The Fetal Matrix: Evolution, Development and Disease - part 9 doc

We have already described animal experiments that suggest that nutrition aroundthe time of conception has important inductive effects.11Only now are the neces-sarily very complex studies

201 Primary prevention

As women delay pregnancy longer, the risks of infertility grow Likewise, thedrastic fall in sperm count in Western males produces unwanted infertility even inmany younger couples Both trends have occurred at a time when assisted repro-ductive technologies have been broadly available This has led to a great increase

in multiple pregnancies and to the associated maternal constraint The availabledata also raise the probability that the abnormal environment of the pre-embryocultured in vitro may have long-term effects – it is too soon yet to know

While the evidence is still preliminary, there are increasing clues that for both

mother and father there are factors to consider prior to conception We have already

reviewed the evidence in chapter 6 that inappropriate PARs may have been triggered

in this period There may also be sperm effects as well as the egg effects that wehave described in some detail Exposure of the sperm during spermatogenesis toless than optimal conditions may influence the imprinting status of paternal genesafter fertilisation and thus produce epigenetic effects As we noted in chapter 5, datafrom Sweden reveal that the risk of Type 2 diabetes in men is determined in part

by the diet of their grandfathers in the period before they reached puberty So thechallenge may extend to both parents

We have already described animal experiments that suggest that nutrition aroundthe time of conception has important inductive effects.11Only now are the neces-sarily very complex studies being started, in which data on nutritional status at thebeginning of human pregnancy are collected and then related to the outcome ofpregnancy.12This is information we badly need Such data create a real challenge Itmay not be sufficient to focus on nutrition once the woman knows she is pregnant –much may have already happened before then that will have lifelong consequences

If that is the case we then need strategies to address how to improve the nutritionalstatus of women prior to conception – a difficult and culture-specific challenge

It is apparent that nutritional factors are the most important environmentalfactors We have seen that the nature of the impact of nutritional signals maydiffer at different times in pregnancy and that nutritional information is a majorpathway of signalling to the fetus It can reflect the actual status of the mother andher environment or it can be a false signal that arises from placental dysfunction

or maternal metabolic disease (e.g diabetes) What this suggests is that, maternaland placental disease apart, maternal nutrition may be by far the most important

11 Sheep that are undernourished in the time around conception have abnormal placental and fetal ment Indeed these fetuses appear to grow normally for the first 80 per cent of pregnancy then slow their growth Extrapolated to the human, if one had not known that the cause of the growth failure had occurred

develop-at the time of conception one would have assumed it to be a ldevelop-ate-pregnancy problem In the sheep study there was also a higher incidence of premature labour and there is a well-described relationship between being born small and being born premature In Indian women with a reduced body mass at the start of pregnancy there is a tendency to have shorter gestational lengths by about 10 days compared to European women with high body weights at the start of pregnancy.

12 The most comprehensive and detailed is the Southampton Women’s Survey.

component of any preventative strategy to reduce the probability of inappropriatePARs.13

But we face a knowledge crisis We actually know very little about what constitutesthe optimal nutrition for a woman at different stages before and during pregnancy

We have already discussed what little we know and perhaps should summarise itagain now The studies we have reviewed show clearly that maternal nutrition is notjust a question of adequate calories, unless the mother is under starvation condi-tions Assuming that the total caloric intake is adequate, the balance of carbohydrate

to protein and the source of that protein (dairy, meat or vegetable) have importantinfluences on optimal fetal growth Folate, an important vitamin regulating aminoacid metabolism, has an essential interaction with the amount of protein ingested.High maternal protein intakes without folate will lead to fetal growth retardation.The major source of folate is green leafy vegetables We also know that some criticalmicronutrients are important to the development of specific organs, e.g iodine forthe thyroid gland, calcium for bone Based on animal experiments we predict, but

do not know for sure, that many other micronutrients such as zinc, vitamin D andmembers of the vitamin B family are also important for optimal fetal development.While we know that the most likely focus for a preventative strategy will benutritional, we are also very cautious There are many examples of well-meant pre-ventative measures that, when implemented, did not reduce the size of the problembut in fact made it worse For example babies were put to bed on their bellies forapparently logical reasons during the mid-twentieth century and yet this caused

an increase, not the predicted decrease, in cot death In a sense the public healthemphasis on promoting big infants as healthy babies, and the consequent overfeed-ing of infants leading to obesity has contributed to the problem of inappropriatePARs These of course are strategies that assume a level of choice in the degree ofuptake by the population, and humans are by nature somewhat irrational and like

to take risks But they can also take advice too seriously and exacerbate problems byoverdoing things There are many health warnings about the dangers of excessiveuse of vitamin and mineral supplements, for example

A good deal of attention to date has been on folate because it plays a range ofmetabolic roles (assisting in DNA synthesis, breakdown of toxic substances such

as homocysteine and DNA methylation) and we have growing belief that suchmechanisms play a major role in the origin of PARs The Indian data and theMotherwell data both point to folate deficiency in pregnancy as a cause of inap-propriate PARs Folate supplementation at the beginning of pregnancy is known toreduce the risk of spina bifida But what is the appropriate amount and timing of

13 And the emerging information on the influences of maternal body composition and metabolism in initiating PARs are related to nutrition on a longer time scale.

203 Secondary prevention

folate administration? We do not know enough to be certain Preliminary ments show that the beneficial effects of folate supplementation in pregnancy onthe offspring of undernourished rat dams are in fact converted into deleteriouseffects on the offspring if well-nourished dams are given excessive dietary folatesupplementation during pregnancy One reason for this might be that such nor-mal offspring will become adapted to high folate levels before birth, make PARsaccordingly and may develop pathology unless they are able to access such highfolate levels postnatally

experi-So we are left with a gaping hole in our knowledge We simply do not knowthe optimal or normal nutritional profile either for a pre-pregnant or a pregnantwoman If we have to focus on the periconceptional period, then for most of theworld we must focus on optimising the nutritional status of adolescent females.For cultural and political reasons this is a real challenge in many societies, butparticularly so in developing societies Many studies have shown that the nutrition

of girls in developing countries is worse than that of boys, that they grow less inconsequence and that this is not necessarily rectified when they marry or commencechildbearing But even there we are uncertain whether the focus should just be onmacronutrients (fat, carbohydrate, protein) or whether there needs to be a greateremphasis on micronutrients such as folate, zinc and vitamins In other words, canmicronutrient supplementation correct dietary macronutrient imbalance? It mayeven be that the optimal nutritional balance is different prior to, in early and in latepregnancy

But some progress is being made Knowledge about treating the fetus as a tial patient is growing.14 Could fetal growth be enhanced or placental functionimproved in situations where it appears to be inadequate? There are data, partic-ularly in the sheep, suggesting that experimental hormonal therapies can enhanceplacental function and thus might promote fetal growth But we still do not knowwhether enhancing fetal growth in late pregnancy would reverse the impact of inap-propriate PAR events earlier in pregnancy This is a 64-billion-dollar question forwhich there are no data at present

predicted in utero, the risk would have been lower This is an extreme example andraises other ethical and philosophical issues – it is clearly unacceptable to maintain aFalasha child in Israel on near-famine rations just to stop him or her getting diabetes.But take the more common scenario The nutritional energy burdens we allow ourchildren to be exposed to are far higher than the nutritional status predicted by thefetus of any normal pregnancy and needed for optimal postnatal growth Pregnancy

is a constraining environment in which the amount of nutrients that reach the fetus

is normally limited by placental function So we can see that the consequences ofinappropriate PARs operate in both the developed and developing world, just atdifferent levels of postnatal nutrition – recall Figure 7.1 In the developed countriesthe philosophical issues of managing childhood nutrition to a more appropriatelevel are certainly challenging but the dilemmas are less

We suspect the time is approaching when the birth-size phenotype and thepregnancy history (supported by markers of the level of PARs, e.g the degree ofepigenetic change on specific genomic regions, the amount of omental fat, bloodhormonal levels etc.) may lead to individual-specific recommendations about theoptimal growth curve for a given child A neonate born weighing 3800 grams and

52 cm in length will have a different postnatal nutritional range from one weighing

3300 grams and 58 cm in length, even if both are born at the same gestational age

We need to use the available data, and collect more prospective data, to identifyoptimal growth curves for these two children – for example the latter may wellrequire a lower postnatal nutritional level to be optimally healthy These will giveenormous but important challenges to public health and nutrition scientists and

to pediatricians

Treating the inevitable

How should we approach intervention beyond the perinatal period? Obviously there

is the ‘ambulance at the bottom of the cliff’ approach, in which we treat hypertensionwith antihypertensive drugs and diabetes with insulin-sensitising drugs or insulinitself But after all that we have said above, couldn’t we intervene earlier? Thesimple answer is that we just do not know It would seem probable that if we canstop childhood obesity developing we would have an effective intervention in thecascade Whether this justifies the use of agents that aggressively reduce fat mass

in children is far from certain In the meantime, where can we turn for support?

We cannot count on the food industry for help, as many of the larger companieshave little interest in stemming the enormous flow of junk-food consumption byyoungsters worldwide To be fair to these companies, many of them merely providewhat their customers wish for, and it is sad that such fast foods (considered 30 yearsago almost a luxury item) are now among the cheapest forms of nutrition, offering

205 Treating the inevitable

enormous appeal in low socio-economic societies and to parents who are pressedfor time

Nor, unfortunately, can the pharmaceutical companies be expected to lend theirwealth and resources to the endeavour Their market models and the requirements

of regulating agencies mean that they just simply cannot afford to invest in maceutical approaches that have only a long-term benefit To be cynical, we wouldsay that the very reverse might be true: the prospect of as much as 25 per cent ofthe vast populations of the developing world suffering from Type 2 diabetes by theage of 30, and needing to be maintained on glycaemic-control drugs for decades,spells profits on an unheard of scale Moreover, the health effects of inappropriatePARs in the increasingly ageing populations of the developed world should fill thecoffers too We have to temper this cynicism once again by reminding ourselves thatthe pharmaceutical industry provides what is acceptable and usable We cannot atpresent see the prospect of a unique drug that would prevent the consequences ofinappropriate PARs, and even if such a drug were available, the ethical and practicalissues of using it would have to be resolved There have been too many examples

phar-of unforeseen side effects phar-of new drugs, especially if they are used during humandevelopment

These gaps in our knowledge are enormous The challenges for differentsocieties – those of the developed and those of the developing worlds – are verydifferent We have no doubt that if we knew how to address these challenges theburden of disease would fall dramatically

Fetal futures

We have seen how the theory of PARs helps us to understand the aetiology ofsome of the common chronic diseases of adulthood, especially components of themetabolic syndrome (Syndrome X), which include high blood pressure, Type 2diabetes, a disordered blood-lipid profile and clotting-factor levels, obesity andincreased risk of atheroma, coronary heart disease and stroke Such diseases have

a high prevalence in the developed world and are increasing at an alarming rate inpopulations in transition in the developing world The humanitarian and financialburden they convey is enormous The growing epidemic of obesity in young peoplefurther magnifies the problem

But what of other common chronic diseases – breast and prostate cancer, asthma,Alzheimer’s disease? Is it possible that the biological phenomenon of PARs couldunderlie the aetiology of such diseases? We can only speculate For each of these con-ditions there is some evidence but as yet it is either preliminary or unconfirmed, so

we felt that it would not be responsible to include it in our discussions at this stage.Not everything that happens in early life and has lifelong consequences is a result

of PARs Teratogenesis is an irreversible developmental disruption which can have

no predictive or adaptive value In addition, some responses the fetus must make tosurvive (e.g preterm delivery in the face of amniotic infection) must have inevitablecosts after The obesity of adults who had diabetic mothers may similarly just be

a consequence of increased fat mass laid down in fetal life by virtue of the highinsulin levels, or alternatively there may be a predictive element – we do not yetknow, but an experimental approach to answer this question is possible

It is important that those who think about PARs do so within the framework ofthe definition we laid down at the start of chapter 7

Predictive adaptive responses: the evolutionary perspective

Our primary goal has been to examine how the environment and genome interact

at critical points in early development, thus determining biological destiny and,206

207 PARs: the evolutionary perspective

in particular, having consequences for subsequent postnatal gene–environmentalinteractions Our theory is that the mechanisms underlying PARs have been pre-served through evolution because they provide a species with a way of survivingshort-term environmental challenges while still retaining maximal genotypic vari-ation Many environmental changes are transient, and because PARs can be reversedover a few generations, they permit phenotypic changes to follow these environmen-tal changes We saw in chapter 7 that while the Darwinian processes of natural andsexual selection provide a way of surviving an environmental challenge for somemembers of the species, these members will only be those with a genotypicallydetermined phenotype that confers advantageous characteristics; other memberswhose genotypes produce less advantageous characteristics will be more likely toperish, and their genetic information will not be passed on to future generations.Thus this component of the genetic variation of the species will be lost and maynever be recaptured unless a random mutation throws it up again at some futuretime Darwinian evolution is not only usually very incremental, but also permanent.Darwinian responses to an environmental challenge that turns out to be transientare extremely costly as they produce (probably irreversible) loss of a component ofthe gene pool

We have argued that PARs are a natural phenomenon whereby the developingconceptus makes some short-term choices to allow intrauterine survival, but that

it also makes a series of choices that will establish biochemical and physiologicalphenotypes intended to assist survival to adulthood and to promote species survivalthrough reproduction By preserving the maximal amount of variation in the genepool, PARs are likely to assist population survival by preserving genotypic variation

In one sense this aids the evolutionary process by sustaining genetic variation –the substrate for selection – and certainly this will allow a species to survive in

a variety of ecological niches In another sense it dilutes selection by preservinggenotypic variation rather than concentrating gene frequencies However, it is thecapacity to sustain a population through transient change as well as maximising theindividual’s chances of reproductive success that confers the overall advantage ofPARs

In most situations in most species, the predictions made by the fetus will becorrect and these mechanisms will be largely silent players In humans, the evolu-tionary echo is seen in the phenomena, driven by maternal constraint, that created adefault phenotype to protect against the historically most likely and risky scenario –

a transient lack of food Predictive adaptive responses become important when theenvironment changes because they allow the developing conceptus to makes choicesthat should turn out to be appropriate for the changed postnatal environment Butmisinformation from the mother or a rapidly changing environment can lead to the

in utero prediction of the postnatal environment being incorrect In animals this

usually means that the affected offspring dies, but in humans there are enormouslonger-term consequences.

The nature of the problem

Diseases are always easier to treat if we understand their causes The problem withPAR-related disease is that we do not know the underlying mechanisms in sufficientdetail There are several reasons for this

First, the phenomenon has only recently been described – the major impetus

to its study arose from epidemiological observations made in the last 15 years

or so It has taken time for the scientific community to catch up, and of courseduring this period there have been debates about the validity of the epidemiologicalobservations themselves

Second, it has taken considerable time to generate animal models to demonstratethat PARs are a biological phenomenon in other mammalian species, especially lab-oratory animals such as rats, mice and guinea pigs, but also in large animals such

as the sheep and pig This has been necessary not only to confirm the validity ofthe concept, but also to show that it occurs in species with very different degrees

of maturity at birth It has also been necessary to rule out purely genetic tions Now established, such animal models should speed the investigation of themechanisms underlying PARs and PAR-related disease

explana-But herein lies the third problem: the discipline of integrative biology, which

is needed to investigate such mechanisms, has been chronically under-funded fordecades and the expertise needed to make rapid progress in this area has been allbut lost As active researchers in this field, the authors know only too well how hard

it has been to get funding for research in the area and, once funded, how difficult

it is to recruit young scientists to it Biomedical research has taken an increasinglyreductionist approach, and this has been fuelled recently by the human genomeproject While we do not deny the importance of such work, its fashionable styleand the sheer resources it has commanded have drawn bright young scientists andfunding away from the integrative biology that will now be needed to put the genesback into the environment in which they operate Nowhere is this more true than

in the field of developmental origins of disease

The experiments necessary to advance our knowledge are complex and of verylong duration To prove experimentally aspects of the theoretical construct we haveput forward will require very difficult and careful experimental design Provingthe relationship we model in Figure 7.1 will be difficult but is feasible It is worththe effort in clarifying both our biological understanding and its application todisease prevention

209 What we do know

Defining the mechanisms requires good models and good clinical studies But

to plan intervention studies will require both expensive and lengthy experimentalstudies and clever epidemiological and clinical enquiry

Last of all, to make real progress in this field requires a marriage of disciplines,from theoretical biology to health policy and economics As we discussed in thepreface, there has been a large gulf between the theoretical and evolutionary biolo-gists on one hand and medical scientists on the other Humans are animals and weneed to get better at bridging this gap in approach

What we do know

Despite this somewhat gloomy prelude to the final chapter of this book, the fact isthat the knowledge we have in this field is far from negligible The phenomenon ofso-called ‘fetal origins of adult disease’ is now widely accepted,1as a result of theplethora of experimental, epidemiological and clinical studies conducted aroundthe world by many groups Particularly as a result of animal studies, we now havesome idea of the maternal processes that initiate PARs, primarily via the balance ofnutrition but also through body composition and hormonal status For hormonalstatus a range of stimuli loosely grouped under the heading ‘stress’ – emotional,environmental and nutritional – may be included, and it is possible that changes

in hormones such as the steroid cortisol in the maternal (and hence potentially thefetal) bloodstream are of key importance

The explosion of knowledge about genes and cells, introduced in the first twochapters, has fuelled a rapid expansion of knowledge in developmental and peri-natal biology This has helped us to improve our knowledge of the very early lifecomponents of PARs more than those that operate in late gestation It is clear that theinduction of a PAR can occur in the periconceptional period, before implantation

of the blastocyst This suggests that induction involves changes in the expression ofgenes that control this early development, and we have now progressed to the pointwhere we have some good candidates for such genes, especially those that can beimprinted, and that regulate the growth and development of the early embryo andthe placenta This in turn has led to investigation of epigenetic changes induced byenvironmental stimuli in such responses Because DNA methylation depends onthe provision of methyl groups from amino acids such as glycine and on cofactorssuch as folic acid and vitamin B12, the links between such early embryonic PARprocesses and maternal nutritional balance can be drawn But there is much yet to

1 These days usually referred to as developmental origins of health and disease (DOHaD), for which an international learned society now exists.

understand about the scope of these changes, of the environmental cues and theimpact on cell growth and differentiation.

Then we will start to get an idea of the range of disorders that might be caused byinappropriate PARs There are many gaps – understanding the effects on the nervoussystem, on the origins of obesity, and on immune function and the susceptibility

to infection – would appear to be priorities This cannot be done in isolation fromunderstanding the window in which environmental cues might act and the nature ofthe cues involved Even with something apparently as fundamental as nutrition, westill do not really know how to feed the mother optimally and what guidance to giveabout nutrition for children of different birth phenotypes in different populations.Research in this area must be given a high priority

PARs and the genome

In chapter 1 we pointed out that every phenotypic response depends on interactionbetween the genome and the environment This is clearly the case with respect tothe special set of interactions we have termed PARs

What PARs do is to set up a two-stage interaction An irreversible choice is firstmade in fetal life as a result of an environmentally cued prediction and that irre-versibility is reflected, at least in part, in epigenetic change That choice (e.g ingrowth trajectory), once established, becomes the framework on which the post-natal interactions between the organism and its environment occurs Depending

on the choice made in early development, the consequences of that second action will be different Because ultimately all such interactions involve changes ingene expression, one would expect there to be genetically determined differences

in the nature of these responses, both in the primary and in the secondary actions We are starting to find out that this is indeed the case Polymorphismsinfluence the relationship between the fetal environment and long-term conse-quences2, and over time we expect to identify many more polymorphisms influ-encing both reactions

inter-Many clinicians and clinical researchers have fallen into the trap of thinking aboutfamilial disease either in terms of purely genetic considerations (on the model of

a monogenic inherited disease such as cystic fibrosis) or in terms of clustered andcommon environmental factors (e.g all members of a family being exposed tocigarette smoke) There is a third pathway which is well recognised in plant andcomparative biology but has received little attention in human medicine – that ofmaternal effects or epigenetic inheritance Studies arising from PARs research show

2 Good examples of this were given in chapter 4, where we showed how the risk of Type 2 diabetes associated with low birth weight was greater in individuals with a PPARγ polymorphism.

211 Preventing PAR-related disease

that this phenomenon is relevant to human disease We described both animalstudies of dietary-induced PARs and human population studies (e.g the Dutchfamine) that demonstrate its relevance The extent to which familial clustering ofdiseases such as Type 2 diabetes mellitus is epigenetic and PARs-related rather thangenetic therefore needs to be given much more attention

Preventing PAR-related disease

In chapter 9 we examined how this developing knowledge might be applied toreduce the burden of disease We are reluctant to make sweeping statements thatextend beyond the available data First, with the exception of one estimate based onone population (Finland), there are no data available on which we can predict thesize of the impact of successful intervention – we suspect it to be large Second, as

we have been careful to point out, societies by definition have very different geneticand environmental histories, and therefore the nature of any intervention and themagnitude of the resulting effect may be very population-specific

From first principles we can identify two preventative approaches – primary vention by optimising the fetal environment, and secondary prevention by adjustingthe postnatal environment to make it closer to that which the fetus predicted But

pre-we are immediately faced with difficulties One is that pre-we simply do not know theoptimal nutrient delivery for an embryo/fetus at different points in development.Nor do we have good tools for identifying the level of nutrition a fetus is receiving

We certainly do not know how to intervene in populations to optimise nutritionprior to conception We do not have markers of when a neonate has made a PAR and

we have yet to understand the optimal range of childhood growth for a given birthphenotype These are all good research questions and there is an understandableurge to leap ahead

But at this stage we can but be cautious It would appear to us that a balancedand micronutrient-adequate nutritional intake starting well before conception isessential for optimal fetal outcome So is optimal maternal age, avoiding smoking,minimising the risk of disease and perhaps moderating maternal exercise We stillhave insufficient data about placental disease and its management and even aboutgestational diabetes And of course we must remember that some factors can never

be avoided, e.g primiparous pregnancy

It would also seem to us that excessive catch-up growth and excessive weightgain in infancy and childhood must be avoided Much more work is needed tounderstand optimal nutritional regimes for childhood growth and these must beindividualised to the birth phenotype The challenging possibility remains thatchildhood obesity constitutes such a risk factor that it merits therapeutic interven-tion with agents that will inhibit fat deposition – a research question that should not

be ignored We do not know whether epigenetic change can be specifically reversed,

or indeed the consequences of doing so

We have focused the discussion in this book largely on the metabolic syndromeand its components At its conclusion we should note that we suspect a muchbroader set of PARs might be associated with altered disease risk The study of theJapanese soldiers,3 for example, illustrates PARs operating in a different physio-logical framework

The scientific agenda

As we hope will be clear from this book, the theory of PARs has considerableimplications for science This applies not only to developmental biology and fetalphysiology, but also to epidemiology, clinical research and evolutionary biology.Progress in research will be much faster with greater collaboration between thesefields, and happily such wide-ranging collaborations are now becoming more com-mon in science Achieving funding for them is of course vital We have alluded tothis earlier, but here would make two points The first is that such research reallydoes need to break the mould of current concepts about collaborations It will have

to be based on a true exchange of staff, ideas and expertise, not just on Laboratory

A doing some experiments and collecting some samples which are of interest tothose in Laboratory B, who proceed to analyse them and publish the results as adissociated exercise This gives no added value to the endeavour

The second issue is that the work now needed involves conducting very term studies, whether in the following up of children born to mothers in whompre-pregnancy characteristics have been measured in detail, or in establishing howexperimental diets in pregnancy influence cardiovascular function and responses

long-to stress in subsequent generations of experimental animals There is a matrix ofquestions that need to be answered and there is a need to develop a coordinatedapproach to filling in this matrix This requires a ‘big science’ approach as used inhigh - energy physics or in the human genome project, rather than ad hoc individualisolated research efforts where key questions are left unanswered The researchparadigm is changing and research agencies need to adjust their approach to fit it

or the public health applications of potential collaborations will never be realised

A final commentary

Understanding PARs changes our view of prenatal development and health Logicwould suggest that a greater emphasis on the well-being of women of reproductive

3 See chapter 1.

213 A final commentary

age, even before pregnancy, must be made in medical research, in healthcare delivery,

in economic policy and in the political process It is no longer possible to see theembryo or fetus as the larval stage of human development, not needing particularcare or attention because it will be nourished, nurtured and defended from a hostileenvironment by its mother Instead it is now apparent that by taking a developmentalperspective, radical changes in priorities are demanded that will impact on manycomponents of our lives

We believe that this has implications both for individuals, be they parents orpoliticians, and for society It has immediate implications for how potential parentsmake choices about their lifestyle, health care and rearing of their children In soci-ety, it must influence scientific, medical, philosophical and economic thought This

in turn will have implications for decisions about priorities made by ers and politicians, whether with respect to fiscal allocations to medical research,healthcare delivery, education, or aid to the developing world

policymak-We now know that even periconceptional events can have an impact that extendsthrough and beyond fetal life and early childhood While this has most easily beendemonstrated in the animal experiments we have discussed, the careful observa-tions on the offspring of those exposed to famine in the Dutch Hunger Winterdemonstrated that the biology is equally applicable to the human The implications

of this to human healthcare and to preventative medicine are immense It should

be obvious that it is critical to focus on the health of the developing adolescent and

to promote planned pregnancy within an optimal window of maternal age nal care must start before pregnancy In both developed and developing societiesachieving this goal will challenge prevailing attitudes

Mater-This discussion has implications for how people live and the choices they make.Many of these choices are those of the mother The modern lifestyle creates atti-tudes and a self-centered focus that is not necessarily in accord with an optimaloutcome for a pregnancy The fad for thinness, the tendency to delay first preg-nancy and advance maternal age in the Western world, the converse prevalence

of adolescent pregnancy in many populations, the use of tobacco, of recreationaldrugs, are all relevant These are all factors that impact on prenatal development.Most people, and particularly our youngsters, do not want to think too muchabout longer-term implications of their behaviour We know that simply statingthat the body composition and diet of young women may affect the health of theirchildren in middle age is unlikely to provide much of an incentive for them tochange their lifestyle We have to find new ways of educating them, to make theircurrent health status itself fashionable and to focus on the short-term benefits tothem

Just as those controlling medical research need to take greater account of theissues we have addressed, so too must those determining healthcare policy In the