The phantom epidemic of child diabetes

We’ve heard it repeated so many times that consumers and healthcare professionals, alike, assume it must be true: that the epidemic of childhood obesity has created a new epidemic among today’s young people — of type 2 diabetes. Once thought to be an adult disease, type 2 diabetes is said to have alarmingly increased among children and adolescents because rates of overweight and obesity have tripled in the past two decades.

But after twenty years, has anyone actually checked to see if this is really true? Is there really an epidemic of childhood diabetes? Physical assessments and blood tests are done regularly on representative samples of the American population through the National Health and Nutrition Examination Surveys (NHANES), and the prevalence of type 2 diabetes among U.S. children has been tracked since 1988. Unlike obesity statisticulations, no one even has to think or understand statistics to do a simple check of these facts, if they cared to.

Incredibly, virtually no one has.

Back when the government declared a nationwide war on obesity, with a focus ‘on the children’, the public was easily convinced that obesity was deadly. Facts to the contrary were already available to public health officials, though, and had even been published for medical professionals. So — as young people continue to be healthier with each passing decade and the CDC estimates that today’s children will live longer than at any other time in our nation’s history — as every childhood obesity initiative has been unable to demonstrate effectiveness in reducing obesity rates or disease — and as even the U.S. Preventive Services Task Force concluded, after a comprehensive review of 40 years of evidence on childhood obesity screening and interventions, there is no sound evidence to support them or that childhood ‘overweight’ or ‘obesity’ is even related to health outcomes in adults or children — the belief continues that childhood obesity is so unhealthy and epidemic, that it’s driving increases in type 2 diabetes among young people.

These claims of dual ravages of population-wide obesity and diabetes epidemics among youngsters, and predictions of dire consequences for the health of the nation, are what’s supporting calls for urgent nationwide public health measures, especially focused on the purported bad diets of today's young people. An epidemic of diabetes in children is at the core to the claim that this will be the first generation to live shorter lives than their parents. The importance of this childhood diabetes claim and the massive federal, state and local funds being spent to address it, make it even more incredible that no one has done a simple fact check to see if is even true.

The experts tell us

So many public health officials and medical experts have repeated the same information, that, perhaps, no one has thought to or dared to question or critically examine it:

Type 2 diabetes has changed from a disease of our grandparents and parents to a disease of our children. As more and more children and young adults develop this devastating disease, it has become apparent that we have much to learn about… how to prevent this "new epidemic" from destroying future generations of Americans. Type 2 diabetes has been described as a new epidemic in the American pediatric population … In 1992, it was rare for most pediatric centers to have patients with type 2 diabetes… and by 1999, it accounted for 8–45% of new cases depending on geographic location. — Dr. Francine Ratner Kaufman, M.D., president of the American Diabetes Association. (Clinical Diabetes, 2002)

In the past 25 years, the health status of our children has deteriorated and with it, their ability to learn and become productive members of society. The proportion of overweight children has tripled, and the number of children who suffer serious medical, psychological and social repercussions has increased. Type 2 diabetes, also known as adult onset diabetes, has increased by a factor of ten. — Dr. Maria Goldstein, M.D., as chair of the New Mexico Pediatric Society and member of the Childhood Obesity Task Force and steering committee creating New Mexico’s Strategic Action Plan, Physical Activity and Nutrition in the Schools. (Albuquerque Tribune, February 1, 2005)

The prevalence of overweight among adolescents in the United States has nearly tripled in the past two decades… Type 2 diabetes, previously considered an adult disease, has increased dramatically in children and adolescents. Overweight and obesity are closely linked to type 2 diabetes. — U.S. Surgeon General, Call To Action To Prevent and Decrease Overweight and Obesity Overweight in Children and Adolescents. The term “diabesity” is even a registered trademark of ShapeUp America!

During the mid-1990s, type 2 diabetes in youth increased ten-fold in the US, and mirrored the childhood obesity epidemic. — Dr. Kaufman (“Childhood Obesity: The Declining Health of America’s Next Generation,” testimony before the U.S. Senate Subcommittee on Children and Families, July 16, 2008)

With rising cases of childhood obesity, children are now getting Type 2 diabetes, and the numbers are increasing rapidly. — Dr. Jennifer Shine Dyer, M.D., assistant professor of pediatrics at Nationwide Children’s Hospital and the College of Medicine at Ohio State University.

We know that obesity and overweight are important health threats… Another very scary fact for children. Type 2 Diabetes, which used to be known as adult onset diabetes, is now increasingly being diagnosed and adding to the cardiovascular risk profile of our children.— Dr. Julie Gerberding, M.D., Director of the Centers for Disease Control and Prevention, June 2, 2005.

Until recently, the majority of cases of diabetes mellitus among children and adolescents were immune-mediated type 1a diabetes. Obesity has led to a dramatic increase in the incidence of type 2 diabetes among children and adolescents over the past 2 decades. — Dr. Tamara S. Hannon, M.D. and colleagues at the Division of Weight Management and Wellness, University of Pittsburgh School of Medicine (Pediatrics, August 2005).

What we're seeing is an alarming increase in diabetes among children. — John M. Auerbach, MBA, Commissioner of the Massachusetts Department of Public Health (Boston Globe, 2006).

The 2008 annual F as in Fat: How obesity policies are failing in America report, authored by Trust for America's Health, and the Robert Wood Johnson Foundation, says that type 2 diabetes is the new epidemic among American children and now accounts for 8 to 45 percent of new pediatric diabetes cases, depending on geographic location. In describing the trend, it states:

Although there are a number of genetic risk factors, obesity is largely driving the increase in childhood type 2 diabetes. The problem is especially severe among children and youth of African, Hispanic, Asian, or American Indian ancestry. In 2000, SEARCH for Diabetes in Youth, a 5-year, $22 million research project funded by CDC and the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), was launched to identify the number of children under age 20 with diabetes by type, age, sex, and race or ethnicity… Initial results from the study show that while type 1 diabetes remains the most common form of diabetes among children and adolescents, type 2 diabetes becomes more common after the age of 10, with minority children more affected than non- Hispanic white children.

According to Francine Ratner Kaufman, president of the American Diabetes Association, “there is no doubt that the emergence of this epidemic in children and young adults is a major public health problem.”

The facts are stunningly different from the assuredness of these claims. As we look at the prevalence of diabetes among U.S. children and teens since 1988, when type 2 diabetes began to be tracked, we’ll also dig up the sources for these statements and the doublespeak behind them.

Population-based evidence

Dr. Joyce M. Lee, M.D., MPH, with the division of Pediatric Endocrinology, Child Health Evaluation and Research Unit at the University of Michigan in Ann Arbor, examined the soundest sources for national data on childhood diabetes and obesity and reported them in the latest July issue of Archives of Pediatric and Adolescent Medicine.

As she noted, when increased rates of childhood obesity were publicized in the mid-1990s, reports surfaced about corresponding increases in childhood diabetes. These reports, however, were based on small studies of select, high-risk populations in clinical settings, rather than national data, and could not credibly be extrapolated to population figures to make trend claims.

On the other hand, rates of childhood diabetes in the U.S. have been assessed through larger studies, beginning with NHANES, and drawn from large population samples. Negating concerns that NHANES may underestimate type 2 diabetes, often said to be higher among racial/ethnic minorities, she reminded readers that NHANES oversamples racial/ethnic minorities. Between 1988 and 2000, these population studies found rates of type 2 diabetes among adolescents had remained relatively stable, “despite the notable increase in obesity prevalence among children” during that same period. The results of the five large, population-based studies that followed children and teens through 2004, remained consistent with NHANES data.

In other words, she concluded that, “despite significant increases in prevalence rates of childhood obesity in the United States during the past two decades,” the medical literature shows that rates of type 2 diabetes among children and adolescents at the population level have remained steady. There is no epidemic of childhood diabetes, or any measurable change at all.

What may be most surprising is how rare type 2 diabetes is, and has consistently been, among young people. The first large population study had been done by CDC researchers at the National Center for Health Statistics and published in a 2001 issue of Diabetes Care. Led by Dr. Anne Fagot-Dampagna, M.D., these researchers examined actual fasting blood glucose levels measured as part of NHANES from 1988-1994 on a nationally-representative sample of nearly 3,000 adolescents. They found that 0.29% of young people had type 1 diabetes and only 0.12% had type 2 diabetes. In fact, the prevalence was so low, they couldn’t be precise in their estimates.

Professor Glen E. Duncan, MS, Ph.D., at the Nutritional Sciences Program, Department of Epidemiology at the University of Washington in Seattle, then examined NHANES data from 1999 through 2002, a general population sampling of nearly 1,500 teens, as well as fasting blood sugars among more than 4,000 self-reported diabetic teens. As he reported, overall, only 0.35% of U.S. teens had type 1 diabetes and 0.15% had type 2 diabetes.

As Dr. Lee noted, although slight differences in type 2 diabetes prevalence in population-wide studies have been reported because of sampling variability and differences in classification methods, they’ve been consistent with NHANES. And “similar rates were reported in a racially and socioeconomically diverse cohort of adolescents within an urban-suburban school district.”

This third study was led by Dr. Lawrence M. Dolan, M.D., a pediatric endocrinologist at Cincinnati Children’s Hospital Medical Center. He measured the fasting blood sugars and glucose tolerance tests in a random sampling of around 2,500 black and Caucasian teens in 2003. He found similarly low rates of diabetes among the teens as had been reported in NHANES 1988-1999: 0.24% had type 1 diabetes and 0.12% had type 2 diabetes. “Undiagnosed diabetes mellitus was rare,” he added. He also found increased fasting glucoses correlated with puberty and decreased BMI z-scores. “Impaired fasting glucose levels,” using the NHANES’ definitions, were also similar: 2%, compared to 1.7% in NHANES.

The SEARCH for Diabetes in Youth Study was a six-center population study based on what Dr. Lee described as the “gold standard classification method,” physician-diagnosed cases of diabetes. Among a population of 3.5 million children (0-9 years old), and 1.7 million adolescents (10-19 years old), it reported even slightly lower prevalences for diabetes. Among the young children, rates were miniscule, with 0.15% diagnosed with type 1 diabetes and 0.02% with type 2 diabetes. Among the pre-teens and teens, the prevalence of type 1 diabetes was 0.23%, with type 2 diabetes only 0.04%.

As Dr. Lee also noted, the 2003-2004 National Health Examination Survey data on 102,353 young people of all ages (0-17 years), the parental reported cases of all types of diabetes together estimated an overall prevalence of 0.32%.

Here’s the 'epidemic' of type 2 diabetes among America’s youth:

Over the past two decades, the estimated prevalence of type 2 diabetes among U.S. teens has been:

0.12% — 0.15% — 0.12% — 0.04%.

Estimates among children as a whole (age 0-19 years) are lower, 0.02%.

The consistency of the population findings across two decades regardless of the methodology, said Dr. Lee, clearly shows very low rates of type 2 diabetes among young people and a lack of any notable increase, “despite increases in obesity.” The overall burden of type 2 diabetes remains concentrated in older adults, she said.

These findings received little media attention. When the review was given brief mention, the findings were still spun to make doomsday predictions, largely based on beliefs that obesity is bad, so the adverse effects may be just around the corner. Examining the facts, however, finds no support for such speculations.

There is no epidemic of diabetes destroying the lives of children. There’s not even a hint of an impending epidemic in the population data. And diabetes most certainly hasn’t increased among the population of American children by a factor of ten. So, where did that oft-repeated claim of a ten-fold increase come from?

Remember, once upon a time...

The source of this claim turns out to be a small paper published more than 12 years ago by doctors at the University of Cincinnati College of Medicine.

They had examined the medical records of 1,027 patients who had been sent to their regional pediatric diabetes referral center. Over 12 years, from 1982 to 1994, a total of 54 children met their clinical criteria for type 2 diabetes. They said that the number of children who were newly diagnosed with type 2 diabetes at their clinic had gone from 4% prior to 1982 to 16% in 1994. Extrapolating the rate of heightened diagnoses made at their clinic to the entire population of teenagers in the Greater Cincinnati area, they estimated the incidence of adolescent type 2 diabetes had increased ten-fold, from 0.7/100,000 youngsters in 1982 to 7.2/100,000 youngsters in 1994.

This is another example of the Land of Incognita fallacy of logic.

As professor Michael I. Goran, Ph.D., at the University of Southern California in Los Angeles, cautioned in a New England Journal of Medicine letter to the editor in response to another study that had estimated population prevalence also using data on sick children referred for medical care, “it is important to note that the study sample was derived from a clinic population that may not be the most representative sample suitable for deriving [population] prevalence estimates.” Referrals of children for type 2 diabetes evaluations can reflect selection bias of physicians, as well as clinical guidelines calling for such work-ups for the fat children — it can be a “seek and ye shall find” phenomenon.

The Cincinnati doctors had found no change in the average age of their patients at diagnosis. However, nearly twice as many diagnosed with type 2 diabetes were female and in puberty, which, they said, suggests that the normal insulin resistance characteristic of puberty may contribute to the appearance of type 2 diabetes among teen girls earlier than in young men. While they reported no change in the racial/ethnic or socioeconomic distribution of the young people referred to their clinic, they didn’t mention if more of the referrals were fat, only that there was no significant increase seen in the average BMIs among those newly diagnosed with type 2 diabetes, except for a slight increase in black females. But nearly all (85%) of those diagnosed with type 2 diabetes had a first- or second-degree relative with type 2 diabetes, they said, illustrating the high genetic risks associated with this condition.

That other study addressed by professor Goran is often cited as evidence of higher rates of type 2 diabetes and impaired glucose tolerance among young people who fall in the obese category. It was conducted by researchers at Yale University School of Medicine Pediatric Obesity Clinic, who based their estimates on 55 fat children and 112 fat teens who had been referred to their medical weight loss clinic between 1999 and 2001. These young people were especially not representative of the general population, with 40% of the girls having polycystic ovary syndrome, for example.

Using the American Diabetes Association definition for impaired glucose tolerance as fasting blood sugar levels under 126 mg/dl and 2-hour glucose tolerance levels of 140-200mg/dl; and the definition of type 2 diabetes as fasting blood sugars of 126 mg/dl and higher, or 2-hour glucose levels over 200 mg/dl, they found 25% of the children and 21% of the teens had impaired glucose tolerance — more were girls.

But two points suggest extra caution when extrapolating their percentages to all obese young people. While their entire cohort was ‘obese’, they reported that “the degree of obesity was not found to be a significant risk factor [for impaired glucose tolerance].” As doctors with the National Institute of Health in Bethesda, Maryland, led by Dr. Gabriel I. Uwaifo, M.D., responded in the New England Journal of Medicine writing, “we suggest that the unexpectedly high prevalence of impaired glucose tolerance in the group of children who were 4 to 10 years old may be due to referral bias in favor of extremely overweight children” who may have had medical problems leading to the referrals. The NIH researchers said that when they had recruited even overweight black and white children from their local community “whose parents were not seeking treatment for weight problems,” they found a much lower prevalence of impaired glucose tolerance. Examining the children in their cohort with similar BMIs to those in the Yale report, they found only 3 of the 48 obese children had “impaired glucose tolerance.”

Definitions are everything

How many read the earlier statement that “type 2 diabetes accounts for 8-45% of new cases, depending on geographic location” — and incorrectly thought it meant that 8-45% of kids tested have type 2 diabetes?

You would have over-estimated the incidence by up to 375-fold!

Instead, it is simply referring to the proportion of type 1 to type 2 diabetes and that’s flipped around a lot over the years, as definitions change and some studies and institutions place some types of diabetes in the type 1 category and others into type 2, and some break them down into more precise detail. We have to remember that just ten or fifteen years ago, no one even realized children got adult-type diabetes and weren’t watching for it. There is no population-based data on type 2 diabetes in young people available prior to 1988.

Until recent years, when diabetes was found in children, it was assumed to be type 1 juvenile, insulin-dependent diabetes. Then, when it became apparent that young people also get non-insulin dependent diabetes, for awhile, a lot of healthcare professionals assumed that fat kids had type 2 and the thin kids had type 1 diabetes. When that turned out to not be the case, and fat kids were also found with type 1 diabetes and that thin kids can get type 2 diabetes, the category kids were placed in became even more varied. Even a child with other forms of non-insulin dependent diabetes may be incorrectly lumped into the type 2 diabetes category.

In other words, diabetes is not a single disease and unless you know precisely how the authors defined it, one cannot assume that they used the same definitions. You can come away with very wrong impressions.

And, to confuse things even more, there’s the downward migration of blood sugar cut-offs used to define diabetes and glucose intolerance. Countless young people, as well as adults, for example, wrongly believe they have diabetes, especially if they’re fat. They’ve been told they have “mild” diabetes, are “at risk” for diabetes, or have “pre-diabetes” if their fasting blood sugar levels are in the adult range of normal, 100 mg/dl, often in an attempt to incentivize them to lose weight. That’s one problem of relying on studies that use self-reported diabetes, as they can overestimate the prevalence of diabetes in these cases.

At the 68th Annual Scientific Sessions of the American Diabetes Association, this past June, the merit of this newly proposed “prediabetes” label was debated, with several researchers concluding there doesn’t appear to be natural thresholds for abnormal glucose levels. Dr. Saul Genuth, M.D., with Case Western Reserve University in Cleveland, Ohio, and one of the Diabetes Control and Complications Trial investigators, noted that only about 25% of patients with “prediabetes” actually progress to diabetes, another 25% revert to normal blood sugar control, and the majority benignly remain in the “prediabetic” state. In other words, ‘prediabetes’ isn’t really predictive of anything.

The risks that have been reported as being associated with diabetes are based on studies using definitions with fasting blood sugars well over 140 mg/dl and there is no evidence that these lower average blood sugars in these normal ranges carry similar risks. Nor have there been any clinical trials to show that heart disease can be prevented by maintaining low blood sugars. But plenty of people have been frightened by thinking they’re at risk, after being labeled as prediabetic by their doctor or insurance company.

But the most tragic and heartbreaking result of the misinformation among the public about diabetes in children is the cruel blame and condemnation, especially towards fat kids with diabetes. Popular culture generally thinks there’s only two kinds of diabetes: one that’s not the patient’s fault and the other form that is. Rightly, most people would never think to blame thin children with type 1 diabetes for their condition, as it’s generally recognized that this form is from an immune disorder resulting in damage to their pancreas and can run in families. Type 2 diabetes, on the other hand, is perceived as the person’s fault for eating sugar and junk food, having a sedentary lifestyle and being fat; when, in fact, type 2 diabetes has considerably stronger genetic roots. And several other forms of non-insulin dependent diabetes also have even stronger genetic links or can be due to traumas and other syndromes beyond a child’s control.

Lots of forms of diabetes

It’s outside the scope of a single blog post to describe all of the different types of diabetes in detail, but just introducing the fact that there are different types and the degree to which they run in families may be of help in countering the blame and pain that misinformation has caused young patients and their parents. No doubt these classifications will continue to evolve and change, too, but briefly, diabetes mellitus can refer to many different conditions.

Type 1 is when the pancreas doesn’t produce enough insulin, often due to immune system damage to the pancreatic beta cells, and these patients always need insulin injections. It typically presents with weight loss, leading to its common association with thinner kids. It can also run in families and several genes have been identified associated with this condition. There is also nonimmune forms, including “idiopathic type 1 diabetes” the result of auto-immune destruction to the function of pancreatic cells.

Within the type 2 diabetes category, there can be found type 2 diabetes with insulin resistance, atypical diabetes mellitus, maturity onset diabetes of youth (MODY), genetic defects in insulin action, and secondary diabetes. With type 2 diabetes with insulin resistance, a symptom can be weight gain. It is more common among non-European populations, as is atypical diabetes mellitus, which isn’t associated with obesity and where insulin resistance is normal. Both are found in those with strong family histories for diabetes.

MODY was first described in 1974 and, like type 1 diabetes, is often found in thinner white children. So, for many years, MODY was misclassified and lumped in with type 1 diabetes, too. Atypical diabetes mellitus is also sometimes classified as a kind of MODY. People with MODY don’t produce enough insulin, however, they don’t always need insulin. MODY is the result of genetic defects and is believed to be an autosomal dominant inheritance. Each type of MODY is caused by a single gene not working correctly, whereas type 2 diabetes is caused by problems with several genes at once, explained Dr. John Porter, at Birmingham Children's Hospital and Diabetes U.K. Clinical Research Fellow. There are now six genes which are known to cause MODY when defective, he said. Among young people with MODY, 85% have a parent with diabetes, compared to only 11% of young people with type 1 juvenile diabetes having a parent with diabetes. Forms of MODY have now been identified among nearly every ethnic group around the world.

There’s also a list of other genetic defects that affect the insulin action, such as Type A insulin resistance, Leprechaunism, Rabson-Mendenhall syndrome, and the Lipoatrophic syndromes.

Secondary diabetes is another type which can follow certain conditions, such as secondary diabetes as a result of cystic fibrosis or trauma to the pancreas.

A point that never seems to reach consumers is this: Type 2 diabetes is strongly genetic, considerably moreso than type 1. Among identical twins, concordance rates are nearly 100% for abnormal glucose metabolism and have been estimated as high as 70-90% for type 2 diabetes. [To put the heritability of diabetes into perspective, 60-80% of human height is attributed to genetics.]

Siblings of a type 2 diabetic have been shown to have about a 40% chance of developing type 2 diabetes, compared to 80% chance for an identical twin. The genetic component of type 2 diabetes results in multiple gene variants working together, but numerous different genes have been identified in different families and groups.

Type 2 diabetes is not all the same. According to Dr. John E. Gerich, M.D., professor of Medicine, Endocrine-Metabolism Unit at the University of Rochester Medical Center in New York, about 5-10% of type 2 diabetics have MODY, another 5-10% have latent adult-onset autoimmune type 2 diabetes, and another 5-10% have type 2 diabetes secondary to rare genetic disorders. The etiology for the remaining three-quarters is still not known and “a matter of great controversy” but incidence patterns indicate that it too is heterogeneous, he said. It is generally agreed upon, he said, that its inheritance is polygenic, which means several abnormal genes or polymorphisms are necessary at once for the disease to develop; impaired insulin sensitivity and insulin secretion is also each under genetic control and are both important elements in its pathogenesis; and obesity itself, popularly linked to the garden-variety type 2 diabetes is also under genetic control.

[Type 2 diabetes is also a disease of aging, of course. Forgetting this fact can lead to misperceptions of an epidemic, as many population figures don’t adjust for age despite the fact our population is aging and elderly make up an increasing percentage of the population. “Almost everyone will become diabetic if they live long enough,” said Dr. Paul Ernsberger, Ph.D., at Case Western Reserve School of Medicine in Cleveland, Ohio. “Fasting blood sugar rises steadily with age, when it crosses the 126 mg/dL line, even just once (or perhaps twice), you are considered diabetic.”]

Sadly, most everyone thinks a fat child with diabetes has brought the condition on himself by eating sweets or junk and not exercising. Even thin people with type 2 diabetes can be stigmatized by beliefs that type 2 diabetes is just a lifestyle disease, rather than primarily genetic, as well as seen in aging. “In this day and age, I still get patients who believe that they developed diabetes from eating too much sugar,” said Dr. Gerald Bernstein, M.D., of New York City and former director of the Beth Israel Health Care Systems Diabetes Management Program. “Bad eating habits such as too much refined sugars, empty carbohydrates and fructose do not cause diabetes.”

The next time you hear an expert claim — typically to support healthy eating and lifestyle programs to combat childhood obesity — that there’s an epidemic of diabetes among children, check the facts. It's a very good place to start.

© 2008 Sandy Szwarc. All rights reserved.