At the age of just 23, Kristin mentioned worries about a lump in her breast to her mum and twin sister Maren on a trip to Barcelona. Upon their return home, Kris went to see her GP, who examined her and said that it might just be a hormonal reaction to the contraceptive pill. They briefly discussed breast cancer – Kris’s grandmother had had the disease in her thirties but lived till 75 – but the GP was not concerned and prescribed evening primrose oil. The pain continued, and six months later after a long work trip to China, Kris returned. A different GP, without examining her, suggested that she change her contraceptive pill – which she stopped, to no effect. At last, several weeks later, at the insistence of her mother, Kris returned to her original GP and insisted on a referral to the local hospital, where the diagnosis was eventually made. It was breast cancer.
‘It was a real shock for me but also very hard for me to tell Maren, but I knew it would be even harder for her to hear. But she needed to be tested too,’ she said. Maren quickly had an MRI scan, which came up clear. Further tests brought more bad news: Kris’s tumour was already very large and had spread to her spine. As she put it bluntly: ‘I’ve got stage-four cancer. There is no stage five.’
Three years on, Kris amazingly is still very much alive and has now had chemotherapy and radiotherapy to shrink the tumour, followed by a mastectomy. Having been totally bald when I last saw her, she has now regrown her hair and is having cement put into a vertebra of her spine damaged by radiotherapy and the tumour. She takes Tamoxifen, the anti-oestrogen cancer drug. ‘One of the worst things about being bald and sick was not being recognised as twins any more – that was a real low point – but I’m much better now, and look like Maren again.’
Kris and Maren are the daughters of Anglo-German parents. They spent the first half of their childhood in Germany, but moved to the Midlands after their parents divorced in the mid-1990s. At school the tall skinny blonde Germans (as they were known) were very close, sat in the same classes, studied the same A levels, and achieved the same top grades (three As). They had very similar upbringings and environment until the age of 19, and even travelled round Australia in their gap year together. They then chose different universities. They kept in close contact through student life and spoke every day.
‘I did think: “Why her, why not me?”’ says Maren. ‘But then I realised that had I been asked which one of us was more likely to get cancer, it was always going to be Kris. She was always the one with most health problems. She alone had headaches and was homesick, she had her appendix and wisdom teeth removed, while I still have mine.’ ‘I guess I was the slightly weaker more anxious one, who got more stressed,’ agrees Kris. ‘My sister was always a little bit stronger.’
Apart from this, there were no major clues to why Kris was affected and not Maren. They weighed the same at birth – 7 lb with no complications. They both grew well, had their first periods late – at 14 and 15 – had their first boyfriends at 17, and both had irregular periods. Maren didn’t get on with the contraceptive pill and stopped after a few weeks, whereas Kris was on it for two years. They both had the same identical ‘Wurst-und-Kartoffeln’ (sausage and chips) diet for their early German years, then continued with their mum’s fish fingers. They are now both non-dairy vegetarians since the diagnosis.
Genetic tests performed in Oxford have established that Kris is not a carrier of the rare breast cancer gene mutations called BRCA1 and BRCA2, which is a relief to Maren, who now has yearly scans. Kris is positive and upbeat: ‘I could let those questions about my cancer get to me and stop me from doing things. But I can’t see into the future and I definitely can’t live every day clouded by the thought of death. So the only answer is to live – and live well.’ Kris has set up a charity called CoppaFeel, which is educating young women about self-screening early in life, to reduce delayed diagnoses like hers.
Most people, if asked, would assume that breast cancer is strongly genetic. However, although it runs strongly in some unusual families with rare gene mutations such as BRCA1 and 2, it is overall only around 25 to 30 per cent heritable. This explains why identical twin sisters, who share the structure of their DNA, are often differently affected by breast cancer, as with Kris and Maren. The disease also gets less heritable with increasing age, so there is hardly any gene influence when women are over 70 years old, when the majority of cases occur. Although many think of breast cancer when we think of heritable diseases, breast cancer is in fact only about a third as heritable as a common but dull disease like back pain.1
When we say that breast cancer is overall only about 25 per cent heritable, obviously it doesn’t mean that genes aren’t involved – in fact they are crucial in the cancer process and geneticists have discovered at least 10 different sub-types depending on the genes involved2 – it’s just that these changes or genetic risks are only weakly passed from one generation to another. Maren now has less than a 50 per cent chance of developing Kris’s form of breast cancer, whereas if Kris had one of the rarer BRCA1 or 2 genetic forms of the disease she would have a nearly 90 per cent chance, as these forms are much more heritable.3
Kris’s GPs would never have seen a case in someone of her age, whereas they commonly see breast-tissue changes due to the contraceptive pill, so Kris was unlucky in that regard. We now know too that women with firmer, high-density breast tissue (which is mainly genetic) have higher cancer risks and are also harder to examine and feel lumps in. Breast cancer is very rare in the young and produces only around 20 cases a year in the UK in women under age 25. It then increases slowly with age. Eighty per cent of breast cancer is diagnosed over the age of 50, when most routine screening starts. It affects around one in eight women in the UK and USA, and rates have generally been steadily rising over the last 30 years, with numbers doubling in the UK.4 Although there is a real increase, some inflation comes from better diagnosis and screening and a recent trend of labelling early reversible changes (called cancer in situ) as definite cancer. Breast-cancer rates also vary widely across countries. Northern Europe has four times the rates of Africa and Asia, although there have been recent dramatic rises in Japan, and in cities in China experiencing social and economic changes, which can’t be explained by differences in screening.5
What exactly caused Kris’s cancer – the most common cancer in women – that the odds say will not affect her identical sister? Epidemiologists have thought they know most of the associated risk signals for several decades. They include a number of hormonal factors, including an early menarche (start of periods and puberty), being older at first pregnancy, and a late menopause. All of these factors, as well as not breastfeeding, increase the number of periods and reproductive cycles a woman undergoes in a lifetime. The contraceptive pill (invented mainly by Carl Djerassi 60 years ago and ironically put into clinical trials by a staunch Catholic, John Rock) strangely has both good and bad effects for cancer.6 On the plus side, they reduce the total lifetime number of normal periods a woman has and mimic pregnancy, which is protective. This ‘natural’ effect was why John Rock thought it would please the pope.
However, the hormones, oestrogen and progesterone overstimulate growth of breast tissue, which is a risk factor. The two effects more or less cancel each other out, leading to only a small – 15 per cent – increased risk which luckily disappears quickly on stopping taking the pill. Natural oestrogen is important throughout life. Fat tissue stores oestrogen and obese women have a 30 per cent greater risk of breast cancer.7 Women who as newborns were chubbier than average also have an increased risk. This is probably due to higher levels of oestrogen in their mother’s blood. Those whose mothers suffered high blood pressure in pregnancy had reduced risks and low oestrogens.8 So changes to older childbirth, higher birth weights, reduced breastfeeding and rising obesity all influence early oestrogen levels and could explain some of the recent increases in the cancer.
As usual there is controversy as to the effects of dietary factors. The strongest evidence seems to be for fat intake: a meta-analysis of 45 studies reported that higher total fat intake increased breast-cancer risk by 13 per cent, while other large studies implicate saturated fats.9 Phyto-oestrogens – plant compounds that mimic oestrogen – have been extensively studied and marketed as a ‘natural’ menopause cure. The main type consumed in Western diets are lignans, found in seeds, broccoli and strawberries, which if you eat a lot could theoretically reduce your risk by 15 per cent.10 The other type of phyto-oestrogens are isoflavones, found in soybeans, common in Asian diets. As HRT has gone out of favour because of its breast-cancer risk, many women are taking unlicensed ‘alternative’ medications containing soya isoflavones, in particular genistein. Urged on by the media and marketing, these are perceived as ‘natural, safe and protective’ against cancer. The data show otherwise. When given to young animals genistein seems to protect against later development of breast cancer, but in older animals it will actually accelerate tumour growth.11 Clearly, as the actual doses taken are unknown in natural products, women who choose the tablet and not the plant are taking a risk.
The last few years have seen a revolution in our view of cancer. Till lately it was thought that single mutations in a ‘cancer’ gene usually started the process. This is unlikely. Epigenetics is now understood as the key to the cancer-forming process, by ensuring the survival of the abnormal cancer cells.
When researchers look at the whole DNA of cancer cells, they find that generally they are under-methylated – enabling many genes that are normally suppressed literally to run wild. This leads to a general instability of the chromosome that can lead to gene mutations and also to a breakdown in the imprinting system by which (in key growth genes) one copy is normally turned off. In parallel a few DNA areas show the opposite: they are hyper-methylated and the genes suppressed. These genes are the body’s built-in protection system, the tumour-suppressor genes that keep the DNA under control. One of these is BRCA1, which when the gene is faulty causes a highly genetic rare form of breast cancer in the majority of women who carry it.
A group from Johns Hopkins University in Baltimore have shown that, contrary to the idea that cancer was often random, there are amazing similarities in methylation regions among different cancer cell types, which suggest a common epigenetic mechanism ensuring these cells’ survival at our expense. Thus each group of cells is perfectly placed to divide and grow rapidly and have a flexible system of cellular evolution. This allows them to adapt to new surroundings when they spread elsewhere – ‘metastasise’ – and to fight off a lack of blood supply and chemotherapy drugs.12 To understand the epigenetics of cancer, let’s look at another pair of twins, and how they and their cancer cells fought each other for survival.
Fifty-five-year-old Heather and Judith were unlucky enough to be identical twins who both had breast cancer. Heather was diagnosed 20 years ago, and Judith ten years previously. ‘We’ve done everything together,’ said Judith, a retired teacher. ‘We had the same childhood illnesses, so in a strange way it does not surprise me that we both have the same illness now.’ After successful treatment they thought all was well, until Judith went on holiday to Bermuda last year and noticed some worrying signs. On returning she found out she had a recurrence. She had to tell her sister – who within a few weeks was told exactly the same thing.
Both twins went through the chemotherapy process and further tests and scans together in Manchester. Unfortunately Heather’s scans revealed extensive liver metastases. ‘The doctors said this was likely to be terminal.’ As Judith started to respond well to her treatment, the new but expensive drug Herceptin, her sister sadly did not, and deteriorated. ‘It’s the worst blow you could be dealt, but seeing Heather so happy for me makes me appreciate life and want to live every moment for her.’
Heather, a retired nurse, said: ‘It was wonderful news when she told me she was okay and in remission, even though I knew it would be tough for her to tell me. It’s strange, but it felt like a relief as much as anything, as I knew she would be there for my son when I’m gone.’
There is a history of breast cancer in their family. Their mother Sheila, now 81, was diagnosed with the disease 13 years ago and their aunt at 37, though both survived. But like the younger twins they did not have a rare cancer-gene mutation. ‘We are unlucky but feel philosophical – I’ve had a good twenty years of life since my first diagnosis and have seen my son grow up. I am grateful for every day I get with my wonderful family.’
Two years after her re-diagnosis and 22 years after the original tumour, Heather died in the hospice – the one the twins had been raising money for.
Breast-cancer response to treatment, as in these twins, is very unpredictable. Scientists are only just beginning to understand why. The drug Herceptin, which both the twins took, is effective only in blocking expression of the cancer-promoting (HER-2) gene which 20–25 per cent of tumours produce. In these cases it usually increases survival and remission. However, it has a price: it costs £20,000 for a single year’s supply in the UK and over double that in the US. As we saw with Heather, it doesn’t always work: the tumour mutates to avoid being killed as other abnormal genes are switched on or off. While it can be used with other anti-cancer drugs, it looks again as if epigenetics is crucial in giving cancer cells protection against chemotherapy and allowing them to metastasise. The Sloane Kettering Cancer Center in New York can now predict the early response to these drugs by looking at the methylation of tumour-suppressor genes in the biopsy.13 These profiles are likely to be used routinely in the future in clinical practice.
The fact that cancer cells use epigenetics to quickly adapt and mutate is also their Achilles heel. Because epigenetic changes are temporary they can be reversed. Epigenetic anti-cancer drugs have been rapidly produced, and there are at least five now available on the market in the US including Vidaza, Dacogen, Zolinza and Istodax.14 At least another 30 are in development. Although used so far mainly in leukaemia, they seem to be working well and are better targeted with fewer side effects than conventional anti-cancer drugs.
These drugs target and try to unblock (unmethylate) the body’s natural defence genes, allowing them to do the natural suppressing job that the cancer cells are preventing.15 Other clever mechanisms have been found called TET proteins, which add to methylated genes a marker (a hydroxyl group) that then signals for the whole gene to be repaired and reset in fresh unmethylated mode.
These are exciting new prospects,16 but the field is moving fast. Even more specific targets are also being developed. In breast cancer the tumour, in order to metastasise to bone, has first to epigenetically suppress the E-cadherin gene (responsible for cells sticking together), then once the tumour is in its new site it has to reactivate the gene so it can feed off the local blood supply. Drugs are being developed that block this, and so would stop tumours spreading.17 One German company has developed an FDA-approved screening test in blood for colon cancer targeting a single gene, SEPT6, that changes methylation.18 We have some encouraging early results with colleagues in Barcelona exploring 20 pairs of identical twins discordant (i.e. unlike each other) for breast cancer where we find some cancer genes are methylated in one twin and not the other.19 Several interesting genes are regularly emerging as clearly different in their methylation patterns. These are likely to be valuable biomarkers and possible drug targets.
If we go back to the accepted hormonal risk factors for cancer for a moment, could some of these risks be passed on across generations? Two studies suggest they could. The first comes from recent studies in pregnant rats fed oestrogen, which had daughters, granddaughters and great-granddaughters with more breast tumours.20 The other evidence comes from humans who took diethylstilboestrol (DES), a synthetic oestrogen medication that is famous for the wrong reasons. Human exposure to DES came partly from supplemented cattle feed but mainly from its medical use. From 1947 to 1971, DES was always a controversial drug. Without good evidence, it was given to pregnant women under the mistaken belief it would reduce the risk of pregnancy complications and miscarriages. In 1971, DES was shown to cause a rare vaginal tumour in girls and women who had been exposed to the drug in the womb. After much fighting and delay, it was eventually withdrawn.
For the women exposed to DES, commonly referred to as ‘DES daughters’, the problems didn’t end there. As well as getting these rare tumours in childhood they also had an increased risk of breast cancer 40 years later. The risk even increased for women over 50. Current research is now looking at the third generation, the grandchildren of women who were given DES during pregnancy. Results so far are worrying and show they have less regular menstrual periods and reduced fertility, with a possible increased risk of ovarian cancer. All this is likely to be due to epigenetic change across generations.21 Giving rodents DES prenatally produced the same breast cancers in the daughters as in humans, but here, interestingly, the timing is important. When given after birth DES has no effect, or may even be protective. Although not yet proven, this is highly likely to be via epigenetic modification of the oestrogen receptor genes at a critical stage. Like most epigenetic mechanisms there is considerable variation in the effects, so not all the DES daughters or granddaughters have problems.
There is a group of chemicals (called hormone disruptors) that resemble oestrogen and can mimic and influence the delicate hormone systems. They can be natural like soya beans or resveratrol from wine, or synthetic like DES or fungicides. These chemicals can affect any pathway controlled by hormones and produce defects of early development.22 We are slowly realising how many different ones there are (we know of at least a hundred) and how frequently they appear in the modern environment. The fact that they are so widespread and can have effects at very low dosage that can occur 50 years later or even in a different generation could be a major reason why they haven’t been spotted before. These hormone disruptors could be a crucial part of the puzzle not only of the increase recently in breast cancer, but also of other mysterious recent increases such as asthma, allergies, heart disease and even autism, which we will explore further.
Flo and Kay are identical twins with a special talent. They can recall the day of the week for any date in the last 100 years, or when given the song title can provide the artist, date and record label of every music recording of the 1960s, 70s and 80s. They can also remember the time and date of every major US event that happened in their lives from about the age of ten. This means for any particular date you name, they can recall the weather, recall what they each had to eat, and the tie and shirt colours of their favourite TV presenters. They are unique identical twin autistic savant sisters.
On the autism spectrum, Flo and Kay fall right around where Dustin Hoffman’s character in Rain Man does – and they sound just like his character. They have problems socialising, can’t cook for themselves, can’t drive and aren’t able to live independently – yet are in some aspects technically geniuses. Unlike many autistics, most of whom have no savant abilities, they have a quirky sense of humour and are passionate about music, which they like to see live. However, and classically, they avoid eye contact, handle emotions and communications poorly and have a number of obsessive behaviours. The twins are exceedingly rare, as they are female (four to five times more males than females are diagnosed with autism), they are prodigious savants (the highest level), and are strikingly similar – their abilities and disabilities appear as identical as the way they look. They have been inseparable since they were born, and still share a bedroom.
In 1996, when the popular long-running daytime US quiz show $100,000 Pyramid was cancelled, the two went through a major personal crisis. They had watched virtually every show and memorised every buzzer and bell pressed and remembered the colour patterns of the participants and host. Suddenly one of the things that endowed their life with some sense and order was lost. It was made worse when the host they obsessed about, Dick Clark – ‘who we call our personal saviour’ – suffered a stroke a few years later. The twins actually got to meet him before and after his stroke and he sends them Christmas cards.
They had a troubled early life, with bad squints, thick glasses and odd behaviour that they and their family were bullied and teased for. Their mother suffered from depression, and on one sad occasion tried to kill them using the gas cooker. She was stopped at the last minute by their father. Their parents, lacking a clear diagnosis, just felt they had odd children and kept them sheltered from the world, mainly from shame and embarrassment.
This went on until, when they were 27 years old, the family with the twins and their younger sister Jane (who had no symptoms) moved from New Jersey to sunny Florida. Their life changed dramatically when both their parents died within four months of arriving in Tampa. Luckily their younger sister was fond of them, and although she had a family of her own, continued to look after them in her house, allowing them much more freedom. Tragically, after only a few years she too suddenly died of a heart attack and they were on their own, confused and lost.
They moved back to New Jersey with their older brother and his family as a temporary measure, but they were having major problems coping, and a care home was the next option. As they were discussing all the deaths in their family and possibly even one day that of Dick Clark, they said: ‘We would like to be both buried in a casket surrounded by all the pictures, colour charts and Dick Clark memorabilia.’ The thought seemed to give them great happiness.
Dr Darold Treffert, a world expert in these syndromes, examined the twins a few years ago and confirmed that Flo and Kay had the savant syndrome, of which he knew of only 100 genuine cases and a handful of females. He speculated: ‘Until we can account for the savant, we can’t account for brain function overall. Until we can explain the savant, we really can’t explain ourselves.’23
Autism, along with its ‘milder’ cousin Asperger’s syndrome (grouped together with non-specific autism), is now best referred to as a group of disorders called Autism Spectrum Disorder (ASD), which can affect up to 1 in 100 of us. Until the 1990s it was viewed as a very rare childhood affliction that was caused by some combination of birth trauma, infections, child abuse, or poor parenting. The term ‘refrigerator mother’ was used to describe the common psychologist view that unemotional mothers caused the autistic symptoms of their infants and this was a child’s self-defence mechanism.24 In many Asian countries these ideas are still in vogue today, so that the commonness of ASD is largely denied and mothers blamed.25
Twin studies in the 1990s changed these perceptions dramatically when they clearly showed one of the strongest genetic influences of all diseases. The original estimates (based on small numbers) show autism to have a very high heritability of around 75–90 per cent, putting the refrigerator-mother theory firmly in its place. More recently, the largest study to date in 192 twin pairs with either autism or Asperger’s has suggested that, although still clearly heritable, this figure is likely lower – closer to 40 per cent.26 High rates of shared autism for identical twins of 60 per cent were found. Surprisingly, the study also found fairly high rates of sharing in non-identical twins, suggesting for the first time that there might be a shared environmental or womb factor, as studies have found much lower (10 to 20 per cent) sharing rates among brothers or sisters of children with autism.
Until recently reports of genes and ASD have been unconvincing, as studies have been small and speculative. However the use of genome-wide methods and large numbers (over 2,000 cases) has eventually produced results. A number of key genes are involved either with altered variations or containing duplications (called CNVs) that alter their function.27 Although still of small effect, they highlighted certain key brain pathways, particularly genes (cadherin) which encode molecules that allow neurones to interact, as well as a pathway (ubiquitin) which modifies proteins that influence neuronal connections and plasticity, or rarer mutations in a gene altering neural connections that also leads to mental retardation.28 These gene networks are involved in neuronal function, sensitivity of the synaptic nerve connections, as well as plasticity – the ability of cells to alter function.29 Some of the changes to these pathways are even reversible in animal models. Overall the gene pathways tell us that minor alterations in neuronal development are crucial in the growing brain. But how could this cause the very special characteristic symptoms of autism?
One of the latest and most elegant theories to explain the lack of empathy and ability to gauge emotions that is seen in autism comes from the eminent neurologist V. S. Ramachandran. Originally an Indian Tamil, he now works in San Diego. His idea derives from the recent discovery that the brain contains groups of highly specialised nerve cells called mirror neurones. These function as a virtual-reality network and are key to the way the brain develops by mimicry.
These mirror neurones can imagine actions and fire off when looking at someone else performing a precise action. The fundamental characteristic of these neurones was that they discharged both when the subject himself performed a certain motor act (e.g., grasping an object) and when the brain observed another individual performing the same act, thus internally rehearsing that act by mirroring the body’s own brain response in the same areas. Originally these neurones were shown to be firing in the prefrontal cortex of monkeys by Italian Giacomo Rizzolatti in 1992. The field was slow to accept their importance until the same neurones were shown in humans.30 Ramachandran and others found that one of the specific brain waves (called mu waves) of a normal person will be suppressed when watching someone else performing an action or expressing an emotion. Autistics however produce no such suppression of these brain waves, or any signs of emotion as tested by skin blood flow or functional MRI brain scans.31 This suggests that mirror-neurone deficiency (or broken mirror syndrome as Ramachandran called it) is due to disorganised development of the brain, occurring at an early stage.32 This coincides nicely with the evidence coming from the recent genetic studies.
The fact that most of us have brains that are so full of these mirrors suggests that many of our learning patterns and behaviours revolve around mimicry of others. This gives us ability both to read the emotions and intentions of other humans and to reflect on our own, which is perhaps the basis of human consciousness. We all know about how we mimic other people’s body postures subconsciously, depending on how much we like or fear them. Many salesmen now go on courses as part of their training so that they can improve their mimicry skills and sell us even more double-glazing. Autistics, lacking the mirror neurones, don’t have this ability to mimic body postures and are generally poor salesmen. In his latest book Ramachandran suggests that the great leap forward in human evolution happened around 60 to 70,000 years ago, after the brain had been at its current size for over 200,000 years without much happening. At that point: ‘By hyper-developing the mirror-neurone system, evolution in effect turned culture into the new genome. Armed with culture humans could rapidly adapt to new hostile environments …’ He suggests more speculatively that this change could also have been the origin of human language.33
Ramachandran and others have successfully treated many patients with phantom pain from amputated limbs with simple mirror therapy, enabling the brain to be tricked into rewiring internal pain signals possibly by increasing their mirror neurones. Work has started in autistic patients to see if their few existing mirror neurones can be increased and activated using synchronised dance therapy which trains the brain to mimic others.34 Currently around 20 per cent of neurones are believed to have this mirror function in humans and are concentrated in a unique and large area, the inferior parietal lobe, at the crossroads of visual, touch and hearing centres of the brain.
The recent optimism for treatment of autism comes from the discovery that the brain is much more flexible than previously believed. Just as we saw with the black-cab drivers whose brain areas altered at the expense of other areas, specialised areas can alter function and cells can alter specialisation through epigenetic change. After strokes or injuries, different brain areas can also take over moving different bits of the body, and visual areas in blind people can convert to hearing or smell. It is likely that the savant skills shown by Flo and Kay operate because other parts of their brain are being underused or the distracting signals flooding in from our normal senses are suppressed, so that they can focus exclusively on remembering trivial details or amazing mathematics. Other savants who may possess, say, great artistic skills when young tend to lose them when their autism improves and presumably they start to use other bits of the brain.
Margaret was looking forward to having twins and had an uneventful pregnancy until the end, when she developed the common complication of high blood pressure (pre-eclampsia) and had to rest. She eventually delivered identical twin boys at 40 weeks in February 1973. Kevin came out first, Shaun ten minutes later, and both looked reasonably healthy.
After a few days, Margaret’s happiness dissolved, to be replaced by anxiety. Shaun was not well and having problems feeding. He had also had a few very frightening grand mal epileptic seizures. The doctors didn’t know the reasons but advised keeping him in hospital for a few weeks until he stabilised. Six weeks later, after reuniting them at home, his parents Margaret and Patrick both noticed that Shaun wasn’t right, crying very loudly and strangely, whereas Kevin was quiet and placid.
The parents noticed that although the twins looked similar, there were clear differences in how they responded to their environments. By the age of two it was clear there was a permanent problem with Shaun. He couldn’t sit or hold things or respond to his name. He wouldn’t play or talk with his siblings or use any toys except his unusual favourite objects: plastic bottles filled with water, which he played with for hours. He was fascinated by particular music and records and the only word he spoke for many years was ‘Elvis’, relating to his favourite record. At the age of six he was given the diagnosis of ‘learning disability’ and then sent to a series of special schools. Only when he was 18 did his doctors use the label of autism.
His parents were stressed by Shaun, who took up most of their time and energy for the first few years, so they were relieved to see that their older son Mark and Kevin appeared normal, healthy, and played together. In retrospect all was not perfect with Kevin. After the age of five or six, ‘He didn’t use to look at people directly or greet them in the street, and if asked a question would reply in monosyllables,’ remembers Margaret. ‘He was also from an early age obsessed with tidiness and washing his hands many times each day.’ This was not seen as a problem in a busy family, but thereafter Kevin’s childhood, and school in particular, was not easy. He was bullied regularly, often because he wouldn’t play sports and would annoy other kids by turning away when a football was kicked at him. In one strange episode he was being picked on at school by a boy. When his dad Patrick went to see the parents he found out the ‘bullying’ kid was five years younger and half his size.
Kevin often got into trouble, usually because he misunderstood teachers and classmates. ‘From an early age I was obsessed with the TV science fiction programme Space: 1999 [a UK equivalent of Star Trek] and the lead female alien “Maya”, who could metamorphose, and wore out videotapes watching repeats. I was also fascinated by plane crashes and disasters and knew (and still know) the words to every Abba song’.
He gladly finished school and got a job selling TVs and electronics, and a few years later he bought his own corner shop. He was doing well until one day two masked men robbed him at knife point. This left him very traumatised and he closed the shop. He became depressed for several years and got other jobs, with relative success despite bullying by managers. After fighting with various GPs over his diagnosis and treatment, he consulted an autism support group and was finally referred to Professor Baron-Cohen in Cambridge, where he was formally diagnosed as having Asperger’s syndrome. For the last eight years Kevin (now 38) has coped well with his problems with the help of cognitive behavioural therapy. He has dedicated himself to raising money for autism spectrum disorders (ASDs), formed a local support group in Staffordshire, and wrote up his story to help finance it.35 ‘My diagnosis was the turning point in my life. I am a positive and mentally stronger person. I am confident enough to take the stage. I may be nervous – but I have to do this.’
Shaun and Kevin’s story is not an uncommon one among twins. They grew up at a time in the 1970s when the diagnosis was rare amid widespread ignorance. They also show clearly that their identical genes are not the only factors and can produce a wide variation in outcomes: from an intellectually impaired and totally dependent brother to another who is a high-functioning ASD with independence and a successful life. While it’s possible that some birth-trauma difference between the twins accounts for these brain differences, it is more likely that these changes in Shaun occurred before birth, causing brain damage and making the brain liable to seizures, which in turn can make things worse. The genes may have given then both a mild susceptibility, but something else in the environment had accelerated the process and altered the way the genes functioned at a crucial stage.
Of all the changes in human diseases, recent changes in autism rates have been the most dramatic, puzzling and controversial. Over the last three decades autism has increased tenfold, from 4 per 10,000 children in the 1960s to around 40 per 10,000 today in the US. More than 3,000 new cases of autism were reported in California in 2006, compared with only 205 in 1990.36 The numbers have continued to rise since then and are mirrored in most but not all countries. Although some of this increase (maybe half) is due to changes in diagnosis and medical attitudes, there appear to be real increases of 2–3 per cent per year.37 This rapid increase and the odd twin results can’t be explained by traditional genetics, so what is going on?
Previous research has implicated in a few cases anti-epileptic drugs, thalidomide and measles infections in mothers in early pregnancy, as well as a link with cerebral palsy. Up to a third of autistic children have had some form of temporal lobe epilepsy when young, which could certainly also upset brain development. Researchers recently found a higher risk of autism among children born to mothers who took SSRI antidepressants (like Prozac or Zoloft) during the year before birth, particularly in the first three months of pregnancy.38 But none of these exposures have increased dramatically enough recently to explain the present trends.
One popular theory was that the rises were related to increased childhood vaccinations since the 1960s. In 1998 The Lancet published a UK study implicating a triggering role of vaccination with the triple combined MMR (measles, mumps, rubella) vaccine. The paper, and media frenzy promoted by the authors, caused millions of parents worldwide to boycott the vaccine and led to a major recurrence of measles and a number of deaths. The paper was later retracted, due initially to the main author, gastroenterology surgeon Andrew Wakefield, having a major conflict of interest.39 He had been working as a paid consultant for the law firm trying to sue the vaccine manufacturers via a class action, and had acquired his 12 classical autism cases mainly through this route. Much larger population studies have since shown no association between use of the vaccine and ASD. Wakefield was subsequently investigated by the GMC and stuck off, with a suggestion of fraud. It appears on closer scrutiny by investigative journalists like Brian Deer of the Sunday Times that at least five of the 12 cases already had developmental problems before the vaccine.
This story was a tragedy all round. Not only for the normal kids who developed some severe complications of measles like blindness and mental problems, but also for the autistic families who had clung to the idea that this was the cause of the problem – which made it much easier to accept. Unfortunately it was just a coincidence: autistic symptoms happen to become apparent at much the same time that vaccinations are often given. But after this major red herring, are we any closer to finding a cause of the increase?
Another current theory proposes that the increased mating of nerdy men and brainy women explains reported high clusters of disease in areas of the US (Silicon Valley and Cambridge, Mass.) and UK (Cambridge) with hi-tech industries and computing. This is hard to prove and prone to bias, although it would go along with the extreme male brain theory originally proposed by Simon Baron-Cohen.40 The theory goes that ASD sufferers are at one end of a spectrum of male–female differences, whereby typical male (highly technical or systematised) brain activity is favoured over emotional or empathetic activity (typically female). As males with high systematised intelligence (sometimes called geeks) are now finding great financial success in biotech, computing, electronics and gaming, they have apparently become more attractive to some females. Indeed, if you are interested there are growing numbers of websites dedicated to women seeking nerdy husbands (such as www.Nerd Passions, Geek2Geek, IQcuties, Sweet on Geeks etc.).
The theory would suggest that whereas in the past these men would have been monks sitting alone on high wooden stools copying Latin texts and performing amazing calligraphy, now they are back in the gene pool. These matings are more likely to produce high-IQ males with ultra-male-orientated brains and at risk of ASD. The nerd mating theory is pretty much impossible to prove or disprove, although it might explain a few cases.
A better bet for a broader explanation of the increase in ASD is hormones. The extreme-male-brain theory has fostered ideas that levels of sex hormones in the amniotic sac of the womb –in particular testosterone, produced by both the fetus and mother – are quite variable and could influence the brain. A study in Cambridge has been following up 635 children for ten years having obtained amniotic fluid samples at birth. So far they have found good correlations between levels of high testosterone and mild features of ASD such as avoiding eye contact, reduced empathy and reduced communication skills. Contrary to the views of half the population, men actually have larger brains than women. Although size isn’t everything, men also have much larger amygdalas – paradoxically a key area in emotion and empathy. ASD sufferers have larger-than-average brains and amygdalas – again supporting the idea that ASD males have excessive male characteristics and that sex hormones play a role. It is also possible that the X chromosome has a protective role, as females have two copies of all these genes and males only one, allowing females more leeway if one copy of their genes gets modified epigenetically or rare mutations occur.41
There is now good evidence of an epigenetic influence in ASD.42 First, two common single gene imprinting disorders (which you will remember are extreme forms of epigenetics by which one copy of the gene is totally switched off) – fragile X syndrome and Rett syndrome – are both associated with autism and ASD. Second, a few ASD patients have gene duplications in a key area of chromosome 15 responsible for other similar disorders (Prader-Willi syndrome and Angelman syndrome). Finally, a small study of three discordant identical twin pairs with autism found small methylation differences in key candidate genes for ASD, and methylated suppression of a gene that reduces a protein in autistic brains.43
But if epigenetics is a mechanism by which genes could be modified, what unexplored external factors could be responsible? Some studies have implicated older parental and grandparental age at conception.44 As women (and men) are gradually increasing in age at the time of having children, this could cause genetic and epigenetic instability in sperm and eggs.45
Another study looked at the month of conception. The study included more than 6.6 million children born between 1990 and 2002 in California, of whom 19,238 were diagnosed with autism before the age of six. Children conceived in the winter season (December, January or February) had a 6 per cent greater risk compared with those conceived in July. Greater numbers of winter infections or vitamin deficiencies are obvious suspects.46 Infections like viruses have long been implicated in ASD, but the link has never been proven. Moreover childhood infections have actually reduced in recent years.
Vitamin levels before birth may have changed. Some nutritionists believe that the nutrition content of most fruit and vegetables we eat has decreased over the last 50 years. If true, this could account for a number of recent trends. Twenty-seven vegetables, 20 fruits and ten meats were compared over 50 years by careful surveys in the UK from the MRC and Ministry of Agriculture. They showed average falls of 48 per cent in calcium and 27 per cent in iron, with similar or greater losses for other minerals like copper or magnesium.47
A recent case-control study (not a proper trial) of 288 autistic children in California shows that use of prenatal vitamins around the time of conception was associated with a reduction in the risk of having children with autism.48 They found an overall 40 per cent reduced risk if mothers took vitamins. Some influence of these vitamins on gene methylation was also shown. A few women not taking any vitamins who also had certain gene variants influencing levels of methylation had a risk of autistic children twice to seven times as high.
There are probably more theories for the cause of autism than for any other disease, each with its own website and support group. Those mentioned here are just the mainstream ones. Another has appeared recently suggesting that paracetamol (acetaminophen) could be the deadly cause, as use has increased since the 1980s. This sounds crazy, but supporting it is the strange phenomenon that when autistics develop a fever their autistic symptoms often lessen or disappear, implicating the fever centre of the brain (the locus coeruleus). Artificially inducing fever has now been proposed as a therapy, but as you can imagine sounds medieval and will be tricky in practice.49
There is a nearly limitless number of theories and conflicting opinions in the field. Some sceptics are not yet convinced there is even a real increase in cases of autism. However, if we accept the majority opinion that it is increasing and agree that males are four to five times more at risk, then the clear link must be a hormonal influence of some sort on the developing brain. This would have to have changed rapidly in the last 30 years.
We have discussed vitamins, which might partly contribute, but what about chemicals? We talked already about some – called hormone disruptors – that alter hormones epigenetically. These include phyto-oestrogens from plants that can influence breast-cancer risk, cadmium in tobacco and low levels of arsenic. But there are others, that if you don’t live in Canada you may be less aware of. A year after Walter Brooke’s famous quote to Dustin Hoffman discussing his future in The Graduate – ‘I have only one word to say: “plastics”’ – the world changed, and became full of them. One plastic component under the microscope is called BPA or bisphenol A, found in most plastics used today since the 1950s,50 manufactured in vast quantities in the US and other large countries. It has mainly oestrogen-like properties and has been tested quite extensively over the last few years in rodents in realistic (i.e. not ridiculously high) doses.
Like other hormone disruptors, it has some cancer-inducing effects in hormone-related cancers, as in the prostate and breast. However, its role in the brain is most interesting. BPA when given as a supplement to our old friends the Agouti mice will alter their coat colour by epigenetically removing the blocking methylation signals on the gene (and therefore activating it), an effect that can then be reversed by giving methyl or folate vitamins.51 More worryingly, mothers’ exposure to BPA has also been shown to alter methylation in the fetal mouse forebrain. This produces changes in the neuronal synaptic connections, and in brain development as well as behaviour in the offspring.52 One of the clearest effects is increasing anxiety in pups and altering the mother–infant behaviours, making mother and pup more introverted and the mothers less caring. One study was performed on pregnant monkeys given BPA for the last 28 days.53 Infants of these BPA mothers were physically similar, but less attached physically and visually to their mothers. Also relevant to autism, BPA in the low doses that might be found in humans has been shown recently in male rodents to reduce two different learning skills in a maze test. It may do this by altering the hormone receptors in the hippocampus.54
Another study looked at mice offspring of mothers fed BPA in pregnancy. These were special male deer mice who in the wild need great visual–spatial skills to find and mate with females across long distances.55 The BPA mice looked identical, but were a bit geeky underneath and had poor female-hunting skills compared with normals. The females sensed this somehow, and even when the BPA mice did eventually locate them they found them less attractive. As well as giving BPA in pregnancy, studies have also looked at BPA in the male line where it affects sperm and promotes testicular disease and can be passed on epigenetically across at least three generations in rats.56 These are just the results for BPA. Similar results occur with other chemicals such as those found in flame retardants called BDE-4 – shown to alter rodent learning and induce behaviour inactivation57 – or with vinclozolin, the chemical used in many common fungicides, which reduced male fertility even in the fifth generation.58
So should we all be worried by plastic? After all, it is now pretty hard to avoid. Try it yourself – BPA is the type 3 and type 7 marked on recycling codes. Over 2 million tonnes are produced every year and it is the major hormone that leaks into our rivers and water supplies. When tested 95 per cent of us have detectable levels in our urine.
Although the risks in animals have been known for over five years, governments have been slow to react. Canada has as of September 2010 banned BPA as a toxic substance, and the EU has recently banned its use in baby bottles. In the US, despite a worrying report to the FDA in 2010, no action has yet been taken. Those with faith in government will rest assured because the director of the US Food Standards Agency and the UK Medical Research Council rated it ‘quite safe in humans’ as doses are low and it was all excreted. But it’s not just baby bottles: BPA is everywhere – sports equipment, CDs, DVDs, medical devices and tubes, recycled paper, dental fillings, even the inside of most canned drinks, and on plastics in contact with food. Worryingly, the highest levels (400-fold) so far recorded are in babies in intensive care units, surrounded by plastics, where it can be stored in fat tissue for years.59
In humans observational cross-sectional studies of BPA levels in the urine in 1,445 Americans measured at different time points from the NHANES study linked them to an increased risk of breast and heart disease,60 diabetes, altered liver function and many other slight changes in blood tests.61 The chemical triclosan, found in deodorants, is similar and in humans is associated with altered immune function.62 Although these kinds of observational studies can be misleading, they are worrying.63 Although the levels detected are within the safe limits suggested by US and UK governments, doses lower than that have been shown to affect rodent behaviours. But let’s not forget the other 104 known hormone disruptors, including pesticides, herbicides, fungicides, hydrocarbons and flame retardants, that could be having similar or unknown effects on us or our grandchildren. Food for thought as you swig from your plastic bottle of ultra-healthy mineral water – which when tested has oestrogenic properties in 61 per cent of samples, unlike glass bottles or tap water. So plastic bottles may be one reason why women’s breasts (and men’s) are getting larger.