Table 24.2 presents the known modifiable and non-modifiable risk factors or aetiological determinants associated with type 2 diabetes (27). The overall risk of type 2 diabetes must be assessed on the basis of all of these. Because of the additive effect of different risk factors and determinants, individuals with high levels of non-obesity risk may develop type 2 diabetes without becoming obese, while in other cases obesity alone may be sufficient to lead to diabetes. Generalized and central obesity are just two of the interrelated risk factors associated with type 2 diabetes, and of the modifiable lifestyle factors are probably the most important in terms of size and consistency of effect.
Table 24.2 Aetiological determinants and risk factors of type 2 diabetes
A. Genetic factors
Genetic markers, family history, thrifty gene hypothesis etc.
B. Demographic determinants Sex, age, ethnicity
C. Behavioural and lifestyle-related risk factors Obesity (including distribution of obesity, duration) Physical inactivity
D. Metabolic determinants and intermediate risk categories of type 2 diabetes
Impaired glucose tolerance, impaired fasting glucose Insulin resistance
Other components of the metabolic syndrome Pregnancy-related determinants (parity, gestational diabetes, diabetes in offspring of women with diabetes during pregnancy, intrauterine environment)
The association between obesity and type 2 diabetes has been observed in both cross-sectional (13-19, 28-33) and prospective studies (20-23, 31,34-38) across various populations. Obesity confers a minimum 3- to 10-fold risk of type 2 diabetes
(39) and it is estimated that type 2 diabetes risk could be reduced by 50-75% by control of obesity
(40). Weight loss associated with the onset of diabetes (36,41) means that the association of obesity with type 2 diabetes prevalence is generally weaker than its association with incidence. For example, in an Israeli study, the main determinant of the incidence of type 2 diabetes over a 10-year study period was the BMI at baseline, rather than the BMI at follow-up when glucose tolerance was measured (38). Incident impaired glucose tolerance was associated with both concurrent and prior BMI, as would be expected if weight loss only occurs after glucose tolerance has deteriorated to frank diabetes.
In a study of Pima Indians baseline BMI was also strongly related to the incidence of type 2 diabetes but there was little association between diabetes prevalence and concurrent obesity (42). In a small subgroup of subjects of that study with BMI data from 4 years before diagnosis to 2 years after, there was a clear pattern of weight gain preceding diag nosis, followed by weight loss after diagnosis. Older participants developed diabetes at a lower BMI than younger individuals, suggesting that age-related deterioration in insulin sensitivity enables the development of diabetes at lower levels of adiposity than required for the development of diabetes in younger subjects.
In contrast to the results in Pima Indians, Tai et al. (31) found that BMI was similarly associated with the prevalence or 4-year cumulative incidence of diabetes, with odds ratios of 1.12 and 1.14 respectively for a 1 unit increase in BMI in slim (mean BMI 23 kg/m2) Chinese.
If diagnosis is made early in the natural history of diabetes before weight loss occurs it could be expected that a stronger positive association between BMI and prevalent type 2 diabetes would be observed.
The Importance of Duration and Changes in Obesity for Risk of Type 2 Diabetes
Although the duration of obesity is considered important in determining the risk of obesity associated conditions, including type 2 diabetes, there is little information available to quantify this relationship. Even in most prospective studies that actual onset of obesity is not measured and can only be estimated by recall. Moreover, if weight is changing it is difficult to differentiate between the effects of the degree and duration of obesity.
In the above mentioned large study in Israel, Modan et al. (38) found that obesity lasting for less than 10 years was not associated with a major increase in diabetes incidence compared with that in subjects who had remained slim (BMI < 23 kg/ m2). The risk of type 2 diabetes was increased in subjects who had lost weight to reach a specific BMI class relative to those who had remained stable within that class, while those with a stable BMI had in turn a greater risk of type 2 diabetes than those who had increased their BMI class, indicating that weight gain per se was not associated with increased risk of type 2 diabetes (38). Similar results were observed in the multiethnic population of Mauritius. For any level of BMI at 5 years of follow-up, the highest prevalence of type 2 diabetes was associated with weight loss since baseline, and the lowest with weight gain, while those whose weight had remained stable had an intermediate risk. Comparing the weight gainers with weight maintainers in both studies suggests that the stable group were at greater risk due to their longer duration at the higher current BMI.
In one of the few reports to actually examine the levels of glucose tolerance associated with different duration of self-reported obesity (based on percentage of standard weight ranging from 14 to 137% overweight), Ogilvie (43) observed that it took 5 to 18 years of obesity for glucose intolerance to develop, and 12 to 38 years for diabetes to occur. In contrast to other studies, the degree of obesity was not associated with glucose tolerance. In a more recent study, Felber et al. (44,45) examined fasting plasma glucose levels and glucose storage capacity in relation to obesity duration cross-sectionally in 67 moderately obese subjects (mean ideal body weight 150 + 3%). Participants with a duration of obesity less than 17 years generally had low fasting glucose concentrations and high rates of glucose uptake, while beyond this time fasting glucose began to increase and glucose storage capacity to fall. The time when fasting glucose began to rise and glucose uptake declined was considered the onset of type 2 diabetes; thus the results were interpreted as showing that at least 17 years of moderate obesity was required before type 2 diabetes developed. It seems likely that the actual level of obesity and other risk factors would affect the duration of obesity that had to be experienced before glucose tolerance deteriorated to diabetes.
Evidence for a specific effect of weight gain on type 2 diabetes comes from two American studies, where self-reported weight gain throughout adulthood or immediately prior to the study period was associated with increased risk of diabetes independent of BMI in early adulthood (34,35); although weight gain was no longer significant if attained BMI was controlled for. However, if the effect of weight gain is modelled controlling for attained BMI, it is effectively modelling duration, as obese people who have not gained weight also have a longer duration of obesity (46). This emphasizes the difficulty in delineating the effects of current BMI, duration of BMI, and weight gain. Harris (47) also indicated that weight gain between 25 and 50 years of age was a risk factor for type 2 diabetes, and Di Pietro et al. (48) have shown a rapid weight gain between puberty and age 25 years in a cohort of Swedish adults who were overweight in childhood and went on to develop diabetes. Weight gain also preceded type 2 diabetes in Pima Indians (36,42).
However, as mentioned earlier weight change had little effect on risk of type 2 diabetes in the Israeli study, and most of the incident cases of type 2 diabetes had not changed weight over 10 years of follow-up, and BMI had been at 27 kg/m2 or greater for at least the period of the study (38).
The effect of weight gain on diabetes incidence was quantified in a follow-up study of the baseline NHANES (National Health and Nutrition Examination Survey) cohort in the USA (49). The 9-year diabetes incidence, as determined from death certificates, hospitalization and nursing home records, increased by 4.5% for every kilogram of weight gained. This was after controlling for a number of factors including baseline BMI, but not BMI at diagnosis. From this result it was estimated that the average weight gain of 3.6 kg recorded between NHANES II and NHANES III could theoretically give rise to a 16% increase in diabetes incidence between 1990 and 2000.
Numerous mechanisms have been proposed linking obesity and glucose intolerance. Most obese individuals with type 2 diabetes are also insulin resistant, while lean subjects with type 2 diabetes are likely to have a defect in insulin secretion. A continuum between obese glucose tolerant, obese glucose intolerant, obese diabetic with hyperins-ulinaemia, and obese diabetes with hypoins-ulinaemia has been proposed by Golay et al. (45) and is supported by the work of others (50,51). Deficiency in glucose storage as glycogen is evident in each of these groups of obese subjects. The first step in both glucose storage and oxidation is through glucose 6-phosphate. Glucose storage proceeds under the action of glycogen synthase and mobilization is controlled by glycogen phos-phorylase. Glucose oxidation proceeds via glycoly-sis to the citric acid cycle. In obesity there are increased circulating levels of free fatty acids (FFAs) and elevated lipid oxidation. This results in metabolic products (acetyl-CoA and citrate) which inhibit glucose mobilization. Intracellular glycogen therefore accumulates, inhibiting glycogen synthase and glucose storage (45). This occurs independent of the positive effect of hyperglycaemia and hyperin-sulinaemia on glucose storage.
As long as glucose tolerance is normal in obesity, the negative effect of increased intracellular glycogen on glucose uptake is smaller than the positive effect of hyperinsulinaemia and hyperglycaemia. Glucose intolerance occurs when the stimulating effects of increased glucose and insulin can no longer overcome the resistance to glucose storage— hence resulting in continuous hyperglycaemia. A vicious cycle develops with higher fasting glycaemia inhibiting glucose storage and inhibition of glucose storage causing hyperglycaemia (45).
Initially hyperglycaemia is accompanied by an increased insulin response to a glucose load, but eventually ^-cell response becomes insufficient, and although basal insulin levels may still be elevated in comparison to lean subjects, insulin response to a glucose load or meal is diminished and hyper-glycaemia persists (45,50,51).
This simplified account provides a framework within which the association between obesity and type 2 diabetes can be understood. The lowering of fat stores with weight loss enables mobilization of glycogen and therefore increased uptake of glucose for storage. In the short term, an energy-deficient diet or exercise will also improve glucose uptake by facilitating glycogen mobilization without any change in body fat content. Weight gain, as well as current weight, has been considered a risk factor for type 2 diabetes. At any level and duration of BMI, extra body fat could tip the balance to impaired glucose uptake. Behavioural factors resulting in weight gain, such as dietary changes or reduced physical activity, may also promote the development of type 2 diabetes.
Abdominal fat, especially the visceral rather than subcutaneous depots, is strongly associated with the metabolic complications of obesity (52). Kissebah has recently reviewed the current understanding of the relationship between abdominal adiposity and metabolic changes leading to insulin resistance and glucose intolerance, elevated blood pressure and dyslipidaemia (52). Microcirculatory changes in blood flow may contribute to insulin resistance, along with a primary neurohumoral dys-regulation, enhanced by genetically or environmentally overactive 'arousal systems'. The increased lipolysis associated with visceral adipose tissue, and the resultant increase in FFA flux to the liver, may impair hepatic insulin extraction, while increased androgenic hormones in centrally obese women could also contribute to insulin resistance (52). There are several issues pertinent to the relationship between fat distribution and diabetes.
Anthropometric measures of body fat distribution (e.g. waist-to-hip ratio (WHR), subscapular-to-triceps skinfold ratio (STR), waist-to-thigh ratio) or computed tomography (CT) scan measures are associated with risk of diabetes, in both longitudinal (20-23) and cross-sectional (13-19,24,32,53) studies. The effects of fat distribution are generally independent of measures of overall fatness (13-23) and may be even more important.
Using a variety of markers of fat distribution, prevalence studies in Asian Indians (17,18), English Caucasians (18), Pima Indians (36), American Caucasians (24) and American Hispanics (24) have demonstrated that overall body fat is less closely associated with type 2 diabetes than is body fat distribution.
In addition, some studies indicated that the effect of fat distribution is greater in more obese individuals (16,21,53). Exceptions occurred in Mauritius, where there was no apparent effect of BMI on the prevalence of type 2 diabetes across tertiles of WHR (14), and in American men (16), where the effect of WHR was higher in the leaner individuals. In both these studies the general level of BMI was not extreme. In the Pima Indians, an extremely obese population, the effect of fat distribution was diminished with increased BMI or age (36). It may be that a threshold amount of body fat is required before the effects of fat distribution become apparent, and that after a certain level of obesity, the deposition of fat in peripheral depots diminishes the importance of central fat.
Longitudinal studies are less consistent. Among men of the Normative Aging Study who were followed over 18 years, fat distribution as measured by the ratio of abdominal circumference/hip breadth was a stronger predictor of both type 2 diabetes and impaired glucose tolerance than was BMI (23), but in prospective studies of Swedish men (20), and women (21), BMI and WHR were of similar importance.
The tendency for markers of fat distribution to be more strongly associated with prevalence of type 2 diabetes than is BMI in cross-sectional studies could be explained by decreases in BMI, but not WHR, associated with the onset of diabetes. Unpublished analyses from a prospective study in Mauritius show a fall in BMI but not WHR, over 5 years in people with diabetes.
Gender, Fat Distribution and Diabetes
Overall the literature suggests that both general adiposity and distribution of fat deposits are independently important risk factors for type 2 diabetes, in both men and women. However, some studies suggest that there are gender differences in the relative importance of overall fatness and fat distribution. In general, fat distribution appeared to be less important in men than women in comparison with overall fat measure (14-16,54), This observation led Haffner et al. (54) to propose a plateau effect of centrality, in this case measured STR, whereby above a certain level of STR, i.e. that achieved in most men, there was no further increase in rates of type 2 diabetes.
Fat deposition in men is generally abdominal; thus waist circumference or WHR will correlate more strongly with overall obesity, and the range of fat distribution may be limited, compared to women. These factors may explain statistically the lack of an independent association of prevalence of type 2 diabetes with fat distribution in men, rather than their higher degree of abdominal obesity as suggested above.
In summary, both overall obesity and fat distribution contribute to the risk of type 2 diabetes, but their relative importance may vary in relation to whether incidence or prevalence is considered, the gender of the individuals examined or their degree of obesity. However, strategies to reduce type 2 diabetes risk via diet and physical activity can reduce both overall and abdominal obesity and improvements in both should be sought. For management of obesity, waist circumference rather than WHR is probably a better measure of benefit, as weight loss and decrease in abdominal adipose tissue, can occur without changes in WHR.
There is evidence to suggest that in some ethnic groups the risk associated with a central distribution of body fat is relatively low. Among non-His-panics in Colorado the diabetes risks associated with BMI, triceps and subscapular skinfold thicknesses, family history and income were similar to those found in Hispanics. However, a 1 unit increase in either WHR or STR was associated with a greater increase in risk of type 2 diabetes among non-Hispanic whites than among Hispanics (28). Similarly, upper body obesity was more closely associated with higher levels of plasma insulin and glucose, and reduced insulin sensitivity, in obese Caucasian compared to African-American women (55). This metabolic study, along with the study of Marshall et al. (28), suggest that Caucasians may be more susceptible to the effects of fat distribution than some other ethnic groups.
There is clearly a genetic component to type 2 diabetes as indicated by twin studies, familial clustering of cases, and at the population level, by apparent ethnic differences in diabetes susceptibility with several genes assumed to contribute.
At the individual level, family history of type 2 diabetes can be used as an index of genetic predisposition to diabetes. A number of studies, examining different populations, have suggested that lean individuals with type 2 diabetes have a greater load of susceptibility genes. Thus individuals with a strong family history of type 2 diabetes do not need to accumulate large fat depots to achieve the same level of risk as those with less genetic susceptibility but a higher degree of obesity.
In Japanese men with a family history of type 2 diabetes it appeared that elevated prior or current obesity was not as strongly associated with type 2 diabetes as in men with no family history (56). Lemieux et al. (57) showed that among 'normal' glucose tolerant men with relatively high levels of insulin or glucose or both following an oral glucose tolerance test (OGTT), that is those at high risk of progression to diabetes, those with no family history of diabetes had a mean BMI at age 20 which was 2.6 kg/m2 higher than those with a family his tory of type 2 diabetes. Using a different approach, Kuzuya and Matsuda found that patients with type 2 diabetes who had a definite history of obesity had a significantly (P < 0.01) lower prevalence of family history of diabetes (32%) than those who had not been obese (50%) (58). Similarly, the siblings of lean diabetics tended to have a higher prevalence of type 2 diabetes than the siblings of obese diabetics in the study of Lee et al. (59), while Hanson et al. (60) observed that the odds ratio for type 2 diabetes in the offspring of obese diabetic parents was 0.6 compared with offspring of lean diabetic parents. Among elderly diabetic men in the Zutphen (Netherlands) study, there was no difference in the degree or duration of obesity in those with a family history of type 2 diabetes or those with no family history (61). This lack of interaction could be related to the older age of these men, 69-90 years. If, as reported by Kuzuya (58) and Lee (59), family history of diabetes is associated with earlier onset of diabetes, it is possible that men with a strong family history and young onset of type 2 diabetes will have already died, which could bias the results. Ohlson et al. (21) also did not find any statistically significant interaction between obesity and family history of diabetes in Swedish men.
Comparisons between different ethnic groups indicate residual differences in the prevalence of type 2 diabetes even after adjusting for BMI and other risk factors (28,62). Such differences may be attributed to a variety of factors, including increased genetic susceptibility, increased levels of other risk factors that were not considered, or the inability of anthro-pometric methods to accurately assess overall fat mass and distribution. Asian Indians, for example, appear to have an elevated risk of type 2 diabetes compared with members of other ethnic groups, even when BMI is at a similar or lower level (63-65). This is explained to some extent by differences in body fat distribution (66) which mean that Indian BMIs are not equivalent, regarding risk for diabetes, to BMIs of other ethnic groups. Fijian Indians had lower BMI than Melanesians, but their triceps skinfold thicknesses were greater, suggesting a higher body fat content in Indians (65). Indians also had a greater mean WHR than Europeans for the same level of BMI (18). In a comparison of young European and Polynesian women in New
Zealand it was found that a BMI of 30kg/m2 for a European was equivalent, in terms of body fat content, to a BMI of about 34 kg/m2 for a Polynesian (67); therefore the effect of similar BMI on risk of type 2 diabetes may differ between these ethnic groups. Swinburn et al. (68) previously reported that resistance measured by a bioelectrical impedance device was lower in Polynesians than Caucasians at any level of weight (adjusted for height and age), also implying a lower body fat content. On the other hand, a recent study of Chinese and Dutch adults found similar relationships between BMI and body fat in both populations (69), suggesting that differences between ethnic groups are not universal.
Diet, Obesity and Type 2 Diabetes
Diet is associated with risk of type 2 diabetes through its effect on obesity, and more specifically, dietary components such as fat, sugar and fibre have long been implicated in the development of type 2 diabetes (70). The importance of diet, independent of obesity, in at least some studies, indicates that among subjects with similar BMIs, the risk of type 2 diabetes could be modified by their usual diet.
In the San Luis Valley Diabetes Study, total fat intake was predictive of type 2 diabetes over 1-3 years of follow-up in 123 individuals with impaired glucose tolerance, independent of BMI (71). After adjusting for a number of dietary, anthropometric and metabolic risk factors the odds ratio for type 2 diabetes associated with a 40 g increment in total fat intake was around 7; however, saturated fat intake was not a predictor. In a cross-sectional analysis of baseline data from the same study the evidence was also in support of total fat being associated with type 2 diabetes (72).
Tsunehara et al. have published cross-sectional (73) and prospective (74) data on diet and type 2 diabetes in Japanese American men indicating an association with animal fat, which is consistent with the results of a study on Seventh Day Adventists showing relationships between meat consumption and diabetes (75), although these studies did not control for obesity. In the Finnish and Dutch cohorts of the Seven Countries Study, total fat, saturated fat, monounsaturated fat and cholesterol were higher 20 years before diagnosis in men with diabetes compared with men who were normal or had impaired glucose tolerance (76). After adjustment for past BMI and other factors, total fat was still significantly associated with 2-h post-load glucose level.
Beneficial effects offish on glucose tolerance have also been noted and it is believed that ro-3 fatty acids may be protective (77), although in one study this was only if accompanied by moderate physical activity (78). Thus the evidence for fat quantity and type being involved in the development of type 2 diabetes is expanding and is supported by studies on the effects of lipid composition of cell membranes on insulin sensitivity (79).
Two reports published in 1997, based on the long-running Health Professionals Follow-Up Study (80) and Nurses Health Study (81), concluded that diets with a high glycaemic load and low content of cereal fibre increased the risk of type 2 diabetes independently of BMI, in men and women respectively; the authors recommended that grains be consumed in a minimally refined form to reduce the risk of diabetes. The risk of developing type 2 diabetes for men and women in the top tertile of glycaemicload and the lowest for cereal fibre intake was twice that of those with low glycaemic load and high cereal fibre intake.
Nevertheless, a number of other studies have been unable to show any associations between diet and type 2 diabetes (82-85). However, it is clearly important to get some idea of dietary composition in order to evaluate the type 2 diabetes risk in individuals. The magnitude of the effect of diet on risk of type 2 diabetes is estimated as being of similar order as that of BMI; thus it should not be ignored.
Physical Activity, Obesity and Type 2 Diabetes
Obesity and physical activity have been found to be independently associated with both prevalence (14,86) and incidence (87-91) of type 2 diabetes in men and women. Physical activity may lower the risk of type 2 diabetes via reduced total body fat (86,87,89-93) and less abdominally distributed fat (86,89,92,93), and/or through its action in improving insulin sensitivity (92,93). These mechanisms are closely linked, but the independent effects of activity and obesity suggest that physical activity can modify the risk of type 2 diabetes at any given level of obesity.
In prospective studies where the effect of physical activity on type 2 diabetes risk has been shown to be independent of obesity, the benefits of physical activity were similar for lean and obese subjects among men and women in Malta (91), and women in the Nurses Health Study (87). In two studies of Caucasian American men, the effect of physical activity was strongest for more obese men and nonexistent in the leanest men (88,90), in contrast to the findings of the Honolulu Heart Study, where only the leaner men were protected by physical activity (89). These conflicting results cannot be explained by very different BMI levels in the different populations, but levels of other risk factors may be important. Gender may also modify the interaction between obesity and physical activity.
The relative risk of type 2 diabetes in less active individuals depends also on the levels of activity being compared, as well as many other factors. In the studies discussed above the relative risks ranged from 1.1 for the Pennsylvania Alumni (88), to 2.6 in Malta (91), of similar magnitude to the risks associated with obesity.
Fetal and Early Infant Nutrition, Obesity and Type 2 Diabetes
Hales and colleagues have proposed that poor in-trauterine nutrition, and perhaps poor nutrition in early infancy, reflected in low birthweights and decreased rates of postnatal growth, can increase the risk of type 2 diabetes and cardiovascular diseases in later life, through either reducing pancreatic cell capacity, or increasing insulin resistance (94,95). The effects of low birthweight on insulin resistance (96) or the insulin resistance syndrome (97) appear to be enhanced by adult obesity. Thus for adults with a ponderal index at birth of < 20.6 kg/m3, the degree of insulin resistance (t1/2 min) ranged from 18 to 31 at adult BMI increased, compared with a range of 14 to 19 for those with a ponderal index at birth of > 25 kg/m3 (96). Data in support of the so-called 'thrifty phenotype' hypothesis have been reported in Europeans (96-101), Indians (102), Mexican Americans (97), Australian Aborigines
(103) and Native Americans (104), although at least two studies in Europeans have failed to show the expected associations (105,106).
A recent review of studies into associations between fetal and/or infant growth and adult chronic disease concludes that the reported associations may be biased rather than causal, with possible selection bias due to loss to follow-up and confounding by socioeconomic factors (107). New evidence for a genetic explanation of the link between fetal growth and adult diabetes comes from Dunger et al. (108), with the observation that variation in the expression of the insulin gene is associated with size at birth.
Whatever the mechanism for the association of low birthweight with adult disease, and McCance et al. have suggested that the observations can be explained by more conventional genetic hypotheses
(104), the association is strong in a number of ethnically varied populations and hence its effects may contribute to differences in risk of type 2 diabetes seen in people with similar levels of adult obesity.
Leptin, Obesity and Type 2 Diabetes
While it is now accepted that human obesity is generally associated with elevated circulating leptin levels (109), and in many studies leptin is correlated with insulin levels (110-114) or insulin resistance (115,116), it is not clear whether leptin has a role in glucose intolerance in humans. In the leptin deficient ob/ob mouse treatment with leptin lowered glucose and insulin concentrations in the blood independently of weight loss (117,118). The leptin treated animals also increased physical activity and metabolic rate to 'normal' levels (117) and this may have enhanced insulin sensitivity and glucose uptake via increased glycogen mobilization as discussed above, or increased glucose uptake by some insulin-independent mechanism which would result in reduced insulin levels (118). Leptin may also influence insulin secretion in ob/ob mice via neuropeptide Y (NPY) (119) or by direct action on ^-cells (120,121).
In humans, in contrast to rats and mice, insulin does not have an acute effect on leptin levels (111,113,122), although chronic hyperinsulinaemia appears to be associated with elevated leptin levels (111,123,124), perhaps due to adipocyte hyper trophy (111). On the other hand, there is some evidence for leptin influencing insulin sensitivity. In isolated human liver cells leptin antagonizes insulin signalling (125), and in the Israeli sand rat (Psam-momys obesus), a polygenic animal model of type 2 diabetes, leptin has been reported to inhibit insulin binding to adipocyte insulin receptors (126).
A number of studies have reported on leptin levels in diabetic versus non-diabetic individuals; after adjusting for obesity there is no consistent picture with no difference being found in Polynesians in Western Samoa (110), Mexican Americans (127), Finnish men (123), American men and women (128) and German men and women (129) who mostly appear to have type 2 diabetes. Clement et al. (130) have found lower leptin levels in morbidly obese, poorly controlled diabetes compared with controlled diabetes or non-diabetics with similar levels of obesity. The low leptin levels in poorly controlled subjects may have been associated with lower insulin levels in this group.
Although the development of obesity in humans is unlikely to be linked to a defect in the OB gene as in the ob/ob mouse, in two related cases OB mutations in the homozygous form were reported to result in severe leptin deficiency and obesity (131). Recently, a rare mutation in exon 16 of OB-R was identified in humans. This alteration was found to result in morbid obesity and endocrine abnormalities in individuals homozygous for the mutation (132).
Leptin resistance, as observed in the db/db mouse and fa/fa rat which have a single mutation in the leptin receptor gene, has not been demonstrated yet in humans (133) and it may be that leptin resistance is due to a defect in body mass regulation downstream of leptin. It may also be that leptin is not important in preventing obesity in humans as it is in ob/ob mice, and merely reflects fat stores. In a small study of Pima Indians, low leptin levels predicted weight gain (134), but in a population-based study of Mauritians we were unable to find any association between high (leptin resistance) or low (leptin deficiency) leptin levels and weight gain over 5 years (135). Consistent with human evolutionary pressure, it has been suggested that leptin may have a more important role in protecting against the effects of undernutrition (136-138), especially in relation to reproduction (139-141), rather than preventing overnutrition. In this case it may have little relevance to type 2 diabetes, but more research is
► Direct relationships
Figure 24.1 The complex relationship between obesity and type 2 diabetes mellitus, illustrating the roles of other risk factors. IGT, impaired glucose tolerance
► Direct relationships
Figure 24.1 The complex relationship between obesity and type 2 diabetes mellitus, illustrating the roles of other risk factors. IGT, impaired glucose tolerance needed to determine whether leptin has a role in the development of diabetes.
Was this article helpful?