The rise of diabetes in American youth was the focus of a session at the 2012 ADA annual meeting that was devoted to the Treatment Options for Type 2 Diabetes in Adolescents and Youth (TODAY) study. Philip S. Zeitler, MD, PhD, Section Head, Endocrinology, University of Colorado, Denver, presented updated data from new analyses released after the publication of the key results in April 2012.
The TODAY trial, which was sponsored by the National Institute of Diabetes and Digestive and Kidney Diseases, followed 699 young individuals who were between the ages of 10 and 17 years at study entry (mean age, 14 years).
The average duration of diabetes among study participants was 7.8 months; the average body mass index (BMI) was 34.9 kg/m2; and the average hemoglobin (Hb) A1c was 6.0% at baseline. Nearly 90% of the participants had a family history of diabetes; 65% of the participants were female, 41% Hispanic, 32% black, 20% white, and 6% were American Indian.
The 3 treatment groups were (1) metformin-alone, (2) metformin-plus-lifestyle, and (3) metformin-plus- rosiglitazone; participants were followed for 2 to 6 years. (Rosiglitazone is not approved by the US Food and Drug Administration for the treatment of type 2 diabetes in young patients.)
Nearly 46% met the primary end point of failure to maintain HbA1c <8% for 6 months.
Outcomes and Implications
Metformin monotherapy is inadequate in 50% of young patients with type 2 diabetes, said Dr Zeitler, based on the TODAY study. This percentage is greater than has been shown in studies of metformin in adults. The addition of rosiglitazone did not change the median time to failure. Lifestyle intervention increased weight reduction at 6 months but did not improve sustained glycemic control.
Early addition of a second agent to metformin may be required in this patient population, but additional studies are needed to determine the best agent for this purpose.
The role of intensive lifestyle intervention in young patients with type 2 diabetes is uncertain. Despite increased weight loss at 6 months in the metformin-plus-lifestyle group, glycemic control was not improved.
Dr Zeitler noted that treatment assignment did not affect long-term control of HbA1c, which was nearly the same in each arm in the almost 50% of young patients who maintained glycemic control over the course of the study. Furthermore, the onset of elevated HbA1c is quick and therefore cannot be predicted before HbA1c levels are first elevated.
Treatment of type 2 diabetes in youth may need to be individualized. There were “very intriguing race and gender differences that need further analysis, including adherence, socioeconomic status, and other predictors” of treatment failure, said Dr Zeitler. Slightly more males (48%) than females (44%) failed to maintain HbA1c control. In addition, failure rates were 53% in non-Hispanic blacks, 45% in Hispanics, 37% in non-Hispanic whites, and 40% in American Indians.
Furthermore, the response to treatment differed: the metformin-plus-rosiglitazone combination was more effective in controlling HbA1c in females, and greater improvement was seen in males with metformin plus a lifestyle intervention.
In non-Hispanic blacks, there was a “strikingly poor response to metformin,” said Dr Zeitler, with 50% losing glycemic control by 6 months. No treatment difference was found among Hispanics.
A snapshot of the comorbidities in the TODAY study raises concerns. Comorbidities were apparent early in the course of type 2 diabetes, within 2 to 6 years of the initial diagnosis, said Neil H. White, MD, CDE, Professor of Pediatrics, Washington University School of Medicine in St. Louis, MO.
Between 10% and 30% of participants had dyslipidemia (low-density lipoprotein cholesterol ≥130 mg/dL and triglycerides ≥150 mg/dL), and 34% had hypertension, which was defined as a blood pressure (BP) of ≥95th percentile, systolic BP ≥130 mm Hg, or diastolic BP ≥80 mm Hg. In addition, 17% had microalbuminuria (defined as an albumin-creatinine ratio ≥30 mg/g creatinine), a risk factor for nephropathy along with hypertension.
The predictive factors for hypertension were sex, age at baseline, and an increase in BMI over time, and for microalbuminuria, a predictive factor was an increase in HbA1c over time.
The only difference in adverse events between the treatment groups was hepatotoxicity, which was lower in the metformin-plus-rosiglitazone group (3.1 events per person-year) than in the metformin-only (8.1 events) and the metformin-plus-lifestyle (5.9 events) groups.
Fundus photography revealed a 14% rate of nonproliferative diabetic retinopathy, Dr White noted, after an average of 4 to 5 years of diabetes.
Increased levels of echocardiography measures were found as well: 85% to 90% for left ventricular (LV) mass and 75% for left atrial (LA) diameter. LV mass and LA diameter did not differ between the treatment arms.
“Overall, these findings show that these young people with type 2 diabetes have large hearts, suggestive of risk for cardiovascular disease,” said Dr White. “Glycemic control appears to matter in the development of left-ventricular hypertrophy.”
Factors Predicting Poor Glycemic Control
Increased obesity and HbA1c levels, as well as lower insulinogenic index (IGI) levels at baseline, were associated with failing to maintain glucose control for 6 months, according to an analysis of 2 subsets of patients, said Kenneth C. Copeland, MD, Associate Director, University of Oklahoma Diabetes Center. (IGI is an indicator of insulin secretion.)
The HbA1c rate increase (even within the normal range) was the only longitudinal factor associated with the primary end point.
The changes from baseline values over time were evaluated to determine their relationship to durable glycemic control in 172 participants who maintained glycemic control for at least 48 months and in 305 participants who lost glycemic control before that time point.
Treatment assignment was not significantly associated with maintaining or losing glycemic control. Race/ethnicity predicted durable glycemic control, said Dr Copeland. A lower failure rate was found in non-Hispanic whites (16%), and a higher failure rate was seen in non-Hispanic blacks (38%).
Failure rates for glycemic control decreased with increases in income and education. Only 20% of participants failed to maintain glycemic control in the >$50,000 category compared with 45% in the <$25,000 category and 35% in the $25,000 to $49,999 category.
A significantly higher HbA1c at baseline predicted failure. Participants with a 6.4% HbA1c at baseline failed compared with participants with an HbA1c of 5.7%, who maintained glycemic control. In addition, a significantly lower IGI (1.12 vs 2.04) predicted failure.
Factors that were associated with failure over time were a rise in HbA1c, a lower IGI at baseline, and a higher BMI. Baseline HbA1c, even within the nondiabetic range, was the best predictor of failure to maintain glycemic control for 6 months, regardless of the treatment arm. During treatment, a rapidly rising HbA1c, even within the normal range, is associated with failure and may suggest the need to intensify treatment early.
Combination treatment may be required for young patients with type 2 diabetes “to address the profound degree of insulin resistance observed in the study,” said Sonia Caprio, MD, Professor of Pediatrics (Endocrinology), Yale School of Medicine, New Haven, CT.
Beginning aggressive therapy very early may slow the decline in beta-cell adaptation and may achieve more durable glycemic control.
Dr Caprio presented results from an analysis of 674 participants in the TODAY study who had a slower rate of decline in beta-cell function in the metformin-plus-rosiglitazone group, which was relatively stable from 6 months to 24 months compared with the metformin-alone or the metformin-plus-lifestyle cohorts. This favorable change accounts for the lower rate of glycemic control loss observed in the metformin-plus-rosiglitazone group, she said.
Treatment with metformin plus rosiglitazone resulted in a 20% significant increase in insulin sensitivity in the first 6 months, but no change was found in the other 2 groups. However, insulin sensitivity decreased thereafter to approximately 10%, and no difference was seen between the groups at the end of the study.
The metformin-alone and metformin-plus-lifestyle groups had no significant effect on insulin sensitivity or beta-cell adaptation, both of which decreased progressively across 48 months.