The proportion with previously undiagnosed diabetes varies with age. It seems to be highest at 30–39 years of age (70–80%), and lowest in the elderly (around 40%) [10,12]. The only exception was seen in European women where the proportion of undiagnosed diabetic cases was around 40–45% in all age groups. The proportion of undiagnosed diabetes was higher in European men than in European women. In Asians, it was slightly higher in women than in men in the youngest age groups.
Prevalence of impaired glucose tolerance and impaired fasting glycemia in different ethnic groups
A category of nondiabetic fasting hyperglycemia was defined only recently, by the ADA in 1997  and adopted also by WHO in 1999 , and named impaired fasting glycemia (IFG). It was introduced by consensus to define impaired glucose homeostasis intermediate between diabetes and normal glucose homeostasis and to be analogous to impaired glucose tolerance (IGT), but without any epidemiologic evidence of possible risks associated with it. Since the introduction of the category of IFG, prospective studies have examined the relationship between IFG and future morbidity and mortality with a comparison to IGT, and shown that the risk of cardiovascular disease (CVD) morbidity and mortality is higher for IGT than for IFG [44–47].Thus far the data are scarce on the risk of progression to diabetes in subjects with IFG as compared with those with normal fasting glucose or those with IGT. A few studies, which have examined the issue, agree that the risk of developing diabetes is high in subjects with either IFG or IGT and highest in those with both IFG and IGT, as compared with subjects with normal fasting and normal 2-h glucose [48–52]. At present neither IFG nor IGT is considered a clinical entity, but as a risk category for the future development of diabetes . Each represents a metabolic state intermediate between normal glucose homeostasis and diabetic hyperglycemia, and they were combined and defined formally as impaired glucose regulation (IGR) by WHO Consultation in 1999 . All studies agree that only IGT but not IFG is a risk factor for CVD.
The prevalences of IGT and IFG in Europe and Asia were reported recently by the DECODE and the DECODA Study Groups [8,12].
The prevalence of IGR rose with age in each study  (Figure 3.6). In most of the study populations, the prevalence of IGR was less than 15% at 30–59 years of age and between 15 and 30% after 60 years of age. The prevalence of IGT increased linearly with age, but the prevalence of IFG did not (Figure 3.6). The increase in the prevalence of undiagnosed diabetes and IGR in the elderly population resulted mainly from the proportionately larger increase in postload hyperglycemia than in fasting hyperglycemia.
The prevalence of IGR rose with age up to the 70s and 80s in most of the study cohorts  (Figure 3.7a,b). The increase was graded with aging in Chinese, Japanese, and Singaporean populations, as observed in Europeans, but not in Indians where the prevalence of IGR started to increase by the age of 30–39 years and did not change much with increasing age. The peak prevalences of IGR were not different among different populations, but the age- specific prevalence of IGR was higher in Indians than in Chinese and Japanese at 30–49 years of age for both men and women. In the urban populations the prevalence was higher than in the rural populations aged 40–69 years in men and 50–59 years in women in the Chinese and Japanese populations (Figure 3.7). The difference in the prevalence pattern in different ethnic groups may not be completely explained by living environments and geographic locations, suggesting that genetic differences also play a role.
IGT was more prevalent than IFG in almost all age groups in Asian subjects (Figure 3.7). The prevalence of IGT increases with age whereas IFG does not. This pattern is consistent with that among the European populations . The concordance for IFG and IGT was very poor in all populations, particularly in Asians [10,12,54–56]. The finding that postload hyperglycemia was more prevalent in the elderly in Europe and Asia is consistent with the report from NHANES III . Thus, the prevalence of undiagnosed diabetes and IGR would be underestimated to a large extent, especially in female and elderly populations, if only fasting glucose determination were used. The primary purpose of population-based testing for blood glucose is to detect previously undiagnosed diabetes and IGR in order to apply early intervention to reduce the serious diabetic complications and to prevent progression from IGT to diabetes as demonstrated by the recent diabetes prevention trials [58–62].
Sex differences in prevalence of diabetes, IGT, and IFG
The ratio of prevalences of glucose abnormality between men and women has been estimated in many studies, but so far there has been no consistent trend [16,17]. In the DECODE study, we found there is a clear pattern in the prevalence of postload hyperglycemia and the prevalence of fasting hyperglycemia by sex . Undiagnosed diabetes and IFG defined by isolated fasting hyperglycemia was more common in men than in women at 30–69 years of age, whereas the prevalence of isolated postload hyperglycemia was higher in women than in men and was particularly high in the elderly population . In the DECODA study, sex difference was not as clear as in Europe.The prevalence of IFG also seems higher in Chinese and Japanese men than in women, whereas it was higher in Indian women than in men. IGT was more prevalent in Chinese and Japanese women than in men, but such a difference was not observed in Indians . Sex differences in the prevalence of diabetes and IGR depend on how the prevalence was estimated, by fasting or by postload hyperglycemia, on the age distributions, and on the ethnic groups. Asian Indians, who have a very high risk of diabetes, show abnormalities in fasting glucose values at an earlier age than other populations.
The ratio of IGT to diabetes
The ratio of IGT to diabetes has been reported to decrease as prevalence of diabetes rises  and may have some predictive value in determining the stage of a glucose intolerance epidemic within a population . When the ratio is high but the prevalence of diabetes is low, the early stage of a diabetes epidemic may be occurring . The age- and sex-specific ratios of IGR to diabetes according to the newly revised diagnostic criteria for diabetes  are shown in Figure 3.8a for Asian and Figure 3.8b for European populations [10,12]. The ratio of IGR to diabetes declined when the prevalence of diabetes increased in both Asian and European populations.
Accumulating evidence shows that the prevalence of diabetes is increasing with time over recent decades.This upward trend has been seen primarily in developing countries [3,64] (Figure 3.1). A series of studies in the southern Indian city of Chennai showed a steady increase in the prevalence of diabetes in the Indian population. The age- standardized prevalence increased from 8.2% in 1988–1989  to 11.6% in 1994–1995  and reached 13.5% in 2000 , a 65% increase within the two decades. During the last two decades a considerable amount of information has been obtained from China. Studies conducted in China between 1980 and 1990 consistently show low diabetes prevalence rates of approximately 1.5% or less, even in urban populations such as in Shanghai in 1980 [68–71]. The prevalence of diabetes in Shanghai in 1980 was close to 1%. In rural Guangdong province it was 0.33% . Studies undertaken in the late 1990s, however, indicate sharply rising prevalence rates in China [72–74] and the rates estimated at the beginning of the current century show that diabetes in an urban Chinese population in mainland China  is already as prevalent as in Hong Kong and Taiwan in the mid 1990s (Figure 3.9) [76–78]. In 2007, the prevalence of T2DM in China was almost 10% indicating a three-fold increase in three decades . It Turkey, the prevalence of T2DM doubled during a 12-year interval from 1998 to 2010 . The prevalence of both diabetes and its microvascular complications in a Pacific Island population (20 years or older) of Western Samoa was examined in 1978 and 1991 . In 1978, the crude prevalence rates were 3.4 and 8.7% in rural and urban populations, respectively. By 1991, these rates had risen to 6.5 and 9.0% in two rural communities and to 16.0% in the urban setting of Apia.
Recent studies indicate that diabetes prevalence continues to increase even in developed countries. According to the data from the US NHANES II and NHANES III surveys, using ADA criteria, the prevalence of T2DM in the US adult population aged 40–74 years of age increased from 8.9% in the period 1976–1980 to 12.3% by 1988–1994. A similar increase was found when WHO criteria were applied (11.4 and 14.3%) . In Australia, the total prevalence of diabetes had increased from 3.4 to 7.2% from 1981 to 1999–2000 and the difference persisted after adjustment for BMI  (Figure 3.10). In an adult Norwegian population , the crude prevalence of known diabetes in men increased from 2.6% in 1984–1986 to 3.3% in 1995–1997, an increase of 24%, but the increase was not found in women. Over the same time period, an increase of 86% in the prevalence of obesity defined by BMI ≥30 kg m−2 was observed in men, which was much higher than the increase of 38% in women. In a Danish study of a 60-year-old cohort over a 22-year period an increase of 58% in men and 21% in women in the prevalence of diabetes was observed, which was fully explained by a concurrent increase in BMI . Rising trends in the prevalence of diabetes and obesity have also been reported in other European countries [81,82]. In addition to the increase in obesity, reduced physical activity resulting from changes in work-related activity and sedentary lifestyle has contributed to the rising trend in T2DM.
Prevalence, which reflects the accumulation of the patients at any given time, can be influenced by many factors such as an increase in the number of new cases and a reduction in the mortality attributed to the disease. There is evidence that mortality in diabetes has declined in men in the United States .Thus, a rise in prevalence could be a result of an improved survival of diabetic subjects. However, studies also show an increasing trend in diabetes incidence due to the increase in obesity and decrease in exercise. Knowler et al.  compared the incidence rates over two 10-year time periods, 1965–1975 and 1975–1985, in Arizona Pima Indians, and found that over the 10-year period the incidence rates increased by 50% in most age and sex groups. The San Antonio Heart Study revealed an increasing secular trend in the 7- to 8-year incidence of T2DM occurring from 1987 to 1996 in Mexican American and non-Hispanic Whites . Therefore, both increased incidence and decreased mortality among diabetic subjects have contributed to the increased trend in the prevalence of diabetes.