Hearts, famines and grandparents
When my father went to bed one night complaining of indigestion and failed to wake up the next morning it was a shock. The subsequent inquest and post mortem showed he died of a heart attack. He was only 57 years old and had no obvious risk factors except a lifelong dislike of exercise. He didn’t smoke, had a normal diet and had no previous heart problems; both his parents had lived twenty years longer than him. The underlying reason for his premature death was a mystery. He was a medical pathologist, an expert on heart disease and cot death syndrome. At the time of his death in 1981 he had been writing the outline draft of a book entitled Social Pathology: The Social and Psychological Causes of Disease. Executive stress was the trendy subject of the moment, and the role of genes in causing death and disease was barely discussed.
When twins die within minutes of each other, as does occasionally happen, it is the classic type of news story that makes headlines around the world. Such stories help to reinforce the traditional view that you can’t escape your genes - and the timing of natural death is pretty much pre-ordained. But the vast majority of identical twins do not die within hours of each other, and one usually survives the other by several years. One of the world’s largest twin studies followed up 40,000 Swedish twins for 36 years. In the pairs where one died of heart disease, only 40 per cent of male and 30 per cent of female identical twins also died the same way.1 In the cases of non-fatal heart attacks, the correlation between identical twins was even weaker.
Despite sharing the same DNA and genetic make-up as well as similar environments, why is it that most identical or non-identical twins, let alone siblings, do not usually suffer from the exact same diseases? The structural DNA in every one of the trillions of cells in our body is, for practical purposes, identical, as are the individual genes and gene variants. As most cells in our body replicate and divide to keep us growing or alive, this genetic material is passed on exactly to the next generation of cells. In this way our cells can be thought of as mini-replicas of ourselves. Now DNA makes proteins, themselves made up of amino acids, which are the main drivers of all the chemical processes in the body. There are four times as many proteins as genes, and these proteins can also activate genes themselves. In other words, it is not the case that every gene makes one protein, but some genes make several proteins and some proteins are produced by the joint work of several genes. The possible proteins combinations are endless and the same genes can generate very different messages in the cells.
The cells that make up the 200 different tissues in our body (by which I mean heart, lung, bone, skin and so on) all behave differently because they have different proteins which vary in each cell type with different functions. How can this be? Clearly, although every cell has exactly the same DNA and the full complement of 25,000 genes, they are not all used or expressed equally. It is this variable gene expression that distinguishes the cells from each other. This is caused by what are called epigenetic signals like methylation (adding a methyl group) which we have discussed before, or the more complex histone modification (chromosome folding), which controls the expression of the genes. These messages get passed down the generational line each time a cell divides, making sure that a heart cell as it grows or ages, importantly keeps behaving like a heart cell and doesn’t get distracted and revert to some other kind of cell.
We believe that hundreds if not thousands of genes are involved in most complex traits like heart disease, each with a tiny individual effect.2 To use a musical analogy, the genes are like the 25,000 pipes of an organ used in some concert halls (our cells), whereby hundreds function in harmony, being opened and closed by the mysterious (epigenetic) organist to control the music. It no longer makes any sense to talk of a ‘heart attack gene’ any more than you can say there is a single note for Bach’s Fugue in D minor.
Because multiple genes work together their effects are hard to predict and minor subtle changes in a few genes can have major consequences. This is where epigenetics and methylation feature. Small changes in a risk factor like smoking or gaining weight can alter methylation in a large number of genes and so affect their function in many ways. This is why if we want to better predict clinical outcomes like heart attacks we may need to look elsewhere.
At 4 a.m. on a cold winter’s morning close to the Scottish border a strong wind was battering the windows of Andre’s house, but it was the strange aching down his neck and left arm that woke him. He hadn’t done anything out of the ordinary the previous night, just sat in the pub as usual, had a few drinks and his weekly Friday night curry. When the pain kicked in again, harder and vice-like, he suddenly felt like retching and prodded his sleeping wife. She woke to find that he seemed to have stopped breathing.
At the age of 67, Andre had just suffered a heart attack; his doctor would later note in her files that he had suffered a myocardial infarction due to blockage of his coronary arteries. Like two thirds of victims nowadays he survived this first attack, giving him a chance to re-evaluate his life. He would describe himself as well-built, at nearly six foot tall, but with a bit of a weight problem that had afflicted him since childhood, despite his occasional attempts at cycling. His older brother, who had moved to the US, and his sister, who still lived in the Netherlands, where he had been brought up, were both fitter and slimmer than him and had no heart problems.
Until the twentieth century heart disease was unusual. In the West it is now the commonest form of death. In countries such as China and India, where death from heart problems was virtually unknown in rural areas until the last thirty years, the situation is changing rapidly. 60 per cent of all worldwide deaths are now due to so-called Western illnesses, such as heart disease and diabetes. Genes cannot explain the massive increase in the rates of heart disease over the last seventy years, nor can it be down to slow Darwinian selection. Though identified lifestyle risk factors such as obesity, poor diet and smoking, have increased.
So Andre’s problems could just be a result of his unhealthy lifestyle. Although trends are improving, Scotland, where he lives, has high levels of heart disease and instances of poor diet. But let’s not blame the Scots alone. Could we also blame the Germans? Andre’s destiny could well have been altered before he was born when his Dutch mother was pregnant.
In the winter of 1944 the forces of Nazi Germany were in controlled retreat after the Normandy landings in the West and the Soviet offensive in the East. Attempts by the Dutch population to help the British advance from Arnhem with attacks by partisans and a National railway strike had failed. The Germans took their revenge. They stopped food reaching the west of the country and flooded huge areas of farmland at the same time as an unusually cold winter hit transport and natural food supplies.
Roughly half the Dutch population starved, living on less than 1000 calories a day. Over 20,000 did not survive, those that did, like Andre’s mother, had to beg for food and resorted to eating pets, rodents and tulips to stay alive. After the war a group of British and American doctors were flown in and saw the opportunity to look at the effects of starvation on long-term health, in what became the Dutch Famine Birth Cohort Study. In the rubble of devastated cities like Rotterdam, they managed to find key obstetric records, which showed the markedly lower birth weights of babies exposed to maternal malnutrition in the crucial last few months of pregnancy and the paradoxically larger birth weights of babies exposed in the first three months.
Following the lives of these children, the researchers noticed that when the boys born in the famine had reached 18 years of age and enlisted in the army they were considerably fatter than their friends born in the east of the country or even neighbours or relatives born just six months earlier or later, who had not suffered from prolonged fetal malnutrition. As they grew older in their 1950s and 1960s they found that rates of obesity, heart disease, diabetes and schizophrenia were also greater. Similar research in the 1980s found that 5,000 British men born in Hertfordshire who had the lowest birth weights, also had the highest risks of heart disease 60 years later.3 So some of Andre’s current propensity to obesity and heart disease could well be linked to his poor nutrition as a fetus. The studies are still on-going into whether the risk of famine extends to the third generation.
Other such tragic ‘natural experiments’, along with good historical detective work, could help support these findings. Hidden from the West for many years, the Chinese famines during the so-called Great Leap Forward affected large parts of many rural provinces. Like in Stalin’s Russia twenty years before, this was caused by collectivisation policies and the pseudo-science of Lysenko’s crazy agricultural ideas. Up to 50 million Chinese peasants are believed to have died.4 Mao used the reforms to also pay off massive debts to Russia in the way of grain, which Russia ironically needed because of its farming failures due to his own collectivisation policies. A third of all the grain from villages was requisitioned for this purpose and Mao was quoted as saying, ‘When there is not enough to eat, people starve to death. It is better to let half of the people die so that the other half can eat their fill.’ Fifty years later, the official Chinese position on the famine remains a sensitive issue.
Although only very basic official records are available recording births and deaths from the provinces, there are huge numbers of them. They show that starving women had massively reduced fertility rates in the following year. In regions with sudden drops in fertility, researchers surmised that famine conditions were present for at least 12 months before. They then followed up the rates of psychiatric diseases in these women’s children twenty or more years later compared to children from non-famine areas. Children born from starving areas showed twice the normal rates of schizophrenia as adults. The same pattern was seen in the Netherlands, and in the Anhui and Guangxi provinces in China, regions 1,000 miles apart - showing the risk was worst in rural areas where the famine hit hardest. Epigenetics could be the mechanism for both these remarkable observations in China and the Netherlands.
It is possible, however, that the consequences of both the Chinese and Dutch famines were simply a direct effect on the fetus due to lack of nutrition. Knowing whether such risks could be passed on to the third generation would therefore be crucial to determining whether epigenetics played a role. Although not many countries keep detailed historical records of harvests, obstetric and nutritional details as well as causes of death across generations, luckily for us, the Swedes have long been keen record keepers.
The small farming community of Norrbotten in the north of Sweden was so isolated and self-sufficient in the nineteenth century that if the harvest was bad, as it was in 1800, 1809, 1812, 1828 and 1856, people starved. If it was good, they gorged themselves excessively. One study focussing on the parish of Överkalix and a cohort of 330 adults born in 1905, estimated the food availability for their parents and grandparents on the basis of harvests and food prices over the period from 1803 to 1849. The oocytes (eggs) in the girls and sperm cells in boys (known as germ cells) are formed differently to normal cells with half of the complement of chromosomes and stored separately to the rest of the body, eventually in the gonads. The study focussed on the period between the ages of 9 and 12 years, just before puberty, when the DNA in sperm, which is usually well-protected in the scrotum, is moving downwards and probably most susceptible to epigenetic modification.
The first results were a surprise. They showed that adults whose grandparents had gorged themselves in feast years, died on average six years earlier than those whose grandparents survived famine years. When they expanded the group and looked at their genders, the findings became even clearer. The paternal grandsons of men exposed to famine before the age of 12 did rather well and were actually less likely to die of heart attacks.5 Over-eating grandfathers seem to have passed on to their grand-children not only an elevated risk of earlier cardiac death but also a four-fold risk of diabetes.6 The findings were most clearly seen in males but a similar trend was found in women. Yet again the trend was found only within genders: women were affected only by the habits of their grandmothers, men only by their grandfathers. So something happened to the eggs or sperms of the respective grandfathers and grandmothers during the famine which affected their grandchildren two generations later. A follow-up study in Bristol showed that early childhood smoking (before 11 years) in 166 fathers produced fatter sons, but not daughters.7 Over-eating before puberty had the same toxic effect on the next generation.
The only other similar study in humans comes from the study of Betel nuts. These are commonly chewed around the world by 600 million people in Asia, and Taiwan in particular. Taiwanese researchers already knew that its regular consumption was linked to increased rates of cancer of the mouth, but now found it was also related to increased risk of later obesity and diabetes in men and importantly their sons.8 This result suggested that even exposure after puberty could still cause epigenetic changes in the sperm when it had reached the relative safety of the testicles. This may be because men continue to produce sperm all their lives and sperm DNA can be influenced over many years. Similar experiments in male mice fed betel nuts produced the same results: offspring with increased rates of diabetes, particularly males.9
The fact that changes in weight and diabetes risk occurred through the male line via sperm, rules out a direct effect of fetal nutrition as the cause. The environmental signals or changes to the genes must have somehow been passed on genetically to the grandchildren. One of these key genes that researchers examined is involved in the growth of the fetus (IGF2). This was abnormally methylated in sons and daughters of Dutch famine victims tested 50 years later when compared to their siblings born shortly before or after.10 Other studies of Dutch famine babies suggest they have a significantly lower risk of colon cancer, one of the commonest cancers in adults.11 They found the important promoter area of the cancer enhancing gene was over-methylated, effectively turning it off.
These historical studies tell us is that what your parents or grand-parents were doing or consuming before you were conceived can have a major influence on your life at least two generations later. What we thought might be an adverse environment like starvation, while leading to a high schizophrenia risk, possibly through the stress pathways we discussed earlier, could strangely also confer some advantages in protecting against cancer. By the same rules, over-eating at different times can also have good and bad consequences in future generations. Separating cause and effect is not easy when studying socio-economic factors in health and disease, but twin studies can again help us tease them apart.
Joyce and Margaret were identical twins born in post-war Dulwich, in south London. Their family moved to Reading, 30 miles away, to a very average middle-class, three-bedroom semi-detached house, where they grew up with another sister. Their father worked for the post office and their mother was a part-time nursery teacher. They had a fairly happy childhood, didn’t get into trouble and had no major illnesses that they remember. They looked alike and shared most things, until they left home at 18. They both first went to secretarial college for a year, but Margaret got bored and found a job at Selfridges, while Joyce completed her training and got a job as a junior secretary in a law firm. They started to move om different social circles. Margaret was the first to announce she was getting married – to Frank, a friendly plumber from Walthamstow. Joyce followed the next year and married Anthony, one of the junior lawyers in her firm.
They had five children between them and led generally happy lives, but their financial statuses started to diverge when Anthony was head-hunted by a big city law firm and unexpectedly early became a partner. Frank had some recurrent back problems and several times had to take 6 months off from his own plumbing business, which suffered badly. Margaret went back to work to keep the family going, and thoughts of buying their own house were put on hold, while Anthony and Joyce moved into a large family house. Although less than 5 miles apart, they lived in different social worlds, causing a strain on their relationship.
When we saw the pair for a twin examination, 35 years since their paths had diverged, some of these differences were now evident. Both twins had smoked continually in their twenties, but Joyce had managed to quit after getting married, while Margaret admitted to a 15-a-day habit until she was 40. Both still liked a drink: Margaret preferred port and brandy, Joyce white wine. Both were the same height, 5′ 4″, but Margaret was about 15lbs (7kg) plumper than Joyce, who went to exercise classes. Margaret, despite her plumper cheeks, looked older than her sister, who also had more of her teeth. When we tested their DNA we found that Margaret had a shorter average telomere length (by 200 base-pairs) in her white blood cells than Joyce. This suggested that although they had the same chronological age to within a few minutes, the biological age difference between them now reflecting the age of their cells was around 7 years, even accounting for their smoking.
We found the same social effects as in Joyce and Margaret in a formal study of several thousand British twins, which allowed us to remove the effect of genes by comparing within identical pairs. Social status was clearly correlated to a biological marker of early ageing, or telomere length. 12
A separate study of 10,000 British civil servants showed a three-fold increase between those at the top and bottom of the administrative tree in terms of risk of heart disease.13 However there were, as expected in the lower-level messengers and cleaning staff, many more bad habits, like smoking, poor diet, and lack of exercise. Yet even when all known risk factors were combined and accounted for, they only explained one third of the excess mortality: the majority remained unexplained. The authors of this study believed the missing factor could simply be stress. This in turn may be due to a perceived feeling of loss of control caused by the relative position of an individual in a society – something unaffected by absolute wealth. Epigenetics could explain how stress translates into health problems. It is well known that the lower in the perceived social pecking order someone is the more long-term stress they are under. This is true in rodents, primates and humans.14 A recent study in monkeys placed females in different social environments and groups and noted their level of stress and health. The worse their social position in the new group they joined, the worse their stress levels and also their immune system. The mechanism at work appears to be methylation, epigenetically deactivating their immune genes.15
The telomeres we measured in the DNA of the twins are a marker of the accumulated stress in a cell: they act as protective caps (like plastic on boot-lasses) on the ends of the chromosomes to stop them being eroded. As you age or get cellular stress due to smoking, obesity, chronic illnesses or social status, these telomeres get progressively shorter.16 Once they shrink below a certain number of base-pairs (around 13), they no longer protect the chromosome and the cell auto-destructs. The more the relative social differences between groups, the more the cells are subject to stress (often called oxidative stress) and the more they lose telomeres.17 This increases the risk of early heart problems and many other common age-related diseases.
The male residents of central Washington can expect to live 17 years less than their neighbours in suburban Maryland, and there are up to 25-year differences between residents of some areas of Glasgow.18 These mortality differences cannot be explained by the classic risk factors of nutrition and lifestyle alone. The poor groups in Maryland and Glasgow are under more stress and have less control over their lives. They have the same low life expectancy as the average male in rural India. Chronically stressed cells and their owners do not grow and replicate to the same degree. Although the human studies haven’t yet been performed, animal studies suggest it is likely that chronic stress epigenetically switches off the genes that are important for normal development and the heart. The same epigenetic stress message will continue to be transmitted through the next few generations, perpetuating the tendency – and possibly the social divide.
The key message here is that the environment and exploits of our parents and grandparents influence us in a number of ways – modifying our growth, altering our brain development and affecting our risk of diabetes and heart disease. These environmental stresses have been passed to us epigenetically, by so-called ‘soft inheritance’. The reason that lifestyle and environmental risk factors for disease and our mortality have been so hard to pin down may be that we have been looking in the wrong place and wrong time. We should have been sending out questionnaires and surveys to our grandparents 100 years ago.
The bigger question has to be how this affects the future. What we eat today will have a major impact on future generations, regardless of what they eat. The answer as to how they will fare can be found in our present environments and diets – something we will explore in the next chapter.