Tanner (T) stage was assessed every six months and most participants underwent three study visits: at baseline (T2-T3) and when they reached T4 and T5.
Before each visit, participants ate a standard macronutrient diet and restricted exercise for three days. The visits included intravenous glucose tolerance testing, fasting laboratory studies and dual X-ray absorptiometry to measure percent fat mass.
The investigators calculated insulin sensitivity, acute insulin response to glucose (AIRg), and disposition index (DI). They used AIRg to estimate insulin secretion and DI as a measure for beta cell response in relation to insulin sensitivity. A decline in DI precedes T2D onset in both youth and adults.
The interventional arm was a double-blind randomized placebo-controlled trial of metformin in children with obesity and normal blood sugar levels at baseline. Investigators randomized these children to treatment with metformin or placebo starting at the baseline visit and continuing through the T5 pubertal stage. Youth in the clinical trial had a fourth study visit six months after stopping the intervention so investigators could check for any lasting effects of active treatment.
Research results: insulin and glucose dynamics affect who develops TD2 during puberty
HIP Aim 1: obese youth had lower insulin sensitivity and higher AIRg at every pubertal stage
Fifty-four participants completed the study. Insulin sensitivity and DI decreased and AIRg increased in both youth with normal weight and obesity as they advanced through puberty, with no significant group differences in the change over time. Obesity had a significant impact on insulin sensitivity and response at any given time point, however. Insulin sensitivity was almost 3-fold lower and AIRg 3-fold higher in early puberty in youth with obesity.
There were no group differences in DI at any time point or in DI change over time.
Despite the impact of obesity on insulin sensitivity, youth with obesity mostly maintained enough insulin secretion to compensate for insulin resistance. This is demonstrated by a similar DI compared to normal-weight youth despite the high prevalence of risk factors for youth-onset T2D (ie, family history, Hispanic ethnicity, female sex).
HIP AIM 2: effects of two years of metformin treatment
The study included 20 children with obesity who were randomized to metformin treatment and 24 youth who received placebo. The two groups were well matched for race/ethnicity and family history of diabetes and for baseline age, Tanner stage and BMI; 80% of participants were Hispanic.
Both groups had an overall decline over time in insulin sensitivity, AIRg and DI. There were no significant differences in changes in these factors between the metformin- and placebo-treated youth.
DI was higher in the metformin group at the T4 visit after a mean of 1.4±0.58 years of treatment but the benefit was absent at the T5 visit.
Children with the highest DI at baseline had the greatest drop in DI. Only two girls developed diabetes (one in the metformin group and one in the placebo group). Both were outliers at baseline, with a high DI at baseline and a rapid decline in DI during the study.
Metformin treatment did improve BMI, body fat percentage and waist circumference:
- BMI adjusted for children’s ages and gender (BMIz) decreased in the metformin group and increased in the placebo group; youth who took metformin retained the decrease 6 months after stopping treatment.
- The increase in waist circumference was much smaller in the metformin group versus the placebo group; this effect was still apparent, though reduced at 6 months post-treatment.
- Body fat percentage was lower in the metformin group, though this improvement was lost at 6 months post-treatment.
Improvements in body composition and weight status did not result in improvements in other health outcomes like cholesterol, blood pressure or blood glucose levels during the time of the treatment.
HIP Hba1C 1: puberty associated with rising HbA1c
Youth with obesity had higher BMIs and body fat percentage and lower insulin sensitivity and adiponectin. HbA1c increased significantly in normal weight youth at T4 and T5 compared with baseline, but not in youth with obesity.
HbA1c was higher in youth with obesity versus normal weight youth at T4 and T5. Although the overall HbA1c distribution was wider in youth with obesity, it overlapped between youth with normal weight and obesity.
The increase in HbA1c did not relate to DI. Study authors noted this was a surprising finding, particularly given that their earlier analysis found that DI decreases during puberty in both study groups. Lower adiponectin was significantly associated with higher HbA1c.
Research discussion and conclusions: challenges of early intervention and prevention
The HIP study found insulin resistance and insulin hypersecretion is already present in youth who are at an unhealthy weight at the earliest stage of puberty and before development of significant glucose abnormalities.
At any given time, insulin sensitivity was almost three-fold lower and AIRg three-fold higher in early puberty in youth with obesity versus normal weight youth. However, the trajectories of change in insulin sensitivity and beta cell function were unexpectedly similar between youth with normal weight and obesity.
Inclusion of known risk factors for diabetes, such as race/ethnicity, sex and percentage body fat in the predictive models failed to uncover predictors of change in DI during puberty, demonstrating the challenge of finding youth most in need of early intervention to prevent T2D. The HIP study did find that a high DI early in puberty may be a risk factor for a rapid decline in beta cell function.
Both youth with normal weight and obesity had a mild rise in HbA1c during puberty. If HbA1c, and by extension glycemia, increases in all youth during adolescence, then current criteria for prediabetes in adolescents based on adult norms may not be valid. Although HbA1c was, on average, higher in youth with obesity, the overall distribution overlapped between youth with normal weight and obesity.
Even when started early in puberty, before dysglycemia develops, metformin treatment did not prevent beta cell function decline during adolescence. Metformin treatment did confer some potential cardiometabolic benefits, including improvements in BMI, body fat percentage and waist circumference.
Additional studies are needed to better identify youth at highest risk for early onset diabetes, to improve understanding of glucose homeostasis mediators during puberty, and to define normal glycemia during adolescence. In addition, given the rapidly progressive nature of youth-onset T2D, further research is needed to develop preventive strategies that can be implemented before at-risk children begin puberty.