The epidemic of obesity, and the difficulty in losing accumulated weight suggest that there may have been an advantage to this metabolic phenotype during human evolution. The development of type 2 diabetes in susceptible individuals would not have been a disadvantage in the absence of opportunities to become obese.
The "thrifty" genotype hypothesis was first advanced by JV Neel (24) nearly 40 years ago and recently updated (32,33). This hypothesis explains the insulin resistance and relative beta cell insufficiency associated with the development of type 2 diabetes as an adaptation to conserve energy in times of famine. Changes in gene frequency or in the genetic pool cannot explain the rapid increases in type 2 diabetes prevalence within one or two generations in some populations, emphasizing the importance of environmental factors operating on this genetic background (Fig. 2).
When it was noted that impaired glucose tolerance (IGT) or type 2 diabetes occurred in adults who had lower birth weight, smaller head circumference, and were thinner at birth, it was thought to indicate in utero programming that limited P-cell capacity and induced insulin
resistance in peripheral tissues. Maternal malnutrition was considered the cause of islet cell hypoplasia (34,35). Later study demonstrated that the glycemic response to insulin was also reduced in individuals who had been thin at birth (44). Large studies in Sweden and the US have confirmed the association of fetal undernutrition with later type 2 diabetes risk (36,37). The adult offspring of women who had starved during the last trimester of pregnancy during the Dutch famine at the end of World War II have also been found to have increased risk for IGT (38). Underweight for gestational age has been associated with increased cortisol axis activity in urbanized South African 20-year olds who were not obese. They also had IGT compared to normal birth weight controls (39).
Two studies in young subjects from high-risk populations support findings in older subjects on the effect of fetal nutrition on the risk for development of the insulin resistance syndrome (type 2 diabetes, hypertension, hyperlipidemia) in adulthood. The relationships between birthweight, present weight, fasting and post-load glucose and insulin concentrations were examined by multiple regression analysis in 3061 Pima Indians aged 5 to 29 years. Their current weight correlated with their birthweight. A U-shaped relationship was noted between two-hour glucose concentrations and birthweight in those over 10 years of age, unrelated to present weight. When adjustment was made for height and weight, negative correlations were found between birthweight and insulin concentrations at baseline and two hours, and insulin resistance in the 2272 subjects without diabetes. These observations supported the hypothesis that insulin resistance has a survival advantage for low birthweight babies (40). In a study of 477 8-year old Indian children, the cardiovascular risk factors of insulin resistance and plasma total and LDL cholesterol concentrations were strongly related to current weight. With adjustment for current weight, age, and sex, lower birthweight was associated with elevated systolic BP, fasting plasma insulin and 32-33 split proinsulin concentrations, glucose and insulin concentrations 30 minutes after glucose, and plasma lipids. Lower birthweight was also associated with increased calculated insulin resistance. Children who had low birthweight but high-fat mass at 8 years had the highest risk for insulin resistance syndrome variables and hyperlipidemia (41).
The thrifty phenotype hypothesis has been developed to explain how low birthweight, reflecting fetal undernutrition, is a risk factor for the later development of the insulin resistance syndrome. Poor nutrition in fetal and early infant life would restrict the development and function of the beta cells and insulin sensitive tissues, primarily muscle, leading to insulin resistance. Obesity in later life, with the attendant insulin resistance, would overcome the limited beta cell capacity, leading to type 2 diabetes. These findings could, however, be interpreted as a reflection of the thrifty genotype, that genetically determined defective insulin action in utero results in decreased fetal growth and obesity-induced IGT in later childhood or adulthood (42).
A number of studies comparing African-American and European American children suggest a genetic basis for the apparently greater susceptibility to type 2 diabetes in certain racial/ethnic groups. African Americans had greater insulin responses to oral glucose then European-Americans after adjustment for weight, age, ponderal (obesity) index, and pubertal stage, in a study of 377 children aged 5 to 17 years (43). In another study of nearly 1200 11- to 18-year olds, African Americans had higher insulin levels and lower glucose-to-insulin ratios than did European Americans, after correction for ponderal index, further indicating reduced insulin sensitivity in African-American youngsters (44). African-American prepubertal and pubertal youngsters have higher fasting and stimulated insulin concentrations during glucose clamp studies than do European-American youngsters (45). Rates of lipolysis have also been found to be significantly lower in African-American than in European-American children, further suggesting an energy conservation phenotype that would be detrimental with a surfeit of nutrition (46).
Prepubertal healthy children with a family history of type 2 diabetes (n=9) matched for age, pubertal status, total body adiposity determined by dual energy X-ray absorptiometry, abdominal obesity determined by computed tomography scan, and physical fitness measured by VO2maxwith those without such history (n=13) had three hour hyperinsulinemic clamp studies to assess insulin sensitivity. Those with a family history of type 2 diabetes had lower insulin stimulated glucose disposal and nonoxidative glucose disposal; there were no differences in glucose oxidation, fat oxidation, or FFA suppression (47). These data indicate that family history of type 2 diabetes is a risk factor for insulin resistance.
The familial clustering of type 2 diabetes can indicate environmental rather than genetic causation. In a study of physical, behavioral, and environmental characteristics of 42 parents and siblings in 11 families of adolescents with type 2 diabetes, 5 mothers and 4 fathers had diabetes before the study and it was diagnosed in 3 of the remaining fathers during the study. All 42 relatives had BMI > 85th percentile and skin fold measurements > 90th percentile. Fat intake was high and fiber intake low; physical activity was nil to low. Eating disorders were common and diabetes control poor (48).
Fetal beta cell function was assessed by amniotic fluid insulin (AFI) concentration at 32 to 38 weeks gestation in 88 pregnancies with pre-gestational or gestational diabetes, The offspring had oral glucose tolerance testing annually from 18 months of age. At < 5 years of age IGT was present in 1.2%, in 5.4% at 5 to 9 years, and in 19.3% at 10 to 16 years of age. There was no association between IGT and the type of maternal diabetes or macrosomia at birth. One-third of those with elevated AFI had IGT at adolescence in contrast to only one of 27 with normal AFI (49). Studies in the Pima Indian population have also indicated that the diabetic intrauterine environment is an important contributor to the risk of type 2 diabetes. The prevalence of diabetes in the offspring of Pima women with diabetes during pregnancy is significantly greater than in nondiabetic mothers or those who develop diabetes after delivery (50). These studies of the effect of diabetic pregnancy on altered p-cell function and glucoregulation later in life are of great concern because of the possible cumulative effect from generation to generation.
The mean age at diagnosis in all studies of type 2 diabetes in children, including the Florida series, is approximately 13.5 years, corresponding to the time of peak adolescent growth and development (9,51). Puberty is a time of relative insulin resistance, with normally a 2- to 3-fold increase in peak insulin response to oral glucose and for those with type 1 diabetes, substantial increase in insulin dose (52). Insulin mediated glucose disposal averages 30% less in adolescents compared to prepubertal children or to young adults (53). This physiologic insulin resistance of puberty is readily countered by increased insulin secretion in the absence of predisposition to type 2 diabetes and the additional stress of obesity. Increased activity of the GH-IGF axis is the likely cause of this physiologic insulin resistance of puberty, because it is transitory and coincident (30).
Approximately 55% of the variance in insulin sensitivity can be explained by total adiposity. Obese children have hyperinsulinism and 40% decrease in insulin stimulated glucose metabolism compared to the nonobese (53). There is a direct correlation between the amount of visceral fat in obese adolescents and basal and glucose stimulated insulinemia and an inverse correlation with insulin sensitivity. Body mass index increase results in decrease of insulin stimulated glucose metabolism and increase of fasting insulinemia. This inverse relationship between insulin sensitivity and abdominal fat is greater for visceral than abdominal subcutaneous fat (53).
Polycystic ovarian syndrome (PCOS) is being increasingly recognized in adolescents, often as part of the metabolic or insulin resistance syndrome. The syndrome includes, in addition to obesity and hyperinsulinism, hypertension, hyperuricemia, PCOS, acanthosis nigricans, dyslipidemia, and elevated plasminogen activator inhibitor-1 (54). Adolescents with PCOS have an approximate 40% reduction in insulin stimulated glucose disposal in comparison to body composition matched nonhyperandrogenic control subjects (55,56). Girls with premature adrenarche are at increased risk for ovarian hyperandrogenism and PCOS (57).
It is of considerable interest that children born small for gestational age are at increased risk for premature adrenarche, similar to the increased risk for insulin resistance from intrauterine undernutrition (58-60). This link between premature adrenarche and insulin resistance has been further explored by examining 60 first-degree adult relatives of girls with precocious adrenarche. Seven of the relatives (11.6%) had type 2 diabetes and another 14 (23.3%) had glucose intolerance, compared to the reported figures for the population of the same age of 2.5% and 7.5%, respectively. At least two abnormal lipid levels were found in 40% of subjects. Gestational diabetes was common and female relatives had lower steroid hormone binding globulin levels than did population controls (61).
Screening is testing applied to a group of individuals to separate those who are well from those who have an undiagnosed disease or defect, or who are at high risk. Considerations of testing for type 2 diabetes in children begin with the assumption that this will be done in obese youngsters. Thus, determination of obesity is the screening test. Case finding, the more appropriate designation than 'screening' for testing obese children for type 2 diabetes, is defined as diagnostic testing in a population at risk (62).
Case finding is justified if the condition tested for is sufficiently common to justify the investment and type 2 diabetes is sufficiently common in obese children and youth to justify testing such youngsters, especially those with high-risk ethnicity or family history. Another criterion for case finding is that the condition tested for be serious in terms of morbidity and mortality, which is unquestionably true of type 2 diabetes in children because of the association with increased cardiovascular risk factors of hypertension and dyslipidemia, hyperandrogenism/infertility, and early onset of microvascular disease. The condition tested for should have a prolonged latency period without symptoms, during which abnormality can be detected. Type 2 diabetes in children is often detected in the asymptomatic state, and albuminuria may already be present at the time of diagnosis, indicative of a prolonged latency (63).
Further requirements for case finding include the availability of a test that is sensitive (few false negatives) and accurate with acceptable specificity for the test (minimal number of false positives). The fasting plasma glucose (FPG) and two-hour plasma glucose (2HPG) have been applied to risk populations and are acceptably sensitive and specific, depending on criteria selected. There must also be an intervention able to prevent or delay disease onset or more effectively treat the condition detected in the latency phase (63). Intervention to reverse hyperglycemia and associated dyslipidemia, or to prevent the development of overt disease in those with IGT involves the daunting challenge of changing lifestyle in asymptomatic individuals, who are at an age when long-term health goals are not on their agenda.
A consensus panel of the American Diabetes Association recommended that individuals overweight as defined in Table 3, and with any two of the other risk factors indicated in the table should be tested every two years starting at age 10 or at the onset of puberty if that begins earlier (30). In the absence of data making definitive recommendations possible, the consensus panel considered it appropriate for the individual physician to test a specific child with any of the risk factors noted. Most instances of type 2 diabetes in children have occurred in the 10 to 19 year age group, although patients have been reported as young as five years. The FPG and the oral glucose tolerance test (FPG + 2HPG) were both considered suitable means of testing and the FPG was thought preferable by the ADA consensus panel because of lower cost and greater convenience (30). If one is testing for glucose intolerance in those at risk, however, the 2HPG will be elevated before the FPG. If necessary for convenience, PG can be measured in individuals who have taken food or drink shortly before testing. A random PG concentration > 140 mg/dL (7.8 mmol/L) is considered an indication for further testing, requiring FPG or 2HPG for confirmation on a different day (27).
The first U.S. population-based study outside the native North American population has recently been reported, involving ~1500 subjects without diabetes aged 12 to 19 years from the National Health and Nutrition Examination Survey of 1999 to 2002 (64). Applying the contemporary criteria for IFG, 11% were abnormal with 95% confidence intervals of 8% to 14%; as expected, there was a significant association between glucose levels and BMI, HbA1c, insulin, and C-peptide levels. This surprisingly high frequency of IFG in the random population not selected on the basis of risk factors might raise questions about the specificity of the IFG criterion. This criterion, however, was based on sophisticated epidemiologic analysis, as noted above. Recent studies have emphasized that FPG, similar to other biologic measures such as blood pressure and lipidemia, exists on a continuum from absolutely normal to absolutely abnormal, with predictive value for the eventual development of disease increasing as one moves toward the abnormal end of the spectrum. In a long-term follow-up of 13,000 Israeli Defense Forces recruits aged 26 to 55 years, those with normal FPG levels of < 100 mg/dL (5.6 mmol/L), but in the upper range of 91-99 mg/dL (5.1-5.5 mmol/L) were at much greater risk of developing type 2 diabetes during the years of follow-up than those with lower levels and this risk was heightened by greater BMI and serum triglyceride levels > 150 mg/dL (8.3 mmol/L) (65).
TABLE 3 Testing for Type 2 Diabetes in Children Criteria
Overweight (BMI 85th percentile for age and sex, weight for height > 85th percentile, or weight > 120% of ideal for height) PLUS any two of the following risk factors: Family history of type 2 diabetes in first or second degree relative Race/Ethnicity (American Indian, African American, Hispanic, Asian/Pacific Islander)
Signs of insulin resistance or conditions associated with insulin resistance (acanthosis nigricans, hypertension, dyslipidemia, PCOS)
Age of initiation: age 10 or at onset of puberty, if puberty occurs at a younger age Frequency : every two years Preferred test: fasting plasma glucose
Note: Clinical judgment should be used to test for diabetes in high-risk patients who do not meet these criteria.
More extensive screening programs for investigational purposes are needed. Such studies could establish the strength and risk level of various factors that might influence the development of type 2 diabetes (blood pressure, BMI, fat distribution (waist circumference, skinfold thickness), acanthosis nigricans, family history, race/ethnicity, socioeconomic status). They would also provide useful information about the testing tools, including FPG, 2HPG, random glucose, and HbA1c. These school based studies should be carried out in populations with sufficient numbers of high-risk youth and must be ongoing for several years, in order to track subjects with IGT, as well as those with risk factors who test normal, and to establish the predictability of various concentrations of PG and HbA1c (30).
Was this article helpful?