June 2010
Feature Story

Kevin F. Foley, PhD, DABCC
Shirley L. Welch, PhD, DABCC

Recently our clinical chemistry laboratory has received phone calls from nurses and physicians inquiring whether their patients who need to be screened for gestational diabetes can use jelly beans instead of the standard 50-g glucose beverage. We were amused the first time we were asked this question and responded that all patients should be administered the usual, established glucose drink provided by the lab. After the second and third times we were asked about jelly beans, we decided to look into this more deeply. If you Google “glucola,” or “gestational diabetes alternatives,” you will find multiple Web sites that cite jelly beans as an alternative for gestational diabetes testing.

Gestational diabetes mellitus, or GDM, remains a significant health problem in the United States. It is recommended that all pregnant women be screened with an oral glucose tolerance test between 24 and 28 weeks of gestation. Some women complain about the thickness and unpalatable taste of 50-g glucose beverages. Others complain of nausea, vomiting, bloating, and GI pain. One source says noncompliance is high and 10 to 15 percent of patients can experience vomiting.¹ Some women claim to be allergic to the flavorings or coloring agents in the drink. It is therefore not surprising that nurses, physicians, and even dieticians have weighed in on alternatives that would be more tolerable and presumably contain the same amount of glucose.

Clinical chemists are more likely to be concerned about obvious analytical issues: Which brand of jelly bean should be used? How many will be taken? What sugars are in the jelly beans? Are the sugars as bio­available as the glucose solution, and will they be absorbed with similar kinetics? Other questions about who would supply and keep fresh inventory on the jelly beans are also pertinent. (It can be hard to find the “right” jelly beans in any season other than spring.) So where did the jelly bean idea come from, and are jelly beans being used?

Boyd and colleagues published in 1995 the results of a small study in which ingestion of 18 Brach’s brand jelly beans was compared with the standard 50-g oral glucose solution.² Boyd, et al., took 157 women who had no history of diabetes of any kind and administered the standard 50-g liquid challenge (followed by a one-hour venous plasma glucose). Within two weeks of the 50-g liquid challenge they also had patients consume 18 jelly beans within two minutes, with another one-hour plasma glucose level drawn. The jelly bean dose was cal­culated to provide 50 grams of carbohydrate. Each contained 53 percent sucrose, 14 percent dextrose, six percent cornstarch, and 18 percent higher saccharides. Patients also underwent a 100-g, three-hour OGTT within two weeks of the jelly bean challenge. The American College of Obstetricians and Gynecologists cutoff value of 140 mg/dL was used to determine the need for a further confirmatory three-hour OGTT. Using this standard, the liquid 50-g dose was 46 percent sensitive and 81 percent specific at detecting gestational diabetes. The jelly bean test was 31 percent sensitive and 96 percent specific. The positive predictive value using jelly beans was 40 percent versus 18 percent with the liquid dose. The incidence of GDM in the study was 8.3 percent. Not surprisingly, 88 percent of the women in the study preferred the jelly beans to the beverage. Using ROC curve analysis, Boyd, et al., identified 120 mg/dL as a cutoff that would match or exceed the sensitivity of the standard liquid beverage. When using a cutoff of 120 mg/dL instead of 140 mg/dL, the jelly beans were 54 percent sensitive at detecting those who had GDM (and would have needed the confirmatory three-hour OGTT).

Four years later, in 1999, Lamar, et al., conducted another jelly bean study.³ This group extended the previous study by attempting to increase the jelly bean dose such that it would be usable at the 140 mg/dL cutoff. Lamar’s study was conducted in a similar fashion but it randomized the order of the jelly bean versus liquid tests and scheduled the second test within one week of the first. The 100-g three-hour OGTT was performed seven to 10 days later. Using the same brand of jelly beans, the researchers instructed patients to eat 28 jelly beans instead of 18 and to consume them within 10 minutes. The study showed good correlation between serum glucose values after the jelly beans compared with serum glucose values after drinking the 50-g liquid dose. However, in this study of 136 wo­men, gestational diabetes was present in only five women. Two of these women screened positive with the jelly bean test while four were positive with the beverage. Thus, statistical power was lacking to compare the sensitivity of both assays at the 140 mg/dL cutoff.

The American Diabetes Association criteria now say it is not necessary to screen low-risk women who fulfill all of the following criteria: <25 years old, normal body weight, no family history of diabetes, no history of abnormal glucose metabolism, no history of poor obstetric outcomes, and not members of high-risk ethnic groups (Hispanic, Native American, Asian, African American, Pacific Islander). For those women who need screening (average- or high-risk patients), the standard cutoffs of a fasting plasma glucose ≥126 mg/dL or a casual plasma glucose ≥200 mg/dL are diagnostic of GDM. However, if these average- or high-risk patients don’t exhibit hyperglycemia with a fasting or casual plasma glucose, then a challenge test on a later day is required. A one-step or two-step approach for gestational diabetes screening can be used. The one-step approach uses a 100-g glucose load subsequent to an eight- to 14-hour fast. Patients must exceed at least two of the time-dependent cutoffs below.

The two-step approach uses an initial screen which, if abnormal, reflexes to the 100-g challenge described above. For the first screening step no fasting is required, and for this reason this approach has become popular. A fast is required only for those who fail the screen and need the subsequent three-hour OGTT. It is the first step using a 50-g challenge that has been tried with jelly beans. The initial screen is performed using a 50-g glucose challenge followed by a one-hour plasma glucose. Using this 50-g screen, a cutoff of 140 mg/dL identifies 80 percent of women with gestational diabetes. A cutoff of 130 mg/dL increases sensitivity to 90 percent.⁴

Although it is likely that a jelly bean dose could be found that would be acceptable at the commonly used 140 mg/dL cutoff, this is only extrapolation. A jelly bean alternative that has been shown to be as sensitive as, or more sensitive than, the liquid dose at the 140 mg/dL screen cutoff has not been verified or validated. As two physicians commented in a letter published in Lancet: [Given the positive predictive values] “the jelly bean jar should remain firmly shut, and any temptation to develop confectionary screening should be discouraged.”⁵ However, physicians, nurses, and dieticians have been assuming that a fistful of jelly beans “should” replicate a 50-g liquid dose. We have talked to clinicians who have been using jelly beans (the number and brand vary) and a cutoff of 130 mg/dL for screening those patients who need to undergo the confirmatory three-hour OGTT. Although the more sensitive 130 mg/dL lower cutoff is preferable to the 140 mg/dL cutoff, jelly beans as the source of the glucose still has not been formally established or recommended. This practice also causes one to question further: If a patient cannot tolerate the 50-g challenge, how would she complete the 100-g challenge if required? Is the three-hour OGTT also being conducted with jelly beans? We hope not.

There is a non-glucose challenge alternative for the diagnosis of GDM. A fasting plasma glucose ≥126 mg/dL or a casual plasma glucose ≥200 mg/dL is diagnostic (when confirmed with repeat testing on a different day). GDM is defined as any degree of glucose intolerance with onset or first recognition during pregnancy. This definition applies re­gardless of whether the condition persists after pregnancy or if the condition predated or began concomitantly with the pregnancy. Thus, diagnostic criteria for GDM include the standard diagnostic criteria for type 1 and type 2 diabetes mellitus. A 2000 study of 1,031 women demonstrated similar c-statistics using the OGTT and a fasting glucose and actually concluded that fasting glucose measurement could replace glucose challenge testing in GDM screening.⁶ Currently, however, either a positive three-hour OGTT or a fasting glucose ≥126 mg/dL or a casual glucose ≥200 mg/dL can serve as diagnostic criteria for GDM. But note that patients who do not exhibit a fasting glucose ≥126 mg/dL or a casual glucose ≥200 mg/dL still require GDM screening with a glucose challenge. Only if they already meet diabetes diagnostic criteria of a fasting glucose ≥126 mg/dL or a casual glucose ≥200 mg/dL are they exempt from further GDM testing.

Decisions surrounding significant changes to a laboratory test must be made only after consultation and approval of the laboratory’s directors. Although well intentioned, many health care workers are unaware of the rationale and empirical evidence that goes into laboratory guidelines. They feel empowered to alter preanalytical criteria if it seems reasonable to them. The urban-legend status of the jelly bean approach lends itself to be nonstandardized and ambiguous. Those who are using this “off-label” method don’t appear to be consistent in the number or brand of jelly beans used. Preanalytical testing issues are, as CLIA mandates, under the purview of laboratory directors and managers. In our opinion, the advantages of jelly bean administration do not outweigh the risks of a false-negative screen in an expectant mother. Until this approach has been validated, the jelly beans should stay in the Easter basket and out of the clinics.

References

1. Helton D, Martin R, Martin J, et al. Detection of glucose intolerance in pregnancy. Perinatol. 1989;9:1259–1261.
2. Boyd KL, Ross EK, Sherman SJ. Jelly beans as an alternative to a cola beverage containing fifty grams of glucose. Am J Obstet Gynecol. 1995;173:1889–1892.
3. Lamar ME, Kuel TJ, Cooney AT, et al. Jelly beans as an alternative to a fifty-gram beverage for gestational diabetes screening. Am J Obstet Gynecol. 1999;181:1154–1157.
4. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care. 2008;31(suppl)S55–S60.
5. Dornhorst A, Frost G. Jelly-beans, only a colorful distraction from gestational glucose-challenge tests. Lancet. 2000;355(9205):674.
6. Tam WH, Rogers MS, Yip SK, et al. Which screening test is the best for gestational impaired glucose tolerance and gestational diabetes mellitus? Diabetes Care. 2000;23:1432.