The impact of rapid urbanisation on the diabetes epidemic is exemplified by the parallel increase in gross domestic product (GDP) and diabetes prevalence in China (Fig. 5.1). In less than three decades, while the GDP has increased by 100-fold, the prevalence of diabetes has also increased by tenfold. In the 2010 National Survey of 98,658 adults and using both 75 g oral glucose tolerance test (OGTT) and glycated haemoglobin (A1c) as diagnostic criteria, one in nine Chinese had diabetes, one in three had obesity (general or central) closely associated with hypertension and dyslipidemia and one in two had prediabetes. Amongst those with diabetes, only 30 % were diagnosed and amongst the diagnosed, only 30 % were treated and amongst the treated, only 40 % were controlled, defined as A1c less than 7 % [4].

According to the Diabetes Epidemiology Collaborative Analysis of Diagnostic Criteria in Asia (DECODA) data, 30–50 % of subjects in the WP Region had metabolic syndrome, characterised by central obesity and clustering of cardiometabolic risk factors, and a strong predictor for future development of diabetes [5]. In 2008, 60–70 % of people in Tonga, Nauru and Cook Islands were considered obese [6]. In 2004, the sex-standardised diabetes prevalence was reported to be 13.0 % in men and 14.4 % in women in Nauru. The age-specific prevalence rose from 4.5 % in the 15–24 age group to 42.7 % in the 55–64 age group [7]. Tables 5.2 and 5.3 summarise the prevalence and number of people with diabetes in the WP countries based on the IDF figures published in 2013 [8].

Rapid quantitative and qualitative changes in nutritional intake, notably a shift from a traditional high-fibre, low-fat diet to an energy-dense diet with increased meat and fat consumption as well as reduced physical activity due to mechanisation and increased use of mobile vehicles, are important drivers for this global epidemic [9]. In the islands of Vanuatu, researchers reported close associations of economic development with increased consumption of animal proteins and simple carbohydrates, as well as increased tobacco and alcohol consumption in men in areas with high tourism [10]. In China, during the last two decades, changes in food technology with production of high-calorie foods together with reduced physical activity have contributed to the rising burden of obesity in both adults and children [11].

Age, family history and obesity are the main risk factors for diabetes [4] although inadequate beta-cell response to overcome insulin resistance due to factors such as ageing, obesity and inflammation remains the primary culprit [12]. In an analysis of 74 study cohorts comprising 3813 individuals (19 African, 31 Caucasian and 24 East Asian cohorts), the authors examined the hyperbolic relationship between insulin sensitivity index (IS) and acute insulin response to glucose (AIRg) in healthy cohorts. Amongst these three subpopulations, Asians had a steep and non-linear relationship with small changes in one variable having large effect sizes on the other. Besides, Asians had the lowest body mass index (BMI) and AIRg and developed diabetes only with a small increase in BMI compared to the Caucasians and Africans [13].

Compared to their Europid counterparts, people of Asian ethnicity are more prone to develop diabetes for the same level of BMI or waist circumference due to their propensity to store excessive body or visceral fat [14]. Even amongst lean subjects and given the same amount of visceral fat, Asians were more insulin resistant than Europids due to increased concentration of free fatty acids and inflammatory markers [15]. Using insulin clamp studies, researchers from Thailand reported insulin deficiency as the predominant feature in lean subjects and insulin resistance and in the overweight and obese subjects [16].

In autopsy studies, Korean researchers have reported close correlations between BMI and relative percentage of beta cells in both diabetic (r = 0.55) and nondiabetic specimens (r = 0.35). However, diabetic subjects had a lower percentage of beta cells than nondiabetic subjects for the same BMI [17]. In healthy subjects, both Chinese and South Asian subjects with normal glucose tolerance exhibit higher glucose excursion during an OGTT with reduced rate of glucose disposal than their Caucasian counterparts [18]. This inherently low beta-cell function together with propensity to develop obesity which will impose metabolic stress has provided a highly plausible explanation for the high risk of diabetes in Asian populations undergoing rapid transition (Fig. 5.2).

According to the estimates by IDF in 2013, 36 % of global deaths in adults are due to diabetes affecting 5.1 million people. The most populous WP Region had the highest number of diabetes-related deaths affecting more men (1,080,000) than women (789,000). Of note, 44 % of deaths due to diabetes occurred in people under the age of 60 [1]. On average, diabetes reduces life expectancy by 6–12 years, especially in people diagnosed young, and increases the risk of vascular, cancer, nonvascular, non-cancer deaths by 1.3–3-fold. The nonvascular, non-cancer-related deaths are mainly due to renal failure, mental illnesses, hepatobiliary disease and sepsis with fasting plasma glucose having linear relationships with all clinical outcomes, even after adjustment for confounders such as age, sex, BMI and lifestyle factors [55].

In the WP Region, in contrast to Europids in whom cardiovascular disease is a leading cause of mortality and morbidity, diabetic kidney disease is a major complication in Asia [26], affecting 50–60 % of type 2 diabetic patients in different clinic settings [56]. In prospective studies, the annual incidence of diabetic kidney disease in Chinese population was 2–4 %, driven primarily by disease duration, smoking, metabolic syndrome, blood pressure and glycaemic control [57]. Other factors such as the use of over-the-counter medications including herbal medicine [58] and chronic hepatitis B infection [50] might also contribute to this high rate of diabetic kidney disease in the WP Region. While many patients in the WP Region die from end-stage renal disease due to lack of access to renal replacement therapy, this devastating condition can be prevented by intensive control of all risk factors using a protocol-driven and collaborative approach [59].

Countries in the WP Region differ considerably in the ecosystem of health-care provision and health literacy. In low-income areas such as some Pacific Islands where health illiteracy is common and access to care is poor, leg amputation continues to be a major cause of morbidity and mortality. In a recent survey of amputations involving 85 people with diabetes from Solomon Islands, Nauru and Vanuatu, delayed treatment (42 %), the use of traditional treatments (18 %) and insufficient knowledge about foot care (11 %) [60] are the main reasons for amputation.

On the other hand, several local, national and regional registries have provided insights regarding the interethnic differences in diabetic complications and the impact of ageing and care provision on the secular changes of these clinical outcomes. Apart from renal failure, stroke is a major disease burden in Asian diabetic populations [61]. Depending on the demographic pattern and stage of societal transition, cancer is now emerging as a leading cause of death especially in ageing societies and areas with high rate of endemic infections (e.g. hepatitis B for liver cancer) and high survival rates from cardiovascular-renal complications [62]. In the Australian National Diabetes Registry of nearly one million subjects, researchers have reported 1.3-fold increased standardised incidence risk of all-site cancer in both type 1 and type 2 diabetes except for prostate and melanoma [63]. In a national insurance database from Taiwan, diabetes-related death rates have declined over time (men, women: 3.92 %, 3.29 % in 2000; 3.64 %, 3.11 % in 2005; and 3.12 %, 2.71 % in 2009). In 2009, the estimated loss of life due to diabetes was 6.1 years in women and 5.3 years in men amongst people diagnosed at the age of 40. The four major causes of death were diabetes, malignancies, heart disease and cerebrovascular disease [64].

In the Hong Kong Diabetes Registry established as a quality improvement programme since 1995, researchers were able to track the secular changes in clinical outcomes in Chinese adults with diabetes, captured by a territory-wide clinical management system. In the early 1990s when universal health-care coverage was still being developed, stroke and end-stage renal failure were the leading causes of death. With ageing, societal affluence and access to interventions, coronary heart disease, cancer and heart failure became the leading causes of death. In 2005 and after a follow-up period of 5.5 years, amongst the 7534 type 2 diabetic patients enrolled in the registry since 1995, 763 died with the main causes of death being neoplasms (24.5 %) and cardiovascular (23.5 %), respiratory (15.3) and renal disease (15.3 %) [65].

The criteria for diagnosing diabetes and prediabetes in Asia for the most part take references from international guidelines including those of the World Health Organization (WHO) and American Diabetes Association (ADA) [69, 70]. The measurement of fasting plasma glucose for diabetes screening is widely accepted and practised. On the other hand, there are differences in regional recommendations for oral glucose tolerance test (OGTT) and glycated haemoglobin (A1c) in the routine diagnostic process. Approaches to screening are also closely related to health-care access and resources available. For instance, the Chinese Diabetes Society guideline for the prevention and control of type 2 diabetes last updated in 2010 has not yet endorsed the use of HbA1c as an alternative to glucose-based tests for diagnosing diabetes [71]. Similarly, the UNITE FOR Diabetes Philippines in its 2014 practice guideline also recommended against using A1c. They also proposed the use of OGTT for diabetes screening only in individuals with known impaired fasting glucose or metabolic syndrome [72]. In the latest report published in 2010, the Japan Diabetes Society adopted the use of A1c to complement glucose-based tests in diabetes screening but stipulated that an individual meeting the A1c threshold of 6.5 % must concurrently fulfil one other criteria of either fasting plasma glucose ≥7.0 mmol/L or 2-h post-OGTT plasma glucose ≥11.1 mmol/L to establish the diagnosis [73]. This differs from international guidelines which assign similar weights to glucose-based studies and A1c in the diagnostic schema and allows the diagnosis to be made on the basis of A1c measurement alone. The Korean Diabetes Association fully applied the ADA criteria for diagnosing diabetes since 2011 [74].

Glycated haemoglobin has a number of advantages over plasma glucose in that it summarises long-term glucose exposure, has less intra-individual variability and does not require testing during a fasted state [75]. The recommendation to include A1c in diagnostic criteria of diabetes by the international expert committee was based on epidemiological studies which demonstrated a tight association between A1c and diabetic retinopathy in mixed populations [75–77]. Progress in analytic procedures and global movement to certify A1c assay to the National Glycohemoglobin Standardization Program facilitated its wider use as a diagnostic tool as well as in disease monitoring [75, 78]. In developing countries such as China and the Philippines, however, the measurement of A1c has not yet achieved standardisation across the nation, and this remains the key limitation to its application [72].

Notwithstanding logistics of costs and instrumentation, inclusion of A1c in diagnostic criteria will uncover more patients with diabetes than using glucose-based criteria alone. In large population surveys in China and Japan, over half of the participants with newly diagnosed diabetes and prediabetes were detected with A1c. This may be due to the high prevalence of high post-glucose challenge blood glucose levels which may be detected by an integrated value like A1c but missed by using fasting plasma glucose only [4, 79].

Increasingly, it is recognised that ethnicity may modify the relationship between measured A1c and blood glucose levels. Several studies have shown that for similar distribution of blood glucose, A1c levels were consistently higher in Asians compared to Caucasians and that amongst Asians, higher in Indians and Malays relative to Chinese [80, 81]. A few factors may contribute to such racial disparities such as differences in haemoglobin glycation rate and in population prevalence of haemoglobinopathies which interfere with A1c measurement.

Against this background, concerns are raised regarding the optimal A1c threshold for detecting diabetes in Asian groups and whether ethnic-specific standards are required. Studies carried out in Asians have suggested that lower cut-offs of A1c correlated more closely with diabetes as defined using the gold standard of OGTT [82–84]. In a recent population-based study comprising three ethnic groups of Malay, Chinese and Indian which examined the cross-sectional relationship between A1c and moderate retinopathy, it was found that although A1c differed slightly in its sensitivity and specificity in predicting retinopathy amongst different ethnic groups, eye complication was generally rare when A1c fell below 6.5 % [85]. Results from this Asian study concurred with those of a large multinational database which confirmed that an A1c threshold of 6.5 % has similar discriminatory power for the presence or absence of diabetic retinopathy in Caucasians and Asians [77, 85].

Oral glucose tolerance test is the gold standard test for the diagnosis of diabetes and is necessary in making a diagnosis of impaired glucose tolerance (IGT). The requirement of a second blood sample and being more time consuming has greatly limited its practice in many Asian countries. However, the importance of the 2-h plasma glucose has been highlighted in studies conducted in Asia which showed that fasting plasma glucose alone could only identify half of the people with diabetes and that individuals with predominantly post-load hyperglycaemia are phenotypically different from those with fasting hyperglycaemia [86]. As discussed, Asian diabetic patients have lower BMI and less beta-cell reserve than their Caucasian counterparts and thus are more likely to fail glucose stress test. On this premise, a lower threshold to perform OGTT in Asians should be considered. In Chinese, a paired value of A1c ≥6.1 % and fasting plasma glucose ≥6.1 mmol/L had 13–17 times increased likelihood to occur in diabetic subjects compared to nondiabetic subjects [87, 88]. In resource-limited setting, OGTT will have the greatest yield in detecting diabetes when paired A1c and fasting plasma glucose are above these cut-off values.

The chronic, silent and nonurgent nature of diabetes often leads to late diagnosis and delayed treatment. Once diagnosed, people with diabetes have pluralistic needs including information, social and psychological support, education, medical and surgical treatment, to name but a few. Given these multiple demands, large population size, lack of capacity and resource restraints, prevention and control of diabetes is often challenged with care fragmentation, clinical inertia and treatment non-adherence with insufficient community support [3, 89, 90]. In both randomised disease management programmes and observational cohorts, attaining multiple treatment targets and the use of organ-protective drugs such as statins and RAS inhibitors have been shown to reduce cardiovascular-renal complications and related deaths in the WP Region [91, 92]. Yet large-scale surveys have demonstrated only 30 % of patients with diabetes ever had assessments for risk factors and complications. Amongst those who had assessments, only 30–40 % attained one of the three ABC targets (A1c <7 %, BP <130/80 mmHg, LDL-C <2.6 mmol/L) and 5–10 % attained all three targets [93] with particularly low rates amongst people with young-onset diabetes [30, 94].

Diabetes is a lifelong disease which requires acquisition of knowledge, change of attitudes and formation of new habits in order to reduce disease burden. It is also a biological problem that requires medications, technologies and monitoring which have resource implications. In a survey from Guam involving 125 patients, 40 % were not aware of the type of diabetes they had, 20 % had not received education on self-management and 30–60 % had not received advice on tobacco cessation, immunisation services, nutritional counselling and/or regular eye and foot examinations. Over 50 % of patients expressed their desire to have preventive and education services with enhanced access to specialists and specialised care as well as financial support to cover the costs of chronic care and medications [95].

There is now consensus that a chronic care model involving community mobilisation and engagement supported by an integrated health-care system with country/area appropriate financing and health-care organisation is needed to promote interactions between an informed person with diabetes and a proactive health-care team to achieve positive clinical outcomes [96]. In this information and technology era, there is a growing advocacy to use improvement science, often entailing the use of real-world data, to inform and change clinical practice [97]. In the case of diabetes, changing the workflow and using protocols to collect data systematically for establishing registry can generate personalised report to promote shared decision-making. On a broader perspective, data transparency and comparison of performance indexes may also motivate practice and policy changes through ongoing evaluation and sharing of best care models [98–100].