The conversion rate of individuals from prediabetes to diabetes changes with population characteristics and the criteria used to define prediabetes[14,15]. In a meta-analysis evaluating the progression of prediabetes to diabetes published in 2007, the annual incidence rate of diabetes was found to be 4%-6% for isolated IGT, for isolated IFG 6%-9% and for both IGT and IFG was 15%-19%[16]. This meta-analysis only consisted of studies published prior to 2004. In subsequently reported major studies, the annual incidence rates of conversation from prediabetes to diabetes were similar. In the Diabetes Prevention Program (DPP) Outcomes Study, the incidence of diabetes was noted to be 11% in the control group[17]. In the United States Multi-Ethnic Study of Atherosclerosis the annual incidence of diabetes in IFG group slightly above 4%[18]. In the The Toranomon Hospital Health Management Center Study the incidence of diabetes was reported as 7% in the group with an HbA1c 5.7%-6.4% and 9% in the IFG group[19]. In the China Da Qing Diabetes Prevention Study (CDQDPS), the cumulative incidence of diabetes over a 20 years period, was noted to be higher than 90% among subjects with IGT defined by repeated OGTT in the control group[20]. The use of ADA vs WHO criteria to define prediabetes has also been shown to affect the incidence rate of diabetes with lower incidence in individuals defined by ADA criteria compared to WHO criteria[21].

According to an expert panel, continuous rather than dichotomous risk scores are more useful for predicting the risk of developing diabetes[22]. A diabetes risk score based more easily accessible variable such as age, sex, ethnicity, fasting glucose, systolic blood pressure, HDL cholesterol, BMI and history of diabetes in parents or siblings has been shown to have better predictive value than either IFG or IGT[23].

Several studies have shown an association of increased risk of chronic kidney disease and early nephropathy with prediabetes[24-28]. The causal nature of this relationship remains unclear as this association may be due increased incidence of diabetes in this group or the presence of other factors associated with both hyperglycemia and nephropathy rather than the effect of prediabetes itself[29,30].

Prediabetes is found to be associated with dysfunction of cardiac autonomic activity, reflected by reduced heart rate variability[31-35], decreased parasympathetic modulation of the heart[35] and increased prevalence of male erectile dysfunction in individuals with prediabetes[36]. Non-invasive evaluation of neural impairment in subjects with IGT has shown significantly greater abnormalities detected by four of five cardiovascular reflex tests, increase prevalence of both hyperesthesia and hypoesthesia, and increased heat detection thresholds[37]. There is also increasing evidence to demonstrate a higher frequency of idiopathic polyneuropathy, painful sensory neuropathy[38-43] and small fiber neuropathy[38,40,41] among prediabetic individuals with IGT. These findings suggest an involvement of the small unmyelinated nerve fibers that carry pain, temperature, and regulate autonomic function during prediabetes, prior to development of diabetes.

Nearly 8 percent of participants with prediabetes in the DPP study were found to have evidence of diabetic retinopathy[44]. While prediabetes has been associated with an increased risk of diabetic retinopathy in some studies, these findings vary depending on the method used for detection[24,45-49].

Prediabetes has been associated with increased risk of developing macrovascular disease but whether this elevated risk is due to prediabetes itself or due to development of diabetes remains unclear[50,51]. While cross sectional studies have shown an increased prevalence of coronary heart disease in individuals with prediabetes[52,53] but this relationship may be confounded by the common risk factors present between cardiovascular diseases and prediabetes.

The encompassing theme of lifestyle intervention programs is to change the modifiable risk factors of prediabetes and diabetes by targeting obesity with increase in physical activity and dietary changes. The two largest diabetes prevention studies, the United States DPP and the Finnish Diabetes Prevention Study (DPS) have both shown beneficial effects of lifestyle interventions[54,55]. In the DPP study, after a 3 year follow-up, intensive lifestyle interventions (ILS) lead to a 58% risk reduction. The ILS involved changes in diet and physical activity aimed at producing weight. The biggest determinant of risk reduction was note to be weight loss. This study showed that for every 1 kg decrease in weight, the risk of developing diabetes in future was reduced by 16%[56]. In the DPS, the benefits were found to be dependent on achievement of the number of pre-defined goals of the intervention by the participant. These goals consisted of weight reduction greater than 5 percent, total fat intake less than 30 percent of energy intake, saturated-fat intake less than 10 percent of energy intake, fiber intake greater than or equal to 15 g per 1000 kcal , and exercise greater than 4 h/wk[55]. While both of these studies were largely among Caucasians, studies in Asian population have also shown similar benefits[57,58].

Several groups of antidiabetic drugs such as Biguanides, Thiazolidinediones, α-Glucosidase Inhibitors, GLP-1 analogies and non-antidiabetic drugs and therapies such as anti-obesity drugs, and bariatric surgery have been studied in context of prediabetes.

Metformin has been used for several decades for treatment of diabetes and has been noted to have additional favorable outcomes such as body mass index (BMI) reduction and improved cholesterol profile. The collective evidence of trials among subjects with IGT, suggests a 45% risk reduction for development of type 2 diabetes[59]. Metformin was noted to be less effective than lifestyle in the United States DPP trial but in the indian DPP (IDPP) trial it was noted to be as effective as lifestyle intervention[54,57]. Metformin has been found to be more beneficial to individuals with higher BMI and higher FPG[54]. Metformin has also been studied in obese children by several investigators. The collective evidence shows a slight benefit in BMI reduction over lifestyle interventions, while the benefit was statistically significant it was noted to be only short term with the greatest benefit at 6 mo and no difference at 12 mo duration[60].

The glitazones are synthetic ligands for peroxisome proliferator-activated receptors-γ. They increase glucose uptake and utilization in the peripheral organs and decrease gluconeogenesis in the liver, thereby reducing insulin resistance[61]. In the double blind placebo controlled Diabetes Reduction Assessment with Ramipril and rosiglitazone Medication study, rosiglitazone was found to be effective in decreasing incidence risk of diabetes by 60% over a 3 year period but was associated with significant side effects such as an additional average weight of 2.2 kg in intervention group compared to controls and a higher incidence of heart failure (0.5% vs 0.1%) and total cardiovascular events (2.9% vs 2.1%)[62,63]. Pioglitazone was found to decrease the risk of diabetes by > 70% in obese subjects with IGT in the ACT NOW study. Some of the added benefits were, decrease in diastolic blood pressure, reduction in rate of carotid intima-media thickness and a greater increase in HDL cholesterol but it was associated with increased weight gain (approximately 3 kg more than placebo) and edema (13% vs 6% in controls)[64]. In the double blinded placebo controlled 3 years prospective IDPP-2 study, there was no difference in incidence of diabetes between subjects receiving lifestyle intervention and placebo and subjects receiving lifestyle intervention and pioglitazone[65]. In the more recent Canadian Normoglycemia Outcomes Evaluation trial, low does combination of rosiglitazone and metformin was tested against placebo to investigate whether low dose combination therapy would decrease the incidence of type 2 diabetes with a lower risk of adverse events. Incident diabetes occurred in significantly fewer individuals in the active treatment group (14%) than in the placebo group (39%). The relative risk reduction was 66% and the absolute risk reduction was 26%, and 80% subjects in the treatment group reverted to normoglycemia compared to 53% in the control group, but the subjects in active treatment group had increased reports of diarrhea (16% vs 6% in controls)[66]. Overall, there are safety concerns for thiazolidinedione such as weight gain, liver toxicity, increased cardiovascular risk and possible link with bladder cancer which have limited the use of these mediations for treatment of prediabetes.

α-glucosidase inhibitors such as acarbose and voglibose, prolong the overall carbohydrate digestion time, and reduce the rate of glucose absorption, thus decrease the postprandial rise in blood glucose[67]. In the STOP-NIDDM trial, acarbose was found to decrease the relative risk for diabetes by 25% among subjects with IGT during a 3.3 years of follow-up[68,69]. The medication was associated with several gastrointestinal side effects such as flatulence and diarrhea and 31% of the participants in the acarbose arm dropped out before completion of the study[68]. A Japanese trail found a 40% risk reduction in incidence of diabetes in high-risk individuals with IGT with voglibose over a 48 wk period. Voglibose was noted to have a similar side effect profile as acarbose but only 7% subjects discontinued the use of drug due to adverse effects[70].

GLP-1 analogs exploit the physiological effects of GLP-1, they have been shown to augment post prandial insulin secretion, suppress glucagon and hepatic glucose production, slow gastric emptying, and reduce appetite[71]. Exenatide and liraglutide have been demonstrated to have long term efficacy for sustained weight loss in obese subjects and reduce prevalence of prediabetes over a follow-up period of 1-2 years. The most common side effects with these drugs are nausea and vomiting and they remain injectable preparations[72-74].

Anti-obesity drugs Orlistat has also been studied in context of prediabetes. Orlistat is a gastrointestinal lipase inhibitor used for treatment of obesity that acts by inhibiting the absorption of dietary fats by approximately 30%. Research has shown that over a 1.5 year follow-up period, use of Orlistat in conjunction with low energy diet is associated with greater weight loss as compared to placebo (6.7 kg vs 3.8 kg) and a decrease in conversion rate from IGT to overt diabetes (7.6% vs 3.0%) in obese adults[75]. Similar finding have also been reported by the XENDOS trail regarding the efficacy of Orlistat with a 37% relative risk reduction in development of diabetes after 4 years of treatment[76].

Bariatric surgery includes a variety of procedures aimed at either creating a mal-absorptive state, a restrictive state or a combination of the two to limit caloric intake. The procedures commonly used include Roux-en-Y gastric bypass, Laparoscopic adjustable gastric banding, Sleeve gastrectomy, and Duodenal switch with biliopancreatic diversion. In the Swedish Obese Subject, bariatric surgery was found to result in sustained weight loss (23.4% at 2 years and 16.1% at 10 years) and a 75% relative risk reduction of diabetes compared to controls[77]. Bariatric surgery was also associated with a lower 2 year and 10 year rate of development of type 2 diabetes, cardiovascular disease and reduced number of cardiovascular deaths in obese adults[77,78]. A previous study had demonstrated that after gastric bypass surgery, 78% subjects with previous diabetes and 98% subjects with IGT reverted to normoglycemia[79].