Epidemiology
About 1 in 5 children with new-onset diabetes has abnormal Lp(a), and levels can fluctuate within months, challenging the 'measure once' assumption (J Endocr Soc 2025)
Original title: Characterizing Lipoprotein(a) in Children With New-Onset Diabetes and Implications for Cardiovascular Risk Assessment
This study measured Lp(a) and HbA1c at baseline and again 3 months later in 76 children and adolescents aged 5-18 with incident diabetes mellitus (76% type 1, 23% type 2), testing whether Lp(a), often overlooked in paediatric diabetes screening, is stable as classically assumed. Baseline median Lp(a) was 43.3 nmol/L (Q1-Q3 13-73.7), with 17 of 76 children (roughly 20%) having elevated levels (75 nmol/L or above). Among the 22 children with follow-up measurements, 4 whose baseline Lp(a) was abnormal remained so, but 4 with normal baseline levels became abnormal at 3 months, and 3-month Lp(a) correlated positively with HbA1c (P = .004). The authors conclude that roughly 1 in 5 children with new-onset diabetes has abnormal Lp(a) warranting inclusion in cardiovascular risk stratification, and that the observed fluctuations argue for serial rather than one-time Lp(a) assessment in this population.
Original abstract
Context: Children with diabetes mellitus (DM) have an increased risk for cardiovascular disease (CVD), a risk potentially exacerbated by elevated lipoprotein(a) (Lp(a)). While other cholesterol parameters are screened in this population, Lp(a) is often overlooked despite being an independent CVD risk factor. Lp(a) levels are historically believed to not change over an individual's life and are genetically determined, but newer literature suggests variation.
Objective: This study investigated Lp(a) levels and their relationship with glycated hemoglobin A1c (HbA1c) in children with incident diabetes mellitus (DM).
Methods: Children and adolescents aged 5 to 18 years with incident DM had baseline Lp(a) and lipid profiles. Repeat Lp(a) and HbA1c were obtained 3 months later. Descriptive statistics (frequencies, proportions, means, medians) and nonparametric tests (Spearman correlation, Wilcoxon rank-sum/Kruskal-Wallis) were used. Statistical significance was set at P less than .05.
Results: Seventy-six children were included for evaluation: 76% with type 1% and 23% type 2 DM. Baseline median (Q1-Q3) Lp(a) was 43.3 nmol/L (13-73.7 nmol/L), 17 of which were elevated (≥75 nmol/L). Of the 22 participants with follow-up, 8 were abnormal: A total of 4 whose baseline Lp(a) were abnormal remained so and 4 with normal levels became abnormal. A positive correlation was found between 3-month Lp(a) values and HbA1c (P = .004).
Conclusion: Children with DM have abnormal Lp(a) levels at a prevalence of approximately 20%, so this should be considered in CVD risk stratification. Further, observed Lp(a) fluctuations suggest value in serial Lp(a) assessments due to nongenetic influences. Without Lp(a) quantification, CVD risk characterization in children with DM may be inaccurate and should be considered for a comprehensive assessment.
Summary written by lp-a.org from the published abstract; figures as published. Page updated 17 August 2026. Methods.