Glucose Tolerance Tests Accuracy In Diagnosing Diabetes

According to the World Health Organization (WHO), more than 220 million people worldwide have diabetes. An estimated 1.1 million people died from diabetes in 2005, and almost half of diabetic deaths occurred in people under the age of 70 years of age. WHO projects that the number of diabetic deaths will increase to 366 million by the year 2030 (8).

Diabetes Mellitus Type 2 is a prevalent disorder that causes one to have high blood sugar, or hyperglycemia. This hyperglycemia can be the result from one or a combination of 1) decrease production of insulin from beta cells of the pancreas; 2) increase sugar production from the liver; 3) decrease sugar uptake by cells secondary to insulin receptors. Symptoms of DMII are excess urination, excess thirst, dizziness, blurred vision, sweating, and fatigue. Patients presenting with these symptoms should be screened by a finger stick, where a blood sample is taken from a quick prick of the finger, to be tested for hyperglycemia. Normal blood sugar should range from 70-100mg. If one has a fasting sugar of >126mg or an after eating sugar level > 200mg, then an oral glucose tolerance test (OGTT) should be performed. During an OGTT, a patient consumes a 150-200g carbohydrate diet for three days and fasts from midnight prior to test date. The morning of test, the patient consumes 75g sugar mixed with 300ml of water within a 5 minute period. The patient’s blood sugar level is be measured at baseline, and then again at 120 minutes. A diagnosis of DMII is made if the baseline level is >126 mg and the 120 minute level is >200mg. These guidelines are set by the American Diabetic Association (ADA) and the World Health Organization (WHO) (2,8).

Another option for obtaining a blood sugar level is measuring the percent of glycosylated red blood cells, or the percent of sugar attached to a RBC. RBCs live for approximately 90 days in the human body. By measuring this percentile one can observe the patient’s blood sugar level over the previous 3 months and not just at the moment an OGTT is performed. Today, HbA1c is a main tool for following metabolic control in persons with diabetes(5). A HbA1c > 6.0 percent should permit a diagnosis of DMII, but is not at this time a definite diagnostic tool.

Diabetes can cause complications of multiple organ systems. WHO defines consequences of diabetes as follows:

Diabetes increases the risk of heart disease and stroke. 50% of people with diabetes die of cardiovascular disease (primarily heart disease and stroke).

Combined with reduced blood flow, neuropathy in the feet increases the chance of foot ulcers and eventual limb amputation.

Diabetic retinopathy is an important cause of blindness, and occurs as a result of long-term accumulated damage to the small blood vessels in the retina. After 15 years of diabetes, approximately 2% of people become blind, and about 10% develop severe visual impairment.

Diabetes is among the leading causes of kidney failure. 10-20% of people with diabetes die of kidney failure.

Diabetic neuropathy is damage to the nerves as a result of diabetes, and affects up to 50% of people with diabetes. Although many different problems can occur as a result of diabetic neuropathy, common symptoms are tingling, pain, numbness, or weakness in the feet and hands.

The overall risk of dying among people with diabetes is at least double the risk of their peers without diabetes (8).

Previous studies have showed that better control of plasma glucose levels reduced the risk of developing long-term complications pertaining to diabetes (4). A higher HbA1c correlates well with the likelihood of developing chronic complications such as the ones listed above.

This study is designed to explore if a HbA1c be used to diagnose diabetes. Observations suggest that a reliable measure of chronic glycemic levels such as HbA1c, which captures the degree of glucose exposure over time and which is related more intimately to the risk of complications than single or episodic measures of glucose levels, may serve as a better biochemical marker of diabetes and should be considered a diagnostic tool (2). As for the current gold standard for diagnosing diabetes, the oral glucose tolerance test (OGTT) has its limitations (2). These include high interindividual variability, low reproducibility compared to FPG, poor compliance with the conditions needed to perform the test correctly, and is cumbersome and time-consuming for medical staff and patients (4). Due to these factors one may ask, “Is a HbA1c or an OGTT more accurate at diagnosing new onset diabetes mellitus type 2 in a patient presenting with hyperglycemia?” By exploring this question and answering it from an evidence-based approach, the answer may help clinicians advance to an easier and less time consuming way to diagnose diabetes mellitus type II.

CLINICAL CASE

A 57 year old African American male presented to the outpatient office with symptoms of dizziness, blurred vision, polydipsia, and polyuria. He has a significant history of hypertension and hyperlipidemia. The patient was unclear when his symptoms started. Upon evaluation in the office, the patient was noted to have a marked glucose elevation of 420. An in-house HbA1c was also noted at 13.0. Upon further questioning, the patient has not been taking any medications for diabetes, and is currently taking Lisinopril and Zocor for his other medical conditions. Due to the presenting symptoms and lab results, the patient was admitted to the hospital for hyperosmolar nonketotic hyperglycemic state.

METHODS

A PubMed search was performed by using the “Clinical queries” and “Diagnosis”

filters. The terms “A1c AND diagnosis AND diabetes” and “glycosylated hemoglobin

AND diagnosis AND diabetes” were used to search the site for relating articles. With these

search terms, a total of 176 hits revealed articles pertaining to the requested information.

Articles that met all inclusion criteria for the research were evaluated and assigned a

type/level of evidence.

In order to be included in this evidence-based study, articles had to meet the following inclusion criteria:

Articles must be cohort studies.

Studies must not be > 6 years old.

Articles must have participants with impaired glucose levels or symptoms of impaired glucose.

Studies must include evidence of OGTT or FPG and HbA1c.

Studies must have a significant number of participants to produce a significant result (n > 375).

Any articles that did not specifically relate to diagnosing DMII with a HbA1c were excluded. Articles that were not cohort studies, were older than six years, did not have participants with impaired glucose, or did not have a significant amount of participants were excluded. Certain articles that appeared in the PubMed search were strictly review articles. These papers were reviewed, and if applicable, may be used to provided supporting factors about pathophysiology/ epidemiology of diabetes type II and its diagnostic criteria. Articles that met all inclusion criteria were evaluated and assigned a level of evidence using the Oxford Centre for Evidence-based Medicine Levels of Evidence worksheet.

RESULTS

Study #1: Diagnosing Type 2 Diabetes Mellitus: in Primary Care, Fasting Plasma Glucose and Glycosylated Hemoglobin Do the Job

Study Design: This study was performed at the Raval Sud Primary Care Center in Barcelona, Spain and was begun in 1992. The purpose of this study was to determine the validity of glycosylated hemoglobin values as a method to diagnose type 2 diabetes mellitus in a population at risk seen in primary care. Four hundred fifty four subjects were selected to participate in the study. The population served by the Raval Sud Center is characterized by it low evonomic level, high rate of immigration, and high rate of morbidity and mortality for certain diseases and disorders. Inclusion criteria for eligible participants had at least on e of the risk factors for developing DMII described in the ADA guidelines. These included family history of DMII, personal history of carbohydrate intolerance or gestational diabletes, prolonged use of a drug able to raise glucose levels, obesity with a body mass index > 30, hypertension, HDL-cholesterol levels < 35 mg/dL, or triglyceride levels > 250 mg/dL. Persons who did not wish to take part in the study were excluded. For the purpose of this particular study, data was recorded from the time the patient was included in the Raval Sud Care Center. The study then used a cross-sectional analytical design to validate a diagnostic test. (4)

Study Conduct: Subjects were interviewed and variables were recorded for each participant. These included sociodemographic characteristics such as age and sex, clinical characteristics such as BMI and blood pressure, and laboratory values including fasting plasma glucose in a venous blood sample, oral glucose tolerance test after a 75g glucose overload, and a HbA1c measured by high pressure liquid chromatography. To standardize the results for the HbA1c, the absolute values were recalculated in terms of the number of standard deviations above the mean. FPG and OGTT values were based on the WHO criteria as having normal, impaired, or DMII glucose levels. (4)

Study Results: The distribution of demographic characteristics and laboratory findings are shown in Table 1. The study found that plasma glucose levels were significantly lower in normal subjects than in subjects with abnormal glucose levels (IFG or OGTT) and even lower in subjects with abnormal glucose levels than in patients with diabetes (P< 0.001). Mean HbA1c values were significantly higher in patients with diabetes than in all other categories: 7.04% versus normal glucose- 4.98%, IFG- 5.42%, and IGT- 5.12% (P< 0.001). Table 2 shows the distribution of HbA1c values according to the diagnostic classification. Patients with HbA1c values above 6.37% had DMII, where none of the patients with a HbA1c below 4.22% had diabetes. Patients with HbA1c values ranging from 4.56% to 6.37% were considered to have IFG. Table 3 shows the validity for the different cutoff values of HbA1c in establishing a diagnosis of DMII. The study found that the validity of the HbA1c cutoff points increased markedly as the percentage value increased. From this table we can conclude that if a HbA1c value is > 5.94% (mean, +3SD), the diagnosis of DMII is reliable and accurate in 93% of the cases. Table 4 shows the diagnostic validity of a combined strategy of FPG and HbA1c values: patients were considered to have DMII when FPG > 125 mg/dL, or when FPG >110 mg/dL and HbA1c was greater than the cutoff value. Maximal efficacy (93% GV) was found for HbA1c > 5.94% (x +3SD), with a sensitivity of 92.2% and a specificity of 95.1%. (4)

Study Critique: It has been confirmed that the relationship between circulating glucose values and the onset of chronic complications exists. Thus, it is logical for the diagnosis of DMII to be based on glucose values. One of the main problems in this particular study was to define and establish a cutoff point for this continuous quantitative variable. This study analyzed different cutoff points for the whole sample of patients at risk for DMII. When HbA1c values > 5.51% (x +2SD), were used for the cutoff point for diagnosis of DMII, the sensitivity (76%) and specificity (85%) were acceptable. However, when a higher cutoff point was used, specificity increased, but only at the expense of reduced sensitivity. Due to this situation, the study designed a strategy for diagnosis based on the FPG values and the validity of HbA1c. (4) Level of Evidence: 1c

Study #2: Comparison of A1c and Fasting Glucose Criteria to Diagnose Diabetes Among U.S. Adults

Study Design: This study included participants from the 1999-2006 National Health and Nutrition Examination Survey. Participants included 6,890 adults (>20 years of age), without a self-reported history of diabetes. The subjects attended a morning examination, fasted for > 9 hours at the time of their blood collection, and had valid plasma glucose and HbA1c values taken. Participants were categorized into one of the four groups by presence or absence of fasting plasma glucose > 126 mg/dL and HbA1c > 6.5%. The distribution of the population into these groupings was determined and the K statistic value was calculated. Also, the distribution of U.S. adults by fasting glucose and different HbA1c cutoff points (6.0-6.7%) were calculated. The objective for this study was to compare A1c and fasting glucose for the diagnosis of diabetes among U.S. adults. (6)

Study Conduct: Data was collected through questionnaires (demographics, medical history), a physical examination (blood pressure, BMI, and waist circumference), and blood collection (lipids, plasma glucose, HbA1c). The plasma glucose was measured by using a modified hexokinase enzymatic method and the HbA1c using a high-performance liquid chromatography. (6)

Study Results: This study concludes that an HbA1c of > 6.5%, along with a FPG >125 mg/dL demonstrates reasonable agreement for diagnosing diabetes. 1.8% of the participants were classified as having diabetes with a HbA1c > 6.5% and a fasting glucose >126 mg/dL. Among participants with a HbA1c < 6.5% and a fasting glucose > 125 mg/dL, 45% had an A1c value > 6.0% but less than 6.5%. According to A1c guidelines, this value poses an elevated risk for diabetes. Table A1 shows a distribution of adults by fasting glucose and different HbA1c cutoff points. From this table, the lower the HbA1c cutoff points results in higher sensitivity and lower specificity. (6)

Study Critique: In this study, certain participants had discordant results such as a HbA1c > 6.5% and a fasting glucose of < 126 mg/dL. These results may have been due to the fact that assessment of different aspects of glucose metabolism. For example, subjects with these results may have been diagnosed with an OGTT, which was not available for the majority of participants in this study. A comparison of these participants using the OGTT would have been a interesting assessment done by this study to compare with the FPG and HbA1c. (6) Level of Evidence: 1c

Study #3: A1c and Diabetes Diagnosis: The Rancho Bernardo Study

Study Design: The Rancho Bernardo Study included 2, 107 participants without known DMII, who had an OGTT and a HbA1c between 1984 and 1987. This cross-sectional study of community dwelling adults was provided written informed consent and laboratory data was performed. (3)

Study Conduct: HbA1c was measured with high performance liquid chromatography using an automated analyzer. Ophthalmologic evaluation was also performed on the subjects. This was done by using nonmydriatic retinal photography. Sensitivity and specificity of HbA1c cutoff points for DMII were calculated, along with K coefficients which were used to test for agreement between A1c values and diabetes status. The objective for this study was to examine the sensitivity and specificity of HbA1c as a diagnostic test for DMII in older adults. (3)

Study Results: For this study the HbA1c cutoff value was 6.5%. This value had a sensitivity of 44% and a specificity of 79%. A lower A1c cutoff point of 6.15% yielded the highest sensitivity at 63% but a lower specificity at 60%. If one were to use this cutoff value, it would miss one-third of those with DMII by the American Diabetes Association guidelines. It would also misclassify one-third of those without DMII. Using the HbA1c value of 6.5% as the cutoff point, the agreement with DMII diagnosis was low (K coefficient was 0.119). In order to compare A1c and ADA criteria with DMII complications, the study looked at participants with some degree of retinopathy. Of the participants who had retinopathy, 40% had and A1c > 6.5% and none had DMII by ADA criteria. This study concluded that the limited sensitivity of the A1c value cutoff may result in missed or delayed diagnosis of DMII, whereas the use of current OGTT criteria will fail to identify a high proportion of individuals with high A1c values, which correlate with long term complications of DMII. (3)

Study Critique: This study was performed on a much older population than the other studies examined in this paper. It has its benefits and disadvantages for surveying a population in which there mean age was 69.4. The advantage is that the U.S. elderly population has the greatest current burden and is expected to have the greatest increase in the prevalence of DMII. On the other hand, the disadvantage to having such an older subject population is that it limited the HbA1c cutoff values to that particular population. In a previous critique of an article one of the concerns was the fact that there are different aspects of glucose metabolism. It would have been supportive if the article addressed the age of their participants and compared them with the study results. (3) Level of Evidence: 1c

Study #4: Diagnostic value of glycated haemoglobin (HbA1c) for the early detection of diabetes in high-risk subjects

Study Design: This study was performed by collecting data from the Bundang CHA General Hospital database. A total of 392 subjects who had an abnormal random plasma glucose, a history of gestational diabetes mellitus, a macrosomic baby, or a severe obesity were selected to participate in the study. Exclusion criteria included a previous history of diabetes of other endocrinopathies, pregnancy, abnormal liver or renal function tests, a history of major surgery, severe illness, blood transfusion within the previous 6 months, and weight loss > 3kg during the past three months. After an overnight fasting, blood samples were drawn from all participating subjects to include FPG and HbA1c values. (7)

Study Conduct: Glucose concentrations were measured using the glucose oxidase method on a autoanalyzer. The HbA1c values were measured by the high-performance liquid chromatography method. All statistical analysis was performed and the best predictive cutoff values for FPG and A1c for detecting patients with new diabetes were identified using the optimal sensitivity/specificity values determined by the receiver operating characteristic curve. (7)

Study Results: Figure 1 shows the ROC plot representing the sensitivity and specificity for the HbA1c and the FPG in detecting undiagnosed DMII. From this study, the optimal cutoff value for HbA1c was 6.1% and for FPG was 6.1 mmol/l. The sensitivity/specificity for the HbA1c cutoff value was 81.8% and 84.9% respectively. Table 1 shows the results from the combination of using FPG and HbA1c. This study demonstrated that HbA1c was very useful to screen for diabetes in high-risk patients and the combined use of HbA1c and FPG made up for the lack of sensitivity in FPG alone. (7)

Study Critique: This study’s subjects were only Korean, therefore making the population very ethnically limited. It would have been beneficial to have seen the population more diverse and to notice the change in results. Also, the study stated that an OGTT was performed, yet a confirmation status of repeat testing was not recorded. This would have been beneficial to have in order to compare results to the FPG and HbA1c values obtained for cutoff for diagnosing DMII. (7) Level of Evidence: 1c

DISCUSSION

The purpose if this study was to assess if a HbA1c was sufficient enough to make a unknown diagnosis of diabetes mellitus type 2. From these studies one can gather that a HbA1c is adequate for making a new diagnosis for DMII. The following chart compares the specificity and sensitivity of each HbA1c from each study critiqued in this study. Also, each study uses a different HbA1c cutoff that they gathered from their cohort or cross-sectional study which is also included in the chart below.

Study

Sensitivity

Specificity

HbA1c used for Diagnosis

Diagnosing Type 2 Diabetes Mellitus: in Primary Care, Fasting Plasma Glucose and Glycosylated Hemoglobin Do the Job

63.3%

93.4%

5.94%

Comparison of A1c and Fasting Glucose Criteria to Diagnose Diabetes Among U.S. Adults

72.5%

96.5%

> 6.0%

A1c and Diabetes Diagnosis: The Rancho Bernardo Study

44%

79%

6.5%

Diagnostic value of glycated haemoglobin (HbA1c) for the early detection of diabetes in high-risk subjects

81.8%

84.9%

6.1%

Study #1 discussed the option of performing a combination of HbA1c and a FPG test. This exhibited to be most the most poignant result with a specificity/sensitivity of 92.2 and 95.1, respectively. In study #2, it also agreed that a HbA1c and a FPG level provided the most assured diagnosis for DMII. However, this study had the most discordant results and was probably due to the fact of its subject population. It stated that the results may have been due to the fact that assessment of different aspects of glucose metabolism was present (6). Study #3 was performed on a much older population, and focused on the importance of following HbA1c levels to help prevent long term complications of DMII. However, it also stated that a HbA1c would also have a higher sensitivity and specificity if it were performed along with a FPG test. Finally, study #4 agreed on the fact that a HbA1c was very sufficient for screening for DMII, and that it provided much support for diagnosing DMII along with a FPG.

CONCLUSION

This study provided that a HbA1c of approximately 6.0% is a great support to help making the diagnosis of DMII along with a FPG > 125. Some studies have suggested that a HbA1c of this value is suggestive of a diagnosis, however, the studies above advocate that FPG levels should also be obtained to solidify the actually diagnosis of DMII. However, in a recent publication from the JAAP, it states that”an A1c value of 6.5% higher as diagnostic. This value appears to be the level at which a person is at risk for developing the complications of diabetes. A diagnosis should be confirmed with a repeat A1c test, unless clinical symptoms and a glucose level higher than 200 mg/dL are present (5).” From this statement one can confer that the patient described above in the clinical case portion of this paper, does indeed warrant the diagnosis of DMII on the basis of a HbA1c of 13.0%, the presence of clinical symptoms, and the glucose elevation of 420 mg/dL.