Chapter 1
Americans Weigh in Over Time

More die in the United States of too much food than of too little.

—John Kenneth Galbraith, The Affluent Society (New York: Houghton Mifflin Co., 4th ed., 1984)

Americans in 2006 are fatter than ever, the heaviest since the government started tracking patterns of body weight for the U.S. adult population in the first half of the twentieth century. An estimated 127 million adults weigh more than is considered healthy, and of this group, more than sixty million are considered obese. According to the Centers for Disease Control (CDC) and U.S. Surgeon General, overweight and obesity afflict more than two-thirds of Americans and constitute a public health problem of epic and epidemic proportions. (An epidemic is not a specific number of cases of a disease or condition; an epidemic exists when the number of cases exceeds that expected based on past experience for a given population.) Despite billions of dollars spent on diet programs, overweight and obesity are widespread and increasingly prevalent throughout the United States.

Although Americans' body weights had been incrementally increasing during the last century, overweight and obesity skyrocketed between 1985 and 2003. The CDC reports that during that time obesity among adults more than doubled, and obesity among adolescents tripled. Normal-weight adults are now a minority in the United States; nearly one-third of the adult population is obese, and childhood obesity is at an all-time high. In 1991, four states had obesity prevalence rates of 15%-19%, and no states had rates at or above 20%. By 2003, fifteen states had obesity prevalence rates of 15%-19%; thirty-one states had rates of 20%-24%; and four states reported rates of more than 25%. (The prevalence rate is the number of cases of a disease or condition present during a specified interval of time, usually a year, divided by the population.) Figure 1.1 maps the geographic distribution of obesity throughout the United States in 1991, 1996, and 2004.

The prevalence of obesity varies by state. An analysis of 2004 data from the CDC's Behavioral Risk Factor Surveillance System (F as in Fat: How Obesity Policies Are Failing in America 2005, Washington, DC: Trust for America's Health, 2005) revealed that Colorado reported the lowest percentage of obesity (16.8%) in 2004, followed by Massachusetts (18.4%), Vermont (18.7%), Rhode Island (19%), Montana (19.7%), and Connecticut (19.7%). More than 25% of adults in twelve states were obese in 2004. Mississippi reported the highest rate of obesity (29.5%), followed by Alabama (28.9%), West Virginia (27.6%), Tennessee (27.2%), and Louisiana (27%).

Analysis of self-reported data collected by the CDC Behavioral Risk Factor Surveillance System reveals that the obesity epidemic affects men and women of all ages, races, ethnic origin, smoking status, and educational attainment. Table 1.1 shows that while the prevalence of obesity among U.S. adults disproportionately affects older age groups, African-Americans, and Hispanics, and declines with increasing educational attainment, from 27.4% among people with less than a high school education to 15.7% among those who attended college, no group remains untouched by this epidemic.

Researchers Khoa Dang Truong and Roland Sturm looked at various sociodemographic groups to assess whether U.S. weight-gain trends varied in response to education, income, race/ethnicity, and gender. An analysis of data from the Behavioral Risk Factor Surveillance System found that overall, each sociodemographic group experienced generally similar weight gains, although women gained more weight than men. On average, individuals in the lowest-income group gained as much weight as those in the highest-income category, and there were no differences across racial or ethnic groups except that non-Hispanic blacks gained more weight than other groups. The only significant difference in terms of weight-gain trends was that people with college degrees gained less weight than did others ("Weight Gain Trends across Sociodemographic Groups in the United States," American Journal of Public Health, vol. 95, no. 9, September 2005).

In the United States obesity is the second-leading cause of preventable death after smoking. In 2005 there were about 35,000 more deaths attributable to smoking than to obesity. However, obesity is poised to overtake smoking as the leading cause of preventable death in 2006, according to Ali H. Mokdad and others in their study, "Actual Causes of Death in the United States, 2000," published in the Journal of the American Medical Association in March 2004. There is conclusive scientific evidence that mortality (death) risk increases with increasing weight and that even slightly overweight adults—people of average height who are ten to twenty pounds above their ideal weights—are at increased risk of premature death. The rising prevalence of overweight and obesity not only foretell increasing adverse effects on health and longevity but also guarantee increased costs for medical care. Overweight and obesity increase the risk of developing a range of ailments including heart disease, stroke, selected cancers, sleep apnea (breathing stops for multiple, brief periods during sleep), respiratory problems, osteoarthritis (loss of joint bone and cartilage), gallbladder disease, fatty liver disease, and Type 2 diabetes. (Insulin is necessary for the body to be able to use sugar, the basic fuel for the cells in the body. People with diabetes do not produce enough insulin or their cells are resistant to the effects of the insulin.) The CDC estimates that the annual medical costs of an obese person are nearly 38% higher than those incurred by a person of normal weight.

Overweight and obesity also exact a personal toll, with affected individuals at increased risk for emotional, psychological, and social problems. Overweight children, teens, and adults suffer from depression, low self-esteem, and other mental health and emotional problems more than their normal-weight counterparts. Along with a physical inability to participate in many activities, people who are overweight or obese may encounter weight-based stigmatization, bias, and discrimination in school and at the workplace and may be excluded from opportunities for socialization.

TRENDS IN U.S. BIRTH WEIGHTS

Americans are not born overweight. In fact, the mean birth weight of infants born as singletons (births of one infant as opposed to twins or other multiple births) has steadily declined since 1990, according to the CDC National Center for Health Statistics. In 2003 the mean birth weight of all singletons was approximately 7 pounds, 5 ounces (3,325 g), and the average white non-Hispanic singleton (3,384 g; 7 pounds, 7 ounces) weighed a full nine ounces more than the average non-Hispanic black singleton (3,122 g; 6 pounds, 14 ounces). (See Table 1.2.) The percent of infants born with higher-than-average birth weights (4,000 g or more, or at least 8 pounds, 13 ounces) has been declining for more than a decade, as reported by the U.S. Department of Health and Human Services in Pediatric Nutrition Surveillance: 2003 Report. In 1994, 8.5% of births were at 4000 g or above, compared with 7.3% in 2003.

While ideal birth weight varies based on the expectant mother's ethnicity, for women in the United States, the average ideal birth weight is approximately 7.5 pounds, close to the average weight of singletons born in 2003. In the United States, the percent of babies born with low birth weight (LBW)—less than 2,500 g (5 pounds, 8 ounces) has risen steadily since the mid-1980s. (See Figure 1.2.) According to data from the CDC's National Center for Health Statistics the LBW rate rose from 7.6% in 2000 to 7.9% in 2003, the highest level reported in more than three decades. The percent of infants with very low birth weights (VLBW; weighing less than 1,500 g or 3 pounds, 5 ounces) remained nearly steady between 2000 (1.43%) and 2003 (1.45%).

LBW and VLBW are major predictors of infant morbidity (illness or disease) and mortality. For LBW infants, the risk of dying during the first year of life is more than five times that of infants born at normal weights; the risk for VLBW infants is nearly 100 times higher. The risk of delivering an LBW infant is greatest among the youngest and oldest mothers; however, many of the LBW births among older mothers are attributable to their higher rates of multiple births. CDC data for 2002 showed that multiples accounted for nearly two-thirds of all LBW infants delivered to mothers age forty-five and older that year. Close to 10% of singletons born to mothers age forty-five or older were LBW as were 8.7% of births to mothers less than twenty years old in 2002.

In 2003, 324,064 babies were born at low birth weights in the United States, according to the CDC (Births: Final Data for 2003). That number represented 7.9% of all births. However, the percent of LBW babies varied by state. In 2003 Alaska reported the lowest percent (5.2%) and Wyoming the highest (8.9%) of LBW births to non-Hispanic white mothers. Of the states that reported more than 1,000 births to non-Hispanic black women, LBW ranged from a low of 9.1% in Alaska to a high of 16% in New Mexico.

Birth Weight Influences Risk of Disease

Although the relationship between birth weight and development of disease in adulthood is an emerging field of research, and scientists cannot yet fully explain how and why birth weight is a predictor of health and illness in later life, mounting evidence indicates that both low birth weight and higher-than-average birth weight are linked to future health problems. Research reveals that LBW infants are more likely than normal-weight infants to develop disease in later life. Male infants with LBW who gain weight rapidly before their first birthdays appear to be at the highest risk. Investigators hypothesize that LBW infants have fewer muscle cells at birth and that rapid weight gain during the first year of life may lead to disproportionate amounts of fat to muscle and above average body mass. People with LBW who later develop above average body mass are at increased risk for developing diseases such as Type 2 diabetes, hypertension (high blood pressure), cardiovascular disease, and stroke. A 1997 study published in the British journal The Lancet examined the medical records of 13,249 men and found the risk of dying from stroke or heart disease was highest for those who weighed 5.5 pounds at birth or less. Those who weighed more than 8.5 pounds at birth had the lowest rate of mortality from heart attack or stroke.

A 2005 study published in the American Heart Association Journal Circulation found an inverse relationship between birth weight and cardiovascular disease (heart disease and stroke). In general, rates of both coronary heart disease and stroke decreased with increasing birth weight. The association was strong, did not depend on adjustment for size in later childhood; and was independent of social class and other maternal and pregnancy characteristics.

Low birth weight also was linked to childhood asthma in a U.S. study published in a 2001 issue of Archives of Pediatrics and Adolescent Medicine, which found that babies born at 5.5 pounds or less faced the greatest risk of respiratory complications such as asthma. Research also has demonstrated that both LBW and abnormally high birth weight are associated with risk of developing diabetes later in life.

Evidence also indicates that birth weight is related to risk of developing breast cancer. Valerie A. McCormack and her colleagues at the Department of Epidemiology and Population Health, London School of Hygiene and Tropical Medicine, investigated whether size at birth and rate of fetal growth influenced the risk of developing breast cancer in adulthood. The results of the study were published in "Fetal Growth and Subsequent Risk of Breast Cancer: Results from Long Term Follow up of Swedish Cohort" (British Medical Journal, vol. 326, no. 7,383, February 2003). By examining birth and medical records of 5,358 singleton females born from 1915 to 1929, the investigators determined that size at birth was associated with breast cancer in premenopausal (the stage of reproductive life immediately before the onset of menopause) women age fifty or younger—the larger and longer the baby, the greater the risk. Birth weight or size was not associated with rates of breast cancer in postmenopausal women. Among premenopausal women who weighed 4,000 or more grams at birth (8 pounds, 14 ounces) rates of breast cancer were 3.5 times higher than those women who weighed less than 3,000 g at birth (about 6 pounds, 10 ounces). The investigators concluded that size at birth, including birth weight, length, and head circumference, is associated with risk of breast cancer in women under age fifty.

An analysis performed by Canadian researchers found that infants born either prematurely or with an extremely low birth weight (ELBW; 800 g or 1 pound, 12 ounces) were significantly more likely to suffer a lower level of fitness later in life, including less strength, endurance, and flexibility, and a greater risk of health problems as adults. When compared with teens born at normal weights, the ELBW teens had lower aerobic capacity, grip strength, leg power, and vertical jump. They were unable to perform as many push-ups, had less abdominal strength as measured by curl-ups, showed less flexibility in their lower backs, and had tighter hamstrings. The ELBW teens reported less previous and current sports participation, lower physical activity level, and poorer coordination compared with term-born control subjects. ELBW teens also had more trouble maintaining rhythm and tempo than their peers who were born at normal weights (Marilyn Rogers et al., "Aerobic Capacity, Strength, Flexibility, and Activity Level in Unimpaired Extremely Low Birth Weight [≤800 g] Survivors at Seventeen Years of Age Compared with Term-Born Control Subjects" Pediatrics, vol. 116, no. 1, July 2005).

The only action able to alter the birth weight of an infant is to modify weight gain during pregnancy. In 2006 health professionals concur that for normal-weight women the ideal weight gain during pregnancy ranges from twenty-five to thirty-five pounds of fat and lean mass. Further, research published in 2003 revealed that a newborn's birth weight and mother's post-pregnancy weight are influenced not only by how much weight is gained during pregnancy, but also by the source of the excess weight. In "Composition of Gestational Weight Gain Impacts Maternal Fat Retention and Infant Birth Weight" (American Journal of Obstetrics and Gynecology, vol. 189, no. 5, November 2003), researcher Nancy F. Butte and her colleagues conducted body scans of sixty-three women before, during, and after their pregnancies and recorded changes in women's weight from water, protein, fat, and potassium—a marker for changes in muscle tissue, one component of lean mass. The researchers found that only increases in lean mass, and not fat mass, appeared to influence infant size. Independent of how much fat the women gained during pregnancy, only lean body mass increased the birth weight of the infant, with women who gained more lean body mass giving birth to larger infants.

Breastfeeding is linked to improved health outcomes for all infants; however, it is especially advisable for LBW infants. For these infants, breastfeeding can reduce the risk that they will develop chronic diseases in adulthood by preventing the development of above average body mass. LBW infants who are breastfed for at least twelve months have about half the risk of developing above-average body mass during childhood.

FIRST WEEK OF LIFE MAY DETERMINE ADULT OBESITY

Research has demonstrated that low birth weight and low weight gain during infancy are associated with coronary heart disease. Similarly, rapid weight gain in infancy has been shown to predict obesity in childhood. In 2004 research funded by the National Institutes of Health (NIH) and conducted at the Children's Hospital of Philadelphia, University of Pennsylvania School of Medicine, and the Fomon Infant Nutrition Unit at the University of Iowa sought to determine which periods of weight gain in infancy might be associated with adult obesity.

The investigators reviewed data for 653 subjects who had been weighed on seven occasions during infancy and were contacted when they were young adults, ages twenty to thirty-two, when they again reported their heights and weight. The researchers pinpointed the period between birth and age eight days as potentially critical since weight gain during the first week of life was associated with adulthood overweight status. The formula-fed babies who gained weight rapidly during their first week of life were significantly more likely to be overweight decades later. They concluded that, "In formula-fed infants, weight gain during the first week of life may be a critical determinant for the development of obesity several decades later." The investigators also observed that their findings reinforced the American Academy of Pediatrics recommendation that infants should exclusively be breast-fed for the first six months of life. Among the numerous health benefits associated with breastfeeding is the fact that breast-fed babies are much less likely than formula-fed babies to become obese adults (Nicolas Stettler et al., "Weight Gain in the First Week of Life and Overweight in Adulthood: A Cohort Study of European American Subjects Fed Infant Formula," Circulation, vol. 111, no. 15, April 2005).

Another study, conducted by Janis Baird and her colleagues at the MRC Epidemiology Resource Centre at the University of Southampton in England found that big babies who grow quickly in the first two years of life risk being obese in childhood and adulthood. To determine whether obesity may begin in infancy, Baird and her colleagues looked at twenty-four studies that found an association between infant size or growth during the first two years of life and obesity later in life. They found that the heaviest infants and those who gained weight rapidly during the first and second year of life faced a nine-fold greater risk of obesity in childhood, adolescence, and adulthood. Their findings suggest that factors in infant growth are probably influencing the risk of later obesity. Baird and her colleagues do not know why big and fast growing babies had a higher risk of obesity, but they believe that some factors related to how an infant grows are important in influencing their later risk of obesity and suggested that infant feeding, being bottle or breast fed, the timing of weaning and social circumstances were factors that merit further investigation (Janis Baird et al., "Being Big or Growing Fast: Systematic Review of Size and Growth in Infancy and Later Obesity," British Medical Journal, vol. 331, no. 7,522, October 22, 2005).

DEFINING AND ASSESSING IDEAL WEIGHT, OVERWEIGHT, AND OBESITY

Historically, desirable, healthy, or ideal weights have been derived from demographic and actuarial statistics (data compiled to assess insurance risk and formulate insurance premiums). The National Center for Health Statistics compiles and analyzes demographic data—the heights and weights of a representative sample of the U.S. population to develop standards for desirable weights. In 1943 the Metropolitan Life Insurance Company (MetLife) introduced standard weight-for-height tables for men and women based on an analysis of actuarial data. The MetLife weight-for-height tables assisted adults in determining if their weights were within an appropriate range for height and frame size. Revised in 1959 and 1983, the tables were based on actuarial data, in which desirable or ideal weight is defined as the weight for height associated with the lowest mortality rate, or longest life spans, among the client population of adults (policyholders) insured by MetLife.

Although the MetLife and other weight-for-height tables remain in use in 2006, many health professionals and medical researchers believe they have limited utility. Nearly every version of desirable weight-for-height tables shows different acceptable weight ranges for men and women, and considerable debate continues among health professionals over which table to use. The tables lack information about body composition, such as the ratio of fat to lean muscle mass; their data were derived primarily from white populations and do not represent the entire U.S. population; they generally do not take age into consideration; and it is often unclear how frame size was determined. Further, it is now known that ideal, healthy, or low-risk weights vary for different populations, and vary for the same population at different times and in relation to different causes of morbidity and mortality.

The limitations of weight-for-height tables have prompted health-care practitioners and researchers to adopt other measures that allow comparison of weights independent of height and frame across populations to define desirable or healthy weights as well as overweight and obesity. For example, the 2005 Dietary Guidelines

for Americans published jointly by the U.S. Departments of Agriculture (USDA) and Health and Human Services (HHS), and weight-control information published by the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health (NIH) include updated weight-for-height tables for adults that incorporate height, weight, and body mass index (BMI). (See Table 1.3.) Table 1.4 is an example of the weight ranges that are considered underweight, healthy weight, overweight, and obese for a person who is 5′9′ tall.

Overweight is generally defined as excess body weight in relation to height, when compared to a predetermined standard of acceptable, desirable, or ideal weight. One definition characterizes individuals as overweight if they are between ten and thirty pounds heavier than the desirable weight for height. Overweight does not necessarily result from excessive body fat; people may become overweight as the result of an increase in lean muscle. For example, while muscular bodybuilders with minimal body fat frequently weigh more than non-athletes of the same height they are "overweight" because of their increased muscle mass rather than increased fat.

Rather than viewing overweight and obesity as distinct conditions, many researchers prefer to consider weight as a curve or continuum with obesity at the far end of the curve. People who are obese constitute a subset of the overweight population—using this definition it becomes clear that while only some overweight people are obese, all obese people are overweight.

Similarly, there is still no uniform definition of obesity. Some health professionals describe anyone who is more than thirty pounds above his or her desirable weight for height as obese. Others assert that body weight 20% or more above desirable or ideal body weight constitutes obesity. Extreme, or clinically severe obesity, is often defined as weight twice the desirable weight or 100 pounds (45 kg) in excess of the desirable weight. Obesity also is defined as an excessively high amount of adipose tissue (body fat) in relation to lean body mass such as muscle and bone. The amount of body fat (also known as adiposity), the distribution of fat throughout the body, and the size of the adipose tissue deposits also are used to assess obesity because the location and distribution of body fat are important predictors of the health risks associated with obesity. The location and distribution of body fat may be measured by the ratio of waist-to-hip circumference. High ratios are associated with higher risks of morbidity and mortality.

Overweight and obese body types may be characterized as "pear- or apple-shaped," depending on the anatomical site where fat is more prominent. In the apple or android type of obesity, fat is mainly located in the trunk (upper body, nape of the neck, shoulder, and abdomen.) Gynoid obesity or the pear-shape, features rounded hips, and more fat located in the buttocks, thighs, and lower abdomen). Fat cells around the waist, flank, and in the abdomen are more active metabolically than those in the thighs, hips, and buttocks. This increased metabolic activity is thought to produce the increased health risks associated with android obesity. In general, women are more likely to have gynoid obesity. However, those with the android type of obesity are subject to similar health risks as males with android overweight.

There are many ways to measure body fat. Weighing an individual underwater in a laboratory with specialized equipment provides a highly accurate assessment of body fat. By performing hydrostatic or underwater weighing, an examiner obtains an estimate of whole-body density and uses this to calculate the percentage of the body that is fat. First, the subject is weighed on a land scale. The subject puts on a diver's belt with weights to prevent floating during the weighing procedure, sits on a chair suspended from a precision scale, and is completely submerged. When maximum expiration of breath is achieved, the subject remains in this submerged position for about ten seconds while the investigator reads the scale. This procedure is repeated as many as ten times to obtain reliable, consistent values. The weight of the diver's belt and chair are subtracted from this weight to obtain the true value of the subject's mass in water.

Simpler, but potentially less accurate assessments of body fat include skinfold thickness measurements, which involve measuring subcutaneous (immediately below the skin) fat deposits using an instrument called a caliper in locations such as the upper arm. Skinfold thickness measurements rely on the fact that a certain fraction of total body fat is subcutaneous and using a representative sample of that fat, overall body fatness (density) may be predicted. Several skin-fold measurements are obtained, and the values are used in equations to calculate body density. Using a caliper, the examiner grasps a fold of skin and subcutaneous fat firmly, pulling it away from the underlying muscle tissue following the natural contour of the skin. The caliper jaws exert a relatively constant tension at the point of contact and measure skinfold thickness in millimeters. Most obesity researchers believe there is an acceptable correlation between skinfold thickness and body fat—that it is possible to estimate body fatness from the use of skinfold calipers. Skinfold thickness measurements are considered more subjective than underwater weights because the accuracy of measurements of skinfold thickness depends on the technique and skill of the examiner, and there may be variations in readings from one examiner to another.

Another technique used to evaluate body fat is bioelectric impedance analysis (BIA). BIA offers an indirect estimate of body fat and lean body mass. It entails passing an electrical current through the body and assessing the body's ability to conduct the current. It is based on the principle that resistance is inversely proportional to total body water when an electrical current (75 MHz) is applied through several electrodes placed on body extremities. Since greater conductivity occurs when there is a higher percent of body water, and fat cells contain less water than muscle cells, and a higher percent of body water indicates larger amounts of muscle and other lean tissue, people with less fat are better able to conduct electrical current. BIA has been shown to correlate very well with total body water assessed by other methods.

Other means of estimating the location and distribution of body fat include waist-to-hip circumference ratios, or imaging techniques such as ultrasound, computed tomography, or magnetic resonance imaging.

Waist Circumference and Waist-to-Hip Ratio

Along with height and weight, waist circumference is a common measure used to assess abdominal fat content. An excess of body fat in the abdomen or upper body is considered to increase the risk of developing heart disease, high blood pressure, diabetes, stroke, and certain cancers. Like body fat, health risks increase as waist circumference increases. For men, waist circumference greater than 40 inches (102 cm) is considered to confer increased health risks. Women are considered at increased risk when their waist measurements are 35 inches (88 cm) or greater. Figure 1.3 shows how waist circumference is measured to obtain accurate measurements of abdominal girth. Waist circumference measures lose their incremental predictive value in people with a BMI greater or equal to 35 because these individuals generally exceed the cutoff points for increased risk. Table 1.5 shows the relationship between BMI, waist circumference, and disease risk for people who are underweight, normal weight, overweight, obese, and extremely obese.

Waist-to-hip ratio is the ratio of waist circumference to hip circumference, calculated by dividing waist circumference by hip circumference. For men and women, a waist-to-hip ratio of 1.0 or more is considered to place them at greater risk. Most people store body fat at the waist and abdomen (android body fat distribution) or at the hips (gynoid body fat distribution). Interestingly,

while overweight and obesity both increase health risks, body fat that is concentrated in the lower body, such as fat deposits at the hips and thighs, is thought to be less harmful in terms of morbidity and mortality than abdominal fat—excess fat in the upper body.

Body Mass Index (BMI)

BMI is a single number that evaluates an individual's weight status in relation to height. It does not directly measure the percent of body fat; however, it offers a more accurate assessment of overweight and obesity than weight alone. It is a direct calculation based on height and weight, and it is not gender specific. BMI is the preferred measurement of health-care professionals and obesity researchers to assess body fat and is the most common method of tracking overweight and obesity among adults. BMI, calculated by dividing weight in kilograms by the square of height in meters (BMI 1/4 kg/m2) classifies people as underweight, normal weight, overweight, or obese. Table 1.6 shows the formula used to calculate BMI when height is measured in either inches or centimeters and weight is measured in either pounds or kilograms.

The World Health Organization and National Institutes of Health consider individuals overweight when their BMI is between 25 and 29.9, and they are classified as obese when their BMI exceeds 30. Table 1.3 shows the relationship between height, weight, and BMI. Table 1.5 shows the classification of overweight and obesity by BMI and distinguishes between three levels of obesity.

Although BMI is a simple, inexpensive tool for assessing weight, it has several limitations. BMI calculations may deem a muscular athlete overweight, when he or she is extremely fit, and excess weight is the result of a larger amount of lean muscle. It may similarly misrepresent the health of older adults who as the result of muscle wasting—loss of muscle mass—may be considered normal or healthy weights when they may actually be nutritionally depleted or overweight in terms of body fat composition. While it is an imperfect method for assessing individuals, BMI is extremely useful for tracking weight trends in the population.

Definitions and Estimates of Prevalence Vary

Historically, varying definitions of, and criteria for, overweight and obesity have affected prevalence statistics and made it difficult to compare data. Some overweight- and obesity-related prevalence rates are crude or unadjusted estimates; others are age-adjusted estimates that offer different values. Early efforts to track overweight and obesity in the U.S. population relied on the 1959 or 1983 Metropolitan Life Insurance tables of desirable weight-for-height as the reference standard for overweight. During the last three decades, most government agencies and public health organizations have estimated overweight using data from a series of surveys conducted by the CDC's National Center for Health Statistics. These surveys include the National Health Examination Surveys, National Health and Nutrition Examination Surveys (NHANES), and the Behavioral Risk Factor Surveillance System (BRFSS).

Despite changing definitions of overweight and obesity and various methods to track changes in the U.S. population, there is irrefutable evidence that the prevalence of overweight and obesity have steadily increased among people of both genders, all ages, all racial and ethnic groups, all educational levels, and all smoking levels. The prevalence of obesity in the United States was first reported in the National Health Examination Survey in 1960, and subsequent reports were derived from three National Health and Nutrition Examination Surveys: NHANES I, 1971; NHANES II, 1976–1980; and NHANES III, 1988–1994. Most obesity data referenced in the medical literature in 2006 are drawn from the NHANES study conducted between 1999 and 2002, along with several other national studies. Data from the National Health Examination Survey, NHANES I, and NHANES II indicated that the prevalence of obesity was relatively constant from 1960 to 1980; however, the results of the NHANES III indicated a sharp increase in the prevalence of obesity.

Overweight and obesity have steadily progressed at an alarming rate over the course of the past two decades. From 1960 to 2000 the prevalence of overweight (defined as BMI greater than 25 but less than 30) increased from 31.5% to 33.6% in U.S. adults aged twenty to seventy-four. The prevalence of obesity (BMI of 30 or more) during the same years more than doubled from 13.3% to 30.9%, with most of the rise occurring in the past twenty years. From 1988 to 2000 the prevalence of extreme obesity (BMI greater than or equal to 40) increased from 2.9% to 4.7%, up from 0.8% in 1960.

The annual prevalence of obesity among U.S. adults age twenty and older increased from 19.4% in 1997 to 20.6% in 1998, 21.5% in 1999, 21.8% in 2000, 23% in 2001, 23.9% in 2002, and 23.7% in 2003. In 2004, 24.5% of adults were obese, comparable to the 2003 estimate of 23.7%. (See Figure 1.4.)

The prevalence of overweight and obesity generally increases with advancing age, then starts to decline among people over sixty. In 2004, for men and women combined, the prevalence of obesity was highest among adults aged forty to fifty-nine (27.7%) and lowest among adults aged twenty to thirty-nine (21.2%). There was no significant difference in the prevalence of obesity between men and women in all three age groups. (See Figure 1.5.)

The age-adjusted prevalence of obesity (BMI of 30 or more) in racial and ethnic minorities, especially minority women, is generally higher than in whites in the United States. In 2004 for both genders, non-Hispanic black people were more likely than Hispanics and non-Hispanic whites to be obese. (See Figure 1.6.) The age-adjusted prevalence of obesity was highest among non-Hispanic black women (39.6%) and lowest among non-Hispanic white women (21.3%). Earlier studies, including the NHANES, reported a high prevalence of overweight and obesity among Hispanics and Native Americans and lower prevalence of overweight (BMI greater than 25) and obesity (BMI of 30 or more) in Asian Americans than in the U.S. population as a whole.

WHY ARE SO MANY AMERICANS OVERWEIGHT?

Historically, overweight and obesity were largely attributed to gluttony—solely the result of inappropriate eating. The scientific study of obesity has identified genetic, biochemical, and metabolic alterations in humans and experimental animals, as well as the complex interactions of psychosocial and cultural factors that create susceptibility to overweight and obesity. Although obesity is thought to result from multiple causes, for the overwhelming majority of Americans, overweight and obesity result from excessive consumption of calories and inadequate physical activity—eating too much and exercising too little.

Some observers maintain that Americans were destined to become overweight when their diets remained unchanged even as products of the industrial revolution such as cars, automation, and a variety of laborsaving devices sharply reduced levels of physical activity. The widespread availability of high-calorie foods and less physically demanding jobs conspired to make Americans fatter. Others contend that the rise in overweight and obesity began during the 1970s, when women entered the work force in large numbers and had less time to cook, so Americans came to rely instead on processed, convenient, and calorie-dense, saturated-fat-laden fast foods. The CDC reported that in 2000, women ate 1,877 calories per day, 335 calories more per day than they did in 1971. Men, averaging 2,618 calories per day, consumed 168 calories more per day than their counterparts in 1971 ("Trends in Intake of Energy and Macro-nutrients—United States, 1971–2000," Morbidity and Mortality Weekly Report, vol. 53, no. 4, February 6, 2004).

The American Diet Has Changed

The American diet has changed dramatically since the middle of the twentieth century. According to the U.S. Department of Agriculture (USDA) in the Agriculture Fact Book 2001–2002, during the 1950s food production in the United States provided about 800 fewer calories per person per day than in 2000. Of the 3,800 calories produced per person per day in 2000, the USDA estimates that about 1,100 calories were wasted, either through spoilage, plate waste, or cooking, leaving an average of about 2,700 calories per person per day. The USDA data reveal that average daily calorie intake increased nearly one-quarter (24.5%) or about 530 calories between 1970 and 2000. Of that 24.5% increase, grains (primarily refined grain products) accounted for 9.5%; added fats and oils, 9.0%; added sugars, 4.7%; and fruits and vegetables together, 1.5%; and meats and nuts together 1%; while dairy products and eggs together declined by 1.5%.

According to the Agriculture Fact Book 2001–2002, Americans consumed an average of seven pounds more red meat than in the 1950s, forty-six pounds more poultry, and four pounds more fish and shellfish per person per year in 2000 than they had during the 1950s. Table 1.7 shows that meat consumption is at a record high—Americans consumed more meat—fifty-seven pounds more per year in 2000 than they had during the 1950s. Despite record-high per capita consumption of meat in 2000, the proportion of fat in the U.S. food supply from meat, poultry, and fish declined from one-third (33%) in the 1950s to one-quarter (24%) in 2000. This decline resulted from marketing of lower fat ground and processed meat products, a shift away from red meat to poultry and closer trimming of outside fat on meat, which commenced in 1986.

The USDA also reported that in 2000 Americans drank an average of 38% less milk and ate nearly four times as much cheese (excluding cottage, pot, and baker's cheese) as they had in the 1950s. Consumption of milk dropped from an annual average of 36.4 gallons per person in the 1950s to 22.6 gallons in 2000, although consumption of lower fat milk increased during this period. The USDA posited a link between the trend toward dining out and the reduction in beverage milk consumption. According to the USDA, soft drinks, fruit drinks, and flavored teas appear to be displacing milk as the beverages of choice of Americans. (See Table 1.8.)

Average use of added fats and oils in 2000 was two-thirds higher (67%) than the average use in the 1950s. Added fats included those used directly by consumers, such as butter on bread, as well as shortenings and oils used in commercially prepared foods. All fats that naturally occur in foods, such as those in milk and meat, were excluded from the USDA analysis. In 2000 Americans consumed, on average, 259% more salad and cooking oil than they did annually in the 1950s, and more than twice as much shortening, however, use of table spreads (butter and margarine) declined by 25% during the same period. (See Table 1.9.) During the 1950s, the added fats and oils group contributed the most fat to the food supply (41%), followed by the meat, poultry, and fish group (32%). By 2000, the fats and oils group's contribution to total fat had jumped twelve percentage points to 53%, probably due to the higher consumption of fried foods in fast food outlets, the increase in consumption of high-fat snack foods, and the increased use of salad dressings. Margarine, salad dressings and mayonnaise, cakes and other sweet baked goods, and oils continue to appear in the top ten foods for fat contribution, according to USDA food intake surveys, which examine the ongoing prevalence of discretionary fats in Americans' diets.

According to the USDA in Agriculture Fact Book 2001–2002, Americans in 2000 consumed 20% more fruit and vegetables than did their counterparts in the 1970s. Fruit consumption in 2000 was 12% above average annual fruit consumption in the 1970s. Fresh fruit consumption rose 28% during the same period, outpacing processed fruit consumption, which increased by just 2%.

Total vegetable consumption in 2000 was 23% higher than the average annual vegetable consumption in the 1970s. Like fruit, fresh vegetable use rose 26%, surpassing processed vegetable use, which increased by 21%. The USDA attributed some of the increase to the introduction of convenient, ready-to-eat, pre-cut, and packaged vegetables and to increasing consumer health awareness. (See Table 1.10.)

Per capita use of flour and cereal products reached 200 pounds in 2000 from an annual average of 155.4 pounds in the 1950s and 138.2 pounds in the 1970s, when grain consumption was at a record low. The increase reflects plentiful grain stocks, robust consumer demand for store-bought bakery items and grain-based snack foods along with increasing consumption of fast-food products such as buns, pizza dough, and tortillas. Despite the overall increase in grain consumption, the average American's diet contained mostly refined grain products and fell short of the recommended minimum three daily servings of whole grain products. (See Table 1.11.)

The USDA analysis in Agriculture Fact Book 2001–2002 revealed that in 2000 Americans favored sweets more than ever before. Per capita consumption of caloric sweeteners—sucrose (table sugar made from cane and beets) and corn sweeteners (high-fructose corn syrup) soared forty-three pounds, or 39%, between 1950 and 2000. In 2000 Americans consumed an average 152 pounds of caloric sweeteners per person, or about two-fifths of a pound per day. (See Table 1.12.)

The USDA cites a variety of factors that have contributed to the changes in the American diet over the past fifty years, including fluctuations in food prices and availability, increases in real (adjusted for inflation) disposable income, and more food assistance for the poor. New products, particularly the expanding array of convenience foods, also alter patterns of in consumption, along with more imports, growth in the away-from-home food market, intensified advertising campaigns, and increases in nutrient-enrichment standards and food fortification. The social and demographic trends driving changes in food choices include smaller households, more two-wage earner households, more single-parent households, an aging population, and increased ethnic diversity.

Americans Enjoy Eating Out

A variety of societal trends are thought to contribute to Americans' propensity to overeat, including eating outside the home, as well as ready access to and preference for sugar- and fat-laden foods. Table 1.13 shows how expenditures for eating away from home have increased steadily, and more than doubled between 1989 and 2004. In addition to less strenuous work, many Americans spend their leisure time in relatively sedentary pursuits—watching television, using computers, or playing video games—that not only do not expend calories but also, as in the case of television, actually encourage excessive eating.

Dining out is an important part of American culture, and market research conducted by Mintel International Group Ltd. found that Americans spend nearly half of their household food budget on eating out. In 2001 Americans spent nearly $30 billion on casual dining (as opposed to fine dining restaurants or fast-food outlets). The Mintel Report (Dining Out Review Market, Volume II: Casual/Family Restaurants, U.S. Report, July 2004) revealed that the upsurge in overall restaurant dining slowed to single-digit growth in 2001 from the high teens between 1995 and 2000; the casual dining market segment saw increases of about 5% to 7% per year from 2001 through 2003. The researchers anticipate continued growth in this sector in response to increasing time pressures in many households, which prevent people from preparing meals at home as well as an overarching cultural trend toward eating more meals outside the home.

Many nutritionists and obesity researchers assert that controlling portion size, which is key to controlling calorie consumption, is more difficult in restaurants, where portions are frequently quite large. Increasingly, restaurants have translated consumer demands for value into more food for less money. Since humans are genetically programmed to eat when food is abundant, larger portions trigger the natural impulse to eat more.

Pennsylvania State University researchers confirmed the notion that presented with larger portions, people will generally consume more. When they offered research subjects a five-cup portion of macaroni and cheese, the subjects all responded by eating 30% more than they had when they were given portions half that size. In "Portion Size of Food Affects Energy Intake in Normal-Weight and Overweight Men and Women" (American Journal of Clinical Nutrition, vol. 76, no. 6, December 2002), Barbara Rolls and her colleagues observed that both "restrained and unrestrained eaters" ate more when offered larger portions and asserted that Americans have become accustomed to eating too much at one sitting. The problem of portion size is compounded by the observation that Americans are eating larger portions of foods that are high in calories and fat.

David Grotto, a registered dietitian and spokesperson for the American Dietetic Association (ADA), asserted that restaurant portions have not changed as much as Americans' eating habits. In an interview published in the June 16, 2003, issue of the Miami Herald, Grotto noted that "A generation ago, dining out was pretty much limited to special occasions. Then, over time, the frequency of eating out increased and going to restaurants became part of Americana, especially in urban areas."

BIGGER PORTIONS IN RESTAURANTS

Researchers Barry Popkin and Samara Joy Nielson from the University of North Carolina at Chapel Hill looked at portion size consumed in the United States to determine whether average portion sizes had increased over time. They analyzed data collected by national nutrition surveys—the Nationwide Food Consumption Survey and the Continuing Survey of Food Intake by Individuals—conducted in the United States in 1977, 1989, 1994, and 1996, detailing the consumption habits of more than 63,000 people. For each survey year the researchers analyzed average portion sizes consumed of specific food items (salty snacks, desserts, soft drinks, fruit drinks, French fries, hamburgers, cheeseburgers, pizza, and Mexican food) by eating location—home, restaurant, or fast-food outlet. In "Patterns and Trends in Food Portion Sizes, 1977–1998" (Journal of the American Medical Association, vol. 289, no. 4, January 22, 2003), the researchers reported that over the past two decades, the average portions of such salty snacks as popcorn and chips have increased by 60%, and soft drinks have grown by 50%. The average dispensed soft drink measured 13 ounces (370 milliliters) in 1977, but by 1996, it was 20 ounces (570 milliliters). During the same period, an average bag of chips grew to 1.6 ounces (45 g) from 1 ounce (28 g). As a result, the average chips-and-soda snack contains 150 more calories than it did two decades before.

The portion-size changes were observed with many fast-food offerings. During the twenty years studied, the size of the average hamburger grew by 23%, to 200 g (7.05 ounces), while servings of fries grew by 16%, to 100 g (3.52 ounces). A regular-size burger-and-fries meal contained 155 calories more than it did in 1977. The researchers explained that increasing portion sizes reflected the fast-food industry's practice of "supersiz-ing" or "value adding"—offering larger sizes at discounted rates. Worse still, the researchers found that portion size also had expanded in Americans' homes, indicating widespread ignorance about appropriate portion size. Interestingly, portion sizes were smallest in restaurants, although they too had increased during the study period. For example, the average restaurant portion of spaghetti with tomato sauce and meatballs doubled in size from 500 to 1,025 calories.

Another study, conducted by University of Pennsylvania researchers found that larger portions served in restaurants resulted in patrons consuming more calories. The investigators covertly recorded the food intake of patrons who selected a pasta entrée over a ten-day period of a cafeteria-style restaurant on a university campus. On five days, the portion size of the entrée was the standard portion, and on five different days, the size was increased to 150% of the standard portion. Subjects were also asked to complete a survey to deter-mine perceptions of the portion size of the entrée and of the amount that they ate. The subjects who completed a survey were unaware that their intake was being monitored.

The investigators posited that when the portion size of an entrée was increased by 50%, the subjects would consume significantly more than when the standard portion was offered. They also sought to determine whether the subjects would compensate for the increased intake from the entrée by reducing their consumption of other foods at the meal and whether they could identify any characteristics of subjects that would predict how they would respond to increased portion size.

When the larger portion size was offered, subjects who purchased it consumed 43% more of the entrée than those who purchased the standard portion size. Subjects given the larger portion also ate significantly more of the entrée accompaniments (tomato, roll, and butter) than those who purchased the standard portion, even though the portion size of the accompaniments was the same for all subjects.

Overall, ratings of the appropriateness of the portion size of the entrée did not differ between subjects given the 150% portion and those who received the standard portion. There was, however, an effect of subject body size on this rating. Underweight and normal weight subjects who purchased the 150% portion rated it as closer to the "too large" end of the seven-point scale than those who purchased the 100% portion. In contrast overweight and obese subjects did not rate the portion size as "too large". The investigators concluded that subjects ate significantly more when the portion size was increased, and their responses to the survey indicated that they were unaware that the portion was larger than normal or that they had consumed more food (Nicole Diliberti et al., "Increased Portion Size Leads to Increased Energy Intake in a Restaurant Meal," Obesity Research, vol. 12, no. 3, March 2004).

AND BIGGER PORTIONS AT HOME

Increased portion sizes at home are reflected in recipes and cookbooks. Lisa Young reported in The Portion Teller (New York: Morgan Road, 2005), that recipes call for bigger portions using the same ingredients than they did in past decades. For example, a brownie recipe from a classic cookbook, Joy of Cooking (New York: Penguin Books, 1964), recommended dividing it into thirty servings, while the same recipe in the 1997 edition of the book is divided into only sixteen servings. Similarly, a 1984 recipe for Toll House cookies yielded 100 servings, whereas today the same recipe yields only sixty. Other popular food items have increased in size and calorie content. When the National Heart, Lung, and Blood Institute compared portion sizes and the corresponding calories of several popular foods from 1983 and 2003, researchers found that two decades earlier a bagel measured three inches in diameter and contained 140 calories. In the early twenty-first century, six-inch bagels contain more than twice as many calories—a whopping 350.

The University of North Carolina researchers also noted other changes in eating behavior. For example, the study found that Americans obtain 19% of their total calories from snacks—double the amount of 1977—and 81% from meals. They concluded that "control of portion size must be systematically addressed both in general and as it relates to fast food pricing and marketing. The best way to encourage people to eat smaller portions is if food portions served inside and outside the home are smaller."

Technology Satisfies the Hunger for Quick, Inexpensive Food

In Why Have Americans Become More Obese? (Cambridge, MA: National Bureau of Economic Research, Inc., 2003), Harvard University economists David Cutler, Edward Glaeser, and Jesse Shapiro refuted the notion that increased portion sizes, increasingly sedentary lifestyles, or restaurant dining were responsible for Americans' widening waistlines. After examining nearly 100 years of nutritional data, the researchers determined that technological advances have increased the efficiency of food production and made food more varied, convenient, tastier, and cheaper.

The economists illustrated how efficiencies in food preparation have revolutionized Americans' eating habits. They compared the speed and ease of preparation of commercial French fries with the previously time-consuming, labor-intensive process of scrubbing, peeling, paring, and frying required to prepare French fries. They observed that during the 1960s women spent an average of two hours a day on meal preparation—twice as long as the average American nonworking woman devotes to meal preparation today. It takes considerably less time today to prepare food because of advances in food processing and packaging. Further, technology improvements in the home, such as the microwave oven, have made it easier to eat quickly on demand.

The Harvard researchers' conclusion was that increased food consumption is the direct "result of technological innovations which made it possible for food to be mass prepared far from the point of consumption, and consumed with lower time costs of preparation and cleaning. Price changes are normally beneficial, but may not be if people have self-control problems." The study found that the average number of daily snacks between meals has risen by 60% since the late 1970s. Unable to resist the tempting, affordable variety of foods, Americans engage in more frequent snacking, consuming the excess calories that ultimately result in overweight.

Is the Food Industry the Culprit?

Kelly Brownell, director of the Yale Center for Eating and Weight Disorders and co-author, with Katherine Battle Horgen, of Food Fight: The Inside Story of the Food Industry, America's Obesity Crisis, and What We Can Do About It (New York: McGraw-Hill, 2004), cited a "near-total surrender to a powerful food industry" as one of the main causes of the obesity epidemic in the United States. Brownell and Horgen contend that the obesity epidemic represents more than a failure of Americans to assume personal responsibility and exercise willpower over their appetites. They exhort consumers to agitate against a food industry intent on fattening them and to work to counteract a variety of unhealthy social trends. The authors lament the super-sized meals and sedentary lifestyles, including Americans' "car-centric" culture that actively discourages walking and encourages children to sit in front of television, video games, and computers while eliminating physical education classes from schools, but they insist that the food industry bears the lion's share of responsibility for the rise in obesity. They argue that America feeds its pets better than its children, and that children are induced and manipulated by food industry media advertising to adopt poor eating habits and to consume high-calorie, low-nutrition junk food.

Brownell and Horgen cite toy giveaways, movie tieins, and in-school promotions as evidence of effective strategies employed by the politically powerful food industry to promote fast-food consumption. They feel that the battle against these pervasive influences is one that parents cannot win because even children receiving consistent, sound nutritional counseling from parents are not immune to the effects of multiple, powerful exposures to media advertising. The authors call for a nationwide, grassroots movement to reverse these trends and advocate specific measures such as junk-food taxes and banning advertisements that target children.

Greg Critser also indicts the food industry in Fat Land: How Americans Became the Fattest People in the World (New York: Houghton Mifflin Co., 2003). The nutrition journalist presents a critical analysis of the many social and economic factors that make Americans among the most overweight people in the world. Critser believes that chief among these factors is high fructose corn syrup, a low-cost sweetener that was developed by Japanese scientists in response to an overabundance of cheap corn. Corn syrup does more than sweeten, it also acts as a preservative, giving sweet foods longer shelf lives. Since the 1970s, high-fructose (a very sweet sugar) corn syrup has been used to sweeten nearly every product on supermarket shelves, from cereal to soda. Some researchers feel that because it is so ubiquitous, many Americans are unknowingly consuming excessive amounts of fructose. Table 1.14 shows increasing per capita consumption of high fructose corn syrup, which more than tripled from 1980 to 2002, peaked in 2003 and declined only very slightly in 2004.

Unfortunately, fructose also appears to trigger fat storage more efficiently than other sugars do. New studies are showing the body does not metabolize high-fructose corn syrup well. Although all sugars are stored in the body as fat, some researchers think that fructose is more readily converted into fat than other sugars. The fructose encourages the liver to promote fat by activating enzymes that create higher levels of cholesterol and triglycerides, and make muscles more insulin-resistant. Elevated levels of cholesterol and triglycerides, fatty substances normally present in the bloodstream and all cells of the body, increase the risk of coronary heart disease. Insulin resistance can lead to diabetes.

Critser also explained that once the staples used to produce fast foods became cheaper, the industry intensified marketing efforts to induce consumers to buy and eat more. Table 1.15 shows that food expenditures have consistently decreased as a percent of disposable personal income, declining from almost one-quarter of personal disposable income in 1930 to just 9.5% in 2004. Critser observes that a serving of McDonald's French fries "ballooned from 200 calories (1960) … to the present 610 calories" and that Americans' appetites grew to expect and demand the bigger servings. Critser noted that changing values and lifestyles conspired to fatten Americans. He described the rise of a "new boundary-free culture" that promoted consumption of sugary and fat-laden foods. Traditionally, families convened for home-cooked dinners, but Critser described the rushed parents of the 1980s as preferring to eat out or take in prepared foods. Childcare experts popularized the theory that children instinctively knew when they were sated and encouraged busy parents to relinquish control over their children's food consumption. In some parts of the country, budget cuts prompted schools to allow fast-food franchises to sell lunches and snacks to students on the school campuses. Finally, Critser observed that to accommodate—or even camouflage—Americans' expanding bodies, clothing manufacturers marketed large, loose-fitting clothing.