ONE OF THE LESSER-KNOWN ATROCITIES COMMITTED DURING World War II occurred during its waning months. In September 1944, the Germans were in retreat throughout most of the Europe. They retained, however, a stronghold in the populous northwestern portion of the Netherlands, which was of both strategic and symbolic importance to the fading Nazi cause. But German control of this area was threatened by Allied forces approaching from the south, in support of which the exiled Dutch government ordered a railway strike. Though the allied forces were stopped at Arnhem, the Germans retaliated for the railway strike and other hostile actions by Dutch partisans with a food embargo. Unfortunately, the embargo coincided with the onset of a particularly severe winter during which the canals froze over, disrupting barge transport. Things further deteriorated when, in response to the advance of allied troops from the south, the retreating Germans destroyed what remained of the transportation infrastructure and flooded much of western Holland’s agricultural lands.
By the end of November, the diet for most inhabitants of the major cities in western Holland, including Amsterdam, was reduced to only 1,000 calories per day, a huge drop from the 2,300 calories normally consumed by an active woman and the 2,900 calories normally consumed by an active man.1 At the end of February 1945, rations had dropped to 580 calories in some parts of western Holland. To augment this meager fare—consisting largely of bread, potatoes, and a cube of sugar—city dwellers were forced to walk many miles to the nearest farms, where they traded whatever they owned for food. Those without the means to trade were forced to eat tulip bulbs and sugar beets as a last resort. The worst effects of the famine were largely confined to the major cities of western Holland, particularly the poor and middle class. In the rural areas of the west, farmers were self-sustaining. Eastern Holland—roughly half of the Dutch population—largely escaped the famine.
By the time the Netherlands was liberated by the Allies in May 1945, 22,000 people had died in western Holland. Death by starvation is the traditional measure of a famine’s effects. But that measure, it turns out, is inadequate, for many who survived the famine were also severely affected, not least those who experienced the famine in their mothers’ wombs. This group became part of the Dutch Famine Birth Cohort Study, a pioneering investigation of malnutrition that continues to this day.2
The Dutch famine was unique in that its onset and end could be precisely dated. Moreover, the Dutch maintained and stored meticulous health records for all citizens after this period. These two circumstances comprised what scientists refer to as a natural experiment. Clement Smith was the first person to recognize it as such.3 Smith, of the Harvard Medical School, was among a group of doctors from the United Kingdom and the United States who were flown into the Netherlands in May 1945, immediately after the German surrender. He saw in this tragedy an opportunity to advance our understanding of the effect of maternal nutrition on fetal development.
Some Unexpected Consequences
Smith obtained obstetric records from The Hague and Rotterdam. He found that babies born during the famine weighed considerably less than those born prior to the famine. That this does not seem surprising to us now is due in no small part to Smith’s groundbreaking research efforts. Moreover, as Smith suspected, subsequent research established a strong link between low birth weight and poor neonatal health.
Others wondered about the longer-term effects of the famine. The first long-term effect was identified, retrospectively, in eighteen-year-old military conscripts. Those who were in their mother’s womb during the famine came of age for military service—which was compulsory for males—in the early 1960s. At induction they were given a thorough physical examination. These records were subsequently inspected by a group of scientists in the 1970s.4 They found that those exposed to the famine during the second and third trimester evidenced significantly elevated levels of obesity, roughly double the levels of those born before or after the famine.
A subsequent study, which included both males and females, focused on psychiatric outcomes. Here again the Dutch penchant for detailed medical records made the study possible. The investigators who mined these data found a significant increase in the risk for schizophrenia in those prenatally exposed to the Dutch famine.5 There was also evidence of an increase in affective disorders, such as depression. Among males, there was an increase in antisocial personality disorder.
In the early 1990s, a new series of studies commenced, based on individuals identified at birth from hospital records, most notably, Wilhelmina Gasthuis Hospital in Amsterdam. The first of these studies was restricted to females and focused primarily on birth weight. The investigators again found that those exposed to the famine during the third trimester were abnormally small at birth. But they also found that those exposed during the first trimester were larger than average, suggesting some compensatory response, perhaps in the placenta, to food stress early in pregnancy.6
In the second study of this series, which commenced when the cohort had reached 50 years of age, both males and females were included. For the first time, investigators turned their attention to cardiovascular and other physiological functions. At this age, those prenatally exposed to the famine were more prone to obesity than those not exposed. Moreover, they showed a higher incidence of high blood pressure, coronary heart disease, and type II diabetes. When the cohort was resurveyed at the age of fifty-eight years, these health measures continued to trend adversely.7
But the nature of the adverse effects of the famine on the fetus depended largely on the timing of exposure. For instance, coronary heart disease and obesity were associated with early exposure during the first trimester. Women exposed during the first trimester also had an increased risk of breast cancer. Those exposed during the second trimester had more lung and kidney problems. Altered glucose intolerance was most evident in those exposed during late gestation.8
By the late 1990s, several research groups were independently studying the Dutch famine cohort, studies which continue to this day. Together they provide some of the most compelling evidence for the long-term effects of the fetal environment on our health. Having documented these effects of the famine, some of the scientists involved have turned their attention to the mechanism underlying them. That is, they now seek to understand how mothers’ malnutrition during pregnancy can cause obesity or schizophrenia in their offspring when those offspring are adults.
From Environment to Gene
It will come as a surprise to many that our external environment affects us through our genes, by modulating their activity. Our environment does not affect our genes directly. Rather environmental influences on our genes are mediated by changes in the cells in which our genes reside. Different kinds of cells respond differently to the same environmental factor, whether it is social stress or food deprivation in the womb. As such, and despite the fact that all of the cells in our body have the same genes, any environmental effect in you is cell type–specific. Your liver cells will react one way to poor nutrition, your neurons will react in a different way, and many cell types won’t react at all. Therefore, in determining any environmental influence on gene action, scientists look at specific cell populations, such as neurons in a particular part of the brain, liver cells, pancreatic cells, and such.
The Dutch famine clearly affected many different kinds of cells in the exposed individuals, some in the brain, some in the heart, some in the liver, some in the pancreas, and so forth. If we were to compare, say, the liver cells of those in the Dutch famine cohort with those unaffected by the famine, we are likely to find different patterns of gene activity. Some genes in the liver cells of affected individuals will be more active and some less active than in unaffected individuals. The initial goal is to identify the particular genes in these liver cells that are altered activity-wise by food deprivation in the womb. Then comes the hard work of establishing a causal link between these altered gene activities in the liver cells and diabetes or whatever condition we seek to explain.
The control of gene activity by a cell is called gene regulation. I will discuss gene regulation, especially epigenetic gene regulation, in more detail later in the book. For now, I am painting with a broader brush.
Before the advent of epigenetics, biologists already knew a great deal about short-term gene regulation, that is, gene regulation that occurs over time spans ranging from minutes to weeks. I will refer to this short-term gene regulation as “garden-variety” gene regulation, because this is the form of gene regulation long taught in introductory biology courses. Epigenetic gene regulation is not garden-variety gene regulation. For reasons we will explore later, epigenetic gene regulation occurs over much longer intervals, sometimes spanning an entire lifetime. Epigenetic gene regulation is long-term gene regulation. It is the kind of gene regulation that is most relevant to the Dutch famine cohort.
Epigenetically regulated genes can be identified by characteristic marks in the form of particular chemical attachments. The most common sort of chemical attachment involves the methyl group, which consists of one carbon atom bonded to three hydrogen atoms (CH3). A gene with methyl attachments is said to be methylated. Methylation is not an all-or-none affair; genes can be methylated to varying degrees. Generally, the more methylated a gene is, the less active it is. It is with these facts in mind that scientists have begun to look for epigenetic alterations induced by the Dutch famine. Though these are still the early days, this research has already borne fruit.
In one recent study of the Dutch famine cohort, a number of epigenetically altered genes were identified in blood cells.9 That is, the degree of methylation in these genes differed in those exposed to the famine compared with those who were not exposed. Of particular note were the epigenetic differences in a gene that codes for the hormone insulin-like growth factor 2 (IGF2), so called because it closely resembles insulin and because it promotes growth, through cell division, in a variety of cell types. (The “2” reflects the fact that it was the second of three IGF molecules to be discovered.) IGF2 is essentially a growth hormone, one that is particularly important for the growth of the fetus.
Scientists are far from being able to causally connect the epigenetic alteration in IGF2, the gene for IGF2, to any of the Dutch famine’s diverse health impacts, such as birth weight, diabetes, and schizophrenia. For starters, they will need to determine whether similar epigenetic changes in IGF2 can be found in other types of cells. They will then need to establish a causal link between the cell type–specific epigenetic alterations in IGF2 and these conditions. This result is nonetheless quite significant in demonstrating that the epigenetic effects of the fetal environment can extend over six decades.
Most epigenetic attachments are removed during the production of sperm cells and egg cells. Hence, the fertilized egg commences development with an epigenetically clean slate. Sometimes, though, epigenetic attachments can be passed on, along with the genes to which they are attached, to the next generation. It is noteworthy, in this regard, that the adverse effects of the famine were not confined to those who lived through it. The children of those who experienced the famine through their mother’s womb are more prone to ill health later in their lives than children of mothers not exposed to the famine.10
This is really quite an astounding discovery, a nongenetic mode of inheritance that influences our health. As I will discuss later in the book, scientists are increasingly aware of nongenetic inheritance of various sorts, some of which we can call true epigenetic inheritance. It is far from clear, however, that this grandmother effect of the Dutch famine represents true epigenetic inheritance, that is, the inheritance of methylated genes. As we will see, there are other possible explanations. To better understand whether this grandmother effect is or isn’t true epigenetic inheritance, we need some background. I begin with the stuff to which epigenetic marks are attached: What, exactly, are these things we call genes? And what do they actually do?