Fat, Fate, and Disease : Why we are losing the war against obesity and chronic disease (21 page)

However, the most impressive data on these ideas have come from long-term prospective studies of Western populations, conducted in Southampton. These were originally conceived by David Barker’s team and are now run by Hazel Inskip and Keith Godfrey. Their originality lies in the fact that they start making measurements of women, their partners, and often their parents too, before the women become pregnant. This is a major undertaking because, of course, not all the women do become pregnant. In one of their studies, 12,500 women were recruited and studied in detail and the 3,150 who became pregnant were then followed up.

The Southampton studies have shown clearly that birth weight is not the most important factor in setting the risk of chronic disease. The thickness of the carotid artery of a child at nine years of age, an early and highly objective marker of risk of cardiovascular disease, was statistically related to low carbohydrate intake by the mother in late pregnancy, and this effect was independent of the child’s birth weight. In other words, the fetus can change its biology on the basis of its mother’s diet without necessarily changing its body size overall.

If this phenomenon is important in the epidemic of diabetes and cardiovascular disease, we need to explain why so many apparently unremarkable pregnancies are associated with getting the forecast wrong. If we are right, many fetuses are mismatched to the world because they predict a poor nutritional environment but are in reality born into a world of plenty, a world they did not really expect.

What a mother eats or how stressed she is need not necessarily be transmitted accurately all the time to her fetus. After all, the fetus is
not directly connected to its mother; there is an intermediate organ—the placenta. And so the transduction of signals about the world out there is dampened and can be distorted. It is like trying to see the world with very primitive radar rather than with your own eyes. You will get an impression, but some things will be missed and others may be misleading. But just as radar, even in its most primitive form, was of value to the air forces and navies of the Second World War, so the information that the fetus can glean from its mother can be valuable to it.

There are many ways in which the information can be misleading. One possibility is that the mother is unwell and is consuming nutritional supplies for herself, or she may have high stress hormone levels.

Another possibility is that the placenta is not working well and thus nutrient transfer is diminished even though the mother has ample supplies. These situations can lead to either premature birth or a reduction in fetal growth, which in turn leads to greater risk of obesity and chronic disease. That might add to risk in perhaps 5–10 per cent of the population but, as the incidence of diabetes and cardiovascular disease is much higher than this, there must be more to the problem than just maternal or placental disease.

One of the many things measured in the Southampton Women’s Survey was the food intake of women before conception. What is worrying is that many women do not eat a balanced diet to support their pregnancy. Indeed more that 50 per cent of the women of lowest educational achievement had a diet that was unbalanced. Among those with a university degree it was about 3 per cent. These are very revealing data—they tell us that unbalanced nutrition from the fetal perspective is not just a concern in the least developed world; it is an issue in Western societies as well.

Sometimes the mother’s behaviour can also affect birth weight. Over the last 30 years in Japan birth weight has been falling, as we discussed in
Chapter 5
. One reason for this is that more women
now smoke; another is that they have drastically reduced the weight they put on during pregnancy, from about 12 kg to about 8 kg. In Japan, until recently obstetricians recommended only a low level of weight gain in pregnancy even in women who are already very thin. Because of the possible long-term health effects on the next generation produced by such a mismatch, and because of the problems of maternal obesity (which we will investigate in the next chapter), major medical authorities and researchers are now revising their recommendations for weight gain during pregnancy. In general, except under conditions of gross maternal obesity, we hope that all women would gain at least 10–12 kg during their pregnancy.

But there may be more fundamental reasons why nearly all of us now get the forecast wrong to a certain degree. One is that many of us are first-born. Some years ago we theorized that first-born children would be more at risk of becoming obese, because the nutritional and physical environment of the uterus is more constrained in first than in subsequent pregnancies. So first children would be more likely to predict a nutritionally limited world and thus be mismatched in a modern energy-dense world. This was primarily an evolutionary argument but, if we were right, then it could be playing a major role in the current chronic disease epidemic. More of the world’s population is now first-born because, as developing societies go through socio-economic transition, family size has fallen. As family size falls the proportion of those who are first-born inevitably rises. In Singapore, for example, the average number of children in a family is now less than 1.5 per family. In Europe, where the number of children has fallen on average well below two per family, more than half the children are first-born. In urban China this has reached almost 100 per cent through the one-child policy and this, if we are right,
might have something to do with explaining the rapid increase of diabetes of China.

Recently our idea has been put to the test. In the early 1950s, a doctor in Motherwell in southern Scotland made substantial recommendations about the diet that his women patients should eat during pregnancy. The offspring were studied until they were 30. Keith Godfrey and his team in Southampton analysed how fat these 30-year-olds were from the point of view of whether they were firstborn or not. As we predicted, those who were first-born have about 25 per cent more body fat than those who were second or subsequent children. Most recently, data from Cesar Victora’s group in Brazil have shown that first-born children are more likely to have higher blood pressure later.

How do such effects come about? They relate to a phenomenon known as ‘maternal constraint’. To understand it we need to think back to our evolutionary past. We represent the result of a compromise. We evolved from a quadrupedal into a bipedal primate—in other words, into walking upright—somewhere in East Africa about 5 million years ago when we left the forests and entered the savannah. We can only speculate about the reasons why our distant ancestors started to walk upright. It may have been that it conferred advantages in a savannah setting where it was possible to see further and to run around. Indeed, the human lineage has the only primates that can run for very long distances—as every charity knows that encourages us to run marathons. We cannot, of course, run as fast as the big cats or other predators which might have attacked us, or as fast as the antelopes or other large mammals which might have been our prey. But because we can run for long distances we are able to run these animals to exhaustion if we work as a team, rather like in a relay race. This may have been one of the critical advantages in allowing us to survive in the savannah environment. But running requires a narrower pelvis, otherwise we would just waddle. So the human pelvis evolved to be narrower than that of other primates.

And there are also other problems with standing on two legs. In non-human primates such as gorillas, and even our closest cousins the chimpanzees, the abdominal contents hang downwards as they walk on all fours. This is not inconvenient and does not compromise digestion. However, in the upright position the intestines will tend to descend into the lower abdomen and be squashed into the pelvis. If the muscles of the pelvic floor are not to be used continuously to support them, then the shape as well as the size of the pelvis needs to be modified. So the width of the pelvic canal narrowed substantially, over thousands of generations of selection. The angle of the pelvic girdle, to which our legs are attached, has also rotated—we can see the difference very easily if we look at the way that the legs of a dog or a cat, or for that matter a cow or a horse, are attached. It is very hard for them to walk on only their hind legs.

But now consider the implications of this during the birth process. The fetus must pass through the pelvic canal from the mother’s womb during birth. And as the human pelvic canal is significantly narrower than that of the chimpanzee, our nearest cousin, the birth process is much more difficult. And it’s harder still because humans have evolved to have much larger brains, and therefore much larger skulls to hold them, than any of our primate relatives. This rapid brain growth starts before birth, and is not completed until after birth—otherwise we could never be born through that narrow pelvic canal at all. Our brains are much more immature at birth compared to our monkey cousins—we are born comparatively helpless and unable to move around.

For the large head of the human baby to pass through the narrow pelvic canal is challenging. Indeed, that is why humans are the only animals for which some form of birth assistance is generally required during the natural process of giving birth. Perhaps it is no surprise that in the absence of modern medical care obstructed labour, where the fetus literally becomes stuck and delivery cannot progress, is, sadly, a relatively common problem. In some undeveloped societies up to 60 in every 1,000 births encounter this problem.

We can see the extent of this problem if we return to our ancestral homeland in the Rift Valley. In the highland plains of what is now Ethiopia, women of the Hansa or Mursi tribes are only too aware of the problem of obstructed labour and its disastrous consequences. The tremendous contractions of the uterus during labour, in order to force the fetus through the pelvic canal, can tear the women’s internal organs and produce a rupture—a fistula—which may rip her urinary tract or her rectum. Such labours can go on not just for many hours but for many days, and when this happens very often the fetus dies. Often the mother will die as well.

It may take the woman several days to reach hospital and she will probably not start the journey until obstructed labour is clearly a problem. Despite the agony of the labour and the wrenching obstruction, she may have to walk for more than a day to reach the nearest road, and then possibly travel on a truck for another day to reach the nearest city. By this time the fetus is almost certainly dead, and all that the doctors can do is deliver it and try to repair the woman’s reproductive tract. Up and down Ethiopia, hospitals have been set up—initially by missionaries like Catherine and Reginald Hamlin in the 1960s—and since the establishment of these hospitals, 35,000 women have been treated for reproductive tract fistulas. Many more never made it to hospital—and consequently, they suffer from urinary or faecal incontinence and often live in shame and social isolation.

We have to note here that this problem is made worse by the practice of female genital mutilation. Some groups, in Ethiopia and other parts of the world, demand female circumcision—something we find abhorrent in the West. The scars themselves narrow the opening of the vagina but frequently it is also sewn up, leaving only a very small opening. If this ligature is not removed before labour, the consequences can be dire.

Under rural conditions there is not very much that can be done about the problem of obstructed labour, although having a friend or relative to assist during the delivery process can help. There are also,
of course, many folk remedies and other fables about what a woman might do to reduce the risk of obstructed labour, usually based on the idea that having a smaller baby will help. Undoubtedly this is true—but, as we have seen, the fetus has a surprising amount of control over its own growth. So Ethiopian women who refuse to drink milk during pregnancy, perhaps to reduce their calcium intake and so reduce the growth of the fetal skeleton, may not actually reduce their risk of obstructed labour, even though they will increase the risk of rickets in their offspring.

Obstructed labour would have been a disaster in our Palaeolithic past. We can imagine that it would have been common for big males to mate relatively juvenile girls. Many of the girls who had a fast-growing male fetus would have died in obstructive labour. So now might evolution have helped to protect us against this problem?

It turns out that the growth of every fetus is, to a certain extent, constrained by its mother; this helps to match its size at birth to the size of her pelvis. This means that the mother’s physiology and metabolism will control the growth of the fetus rather more than its genes, which of course have been partly inherited from the father. There could be a major conflict if the genetic component from a tall male had combined with that of a small female to produce the next generation. A pregnancy involving an especially large fetus developing in the womb of a relatively small woman would be far more likely to end in disaster. This is where the maternal constraint of growth exerts its most powerful effects.

Humans are not the only animals in which this process has been observed. The first demonstration was published in 1938 by Arthur Walton and John Hammond. These researchers crossed small Shetland ponies with the large shire horse breed traditionally used to pull the plough, and then looked at the size of the foals born. When the mare was a Shetland pony and the sire was a shire horse, the size of the foal was much closer to that of a Shetland foal. Conversely, when the mare was a shire horse, the foal was nearly as large as a shire foal,
even if the father was a Shetland pony. Needless to say, these experiments were conducted using artificial insemination.

When we look back at Walton and Hammond’s original experiments, it seems surprising to modern scientists that they were ever published, as the number of animals used was relatively small even if the results were decisive. However, the experiments have recently been repeated in Cambridge by Twink Allen and Abigail Fowden using embryo transfer, and once again it is clear that these powerful processes operate in horses.