Evolutionary Medicine
Miriam Nývltová Fišáková
Department of Physiology
The Modern Nutrition Transition
We see then that the advent of agriculture should not be equated with dietary security or abundance; in fact, it was often associated with the
opposite. Agricultural populations were often more prone to shortfalls in both energy and specific micronutrients than were their foraging
ancestors. The nutritional abundance that is now the hallmark of many modern agricultural societies, and which many of us take for granted, is
in fact a far more recent—and in many regions, ongoing—development. It is a product of a set of changes in diet and lifestyle that have been
described as the nutrition transition (see Popkin 2001). In contrast food security remains a critical issue for many lower-income countries with
rapidly growing populations. There are many factors that contribute to the nutrition transition. One relates to the cheaper production of
refined carbohydrates and fats, in particular fats of plant origin. Government subsidies for certain crops such as corn, introduced for political
reasons, further incentivize farmers and food producers. Greater trans-national trade, the development of intensive industrial farming
practices that reduce the cost of crop production, and the burgeoning market of populations moving from rural to urban environments all
make a contribution. Economic considerations show how the processing and ultra-processing of foods increase the value (and thus the retail
price) of foods as well as their preservation. Additional profits lead food companies to focus increasingly on such production rather than on the
provision of minimally processed, usually less energy dense and healthier, foods. Advertising and communication also play a part in making
these products desirable: for evidence we need only look to the success of fast food chains and of the marketing of sweetened beverages. But
we must also recognize that these trends occurred against the backdrop of human societies who were willing to make the transition. It is
unfortunate, for example, that the traditional scarcity of fats for cooking in poor communities has encouraged overconsumption of these items
as their cost has declined and their availability increased. In many such populations, as income rises so does the intake of additional high-fat
foods and sweetened beverages.
• China provides a good example of this transition. During the 1970s, in the
period after the Cultural Revolution, food insecurity was a major problem. There
was no television, limited bus and other mass transportation, and little food
trade with other countries. Very little processed food existed, and most rural and
urban occupations were very labor-intensive: oxen were used for plowing and
factories still required huge amounts of human labor to move stock and
equipment about. However, work and life in China have changed dramatically in
recent decades. Tractors are now available on many farms and fork-lift trucks in
factories; the internet has arrived in offices along with printers, fax machines,
and modern telephone systems. By 2000, soft drinks and processed foods were
consumed everywhere and nearly 90% of all homes had a television. Hong Kongand
Western-based advertising have been on the increase and may be received
on many televisions as well as seen on billboards and in magazines. Use of the
bicycle has declined and public transport and cars increased. In the course of one
generation, the lives of millions of Chinese people were transformed from a
subsistence agriculture-based economy to a modern, industrialized one. Data on
the incidence of obesity in Chinese children demonstrates the consequences of
this. The overall prevalence of overweight and obesity in childhood increased
from 2.2% in 1981–5 to 20.6% in 2006–10 (see Yu et al. 2012). The increase
largely occurred in urban areas, in which children were nearly twice as likely to
be obese compared with their rural counterparts. Similar changes are also seen
in other low- to middle-income countries, for example in Oceania, Latin America,
elsewhere in Asia, and parts of the Middle East (see Ng et al. 2014). Refugees or
migrants who were born in conditions of poor nutrition but who then moved to
nutritionally abundant environments are likewise at greater risk of diabetes,
obesity, and cardiovascular disease.
Well Fed but
Poorly
Nourished
The foraging populations we describe were generally considered to be free from chronic diseases like diabetes and obesity until the adoption
of more Westernized dietary and lifestyle practices in recent generations. In fact, the relatively high meat consumption of foragers may cause
some readers to pause for thought. The amount of animal products in the diet of foragers appears to far exceed the intake in most Western
industrialized nations. In the USA, where typical per-capita food consumption is of the order of 2700 kcal/day, 15% of energy is from meat,
10% from dairy, and the remaining 75% from flour, sweeteners, oils, fruits, vegetables, and other items. Americans appear to consume far
fewer calories from meat than most foragers. If excessive meat consumption is indeed bad for our health, how then do we explain the rise of
chronic metabolic disease in societies like the USA and the foragers’ comparatively disease-free existence? One important point is that not all
meat is alike. Wild meat of the sort that foragers consume is lean (protein dense) and typically provides about half as much of its energy from
fat as does meat from grain- fed domestic cattle.
• Through time, human herding cultures have
selectively bred the animals that produce the
fattiest, tastiest meat. Then in the post- war
period, as meat has been marketed to a consumer
society, the trend towards taste and textural
qualities has rapidly grown. This has maximized
factors that improve flavor, such as marbling,
which involves the deposition of fat droplets
(triglycerides) within and between the muscle
cells. In addition, industrial herding practices in
places like the USA often involve feeding animals
grain, such as corn, which is grown in abundance
with generous federal subsidies, rather than the
grasses that are their natural food resource.
Compared with grass- fed cattle, grain- fed beef is
higher in fat and has more saturated and
monounsaturated fats, and is also very low in
beneficial polyunsaturated fats. As a result of
these changes, what many of us call meat today is
very different from the meat that has probably
been an important component of the hominin
diet for several million years.
• These changes have consequences for our health. Not
only are we eating more calories as a result of the greater
energy density of domesticated meats, but we now
consume higher levels of the saturated fatty acids that
elevate cholesterol, which is implicated in cardiovascular
disease. In addition to the well- known deleterious effects of
an excessive intake of saturated fatty acids, recent work is
revealing with higher resolution the importance of specific
fatty acids, and especially the ratios of fatty acids, as key to
conditions such as atherosclerosis. The omega- 3 fatty acids
are particularly important. These are not produced within
the body to a great extent and thus are essential. Omega- 3
fatty acids have been shown to have a wide range of effects
such as reducing clotting and inflammation. Omega- 6 fatty
acids, which are also essential, can tip the balance in the
other direction, encouraging inflammation. It appears that
the effects of these fatty acids on health depend critically on
their balance within the body, and that inappropriately high
omega- 6/ omega- 3 ratios in the diet can be harmful. It is
notable, then, that grain- fed beef is significantly lower in
omega- 3 fatty acids than grass- fed beef (see Daley et al.
2010). The ratio of omega- 6 to omega- 3 fatty acids has
increased substantially in human breast milk over the past
20 years, demonstrating the subtleties of how influences
from these changing food behaviors could affect our
biology. The public health push to reduce consumption of
meat and meat products in many countries has encouraged
consumption of vegetable oils. Vegetable oils in general
tend to be higher in monounsaturated and polyunsaturated
fats, which have more favorable effects on risk factors such
as cholesterol profiles when compared with the saturated
fats present in meat or poultry. But some commonly used
vegetable oils have fatty acid profiles and other properties
that are also not beneficial. For instance, soybean and corn
oils, which are mass produced and widely utilized in
processed foods, have very high omega- 6/ omega- 3 ratios.
Palm and coconut oil, the most widely used vegetable oils,
have very high levels of saturated fats.
• Cattle are foraging animals that have been bred to keep their noses
down seeking out the most nutritious grasses in the field. Almost all beef
cattle in Australia start their life off in paddocks at their mother’s side as part
of a larger herd. When weaned, they join a new herd of younger animals.
Some remain in the paddocks eating pasture for the rest of their lives and
some will be sent to feed lots, where they have a constant supply of fresh
water and nutrient rich, grain based feed.
• Grain fed beef cattle spend 70 days minimum, often 100 days or more
prior to processing, eating a mix of grain and plant based feed. Animals fed on
grain for 300 days or more are destined for the top end of the market. They
can be housed in purpose-built feed lots, where they can rest and sleep on
beds of straw with shelter from the sun and rain. They can also be grain fed in
open fields where they are free to roam. Australia has some of the strictest
animal welfare and environmental measures in place in regards to feed lots.
Effluent ponds treat their manure and their bedding is used as compost. The
meat from grain fed beef cattle is considered by many to have better
consistency of quality, be more tender and have better marbling. Many
customers prefer grain fed beef, as the marbled fat has a clean white
appearance and the meat has a juicy mouthfeel with a lovely, nutty flavour
from the grain.
• Some diners prefer the flavour the pasture gives to grass fed beef. Grass
fed beef accounts for roughly two thirds of production in Australia. Natural
compounds in the grass, such as beta carotenes, add to the flavour of the
meat. They can also add a yellow tinge to the fat, something some diners like
but others not. Grass fed beef comes from cattle that free-range for their
entire lives. They can have supplementary feeding of hay, lucerne or silage,
but for the most part they live on grass. Many consider this a more natural
way of raising an animal and better for their well-being. While the beef from
grass fed cattle can have more nuanced flavours, grass fed beef is more likely
to be influenced by weather and climate. Good rainfall and a mild climate can
mean better pasture and better beef, while drier times can affect the quality
of the steak on the plate.
• Our diets have changed in other ways that also influence our health and risk of
metabolic disease. Modern foragers acquire most of their carbohydrates from fruits
and starchy tubers. Western societies increasingly rely on refined sugars and
sweeteners, which differ from naturally occurring sugars and starches in that the
body metabolizes them more rapidly, giving them a high glycemic index with
consequent effects on homeostatic processes such as insulin release. The process of
refining involves extracting the sugars from the surrounding food matrix, which is
often the source of important micronutrients and of fiber which also reduces the
glycemic index. Even as our diets increasingly focus on refined energy-dense foods,
our level of physical activity has decreased in recent decades in response to several
important global economic and social trends. As nomadic foragers, our ancestors
spent much of their time on the move, and our modern life increasingly deviates
from this in fundamental ways. Physical activity can be broken down into that
engaged in at work, during leisure time, and in locomotion, and all have seen
dramatic changes in the past few generations. The percentage of individuals who
engage in physical activity at work has greatly declined with the rising importance of
the service sector and the mechanization of manufacturing processes. Leisure time
is also increasingly sedentary, focusing on activities like watching television or
playing video games rather than physically active pursuits. With the advent and
wide distribution of the internal combustion engine, we now rely to a great degree
upon the hydrocarbons in fossil fuels, rather than our own dietary energy, to move
from place to place. These trends, the so-called nutrition transition, are sweeping
the globe. Although there clearly is no single Paleolithic-type diet or pattern of
activities that will keep us healthy, we see that studies of our evolutionary past and
of modern foragers do give us important insights into the aspects of our lifestyle
and diet that have changed the most, and that consequently are most likely to be
involved with the modern epidemic of metabolic disease. The Western diet and
lifestyle, with its chronic imbalance of intake and expenditure, appears to have
exceeded our homeostatic capacity to cope metabolically, helping drive the global
pattern of weight gain and its associated problems. The modern epidemic of
metabolic disease is not merely a function of calories, but also of the composition of
what we eat. Modern dietary change has led to other, no less important,
imbalances and sources of novelty, such as our increasing consumption of nonprotein
calories like processed sugars and alcohol, and foods that are not only high
in fat but also have an imbalanced fatty acid composition. All of these changes have
been rapid, leading to unprecedented global changes in human metabolism and
health. It is also important to remember that even in high-income countries many
members of the population, especially those of low socioeconomic status and living
in parts of urban environments sometimes called “food deserts,” have a diet
deficient in micronutrients and high-quality protein, and have food insecurity.
How Can Change in the
Environment Increase
Disease Risk?
• These lifestyle transitions have been implicated as causal in the changing patterns of obesity and diseases such as type
2 diabetes and cardiovascular disease. The evolutionary question is, why has this change in environment increased the risk
of disease? Many of us are now living in environments that are novel in an evolutionary sense, and the underlying
consensus is that the disease syndromes are the out come of a mismatch between the human capacity to maintain
metabolic homeostasis and the modern energetic and nutritional environments. Two sets of possibilities exist to explain
how this mismatch may arise, and they are not mutually exclusive (see Low and Gluckman 2016). First, because the
contemporary environment is novel in evolutionary terms, there has never been selection to match our genotype to it;
rather it was primarily selected and adapted for the range of Paleolithic environments. But there are additional questions:
was the genotype selected because it allowed the individual to survive bottlenecks produced by famine and now has
deleterious consequences manifest in obesity and its complications, or was it that the rate of change in the environment
outstripped the capacity of selection to match the organism’s physiology to its environment in a more general sense, and
obesity and metabolic disease are the outcomes of this mismatch?
A “Thrifty” Genotype?
• The hypotheses developed to explain the rise of metabolic and related diseases in
recent history were among the earliest attempts to use evolutionary principles to
understand human health and disease. In 1962 the geneticist James Neel proposed that
such diseases arose through patterns of inherited genes that may have been previously
advantageous in human evolution but were now disadvantageous. He referred to these
genes as “thrifty genes” because they were proposed to be associated with energy-saving
characteristics such as insulin resistance in muscle or the propensity to deposit fat (see
Neel 1962). He reasoned that such genes would have conferred an adaptive advantage in
times of famine and perhaps would have conferred little disadvantage in times of plenty,
because, as we saw for many other chronic diseases, there would have been little
selection operating to maintain health into later life, as in general these diseases do not
inhibit reproduction. Neel proposed that these genetically mediated thrifty traits that
were useful to our ancestors would produce inappropriate effects, including obesity, in a
modern world. He argued that this was particularly manifest in populations in parts of the
world where famine had been frequent. Then, when such populations underwent a
nutrition transition from a hunter-gatherer lifestyle to a modern lifestyle of chronic
dietary abundance they would be at a particular disadvantage metabolically. The “thrifty
genotype hypothesis” led to an early focus on predominantly genetic explanations for the
high prevalence of metabolic disorders in some populations. What properties would we
expect of a thrifty gene in the sense originally proposed by Neel? First, we would expect
that the gene would demonstrate (and indeed would probably have been identified by)
significant linkage to some trait which could represent thriftiness: perhaps BMI or glucose
tolerance. Second, we might expect plausibility of gene function, i.e., that it would code
for a protein or regulatory RNA associated with some aspect of control of metabolic
partitioning (e.g., insulin secretion or action, a key metabolic regulatory enzyme, or some
facet of adipocyte biology). Third, and ideally, we might expect population genetic studies
to show some signal of selection on the gene, perhaps for a “risk” allele in populations
particularly exposed to cycles of feast and famine in historical times, or even for a
“protective” modern allele where the ancestral allele is now detrimental.
Clearly, susceptibility to obesity and type 2 diabetes varies between individuals in the population, and there is
evidence for heritability of these traits. Over 800 candidate genes associated with obesity or with type 2 diabetes
have been identified (see Dai et al. 2013). However, of these only a small number have been confirmed by
replicate studies in different populations with large enough samples to give assurance that the associations are
indeed real. Moreover, even the strongest of these associations typically explains only a very small amount of
the variation in susceptibility to obesity or diabetes in a population (of the order of 1-3%). Moreover, the mere
demonstration that variation in obesity is related to genes does not prove that these genes were selected in the
human gene pool due to advantages during famine, as Neel originally proposed. Thus, the thrifty gene
hypothesis remains just that, a hypothesis without any direct support.
• In fact, putting Neel’s ideas under greater
scrutiny gives one reason to pause when consider
in the rise of agriculture ushered in poorer health
and periodic nutritional stress, which is expected
given the greater reliance upon the more
precarious, narrower selection of crops in such
populations. If hunter- gatherers were not
particularly prone to famine, then could we adapt
Neel’s hypothesis to propose that the more
widespread and severe episodes of famine that
have occurred since the introduction of monocrop
agriculture have been a major driver of selection
for thrifty metabolism? We know from studies of
lactase persistence and malaria resistance genes
that selection over such a timescale can leave
physiologically important signatures in the human
genome. Indeed, it has been suggested that
selection in response to major famine events
caused by failure of the Indian monsoon is the
cause of the tendency of individuals of South
Asian origin to deposit more of their energy in
metabolically labile visceral fat (see Wells 2007).
• Others have questioned on a number of grounds whether the selective
advantage of a putative thrifty gene would be sufficient to cause it to spread
widely. It is argued that the frequency and severity of famine have been
insufficient, that relatively few famine events have occurred in most populations,
that deaths during famine are predominantly caused by factors other than
starvation, and that mortality in famines disproportionately affects the young and
old rather than individuals of reproductive age (see Speakman 2006). In addition,
obesity is rare in well- nourished individuals from present day hunter- gatherer
and subsistence agricultural societies, who would have been expected to inherit
such thrifty genes. If foraging populations do not put down excess calories as fat
during times of plenty, how could the putative thrifty alleles underlying human
metabolism have increased survival during periods of famine? Additional doubt is
cast upon the assumed role of famine when we consider the development of body
fat. In humans, fat makes up a larger percentage of weight at birth than in any
other mammal. This is followed by a continued period of rapid fat deposition
during the early post- natal months. In well- nourished populations, adiposity
reaches peak levels during the first year of life before gradually declining in
childhood, when humans reach their lowest level of body fat in the life cycle (see
Davies and Preece 1989). If the threat of famine is what drove the human
tendency to build up fat reserves, it is not obvious why children’s bodies should do
so little to prepare for these difficult periods. The lower priority placed upon
maintaining an energy reserve by middle childhood suggests that the background
risk of starvation faced by our ancestors—“famine”—was small in comparison
with the nutritional stress during the preceding developmental period of infancy.
Indeed, infancy is marked by often intensive nutritional stress associated with
weaning and the related problem of infectious disease, making nutritional
disruption and nutrition-related mortality common at this age (see Kuzawa et al.
2007). Because all individuals who successfully pass on their genes to offspring
must have survived this early-life nutritional bottleneck, there is likely to have
been selection for building up protective fat reserves at this age in large-brained
humans. In summary, while the thrifty gene hypothesis was valuable in developing
evolutionary concepts related to metabolic disease, the hypothesis itself has
largely been discarded.
Does Evolutionary
Novelty Explain
Current Patterns
of Metabolic
Disease and
Obesity?
• The preceding sections provide the background for
considering more recent concepts of how our evolutionary
history helps explain the changing patterns of obesity and
metabolic disease, in particular the recent increases in these
conditions associated with the rapid nutrition and
socioeconomic transition in Western countries over the last 100
years, and the transition that is now occurring in low-and
middle-income countries. To recapitulate, our ancestors
evolved in the absence of agriculture and a stable source of
carbohydrates. Our species is characterized by being able to
cope with an omnivorous diet and with a metabolism and
behavior evolved on the basis of regular access to food.
Selection acted to match our physiology to these characteristics
and to the diet of the Paleolithic foragers from whom we are
descended. Although, as discussed earlier, the concept of a
typical hunter-gatherer is misleading, it is generally accepted
that we were selected within and adapted to environments
characterized by a relatively high protein intake and low intake
of sugars and fat. Frank obesity would have been a remote
possibility. It has been estimated that hunter-gatherers may
expend up to 2500 kcal each day gathering food. Since the
Upper Neolithic, the human diet has changed dramatically, at
first in association with the development of agriculture and
then, since the industrial and technological revolutions, in
association with the development of various forms of highly
refined foods. In parallel the physical work expended to “earn”
this food is declining, again at an accelerating rate.
• These two changes have occurred against a background of a marked
increase in life expectancy across virtually all populations as public
health measures take effect—more dramatically in high-income
countries, but increasingly also in low- and middle-income countries. It is
important to note that because peak reproduction occurs well before
middle age, health and reproductive fitness are not identical. There will
not be great selection pressure against health consequences occurring in
middle age if reproductive competence at an earlier age has not been
affected, although the effects on kin fitness conferred by the presence of
older individuals in social groups may provide some fitness advantage for
longevity. The increasing incidence of metabolic disorders among
younger people refutes one possible reason for the epidemic—that it is
the direct result of a longer lifespan—but does raise the question of how
such early onset disease will influence the reproductive fitness of
affected individuals. The simplest evolutionary explanation of this
epidemic is therefore that our species is facing a nutritional environment
that is entirely novel in evolutionary history and for which our
metabolism has not been selected. Instead, our metabolic repertoire of
genes is based on that which would have been best adapted for the
Paleolithic. For instance, a metabolism selected for a high-protein, lowfat,
and low-carbohydrate diet might result in obesity and metabolic
disease when confronted with a low-protein, high-fat, and high-sugar
diet. This is the simplest form of evolutionary mismatch in which cultural
and dietary change have outpaced the capacity of evolutionary
processes to compensate (see Low et al. 2015). What seems clear is that
the environmental setting is critical, regardless of genetic background.
Pima Indians who have maintained their traditional way of life do not
have high rates of type 2 diabetes, but their contemporaries consuming
a different diet and engaging in reduced levels of physical activity
frequently develop the disease. The risk of developing diabetes can vary
markedly across populations: this also requires an explanation. In some
cases this can be traced to genetic differences in local populations that
have been confronted with distinct nutritional histories, selecting for
unique mutations which influence their disease risk. At a given BMI,
populations from the Indian subcontinent are more likely to develop
type 2 diabetes than are Europeans, and it has been proposed that
monsoon cycles could have created famine conditions that might have
selected for such a trait. In themselves these observations still fit with a
simple model of mismatch caused by evolutionary novelty, but where
the sensitivity to the novel environment might be genetically different
across populations.
Developmental Mismatch—
Contributing Factor
It has been obvious from Hippocratic times that events in early life may have lifelong consequences. During the past 25 years, much research has emerged to
show that early human development is responsive to metabolic signals, and that a mismatch between the signals and the actual environment in which the
individual later lives could contribute to the rise of metabolic disorders such as diabetes and heart disease (see Gluckman et al. 2015). Population- based studies
were the first to suggest that the developmental environment plays a role in risk of later disease. Evidence for such effects came in several forms. The most
common was the finding that individuals of lower birthweight tend to have higher rates of later metabolic disorders, including hypertension, diabetes, and
cardiovascular diseases (reviewed in Godfrey 2006). These studies are by their nature very long term: how else can we relate what happened to a person when
they were a fetus or a growing child to their health as an adult? One of the most important approaches was to find historical records of birthweight and early
growth after birth from a period over 50 years ago, and then to trace the subsequent adults. In the early 1980s, David Barker, an epidemiologist from
Southampton, and his colleagues were investigating the rates of mortality from coronary heart disease and other vascular diseases in England and Wales over
the previous decade (in fact 1968– 78). They noticed that the highest rates of mortality did not occur in the areas of greatest contemporary affluence, in the
southeast of Britain for example. Rather, the highest rates occurred in the northwest of England and South Wales, which were areas of high unemployment and
poor social conditions, particularly at the time the adults they were studying had been born. Furthermore, there was a strong geographical correlation between
ratesn of infant mortality (a marker of poverty, and indirectly of maternal nutrition) in the early 1900s and rates of death from coronary heart disease several
decades later.
What might account for such an association between early life and late- life mortality? These observations were very remote in time, and
likely to be influenced by unmeasured factors, such as differences in smoking, diet, or lifestyle, which were not quantifiable. The next step was
to move from gross correlations over integrated populations and across time to find out what happened to a specific population where
individuals could be studied throughout their lives. What was needed was a set of records about the growth and development
of children in the early part of the century that could be linked specifically to the causes of death in later life. The largest set of records that
were found related to the English county of Hertfordshire. These included weight at birth, weight at 1 year of age, and whether the baby was
weaned at 1 year. The ledgers had been maintained from 1911 to 1945. Barker and his colleagues used the National Health Service
Central Register at Southport to trace 16,000 men and women born in Hertfordshire between 1911and 1930, and to determine their cause of
death. The results were controversial (see Barker et al. 1989; Osmond et al. 1993). What they found, for both men and women, was that risk
of death from heart disease was doubled in individuals born at a weight of less than 5.5 lb (2.5 kg) compared with those born at a weight of
9–9.5 lb (4.1–4.3 kg). While there was much controversy about these initial findings, they in fact built on many earlier observations suggesting
that the conditions during pregnancy and early life had long- term health consequences. Over the next two decades, a large amount of
confirmatory epidemiological, prospective clinical, and experimental studies supported the general model that an adverse start to life, which
might be proxied by a low birthweight, was associated with a greater risk of later obesity, cardiovascular disease, and insulin resistance.
Other workers also showed that excess nutrition in infancy, such as associated with formula feeding, also led to altered disease risk. Given
that poor fetal growth is generally associated with some rebound rapid growth in infancy these are probably different perspectives of the
same phenomenon. However, further evaluation of the developmental data showed that the relationship between developmental growth
patterns and disease risk is not a function of severe intrauterine growth impairment; rather, there is a continuous change in risk across the full
range of developmental growth patterns. This suggests that disease risk is not a consequence of developmental disruption but is part of a
developmentally plastic process that originally had an adaptive purpose. Indeed, subsequent studies have shown relationships between the
conditions of pregnancy such as maternal caloric intake and later pathophysiology of the offspring (see Gale et al. 2006). The first studies
were conducted in high-income countries, so of course the question arises of how relevant they are to low- to middle-income countries.
There have now been studies conducted in several of these countries including India, China, and parts of South America, and they all show
similar trends. Finally, evidence from modern famines further supports the role of the intrauterine environment in determining health in later
life.
Maladaptive Consequences of an
Adaptive Process
• These epidemiological studies could not provide evidence
about the biological basis of the relationship, but a consensus
arose that poor fetal or early life experience altered
development in such a way that growth was affected and also
influenced the propensity to develop disease in later life. The
term “programming,” previously and variously used to reflect
the putative intergenerational effects of gestational diabetes
(see later) and then the later consequences of formula feeding
versus breastfeeding, was adopted to describe the
phenomenon. Subsequently, Barker and Hales developed the
thrifty phenotype hypothesis, in an explicit reference to Neel’s
earlier proposal of a thrifty genotype, to explain programming
(see Hales and Barker 1992).
• Their concept was that the fetus adjusts its biology in
response to signals from its mother of poor nutrition, allowing
it to survive until birth (i.e., an immediately adaptive response)
but then predisposing it to the adverse consequences of such
programmed thriftiness in adulthood. They proposed that the
likely mechanism was the in utero induction of insulin
resistance, as insulin is known to be a key regulator of fetal
growth. However, that model failed to fit several aspects of
subsequent data. It assumed that fetal growth retardation was
key to the process, yet it soon became clear that most people
who later developer obesity and metabolic compromise were
not growth retarded at birth. Further, infants who are small at
birth have insulin hypersensitivity, and insulin resistance only
appears at about 3 years of age (see Mericq et al. 2005).
Nevertheless, the conceptual framework of the thrifty
phenotype has inspired considerable research which has led to
our current understandings of the developmental basis of later
disease risk.
• Animal studies confirmed and extended the results and helped dissect the underlying
molecular mechanisms. The initial studies were performer in rats. If a rat fetus was
undernourished in utero, because the pregnant dam was fed a reduced energy diet or just an
unbalanced diet (e.g., with a low protein content), it became hypertensive and obese as an
adult (Langley and Jackson 1994). These adult rats were also shown to have insulin resistance
and to have shorter life expectancies than those whose mothers had been fed a balanced diet
in pregnancy. The effect was magnified if the rat was placed on a high-fat diet after weaning,
echoing
the human situation of dietary abundance and demonstrating that the interaction between
the fetal and post-natal environments determined outcome (see Vickers et al. 2000). As
experimental work proceeded it became linked to other fields of biological enquiry, especially
evolutionary developmental biology. The realization was growing that development was an
important and under-represented component in the explanation of metabolic disease. But a
key feature of the link with metabolic disease was that this was not just about those with a
low birthweight. The epidemiological studies had shown a continuous relationship between
birth size and later disease risk, present even in those of above average birthweight.
Epidemiologists had also shown that metabolic programming could be induced by changes in
the fetal environment that did not affect birthweight (see Gale et al. 2006). Further
epidemiological and clinical studies showed that smaller infants had relatively more visceral
fat at birth, although subcutaneous fat was reduced. Such observations, together with the
broad incidence of metabolic disease in the population, suggested that the developmental
component did not represent the outcome of a pathological process involving disruption of
fetal development. It was rather a maladaptive outcome of the generally adaptive processes
of developmental plasticity.
• Indeed, there is now ample experimental evidence in
animals, and growing clinical evidence, that lifelong
epigenetic changes in genes associated with systems such
as insulin sensitivity, glucose metabolism, and the
glucocorticoid axis underpin the developmental induction
of metabolic risk (see Low et al. 2014). In humans,
individuals exposed pre-natally to the Dutch Hunger
Winter famine of 1944 showed differential methylation
levels at gene loci implicated in growth and in metabolic
and cardiovascular disease risk nearly six decades after
exposure, showing that a transient environmental
exposure can indeed have long-lasting molecular
consequences (see Heijmans et al. 2008). Other
epigenetic marks have been found in umbilical cord tissue
that relate to body composition in pre-pubertal Children.
Developmental Plasticity and Mismatched
Signals Across the Life Course
• How can these developmental components be understood in evolutionary
terms? The general paradigm we have put forward is that early developmental
cues have induced an adaptive, developmentally plastic response that cues the
individual’s physiology in the expectation of living in a nutrient-poor or
adequate environment, but then the organism ends up in a more nutritionally
plentiful environment than predicted. This general model suggests that these
processes are not pathological in origin, but are fundamental to biological
variation operating in the normal range of ecological cues. There are other
developmental pathways that reflect evolutionarily novel exposures in
development for which an adaptive explanation is inappropriate, including
maternal obesity, gestational diabetes, and feeding with infant formula.
• We explained how the processes of developmental plasticity act to allow
one genotype to give rise to a range of phenotypes and how organisms use
plasticity as an alternative or additional process to adapt to environments,
particularly among species faced with change during their life course.
Developmentally plastic responses are induced by external cues (e.g.,
maternal nutrition), and depending on the fidelity of the relationship between
the cue and the future environment there may be effects on fitness or health.
It was hypothesized that if the developing organism predicts a limited
nutritional environment in the future, it might be appropriate to use the
mechanisms of plasticity to adjust growth patterns so that later body function
is optimized for a limiting nutritional environment. Conversely, if the fetus
predicts a later environment with plentiful nutrition it is appropriate to have a
metabolic system set up with different expectations. In the former situation of
predicted nutritional threat, an appropriate response would include reduced
investment insomatic growth (e.g., reduced muscle mass), a preference for
high- fat foods, metabolic settings that favor fat deposition in times of energy
excess, and altered endocrine, behavioral, and vascular controls such that the
organism has reduced insulin secretion and sensitivity. Given that evolution is
driven by the fitness imperative, anticipation of a threatening environment
might be expected to accelerate the timing of maturation and commit more
resources to reproduction, perhaps even at a cost to other traits that improve
longevity (e.g., by investing less in cellular or DNA repair- see Gluckman and
Hanson 2006a). Depending on the severity of the initial cue, the organism may
only respond with predictive adaptation, but if the challenge in early life is
more severe it will induce the immediate adaptive responses of reduced
growth and/ or early delivery, thus explaining the original birthweight
relationships found in epidemiological studies.
• The general hypothesis of predictive adaptive responses was
initially tested in rats. It was shown that neonatal rats that had
been born to undernourished mothers could be “tricked” into
thinking they were in a high- nutrition environment by being
injected with the adipokine leptin. This prevented the animals
from developing obesity, insulin resistance, and the other
features of metabolic compromise, and the associated epigenetic
changes— even when maintained on a high- fat diet through life
(see Vickers et al. 2005; Gluckman et al. 2007b). But could such a
hypothesis be tested in humans? Studies in a population where
severe undernutrition is common showed that being born small
was protective against the risk of dying from infant
undernutrition, and could be best interpreted as supportive of
the evolutionary hypothesis (see Forrester et al. 2012). The
adaptive advantage of a predictive response depends on the
fidelity of the prediction. If correct predictions lead to greater
chances of growth and survival to reproduce, this would be why
underlying anticipatory and plastic processes have been selected
through evolution. Modeling work shows that developmental
forecasting does not have to be particularly accurate to confer a
selective advantage (see Jablonka et al. 1995). When fetal
nutrition is not a reliable cue of external conditions— as a result
of maternal disease, placental inadequacy or malfunction, or
simply because the environment has changed notably between
birth and later life— this could produce a phenotype that is not
well suited to meeting the challenges of its environment, thus
placing the individual at greater risk of disease. In the metabolic
domain, a phenotype of increased insulin resistance, reduced
muscle mass, and increased propensity to store fat is precisely
the background on which susceptibility to metabolic disease
would be enhanced in a later nutritional environment of high
energy availability.
• Because of maternal constraint, it is reasonable to suggest that most
individuals are sensitized to an obesogenic environment because the
fetus will be biased towards predicting lower- nutritional environments
that may have been typical of earlier epochs. Maternal constraint refers
to a set of somewhat poorly defined mechanisms by which fetal growth is
limited by maternal size (see Gluckman and Hanson 2004). It probably
involves limits on uterine blood flow and placental– fetal hormone
interactions. The outcome is that genetic factors are less important than
postnatal factors in determining fetal growth— indeed embryo transplant
experiments in animals, and studies of humans born to surrogate
mothers, show that the maternal phenotype rather than genotype is the
primary determinant of birth size. These studies also show that fetuses
do not grow to their maximal genetic potential because of these
constraining mechanisms (see Hanson and Godfrey 2008), and thus are
perhaps generally signaled to expect limited nutrition after weaning. The
developmental mismatch model proposes that, in evolutionary terms,
there is an advantage in using the processes of developmental plasticity
to adjust the set points for metabolic homeostasis to match the predicted
environment of the mature organism. This process could have had
adaptive advantage in environments that were reasonably stable over
decades. But the fidelity of the prediction might increasingly be lost
because the mechanisms of maternal constraint limit the forecast that is
possible, particularly as the post- natal environment becomes abundant
in evolutionarily unprecedented ways. Additionally, modern medical care
allows a greater range of fetuses to survive, many with greater degrees of
maternal constraint (such as twins and those with extreme growth
retardation) and perhaps therefore with a greater risk of mismatch.
Maternal ill- health and placental dysfunction is other ways in which the
maternal– placental transduction of environmental information can lead
to a faulty prediction, and progress in obstetrics and pediatrics allows a
far greater range of babies to survive to adulthood.
• While the expansion of experimental and
clinical research and the definition of underlying
epigenetic mechanisms has established the view
that developmental pathways are important, the
relative importance of these pathways in
contributing to the risk of metabolic disease
remains uncertain. Clearly, disease risk would
not exist were it not for the large changes in diet
and lifestyle. It is important to note that this
model does not imply that developmental
factors cause obesity or type 2 diabetes; rather it
explains how developmental factors change the
sensitivity of the individual to the obesogenic
environment in which they will live as older
children and as adults. Indeed, recent
epidemiological studies show that there is an
interaction between developmental and
evolutionary mismatch, such that adult lifestyle
interacts with birthweight to compound the risk
of developing type 2 diabetes (see Li et al.
2015).
Evolutionary
Novelty in
Development
The developing human may, additionally, be exposed to entirely novel exposures, such as infant formula based on cows’ milk. It is well established that feeding with infant formula
predisposes to obesity, probably through mechanisms other than via adaptive developmental plasticity, and that breastfeeding has a protective effect (see Cattaneo and Cogoy 2012).
When infants are breastfed they regulate their own food intake. This is not the case when fed by bottle, and this may change the development of appetite control. Human breast milk
has a number of components including milk oligosaccharides, neurohormones such as leptin, and other hormones such as IGF1, which may have an impact on the infant’s metabolic
and satiation biology and development. Furthermore, human breast milk supports a different gut microbiota from that of formula (see Penders et al. 2006); these differences persist
for some time after birth and may have important long-term effects. Finally, the macronutrient content of infant formula is very different from that of breast milk, and may make
obesity more likely to develop. A second example of probable evolutionary novelty that is of rapidly emerging public health importance is gestational diabetes mellitus. Hyperglycemia
in a pregnant woman leads to fetal hyperglycemia and hyperinsulinemia, thus resulting in adiposity of the offspring. Insulin is known to be adipogeneic in the late-gestation fetus and
infant. As a consequence, infants of diabetic mothers have an increased fat mass, in proportion to the degree of maternal hyperglycemia. The larger number of fat cells in infancy
confers a greater risk of obesity through childhood. Beyond this, there may also be epigenetic changes that also predispose to a later risk of diabetes. It has been suggested that
gestational diabetes is an evolutionarily novel exposure (see Ma et al. 2013). Some degree of insulin resistance is normal in pregnancy, and is induced by placental production of
growth hormone and placental lactogen that induces mild insulin resistance in the pregnant woman. This promotes transfer of maternal glucose to the fetus to promote fetal growth.
But unlike other nutrients such as amino acids and fatty acids, there is no limit to placental glucose transfer. Gestational diabetes itself is more likely in women who had been born
small and are consequently at greater risk in an obesogenic environment, thus reflecting earlier developmental impacts on their own lives. Left untreated, gestational diabetes results
in fetal macrosomia, which in the absence of obstetric intervention leads to dystocia, posing a severe risk to both mother and child. That there is no limit on glucose transfer suggests
that there has been no selective pressure for such a need, which might imply that gestational diabetes is itself a manifestation of modernity. Indeed, maternal obesity and gestational
diabetes often coexist, and correlations have been shown between maternal BMI and offspring adiposity.
• Animal research is now starting to suggest that paternal obesity and
insulin resistance, experimentally induced in male rodents, can also pass
risk to their offspring, even when the mothers are lean and fed a
balanced diet (see Fullston et al. 2013, Wei et al. 2014b). This is one of
the growing pieces of evidence for epigenetic inheritance. One of the
most active areas of current research concerns the role of the gut
microbiome in health. Recent work has shown that women whose gut
microbiota predominantly comprise the bacterial phylum Firmicutes had
different epigenetic marks in genes functionally associated with obesity,
cardiovascular diseases, and inflammation. A human infant that is
delivered vaginally will acquire a microbiota similar to that of its mother’s
vagina, while infants delivered by cesarean section have a very different
microbiota (see Funkhouser and Bordenstein 2013). There is now good
evidence that the risk of later overweight/obesity is up to 25% greater in
babies who were delivered by cesarean section (see Darmasseelane et al.
2014). Although the causative nature of the association remains to be
tested, this probably represents another example of a health risk arising
from exposure to an evolutionary novelty—namely loss of exposure to
the maternal vaginal microbiota. Other evolutionarily novel exposures
include environmental chemicals such as endocrine disruptors, which
mimic or block hormonal action and are found in many common modernday
household products. Many of these endocrine disruptors interfere
with lipid metabolism and adipogenesis, and have been implicated in the
growing obesity epidemic. Recent data in rats suggest that the adverse
effects of maternal obesity on health of the offspring may be mitigated by
supplementing the mother’s diet during pregnancy with methyl donors,
taurine, or a cocktail of antioxidants (see Vickers and Sloboda 2012). The
specific mechanisms by which this occurs remain unclear, but these
findings provide important proof of principle for the reversibility of
developmental metabolic effects induced by evolutionarily novel
exposures. In light of the modern-day obesity epidemic, the potential
public health utility of these findings may be substantial.
Thank you for your
attention!!!!!