Should Babies Be Tested for Diabetes?

When Colby was born in April, her mother, Jennifer Speth, tested the infant for diabetes. "We did it just for curiosity, never expecting she'd come back as a high risk," says Speth.

Colby is part of a national project called TEDDY (The Environmental Determinants of Diabetes in the Young). Parents are asked to consent to diabetes screening of their newborns; high-risk infants — those who test positive for the genes and have at least one close relative with Diabetes 1 — benefit by close follow-up. Participants receive blood tests every three months until age 4 and then every six months until age 15.

In Colby's case, the nurse took a drop of blood from her heel at the same time as she drew a sample blood spot for the testing of rare metabolic diseases such as PKU (phenylketonuria). Scientists analyzed Colby's DNA, looking for two immune-regulating genes related to Diabetes 1 — HLA-DR and HLA-DQ. The test results suggested that Colby was 100 times more likely than most children to develop Diabetes 1.

"It was a shock," says Speth. "But now we'll know to monitor her."

Speth is grateful for the advance warning on Colby. But would all parents feel the same? And this poses the question: Should newborns be routinely tested for diabetes?

The great debate

Many doctors believe screening for Diabetes 1 makes sense due to the increasing prevalence of the disease. The Juvenile Diabetes Research Foundation reports that the rate of Diabetes 1 has risen by 3-5 percent over the past few years.

Scientists don't yet know the reason for the growth, but theories abound. Some researchers attribute the increased rate to a diet of gluten and cow's milk. Others look to viral infections (such as rubella, mumps and Coxsackie’s B4), vaccinations, a germ-free environment, toxic agents, and/or psychological stress. Robert Vogt, Ph.D., a scientist at the Centers for Disease Control's (CDC) Newborn Screening Branch, points to the obesity problem in children.

Another impetus for diabetes screening is the disease's destructiveness. It can cause kidney problems, blindness and even amputations. Early diagnosis, however, can prevent a panicky dash to the ER with a sick child, says Desmond Schatz, M.D., a diabetes researcher at the University of Florida in Gainesville. Because the illness’s "nonspecific" flu-like symptoms often go unrecognized as biomarkers of Diabetes 1, children sometimes wind up in the ICU suffering from ketoacidosis toxicity, which can lead to brain swelling. "Anecdotally, we know there are a few deaths each year," says Dr. Vogt. "Irreversible brain swelling can lead to a dangerous electrolyte imbalance."

Case in point

Knowledge translates to power, says Andrew Muir, M.D., a pediatrics professor at the Medical College of Georgia. A clinician with TEDDY, Dr. Muir describes the case of a high-risk newborn with a diabetic father. During follow-up visits after the initial testing, the mother noted that the baby was wetting his diaper frequently. Three months later, the 18-month-old child was diagnosed with diabetes.

In fact, a recent research project known as DAISY (Diabetes Autoimmunity Study in the Young), headed by Jennifer M. Barker, M.D., at the University of Colorado Health Sciences Center, discovered that only one (or 3 percent) of 33 at-risk infants who were screened and monitored for the disease needed to be hospitalized. (The parents were given a glucometer to measure blood sugar). The control group (also high-risk infants) had a 36 percent hospitalization rate. The conclusion? Parents alerted to their child's predisposition are more apt to recognize symptoms and avoid a pediatric emergency.

Another reason researchers favor screening is to be able to study the genetic causes and pre-diabetic period better, and to identify individuals for prevention trials. Currently, TEDDY and its sister project PANDA (Prospective Assessment in Newborns for Diabetes Autoimmunity), both funded by the National Institutes of Health (NIH) and other health organizations, screen thousands of newborns in Florida, Georgia and Washington. Other countries also participate.

On the other hand

But some scientists say universal screening is premature. The test is not 100 percent accurate and, with its "low predictive value," according to Dr. Vogt, generates many false positives. For one thing, babies with the HLA genes do not necessarily develop diabetes. The same applies to babies with auto-antibodies to pancreatic cells (the second screening test). "The risk level needs to be more specific and precise," says Dr. Muir. So far, he adds, the most exact measure scientists can supply is that a child has a one in seven chance of developing Diabetes 1.

In an effort to improve the worldwide accuracy of laboratory tests, the CDC and the Immunology of Diabetes Society take part in DASP (Diabetes Autoantibody Standardization Program). The goal is to encourage laboratories to evaluate the newest technologies for specificity and sensitivity. The Boston-based company, PerkinElmer Life and Analytical Sciences, reports that auto-antibodies can be early predictors of the disease. So tracking this measure is vital, since babies and children do not exhibit the classic symptoms of diabetes until approximately 95 percent of insulin-producing cells in the pancreas have been destroyed. As an additional benefit, the test results also can be a red flag for celiac disease, a chronic digestive disorder that often precedes Diabetes 1, according to a recent study in Pediatrics.

Detractors protest that the tests heighten anxiety among parents. However, studies also show that the anxiety appears to decrease over time and varies widely with the mother's educational level, ethnic group and marital status.

Another negative experts point out is if testing does not take place within the context of a research study with appropriate follow-up, parents tend to forget their child's risk level, and therefore less monitoring of symptoms takes place. "If parents live near a city with a diabetes research program, I recommend that high-risk children participate," says Dr. Muir. "But if I lived in a small town, say in Utah, I wouldn't bother testing."

Bottom line, the tests are far from foolproof, maintains Dr. Muir, since 85 percent of children who develop diabetes do not have a close relative with the disease. Conversely, a large percentage of high-risk children show auto-antibodies, but never develop diabetes. The upside to this weak genetic link is that since fewer genes are involved, there is, according to Dr. Vogt, a better chance of finding the disease's causes.

Screening for success

The biggest downside to universal screening is the current lack of "intervention" or prevention trials. But, as the saying goes, the times are changing. In the 1990s, prevention trials using insulin failed. However, in 2002, researchers at Columbia University successfully used a drug — a monoclonal anti-body — to slow the progression of diabetes. The drug lengthened the time patients produced their own insulin, improving their overall health with minimal side effects.

Although not strictly an "intervention", TRIGR (Trial to Reduce Insulin Dependent Diabetics in the Genetically at Risk) studies high-risk infants to see if diet is an environmental trigger. International in scope, TRIGR will run till 2007. Its 40 centers in the United States, Canada, Europe and Australia will analyze the effects of a standard cow's milk formula on infants versus a hydrolyzed version. (For more information, call 1-888-STOP-T1D or go to

Experts say that as prevention trials become available in the next five years, various health organizations such as the March of Dimes will exert pressure on state legislatures to support universal diabetes screening.

What to do

Meanwhile, parents of high-risk newborns should follow these recommendations from the CDC:

1. Stay informed about the disease and visit your pediatrician regularly.

2. For the latest listing of prevention trials, access websites such as the American Diabetes Association (, the Juvenile Diabetes Research Foundation (, and the National Institutes of Health (

3. Contact your local diabetes chapter for support and education.

4. Watch for symptoms of thirst and frequent urination in your child.

A last word

While the debate about diabetes screening in newborns is raging, laboratory technology is quietly and rapidly improving. Currently, 55 laboratories in 17 countries are experimenting with automated fluorescent DNA sequencing, radioimmunoassay and remote monitoring to develop better diagnostic technologies. Laboratories not only consider accuracy but also cost-effectiveness.

"Greater efficiency of time is important,” says Harry Hannon, Ph.D., head of the CDC's Newborn Screening Laboratory. "We would like to look at various conditions at the same time."

TrialNet Study welcomes New Yorkers

"Anyone in the country can be involved in TrialNet," says Robin Goland, M.D., an endocrinologist and co-director of the Naomi Berrie Diabetes Center at Columbia University Medical Center. "We will mail kits to doctors who have patients with Type 1 Diabetes."

"By testing for auto-antibodies, we are doing a risk assessment of the future possibility of a person developing diabetes. We talk in terms of higher and lower risk.

"Only four to six percent of people who have siblings or other close relatives with Type 1 Diabetes will get it. So we are reassuring the 94 percent — which reduces anxiety. Also, we follow everyone carefully, and if someone does develop diabetes, we can involve the person in intervention studies.

"Parents who wish to test their children should get involved with a trial. They can call 1-800-HALT-DM1 or contact trial coordinator Ellen Greenberg at Information also is available at the Naomi Berrie Diabetes Center website (

"Parents living in Manhattan can take advantage of the Naomi Berrie Diabetes Center, a member of a research consortium participating in the most up-to-date research."