Chapter  128:  Diagnosis, Prognosis, and Treatment of Impaired Glucose Tolerance and Impaired Fasting Glucose

Objective of This Systematic Review

The goal of this systematic review is to evaluate the state of the evidence in the areas of the diagnosis, prognosis, and treatment of IFG or IGT. This evidence report was requested by the American College of Physicians-American Society of Internal Medicine; other partners were the American Academy of Pediatrics and the American Academy of Family Physicians.

Key Questions

Preliminary questions were subsequently modified and refined in consultation with the partner medical agencies, the Agency for Healthcare Research and Quality, and McMaster University Evidence-based Practice Center. The revised key questions are as follows:

• 1

  Diagnosis—What is the reliability of the diagnosis of IFG or IGT (e.g., does individual variability or measurement error require multiple measurements to ensure reliability of diagnosis)? What is the relationship between IFG and IGT?

• 2

  Prognosis—For those identified with IFG or IGT, what are the short- and long-term risks for developing the following outcomes:

  • a)

    Progression to DM or reversion towards normal glucose tolerance or fasting glucose level.

  • b)

    Cardiovascular events and stroke (fatal and nonfatal).

  • c)

    Microvascular disease, specifically retinopathy and nephropathy as measured by proteinuria, microalbuminuria, elevated creatinine, albumin-to-creatinine ratio in the urine, dialysis, or renal transplant.

• Does this risk vary by subpopulation, such as sex, race, obesity, age or other risk factors (e.g., blood pressure, elevated lipid levels)?

• 3

  Treatment—What is the effectiveness of pharmaceutical and behavioral interventions for reducing the risks associated with IFG or IGT on the following outcomes:

  • a)

    Delay in onset of DM or reversion towards normal glucose tolerance or fasting glucose level.

  • b)

    Reducing risk for cardiovascular events and stroke (fatal and nonfatal).

  • c)

    Reducing risk for microvascular disease, including early markers such as retinopathy/proteinuria.

  • d)

    Improving other metabolic parameters, independently associated with increased risk, such as blood pressure and lipid levels.

    Are some treatments more effective than others for any of the above outcomes, and does the effectiveness of interventions vary by subpopulation (e.g., age, sex, and obesity)?

• 4

  Pediatric population—What is known of the development of IFG or IGT in the pediatric population?

Eligibility Criteria

Primary studies were eligible if they evaluated subjects with IGT or IFG, were published after 1978, and were written in the English language. Excluded publications included systematic reviews, narrative reviews, editorials, letters to the editor, unpublished position papers, consensus conference reports, and practice guidelines.

Study design eligibility varied with the research question:

• Diagnosis—All study designs with a maximum of 8-week retest for reproducibility were eligible.

• Prognosis—Any prospective cohorts, or randomized or controlled clinical trials (RCTs) were eligible for evaluation (with a minimum followup of 1 year).

• Treatment—Only RCTs that analyzed the effects of lifestyle, behavioral, or pharmaceutical treatment (with a minimum followup of 6 months) were eligible.

• Pediatric population—All study designs for children age 0 to 18 years were eligible.

The study had to include an IFG or IGT group as the study population or analyzed as a subgroup. The specific criteria reference (for example, WHO-85) used within a study was checked relative to the procedures described in the methods and results sections of each study. The following inclusion/exclusion criteria were used for the testing procedure for dysglycemia:

• All testing must have been done on venous blood plasma or venous blood serum, not on whole blood or on capillary samples.

• For oral glucose tolerance testing (OGTT), the subject must have been given 75 g of oral glucose (1.75 g per kg to maximum of 75 g for children) and measurement taken at 2 hours post-glucose ingestion (2-hr PG).

• All measurements must have been done in a laboratory and not with a point-of-care device or not undertaken in an acute care setting, such as an emergency ward following a myocardial infarction (MI) or pneumonia.

Study outcomes included glycemic disturbances, nonfatal cardiac outcomes, fatal cardiac outcomes, mortality, lipid and blood pressure disorders, amputation, nephropathy, and ocular problems.

Literature Search Strategy

A comprehensive search was undertaken to capture all relevant studies. In addition to MEDLINE^®, HealthSTAR, CINAHL^®, AMED (alternative medicines), PsycINFO^®, and EMBASE^®, the personal files of the local research team and the reference lists of included articles were searched from 1979 onward.²

Study Selection and Extraction

The title and abstract lists and the full-text papers were screened using the eligibility criteria, standardized forms, and a guide manual. Data from the Access database were summarized into summary tables, which included data about the general study characteristics (study design, location, source of funding, population, mean age, and diagnosis criteria), interventions, and outcomes assessed.

Studies were grouped according to classification of the IFG and IGT status. Five categories were considered, including: 1) isolated IGT (I-IGT), 2) isolated IFG (I-IFG), 3) non-isolated IGT, 4) non-isolated IFG, and 5) combined IGT/IFG.

A classification of I-IGT indicates that 2-h OGTT level was between 7.8 and 11.0 mmol/L and the fasting plasma glucose (FPG) level was less than 6.1 mmol/L. Similarly, the I-IFG classification indicates that the 2-h OGTT level was less than 7.8 mmol/L and the FPG level was between 6.1 and 7.0 mmol/L.

The isolated classifications indicate that both forms of glucose testing were undertaken and only one of the two tests was abnormal. Each eligible study was rated for quality using standardized instruments.^{3, 4}

Key Question 1: Diagnosis

Fifty-three studies provided data on the reproducibility of repeat testing of fasting glucose or OGTTs, comparison of IGT diagnosis by different criteria, and the relationship between IGT and IFG diagnosis in the same population.

Reproducibility of IGT and IFG Tests. Five reports of four studies⁵–⁹ assessed the reproducibility of the OGTT for diagnosis of IGT and three reports of two studies^{6, 7, 9} assessed the reproducibility of FPG for the diagnosis of IFG in publications after 1978. All repeat tests were done within 6 weeks of the first test. The populations studied were mostly Caucasians,⁵–⁷ except for two reports of one study on Hong Kong Chinese.^{8, 9}

The study populations were subgroups of larger studies and did not provide detailed characterization for the subgroup. All studies used the same classification criteria. IGT was FPG < 7.8 mmol/L and 2-hr PG 7.8 to 11.0 mmol/L and IFG was FPG 6.1 to 6.9 mmol/L.

The kappa coefficients for IGT ranged from 0.04 to 0.56, indicating poor to moderate agreement. The proportion of participants classified as IGT by the first OGTT and upon repeat testing ranged from 33 percent to 48 percent. A similar proportion (range 39.3 percent to 46.2 percent) of participants was reclassified as normal glucose tolerance, with the remainder reclassified as DM (range 6 percent to 12.6 percent).

Two studies retested participants based on FPG for IFG.^{6, 9} The kappa coefficients for these studies were 0.22 and 0.44, indicating fair to moderate reproducibility. The proportions of participants classified as IFG by the first FPG and upon repeat testing were 63.7 percent and 51.4 percent, respectively. The reclassified subjects had mostly normal fasting glucose with some newly diagnosed DM. Two studies that evaluated coefficient of variation for biological variation (CV_I) for repeat testing gave similar CV_I for FPG (6 percent and 6.3 percent) and 2-hr PG (18 percent and 16.6 percent) concentrations, indicating consistency in variation between the different populations studied.^{5, 7}

Comparison of IGT diagnosis using different criteria. Only four studies¹⁰–¹³ provided data for a comparison between diagnosis using different IGT criteria (i.e., using both IFG and 2-hr PG concentrations for classification). Studies that assessed IGT based on the 2-hr PG concentration only (WHO epidemiological criteria) were excluded.¹⁴

The characteristics of the study populations represent a broad spectrum of populations (Asian, Dutch, Pima Indians, and women with previous gestational DM). The IGT criteria included were WHO-85, WHO-98, and WHO-99. All of these criteria use a 2-hr PG range of 7.8 to 11.0 mmol/L, but the WHO-85 criteria use an FPG cutpoint of < 7.8 mmol/L whereas both the WHO-98 and WHO-99 criteria use a cutpoint of < 7.0 mmol/L. More IGT diagnoses were made using a FPG cutpoint of < 7.8 mmol/L (13.6 percent to 31.5 percent) than 7.0 mmol/L (8.3 percent to 29.7 percent). There were fewer cases of I-IGT (6.0 percent to 11.9 percent) compared to IGT regardless of the FPG cutpoint.

Relationship between IGT and IFG. Forty-nine studies provided data on the relationship between diagnostic criteria for IGT and IFG. Most studies were prospective cohort studies (n = 14) and cross-sectional studies (n = 31). Data were extracted to give seven classifications:

• 1)

  IGT—2-hr PG 7.8 to 11.0 mmol/L.

• 2)

  IGT—FPG < 7.8 mmol/L and 2-hr PG 7.8 to 11.0 mmol/L.

• 3)

  IGT—FPG < 7.0 mmol/L and 2-hr PG 7.8 to 11.0 mmol/L.

• 4)

  IFG—FPG 6.1 to 6.9 mmol/L.

• 5)

  I-IGT—FPG < 6.1 mmol/L and 2-hr PG 7.8 to 11.0 mmol/L.

• 6)

  I-IFG—FPG 6.1 to 6.9 mmol/L and 2-hr PG < 7.8 mmol/L.

• 7)

  Combined IFG/IGT—FPG 6.1 to 6.9 mmol/L and 2-hr PG 7.8 to 11.0 mmol/L.

Comparison of studies that present both IGT and I-IGT data show that the number of participants in the IGT classification is approximately 40 percent greater than in the I-IGT group (p < 0.0001). Also, IGT classification using the limited criteria, which omits the fasting plasma glucose value, classified 10 percent more participants as IGT (p = 0.0033). Evaluation of studies containing data for all classification categories (n = 16) show a change in proportion between each classification group and between studies. In general, the proportion of participants decreased with increased stringency of the diagnostic criteria—that is, IGT as 2-hr PG > IGT as FPG and 2-hr PG > I-IGT > IFG > I-IFG > IFG/IGT.

The prevalence of IGT and IFG varied greatly among studies ranging from a few percent to over 30 percent. Comparisons between categories of IGT and IFG were significant (p < 0.01) for all combinations except for I-IGT versus IFG and I-IFG versus IGT/IFG. Correlations were much higher for IGT and I-IGT than for IFG and I-IFG.

Key Question 2: Prognosis

A total of 104 studies met the initial eligibility criteria. From these, only some provided sufficient data (frequency counts) versus a reference group of subjects with normal glucose to estimate the following:

• Annualized risk per 100 persons in the exposed group.

• Unadjusted annualized relative risk (RR), with the confidence interval (CI).

• Risk differenceⁱ

• Attributable risk (AR), expressed as a percentage for the observed study duration.

All included studies prospectively followed cohorts; 90 were observational studies and 14 were RCTs (placebo arm only). The duration of followup varied from 1 year to 18 years. Five studies¹⁵–¹⁹ evaluated women only, and nine studies²⁰–²⁸ evaluated men only. The mean age and the ranges varied significantly among studies, but most included middle-aged and older subjects. There was a broad representation of populations.

The measures of risk were calculated from data provided in prospective studies that included both normal and dysglycemic people in either observational studies or the placebo arm of RCTs. The risk estimates for all outcomes were classified into five diagnostic groups: 1) IGT, 2) I-IGT, 3) IFG, 4) I-IFG, and 5) combined IFG/IGT. Findings are summarized for three main outcomes: progression to DM, cardiovascular disease (CVD) outcomes (fatal and nonfatal), and mortality.

Risk for progression to DM. The number of studies that provided data for the five classification groups varied. Studies with IGT subjects (n = 36) were the most numerous, whereas five studies included people with IFG and three studies included people with I-IGT, I-IFG, and both IGT/IFG.

Annualized risk per 100 persons in the exposed groups. The minimum and maximum annualized risk estimates for each of the five dysglycemic classification groups are as follows:

• IGT group—1.83 (minimum) to 34.12 (maximum)

• I-IGT group—4.35 to 6.35

• IFG group—1.60 to 23.44

• I-IFG group—6.07 to 9.15

• IGT/IFG group—9.96 to 14.95.

Two studies and four RCTs had high annualized risk estimates, and these included populations with many risk factors for DM. The variation in the annualized risk per 100 persons observed are likely related to the different populations, mean age, and the sample sizes evaluated within these studies.

Unadjusted annualized relative risk. Three of the 28 studies²⁹–³¹ within the IGT classification group were nonsignificant, indicating no association with IGT and progression to DM. Most of these studies had small sample sizes. In the remaining studies, the unadjusted annualized RR with 95 percent CI varied as a function of the diagnostic groups in the following manner:

• IGT group— range 3.58 (95 percent CI 2.12 to 6.06) to 10.60 (6.38 to 17.60)

• I-IGT group—3.51 (2.22 to 5.54) to 8.63 (5.46 to 13.64)

• IFG group—2.40 (1.71 to 3.37) to 9.04 (6.28 to 13.03)

• I-IFG group 5.05 (2.86 to 8.90) to 9.85 (6.65 to 14.60)

• IGT/IFG group—5.50 (2.86 to 8.90) to 20.69 (12.51 to 34.22).

There were fewer studies in the IFG and I-IFG diagnostic categories than for IGT, and, as such, interpretation across classification groups may be limited.

Meta-analysis was undertaken with unadjusted annualized RR for DM where sufficient numbers of studies were available for combining. The pooled estimates with the 95 percent CI are as follows:

• IGT group—6.02 (95 percent CI 4.66 to 7.38)

• I-IGT group—5.55 (3.15 to 7.95)

• IFG group—4.70 (2.71 to 6.70)

• I-IFG group—7.24 (5.30 to 9.17)

• IGT/IFG group—12.21 (4.32 to 20.10).

All pooled estimates were significant. Heterogeneity tests were significant for all of the dysglycemic groups except for the I-IFG group. Sensitivity analyses did not affect the significance of the Q test for heterogeneity.

Attributable risk in the exposed group. High estimates of AR were calculated for the outcome of DM in dysglycemic individuals. Estimates for each dysglycemic group are as follows:

• IGT group—range 52.8 percent to 97.0 percent

• I-IGT group—68.8 percent to 86.6 percent

• IFG group—57.3 percent to 86.9 percent

• I-IFG group—77.1 percent to 88.5 percent

• IGT/IFG group—78.6 percent to 93.3 percent.

Risk for nonfatal CVD outcomes. Estimates of risk for any nonfatal CVD outcomes were based on six studies. The outcomes characterizing CVD included atherothrombosis, nonstenotic atherosclerosis, clinical MI, percutaneous transluminal coronary angioplasty (PTCA), stroke, unstable angina, heart failure, and combinations of these (major event or any event). Study durations varied from 5 to 9 years and studies were published from 1998 forward. Three of the five studies^{18, 32, 33} evaluating IFG as the risk factor are RCTs.

Annualized risk per 100 persons in the exposed groups. Estimates of annualized risk per 100 persons varied between the types of CVD events. The highest observed annualized risk was within the only IGT study for the outcome of nonstenotic atherosclerosis.³⁴ The lowest observed annualized risk was for stroke in people with IFG.¹⁸ The annualized risk estimates for any nonfatal CVD event are as follows: IGT group—11.58 to 12.39; IFG group—0.63 to 9.68.

Unadjusted annualized relative risk. Only two studies had significant unadjusted annualized RRs. The single study within the IGT group had similar RR and CI estimates: 2.43 (95 percent CI 1.44 to 4.10) and 2.46 (1.46 to 4.12) for both atherothrombosis groups.³⁴ Five studies evaluated subjects in the IFG group and RR estimates varied from 1.24 (1.08 to 1.43) to 1.41 (1.17 to 1.69).

In a meta-analysis, nonfatal CVD outcomes were combined according to these subgroups: 1) PTCA and coronary artery bypass graft, 2) stroke, and 3) any other major cardiovascular event. Tests for heterogeneity were not significant. Only one combination for the IFG of any major cardiovascular event was significant with an overall estimate of 1.28 (CI 1.15 to 1.41, p = 0.0001).

Attributable risk in the exposed group. The AR for CVD outcomes was higher in the IGT group (range 52.8 percent to 52.9 percent) than in the IFG group (0 percent to 32.9 percent).

Risk for fatal CVD outcomes. Eight studies reported fatal CVD outcomes. Some studies subdivided the outcomes into ischemic heart disease, cardiocerebrovascular disease, and coronary heart disease (CHD). There were no eligible studies to evaluate the I-IFG and IFG/IGT combined classifications. The duration of the studies varied from 7 to 18 years. Three of the studies^{21, 25, 27} were based on a male-only cohort (Paris Police study) for the IGT, I-IGT, and IFG groups. Similarly, one study within the IFG group included only postmenopausal women with a history of MI.¹⁸ Two other studies in the IFG group recruited subjects with a history of either MI or CHD.^{35, 36}

Annualized risk per 100 persons in the exposed groups. The annualized risks in the exposed group per 100 persons are as follows: IGT group—0.06 to 0.76; I-IGT group—0.23 to 0.34; IFG group—0.10 to 1.54. The differences in annualized risk are likely a function of the different study populations and categorizations of the CVD mortality subgroup classification.

Unadjusted annualized relative risk. Only four studies^{25, 35, 37, 38} had unadjusted annualized RRs that were significant. The calculated estimates are as follows:

• IGT group— range 1.67 (95 percent CI 1.23 to 2.26) to 3.08 (1.47 to 6.47)

• I-IGT group—1.59 (1.07 to 2.28) to 1.72 (1.23 to 2.41)

• IFG group—1.32 (1.04 to 1.67)ⁱⁱ

In a meta-analysis, within the IGT group, CVD and ischemic heart disease were grouped and the pooled overall estimate of relative risk was 1.66 (95 percent CI 1.21 to 2.11). Within the IFG group, estimates for the CHD/CVD subgroup (1.25 [0.99 to 1.51] and ischemic-related disease subgroup (1.27 [1.06 to 1.54]) were pooled; these estimates did not differ substantively. Overall, the pooled estimates do not provide evidence of a significant association with IFG or IGT and fatal CVD outcomes.

Attributable risk in the exposed group. The AR for fatal CVD outcomes varied as follows: IGT group—range 24.8 percent to 67.3 percent; I-IGT group—36.8 percent to 41.2 percent; IFG group—11.8 percent to 39.5 percent. With the exception of one study,³⁸ the AR did not exceed 41 percent. Tominaga et al.³⁸ evaluated CVD outcome in both the IGT and IFG groups concurrently, and the AR was 67.3 percent and 39.5 percent, respectively.

Risk for mortality. In general, most studies reporting mortality outcomes had the largest sample sizes and the longest followup duration (up to 18 years). Eligible studies included the IGT, I-IGT and IFG classifications; the I-IGT group was based on two studies on the same cohort.

Annualized risk per 100 persons in the exposed group. The annualized risks per 100 persons for mortality (all-cause, cancer, and cirrhosis categories) are as follows: IGT group—0.09 to 2.44; I-IGT group—0.10 to 1.34; and IFG group—0.56 to 1.39.

Unadjusted annualized relative risk. The all-cause mortality estimates along with 95 percent CI are as follows:

• IGT group—range 1.36 (95 percent CI 1.12 to 1.66) to 3.18 (1.79 to 5.63)

• I-IGT group—1.60 (1.33 to 1.92) to 7.19 (3.37 to 15.37)

• IFG group—1.18 (1.03 to 1.35) to 1.45 (1.27 to 1.66).

One study showed a strong association between I-IGT and the outcome of death due to cirrhosis (RR 7.19)²⁵ for male police officers and also showed the highest AR (86 percent) for mortality outcomes. Three studies^{21, 35, 38} evaluated both all-cause mortality and CVD mortality; both the RR and the AR estimates were approximately double in magnitude for the CVD-related mortality relative to all causes with the exception of one study.³⁸ Two studies^{27, 39} compared all-cause mortality to cancer-related deaths, and the RR and AR did not differ substantively for these two mortality outcomes.

Meta-analysis for all-cause mortality in the IGT and IFG groups was undertaken. The overall pooled estimates were 1.48 (95 percent CI 1.09 to 1.86) for the IGT group and 1.21 (95 percent CI 1.05 to 1.36) for the IFG group.

Attributable risk in the exposed group. The ARs for all-cause mortality are as follows: IGT group—range 0 percent to 67.2 percent; I-IGT group—35.1 percent to 86.0 percent; IFG group—13.9 percent to 61.2 percent. The single study with an AR equal to 0 percent has been previously noted for its sample size issues. As with the other risk metrics for mortality, the AR was highest (86 percent) for cirrhosis-related mortality.

Key Question 3: Treatment

Twenty-three reports of 14 RCTs published between 1992 and 2003 evaluated lifestyle or pharmacotherapeutic interventions in adults with IFG or IGT. Duration of followup ranged from 6 months to 6 years. Studies involved 14 to 3,234 participants from Europe, North America, Australia, and Asia with their mean ages ranging from 37.5 to 70 years. Interventions included diet and exercise, oral hypoglycemic agents (metformin, acarbose, and chromium), a statin (pravastatin), and an ACE inhibitor (enalapril). Outcomes included progression to DM or reversion to normal glucose tolerance, cardiovascular events, mortality, and effects on blood pressure and lipid levels.

Progression to DM or reversion to normal. Most studies of the effects of lifestyle or pharmacotherapeutic interventions involved people with IGT.

Lifestyle interventions. Six RCTs evaluated the effect of lifestyle interventions on the risk for developing DM or reverting to normal glucose tolerance in adults with IGT. Intensive combined diet and exercise programs that involved frequent study visits were compared with lifestyle advice alone in five studies.⁴⁰–⁴⁴ One study⁴¹ compared an exercise program with advice alone, and two studies^{16, 41} evaluated the effect of dietary intervention alone. One trial, the Diabetes Prevention Program, also included a metformin arm.⁴⁰

All but one of the trials that evaluated a combined diet and exercise program found a significant reduction in the risk for developing DM, or a higher rate of reversion to normal glucose tolerance, with aggressive lifestyle modification. The absolute risk reduction of progressing to DM per year in the studies was between 1.6 percent and 7.1 percent, corresponding to a number needed to treat for 1 year to prevent a case of DM between 14 and 62. Dietary intervention alone significantly reduced the risk for progressing to DM in one trial⁴¹ but had no effect in a second study.⁴⁵ The trial⁴¹ that evaluated an exercise intervention alone showed a significantly reduced rate of progression to DM (absolute risk reduction 3.9 percent, number needed to treat 25.5, relative risk reduction 37 percent).

Pharmacotherapeutic interventions. Four RCTs evaluated the effects of pharmacotherapeutic interventions on the risk for developing DM in people with IGT.^{40, 46}–⁴⁸ These studies assessed the effect of acarbose and metformin.

The study⁴⁶ with acarbose demonstrated a reduced risk of progressing to DM (32 percent versus 42 percent; relative risk reduction 0.25, 95 percent CI 0.10 to 0.37). This effect did not vary by age, sex, or body mass index (BMI). The study also demonstrated an increased rate of reversion to normal glucose tolerance with acarbose relative to placebo (35 percent versus 31 percent, p < 0.0001).

A large study⁴⁰ found a significantly reduced risk for progressing to DM when taking metformin relative to placebo (7.8 percent versus 11.0 percent per year; relative risk reduction 0.31, 95 percent CI 0.17 to 0.43). Two smaller studies^{47, 48} in people with IGT found no difference in those treated with metformin.

The effect of enalapril in people with IFG and left ventricular dysfunction was assessed in a retrospective post-hoc subgroup analysis. This study⁴⁹ found a decreased risk for progression to DM in the enalapril arm relative to the placebo arm (3.3 percent versus 48 percent, p = 0.0001). The effect of pravastatin on the development of DM in people with IFG and a previous MI was assessed in a retrospective post-hoc subgroup analysis. This study found no effect on the rate of development of DM, based on a fasting blood glucose level of ≥ 7 mmol/L, or reported use of oral hypoglycemic medication or insulin.

Lifestyle versus pharmacotherapeutic interventions in people with IGT. Only one trial to date, the Diabetes Prevention Program,⁴⁰ has directly compared lifestyle intervention with pharmacotherapeutic intervention for the prevention of diabetes in people with IGT. It found a significantly lower risk for progressing to DM with aggressive lifestyle intervention compared with taking metformin (4.8 percent versus 7.8 percent per year; relative risk reduction 0.39, 95 percent CI 0.24 to 0.51), especially in individuals 60 years of age or older.

Cardiovascular event outcomes. No RCTs of lifestyle interventions evaluated cardiovascular outcomes.

Pharmacotherapeutic interventions in people with IGT. A single trial⁵⁰ evaluated the effect of acarbose on cardiovascular event rates in people with IGT.

The primary outcome found a significant reduction in the risk for developing a major cardiovascular event in the acarbose arm compared with the placebo arm of the study (relative risk reduction 0.49, 95 percent CI 0.05 to 0.72, absolute risk reduction 2.5 percent).

Pharmacotherapeutic interventions in people with a previous MI and IFG. Two post-hoc retrospective subgroup analyses evaluated the effect of pravastatin therapy on cardiovascular event rates in people with a previous MI and IFG. In one trial,³² the rate of cardiovascular death or nonfatal MI was significantly lower in the pravastatin group; the relative risk was not significantly different from the values for the post MI patients. In a second trial¹⁸ the relative risk for the outcome of cardiovascular death or a nonfatal MI in individuals with IFG was also not significantly different from those within individuals with normal fasting glucose levels at baseline.

Mortality outcomes. One trial⁴¹ reported the effect of lifestyle intervention on total mortality rates in individuals with IGT. One trial¹⁸ reported the effect of statin therapy on mortality rates in individuals with a previous MI and IFG. In both trials, mortality rates did not differ significantly between groups.

Effects on blood pressure and lipid levels. All studies involved people with IGT.

Lifestyle interventions. Three RCTs^{43, 44, 51} evaluated the effect of lifestyle interventions on blood pressure and lipid levels. Significant differences (decline) in blood pressure (systolic and diastolic) were found in two studies and in lipid levels (ratio of total to HDL cholesterol and serum triglycerides only) in one study.

Pharmacotherapeutic interventions. Four RCTs reported the effects of oral hypoglycemic agents on blood pressure and lipid levels. Two trials^{47, 48} reported the effects of metformin on blood pressure levels in people with IGT and demonstrated no significant effect of metformin on blood pressure or lipid levels. One study^{46 50} reported the effects of acarbose therapy on blood pressure and lipid levels in people with IGT and found significant differences in blood pressure (systolic and diastolic) and hypertension (defined as a blood pressure of at least 140/90 on two consecutive visits or the addition of antihypertensive medications between visits). The trial noted a significant reduction in triglyceride levels. A trial of chromium⁵² found no significant effects on lipid levels.

Key Question 4: Pediatric Population

All articles that met the general criteria (English language, full-text publication, published since 1979, and results for IFG or IGT analyzed separately from other study populations) and included children with IFG or IGT were collected (36 articles). Of these, a subset of five articles met the criteria for diagnosis, prognosis, or treatment according to the criteria outlined in the methodology. These articles are included in the analysis of their respective sections above.

Four articles included within the analysis (one diagnosis, three prognosis) included participants 15 to 18 years of age, but the pediatric data were not presented separately.^{17 53}–⁵⁵ These studies were therefore excluded from the pediatric analysis. Nineteen studies were excluded for the following reasons: nine discussed cystic fibrosis,^{56 64} one discussed endemic fluorosis,⁶⁵ one dealt with Turner's syndrome,⁶⁶ six related to type 1 DM risks,^{55 67 71} and no specific pediatric data could be extracted in two articles.^{72 73}

Thus, 13 of 36 articles had extractable pediatric data in articles relevant to either the prevalence, diagnosis, prognosis, or treatment of IFG and/or IGT. The information from these articles forms the basis of the analysis that follows.

Most studies (12 out of 13) addressed the prevalence of IFG or IGT in various at-risk populations and in the population at large. Two studies compared IFG and IGT diagnosis in children. Four studies examined longitudinal followup of a cohort of children and addressed the prognosis of IFG or IGT. One study examined treatment in an open-label trial with metformin.

Prevalence. As DM in childhood was initially recognized in Aboriginal populations, most prevalence studies examine these groups. Population-based prevalence of IGT in childhood Aboriginal populations varies from 3.5 percent⁷⁴ in Tuvalu to 6.25 percent of Australian Aboriginals aged 7 to 18 years.⁷⁵

The prevalence of IFG has been studied in one population-based study. The Third National Health and Nutrition Examination Survey (NHANES III), conducted from 1988 to 1994, measured fasting glucose in 1,083 adolescents age 12 to 19. IFG (glucose 6.1 – 6.9 mmol/L) was present in 1.8 percent (n = 20/1,083). Of these 20 children, 4 were non-Hispanic white, 9 were non-Hispanic black, and 7 were Mexican American. The majority of the children were overweight (mean BMI at 86th percentile), but the range extended from the 10th to 99th percentile. Prevalence of IGT in children not “at-risk” is available from the control group of a single study in which 2.5 percent of 80 children age 10 to 16 had IGT.⁷⁶

The prevalence of IGT in obese children has been examined in two studies^{77, 78} of children referred to a tertiary care center for obesity management; IGT was found in 25 percent of children (age 4 to 10 years) and 21 percent of adolescents (age 11 to 18 years) in a U.S. study and in 4.2 percent of 6- to 18-year-olds using the same diagnostic criteria in an Italian population. In the U.S. study, 51 percent of those with IGT were non-Hispanic white, 30 percent were non-Hispanic black, and 19 percent were Hispanic (compared to 58 percent, 23 percent, and 19 percent, respectively, in the population studied). “Silent” DM was diagnosed in four participants (two non-Hispanic black and two Hispanic).

Other “at-risk” populations have been identified. These include children with a history of DM in first degree relatives. In a study of 150 Latino children with a family history of DM, 28 percent were noted to have IGT.⁷⁹ Furthermore, 25 percent of Hispanic children whose sibling had type 2 DM had IGT.

Offspring of mothers with pregestational or gestational DM (ODM) also have a higher prevalence of IGT. In a longitudinal study, the prevalence of IGT in ODM was 1.2 percent in children < 5 years (n = 168), 5.4 percent in 5- to 9-year-olds (n = 111), and 19.3 percent (95 percent CI 12.1 to 28.6) in 10- to 16-year-olds (compared to 2.5 percent [95 percent CI 0.4 to 8.1] in controls).⁷⁶ Although the control group was somewhat lighter (BMI 20.3 ± 4.0 versus 22.8 ± 5.4 kg/m²) and had 37 percent of participants other than Caucasian compared to 51 percent in the ODM group, it is unlikely that these differences would account for the difference in IGT prevalence. Within this same cohort, 36 percent of those in the ODM group have had at least one abnormal OGTT result by 14 to 17 years of age.⁸⁰

Finally, 11 of 21 adolescents with polycystic ovary syndrome had abnormal OGTT results (9 IGT, 2 DM).

The prevalence of IGT is related to increasing age in several studies, but few studies have examined children less than 10 years of age. Children under 10 with obesity have IGT rates comparable to adolescents, although type 2 DM is reported with much less frequency in this young group. Two longitudinal studies with repeated OGTT in Aboriginal and ODM children suggest that rates of IGT increase with increasing age, particularly during the peripubertal period.

Diagnosis. A comparison of IGT with IFG is presented in two articles,^{77, 78} and IFG and hemoglobin A1c are compared in the NHANES III study.⁸¹ In obese children, 6.6 percent of children⁷⁸ and less than 0.08 percent of children and adolescents⁷⁷ with IGT had IFG, indicating that this method of screening for IGT is very insensitive.

Similarly, hemoglobin A1c is a poor screen for IFG in children. The reproducibility of OGTT testing has not been well studied. Sinha et al.⁷⁷ showed that, upon retesting, 10 of 10 children (4 with normal glucose tolerance and 6 with IGT) had the same categorization 3 months later. One article⁸² included in the full review for reproducibility of diagnosis that included adolescents concluded that the reliability of test results was likely lower in younger populations.

Prognosis. The prognosis of IGT in childhood and adolescence has not been well studied. Three studies had longitudinal data in IGT, but the numbers were very small and did not allow a prediction or rate of conversion from IGT to DM. All of these longitudinal studies were in high-risk populations (two in Aboriginal populations in the United States and the South Pacific) and one in ODM.

Treatment. Treatment of IGT in childhood has been examined in a single small open-label trial of metformin for 3 months in 15 adolescents with polycystic ovary syndrome and IGT.⁸³ Eight of 15 children had normal glucose tolerance when re-evaluated after 3 months of metformin therapy. This was associated with a significant decline in BMI, although there was no significant change in fat mass.

Diagnosis

An accurate diagnosis of DM is required because the consequences for the individual are considerable and lifelong. The diagnosis of IFG or IGT is used as a risk indicator for future DM and/or CVD. The problem with these arbitrary classifications is that test reproducibility is poor, and this encourages repeat testing that adds to the uncertainty and confusion of the diagnosis when results are different.

Reproducibility of IGT and IFG. The observed reproducibility for both IGT and IFG classification in these studies was roughly 50 percent. The kappa coefficients for the IGT category were quite low and indicate overall fair agreement. The potential factors contributing to the variation and poor reproducibility were not assessed for this review.

The probability that a significant change has occurred in serial measurements can be estimated by calculating the reference change value (RCV). For FPG, the RCV = 2^½ * 1.96 * (1.4² + 6.3²)^½ or 17.9 percent. For 2-hr PG, the RCV = 2^½ * 1.96 * (1.4² + 16.6²)^½ or 46.4 percent. The difference between two fasting glucose values would therefore need to be greater than 17.9 percent to be significantly different. A lower RCV would increase the sensitivity to change, or reduce the variation noise, and could be achieved if the analytical and/or the biological variation are lowered. In the best case scenario, the lowest biological variability reported for fasting glucose was an FPG CVI of 4.8 percent.⁸⁴ If this value is used along with an intra-laboratory imprecision of 1 percent and no bias, the RCV can be reduced to 13.6 percent. This is the very best or lowest amount of variation possible for a fasting plasma glucose measurement.

Comparison of IFG and IGT diagnosis. This review also compared among studies the proportion of participants classified as IGT (2-hr PG), IGT (FPG and 2-hr PG), I-IGT, IFG, I-IFG, and IGT/IFG. Comparisons among these categories were statistically significant except for I-IGT versus IFG and I-IFG versus IGT/IFG. This exemplifies the importance of clearly distinguishing categories as this can affect the proportion of study subjects and the conclusions from prognosis and treatment data.

The reproducibility for both IGT and IFG categorization is poor by both observed and kappa analysis. Because of the large variability in glucose measurement, the absolute FPG and 2-hr PG measurements may be more informative than categorization into IFG and IGT, respectively. Comparison of IGT and IFG categories shows a wide degree of variation among populations. The prevalence of IGT is greater than for IFG in almost all studies. High-risk populations have an equal or greater proportion of IFG compared to IGT diagnoses. Statistically, the proportion of study participants classified as IGT by 2-hr PG alone is greater than if the diagnostic criteria of both 2-hr PG and FPG are used. This will affect the conclusions of prognosis and possibly treatment data in population studies using only the 2-hr PG concentration (WHO epidemiological criteria).

Prognosis

This review provides further evidence of the relevance of the OGTT as a diagnostic test. Despite the many shortcomings of the OGTT reviewed here, it detects a very high-risk group for future DM and may either need to be more accessible to clinicians or replaced by a simpler test that provides comparable predictive information. The OGTT also detects a group at risk for CVD; and if IGT is causally related to CVD, the AR estimates suggest that its treatment may reduce CVD risk by as much as 20 percent to 40 percent.

These studies highlight the relevance of fasting and post-challenge glucometabolic abnormalities to clinically relevant outcomes. Intervention studies have already shown that DM can be prevented in these individuals with some interventions.

Risk for progression to DM. The results of this systematic review clearly show that IGT, IFG, I-IGT, I-IFG, and combined IGT/IFG are strong risk factors for future DM. The combined group has the strongest risk factor, and this observation is not surprising given the fact that the diagnostic threshold for DM is just a farther point along the dysglycemic spectrum than the threshold for either IFG or IGT. Nevertheless, these large risk estimates clearly do suggest that any clinical approach directed at preventing DM should include a policy of detecting IFG or IGT. They do not support suggestions that measures of glucose are not necessary to detect individuals at risk for future DM. However, such a policy may be useful to reduce the number of individuals who require a glucose test.

Risk for CVD outcomes. The reviewed studies provide confirmation that IFG or IGT are risk factors for fatal and nonfatal CVD and are consistent with other studies that were excluded because whole blood or capillary samples were used to assay glucose levels. Moreover, the suggestion that IGT is a greater risk factor for CVD than IFG is supported by this systematic review but is based on the findings of a single study.⁸⁵ This is not surprising given the fact that IGT is detected in response to stressing the physiology with a nonphysiological glucose load, thus exposing a degree of metabolic dysregulation that would not be apparent on the basis of fasting glucose levels alone.

Treatment

Prevention of DM: lifestyle interventions. This systematic review clearly demonstrates that DM can be prevented or delayed with lifestyle modification. All but one of the five studies that evaluated a combined diet and exercise program found significant benefits, with a pooled relative risk of 54 percent for progression to diabetes. The only trial to show no effect of a combined diet and exercise intervention was of short duration (6-month followup). Interventions with diet or exercise alone showed mixed results between studies. Efforts to modify dietary intake and activity levels in individuals at increased risk for developing DM are clearly warranted.

Prevention of DM: pharmacotherapeutic interventions. Only four trials to date have evaluated the effect of pharmacotherapeutic interventions on the risk for developing DM in individuals with IGT. Two of these studies, one involving acarbose and one involving metformin, demonstrated reduced rates of progression to DM with a relative risk reduction of about 25 percent. Given this relative paucity of information, recommendation of pharmacological intervention for the prevention of DM would seem premature at this time.

Pediatric Population

Despite the paucity of population-based studies, several cohort studies in high-risk groups suggest that IGT is a significant and potentially growing problem in the pediatric population. Indeed, larger proportions of children may have IGT than is currently recognized. It is critical to acquire an understanding of the precursors of type 2 DM development in children and youth. However, few conclusions can be made based on the current pediatric literature. Further investigation of prevalence in children and adolescents is necessary to clarify the magnitude of the problem.

Diagnosis. The reproducibility of the diagnosis of IGT with OGTT testing and the clinical significance of IFG versus IGT have not been widely examined in the pediatric literature. Although young age has been implicated as a predictor of poor reproducibility of OGTT results in adults, suggesting that reproducibility may be worse in adolescents and children, this was not the experience in one small pediatric study (n = 10).⁷⁷

Clearly, further investigation of the reliability of diagnostic criteria for IFG and IGT is warranted. Furthermore, given the importance of the prevention of type 2 DM, it may be advantageous to identify children who have disturbed glucose metabolism (insulin resistance and/or beta cell dysfunction) before they develop IFG or IGT.

Prognosis. An understanding of how disturbed glucose metabolism progresses to IGT and to type 2 DM is key to the primary prevention of DM. Currently, details of this progression are completely lacking in the pediatric population. Prevalence data for type 2 DM suggest prognosis may vary with age, pubertal status, and ethnicity. Family history of DM, exposure to a diabetic environment in utero, fitness and physical activity, fat distribution, and characteristics of nutritional intake may also influence the prognosis of IFG and IGT. Longitudinal studies are required to examine mid- and long-term outcomes of IGT and the determinants of outcome in multiple ethnic groups and across a broad age range. Investigation of other metabolic outcomes in children and adolescents with IFG and IGT would further improve our understanding of disturbance in health in this population. Better understanding of the prognosis of IGT in children and adolescents will clarify the need for intervention and contribute to optimal intervention study design.

Treatment. A single study has described the pharmacological treatment of IGT, and no randomly controlled lifestyle intervention has been reported in the pediatric age group. Given the increasing rates of IFG/IGT, research on the optimal approach to the management of these children should be a research priority. This research should compare lifestyle intervention and pharmacotherapy and identify optimal methodologies for young populations and their families. Although glycemic status is a key outcome variable, other metabolic and psychosocial outcomes should also be examined.

Magnitude and Importance of the Problem of IFG/IGT

Diabetes mellitus (DM) and its associated disease outcomes are a growing concern worldwide. The current global prevalence of DM for all ages has been estimated at 2.8% and is predicted to reach 4.4% by 2030.¹ In the United States (U.S.), the prevalence of diagnosed DM was estimated at 5.1% in 1997 for adults between the ages of 40 and 74 years.² In 2002, costs for treating DM and the resulting complications were high, with expenditures and lost productivity estimated at $132 billion in the U.S.³ Due to the high prevalence of DM and the economic and health outcome burdens associated with this disease, there is intense interest in identifying and reducing the impact of risk factors in order to prevent the onset of DM and minimize morbidity.

Impaired fasting glucose (IFG) and impaired glucose tolerance (IGT) refer to the intermediate metabolic states between normal and diabetic glucose homeostasis. One or both of these conditions are thought to be the precursors of DM, but how they progress to overt disease is not well understood. The risk for both macrovascular and microvascular complications increases across the distribution of blood glucose concentrations well below the level for overt DM, and is more strongly associated with post-challenge hyperglycemia than fasting glucose levels. However, it is still unclear from the literature whether this ‘glucose effect’ is independent of other classical risk factors such as blood pressure (BP) and lipids. It is also unclear whether this pathology is related to only blood glucose levels or also involves abnormalities of other metabolites such as free fatty acids. There is accumulating evidence that some microvascular complications emerge in these glucose ‘intolerant’ individuals.

The term IFG was introduced in 1997 to define those individuals with fasting plasma glucose (FPG) between the upper limit of normal FPG and the lower limit of diabetic FPG. IGT refers to those individuals with a 2-hour post-load plasma glucose (2-hr PG) between the upper limit of normal and the lower limit of diabetic.⁴ The International Diabetes Federation IFG/IGT Consensus Workshop concluded, based on these criteria, that IGT and IFG differ in their prevalence, population distribution, phenotype, sex distribution, and risk for total mortality and cardiovascular disease (CVD).⁵ Similarly, the Third National Health and Nutrition Examination Survey found differences in the percentage of the U.S. population with IFG for men (8.8%) versus women (5.0%) and for non-Hispanic white (6.8%), non-Hispanic black (7.0%), and Mexican-American (8.9%) people.² Thus, there is also interest in identifying those subgroups with IFG/IGT who are at greater risk for progression to DM and other health outcomes.

Historical Development of IGT and IFG Classifications

Both IFG and IGT are associated with glycemic disturbances that represent different metabolic processes with different prognostic consequences. Some evidence suggests that IGT is primarily associated with insulin resistance and IFG is associated with impaired insulin secretion and suppression of hepatic glucose output.⁶ Also, IGT is more consistently associated with increased risk for CVD, but IFG is not.⁷ Similarly, the magnitude of the risk for developing DM differs for these two metabolic states.

The historical development of IGT and IFG also has relevance to this systematic review. The diagnostic group of IGT was first defined in 1980 by the World Health Organization (WHO) expert committee on DM.⁸ IGT was not defined as a disease in itself but rather as a metabolic state (or disease process) associated with increased risk for adverse health outcomes. IGT was defined as dysglycemia between normal and DM. Moreover, it was used to identify individuals at high risk for DM and possibly CVD.

The relationship between IFG and IGT has generated controversy about how best to evaluate dysglycemia using either the oral glucose tolerance test (OGTT) or the FPG test. There is particular interest in which of these glycemic disturbance states is a better predictor of progression to DM or CVD. There are also questions regarding the reliability of FPG and OGTT due to the biological variability and the clinical implications for confirming glycemic disturbance and the increased risk for progression to DM and other health outcomes. Moreover, there is a need to evaluate which interventions (pharmacological or lifestyle) are effective in modifying glycemic disturbance in IFG or IGT populations and identify the gaps in the literature to provide direction for future research.

The prevalence of type 2 DM continues to be higher amongst aboriginal peoples of North America (1% of children six to 17 years of age⁹ and 4% of adolescent girls^{10, 11} in a Pima population). Previously undiagnosed DM was also noted in four out of 112 obese adolescents (11 to 18 years) referred to a tertiary-care obesity clinic.¹² Given the increasing trends in childhood obesity, it is imperative that our understanding of the precursors to type 2 DM in children is advanced. Furthermore, as the prevalence of obesity in children and adolescents has increased in North America,¹³ so too has obesity-related co-morbidities, previously thought only to occur in adulthood. Type 2 DM in childhood and adolescence, first recognized in aboriginal populations,^{14, 15} is now also recognized in many other minorities (Black, Hispanic) and in those of European descent.

Diagnosis Question

What is the reliability of the diagnosis of IFG or IGT (e.g., does individual variability or measurement error require multiple measurements to ensure reliability of diagnosis)? What is the relationship between IFG and IGT?

Prognosis Question

For those identified with IFG or IGT, what are the short and long-term risks for developing the following endpoints:

• a)

  Progression to DM or reversion towards normal glucose tolerance (NGT) or fasting glucose level,

• b)

  Cardiovascular events and stroke,

• c)

  Microvascular disease, specifically retinopathy and nephropathy as measured by proteinuria, microalbuminaria, elevated creatinine, albumin to creatinine ratio in the urine, dialysis, and/or renal transplant.

Does this risk vary by subpopulation, such as sex, race, obesity, age, or other risk factors (e.g., BP, elevated lipid levels)?

Treatment Question

What is the effectiveness of pharmaceutical and behavioral interventions for reducing the risks associated with IFG or IGT on the following endpoints:

• a)

  Delay in onset of DM or reversion towards NGT or fasting glucose level,

• b)

  Reducing risk of cardiovascular events and stroke,

• c)

  Reducing risk of microvascular disease, including early markers such as retinopathy/proteinurea,

• d)

  Improving other metabolic parameters associated with increased risk, such as BP and lipid levels.

Are some treatments more effective than others for any of the above endpoints, and does the effectiveness of interventions vary by subpopulation (e.g., age, sex, obesity)?

Pediatric Question

What is known about the development of IFG or IGT in the pediatric population?

The Research Team

A multidisciplinary research team representing epidemiology and systematic review methods (P. Raina, PhD; P.L. Santaguida, PhD), internal medicine and endocrinology (H. Gerstein, MD; D. Hunt, MD), clinical chemistry (C. Balion, PhD), and pediatric endocrinology (K. Morrison, MD) was assembled. The core research team, including experienced staff at the McMaster EPC (L. Booker, BA; M. Gauld, BA; L. Cocking; E. Estrabillo, B.Sc.) and a statistician (H. Yazdi, PhD), participated in regular meetings and reached consensus on key methodological issues. An international Technical Expert Panel (TEP; see Appendix D ^*) was assembled to provide high-level content expertise and participate in conference calls as needed. Participants in this panel include: Vincenza Snow (ACP), Amir Qaseem MD, PhD, MHS (ACP), Rodney Hornbake MD (ACP representative), Tommy Cross MD (ACP representative), Belinda Ireland MD, MS (AAFP), Kevin Patterson MD, MPH (AAFP representative), Francine Ratner Kaufman MD (AAP representative).

A teleconference with the partner organizations, the Task Order Officer (TOO) from AHRQ, the invited technical experts, and the McMaster team was held early during protocol development. The meeting's purpose was to define the scope of the systematic review and to achieve consensus about the preliminary research questions. As a result of these discussions, some modifications were made to the original questions to address, in particular, important gaps in the knowledge needed by family physicians to diagnose patients for IFG or IGT.

Eligibility Criteria

Publication types, year, and language.

These criteria were applicable to all research questions:

• 1)

  Publication year: 1979 forward,

• 2)

  Publication language: English, and

• 3)

  Publication types: primary studies

Excluded: Systematic reviews, narrative reviews, editorials, letters to the editor, theses, unpublished position papers, consensus conference reports, and practice guidelines.

Study design.

Diagnosis question. No study design exclusions for primary studies.

Prognosis question. Primary studies with prospective cohort and randomized control trials (RCT) study designs with at least one year of follow-up.

Excluded: Case-control studies.

Treatment question. Only RCT designs were eligible. However, studies evaluating non-pharmacological interventions (lifestyle, behavioral, or surgical treatment) using non-RCT designs (controlled clinical trials and concurrent cohort trials) were captured in an annotated bibliography and no other data were extracted. (Appendix C)

Pediatric question. All study designs were eligible.

Study population.

General criteria for IFG and IGT classification. Eligible citations had to include IFG or IGT groups as the study population or analyzable subgroup. The criteria for classifying dysglycemia were key to identifying this specific population. The glucose threshold values used to define DM, IFG, and IGT have varied over the past 25 years (see Table 1). The specific criteria reference (e.g., WHO 85) used within a study was noted. For the diagnosis question, additional checks were undertaken to compare the testing procedures described in the methods and results sections of each eligible study.

Laboratory testing procedures.

Laboratory test inclusion:

• 1)

  All laboratory testing for glucose had to be undertaken on venous blood plasma or venous blood serum.

• 2)

  OGTT must have used the following parameters: subject was given 75 g of oral glucose (1.75 g per kg to maximum of 75 g for children) and measurement was taken at two hours post-glucose ingestion.

• 3)

  All measurements must have been done in a laboratory and not with a point-of-care device.

Laboratory test exclusion:

• 1)

  The testing was done on whole blood or on capillary samples.

• 2)

  The laboratory testing was undertaken in an acute care setting (eg. emergency ward, intensive care ward following, for example, a myocardial infarction or pneumonia).

These general criteria for the classification of IFG or IGT were applied to all four questions with the exception of the pediatric question.

Pediatric question. Increased recognition of type 2 DM in children has only occurred within the last two decades. Thus, we anticipated a limited number of articles addressing IFG or IGT in this population. Children were defined as 18 years of age or under. Any study that evaluated children, even if it did not meet all of our eligibility criteria for diagnosis, prognosis, or treatment, was included. We indicated “Include for Children” in the screening form (see Appendix B) if the study met the publication type, language criteria, and testing criteria for IFG and IGT.

Study interventions.

Diagnosis question. The research questions on the diagnosis of IFG or IGT were formulated using two distinct characteristics:

• 1)

  Test-retest reliability, and

• 2)

  The relationship between IFG and IGT.

The reliability of the IFG and IGT diagnostic criteria was assessed using a maximum boundary of eight weeks between the first test and repeated testing. There was consensus among the TEP that true change in the disease status would not likely occur during this time interval, and it represented a typical interval for repeat tests in clinical practice.

For the relationship between the 2-hour OGTT and the FPG and the subsequent diagnosis of IGT and IFG, there was a general consensus that the two tests did not necessarily measure the same population. It was recognized that the literature does not agree as to which test is best or should be used to diagnose glycemic disturbance (IFG versus IGT); thus the degree of association between these two diagnostic tests was of interest.

It was also of interest to evaluate the variation between repeated laboratory measures in subjects. This question was not intended to examine the biochemical basis of the test. Instead, the intention was to describe the change in diagnostic category between having IFG, IGT, normal glycemic levels or DM on repeat testing. It was also of interest to describe any related factors that could contribute to the observed variance.

For the relationship between IFG and IGT, studies were included if they used both the FPG and the OGTT to evaluate subjects for dysglycemia.

Treatment question. There was no restriction on the types of interventions used on an IFG or IGT population. It was expected that these interventions would be categorized into four groups: pharmacological, behavioral, lifestyle, or surgical. Moreover, a minimum follow-up of six months was required.

The specification of interventions was not applicable for the prognosis and pediatric questions.

Study outcomes.

The outcomes selected for this study applied to both the prognosis question and the treatment question. Nine disease categories were selected, and then possible medical or procedural outcomes were further specified within each of these categories (see list below). For example, within the cardiovascular disease category, 11 different cardiac-related outcomes were itemized. Studies were considered eligible if they evaluated at least one of the disease categories or one of the disease outcomes within the category.

Glycemic:

• Progression to DM (if measured with eligible testing criteria).

• Reversion towards NGT (if measured by eligible testing criteria).

Cardiovascular disease:

• Angina requiring a minimum 24-hour hospitalization.

• Myocardial infarction (MI).

• Acute coronary syndrome.

• Cardiac revascularization.

• Peripheral revascularization.

• Cardiac mortality.

• Angiographic percutaneous coronary interventions (PCI).

• Coronary artery bypass grafting (CABG).

• Stent insertion.

• Angioplasty.

• Stroke events.

Mortality:

• All cause.

• Disease specific (cardiac mortality was included in fatal CVD outcomes).

Nephropathy:

• Proteinuria.

• Microalbuminuria.

• Dialysis.

• Renal transplant.

• Elevated creatinine.

• Elevated albumin-to-creatinine ratio in the urine.

Ocular:

• Cataracts.

• Blindness.

• Retinopathy requiring laser photocoagulation.

• Vitrectomy.

• A retinal photograph assessed by standard criteria showing at least a two-step change in retinal images.

Hypertension/blood pressure:

• Concurrent therapy for hypertension or measured BP values. It was noted that studies may give baseline values of numbers of subjects on BP medications, and the number of subjects ending with BP medications. This was an acceptable outcome measure.

Lipid level disturbance:

• If subjects reported baseline and endpoint mean lipid levels for at least one of the following: low density lipids (LDL), total cholesterol levels, high density lipids (HDL), or triglycerides.

Other:

• Amputation (foot, lower limb, or foot digits).

Study Selection

A team of study assistants was trained to apply the eligibility criteria in preparation for screening the title and abstract lists and the full-text papers. Standardized forms and a training manual explaining the criteria were developed and reviewed with the screeners (Appendix B).

For the title and abstract phase, two reviewers evaluated the citations for eligibility. Those articles that met the criteria were retrieved as well as those where there was insufficient information to determine eligibility. The article was retrieved if either one of the two screeners identified it for retrieval. For screening of full-text articles, two screeners came to consensus on the identification, selection, and abstraction of information. Disagreements that could not be resolved by consensus were resolved by one of our McMaster research team members. The level of agreement for inclusion of studies was measured using kappa statistics.

Data Extraction

All eligible studies from the selection phase (full-text screening) were abstracted onto a data form according to predetermined criteria. Appropriate data collection forms were developed for use in the systematic review (Appendix B). The articles were grouped according to the questions they addressed: diagnosis, prognosis, and treatment. One data extractor transferred the data onto data forms, and another data extractor checked the answers for accuracy before they were entered into a Microsoft Access database.¹⁷

Quality Assessment

One member of the research team rated each eligible study within the prognosis and treatment categories for methodological quality (see Appendix B). RCTs were evaluated using the modified Jadad scale.¹⁸ A scale developed by MacKay et al.¹⁹ for non-RCT studies was used to rate the prospective cohort studies. The MacKay checklist had three subscales that could yield a score out of 5 possible points for reporting, 12 possible points for internal validity, and 1 possible point for external validity. The summary scores for methodological quality (Tables 8 and 21) were used to determine the strength of the evidence and to select those studies with the best methodological scores for subsequent meta-analysis.

Summarizing Results: Descriptive and Analytic Approaches

Data from the Access database were summarized in evidence tables, which included data about the general study characteristics (study design, location of study, population characteristics, mean age, and diagnosis criteria for dysglycemia), interventions, and outcomes assessed.

Five classifications of dysglycemia. Studies were grouped according to classification of the IFG/IGT status. Five dysglycemia classifications were considered as risk factors and these included:

• 1)

  isolated IGT (I-IGT),

• 2)

  isolated IFG (I-IFG),

• 3)

  non-isolated IGT,

• 4)

  non-isolated IFG, and

• 5)

  combined IGT and IFG (IGT/IFG).

The threshold values for these five diagnostic groups are detailed in Table 1 as a function of the changing criteria for classification over time. A diagnosis of isolated IGT excludes those diagnosed with IGT who have a FPG between 6.1 and 7.0 mmol/L (110 and 126 mg/dL). A diagnosis of isolated IFG excludes those without a 2-hour OGTT result and those with an OGTT level greater than 7.8 mmol/L (140 mg/dL). For example, a classification of I-IGT using the WHO 98 criteria implies that the FPG was not within the specified range of 6.1 to 6.9 mmol/L, which is indicative of IFG. Thus, this implies that a FPG test was undertaken and deemed negative. However, the OGTT was within the specified range of 7.8 to 11.0 mmol/L and considered positive. A negative FPG and a positive OGTT are required for the classification of I-IGT. Table 1 shows that the classifications of IFG, I-IFG, I-IGT, and combined IFG/IGT are more recent classifications of dysglycemia, commencing with the WHO 1998/99 criteria. Some preliminary evidence suggests that these dysglycemic classifications may represent different subgroups with potentially different mechanisms leading to glucose disturbance.⁶

It should be noted that the criteria for diagnosis of dysglcemia in observational studies have been defined by the WHO²⁰ —specifically, as the epidemiological criteria²⁰ which enable researchers to classify subjects using just their blood glucose concentration, measured after an overnignt fast or 2 hours after a 75 g oral glucose load, without any confirmatory symptoms or blood/plasma determinations. Thus, the recommendation for such large population studies was a single glucose test at the start of the study.

Equations Used To Calculate Measures of Association

Diagnosis question. Kappa coefficients are used to estimate the average rate of concordance between two repeated tests (categorical data) and also take into account chance occurrence.^{21, 22} The equation and methods for calculating variance and 95% CI are shown in Appendix G, Section A.

Prognosis question. The equations for these measures of association for the prognosis question can be derived from a basic 2x2 table. Table 3 shows an example of a 2x2 table with the dichotomous classification (yes or no) for the outcome of interest, in this example DM, for those with the exposure (IFG or IGT) and those without the exposure status (NGT or NFG). From this table, the incidence for the duration of the study is derived. Eligible studies varied in duration from one to 18 years, thus it was difficult to compare measures of association between studies. For this reason, we converted estimates of risk, RR, and risk difference to annualized values.

Annualized risk for those with IFG or IGT and in normal subjects. The incidence was calculated as the rate of those individuals who developed the outcome of interest relative to those at risk, expressed as (a/n₁) in Table 3. Incidence rate is conceptually related to the risk (or probability) for developing an outcome over a specified time period.²³ The method used to convert an incidence rate to the risk for those patients with IFG or IGT developing the outcome of interest for a specified period of time can be seen in Appendix G, Section B. The advantage of these equations is that it does not assume that the rate of change is linear, as would be the case of simple division of the estimated rate by the number of years.²³

The studies evaluated in this systematic review varied in duration from six months to 18 years. In order to allow comparison across studies, the time period was standardized to a one-year period when reporting risk of the exposed for the outcome of interest. Appendix G, Section B allows for conversion of the varied time periods across studies to a common time period of one year. To facilitate presentation of the annualized risk, values are shown as per 100 persons.

The annualized risk for those with IFG or IGT who then progressed to the endpoint of interest, which, in the example of DM, was calculated as follows:²³

= Annualized risk in the exposed group (IFG or IGT) at time t and calculated as:

Relative risk for those with IFG or IGT relative to NFG or NGT. The RR is the ratio of the incidence in the exposed group (IFG or IGT) over the incidence of the unexposed group (NFG or NGT). Details of the derivation of RR are presented in Appendix G, Section B.

Calculation of the CI is presented in Appendix G, Section C.

Risk difference. The risk difference estimates the difference in risk that is attributable solely to the exposure of having IFG or IGT. This estimate is based on the annualized risk estimates and expressed per 100 persons.

Risk difference is calculated as:

Attributable risk for those with IFG or IGT relative to NFG or NGT for the study duration. The attributable risk (AR) represents the proportion of excess risk of the disease outcome (above the background risk) in the exposed group. It is calculated using the incidence in the exposed group (IFG or IGT) and then subtracting the incidence in the non-exposed group (NFG or NGT). This numerator is then divided by the incidence in the exposed group (IFG or IGT) and multiplied by 100 in order to be expressed as a percent. Note that this estimate of the AR was not based on annualized estimates and was therefore not converted to an annualized proportion. The percent AR was expressed for the duration of the study. The AR was estimated using the following equation:

Treatment question. The metrics of association (AR, RR, and AR) presented for studies evaluated in the treatment question are based on annualized risk estimates (from extracted data where permitted). However, many of the studies also presented these same estimates for the study duration but may not have reported sufficient data to permit annualized estimates. Thus, where possible both annualized and study duration estimates were presented.

ARD, NNT, and RRR. The ARD is a metric frequently used in clinical epidemiology that expresses the absolute risk difference between the event rate in the treatment group and that of the control/placebo group. The ARD compares the outcome rates on an arithmetic scale and is expressed in absolute terms.

ARD = absolute value[(R _Et) - (R _Ct)]

An alternative way of expressing the difference between groups is with the number of patients needed to treat (NNT). The NNT expresses the number of patients that a clinician must treat (with the intervention used in the study in question) in order to prevent one patient from having a target outcome Clinicians may find this estimate of the risk difference to be a useful expression of the magnitude of the treatment effect.²⁴ The NNT is calculated as the inverse of the ARD caused by treatment and is detailed, as follows:

NNT = 1/ ARD

Lastly, the RRR is an additional metric used in clinical epidemiology to express the risk that is taken away by the intervention used in the study. It assists in comparing studies with different baseline risks as it considers the ARD and then divides this by the risk for the placebo/control group. The equation used to estimate the RRR is as follows:

RRR = ARD / R _Ct

Meta-Analysis

Quantitative meta-analyses were undertaken within each of the dysglycemia classification groups with a minimum of two studies for the unadjusted annualized RR. Some of the prospective cohort studies were related and not independent. One cohort could have multiple publications that reflected analyses done at different time intervals or on different groups within the same population cohort. Therefore, one representative publication was selected from the series of related studies to be included within the meta-analysis. The representative study was selected by consideration of the methodological quality score, the larger sample size, and the year of publication.

An overall pooled estimate was calculated across all study populations. Tests for heterogeneity were undertaken and, when statistically significant, only the results from the random-effects model (REM) were used to calculate the pooled estimate. Statistical software (SAS, version 8.2)²⁵ was used to calculate the test for heterogeneity and the pooled estimates. Studies were weighted according to the inverse of their variances. Individual study effect sizes were calculated and plotted by year of publication.

Tests of heterogeneity. Tests of heterogeneity are statistical analyses for examining whether the observed variation in study results is compatible with the variation expected by chance alone. The test for heterogeneity selected for this review (Q) is detailed in Appendix G, Section D. The smaller the p value of the Q test (that is the more significant the test), the greater the likelihood that the observed differences between the studies was not due to chance alone. If the value of the Q test is relatively low (for example, one in 10 or one in 20) then the observed differences in the results betweens studies is likely related to factors other than chance.²⁶ The potential factors that account for these differences can be numerous, and caution should be used when attempting to explore the nature of these differences. A single factor may not be the only important source of heterogeneity.

Tests of heterogeneity have some limitations that can make interpretation difficult.²⁷ It has been suggested that the statistical power of the Q test in most cases is low (due to a small number of combined studies). As such, the test may indicate that it is not statistically significant at conventional levels, but, in reality, heterogeneity is present. Similarly, if the sample sizes of the studies are very large, the Q test may be significant even when individual effect sizes do not really differ. Furthermore, design flaws and publication biases can also make the interpretation of heterogeneity tests difficult. For example, if all the studies meta-analyzed have the same design flaw, a consistent bias is present that could make the effect sizes appear more reliable than they really are. Conversely, if the studies have different design flaws, the meta-analysis could show a positive test for heterogeneity, but, in reality, reflect the same underlying population.²⁸

In summary, caution must be used when interpreting the Q statistic. Some have argued that this test should be omitted while others have suggested that it should only be used as a diagnostic tool until further research accounts for variation between studies. One method recommended for dealing with sources of heterogeneity is the REM for meta-analysis.²⁷ Rather than attempting to explain or adjust for the variability between studies, the REM takes into account the variation in the underlying effect sizes. The use of the REM is often used when the source of the variance cannot be identified.²⁷ As such, the REM cannot investigate the causes of the heterogeneity.

When meta-analyses in this systematic review revealed a significant test for heterogeneity (Q), the REM was used to calculate the overall pooled estimate. Exploratory sensitivity analyses were undertaken for those meta-analyses that had five or more studies. In these analyses, each study was removed from the pooled estimate, and the Q test and the overall pooled estimate were reviewed. These data were used to judge whether any individual studies should be removed from the combined estimate.

General Characteristics of the Diagnosis Studies

Fifty-six reports describing 53 unique studies provided data on the reproducibility of repeat testing of FPG or OGTT, comparison of IGT diagnosis by different criteria, and the relationship between IGT and IFG diagnosis in the same population. General characteristics of these studies can be found in Table 4.

Reproducibility of IGT and IFG Tests

Four studies in five reports²⁹–³³ assessed the reproducibility of the OGTT for diagnosis of IGT, and two studies^{30, 31, 33} assessed the reproducibility of FPG for the diagnosis of IFG in publications after 1978 (Table 5). All repeat tests were done within six weeks of the first test. No triplicate testing was done. The populations studied were mostly Caucasians,²⁹–³¹ except for one study on Hong Kong Chinese in two reports.^{32, 33} The study populations were subgroups of larger studies and as such did not provide detailed characterization for the subgroup. The selection of the subgroups was essentially random, but not in all cases. Mooy and de Vegt used different subgroups of individuals from the Hoorn study. Mooy's group²⁹ included participants randomly selected from those individuals with a 2 hour post glucose challenge plasma glucose level (2-hr PG) of < 7.5 mmol/L, stratified by age and sex, plus all participants with a 2-hr PG < 7.5 mmol/L (22% IGT by the first OGTT). de Vegt,³⁰ however, used a reconstructed sample that represented participants without known DM (10.5% IGT by the first OGTT). Ko^{32, 33} reported reproducibility for IGT and IFG from the same random sample of the working Hong Kong Chinese study in two separate publications. Farrer³¹ reported reproducibility of IGT in a high-risk group of mostly male post-CABG surgery patients. No details were given as to how the subgroup sample was selected. All studies used the same classification criteria: FPG ≥ 7.8 mmol/L and 2-hr PG 7.8 to 11.1 mmol/L for IGT, and FPG 6.1 to 6.9 mmol/L for IFG.

The reproducibility of the IGT and IFG tests was assessed by calculating the kappa coefficient, and the percent positive agreement between participants categorized as IFG or IGT on both tests. The kappa coefficients for IGT ranged from 0.04 to 0.56, indicating poor to moderate agreement.³⁴ The proportion of participants classified as IGT by the first OGTT who remained classified as IGT upon repeat testing ranged from 33% to 48%. Most participants were reclassified as NGT (39.3% to 46.2%) with the remainder reclassified as DM (6% to 12.6%). The two Hoorn substudies gave essentially the same reproducibility, even though Mooy used a higher risk group to examine the reproducibility.²⁹ The study that gave the lowest reproducibility (kappa = 0.04) involved mostly men who had undergone CABG and were considered at high risk for IGT and DM.³¹ This study population had a significantly higher prevalence of abnormal glucose tolerance compared to the expected proportion of a population with the same 2-hr PG mean and SD.

Two studies retested participants based on FPG for IFG. The kappa coefficients for these studies were 0.22 and 0.44, indicating fair to moderate reproducibility. The proportion of participants classified as IFG by the first FPG, who were classified as IFG again upon repeat testing, was 63.7% and 51.4% for the Ko and de Vegt studies, respectively.^{30, 33} The reclassified subjects had mostly NFG with some newly diagnosed DM. The low IFG prevalence in the Ko study population of 2.5% compared to 16.5% in the de Vegt study population may have accounted for the difference in reproducibility.

Another measure of reproducibility reported in two studies was within individual coefficient of variation (CV_I) for IFP and 2-hr PG^{29, 31}. The coefficient of variation is the ratio of the standard deviation to the mean and is used as a measure of the relative spread or precision. The CV_I is a measure of random variation around a homeostatic set point for each individual. This variation is distinct from analytical variation (CV_A) of a test method. Both studies gave similar CV_I for FPG (6% and 6.3%) and the 2-hr PG (18% and 16.6%) concentrations,^{29, 31} indicating consistency in variation between the different populations studied. Mooy's study also reported no association between test-retest differences with age, sex, obesity (body mass index [BMI] or waist-hip ratio [WHR]), or BP. However, there was a positive association with heart rate on the difference between the 2-hr PG tests (mean difference 2, 95% CI 1.1 to 2.9 beats/min, p = 0.01).²⁹ Furthermore, when the FPG and the 2-hr PG tests were categorized into NGT, IGT, and DM, the FPG CV_I was greater for participants classified as DM (7.0%) compared to IGT (5.9%) or NGT (4.6%). The 2-hr PG CV_I gave opposite results with a slightly better CV_I for participants classified as DM (12.6%) compared to IGT (14.9%) or NGT (16.3%). No p values were provided, so it is not known if these differences were statistically significant.

Comparison of IGT Diagnosis Using Different Criteria

Four studies³⁵–³⁸ compared diagnoses using different IGT criteria (i.e., both IFG and the 2-hr PG concentrations).³⁵–³⁸ Studies that assessed IGT based on the epidemiological criteria (i.e., the 2-hr PG glucose concentration only) were excluded.²⁰ Study population characteristics varied and thus represent a broad spectrum of populations (Asian, Dutch, Pima Indians, and women with previous gestational DM). The IGT criteria included were WHO 85, WHO 98, and WHO 99. All of these criteria use a 2-hr PG range of 7.8 to 11.0 mmol/L, but the WHO 85 uses an FPG cut point of < 7.8 mmol/L whereas both the WHO 98 and WHO 99 criteria use a cut point of < 7.0 mmol/L.

Table 6 summarizes the data from the four studies. More IGT diagnoses were made using an FPG cut point of < 7.8 mmol/L (13.6% to 31.5 %) than of 7.0 mmol/L (8.3% to 29.7%). There were fewer cases of I-IGT (6.0% to 11.9%) compared to IGT by either FPG cut point.

Of the three studies using FPGs of < 7.8 mmol/L and < 7.0 mmol/L to classify IGT, two large population-based studies showed a similar negative change of 3.1% and 3.4%. One study showed a negative change of 5.8%. The latter study had fewer participants than the other two studies and consisted of a cohort of women with a history of gestational DM. The Pima Indian study group provided cumulative data from 1965 to 1999 for individuals aged ≥ 15 years. The DECODA study included Asian participants from five Asian countries (India, China, Indonesia, Japan, and Singapore) and the U.S. (Los Angeles and Hawaii).

The reduction in IGT classification using an FPG of < 6.1 mmol/L compared to < 7.0 mmol/L was greater in the Hoorn follow-up study (27.9%) compared to the Pima Indian study (19.0 %) and the DECODA study (19.3%). The Hoorn follow-up study excluded participants from the prospective-cohort baseline group who had died, moved away, had missing glucose values, or had DM (1142 of 2484 participants). In contrast, the Pima Indian and the DECODA groups were population-based and included all individuals.

Relationship Between IGT and IFG

Forty-nine studies provided data on the relationship between the diagnostic criteria for IGT and IFG. Most studies were prospective cohort studies (n = 14) and cross-sectional studies (n = 31). The majority of the studies used the WHO 98, WHO 99, ADA 97, or ADA 98 criteria for IGT. These diagnostic criteria are identical to each other (Table 1). The other major set of criteria used for IGT was the WHO 85 criteria that differed only in the FPG value (< 7.8 mmol/L compared to < 7.0 mmol/L). Many of the studies that reported IGT results with the WHO 85 criteria also reported results with the later criteria (WHO 98, WHO 99).²⁰ One study used the WHO 80 criteria for IGT diagnosis.³⁹ The diagnostic criteria for IFG are the same for the ADA and WHO in any year. However, the data were reported differently across studies. For example, studies may have used the term IFG or IGT but actually reported data for the I-IFG or the I-IGT, respectively. Also, studies reporting IGT classifications may have used only the 2-hr PG value rather than FPG and 2-hr PG values. Therefore, to relate studies according to the diagnostic criteria in Table 1, additional calculations were needed in some studies.

The data were extracted to give seven classifications, regardless of which criteria were described in each article. If the data were available to fit into these classifications, they were extracted to provide maximal information for comparison among studies. The seven classifications include IGT (2-hr PG 7.8 to 11.0 mmol/L), IGT (FPG < 7.8 mmol/L and 2-hr PG 7.8 to 11.0 mmol/L), IGT (FPG < 7.0 mmol/L and 2-hr PG 7.8 to 11.0 mmol/L), IGT (FPG 6.1 to 6.9 mmol/L), I-IGT (FPG < 6.1 mmol/L and 2-hr PG 7.8 to 11.0 mmol/L), I-IFG (FPG 6.1 to 6.9 mmol/L, 2-hr PG < 7.8 mmol/L), and the combined group of IFG and IGT (FPG 6.1 to 6.9 mmol/L and 2-hr PG 7.8 to 11.0).

The difference between IFG and I-IFG classifications among the studies shown in Figure 3 is approximately 2.8 fold (p < 0.0001). Comparison of studies that have both IGT and I-IGT data show that classification is approximately 40% greater for IGT compared to I-IGT (p < 0.0001). Also, an IGT classification using the WHO epidemiological criteria, which omits the FPG value, classified 10% more participants as IGT (p = 0.0033).

Table 7 shows a summary of the percent of study subjects with the diagnoses of IGT, I-IGT, IFG, I-IFG, both IGT/IFG, and ratios of IGT to IFG and I-IGT to I-IFG. Studies containing data for all classification categories (n = 16) are expressed graphically in Figure 3. A line connects each classification group within each to show the change in proportion between each classification group and between studies. In general, the proportion of participants decreased with increased stringency of the diagnostic criteria—that is, IGT (2-hr PG) > IGT (FPG and 2-hr PG) > I-IGT > IFG > I-IFG > IFG and IGT. Four studies had a higher proportion of IFG compared to I-IGT,⁴⁰–⁴³ and three studies had a similar proportion of IFG and I-IGT.⁴⁴–⁴⁶ The three studies that had a higher proportion of IFG were hospital-based and selected patients who were at higher risk for DM. The other study⁴³ was conducted on a distinct population of 7018 male police officers in Paris.⁴³ Only one other study⁴⁷ illustrated in Figure 3 was hospital-based, but this study excluded all patients with disorders that would likely affect glucose metabolism.⁴⁷ Other studies described high-risk populations but from non-hospital settings. The three studies with a similar proportion of IFG and I-IGT were from populations in Taiwan,⁴⁴ Ghana,⁴⁵ and Australia.⁴⁶ None of the other 16 studies included participants from these countries.

The prevalence of IGT and IFG varied greatly among the studies, ranging from a few percent to over 30% (Table 7). Comparisons between categories of IGT and IFG among the studies in Figure 3 were significant (paired t-test, p < 0.01) for all combinations except for I-IGT versus IFG and I-IFG versus IGT and IFG. Correlations between IGT and I-IGT and IFG and I-IFG classifications are expected to be high because of the similarity in test criteria. The correlation was much higher in these 16 studies for IGT versus I-IGT than for IFG versus I-IFG. The Passing-Bablok regression equations were IGT versus I-IGT, y = 0.879× – 1.29, r = 0.95, p = 0.0084; IFG versus I-IFG, y = 0.521× – 0.78, r = 0.63, p ≤ 0.0001 (see Figures 4 and 5). Two studies were identified as outliers in the comparison of IFG versus I-IFG: the Paris police study⁴³ and the Bangkok study.⁴⁰ These studies were unique because the policemen study included men only and the Bangkok study included patients who had a history of borderline fasting glucose values.