Type 2 Diabetes in Asian-Indian Urban Children

RESEARCH DESIGN AND METHODS

Children with diabetes diagnosed ≤15 years who had features of type 2 diabetes (n = 18) are reported here. They were seen at the Diabetes Research Centre and M.V. Hospital for Diabetes, Madras, India. Among them, three cases were diagnosed within the last 3 months. The other 15 cases were diagnosed by us at least 1 year ago and patients had come for follow-up from January to June 2002. All of them satisfied the criteria given below.

1. Insidious onset of diabetes at age of ≤15 years

2. Response to treatment with ODAs

3. Presence of insulin secretory capacity comparable to type 2 diabetes in adults, as indicated by serum C-peptide concentrations (stimulated response ≥ 0.6 pmol/ml)

4. Lack of evidence of autoimmune process indicated by the absence of circulating GAD₆₅ antibodies.

Informed consent was obtained from the parents to use the data of the probands and the family members for scientific purposes. The study protocol was also approved by the ethics committee of the Diabetes Research Centre and M.V. Hospital for Diabetes.

Details of clinical history with emphasis on presence of symptoms were recorded. Blood pressure was measured, the details of dietary habits were noted, and waist-to-hip ratio was calculated. Anthropometric measurements included height, weight, and waist and hip circumferences. For each child, the percentage of body weight for the height was calculated using the table for normal height and weight for Indian children (17). Family history of diabetes was noted. We looked for skin manifestation of acanthosis nigricans. A total of 18 parents without known history of diabetes were invited to undergo a standard oral glucose tolerance test. Among them, diabetes was diagnosed in four fathers and five mothers. All children had measurements of fasting and 2-h postprandial plasma glucose (glucose oxidase-peroxidase method), HbA_1c (immunoturbidimetry method; Roche Diagnostics, Mannheim, Germany; normal value <6.1%), fasting total cholesterol, and triglycerides (enzymatic methods; Roche Diagnostics). A total cholesterol value of ≥200 mg/dl (5.2 mmol/l) and a triglyceride value of ≥150 mg/dl (1.7 mmol/l) were considered abnormal.

GAD₆₅ antibodies were measured in fasting serum by radioimmunoassay procedure using a commercial reagent kit that used I¹²⁵-labeled human recombinant GAD₆₅ (RSR, Cardiff, U.K.). The performance of the assay had been evaluated using S³⁵-labeled human recombinant GAD₆₅ as a reference method (18), and it was found to have a significant correlation (r = 0.93) with the standard assay (19). It had a sensitivity very similar to the standard assay, at specificities of 97.5 and 99%. The lowest titer detected was 0.02 units/ml. The intra- and interassay coefficient of variations were <10%. The cutoff value (<1.1 unit/ml) for the normal titer was determined earlier by measuring the antibody titer in 105 nondiabetic healthy individuals (20).

Serum C-peptide in response to a meal stimulus was estimated in all during a review. C-peptide measured at the time of diagnosis could be misleading due to glucotoxicity and/or β-cell suppression by exogenous insulin administration, if any (21). Commercial radioimmunoassay reagents (Diagnostic Products, Los Angeles, CA) were used. The intra-assay coefficient of variation was 5.5% and interassay coefficient of variation was <6.8%, and the lowest detection limit was 0.003 pmol/ml. Stimulated C-peptide value ≥0.6 pmol/ml indicated good β-cell reserve (22). Mean and SDs of the numerical variables are given.

RESULTS

Details of 18 children (5 males and 13 females), belonging to 17 families and identified as having type 2 diabetes, are reported. Among them, 15 had been treated with ODAs for durations ranging from 1 to 12 years; the remaining 3 recently diagnosed girls were being treated with ODAs for 2–3 months. All of the children showed good glycemic control with treatment. The clinical details are shown in Table 1. Monotherapy with metformin or sulfonylureas or a combination of both were prescribed. The age at diagnosis was 9–12 years in seven children, 13–14 years in six children, and 15 years in the other five subjects. All children had a positive family history of type 2 diabetes; parental history was present in 16, and the other 2 children had second-degree family history (grandparents with type 2 diabetes).

At diagnosis, nine children (50%) had ≥120% of ideal body weight and only three children were <100% of ideal body weight. The waist-to-hip ratio was high in 12 of 13 children who had their circumferences measured (normal values males = 0.88, females = 0.80) (Table 1).

At presentation, five children had polyuria and polydypsia, one had ketonuria, one girl had polycystic ovary syndrome, and three children had pruritis. Acanthosis nigricans was present in four children. Asymptomatic children (n = 9) were tested because of the strong family history of diabetes and/or because of obesity.

Before reporting to the center, one girl and one boy were being treated with insulin for >1 month and another three girls and one boy were treated with insulin for <10 days. All of them had symptoms suggestive of hypoglycemia, even with minimum doses of insulin, and responded to ODAs.

The biochemical parameters at the time of registration at the center are shown in Table 2. Isolated hypertriglyceridemia (n = 5), hypercholesterolemia (n = 3), and a combined hyperlipidemia (n = 2) were seen in some children. The C-peptide values ranged from 0.6 to 2.2 pmol/ml with a mean of 1.46 ± 0.4 pmol/ml. Table 2 also shows the fasting and 2-h postprandial plasma glucose values and HbA_1c values.

One 15-year-old girl (subject no. 9) had not attained menarche. One patient was hypertensive and was treated with enalpril. All other children had normal blood pressure. The diets of all study subjects were almost similar, with calories varying from 1,200 to 2,000. They were all consuming the regular Indian diet containing 60–65% of carbohydrate and 20–25% of fat. Physical activity was sedentary in 17 of 18 children; one girl (subject no. 15) previously played tennis regularly.

CONCLUSIONS

Type 2 diabetes in children has been reported from developed countries such as the U.K. (8–10), the U.S. (11–13), and Japan (23). Its rising occurrence has been attributed to the increasing rate of obesity in children (24). The susceptibility to the disorder was more common in ethnic minorities such as the Pima Indians in the U.S. (13,14) and Asian-Indians and Arabs in the U.K. (9). The highest prevalence of type 2 diabetes in children has been reported from Japan (23). Childhood obesity has increased enormously in parallel with rapid westernization of lifestyle in Japan. A high prevalence of glucose intolerance was observed in obese Japanese children. Impaired glucose tolerance/impaired fasting glucose and type 2 diabetes were found in 19.2 and 3.9%, respectively, of 280 children who were overweight by ≥30% of the standard weight. They had high insulin resistance that resulted in subsequent diabetes (23).

The present article highlights the fact that type 2 diabetes in children occurs even in developing countries. In concurrence with the reports from the developed countries (8–14,23,24), obesity, female sex, parental history of type 2 diabetes, and pubertal age appeared to be strongly associated with the disease in Asian children. Unlike in the children with type 1 diabetes who had acute onset of the disease with severe symptoms and ketonuria, lean body weight, and lack of familial aggregation, the type 2 diabetic children showed features similar to classic adult-onset type 2 diabetes. Obesity was present in only half of the probands seen by us. The profile is similar to that observed in adult type 2 diabetes in India—nearly half of type 2 diabetic patients are not obese. Insulin resistance is a common feature even in nonobese Asian-Indian subjects (25).

Clinical conditions associated with insulin resistance like acanthosis nigricans and polycystic ovary syndrome were common in the young type 2 diabetic patients. A similar profile had been reported by Drake et al. (8) and Ehtisham et al. (9) from the U.K. and also in the reports from the U.S. (14). The mean age at onset of the disease was peripubertal coinciding with the relative insulin resistance occurring during puberty. Strong familial inheritance, presence of obesity, nonautoimmune nature as indicated by the absence of GAD antibodies, and good response to ODAs favored the diagnosis of type 2 diabetes in our study group as well as in the reports from other ethnic groups (8–14,23).

The longest duration of diabetes in two probands was 12 years, and both were continuing to show good glycemic control with ODAs. In obese children with diabetes, metformin was found to be effective. Sulfonylureas were also effective in other children. The possible presence of the metabolic syndrome and the long duration of diabetes could favor development of vascular complications at a young age in these subjects.

It is difficult to have a clear cut demarcation of MODY and type 2 diabetes in the young. Although the prevalence of type 2 diabetes in children is not known at present, type 2 diabetes in children is an entity that needs to be recognized and looked for, especially in obese children of diabetic parents in India. Asymptomatic nature may delay the diagnosis in many as it usually does in adult type 2 diabetic subjects.