state for 6–9 hours per day [42]. Therefore, it is logical that the knowledge gained from a standardized glucose load is clinically informative. The 2020 ADA Guidelines state that FPG, 2‐h PG during a 75‐g OGTT, and HbA1c are equally appropriate to test for prediabetes and DM [1]. Many clinicians screen with both a FPG and HbA1c. However, the omission of an OGTT leads to a significant under‐diagnosis of both DM and prediabetes.
Risk of progression from prediabetes to diabetes mellitus
The transition from prediabetes to DM2 is variable and influenced by heredity and lifestyle [18, 19, 43]. Multiple variables have been used to estimate the risk of progression to DM2, including glycemic indices, anthropometric data, comorbidities, metabolomics, and genetic variants. Numerous prediction models have been created based on these variables [43–51].
Baseline FPG is a significant predictor of an individual's risk for developing DM [43]. In nondiabetic adults residing in Minnesota, baseline FPG levels < 100 mg/dL, 100–109 mg/dL, and 110–125 mg/dL were associated with a 7%, 19%, and 39% risk, respectively, with progression to DM over a median of 9 years. A discrete gradient of risk for progression to DM was also observed among subjects with a baseline FPG < 100 mg/dL.
During a 75‐g OGTT, 30‐minute, 60‐minute, and 120‐minute PG concentrations were all significant predictors for future risk of DM [49, 52]. One‐hour plasma glucose during a 75‐g OGTT has been shown to be a better predictor of future DM than FPG, HbA1c, and PG at 30 minute and 120 minutes during a 75‐g OGTT [50]. In one study, the hazard ratio for the development of DM was 9.5 (7.90–11.43) for those with combined IFG‐IGT at baseline, 4.5 (4.03–5.02) for those with isolated IGT at baseline, and 3.98 (3.16–5.02) for those with isolated IFG at baseline [45].
Baseline HbA1c is also a strong predictor for risk of future DM, and the risk of incident DM significantly increased across the HbA1c range of 5.0–6.5% [46]. A systematic review found that the 5‐year incidence of DM was 25–50%, 9–25%, and 5–9% among those with a baseline HbA1c of 6.0–6.5%, 5.5–6.0%, and 5.0–5.5%, respectively.
Numerous studies have shown that demographic data, anthropometric data, and comorbidities are associated with progression to DM. Factors positively associated with progression to DM include: family history of diabetes, former or active smoking, higher BMI, abdominal obesity, increased waist circumference, hypertension, elevated triglycerides, low HDL cholesterol, and elevated high sensitivity C‐reactive protein [45, 53, 54]. Increasing age has also been shown to be a predictor of DM in some studies.
Metabolomics profiles have been investigated as a tool to estimate the risk of developing DM. Numerous metabolites have been found to be positively and negatively associated with progression to DM [51]. More than 400 distinct genetic signals that affect the risk of developing DM2 have been identified [20, 44]. Polygenic scores have been used to estimate the combined genetic risk for the development of DM.
Microvascular complications associated with prediabetes
Retinopathy
In a large cohort of Pima individuals, the frequency of retinopathy – defined as the presence of at least one hemorrhage, one microaneurysm, or proliferative retinopathy – was directly related to baseline FPG and 2‐h PG [55]. Beginning at a baseline FPG threshold of 6.0–6.4 mmol/L and 2‐h PG threshold of 9.0–10.6 mmol/L, there was a significant increase in period prevalence (10‐year interval) of retinopathy. Additional data from the Pima individuals illustrated that beyond a HbA1c threshold of 6.2% there was a significant increase in the prevalence of retinopathy [12]. Similar glycemic thresholds for an increase in the prevalence of retinopathy were observed in a cross‐sectional study of Egyptians and data from the Third National Health and Nutrition Examination Survey [13, 14]. Additionally, in the Diabetes Prevention Program (DPP), diabetic retinopathy was detected in 7.9% of the subjects with IGT and in 12.6% of the subjects who developed DM [56]. The prevalence of retinopathy is significantly higher in subjects with DM, but retinopathy can develop in subjects with prediabetic range dysglycemia.
Pooled data analysis of nine studies from five countries examined glycemic thresholds for diabetes‐specific retinopathy (defined as moderate or more severe retinopathy) in 44 623 subjects [57]. A curvilinear relationship was found to exist for FPG and HbA1c when diabetic retinopathy was plotted against continuous glycemic measures. Diabetes‐specific retinopathy began to increase from a FPG of 6.0–6.4 mmol/L and from a HbA1c of 6.0–6.4%. Based on vigintile distributions, glycemic thresholds for diabetes‐specific retinopathy were observed over the range of 6.4–6.8 mmol/L for FPG, 9.8–10.6 mmol/L for 2‐h PG, and 6.3–6.7% for HbA1C. Compared with the first vigintile, the odds ratios for diabetes‐specific retinopathy for the above vigintile distributions were 2.5 (95% CI 1.2–5.2) for FPG, 10.1 (95% CI 1.3–79.4) for 2 h PG, and 4.5 (95% CI 1.4–15.2) for HbA1c.
A major limitation of many studies that examine the prevalence of retinopathy is the diagnostic criteria used to define diabetes‐specific retinopathy [57]. The use of an overly broad definition for retinopathy leads to the inclusion of subjects with mild retinopathy, which may have etiologies other than hyperglycemia. Overall, the rate of retinopathy increases with the degree and duration of hyperglycemia [56, 57]. Subjects with prediabetes have a higher prevalence of retinopathy than subjects with NFG/NGT, although the prevalence remains relatively low. In many populations, the glycemic threshold for retinopathy occurs in the prediabetic range of dysglycemia [57, 58].
Nephropathy
Analysis of NHANES data from 1999–2006 revealed that chronic kidney disease (CKD), defined as either reduced kidney function or elevated albuminuria (urinary albumin‐creatinine ratio ≥ 30 mg/g), was present in 17.1% of prediabetics compared to 11.8% of individuals with normoglycemia [59]. A multiethnic study found that 16.1% of subjects with IGT had microalbuminuria, compared to 4% of subjects with NGT [60]. After adjusting for multiple variables, glycemic status was found to be the most significant determinant of urinary albumin concentration. A systematic review and meta‐analysis including 9 studies with 185 452 subjects reported that prediabetes was modestly associated with an increased risk of CKD [61]. After adjusting for established risk factors, the relative risk of CKD was 1.11 (95% CI 1.02–1.21) for subjects with FPG between 6.1–6.9 mmol/L.
The prevalence and five‐year incidence of nephropathy increases as FPG, 2‐h PG, and HbA1c rise [12, 58]. Of note, the association of glycemia with nephropathy is weaker than the association between glycemia and retinopathy. When plotting prevalence of microalbuminuria against FPG, 2‐h PG, and HbA1c, there is a visible inflection point and subsequent increase in microalbuminuria prevalence beyond a FPG of 5.5 mmol/L, 2‐h PG of 5.5 mmol/L (and again at 9.3 mmol/L), and HbA1c of 5.8%. In summary, multiple studies suggest that prediabetic‐range hyperglycemia is associated with higher rates of nephropathy.
Neuropathy
Although neural dysfunction is associated with hyperglycemia, clinicians should be mindful that neurologic deficits can be attributable to non‐glycemic causes in individuals with prediabetes and DM [62]. Subjects with isolated IGT were shown to have subclinical neural dysfunction that was generally asymptomatic and characterized by small‐fiber neuropathy and mild impairment of cardiovascular autonomic function [63]. Erectile dysfunction has also been shown to be independently associated with IFG, with an odds ratio of 1.26 (95% confidence interval 1.08–1.46) [64]. Prediabetic‐range hyperglycemia is also associated with chronic idiopathic axonal polyneuropathy (CIAP) [65]. In a study of 100 subjects with CIAP, 36 individuals were found to have IFG, 3 had FPG ≥ 126 mg/dL, 38 had IGT, and 24 had 2‐h PG ≥ 200 mg/dL. Overall, the prevalence of dysglycemia in this cohort was approximately 2‐fold higher than in an age‐matched general population group.
