Type 2 diabetes (T2D) is a heritable disease leading to an abnormal glucose metabolism, influenced by multiple genetic variants across the genome. Although common in adults, T2D was previously rare in childhood, but its prevalence in this age group has been increasing in the past two decades as a result of the childhood obesity pandemic, accounting for up to 45% of cases of diabetes in youth in certain at-risk populations. The prevalence of prediabetes (an early stage of T2D, which is characterized by elevated blood sugar below the threshold to diagnose T2D) is even higher among obese children. Furthermore, individuals who develop both prediabetes and T2D in childhood and adolescence develop early microvascular complications, whose treatment failure is high. Thus, prevention of young-onset T2D using targeted lifestyle modification presents a particularly important yet difficult challenge. As such, identifying children at increased risk early in their lives is critical for these targeted interventions. Additionally, quantifying the genetic liability to youth-onset T2D could help better understand the respective contributions of environment vs genetics in the etiology of this disease. For instance, it is crucial to understand if among children with increased genetic risk for T2D, a favorable environment can prevent the development of T2D.
T2D has a polygenic nature, with an important heritable component explaining between 20% and 80% of the risk to develop this disease. Polygenic risk scores (PRS) have been demonstrated to have an improving ability to identify adult individuals at significantly high/low predisposition towards polygenic diseases. Therefore, it has become possible to identify adults who will lie at the extreme distribution of a trait, such as risk of T2D. T2D is causally linked to obesity. Obesity, as expressed by the measures of body mass index (BMI), is also a highly heritable polygenic trait. A PRS for adult BMI (Khera et al, Cell 2019) has been able to predict differences in body weight in ALSPAC children as early as at birth, with increasing predictive performance as individuals grow older.
Therefore, we posit that, similar to the PRS for BMI, a PRS for adult T2D, in combination with clinical risk factors (such as nutrition and physical activity) may be able to effectively predict individuals presenting abnormal glycemic traits in childhood. Since T2D and prediabetes in both adults and children is causally linked to obesity, we will also test if a PRS for adult BMI predicts abnormal glycemic traits in children. We will then compare the predictive performance of the PRSes and combine them with traditional non-genetic risk factors to enhance T2D risk prediction in youth. Finally, among children at the extremities of the PRS distribution (ie children with the highest and lowest genetic risk for T2D), we will seek to identify which environmental factors mitigate the effects of this genetic risk, and either contribute or prevent the development of T2D. To do this, we will use a large pediatric cohort representing the general population (ALSPAC), as well as a population of children at risk of obesity (QUALITY), both of predominantly European ancestry.