The IDF has also published adult prevalence rates for impaired glucose tolerance (IGT) which closely reflect those for type 2 diabetes. Conversion rates from IGT to diabetes have been reported at 5 – 11% per annum….
Overall prevalence corrected for age for both type 2 diabetes and IGT is set to increase from 6.0% to 7.3% and 7.5% to 8.0% respectively over the 18 years from 2007 to 2025 – an absolute increase from 246 to 380 and 308 to 418 million persons aged 20 – 79 years, respectively (Figure 7.1).
The highest rates are currently in the Eastern Mediterranean and Middle East with North and South America close behind. These reflect the increased life expectancy and overall ageing of the North American population (diabetes is more common in older years). In terms of absolute numbers, the Western Pacific region (particularly China) will have the largest increase of nearly 50%, to 100 million people with diabetes by 2025.
The highest number of people with diabetes is currently in the 40 – 59-year-old age group, but there will be almost parity with 60 – 79 year olds by 2025, at 166 and 164 million worldwide respectively.
There is considerable variation within each region, however. For example, in the Western Pacific, the tiny island of Nauru has a comparative prevalence in 2007 of 30.7%, whilst nearby Tonga has less than half that rate at 12.9%, the Philippines 7.6% and China 4.1%.
In the European region, comparative rates range from 1.6% in Iceland to 7.9% in Germany, Austria and Switzerland. The UK rate is 2.9% age adjusted and 4.0% absolute, increasing to 3.5% and 4.6% respectively in 2025 (representing an increase from 1.7 to 2.16 million in absolute numbers).
There is a global trend for rates of diabetes to increase in populations as they move from a rural to an urban existence. The reasons are unclear but probably relate to both decreasing physical activity as well as dietary changes. For example, rural Chinese have a prevalence of type 2 diabetes of 5%, less than half the rate of Singaporean Chinese (10.5%). Much larger differences are seen in South Asian, Hispanic, African and Polynesian peoples (Figure 7.2).
Impaired glucose tolerance
Comparative prevalence for IGT vary by region with rates almost double those for type 2 diabetes in Africa, but slightly lower elsewhere. These differences are almost certainly a reflection of socio-economic factors as well as a paucity of studies in many African countries where extrapolation is necessary between very different populations. In Europe, the comparative prevalence will increase slightly from 9.1% in 2007 to 9.6% in 2025, representing an absolute change from 65.3 to 71.2 million (UK figures 4.7% to 4.9%, 2.17 to 2.4 million respectively).
Reported incidence rates vary according to population under study and year of observation. For white Europid populations, rates of 0.1 – 1% per annum have been reported. For Hispanic populations in the USA, rates of 2.8% were recorded in the San Antonio Study, similar to those of the Pima Indians in Arizona (approximately 2.5%), and Australian aborigines (2.03%).
Over 20 years, incidence in the Pima has not changed although the age of onset has been declining. The occurrence of type 2 diabetes in adolescence is now a great cause for concern worldwide. In US Asian and Pacific Islanders, for example, rates of 12.1/100,000 patient – years have been reported in 10 – 19 year olds, similar to rates reported for type 1 diabetes. In the UK, the overall incidence for < 16 year olds is much lower, at 0.53 per/100,000 patient-years, but 10 times more common in South Asian or black African compared to white children.
The rural:urban ratio remains for incidence even in the presence of other risk factors such as central obesity. In Japanese, there is an approximately threefold increase in incidence for obese urban compared to rural populations (15.8 versus 5.8% over 10 years). Similarly, there is a twofold increase in incidence for USA versus Mexican Hispanic people corrected for age and economic circumstance, probably a reflection of changes in diet and lifestyle. For the Pima, the contrast is more striking, with a > 5 – fold increase in those living in the US compared to northern Mexico.
The magnitude of these figures has opened a debate on population screening for diabetes, but a Health Technology Assessment report in the UK from 2007 and a report from the US Preventive Services Task Force in 2008 both concluded that there is not enough evidence at present to support such a policy.
Risk factors for development of type 2 diabetes
About 80% of people with type 2 diabetes are obese, and the risk of developing diabetes increases progressively as the BMI (weight (kg)/height (m) 2) increases. A BMI > 35 kg/m 2 increases the risk of type 2 diabetes developing over a 10-year period by 80-fold, as compared to those with a BMI < 22 kg/m 2 Latest data from the NHANES survey in the USA confirm a 6 – 10-fold increased lifetime risk of type 2 diabetes for 18 year olds with a BMI > 35 kg/m 2 compared to those < 18.5 kg/m 2 with an associated average 6 – 7 year reduction in overall life expectancy. Obesity is still widely defined as a BMI > 30 kg/m 2 although BMI is not an accurate reflection of fat mass or its distribution, particularly in Asian people. A simple waist circumference may be better (see metabolic syndrome below).
The pattern of obesity is also important in that central fat deposition has a much higher risk for development of diabetes compared to gluteofemoral deposition. In clinical practice such central obesity can be assessed by measuring the weight:hip circumference ratio, but it is unclear whether this has any advantage over a simple waist circumference. Fat deposition at other sites, particularly muscle, liver and islets, may contribute to metabolic defects and insulin resistance (so-called lipotoxicity).
Physical exercise and diet
Low levels of physical exercise also predict the development of type 2 diabetes, possibly because exercise increases insulin sensitivity and helps prevent obesity (Figure 7.4). Subjects who exercise the most have a 25 – 60% lower risk of developing type 2 diabetes regardless of other risk factors such as obesity and family history.
There has been extensive research into the role of diet as a risk factor for type 2 diabetes. A study in over 10,000 35 – 55 year olds found that a diet containing large quantities of soft drinks, burgers, sausages and low fiber explained 5.7% of insulin resistance as assessed by the HOMA model. There were 77,440 person-years in the study with 427 incident cases of type 2 diabetes.
The Diabetes Prevention Program and Diabetes Prevention Study in the USA and Finland have shown that lifestyle modifications with moderate exercise and modest weight loss can dramatically reduce the number progressing from IGT to type 2 diabetes and reinforce the importance of lifestyle factors in the cause of diabetes.
Insulin resistance can be estimated from the amount of glucose that is infused intravenously in order to maintain a constant blood glucose during a simultaneous intravenous insulin infusion. This method is cumbersome, however, and for population purposes it has been largely superseded by the HOMA (homeostasis model assessment) estimate of steady-state β cell function (HOMA B) and insulin sensitivity (HOMA S) as percentages of normal. These can be derived from a single fasting plasma C peptide, insulin and glucose concentration. Insulin resistance (or, more correctly, diminished insulin sensitivity) precedes the onset of diabetes and can worsen with increasing duration.
Hormones and cytokines
Visceral fat liberates large amounts of non-esterified fatty acids (NEFAs) through lipolysis, which increases gluoconeogenesis in the liver and impairs glucose uptake and utilization in muscle. NEFAs may also inhibit insulin secretion, possibly by enhancing the accumulation of triglyceride within the β cells. In addition, adipose tissue produces cytokines, such as TNF-α , resistin and IL-6, all of which have been shown experimentally to interfere with insulin action. TNF-α has been shown to inhibit tyrosine kinase activity at the insulin receptor and decrease expression of the glucose transporter GLUT-4.
Adiponectin is a hormone with antiinflammatory and insulin-sensitizing properties that is secreted solely by fat cells. It suppresses hepatic gluconeogenesis and stimulates fatty acid oxidation in the liver and skeletal muscles, as well as increasing muscle glucose uptake and insulin release from the β cells. Circulating adiponectin is reduced in obesity and a recent meta-analysis showed that the relative risk for diabetes was 0.72 for every 1-log μ g/mL increment in adiponectin level.
Resistin is an adipocyte-secreted hormone that increases insulin resistance and was first described in rodents, being found in increased levels in experimental obesity and diabetes. In humans, it appears to be derived largely from macrophages, however, and its precise role in human diabetes is uncertain, although higher circulating levels have been found in some people with type 2 diabetes.
Leptin is an adipokine that was found to be absent in the ob/ob mouse model of obesity and diabetes. Its normal function is to suppress appetite, thus providing a candidate mechanism linking weight gain and appetite control. Although abnormal leptin function has been described in humans, these defects are very rare and paradoxically high levels have been found in type 2 diabetes. Ghrelin is a recently described peptide secreted from the stomach and may act as a hunger signal. Circulating levels are negatively correlated with BMI and are suppressed by food intake. It has no known role in human diabetes but antagonism may provide a therapeutic target. Finally there is often increased sympathetic nervous system activity in obesity, which might also increase lipolysis, reduce muscle blood flow and thus glucose delivery and uptake, and therefore directly affect insulin action.
Many of these cytokines are involved in the acute-phase response and it is therefore not surprising that circulating markers such as C-reactive protein and sialici acid are increased in type 2 diabetic patients, as well as in those who later go on to develop the condition. Because these markers have also been found to be elevated in patients with atherosclerosis, a unifying hypothesis has evolved proposing that inflammation may be a common precursor and link between diabetes and coronary artery disease.
Evidence for a genetic basis for type 2 diabetes comes from a clear familial aggregation, but it does not segregate in a classic Mendelian fashion. About 10% of patients with type 2 diabetes have a similarly affected sibling. The concordance rate for identical twins is variously estimated to be 33 – 90% (17 – 37% in non-identical twins), but the interpretation of this is controversial as part of the explanation for the high concordance may be environmental rather than genetic.
Unlike type 1, type 2 diabetes is not associated with genes in the HLA region. So far, 19 gene variants have been described and validated as being associated with type 2 diabetes. Of these, the strongest is TCF7L2; 15% of European adults carry two copies of the abnormal gene and they have double the lifetime risk of developing type 2 diabetes compared to the 40% who carry no copies. Carriers of the T risk allele have impaired insulin secretion and enhanced hepatic glucose output. Nearly all of the other described genes affect either β cell mass or function; few appear to have potential effects on insulin resistance.
A link between low birthweight and later development of type 2 diabetes in a UK population has led to a hypothesis linking foetal malnutrition to impaired β cell development and insulin resistance in adulthood. Abundant adult nutrition and consequent obesity would then expose these problems, leading to IGT and eventually type 2 diabetes. This has been called the thrifty phenotype hypothesis (Figure 7.7).
A meta-analysis of 31 populations involving 152,084 individuals from varying ethnic groups and 6,090 cases of diabetes was published in 2008. This confirmed a negative association between birthweight and diabetes in 23, but found a positive association in eight studies. The combined odds ratio for type 2 diabetes was 0.8 (95% CI 0.72 – 0.89) for each 1 kg increase in birthweight. This relationship was strengthened if macrosomic (birthweight > 4 kg) and offspring of mothers with known type 2 diabetes were excluded (odds ratio (OR) 0.67, 95% CI 0.61 – 0.73). Notably there was a tendency for a positive relationship in North American populations largely due to higher rates of maternal obesity and gestational diabetes. Adjustment for socio-economic status had no effect, but adjustment for achieved adult BMI attenuated the relationship.
With increasing maternal obesity and gestational diabetes mellitus (GDM), it is conceivable that the relationship will change to the pattern currently seen in Native Americans which is more U shaped. However, it is still unclear whether low birthweight is a causative factor or a sign of other potential mechanisms which may predispose to later diabetes.
This is a proposal that type 1 and type 2 diabetes are essentially the same in that both result ultimately from β cell failure. The aetiology obviously differs but superimposed insulin resistance drives the process. Three accelerators are proposed: constitution – individuals have increased β cell apoptosis; insulin resistance – underpinned by physical inactivity and visceral adiposity; and autoimmunity – mainly operative in younger patients and linked to HLA susceptibility alleles. The overlapping driver of obesity would explain increasing rates of ‘type 1’ and ‘type 2’ diabetes. This intriguing idea is currently widely debated and awaits confirmatory studies.
The aggregation of obesity, hyperglycemia, hypertension and hyperlipidemia in people with both type 2 diabetes and cardiovascular disease is now termed the metabolic syndrome (Table 7.1). This concept is not new, it is said to have been first described in 1923, but latterly there have been attempts to standardize its definition.
Since these definitions appeared, there has been considerable debate as to their relative strengths and weaknesses. Indeed, there is some debate as to whether this constitutes a true syndrome at all and whether they add anything to predictive models for type 2 diabetes and coronary artery disease. A major problem is the correlation of many of the features. In prospective studies, fasting plasma glucose (FPG) is overwhelmingly linked to subsequent development of diabetes, but much less so with coronary artery disease. Thus the predictive utility of the metabolic syndrome as a concept adds little to its constituent risk factors when they are used individually. The long – term usefulness of the definition of the metabolic syndrome for identification and intervention in order to prevent diabetes and cardiovascular disease has yet to be demonstrated.
Type 2 diabetes develops because of a progressive deterioration of β cell function, coupled with increasing insulin resistance for which the β cell cannot compensate. At the time of diagnosis β cell function is already reduced by about 50% and continues to decline regardless of therapy (Figure 7.8).
The main defects in β cell function in type 2 diabetes are a markedly reduced first- and second-phase insulin response to intravenous glucose, and a delayed or blunted response mixed meals (Figure 7.9). There are also alterations in pulsatile and daytime oscillations of insulin release. Some researchers have found increases in the proportions of plasma proinsulin and split proinsulin peptides relative to insulin alone. Many of these abnormalities can be found in people with IGT and even in normoglycemic first-degree relatives of people with type 2 diabetes, indicating that impaired β cell function is an early and possibly genetic defect in the natural history of type 2 diabetes (Figure 7.10).
The most common histological abnormality found in the islets of patients with type 2 diabetes is the presence of insoluble amyloid fibrils lying outside the cells. These are derived from islet amyloid polypeptide (IAPP, also sometimes known as amylin). This is co-secreted with insulin in a molar ratio of 1:10 – 50. Although IAPP is reported to impair insulin secretion and to be toxic to the β cell, its precise role in the pathogenesis of type 2 diabetes is uncertain because deposits can be found in up to 20% of elderly people who had completely normal glucose tolerance in life.
β Cell mass is thought to be decreased by only 20 – 40% in type 2 diabetes and this clearly cannot explain the > 80% reduction in insulin release that is observed. There must therefore be additional functional defects in the β cell, perhaps mediated by glucose or lipid toxicity. It is likely that IAPP contributes to this process.
Both insulin resistance and β cell dysfunction are early features of glucose intolerance, and there has been much debate as to whether one is the primary defect and precedes the other. In practice, the contribution of insulin resistance and β cell dysfunction varies considerably between patients, as well as during the course of the disease. Usually, there is a decline in both insulin sensitivity and insulin secretion in patients who progress from IGT to diabetes and undoubtedly environmental and genetic factors contribute to this process (Figure 7.11).
Next Week: Other types of diabetes
For more information and to purchase this book, just follow this link: