all of the things on these lists will create the same response in all people. But I hope that it gives you a clearer idea of the types of ‘stressors’ that I am talking about. My next trick is to explain exactly how an unhealthy stress response (whatever causes it) goes on to cause heart disease.
HOW AN UNHEALTHY STRESS RESPONSE CAUSES HEART DISEASE
In order to explain how this happens, I need to introduce you to the neurohormonal system. This hugely complex system consists of two basic parts: the ‘hormonal’ part and the ‘nervous system’ part. While I have provisionally called this the ‘stress system’, the term is actually horribly inaccurate. Because the system involved in stress is precisely the same system that is involved in relaxation – only in reverse.
In fact, for every hormone in the neurohormonal system that fires you up, there is another one that calms you down; and for every set of nerve fibres that revs you up, there is another network that relaxes you. Eastern philosophy would call this whole shebang Yin and Yang, internal balance, which is a pretty good way of looking at it. Because if the neurohormonal system gets seriously out of balance, you are likely to suffer catastrophic metabolic problems, then heart disease… then cancer, then diabetes, then… well, too much for me to cover in one book.
The hormones involved on the ‘stress’ side include adrenaline, cortisol, growth hormone and glucagon. On the ‘relaxation’ side, for the purposes of this discussion, I shall concentrate on insulin.
Release of stress hormones is controlled by the hypothalamus and pituitary gland acting in unison. Under a stressful situation – for example, a man pointing a gun at you – the hypothalamus sends alarm messages to the pituitary gland, which then fires off hormonal messengers to the adrenal glands to get them to release adrenaline and cortisol, among other things.
This ‘three-part’ hormonal system, consisting of the hypothalamus, pituitary gland and adrenal glands, is often referred to as the Hypothalamic-Pituitary-Adrenal axis, or the HPA-axis for short. The HPA-axis is intimately connected to, and intertwined with, the unconscious or ‘autonomic’ nervous system. The autonomic nervous system has two basic divisions: the sympathetic and the parasympathetic systems. Neither of these divisions is under your conscious control – unless you are a Zen master, or something of the sort.
Fig. 29 The parasympathetic and sympathetic nervous systems
The sympathetic nervous system has a wide range of actions. These include speeding up your heart rate, reducing saliva production and redirecting blood supply to your muscles. It also stimulates the liver to release glucose, thus pushing up blood-sugar levels, and triggers the release of various blood-clotting factors. These are the sort of things you need when physical danger threatens, which is why this whole process is sometimes called the ‘fight or flight’ response.
On the other hand the parasympathic nervous system has directly opposing actions. It slows your heart, stimulates insulin production and the release of bile. It also increases the flow of saliva, and directs blood to the guts to aid digestion.
Another way to look at this is to say that an activated sympathetic nervous system – working in conjunction with raised ‘stress’ hormones – represents the ‘catabolic’ state, a state in which your body is ready to burn up its energy stores, which comes in handy in a fight, or during exercise. You have probably experienced this state after a physical activity such as tennis or squash, when you know you should be hungry but find that when you sit down to eat you have no appetite. The ‘stress’ hormones are still ruling your metabolism, and are telling you that you are not yet ready to eat.
On the other hand, an activated parasympathetic nervous system, working in conjunction with a raised insulin level, represents the ‘anabolic’ state – a state in which you are ready to eat, digest and store energy – and then have a siesta.
In fact, analysing these two metabolic ‘states’ is where I first began in my quest to understand heart disease. Within our bodies, I knew, we have these two systems that are, essentially, directly antagonistic to each other. Anabolism and catabolism. I reasoned that if you were stressed, and then tried to eat, your metabolism would be thrown into confusion. You would be commanding the neurohormonal system to activate catabolism and anabolism simultaneously. This would mean high levels of adrenaline and cortisol, battling against high levels of insulin. Adipose tissue would be under instructions to both absorb and pump out fats into the bloodstream. At the same time, the liver would be trying to store, and release, glucose.
With food inside them, your guts would be automatically switched to ‘absorption’. But the sympathetic system would be fighting to direct blood away from the guts to the muscles. Wherever you looked, a fight for metabolic supremacy would be going on. Perhaps the most important battle would be for control of blood-sugar levels, a battle ending up with ‘spikes’ of blood sugar – as insulin tried, and most likely failed, to overcome the effects of the stress hormones surging about in the bloodstream.
In short, I thought that eating under stress was likely to be pretty damned unhealthy. Equally, taking time over meals, and relaxing while doing so, was likely to be pretty damned healthy. Could this be the reason for the high rate of heart disease in Scotland, and the low rate of heart disease in France? Possibly, probably… it almost certainly represents part of the answer.
More on that later. Now it is time to look at what happens when the ‘stress system’ breaks down. Actually, from now on, I am going to refer to a breakdown of the stress system as a ‘dysfunctional HPA-axis’. Sorry about using this jargon, but it is much more accurate and useful. It also moves the discussion away from the slightly woolly concept of stress, to something that can be measured, i.e. HPA-axis function. (Normally this is done by measuring cortisol levels.)
Causes of a dysfunctional HPA-axis
Probably the most dramatic dysfunction of the HPA-axis occurs when a tumour develops in the pituitary gland, which then proceeds to pump out far too much in the way of stress hormones. Several types of these tumours can develop. A tumour producing too much growth hormone can lead to gigantism and acromegaly; a tumour producing too much adrenaline can cause a condition known as phaeochromocytoma, etc.
However, I am only going to focus on one type: a tumour in the pituitary gland that pumps out too much ACTH (corticotropin). ACTH is a ‘precursor’ hormone which, in turn, stimulates cortisol secretion from the adrenal glands. So, a tumour in the pituitary gland, (secreting too much ACTH) effectively increases blood cortisol levels. This condition is known as Cushing’s disease.
Cushing’s disease, in turn, has a wide range of different effects – which are a direct result of the many actions that cortisol has around the body. For example, cortisol:
• Triggers the liver to release its stores of glucose.
• Stimulates the breakdown of triglyceride stores in adipose tissue, leading to an increase in free fatty acids (FFAs) in the blood.
• Triglyceride breakdown also releases glycerol, which travels directly to the liver, where it is converted to glucose.
• Activates breakdown of muscle protein into amino acids.
• (The amino acids then travel to the liver, where they are converted into glucose.)
• Acts as a direct antagonist to the actions of insulin at most sites in the body.
As you might expect, therefore, people with too much cortisol surging about in the system have high blood-sugar levels and a high degree of what is known as ‘insulin resistance’. In fact, most people with Cushing’s disease develop diabetes.
Another thing that happens to people with Cushing’s disease is that they lose muscle bulk – due to the breakdown of muscle proteins. There is also a redistribution of fat from the periphery (arms and legs) to the trunk, or abdomen. Sometimes this redistribution can be so extreme that it leads to a condition known as ‘buffalo’ hump.
The reason why this happens is because you have two very different types of fat in your body: subcutaneous and visceral. Subcutaneous fat sits just underneath the skin and is found all over the place: arms, legs, neck, even fingers. Sumo wrestlers have lots of this type of fat, and they work hard to build it up. How they do this is a fascinating topic. (Well, at least I find it fascinating, but this is not time to get sidetracked.)
Visceral fat, on the other hand, is mainly found around the organs in your abdomen. It is the type of fat that builds up in those who develop the classic ‘beer belly’. While both types of adipose tissue can each store, and then release, fat, that is the beginning and end of any similarity. From a metabolic perspective, they are as different as different can be. They are to all intents and purposes different organs. One is fat, the other is ‘anti-fat’.
I shall tiptoe around this area because it is both enormous, and enormously complex, and I do not want to get bogged down. Suffice to say, for the sake of this discussion, that cortisol stimulates subcutaneous adipose tissue to release fat, thus making it shrink in size. On the other hand, cortisol stimulates visceral fat to do the exact opposite, i.e. absorb and store fats, leading to an increase in visceral fat mass (This is a horrible oversimplification, but for the sake of this argument it will do.)
There is another reason for bringing these two types of fat into the discussion at this point, which is that a build-up of visceral fat is now recognised as a major risk factor for heart disease. In fact, many people now believe that visceral fat is the primary underlying abnormality in heart disease, as it is thought to create a wide spectrum of metabolic abnormalities that are closely linked to heart disease. These abnormalities have been brought under the umbrella term ‘Syndrome X.’ Also known, among other things, as:
• Metabolic syndrome X
• Reaven’s syndrome
• Metabolic syndrome
• Insulin resistance syndrome
Whatever you choose to call it (and please will someone make up their minds!), to my mind the current thinking is bonkers. Visceral fat doesn’t build up all by itself, just for the hell of it, before going on to create Syndrome X. Something has to cause the build-up of visceral fat in the first place. To argue otherwise is to end up in the mad genetics/magic argument again: ‘Visceral fact accumulation just, sort of, happens. We don’t know why, so it must be due to genetic susceptibility.’ (Listen, guys, it doesn’t just happen. It is caused by HPA-axis dysfunction and abnormal cortisol levels. Hellooo! have a look at Cushing’s disease!)
Anyway, in addition to its effects on raising glucose and insulin levels, and its impact on muscle and fat distribution, a high cortisol level also causes the following abnormalities:
• Raised VLDL level
• Low HDL level
• Raised LDL level
• Raised blood pressure
• Raised fibrinogen levels (clotting factor)
• Raised PAI-1 level (clotting factor)
• Raised Von Willibrand level (clotting factor)
• Raised Lp(a) level (clotting factor)
Does anything seem familiar about this list? If not, it will.
To round off this topic, I should probably mention that people with Cushing’s disease have accelerated atherosclerotic plaque growth, and a gigantically increased risk of heart disease.
Strongly reinforcing the fact that it is the raised cortisol level itself that is causing the damage – rather than some other factor – is the evidence from people who take steroids. Steroids, as mentioned before, are among the most widely prescribed of all medications. They are also called ‘corticosteroids’, because the basic building block of all steroids is cortisol. What this means is that when you take a steroid you are, effectively, giving yourself Cushing’s disease.
Why would anyone want to do this? Well, one effect of cortisol that I haven’t mentioned so far is that it greatly inhibits the immune system. I haven’t the faintest idea why cortisol does this. However, because it does, it is used to treat diseases when you want to shut down an overactive immune response. Such ‘autoimmune’ diseases include rheumatoid arthritis, asthma, eczema and ulcerative colitis. Steroids are also used after a transplant, as they prevent the body from rejecting the organ.
In situations like this, steroids are powerful and life-saving drugs. However, if you keep taking them for too long you will end up with the exact same set of abnormalities found in Cushing’s disease: high blood-sugar and insulin levels, low HDL, high VLDL/ LDL, a whole range of blood-clotting factor abnormalities, and increased visceral fat deposition. In short, the works.
What’s more, people who take steroids long term are at a greatly increased risk of dying of heart disease. Even fit, young, healthy people. And it can happen very fast. To give one example of the abuse of anabolic steroids (a form of cortisol/corticosteroid that has been altered to create muscle build-up, rather than break it down):
Anabolic steroids are frequently abused, thus increasing the risk of cardiovascular disease. We report on a young bodybuilder who presented with ventricular tachycardia as the first manifestation of severe underlying coronary heart disease. Coronary angiogram revealed severe stenotic lesions [narrowings] in the right coronary artery and the left descending coronary artery, and hypokinetic [hibernating] regions corresponding to posterolateral [the back and side] and anterior myocardial infarctions. This young patient had a history without any coronary risk factors, but with a 2-year abuse of the anabolic steroid stanazolol.
Mewis C Clin Cardiol, February 1996
Here is a young man with no classic risk factors for heart disease. Within two years of abusing steroids, however, he had developed severe occlusion in two major arteries in the heart, and he had also suffered two separate heart attacks (which he didn’t actually know had happened). This looks like a fairly clear case of cause and effect to me.
Anyway, we have two different ‘conditions’ where cortisol levels are significantly raised: Cushing’s disease and steroid use (or, rather, abuse). In both of them, exactly the same set of abnormalities develop, followed by heart disease. Clearly though, these two ‘conditions’ represent a very serious form of HPA-axis dysfunction indeed. You would almost certainly expect them to have a major destructive impact on the body. Equally clearly, not everyone who dies of heart disease has Cushing’s disease, or takes steroids. So the next step is to show that other, less obviously severe forms of HPA-axis dysfunction also have the same destructive effect – through the same mechanisms. In order to do this, I want to look at three different initiators of HPA-axis dysfunction:
• Depression
• Smoking
• Spinal-cord injury
Depression first. It has long been known that people with depression are at a greatly increased risk of heart disease, but no one seems to be entirely certain why. However, when it has been studied it is clear that in depression you always find HPA-axis abnormalities.
There is compelling evidence for the involvement of hypothalamic-pituitary-adrenal [HPA-] axis abnormalities in depression. Growing evidence has suggested that the combined dexamethasone [DEX]/corticotropin-releasing hormone [CRH] test is highly sensitive to detect HPA axis abnormalities.
Kunugi H, et al. Neuropsychopharmacology, January 2006; 3
(I left in the stuff about the dexamthasone (DEX)/corticotrophin-releasing hormone (CRH) test for those who do like to see things properly measured, and refuse to believe in things that cannot be measured.)
In addition to the other metabolic problems, depression also leads to a build-up of visceral fat. I popped this observation in to make it clear that visceral-fat build-up is a result of underlying problems with the HPA-axis and raised cortisol levels – it doesn’t happen by genetics. Or, indeed, magic.
We showed that depressive mood is associated with VAT [visceral adipose tissue], not with SAT [subcutaneous adipose tissue], in overweight premenopausal women. These findings may explain some of the association between depression and coronary heart disease. More studies are needed to elucidate the causal relationship.
Lee ES, et al, Obes Res, February 2005; 13
In fact, I think that depression is an almost perfect model to demonstrate that long-term dysfunctions of the HPA-axis – created purely by psychological stressors – works through exactly the same physical, and measurable, mechanisms as Cushing’s disease to cause heart disease. Importantly, if you treat depression, the metabolic abnormalities often disappear – which represents reversibility of effect.
I didn’t need to choose depression to show that psychological upset causes heart disease. I could have presented research on anxiety, or post-traumatic stress disorder, rather than depression. But I can assure you that research in all of these areas shows exactly the same thing. HPA-axis dysfunction, then metabolic abnormalities, then increased risk of death from heart disease. Once again, it is not a coincidence. This is a direct causal chain from HPA-axis upset to heart disease.
Smoking next. Although this may seem to be way out on a limb, it is not, because smoking actually works through exactly the same mechanisms as depression and Cushing’s disease, although the effects are more likely due to repeated short-lived HPA-axis dysfunction, rather than chronic problems.
Two pieces of evidence. The first is taken from a study that looked at the effect of smoking a cigarette on cortisol and DHEA (dehydroepiandrosterone) levels (DHEA is a steroid hormone made in the adrenal glands in response to stress):
Cortisol and DHEA increased significantly within 20 min (P<0.05) and reached peak levels… within 60 and 30 min, respectively. Thus cigarette smoking produced nicotine dose-related effects on HPA hormones and subjective and cardiovascular measures.
Mendelson JH, et al, Neuropsychopharmacology, September 2005; 30
The second study looked at the effects of smoking on ACTH and cortisol levels:
In the control group subjects, cigarette smoking induced a striking increase in the circulating concentrations of ACTH and cortisol, with peak responses 1.4 and 1.5 times higher than baseline at 20 and 30 min, respectively.
Coiro V et al, Alcohol Clin Exp Res, September 1999; 23
In addition to its effects on the HPA-axis, smoking also has a major impact on blood-clotting factors. Whether this is direct effect, or whether it is a result of HPA-axis activation, is not clear.
Finally, in this section, I wanted to mention spinal-cord injury. As with smoking, this may not initially seem to have anything to do with the HPA-axis dysfunction. However, the reality is that a spinal-cord injury impacts with massive force on the HPA-axis. This is because if you break vertebrae, and snap the spinal cord, you (usually) sever many of the sympathetic and parasympathetic nerves at the same time.
Unsurprisingly, this leads to enormous disruption in the entire neurohormonal system. The abnormalities found in spinal-cord injury are wide-ranging and, I regret to say, so complicated that I can’t understand many of them myself. Indeed, most of the papers written in this area discuss hormones, and hormonal axes, that are beyond my ability to describe without tying myself in knots.
So I will use broad brush strokes here. If you want more information, you are perfectly welcome to go to www.pubmed.org and type in ‘spinal-cord injury and/or cortisol levels and/or increased risk of CHD and/or increased visceral fat’. Here, you will find a whole series of papers outlining the same things – namely, that:
• Spinal-cord injury leads to severe HPA-axis dysfunction and raised cortisol levels.
• Patients with spinal-cord injury have low HDL levels (and other lipid abnormalities, e.g. raised VLDL levels).
• Patients with spinal-cord injury have sharply raised blood-clotting factors, including fibrinogen, Lp(a), and plasminogen activator inhibitor-1 (PAI-1).
• Spinal-cord injury leads to insulin resistance, up to and including frank diabetes.
• Spinal-cord injury patients develop visceral obesity.
• Spinal-cord injury patients are at a greatly increased risk of dying of heart disease.
Perhaps I am laying it on with a trowel here; perhaps not. By now, I hope you can see that HPA-axis dysfunction (and abnormal cortisol secretion) ties together a whole series of apparently disparate factors known to cause heart disease. To name but five, these include Cushing’s disease, depression, use of steroids, smoking and spinal-cord injury. (Just try and find another way of linking these things to heart disease other than through HPA-axis dysfunction.)
In addition to this, HPA-axis dysfunction also explains where many of the ‘classic’ risk factors come from, e.g. low HDL, high VLDL/LDL, high blood pressure, diabetes, raised clotting factors and increased visceral-fat deposition. A dysfunctional HPA-axis is the underlying cause of these things.
Do these factors then go on to cause heart disease? Some of them may have a direct impact on heart disease – such as raised blood-clotting factors. Others are probably just signs of an underlying problem, e.g. low HDL levels. When so many things are tangled together, it is not that easy to say which causes what.
Anyway, as a sign-off to this section I want to return to the INTERHEART study. In this study, nine ‘factors’were measured and found to have a close connection with heart disease. Six of them were associated with increased risk, and three of them were associated with reduced risk.
The six factors associated with increased risk of heart disease were:
• Smoking
• Diabetes
• High blood pressure
• Abdominal obesity (increased visceral fat)
• High ApoB/ApoA-1 ratio*
The authors of this paper treated each risk factor as acting in perfect isolation, having no relationship whatsoever to any other factors. However, I would like to point out that every single one of these six risk factors can be directly linked to a dysfunctional HPA-axis and raised/abnormal cortisol levels.
Two of them – smoking and psychosocial stress – are causes of HPA-axis dysfunction. Four of them result from HPA-axis dysfunction: high blood pressure, abdominal obesity, diabetes and dyslipidaemia.
As a quick aside, you may also have noted that, in this 52-country study, a raised LDL level, or raised cholesterol level, was not identified as a risk factor – something that seems to have passed everyone by. They fudged this finding horribly by using the strange concept of the ApoB/ApoA-1 ratio, and using the word ‘dyslipidaemia’ – suggesting the LDL was involved somewhere, but we know that it wasn’t.
Finally, I would like to point out that the three factors in the INTERHEART study that protected against heart disease were:
• A high intake of fruit and vegetables
• Exercise
• Alcohol consumption
Two of these factors – exercise and alcohol consumption – have beneficial effects on the HPA-axis. You think I’m stretching it? Well, have a look at this quote from a study called ‘The effect of a moderate level of white wine consumption on the hypothalamic-pituitary-adrenal axis before and after a meal’:
The results demonstrated a significant alcohol-induced decrease in salivary cortisol irrespective of nutritional status and a significant decrease in salivary DHEAS when alcohol is consumed… It was concluded that moderate white wine consumption may promote a transient alteration in the functioning of the HPA axis.
Pharmacol Biochem Behav, October–November 2001: 70
As for exercise, there is a huge mass of literature demonstrating very clearly that exercise is one of the best things you possibly do to maintain a healthy HPA-axis.
In fact, when you get down to it, the only factor in the INTERHEART study that cannot be related to the HPA-axis, at least not in any way that I know of, is the protective effect of eating fruit and vegetables.
So, while the authors stated that a mere one-third of the risk of heart disease could be due to psychosocial stress, if you look at the evidence in a different way it could be argued that the entire risk of heart disease is due to a dysfunctional HPA-axis – otherwise known as stress.
How do the abnormalities found with high cortisol levels cause heart disease?
By now I hope to have convinced you that a whole range of different ‘stressors’ can upset the HPA-axis. Some operate over an extended period, some are transient but repeated, e.g. smoking. Some are physical; some psychological.
What I need to do now is make the final link in the chain. How do the metabolic abnormalities created go on to cause heart disease – or, to be more accurate, atherosclerotic plaque growth?
To answer this I need to return to the ‘response to injury’ hypothesis, first proposed by Carl Freiherr von Rokitansky more than 150 years ago. It’s a hypothesis that has found support among many scientists over the years, and has an increasing following today – although it has to be said that, as with most hypotheses involved with heart disease, it has fragmented into a number of different versions. However, the basic concept is pretty straightforward, and I happen to think that it is correct.
In the ‘response to injury’ hypothesis, the first step in plaque formation is that a patch of endothelium (the thin, fragile, single-celled layer lining the arterial wall) becomes dysfunctional, damaged, or – more likely – is just plain stripped off.
When this happens, a section of the underlying arterial wall is exposed. This, in turn, acts as a very powerful stimulus to the clotting system to form a blood clot (or thrombus) to plug the gap. Once the thrombus has covered over the area of damage, the clotting process is brought to a halt. This is the basic ‘response to injury’.
Then what happens? Well, for a moment, I would like you to have a think about what happens to your skin if you scratch or cut it. Blood escapes for a bit, then a clot/thrombus forms, which turns into a scab. After a while, the skin re-grows to seal up the scratch under the scab, and the scab falls off. If the same process were to happen in your arteries, then any blood clot that formed on a damaged bit of endothelium would eventually fall off, travel a bit further down the artery, and then jam solid once the artery narrowed. This would cause catastrophic problems – including, for example, strokes. Clearly, this is not a good thing. Therefore, blood clots forming on arteries cannot be allowed to fall off when the endothelial healing process is complete – unlike scabs on your skin.
In order to stop blood clots breaking off artery walls, and causing downstream havoc, they have to be drawn into the artery wall and then disposed of. How does this happen?
Answer: your bone marrow creates millions upon millions (upon billions, probably) of ‘pre-endothelial’ cells (also known as bonemarrow-derived vascular progenitor cells [VPCs]) that travel about in your bloodstream. When they see a breach in the endothelium, it’s their job to cover it up.
Normally, however, a blood clot will have got there first, so these pre-endothelial cells stick to the surface of the blood clot, grow into full-blown endothelial cells, and cover over the damage with a new layer of endothelium. In this way, blood clots are, effectively, drawn into the arterial wall behind a new layer of endothelium. Usually they are then broken down, and removed, leaving no trace that they were ever there in the first place.
Now, you may be thinking, I hope, that this all makes perfect sense. But I have got to admit that the final part of this hypothesis about how arteries deal with blood clots is mine. I just kind of figured that it made sense. Having said this, virtually every other part of this hypothesis is known, and accepted. For example, everyone accepts that the endothelium can be damaged and everyone accepts that blood clots form over areas of damage. The only bit that is speculative is the idea that endothelial re-growth covers over thrombi, pulling them into the artery for disposal, rather than letting them break off and charge downstream. That said, frankly I don’t know what everyone else thinks actually happens to thrombi that form on arterial walls, as there is no other version of events that makes sense. (I get the impression that most people haven’t actually thought about this at all.)
I believe that this extended version of the ‘response to injury’ hypothesis is very strongly supported by some fascinating recent research done at Duke University in the USA:
Scientists at Duke University Medical Center have discovered that a major problem with aging is an unexpected failure of the bone marrow to produce progenitor cells that are needed to repair and rejuvenate arteries exposed to such environmental risks as smoking or caloric abuse.
The researchers demonstrated that an age-related loss of particular stem cells that continually repair blood vessel damage is critical to determining the onset and progression of atherosclerosis, which causes arteries to clog and become less elastic.
I would just ask the question: why would you have pre-endothelial cells circulating in your bloodstream if not to cover up areas of endothelial disruption? What else could they possibly be there for? Until a few years ago, no one even knew that these progenitor cells existed. Now they have been discovered – though, of course, if you think about it, they had to be there. Otherwise we would all be dead, as a damaged artery would never be able to repair itself.
The existence of progenitor cells also explains another issue that mainstream researchers have been grappling with for years. Namely (if you are still clinging to the cholesterol hypothesis), how can plaques form behind an intact endothelium, when LDL cannot penetrate intact endothelium? The answer is, of course, that plaques (which contain Lp(a) – a form of LDL – and LDL itself) start life as thrombi on top of damaged endothelium.
When new endothelial cells re-grow over the top of a thrombus they effectively draw it into the artery wall, along with Lp(a) and LDL. Puzzle solved: plaques don’t actually form behind the endothelium at all. When they first form, that section of endothelium isn’t actually there.
Moving on, if the processes that I have described up to now are ‘healthy’, what makes them become damaging? Or to put this another way, what causes a blood clot to remain stuck inside the artery wall, then grow into a big unstable plaque, instead of being disposed of by the repair systems (as I believe must happen to the majority of thrombi that form)?
The answer to this question is that plaques do not gradually grow by absorbing substances from the bloodstream, molecule by molecule, in some agonisingly slow diffusion-type process. They grow through repeated acute episodes of endothelial damage, followed by thrombus formation, all taking place on top of an existing plaque. In short, plaques grow in sudden, discrete episodes. And you don’t need to take my word for this, because all the evidence I need for this version of events comes from the American Heart Association in their ‘Scientific statement: a definition of advanced types of atherosclerotic lesions and a histological classification of atherosclerosis’:
… 38% of persons with advanced lesions [plaques] had thrombi on the surface of the lesion. These thrombi ranged in size from minimal [microscopic] to grossly visible deposits, and some consisted of layers of different ages. Immunohistochemistry revealed wavy bandlike deposits related to fibrin [a key component of blood clots] within the advanced lesion of an additional 29% of persons. Because of their structure, these were thought to represent the remnants of old thrombi. Similar data were reported by other authors.
The fissure and hematomas [a form of blood clot] that underlies thrombotic deposits in many cases may recur, and small thrombi reform many times. Repeated incorporation of small recurrent hematomas and thrombi into a lesion over months or years contributes to gradual narrowing of the arterial lumen.
As this passage makes clear, repeated thrombus formation over plaques is what makes them get bigger. How else could you find fibrin, a key component of blood clots – and one that absolutely cannot pass through the endothelium – in distinct layers within plaques? How else could you find blood clots of different ages within plaques? You’re right, you couldn’t.
Further supporting the conjecture that thrombus formation is central to heart disease is the knowledge that the final event in heart disease is plaque rupture, with the formation of a very big blood clot on top of the plaque – big enough to completely block a coronary artery.
Almost all of this is accepted by the mainstream – with varying degrees of enthusiasm. What they will not accept is that the thing that gets the plaque started in the first place is endothelial damage, followed by formation of a blood clot. Even though this is exactly the same process that creates plaque enlargement and, eventually, fatal plaque rupture.
Why won’t they accept this? Because it doesn’t fit with the damnable cholesterol hypothesis. No hypothesis is allowed to exist that does not have a raised LDL at its heart. And the ‘response to injury’ hypothesis that I have outlined does not need LDL to make it work and also explains why a significant proportion of people who suffer heart attacks do not have a raised LDL. In fact, it explains everything.
Returning to the ground from my soapbox, I shall now tie a few things together:
• Plaques start life as small areas of damage to the endothelium, which are normally healed by the body’s natural repair mechanisms – thrombus formation and endothelial re-growth.
• Plaques grow through repeated episodes of endothelial damage and blood-clot formation in the same spot.
• Plaques kill you when they ‘rupture’, creating a major blood clot that then blocks an important artery somewhere in the body.
Therefore, factors that cause accelerated plaque growth will be anything that has the capability either to damage the endothelium or cause more dangerous/bigger blood clots to form. Or both.
So what factors have been found to cause ‘endothelial dysfunction’? They include:
• High blood-sugar levels, especially ‘spikes’ of blood sugar following a meal
• High insulin levels
• Acute mental stress
• Smoking
• Cocaine use
• Cortisol
• High levels of adrenaline
Okay so I have mixed up my factors a bit – some are mental disturbances and others are chemicals circulating in the blood – but you get the general drift. These are all factors that I have listed under the title ‘unhealthy stressor’, or else a downstream metabolic abnormality created by HPA-axis dysfunction.
I am not going to provide evidence to support this list. If you wish to check the facts for yourself, go to Google, or www.pubmed.org and type in ‘endothelial dysfunction’, followed by any one of these factors. You can then read the abstracts and papers for yourself. (I believe that this is a more honest form of referencing, rather than just picking the twenty or so papers that support my case, and failing to point out the ones that don’t – though there aren’t any of those anyway.)
Next, I think it is important to look at the factors that make the blood more ready to clot, and more ready to form big and difficult-to-shift blood clots. These are, somewhat unsurprisingly, blood-clotting factors, such as:
• Fibrinogen (Fibrinogen is a small strand of protein. When you stick hundreds of bits of fibrinogen together, it turns into a long, thin, very strong strand of fibrin. This binds blood clots together.)
• Lp(a)
• Plasminogen-activator-inhibitor-1 (PAI-1)
• Von-Willibrand factor
• VLDL (VLDL stimulates blood clots to form.)
I could actually go on giving you a list of clotting factors as long as your arm. Suffice to say that in study after study, you will find that raised blood-clotting factors are directly and consistently associated with an increased risk of heart disease, with no contradictory evidence in any study that I could find.
Although I could give you hundreds of studies supporting this statement, for the sake of brevity I will stick to one, from the New England Journal of Medicine, June 1995:
In patients with angina pectoris, the levels of fibrinogen, von Willebrand factor antigen, and t-PA antigen are independent predictors of subsequent acute coronary syndromes. In addition, low fibrinogen concentrations characterize patients at low risk for coronary events despite increased serum cholesterol levels. Our data are consistent with a pathogenetic role of impaired fibrinolysis [blood-clot break-down], endothelial-cell injury, and inflammatory activity in the progression of coronary artery disease.
The importance of blood clots in heart disease is also supported by the fact that virtually every drug that reduces the risk of dying of heart disease is, essentially, an anti-coagulant. For example:
• Aspirin – stops platelets becoming ‘activated’ and sticking together (activated platelets are critical to thrombus formation).
• Warfarin – reduces various clotting factors in the blood.
• Alcohol – stops platelets sticking together.
• Tissue plasminogen activator – breaks clots apart.
• Statins – have strong, dose-dependent, anti-coagulant activity.
• Streptokinase – a clot-buster.
• Clopidogrel – stops platelets sticking together (see aspirin).
• ACE-inhibitors (used to lower blood pressure) – ACE-inhibitors stimulate nitric oxide synthesis in endothelial cells. Nitric oxide is the most powerful anti-coagulant in the body.
On the other hand, drugs that increase the risk of blood clotting, such as Vioxx, greatly increase the risk of dying of heart disease.
The HPA-axis, response to injury model of heart disease
At this point, I believe it is now possible to put together a reasonably simple model of heart disease that leads from ‘unhealthy stressor’ to heart disease via HPA-axis dysfunction, raised cortisol levels and a series of metabolic abnormalities (see Fig. 30).
Of course, I am not the only person in the world to have recognised most, if not all, of these steps. In fact, a number of researchers are looking very closely at raised cortisol as the primary cause of heart disease.
The main reason for this sudden interest is the knowledge that the metabolic abnormalities of Cushing’s disease are exactly the same as the metabolic abnormalities of Syndrome X. And Syndrome X is increasingly viewed as the number-one cause of heart disease – despite that fact that many in the mainstream refuse to recognise that it exists as a separate entity at all.
Fig. 30 HPA-axis dysfunction ‘response to injury’ model
So what, you might ask, has stopped the model I presented above, or something very much like it, from becoming widely accepted (apart from the fact that it isn’t the LDL hypothesis, of course)?
What has stopped it is the following simple fact. Many people suffering from heart disease, and/or metabolic Syndrome X, do not have a high cortisol level. In fact, it is often found to be low. Which, you might think, completely scuppers this model altogether.
But if you did think this, you would be wrong. The reason for your wrongness is that cortisol secretion usually peaks at about 8 a.m. – a time that coincides with you getting up and getting ready to do battle with the world for the next 16 hours or so. Following its early morning peak, the cortisol level falls during the rest of the morning, then rises a bit, then falls, then rises. This is all very much dependent on what you do during the day.
However, in people with HPA-axis dysfunction, very often what happens is that they lose the normal, healthy, flexible response to various stressors during the day, including the early morning peak. In effect, the HPA-axis ‘burns out’ and just pumps out the same amount of cortisol night and day, with no alteration in response. It becomes inflexible and non-variable. Which means that if you decide to measure the cortisol level at 8 a.m., or 9 a.m. (which are the standard times for such tests to be done), it can often be low in people with HPA-axis dysfunction – although not always, it depends on the degree of HPA-axis dysfunction. In reality, in order to diagnose HPA-axis dysfunction properly, you need to take repeated measurements during the day to look at what the cortisol level is doing. You also need to see if the normal ‘healthy’ response to stressors remains.
This brings me to the outstanding work of the late Per Bjorntorp. Some years ago, he recognised that you need to do more than a solitary cortisol measurement to diagnose HPA-axis dysfunction. He knew that the human body is a flexible and dynamic organism. Health, and healthy systems, are constantly adapting and reacting. When you lose flexibility and responsiveness, you die.
Perhaps the most spectacular example of this is heart-rate variability, i.e. the amount by which the heart rate alters from beat to beat. This is, possibly, the single most sensitive indicator of a healthy heart, and a loss of beat-to-beat variability is one of the most powerful single indicators of the risk of dying of heart disease.
Armed with such knowledge, Bjorntorp wasn’t just looking for high, or low, levels of cortisol at 8 a.m. He was more interested in seeking a loss of HPA-axis flexibility, and ‘burn-out’ of the axis. To give one example of his work in his own, rather distinctly Swedish, words:
The conspicuous similarities between Cushing’s disease and the Metabolic Syndrome X open up the possibility that hypercortisolaemia [high cortisol level] is involved in the latter. Salivary cortisol is possible to measure during undisturbed conditions including perceived stressful events during everyday life.
Such measurements clearly show that normally regulated cortisol secretion is associated with excellent health in anthropometric, metabolic and hemodynamic variables. Upon perceived stress cortisol secretion is increased and followed by the Metabolic Syndrome X [insulin resistance, abdominal obesity, elevated lipids and blood pressure]. In a minor part of the population a defect, ‘burned-out’ cortisol secretion occurs, with decreased sex steroid and growth hormones secretions and strong, consistent, associations with the Metabolic Syndrome X.
Psychosocial and socioeconomic handicaps with tendencies to abuse and depressive-anxious mood changes are consistently associated… [with HPA-axis dysfunction].
We suggest that the Metabolic Syndrome X is due to a discretely elevated cortisol secretion, discoverable during reactions to perceived stress in everyday life.
Bjorntorp, Ann MY Acad Sci, November 1999
Bjorntorp did a great deal more such work, showing exactly the same things, and other researchers have fully confirmed his findings. In my view he should have got a Nobel prize, for he proved beyond doubt that exposure to various stressors causes HPA-axis dysfunction, abnormal cortisol levels and then heart disease – precisely in that order. Has anyone outside of a small, devoted band of followers heard of him? Not likely.
The final thing I need to demonstrate is that the ‘HPA-axis dysfunction response to injury’ hypothesis actually fits the facts, and can explain the enormous variations in heart disease seen around the world. For the sake of brevity from now on I shall call this hypothesis, the ‘stress hypothesis’, as the full, accurate definition is a bit of a mouthful.
*I have tried to find out what they meant by this ratio, but I cannot get an answer from the authors of the study. So, I have to assume this means a low HDL and raised VLDL and LDL level – as both of these lipoproteins have ApoB attached.