Utility of A1C for Diabetes Screening in the 1999 2004 NHANES Population

doi:10.2337/dc07-0585

HOME

HELP

FEEDBACK

SUBSCRIPTIONS

Clinical Care/Education/Nutrition/Psychosocial Research
Original Article

Utility of A1C for Diabetes Screening in the 1999–2004 NHANES Population

Catherine Buell, MD^*^,1^,2, Duclie Kermah, MPH¹ and Mayer B. Davidson, MD¹

¹ Charles R. Drew University, Los Angeles, California
² David Geffen School at University of California, Los Angeles, California

Address correspondence and reprint requests to Mayer B. Davidson, MD, Clinical Center for Research Excellence, Charles R. Drew University, 1731 East 120th St., Los Angeles, CA 90059. E-mail: mayerdavidson{at}cdrewu.edu

Abbreviations: FPG, fasting plasma glucose • HPLC, high-performance liquid chromatography • NHANES, National Health and Nutrition Examination Survey • ROC, receiver operating characteristic

INTRODUCTION

TOP
INTRODUCTION
RESEARCH DESIGN AND METHODS--
RESULTS--
CONCLUSIONS--
References

Thirty percent of people with diabetes are undiagnosed (1),and up to 25% already have microvascular complications at diagnosis(2). The American Diabetes Association recommends screeningadults $≥$ 45 years of age, especially those with a BMI $≥$ 25 kg/m²,at 3-year intervals with a fasting plasma glucose (FPG) measurement(3). However, physicians infrequently use an FPG for screening.For example, in a large health maintenance organization withinthe University of Michigan health system, 184 physicians at22 separate locations screened 5,752 (69%) of 8,286 people withoutdiabetes over a 3-year period. Ninety-five percent of the screeningtests were random glucose measurements, 3% were FPGs, and 2%were A1Cs (4). Random glucose levels depend on the length oftime after the previous meal and the carbohydrate content ofthat meal. There is no agreement on what random glucose valuesshould lead to further investigations.

Because A1C levels reflect average glycemia during the preceding3–4 months, meal issues are not a factor. To determinewhat A1C level should lead to further tests to diagnose diabetes,we examined the 1999–2004 National Health and NutritionExamination Survey (NHANES) population of those not known tohave diabetes to determine the sensitivity and specificity withwhich various A1C levels identified people with diabetes.

RESEARCH DESIGN AND METHODS—

TOP
INTRODUCTION
RESEARCH DESIGN AND METHODS--
RESULTS--
CONCLUSIONS--
References

NHANES oversampled African-American and Hispanic individualsto obtain enough data for analysis in this minority subset.Accordingly, results were weighted to reflect the U.S. population.SUDAAN statistical software was used to obtain weighted estimatesthat account for the complex survey design of NHANES (5). Ofthe 8,881 subjects with a 6-year weighting variable identified,6,036 were 20 years of age or older and 6,012 had fasting bloodsamples drawn. Glycated hemoglobin levels were measured by aboronate affinity high-performance liquid chromatography (HPLC)system and converted to A1C levels (6). The normal range was4.0–6.0% (personal communication, Randie Little, PhD,University of Missouri School of Medicine, Columbia, MO). Subjectswere queried by questionnaire, and 532 were excluded becauseof a previous diagnosis of diabetes. An additional 535 subjectswere excluded due to missing questionnaire data and 10 for missinglaboratory data. The final dataset included 4,935 individuals.Those whose FPG concentrations were $≥$ 126 mg/dl were diagnosedwith diabetes. The results were analyzed by a receiver operatingcharacteristic (ROC) curve.

RESULTS—

TOP
INTRODUCTION
RESEARCH DESIGN AND METHODS--
RESULTS--
CONCLUSIONS--
References

Of the final cohort of 4,935 subjects, 3,280 were normal (FPG<100 mg/dl), 1,485 had impaired fasting glucose (FPG 100–125mg/dl), and 170 had previously undiagnosed diabetes (FPG $≥$ 126mg/dl). The relationship between A1C level and diagnosis ofdiabetes is shown in Fig. 1. An A1C level of 5.8% yielded thehighest combination of sensitivity (86%) and specificity (92%)in the total population. Analyses of subsets of these NHANESsubjects revealed sensitivities of 84, 90, 84, 93, and 95% in2,334 men, 2,601 women, 2,631 non-Hispanic whites, 874 AfricanAmericans, and 1,295 Hispanics, respectively. Specificitieswere 92, 92, 93, 86, and 91%, respectively. The 135 subjectswho listed their race as "other" were not analyzed separatelybecause of their small number.

View larger version (12K):
[in this window]
[in a new window]

Figure 1— ROC curve for A1C levels to screen for diabetes (defined as an FPG $≥$ 126 mg/dl).

	CONCLUSIONS—

TOP INTRODUCTION RESEARCH DESIGN AND METHODS-- RESULTS-- CONCLUSIONS-- References

The American Diabetes Association recommends confirmation ofan FPG level to diagnose diabetes (3). Requiring confirmation(not available in NHANES data) would decrease sensitivity andincrease specificity. However, an A1C level of 5.8% is consistentwith the results of a previous ROC analysis on the 1988–1994NHANES III population utilizing an ion exchange HPLC A1C assay(normal 5.17 ± 0.45% [SD]) with diabetes also diagnosedby an FPG $≥$ 126 mg/dl. They concluded that the best screeningvalue lay between 1 and 2 SDs above the mean (7). The two NHANESstudies relate A1C levels to the prevalence of diabetes. Twostudies have related A1C levels to the incidence of diabetes.In 1,253 veterans aged 45–64 years, the 3-year incidenceof diabetes (by self-report, FPG $≥$ 126 mg/dl or A1C $≥$ 7.0%) at baselineA1C levels of $≤$ 5.5, 5.6–6.0, and 6.1–6.9% was 0.8,2.5, and 7.8 per 100 person-years, respectively (8). In 2,820French people aged 30–65 years whose baseline A1C levelswere split into deciles, the 6-year incidence of diabetes (diabetesdrugs or FPG $≥$ 126 mg/dl) was $~$ 2.5, 5.0, and 10% in the upper threeA1C deciles of 5.7, 5.8, and 5.8–7.1%, respectively (9).

Based on the results of these studies relating A1C levels toboth prevalent and incident diabetes, an A1C value in a DCCT(Diabetes Control and Complications Trial)-standardized assay(10) of $≥$ 5.8% could effectively serve to identify individualsin whom further investigations might be fruitful. Since fastingis not necessary for A1C measurements, this approach would markedlyreduce the number of people required to return for a fastingtest. Given the A1C result, both physicians and their patientsmight be more motivated to further explore the potential diagnosisof diabetes.

Alternatively, an argument can be made to utilize only A1C levelsfor the diagnosis of pre-diabetes/diabetes. The levels of glycemiathat are associated with the microvascular complications ofdiabetes are generally agreed upon as appropriate diagnosticcriteria (11). In both type 1 (12–14) and type 2 (15,16)diabetic patients, these complications did not develop or progressfor 6–9 years when the average A1C level was kept at <7.0%.Supporting this relationship, when A1C levels in three largediabetic populations were divided into deciles, the mean valuein the first decile, in which retinopathy increased, rangedfrom 6.7 to 7.5% (17). Regarding A1C levels and pre-diabetes,people in the DPP (Diabetes Prevention Program) (18) with valuesof 6.1–6.9% were twice as likely to progress to diabetesas those with lower values (19).

Thus, we propose that individuals with A1C levels measured ina DCCT-standardized assay of $≤$ 6.0% are normal, 6.1–6.9%have pre-diabetes, and $≥$ 7.0% have diabetes. Since metformin isrecommended for those with type 2 diabetes (20) and in youngerobese people with pre-diabetes (19), A1C levels should be confirmedin these patients. Alternatively, an A1C level of $≥$ 5.8% couldlead to a glucose measurement (an FPG or even an oral glucosetolerance test). In either approach, diabetes will be diagnosedin those at clear risk for microvascular complications. Thosewith pre-diabetes will be identified so that appropriate measurescan be adopted to reduce their chances of developing diabetes.

Acknowledgments

M.B.D. was supported by National Institutes of Health (NIH)Grant RR014616. D.K. was supported by NIH Grants RR011145, RR019234,and MD000182.

C.B. acknowledges Curt Rohlfing, BES, for his study model andShahrzad Bazargan, PhD, for her guidance and support.

Footnotes

Published ahead of print at http://care.diabetesjournals.orgon 11 June 2007. DOI: 10.2337/dc07-0585.

^* Student research thesis required for medical school graduation.

A table elsewhere in this issue shows conventional and SystèmeInternational (SI) units and conversion factors for many substances.

The costs of publication of this article were defrayed in partby the payment of page charges. This article must thereforebe hereby marked "advertisement" in accordance with 18 U.S.CSection 1734 solely to indicate this fact.

Received for publication March 23, 2007. Accepted for publication June 2, 2007.

References

TOP
INTRODUCTION
RESEARCH DESIGN AND METHODS--
RESULTS--
CONCLUSIONS--
References

Cowie CC, Rust KF, Byrd-Holt DD, Eberhardt MS, Flegal KM, Engelgau MM, Saydah SH, Williams DE, Geiss LS, Gregg EW: Prevalence of diabetes and impaired fasting glucose in adults in the U.S. population: National Health and Nutrition Survey 1999–2002. Diabetes Care 29:1263–1268, 2006[Abstract/Free Full Text]
Harris MI: Undiagnosed NIDDM: clinical and public health issues. Diabetes Care 16:642–652, 1993[Medline]
American Diabetes Association: Clinical Practice Recommendations 2007. Diabetes Care 30 (Suppl. 1):S4–S41, 2007
Ealovega MW, Tabei B, Brandle M, Burke R, Herman WH: Opportunistic screening for diabetes in routine clinical practice. Diabetes Care 27:9–12, 2004[Abstract/Free Full Text]
U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics: Reference Manuals and Reports: Analytic and Reporting Guidelines, 1999–2000. Hyattsville, MD, National Center for Health Statistics
National Health and Nutrition Examination Survey: NHANES 2003–2004 data documentation laboratory assessment: lab 10: glycohemoglobin: years of coverage 2003– 2004. In Documentation, Codebook, and Frequencies: MEC Laboratory Component: Glycohemoglobin: survey years 2003–2004. Hyattsville, MD, National Center for Health Statistics
Rohlfing CL, Little RR, Wiedmeyer HM, England J, Madsen R, Harris MI, Flegal KM, Eberhardt MS, Goldstein DE: Use of GHb (HbA_1c) in screening for undiagnosed diabetes in the U.S. population. Diabetes Care 23:187–191, 2000[Abstract]
Edelman D, Olsen MK, Dudley TK, Harris AC, Oddone EZ: Utility of hemoglobin A1c in predicting diabetes risk. J Gen Intern Med 19:1175–1180, 2004[Medline]
Droumaguet C, Balkau B, Simon D, Caces E, Tichet J, Carles MA, Eschwege E: the DESIR Group: Use of A1C in predicting progression to diabetes in French men and women: data from an Epidemiological Study on the Insulin Resistance syndrome (DESIR). Diabetes Care 29:1619–1625, 2006[Abstract/Free Full Text]
National Glycohemoglobin Standardization Program [Web site]. Available at http://www.ngsp.org.
National Diabetes Data Group: Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance.Diabetes 28:1039–1057, 1979[Medline]
Skyler JS: Diabetic complications: the importance of control. Endocrinol Metab Clin North Am 25:243–254, 1996[Medline]
Krolewski AS, Laffel LMB, Krolewski M, Quinn M, Warram JH: Glycosylated hemoglobin and the risk of microalbuminuria in patients with insulin-dependent diabetes mellitus. N Engl J Med 332:1251–1255, 1995[Abstract/Free Full Text]
Warram JH, Scott LJ, Hanna LS, Wantman M, Cohen SE, Laffel LMB, Ryan L Krolewski A: Progression of microalbuminuria to proteinuria in type 1 diabetes: nonlinear relationship with hyperglycemia. Diabetes 49:94–100, 2000[Abstract]
Ohkubo Y, Kishikawa H, Araki E, Miyata T, Isami S, Motooyoshi S, Kojima Y, Furuyoshi N, Shichiri M: Intensive insulin therapy prevents the progression of diabetic microvascular complications in Japanese patients with non-insulin-dependent diabetes mellitus: a randomized prospective 6-year study. Diabetes Res Clin Pract 28:103–117, 1995[Medline]
Tanaka Y, Atsumi Y, Matsuoka K, Onuma T, Tohjima T, Kawamori R: Role of glycemic control and blood pressure in the development and progression of nephropathy in elderly Japanese NIDDM patients. Diabetes Care 21:116–120, 1998[Abstract]
Davidson MB, Schriger DL, Peters AL, Lorber B: Relationship between fasting plasma glucose and glycosylated hemoglobin: potential for false-positive diagnoses of type 2 diabetes using new diagnostic criteria. JAMA 281:1203–1210, 1999[Abstract/Free Full Text]
Diabetes Prevention Program Research Group: Reduction in the incidence of type diabetes with lifestyle intervention or metformin. N Engl J Med 346:393–403, 2002[Abstract/Free Full Text]
Nathan DM, Davidson MB, DeFronzo RA, Heine RJ, Henry RR, Pratley R, Zinman B: Impaired fasting glucose and impaired glucose tolerance: implications for care. Diabetes Care 30:753–759, 2007[Free Full Text]
Nathan DM, Buse JB, Davidson MB, Heine RJ, Holman RR, Sherwin R, Zinman B: Management of hyperglycemia in type 2 diabetes: a consensus algorithm for the initiation and adjustment of therapy. Diabetes Care 29:1963–1972, 2006[Free Full Text]