Pushing strollers festooned with tote bags and diapers, Kristen and Linda wheel their infant daughters into a sunny lounge on the seventh floor of INOVA Fairfax Hospital, a busy hospital in a Washington, D.C., suburb. Both women are first-time mothers. At the moment, neither is worrying about her child’s long-term health or future risk of becoming fat. They have come to attend a weekly breastfeeding support group, and their biggest worry right now is whether their babies are getting enough to eat. Kristen’s daughter, Katie, is 7 weeks old today. Kristen has been nursing her without much difficulty, but recently she and Katie seem to have hit a bump in the road. “After 15 or 20 minutes, she’ll be slowing down but then she’ll start kicking, arching her back, rooting around. She seems to want more,” Kristen says. “She flips out. She pulls at the nipple.” Katie’s fussing has made Kristen wonder whether there could be a problem with her milk supply.

Linda’s daughter, Grace, is only 5 days old. Lying next to plump, wide-awake Katie, she looks impossibly tiny and fragile. She was born

at 38 weeks’ gestation, 2 weeks before her due date—a little early, but not early enough to be considered premature. She weighs just under 6 pounds and seems to sleep all the time. Like many newborns, especially premature babies, in the days after birth she developed a relatively high blood level of a natural substance, bilirubin, formed by the chemical breakdown of a blood protein. Because her liver—responsible for removing bilirubin from the blood—is still immature, the “bili” level in her bloodstream has so far stayed somewhat high, and it’s a natural sedative. Very high levels of bilirubin can damage a newborn’s brain. Although Grace’s levels are nowhere near that dangerous range, her pediatrician has been tracking them and will start treatment if they don’t come down on their own. Poor Linda has been told by the doctor that breastfeeding every two hours will help Grace excrete the excess bilirubin, but she has to work hard to get the baby to wake up long enough to nurse. Linda looks pale, exhausted, and desperate. “She’s jaundiced. She’s lost a fair amount of weight,” she tells Debbie Tobin, a nurse and professional lactation consultant who runs the breastfeeding support group. “I’m trying to get her enough milk to help with the jaundice and help with the weight loss…. Do you think that as the bili numbers go down she won’t be so comatose?”

Tobin dispenses a stream of information, reassurance, and advice to both mothers. She tells Kristen that Katie’s steady weight gain means her mother has an ample milk supply. Katie’s behavior most likely means that the 7-week-old is developing a “flow preference”: she has decided that she likes nursing better when Kristen’s milk flows out rapidly, as it does first thing in the morning or at the start of a feeding, and gets impatient when the flow slows down, as it does toward the end of a feeding or at the end of the day. Kristen can speed up the flow by interrupting nursing to massage her breasts or compressing them to squeeze out more milk. “Babies figure out that nipple stimulation causes letdown” of milk, Tobin says. “Katie has her own way: pulling on the nipple. Some babies will start humming. Some will start patting. Older babies will actually push” the breast. “They’ll knead, like dough.”

She checks Linda’s breasts and reassures her that she, too, has plenty of milk for her baby; her challenge will be finding ways to keep

sleepy Grace awake and feeding. “Undress her,” Tobin says. “Make a high-pitched noise. Whistle.” Tobin pulls off the sleeping baby’s pink gown, stripping Grace down to her diaper. She rubs the baby’s chest, her back, her feet. She holds Grace against her forearms and rapidly tilts her first upside down, then right side up. Grace finally comes to and starts crying in a tiny voice that sounds like a creaky door hinge. Tobin helps Linda to position Grace on a shelf-like breastfeeding cushion that fastens around the mother with a Velcro belt. She shows her how to get the baby to “latch on” to the nipple properly so she can suck efficiently without hurting her mother. “If she falls asleep, count to 10, then whistle and jiggle her,” Tobin advises.

Linda practices and takes notes. She decides to buy one of the special cushions. Kristen, with the confidence born of seven weeks’ successful experience, assures her that it won’t always be this hard. “Every week is different, trust me,” she says. “Every day is different.” Linda writes down Kristen’s telephone number and struggles to recall her own. “If you [breastfeed] every two hours, you really have no downtime,” she says. “You can’t function on 20 minutes’ sleep every two hours…. I’m so brain-dead I can’t even remember my mobile number.”

Developing the knowledge and skills required to be a good parent isn’t easy. Faced with a newborn, a first-time mother often feels as if she’s trying to interpret the cues and minister to the needs of a tiny alien. That feeling can arise even during pregnancy. Pregnant women are sometimes encouraged to think that building a baby is a lot like building a house: invest in high-quality construction materials (a wholesome diet) and let the contractor—the woman’s own body—do the rest. The development of the unborn infant proceeds invisibly, through complex processes that seem beyond the mother’s control. After the first few months she can feel the fetus moving around in her uterus, and as the long weeks of pregnancy creep by, she devotes herself to maximizing her baby’s well-being by trying to eat healthy foods “for two” and find time to exercise her ever-heavier body.

Even the most detail-oriented mother-to-be may not realize just

how much her activities during pregnancy influence her baby’s development. Her food choices; her body’s fat stores and her rate of weight gain; her overall health; her stress level; the faint sound of her voice transmitted through her abdominal wall; her habits regarding alcohol, tobacco, medicines, and drugs—all of these factors contribute to the fetus’s internal environment. A growing body of research on a relatively new scientific subject known as fetal programming indicates that the fetus has its own system of sensors for picking up signals about what kind of world it will be born into, signals that may influence the activity of various genes, affect the pattern of overall development, even entrain the baby’s future food preferences. “The new paradigm recognizes that much of the development of organ systems and programming occurs in fetal life and during the first few years of life,” says Matthew W. Gillman of Harvard Medical School, who studies factors in the fetal environment that may contribute to future obesity and disease risk. Various studies suggest that the environment in the womb can affect an individual’s later risk of obesity, as well as the risk of heart disease, diabetes, cancer, osteoporosis, neuropsychiatric illness, and asthma.

Before birth, an infant’s future risk of obesity is linked to the health, body size, and nutrition of its mother during pregnancy. Both the woman’s BMI before conception and her rate of weight gain during pregnancy contribute to her infant’s growth pattern before birth and are linked to the child’s future risk of obesity and of certain other medical problems. After birth, the parents’ decisions about infant feeding also play a role in determining future obesity risk. Research evidence suggests that simply by breastfeeding rather than formula feeding a woman can moderately reduce her baby’s chances of becoming overweight later in life. Flavors from foods that a mother eats may also be transmitted to an infant in breast milk, possibly influencing the child’s eventual food preferences. Later, when the baby is introduced to an expanding array of “table foods” during the first two years of life, the foods and beverages offered to the growing child by parents, siblings, grandparents, and caregivers will help establish future eating habits. Careful choices about what and how to feed a baby or toddler can become the basis of a lifelong healthy diet.

The more researchers probe the causes of the obesity epidemic,

the clearer it seems that preventing unhealthy weight gain in America’s children will require adults to make profound changes in many of their own choices about diet, activity, and lifestyle.

In the United States, women are starting pregnancy fatter than ever before. At least 44 percent of women between the ages of 18 and 49 (considered the adult childbearing years) are overweight or obese. Rates among some minority groups are even higher—60 percent among African Americans, 62 percent among Mexican Americans. Between 1960 and the mid-1990s, the rates of obesity and overweight grew faster among women of childbearing age than among older women or among men.

An increasing percentage of women in the United States are also gaining too much weight during their pregnancies. In 2002, 21 percent of women who carried their pregnancies to term (40 weeks’ gestation or more) gained more than 40 pounds, the maximum weight gain recommended by the Institute of Medicine’s current guidelines. The percentage of pregnant women who gain more than 40 pounds during their pregnancies has been rising since tracking of the information began in 1989, when it was 15 percent. The large numbers of women who are either overweight to begin with or who gain unhealthy amounts of weight during pregnancy may be contributing to the rising rates of obesity evident in American children.

Both obesity and excessive weight gain during pregnancy may affect metabolic processes in a woman’s body in ways that can heighten the later risk of obesity or diabetes for her child. Although the genes a baby inherits from its parents are powerful determinants of its future obesity risk, evidence indicates that environmental factors during pregnancy also have an impact. Overweight women tend to give birth to heavier babies, and infants with high birth weights (regardless of their mothers’ weight status) have a somewhat increased likelihood of later becoming obese, compared with newborns who weigh less. With so many women already unhealthily heavy when they become pregnant, some experts see the potential for a vicious cycle: an epidemic of obesity that could worsen with each succeeding generation.

Overweight and obese women also have higher rates of complica-

tions than those who start pregnancy at a healthy weight. These include higher rates of diabetes and of preeclampsia, a condition causing high blood pressure and other abnormalities that can be risky for a woman and her fetus. Obese women (those with a BMI of 30 or higher) are more likely to have difficulty delivering vaginally and to need a cesarean section. They are at greater risk of such pregnancy complications as infection and postpartum hemorrhage. They have a higher frequency of stillbirth than women who are not overweight, and their infants have a higher death rate during the newborn period. Their babies also have higher rates of several types of birth defects, including spina bifida and structural heart abnormalities.

Gaining too little weight during pregnancy can also be unhealthy. Indeed, having a mother who eats too little may be more hazardous for an infant’s future health than having one who eats too much. Although guidelines for healthy weight gain during pregnancy vary considerably depending on a woman’s initial BMI, a minimum of about 15 pounds of weight gain is recommended even for someone who is obese when pregnancy begins. Unfortunately, the percentage of pregnant women who fail to gain even this minimum amount has also been increasing—from 9 percent in 1989 to 12 percent in 2002. If a woman fails to take in sufficient nutrients to adequately nourish her developing fetus, she may have a baby who has been partially starved before birth and as a result is at increased risk of future heart disease, high blood pressure, diabetes, and stroke. Such babies are born smaller than other full-term infants and are referred to as “small for gestational age,” or SGA. Most evidence indicates that they do not grow up to have high rates of obesity. Nevertheless, many studies have found that in later life they often have an unhealthy pattern of fat distribution and an unusually high likelihood of developing the metabolic syndrome, the pattern of abnormalities associated with a high risk of diabetes, heart disease, high blood pressure, and related circulatory disorders.

“We’re looking at trouble on both ends of the birth weight spectrum,” says Matthew Gillman. Research findings suggest that the babies with the lowest risk of future health problems are those whose birth weight is “appropriate for gestational age,” or AGA, falling somewhere in the middle of the normal range rather than at either extreme. The range of appropriate birth weights depends, of course, on whether

a baby is born prematurely or at full term (37 to 42 weeks gestation). Full-term infants are considered appropriate for gestational age if their birth weight is greater than 2,500 grams (about 5.5 pounds) and lighter than about 4,000 grams (about 8.75 pounds).

Experts are increasingly urging women to try to attain as healthy a weight as possible before they try to become pregnant, but when pregnancy begins, weight-loss attempts should cease. Every pregnant woman should promptly seek prenatal care and should ask her doctor’s advice about diet and about the range of weight gain recommended for her. With proper guidance it shouldn’t be too difficult for most women to eat sensibly during pregnancy. A pregnant woman who eats a healthy diet containing a variety of foods that she enjoys is sending messages to her fetus that will help prepare the baby to adopt similar eating habits. Keeping weight gain within the range recommended by her doctor and eating healthy foods may help “program” her fetus’s developing nervous system and digestive organs in ways that will allow the infant, after birth, to optimally regulate her or his own appetite, food intake, and metabolism.

Throughout history, midwives and physicians have recognized that when a fetus grows too slowly inside the womb, something is amiss. Too little food for the mother has historically been the commonest cause, and it remains a frequent factor today in communities where many people are poor. Yet even when food is abundant, some fetuses grow too slowly for other reasons—for example, because a pregnant woman smokes, because she is restricting her food intake in an ill-advised effort to avoid gaining weight, because she is abusing alcohol or drugs, or because there is a medical problem affecting the blood supply to the placenta, the specialized organ that provides nourishment to the fetus. Whatever the cause, the chances of survival and future health are endangered if the fetus does not receive sufficient nutrients during the critical period of development.

But an excess of nutrients during development is also bad for the fetus. Doctors first learned this lesson in the twentieth century by caring for diabetic pregnant women and their newborns. With the discovery of insulin as a diabetes treatment, it was possible for the first time

for young diabetic women to remain healthy enough to have children. However, the disease makes pregnancy high risk, and pregnant women with diabetes need careful medical monitoring. A pregnant woman whose diabetes is not well controlled by insulin or other treatment sends abnormally high levels of glucose and other nutrients to her fetus through the placental circulation. Her fetus responds to this over-abundant energy supply by producing high levels of its own insulin, a hormone that accelerates growth by promoting the storage of glucose and other nutrients in tissues. As a result, the fetuses of women with uncontrolled diabetes often grow very big, with an increased amount of body fat and large internal organs. These infants’ size places them at heightened risk of delivery complications and injuries. They also have higher-than-average rates of intrauterine death and of birth defects.

A small percentage of women who are not diabetic before becoming pregnant develop a pregnancy-related form of the disease known as gestational diabetes. Although these women were previously making enough insulin for their own bodies’ needs, pregnancy places excessive demands on the insulin-producing beta cells of the pancreas. In a woman with gestational diabetes, the pancreas can’t churn out the hormone in sufficient quantities to keep the mother’s blood glucose level within a desirable range. The risk of developing gestational diabetes is higher than average in women who become pregnant relatively late in their reproductive years. The risk is also higher in women who are nonwhite, those who are overweight, and those who smoke.

The body’s added demand for insulin occurs because pregnancy is a time when the interplay of various hormones causes dramatic shifts in a woman’s metabolism, explains Frank R. Witter, director of labor and delivery at Johns Hopkins Hospital in Baltimore. “The amount of insulin goes up, but so do the amounts of hormones that work counter to insulin,” points out Witter. “Overall, the mother’s blood sugar in pregnancy is lower, normally, than it is when she’s not pregnant…. Glucose [in her bloodstream] is preferentially shunted over to the baby. The baby uses it as primary fuel.” Meanwhile, the mother’s metabolism “is geared up to use fats [as fuel] more than she would normally. It’s a shuffling around of the fueling system so that the readily available glucose goes to the unborn child.”

Gestational diabetes is usually a relatively mild form of diabetes, one that can frequently be treated with diet alone and that most often abates when pregnancy ends. Nevertheless, it poses risks for mother and infant if it is not identified and treated. Pregnant women are routinely screened with an oral glucose tolerance test. An hour after drinking a solution containing 50 grams (a little under 2 ounces) of glucose, a woman’s blood glucose level is measured. If it’s above a certain level, she undergoes a more extensive version of the test, with sequential measurements of blood glucose to determine whether her pancreas is producing enough insulin. If a pregnant woman is diagnosed with gestational diabetes, Witter says, she is usually given a diet aimed at preventing sudden rises in blood sugar but not necessarily designed to restrict calories or prevent weight gain. With such a diet, “the patient’s limited insulin response to food intake can result in a continued good range for her blood sugar,” he says.

It was by studying the long-term health of infants born to diabetic mothers that researchers became aware of the possibility that presenting too rich a feast of nutrients to a fetus might program that child for later obesity risk. Infants of diabetic mothers “are born large but decrease their weight into a normal range thereafter; however, they have a significantly increased risk for obesity by 6 to 10 years of age (50 percent prevalence),” writes pediatric endocrinologist Dennis M. Styne. Researchers at Northwestern University monitored children of diabetic mothers into adolescence and found that these children had higher rates of obesity and of impaired glucose tolerance (a prediabetic condition) than did the children of women who had not been diabetic during pregnancy. Whether the chances of future obesity are heightened in children of women who were not diabetic until they became pregnant is controversial. Some studies of infants born to women with gestational diabetes have concluded that they share the same elevated obesity risk as the infants of other diabetics; others dispute this finding.

Why blame the fetal environment at all for causing a heightened risk of obesity or diabetes in the children of diabetic women? Couldn’t those risks be caused instead by genes transmitted by the mother? Some studies have tried to untangle those two possibilities. For example, in

the Northwestern University studies mentioned above, researchers found that the risk of obesity and of impaired glucose tolerance in adolescents whose mothers had been diabetic correlated with high levels of insulin production by those children while in the womb. A high fetal insulin level (measured in the amniotic fluid that bathes the fetus) thus signaled a more severe impact of the mother’s diabetes on the fetus—and also was linked to a greater risk that the child could become obese or diabetic years afterward. That finding provides evidence for an impact of the fetal environmental on future health that is independent of genetically transmitted risk.

The Pima Indians of Arizona, as we have seen, are both genetically and environmentally predisposed to obesity and diabetes. Studying members of this tribe has greatly aided scientific understanding of how genes and environment interact. Researchers found Pima families in which the mothers had been diabetic during pregnancies with some of their children but not during others. Studying these Pima children when they were between the ages of 9 and 24, researchers found that those who had been exposed to diabetes before birth had a higher frequency of obesity than their unexposed siblings, even though the sibling pairs shared many of the same genes. “This tends to isolate the fetal environment as a potentially important period,” says Matthew Gillman.

Screening women for diabetes is a recommended part of routine care during pregnancy. By making sure that all women get timely prenatal care, including diabetes testing and proper treatment for gestational diabetes, it should be possible to reduce the chances of later obesity in infants of diabetic mothers. Still, the impact of the fetal environment on the obesity epidemic doesn’t seem to be limited to children of women with diabetes. Bigger babies, whatever the reason for their large size, appear to be at higher risk in later life of becoming overweight. More than two dozen studies have examined the association between infants’ birth weight and their BMI as children or adults. Almost all have found a direct association: “Higher birth weight is associated with higher attained BMI,” notes Gillman.

Once again, it’s difficult to tease apart the influence of genes from that of environmental factors in the womb. We know that overweight parents are genetically predisposed to have children with a higher-

than-average likelihood of becoming overweight. But Gillman points out that a child’s risk of being overweight is more closely associated with obesity in its mother than in its father. Although this finding could imply sex-linked transmission of certain genes that influence obesity risk (for example, genes located on the X chromosome, which children of both sexes receive from their mothers), it also suggests a possible role for the prenatal environment. Gillman and colleagues studied the relationship between birth weight and later obesity rates in more than 14,000 young adolescents, aged 9 to 14, who were the sons and daughters of women participating in Harvard’s ongoing Nurses’ Health Study of behavioral factors and disease risk. They found a correlation between birth weight and later obesity risk. Average birth weight was about 3.5 kilograms, or 7.5 pounds. Each additional kilogram (2.2 pounds) of weight at birth above this average figure increased an infant’s risk of being overweight during early adolescence by about 50 percent. (For instance, the risk of adolescent obesity was 50 percent greater for a newborn weighing 5.7 kilograms [12.5 pounds] than for an average-weight infant.) Part of that correlation was related to whether the mother was overweight or obese. When the researchers performed a statistical adjustment to take into account the mother’s BMI, the association between high birth weight and later obesity risk was reduced but not eliminated.

Why does being big at birth—sometimes called “large for gestational age,” or LGA—apparently increase a person’s chances of becoming overweight later on? How could factors that prompt a fetus to grow rapidly in the womb somehow “program” that same individual to store extra pounds as fat during childhood or adulthood? One possibility, supported by some studies, is that whatever leads to rapid growth in the womb produces an infant with a higher percentage of body fat than an average-sized newborn. LGA babies have relatively higher levels of body fat and a lower percentage of lean body mass than babies whose weight is appropriate for gestational age. Having a larger number of fat cells, or bigger fat cells, at birth may predispose an individual to later weight gain.

Another theory suggests that exposure to certain environmental factors during fetal life—for instance, high levels of certain nutrients or of specific hormones, such as insulin or leptin, the hormone made

by fat cells discussed in Chapter 2—somehow presets the fetus’s future metabolism in a way that encourages energy storage. Some research in animals supports this notion: newborn rats fed a high-carbohydrate diet respond by producing excessive amounts of insulin even as adults and become obese. Recent studies implicate leptin as a key player during fetal and infant brain development, influencing the wiring of connections among the nerve cells of the arcuate nucleus of the hypothalamus, a brain area that regulates feeding and body weight. A natural surge in leptin levels occurs in mice during the first week of life and in human fetuses before birth. Oregon researchers found that in newborn mice that are genetically lacking in leptin the hormone surge does not occur and nerve cell connections within the arcuate nucleus do not form properly. When researchers treated such baby mice with leptin timed to mimic the natural surge, nerve cells in the arcuate nucleus responded by growing branches and forming normal-appearing connections. The leptin surge had to occur during a critical period in early life; giving leptin later did not have the same effect. “Nutritional or other environmental factors that suppress leptin during brain development may have lasting effects on an individual’s ability to regulate body weight,” suggested Richard B. Simerly of Oregon Health and Science University, one of the authors of the study.

The quantity and mix of nutrients a fetus gets before birth—or receives as an infant during the newborn period—may lastingly influence the development or function of specialized nerve cells in the areas of the hypothalamus that regulate appetite and metabolism. The findings of German researcher Andreas Plagemann imply that an excess of nutrients, either during fetal life or during early infancy, might influence a person’s lifelong regulation of food intake or energy expenditure. In rats that were “overfed” from birth by being raised in unusually small litters, Plagemann and colleagues were able to demonstrate persistent differences in nerve cell activity and in production of chemical messengers within the appetite-regulating areas of the hypothalamus, and these differences were associated with overeating and obesity in the animals as adults.

We have seen that babies who are especially large at birth have an increased likelihood of later becoming overweight or obese. But what

about babies who are especially small, although not premature? The outlook for such infants seems puzzling. If they are not at risk of obesity, why do they nonetheless seem to be at risk of many of the medical problems we think of as being associated with obesity: heart disease, diabetes, stroke, high blood pressure, and the metabolic syndrome? The answer is not yet clear, but several decades of research have provided a partial explanation. It seems likely that SGA babies, who grow poorly before birth, often grow up to develop these diseases not because they are vulnerable to obesity but for other reasons. Most evidence suggests that they are not at heightened risk of becoming overweight: indeed, they tend to maintain a lower-than-average weight for their height throughout childhood and into young adulthood. But the body fat they do store is more likely to be distributed in an unhealthy pattern than the body fat of people whose weight was average at birth. They are prone to central or abdominal obesity, a pattern in which fat is disproportionately stored on the abdomen or trunk rather than the hips or limbs. As I explained in Chapter 1, this type of fat distribution has undesirable effects on metabolism. A tendency to store fat around the abdominal organs is associated with a heightened risk of the metabolic syndrome and an increased likelihood of developing diabetes and heart disease.

It is not known why SGA babies are more likely than others to develop this pattern of fat distribution. Neither is it clear whether their tendency toward abdominal fat storage is the entire explanation for their vulnerability to the metabolic syndrome and diseases related to it. A related theory holds that undernutrition during fetal life alters the development of the fetus’s pancreas, especially the beta cells that make insulin, affecting the way the body processes glucose and other nutrients. Undernutrition may also affect later patterns of secretion of cortisol, a “stress hormone” that influences metabolism and fat distribution. In addition, SGA babies tend to be born with relatively small kidneys, and their kidneys remain on the small side, since this crucial organ grows its full complement of functioning cells during fetal life. Some experts believe that subtle differences in kidney function may help explain why people who were SGA infants are later at greater risk of high blood pressure.

The effects of undernutrition on a fetus may even vary according to the period of pregnancy during which the developing fetus lacks sufficient nutrients. Almost 30 years ago, in the famous “Dutch famine” study, researchers observed that young Dutch men whose mothers had been exposed to wartime famine during the early weeks of their pregnancies were more likely than other men to be overweight as adults. However, men whose mothers had suffered famine in mid-to-late pregnancy were more likely to have the pattern of central obesity and metabolic abnormalities common to SGA babies. Although experts today say that certain aspects of the Dutch study’s design may have reduced the validity of its findings, it provided some of the earliest suggestive evidence of a possible impact of the fetal environment on later health.

David Barker, a British researcher who has searched extensively for possible links between prenatal factors and disease in later life, explains in his book, The Best Start in Life, that a developing fetus sets an overall growth trajectory early in pregnancy, based on the level of nutrients being delivered via the placenta. If the nutrient supply is low throughout pregnancy, a fetus grows slowly and develops into an infant who is small all over. But a fetus can also to some degree alter the trajectory during pregnancy in response to changes. If the nutrient supply starts out adequate but falls off at some point in middle to late pregnancy, brain growth may be preserved at the expense of the growth of kidneys, liver, muscle, or fat stores. Thus, although two infants may be born with identical birth weights, their body shapes, pattern of growth in the womb, and perhaps many other aspects of their development may be quite different.

During the 1950s, says physician Frank Witter, obstetricians were determined to limit pregnant women’s total weight gain in order to prevent them from having big babies. Larger babies posed a greater risk of complicated deliveries and more often needed to be delivered by cesarean section. But by being overly strict about weight gain, Witter says, doctors unknowingly contributed to a crop of SGA babies who grew up to face an elevated risk of heart disease, diabetes, and other disorders: “Regrettably, that’s what we did. With some of our recommendations in the ’50s, we were starving women to lower the birth

weight.” Some doctors even encouraged pregnant women to smoke, he recalls, reasoning that it would reduce their food intake and keep their babies small. In the 1950s, pregnant women were commonly advised to try to limit weight gain to 20 pounds. “That was more than someone who was heavy should have gained, and less than somebody who was underweight should have gained, and we probably produced some of the problems with central obesity and metabolic syndrome by restricting weight gain,” Witter observes. He adds: “Many people are invested in ‘Big baby, big problem. Little baby, all safe.’ In fact, as we’re looking at them longer and longer, both have problems that are of a different nature. The baby who’s at the least risk for obesity is the child who is appropriate for gestational age in weight.”

So far, it’s unclear from U.S. birth weight statistics what impact the obesity epidemic among American women is having on the size of newborns. As women in America have become more overweight, the frequency of diabetes during pregnancy has increased sharply. About 3.3 percent of U.S. women develop gestational diabetes during pregnancy, according to 2002 figures—a 40 percent increase since 1989. It’s difficult to quantify the health impacts on newborns. Fortunately, the number of very large (“macrosomic”) babies, those weighing at least 4,000 grams (about 8 pounds 13 ounces) has actually fallen slightly since the 1980s, perhaps reflecting improvements in screening and prenatal care for diabetic mothers. Just under 10 percent of infants weighed 4,000 grams or more in 2000. Babies in this weight class are at increased risk for complications during delivery, including birth injuries and asphyxia.

The mean birth weight for a “singleton” baby (not a twin or triplet) born in the United States in 2002 was 3,332 grams, about 7 pounds 5 ounces. Mean singleton birth weight has remained almost constant since 1990, according to government data. In the same period the total number of low-birth-weight infants has increased, but this is primarily because fertility treatments have resulted in more multiple births. Babies who are small because they share the uterus with a sibling are biologically different in some respects from SGA babies who are single-

tons. Researchers do not have enough evidence yet to know whether some infants who are part of multiple births face similar health risks.

Interpreting small changes in infant birth weights over time is complicated, because the average birth weight in the United States is affected by whether the frequency of premature births has changed, and even by whether more full-term infants are being delivered after 39 weeks of pregnancy, as opposed to after 40 or 41 weeks. Michael Kramer, a pediatric researcher at Montreal’s McGill University, has analyzed the pattern of birth weight change in Canada in ways that shed light on what may be happening in the United States. There, too, women are starting pregnancy fatter and are gaining more weight during pregnancy. Rates of diabetes during pregnancy have also increased. But an additional important factor influences birth weights in both countries: as a result of public awareness of the risks of smoking, the number of women who smoke during pregnancy has declined dramatically. In the United States, for example, the proportion fell from almost 20 percent in 1989 to 11 percent in 2002. That’s excellent news. Kramer’s analysis shows that in Canada the decline in smoking during pregnancy has been largely responsible for a heartening drop in the number of babies who are born small for gestational age. Eating appropriately and not smoking are the two most important things a pregnant woman can do to prevent her infant from being born SGA.

In Canada the decline in smoking, coupled with the obesity epidemic among women and the increasing frequency of gestational diabetes, contributed to a rise in mean birth weight of about 50 grams (a little under 2 ounces) between 1981 and 1997. The proportion of infants who are large for gestational age has also increased. The data suggest that in Canada “the size of babies at any gestational age is going up,” Kramer says—and this shift toward heavier birth weights may be adding fuel to the epidemic of obesity there. A similar upward trend in birth weights has not been documented in the United States, but some experts worry that it could be on the horizon.

In addition to not smoking, what can a woman do before and during pregnancy to help ensure the best health outlook for her baby? “Getting yourself in the best physical shape, eating well, and getting appropriate rest—those are things that a woman can do,” says Witter.

At best, that effort should get under way well before pregnancy begins. “I talk with all of my patients about their BMI,” says Joan Loveland, an obstetrician-gynecologist who practices in Chevy Chase, Maryland, a suburb of Washington, D.C. For women whose BMI is over 25, placing them in the overweight category, “I tell them that their exercise program and the nutritional changes that they make are like [taking] blood pressure medication or insulin,” Loveland says. Such behavioral changes are part of the prescription for having the healthiest possible baby. “What’s amazing to me is that many physicians don’t bring this up, because they’re afraid they will offend their patients and the patients won’t come back.”

Ideally, any woman who is contemplating pregnancy should schedule a preconception visit with an obstetrician to go over her health history and talk about nutrition and other lifestyle factors such as smoking, alcohol use, and exercise. If she smokes, now is the time to quit. All women of childbearing age should take a daily multivitamin containing 400 micrograms of folic acid, even if they are not intending to become pregnant, because research has shown that adequate levels of this vitamin, if present at conception and throughout early pregnancy, can prevent spina bifida and anencephaly, which are serious birth defects involving the nervous system. If a woman is overweight, she can maximize the chances of good health for herself and her baby by trying to get as close as she can to a desirable BMI before she starts trying to become pregnant. Despite the fact that almost half of nonpregnant women in the United States report that they’re trying to lose weight, research shows that they aren’t using the strategy most likely to be effective: a combination of limiting calorie intake and increasing physical activity.

Loveland wishes that more of her patients would make a point of coming to see her for routine preconception health counseling. “Women who are seriously considering pregnancy soon, they’re very motivated” to make healthy lifestyle changes, she says. “Gestational diabetes and weight gain during pregnancy really do correlate” with future health risks for the fetus. “That’s a pretty compelling argument for trying to get people fit before they become pregnant.” Once a woman begins trying to conceive, she should not be on a highly restrictive or

unbalanced diet. Her meals and snacks should be healthy, with plenty of fruits and vegetables, fiber and calcium-containing foods (such as dairy products), and she should take a prenatal vitamin daily. She should continue to get regular exercise throughout pregnancy, unless her doctor limits her activity for health reasons. If a woman is especially anxious about the weight gain and changes in body shape that are inevitable with pregnancy, she should be sure to share her fears with her doctor, midwife, or other health care provider.

Recommendations for how much weight to gain during pregnancy should be individually tailored, depending on a woman’s BMI at the start of pregnancy. Current recommendations are based on a large body of evidence indicating that adequate weight gain during pregnancy is very important for fetal growth, while excessive weight gain is associated with an increased risk of pregnancy complications and of having a high-birth-weight (LGA) infant. Generally, the Institute of Medicine and other specialty medical organizations recommend a weight gain of 25 to 35 pounds for women of “normal” weight (defined by IOM prepregnancy guidelines as a BMI between 19.8 and 26.0). Underweight women (BMI below 19.8) should aim to gain 28 to 40 pounds. Overweight women (BMI between 26.1 and 29.0) should aim to gain 15 to 25 pounds, and women with a BMI greater than 29.0 should aim to gain about 15 pounds. Every pregnant woman should talk about her own range of desirable weight gain with her obstetrician at her first prenatal visit. At each prenatal visit, she and the doctor or midwife should track her weight and discuss how she is doing.

Women should keep in mind that staying within the weight gain guidelines is not a guarantee that an infant will grow optimally, since pregnancy complications, birth defects, or various problems with fetal development may affect the pattern of growth regardless of the mother’s diet and other habits. At the same time, many infants are carried to term and born healthy, at a weight appropriate for their gestational age, even though their mothers may have gained too much or too little weight during pregnancy.

Also, many pregnancies are unplanned, so it is not always possible for a woman to prepare herself in advance. But every woman, regardless of her weight at the start of her pregnancy, can do a great deal by

eating sensibly, by not smoking, and by getting regular physical activity to give her baby the best possible start before birth.

Kristen, who attended the breastfeeding support group with healthy 7-week-old Katie, says she knew even at the beginning of her pregnancy that she planned to breastfeed. Her friends had done it, and they encouraged her. Her doctor gave her pamphlets to read. When Katie was born, she told the nurse in the delivery room, “I know you’re supposed to breastfeed within half an hour [of birth]. She got me set up and told me how to do it,” Kristen recalls. “I was kind of nervous about it: is this going to work? But once you get the hang of it, you get a lot better.”

Katie is thriving, growing and developing beautifully, and Kristen knows her breast milk benefits her baby in numerous ways. She even playfully scolds Katie when the milk flows too fast and the baby lets some drool out. “I’m like, ‘Don’t drool it. That’s valuable milk,’” Kristen says. “I’m going to breastfeed this child as long as I possibly can. I would like to say at least six or eight months—that would be my goal.”

Breastfeeding has many proven advantages over formula feeding. Breastfed babies are less likely than formula-fed infants to have ear infections, allergies, vomiting, diarrhea, pneumonia, bronchitis, meningitis, and wheezing. They may also be less vulnerable to sudden infant death syndrome (SIDS). Breastfeeding helps the mother, too: it makes it easier for her to get back to her prepregnancy weight, helps the uterus to recover from pregnancy, reduces her risk of breast and ovarian cancer, and may even improve bone strength and reduce her risk of hip fractures after menopause. And breastfeeding enhances the bond between mother and infants. For all these reasons, the American Academy of Pediatrics recommends breastfeeding, preferably exclusively, until an infant is at least 4 to 6 months old.

“Women sometimes say to me, ‘Isn’t formula just as good as breast milk? Won’t my baby do just as well on formula?’” says lactation consultant Debbie Tobin. “I can’t in good conscience tell a woman that formula is just as good as breast milk because it would be lying, based on the research.”

Many new mothers don’t realize that a woman who breastfeeds is probably also giving her baby a degree of protection against future obesity. Of all the various strategies for preventing childhood obesity that have been investigated so far, the scientific evidence supporting a protective role for breastfeeding is “the most compelling,” in the opinion of pediatric obesity expert William Dietz. Based on research results, Dietz has estimated that universal breastfeeding could prevent about 15 to 20 percent of the obesity cases that currently develop in children before the age of puberty. Although research on this topic has produced a somewhat mixed verdict, a number of recent large studies suggest that breastfeeding provides an infant with at least a moderate reduction in the risk of future obesity.

Choosing whether to breastfeed a newborn—and how long to continue—are personal decisions highly influenced by a mother’s social, cultural, and economic circumstances. Mothers of newborns need education, encouragement, and social support in order to adopt and stay with this method of feeding. Many new mothers are relatively uninformed about how to breastfeed or how to solve problems they may experience with nursing. Many belong to families or cultures in which breastfeeding has not been widely practiced recently, so their own mothers or other relatives can offer little help. Women often need to return to work quickly in order to support their families and find it difficult to adjust their schedules, job demands, and child-care arrangements in order to keep nursing.

Fortunately for children in the United States, breastfeeding has come back into vogue in recent years. In 2001, according to a national survey, almost two-thirds of children under the age of 3 had been breastfed at least as newborns, compared with only 54 percent a decade earlier. Still, most women don’t nurse their babies as long as is recommended by health guidelines. At 6 months of age, only 27 percent of babies were still nursing, and at 1 year only about 12 percent were still receiving any breast milk. These figures lag behind national health goals set by the federal government, which specify that by 2010 the proportion of mothers breastfeeding their infants should rise to 75 percent in the newborn period, 50 percent at 6 months, and 25 percent at 1 year. Despite the AAP’s recommendations, exclusive

breastfeeding declines sharply after the first few months of life: only 8 percent of U.S. infants are being exclusively breastfed by the time they are 6 months old.

For various reasons it has been difficult for scientists to determine whether breastfeeding has an impact on later obesity risk. A major one is that women who breastfeed are more likely than those who do not to be white, thin, well educated, and relatively well off. All these characteristics are also associated with having children with lower rates of obesity. Among racial and ethnic groups in the United States, breastfeeding rates are lowest in African Americans, who also have among the highest rates of obesity. In addition, overweight women in general have more difficulty breastfeeding than slender women, so they are less likely to initiate nursing and more likely to discontinue it.

Correlating breastfeeding and future obesity risk is not a straightforward process. The links between breastfeeding and various other characteristics are considered by scientists to be potential “confounding factors,” which means that such factors may misleadingly influence their results. For example, if a study shows that children who were breastfed as infants have lower BMIs than children who were fed infant formula, it could simply be because the mothers of the formula-fed children were fatter and transmitted a genetic vulnerability to obesity—not because of the way their infants were fed. Moreover, most studies on this topic have been retrospective, or “looking backward”: researchers have assessed the BMIs of a group of children at a particular age and have then depended on parents’ memories to categorize how those children were fed as infants. This kind of study design is considered less accurate and informative than a prospective study, in which children are recruited as infants and then monitored as they grow.

A few years ago nutrition researcher Nancy F. Butte of Baylor College of Medicine reviewed evidence from studies published up to 1999 and determined that most (about a dozen) had found “an insignificant effect” of breastfeeding on later obesity. Only four studies suggested that breastfeeding provided some protection, and Butte concluded that confounding factors (other differences between mothers who chose to breastfeed and those who chose to formula feed) may have been re-

sponsible for the apparent benefit. However, Kathryn G. Dewey, a professor of nutrition at the University of California at Davis, recently assessed a group of larger and newer studies and reached a different conclusion. Dewey reviewed 11 studies, 9 of which were too recent to have been considered by Butte. All met certain criteria, such as size: they had to include at least 100 children in each comparison group. She also looked for studies in which children were at least 3 years old when researchers assessed their weight status, reasoning that breastfeeding’s impact on obesity risk might become more evident in older children. And she sought studies in which researchers had adjusted the design and statistical analysis to try to tease apart the influence of breastfeeding from other potentially confounding factors.

Of the five studies on children aged 3 to 6 that Dewey considered, three found that breastfeeding reduced later risk of overweight. Of the six studies on older children and adolescents, all but one found evidence of a protective effect of breastfeeding. In four of these, this effect persisted after researchers statistically adjusted for the most important potential confounding factor, the mother’s BMI. Several studies also used statistical techniques to control for other potential confounders, such as socioeconomic status, maternal smoking, parents’ level of education, children’s diets and physical activity levels, and additional factors. Among studies suggesting a protective effect, Dewey found that the odds of becoming overweight were about 21 to 34 percent lower in children who had been breastfed than in those who had been formula fed. “Although the effect of breastfeeding may not be large, its role in preventing child overweight could still be of significance” given the breadth and severity of the obesity epidemic currently facing the United States, she concluded. The finding of clearer results in older children also suggested that the beneficial effects of breastfeeding on obesity risk are delayed. “Breastfeeding may have a ‘programming effect’ that does not manifest itself fully until the preadolescent and adolescent growth spurt occurs,” writes Dewey.

How might breastfeeding an infant affect that child’s body size later in childhood? One frequently proposed explanation is that a baby who breastfeeds has much greater control over his or her calorie ingestion, which allows the infant’s body to become better at regulating its own

energy intake. Research findings indicate that nursing babies do manage their own intake; for example, babies whose mothers produce higher-fat breast milk generally consume less than the infants of mothers with lower-fat breast milk, Dewey posits. Formula-fed infants, in contrast, are often encouraged by the parent or caregiver feeding them to consume a specified amount of formula, even after the baby shows signs of losing interest in the bottle. Formula-fed babies consume more calories and gain weight faster than breastfed infants. Formula-fed babies “may be modestly overfed,” says William Dietz. “With my own children, when we shifted them to a bottle, I know that I tended to look at the bottle toward the end of the feeding. It’s easy to push a couple of extra swallows.”

As I mentioned earlier, evidence from animal studies indicates that overfeeding in the newborn period is correlated with increased numbers of fat cells, higher insulin levels, and differences in the development of brain regions that regulate food intake. In humans as in animals, overfeeding during infancy might have a lasting effect on brain development, metabolic hormones, and future regulation of body weight, suggests Nicholas Stettler, a pediatric nutrition specialist at the University of Pennsylvania. Stettler and colleagues showed in a study of more than 19,000 U.S. children that rapid weight gain during the first four months of life was associated with a higher risk of being overweight at the age of 7. In another study of 300 young African Americans in Philadelphia, Stettler and other researchers showed that rapid weight gain in the early months of infancy was associated with a doubling of obesity risk at age 20. Although some infants are probably genetically predisposed to rapid weight gain, it’s possible that overconsumption of calories during a critical period of early life might also program a baby for later obesity risk. “You cannot really overfeed a breastfed infant. Whereas with a formula-fed infant, the mother can potentially override the appetite of the child,” Stettler says. “The hypothesis is that if they learn to rely on outside cues rather than their inner cues, maybe that is learned behavior that is leading them, in the long term, to rely more on the environment rather than on their appetite.”

The higher levels of calories and protein ingested by formula-fed

babies also result in higher levels of insulin, a hormone that stimulates fat deposition, in the infants’ bloodstream. Among Pima Indian children, who as we have seen are genetically, as well as environmentally, predisposed to obesity, one study found that those who had been breastfed were less likely to be overweight or to have diabetes than those who had been formula fed. Hormones or other unidentified substances present in breast milk could also play a role. For example, breastfeeding may help program the developing infant’s response to leptin, the appetite-regulating hormone produced by fat cells, either through direct exposure to leptin in breast milk or by other mechanisms.

The only scientifically definitive way to find out whether a behavior (such as breastfeeding) causes a certain outcome (such as reducing obesity) is to conduct what’s called a randomized controlled trial: recruit a large cohort of pregnant women, randomly divide them into two groups similar in racial, ethnic, and socioeconomic makeup, and assign the women in one group to breastfeed and those in the other group to formula feed their infants. But such an experiment would be unethical because breastfeeding is already known to be superior to formula feeding for other reasons. However, an ongoing European trial may be the next best thing. Known as the PROBIT trial, it enrolled more than 17,000 women and their infants in the Republic of Belarus during 1996 and 1997. All of the mothers participating in the trial intended to breastfeed. Women randomly assigned to one group received an experimental intervention aimed at promoting and supporting breastfeeding; women assigned to a second group received standard medical care. Throughout the 12-month period, women in the intervention group were significantly more likely to continue breastfeeding than those in the control group. Researchers are continuing to monitor the health and BMIs of the children, and Stettler is hopeful that future data from the PROBIT trial may help answer the question of whether breastfeeding really helps protect against later obesity.

In the meantime, what recommendations for mothers and infants can be made, based on research so far? As pediatricians “we recommend breastfeeding for at least four months and possibly exclusively for six months,” says Stettler. “That’s really the only safe recommenda-

tion we can make so far. I wouldn’t make the recommendation, with formula-fed infants, to restrict their intake until we know whether the association we’ve shown [between formula feeding and a possibly higher risk of later obesity] is explained by underlying genetic factors or whether there are other causal relationships.”

Women who are planning to breastfeed should ask their obstetricians to check their breasts and nipples during pregnancy. Certain conditions, such as nipples that are inverted or retracted, sometimes make nursing more difficult, but steps can be taken to improve such factors. Many hospitals also offer a breastfeeding class for pregnant women as part of their childbirth and parenting education programs. Most also offer breastfeeding classes and individual coaching for new mothers while they are in the hospital after childbirth. A new mother should take advantage of such opportunities to get all the instruction and support she can before taking her baby home.

AAP guidelines recommend that breastfeeding should begin as soon as possible after birth and that newborn infants should “room in” with their mothers rather than being kept in a nursery. Newborns should be nursed whenever they show signs of hunger and initially will need to feed 8 to 12 times every 24 hours. The guidelines also recommend formal evaluation by a trained observer of how well breastfeeding is going before a mother and her infant are discharged from the hospital and again in the early days at home. Breastfed infants, according to the guidelines, should not be given other fluids as supplements unless there is a medical reason for it. Supplements and pacifiers should be avoided if possible, or at least until breastfeeding is well established. “Exclusive breastfeeding is ideal nutrition and sufficient to support optimal growth and development for approximately the first six months after birth,” the guidelines state. Breastfed infants under 6 months old generally do not need water, juice, or other foods.

Some medical experts argue that the health benefits of breastfeeding are already so well established that there is no point in spending additional research funds to try to quantify breastfeeding’s benefits in preventing obesity, which may turn out to be modest compared to the impact of changing other behaviors that are fueling the epidemic. A recent exchange at a medical conference between Michael

Kramer of McGill and Matthew Gillman of Harvard illustrates the debate on this point. Arguing that breastfeeding probably plays a minor role in obesity protection, Kramer said, “If you had everybody in the United States exclusively breastfeeding for six months … we would still have an obesity epidemic in this country.” Gillman said he agreed. But, he added, “Just think how much the increase in obesity would be worse if we hadn’t seen the increase in breastfeeding over the last 20 years!”

Breastfeeding may have yet another benefit for babies, according to the surprising results of research by Julie A. Mennella, a scientist at the Monell Chemical Senses Center in Philadelphia. Mennella studies the role of early experiences in the development of food and flavor preferences during infancy and childhood. Her findings provide fascinating evidence of how clearly certain smells and tastes are perceived by very young infants. Mennella’s work and that of some other scientists suggests that certain foods impart a perceptible flavor to breast milk and even to the fluid that surrounds the fetus before birth, thereby providing the fetus and the breastfeeding infant with a foretaste of foods that are important parts of the mother’s diet, foods that are prepared and served in that infant’s family and culture. Such findings suggest that food exposures during early infancy and even before birth may help program a child’s eventual dietary likes and dislikes. “Breast milk is almost like a flavor bridge from what the infant experienced in the womb to the foods at the table,” Mennella says.

In evolutionary terms, taste and smell are the oldest senses, linked to deep, primitive areas of the brain. For animals and humans, the proper functioning of these senses contributes to survival, since they help an individual seek out and eat foods that are safe sources of energy and reject others that are potentially dangerous or poisonous. “These senses are well developed in utero,” Mennella says. “However, the infant and child are living in their own sensory world.” Because the senses of taste and smell continue to develop as a child grows, babies and children apparently perceive some smells and flavors more strongly than adults and others less strongly.

Taste refers to the sensory experience that occurs when chemicals stimulate the taste receptors, specialized groups of cells distributed on the tongue and in other parts of the mouth and throat. Each of the five types of receptors detects a different taste: sweet, salty, sour, bitter, and umami (or glutamic acid, a natural substance found in such foods as asparagus, cheese, meat, tomatoes, and kombu, a type of seaweed used in traditional Japanese cuisine that researchers employed to identify the source of the distinctive umami taste).

Smell or olfaction is a separate sense caused by chemical stimulation of olfactory receptors, structures located on a small area of tissue in the nasal cavity. Hundreds and perhaps thousands of odors can be detected by humans, thanks to the existence of a huge number of unique receptor proteins. The flavor of a food or beverage depends on a combination of its smell and taste. While the primary taste of a food may be salty or sweet, its odor is what allows the person eating it to make fine flavor distinctions, such as being able to tell whether a piece of fruit is a strawberry or raspberry, or whether an ice cream is vanilla or chocolate. Smell’s key contribution to our perception of flavors explains why foods taste flat to us when we have a cold.

Taste buds are well developed by the midpoint of pregnancy, and there is some evidence that fetuses can detect sweet, sour, and bitter. Newborn infants will relax their faces and suck when offered a sugar solution and will grimace and even stick out their tongues when given a solution containing citric acid (sour) or quinine (bitter). Infants’ preference for sweet-tasting liquid is an adaptive trait, since human breast milk is noticeably sweet. In contrast, the ability to taste salt does not develop until about 4 to 6 months of age, and perceptions of salt and bitter continue to change during infancy and childhood.

Although an individual’s basic taste preferences are probably partly determined by genes, there is evidence that early experiences, even before birth, can modify them. For example, college students whose mothers had suffered frequent morning sickness during pregnancy (and who, as a result, probably spent part of their pregnancies mildly depleted of water and salt) showed a stronger preference for salty foods than students whose mothers had not had morning sickness. In general, children prefer saltier foods than adults, but experience with salty foods in childhood tends to strengthen an individual’s liking of salt.

Taste is not the only sense that develops before birth. Newborn infants are sensitive to a variety of odors and respond with changes in facial expression, body movements, and heart and breathing rate. An infant can recognize her or his mother by smell alone, and breast-fed infants can distinguish a breast pad worn by their mothers from pads worn by other lactating women. If certain foods, such as garlic or spices, are eaten by a pregnant woman, their odors permeate the amniotic fluid that bathes the fetus. Animal research suggests that foods to which a fetus is exposed in this way may be preferred by the infant after birth.

The foods a mother eats also flavor her breast milk. Garlic, anise, carrot, mint, blue cheese, vanilla, and undoubtedly many other substances can be detected by a breastfeeding infant. Babies even seem to enjoy some of this variety. Mennella and a colleague, Gary Beauchamp, showed that breast-fed infants stayed attached to the nipple longer after their mothers had eaten garlic or vanilla than after their mothers had eaten bland foods. “The flavor varies with the time of day and with what the mother is eating,” notes Mennella. She theorizes that this physiological flavoring of breast milk may serve an adaptive function. Since the infant already recognizes and enjoys the underlying taste of the milk, faintly flavored breast milk may promote the baby’s acceptance of certain foods later on. She points out, for example, that babies are quicker to accept infant cereal if it is prepared using their mothers’ milk. Commercially produced infant formula, in contrast, always tastes the same. It does not expose a baby to new flavors.

In an intriguing test of this theory, Mennella and colleagues conducted a study in which pregnant women who intended to breastfeed their babies were assigned to one of three groups. One group drank carrot juice regularly during the last part of pregnancy; a second group drank carrot juice regularly during the first two months of breastfeeding; a third group (the control group) avoided carrot juice and drank water. When the babies were old enough to be introduced to solid food, their reactions to carrot-flavored cereal versus plain cereal were videotaped and analyzed. Those who had been exposed to the flavor of carrot juice either before birth (in amniotic fluid) or as infants (in breast milk) were less likely to grimace in response to carrot-flavored cereal than were babies in the control group, and they were

rated by their mothers as enjoying the carrot-flavored cereal more than plain cereal. A study by another pair of researchers found that breastfed babies more readily accepted a new vegetable than did formula-fed ones. Mennella suggests that this willingness could be due to their having tasted the specific flavor of the vegetable in breast milk, to the fact that they were simply more used to tasting a variety of flavors, or both.

The idea that flavors in breast milk might educate a baby’s palate and help establish lifelong food preferences gives breastfeeding a layer of significance beyond its established health benefits. “Odors and flavors that have acquired meaning early in life have long-lasting responses,” says Mennella. “These senses can be trained. They are a reward system that encourages us to seek out pleasurable sensations. They underlie the strength of preferences for foods and eating habits.”

During their first two years of life, babies and toddlers make the transition from getting their nourishment from breast milk or infant formula to eating “table foods” like the rest of the family. Their early experiences with food are vital to the development of dietary preferences and eating habits that are likely to persist, at least to some extent, throughout their lives. Ample evidence exists showing that many U.S. families are not doing nearly enough to make sure that their infants or toddlers are learning to eat a healthy diet during this critical time. The Feeding Infants and Toddlers Study (FITS), a recent national survey sponsored by the Gerber Products Company, found that as toddlers in the United States approach their second birthday, their diets start to look more and more like the unhealthy, highly processed, calorie-dense diets of older children and adults. Their intake of yellow vegetables like carrots and squash drops dramatically. French fries become the “vegetable” they most often consume. One-third of them eat no fruit on a typical day, but one-quarter drink soda or a sugar-sweetened beverage. Says Barbara Devaney, a policy researcher who participated in the FITS survey, “We may not be able to affect what infants and toddlers are eating without addressing what older children and adults are eating.”

When the time comes to introduce a baby to solid foods, how should parents handle this important and exciting period in their

child’s development? What can they do to promote good health and encourage lifelong healthy eating habits? To begin with, they should not start solid foods too soon. Infants younger than about 4 months old lack the mouth and tongue coordination needed to swallow solids, and could choke even on infant cereal, which has traditionally been the first solid food a baby is offered. In addition, there is evidence that too early an exposure to foods other than breast milk or infant formula increases a baby’s risk of developing food allergies. Some studies have even suggested that babies given infant cereal before the age of 3 months may have an increased likelihood of producing antibodies (chemicals made by the immune system) against the beta cells of the pancreas. Such antibodies may be involved in causing damage to these insulin-producing cells, increasing an individual’s risk of developing type 1 diabetes.

Many new mothers are urged by relatives and friends to give their babies cereal when the infants are as young as a few weeks old, in the mistaken belief that it will satisfy the baby’s appetite better than breast milk or formula and make the baby sleep through the night. Such beliefs are well intentioned but misguided. Solid foods do not improve the diet of very young infants or make them sleep better. In the 1980s Cleveland researchers scientifically studied the question of whether putting cereal in babies’ bottles can make healthy young infants sleep through the night. Of a total of 106 infants, they randomly assigned half to a group that began receiving rice cereal at bedtime (1 tablespoon per ounce in a bottle) at 5 weeks of age, and the others to a group that received the bedtime cereal at 4 months old. A parent or other caregiver then recorded each baby’s sleep pattern during one 24-hour period each week, from the ages of 4 weeks to 21 weeks. Sleeping through the night was defined as sleeping at least eight consecutive hours with the majority of that time occurring between midnight and 6 a.m. There was no statistically significant difference in sleep patterns or sleep duration between the two groups, nor was there any consistent trend.

Babies do not need solid food before the age of 6 months; breast milk or formula fulfills their nutritional needs. The earliest age that cereal introduction can ever be considered appropriate, according to

feeding guidelines, is at 4 months old. The American Academy of Pediatrics recommends exclusive breastfeeding, if possible, for the first 6 months of life and gradual introduction of solid foods between the ages of 6 months and 1 year old. (If a mother decides to stop breastfeeding before her infant is 6 months old, the baby should receive only formula until the age of 6 months.) Fruit juice is also not recommended for infants younger than 6 months old.

Vegetables and fruits are traditionally the first foods to which babies are introduced after infant cereal. Pureed vegetables and soft or pureed fruits (either prepared at home or bought as commercial baby food) are nutritious and full of vitamins. As we already know, vegetables and fruits are pillars of a healthy diet for children and adults. Pediatricians often recommend introducing one new vegetable or fruit at a time and offering it for several days in a row. Part of the reason for this approach is that babies—as well as toddlers and some older children—are often slow to accept a new food, a response known as “neophobia,” or fear of the new. On the first try a baby may take a mouthful, twist his face into an expression of apparent disgust, and stick out his tongue, letting the green bean paste drip gently down his chin. This doesn’t mean the baby will never again eat green beans. “Really, you expect children, when they taste something strange, to kind of screw up their faces,” says William Dietz. “Parents often mistake the child’s expression of experiencing a novel food as an expression of dislike. Absolutely, repeated introductions do help a child get used to food.”

“Probably because ingesting new substances is a risky business, most new foods are not immediately accepted,” writes psychologist Leann Birch of Pennsylvania State University. A distrustful attitude toward new foods on the part of children may even be a trait that was favored during human evolution, because it was advantageous to survival. Repeatedly exposing a child to a food, however, eventually leads to acceptance in many cases. Children are also more likely to accept a food if they see family members or other children eating it. Birch found that preschool-age children who initially disliked certain vegetables began eating them after they saw other children eat them in a childcare setting. Young children in certain cultures learn to accept even

spicy, “hot” foods after they see family members eating and enjoying them.

As I mentioned previously, young children’s preferences for sweet and salty foods do not have to be learned. A preference for high-fat, calorie-dense foods such as pizza or French fries is apparently not innate, but research has shown that children quickly become conditioned to prefer these types of foods, probably because they’re filling and rapidly satisfy hunger. In one study children were given repeated opportunities to eat soup and yogurt. Some versions of these foods were higher in fat or carbohydrates than others, hence more calorie dense, and their flavors also differed. Although children in the study initially liked the lower-calorie and higher-calorie soups and yogurts equally well, after repeated exposures they became conditioned to prefer the higher-calorie versions, especially when they were hungry.

Young children do vary individually in their liking of high-fat foods. In one study of 3- to 5-year-olds, Birch and a colleague found that youngsters who most strongly preferred high-fat foods also tended to be the fattest in the group and to have parents with the highest BMIs. It’s unclear whether these children’s preference was genetically based or whether it was influenced by exposure to high-fat foods in the home—or perhaps by environmental factors during life in the uterus.

Birch says that for parents of young children the take-home message of much of this research is that they need to focus on giving their babies and toddlers plenty of opportunities to eat and experience healthy foods, especially fresh vegetables, fruits, and whole grain products. Such foods form the foundation of almost all widely accepted schemes for healthy eating. “Kids don’t have to learn to like sweet and salty,” she says. “But if they are not offered a variety of things from the bottom of the Food Guide Pyramid,” such as fresh vegetables, whole grain foods, and fresh fruits, they won’t necessarily learn to like those foods. “Parents should work to get those things into the diet, rather than restricting it to foods that the child is programmed to like and accept.”

This does not mean that parents should try to force a baby or toddler to eat a certain vegetable. That’s a recipe for failure. British researcher Jane Wardle suggests playing a “tasting game” in which the

parent gives the child the tiniest possible piece of the food that is being tried. The child can taste it and then may decide whether to eat it or spit it out. Wardle says research with families using this playful approach has shown that it helps persuade toddlers and young children to try and accept new foods.

It may be a good idea to get a baby accustomed to a varied diet. In one study, Julie Mennella and a colleague assessed several different strategies for introducing 5-month-old, formula-fed babies to carrots. All the infants were given carrots for the first time on day 1 of the study. On days 2 through 10, one group ate pureed potatoes, a second group ate carrots, and a third group, the variety group, ate one of three different vegetables—peas, potatoes, or squash. On day 11, when all the babies were given carrots, the carrots were eaten with equal enthusiasm by the all-carrot group and by the variety group. Group 1, the potato group, was less likely to accept them. Mennella also found that the variety group was the most willing to accept a new food (chicken) and that giving babies some fruit each day enhanced their willingness to accept carrots. This latter finding tends to disprove the belief, common in some communities, that if infants are given fruits at the same time as vegetables, their preference for the sweet fruits will make them more likely to reject the vegetables. Perhaps experience with variety makes babies less likely to fear and reject what is new. Summarizing her findings, Mennella writes, “Early experience with a diversity of flavors may have led to an increased readiness to accept unfamiliar flavors.”

Researchers who have studied various parental feeding styles suggest that mealtimes should be an occasion for positive, cheerful interaction between parents (or caregivers) and babies. A baby should not be left alone with a propped bottle or fed by a distracted adult who does not talk to it or make eye contact. Babies and toddlers should be allowed to refuse to eat a certain food, and they should be the ones who decide when to stop eating. Parents should never use foods as bribes or rewards, nor should they withhold them as punishments. Actively and repeatedly urging a young child to eat a certain food (“Come on, Tommy, eat your broccoli!”) is likely to backfire.

At the same time parents must be in charge of what a baby is of-

fered, and all the presented foods should be healthy and developmentally appropriate for the age of the child. Just because a baby reaches for a hot dog or a candy bar does not mean she should have one! The findings of the Gerber FITS survey suggest that many U.S. parents are regularly feeding babies and toddlers highly processed sweet and salty foods. More than 60 percent of 12-month-olds whose parents were surveyed for the study were being fed dessert or candy at least once a day, and 16 percent ate a salty snack. By the age of 15 months, 30 to 40 percent were drinking sugary fruit drinks daily. Daily fare that contains high levels of sugar and salt is likely to reinforce young children’s innate liking for those tastes and may discourage the development of preferences for healthy foods.

Although 100 percent fruit juice is preferable to fruit drinks because it is rich in vitamins, minerals, and nutrients, it is high in sugars and not as good a fiber source as a piece of fresh fruit. Too much fruit juice may encourage a preference for sweet beverages as a thirst quencher instead of water. The American Academy of Pediatrics suggests limiting juice consumption to 6 ounces a day for children aged 1 to 6 years and no more than 12 ounces a day for those aged 7 to 18.

At a recent medical conference, Jane Wardle contrasted the child-feeding styles of parents in France and the United States, suggesting that cultural differences in how children learn about food may be one reason why the French have much lower rates of obesity than Americans do. Citing research by social historian Peter N. Stearns, she noted that in France the emphasis is on training the appetite and teaching children to learn to eat the foods adults eat. In French families, fussy eaters are generally not allowed to demand substitute foods; instead, if children do not like what is on the menu, they go hungry until the next meal. In the United States, Wardle observes, parents are more concerned with pleasing children and worry more about whether they are getting enough to eat. Snacks are much more commonly offered than in France, and family food choices tend to be child centered. Wardle suggests that American parents might do well to adopt a more typically French attitude in teaching their young children how to eat. “We are responding to the child’s instincts,” she says, “but on the other hand, we are the environment that entrains the child’s instincts.”