Consequences for life of children with in utero exposure to metformin in Finland (CLUE) – a register-based cohort study

Study type

• Study topic:
  Disease /health condition

  Human medicinal product
• Study type:
  Non-interventional study
• Scope of the study:
  Assessment of risk minimisation measure implementation or effectiveness
• Data collection methods:
  Secondary use of data

Non-interventional study

• Non-interventional study design:
  Cohort

Study drug and medical condition

• Anatomical Therapeutic Chemical (ATC) code:
  (A10A) INSULINS AND ANALOGUES

  INSULINS AND ANALOGUES

  (A10BA02) metformin

  metformin
• Medical condition to be studied:
  Gestational diabetes

  Type 1 diabetes mellitus

  Type 2 diabetes mellitus

  Polycystic ovaries

Population studied

• Short description of the study population:
  The study population of the children will be assembled using the following criteria relative to the mothers.; Inclusion criteria:; 1. Singleton pregnancy resulting in live birth, recorded in the Medical Birth Register during the study inclusion period (1996-2016).; 2. Record of GDM during the pregnancy, defined as a diagnosis of GDM recorded in the Medical Birth Register, HILMO or AvoHILMO, or a pathological OGTT in the Medical Birth Register, or dispensation of metformin and/or insulin recorded in the Prescription register during the pregnancy, i.e. on the first day of the last menstrual period (LMP) or any time after it until the date of delivery.; 3. Age between 18 and 45 years at delivery, recorded in the Medical Birth Register.; 4. Registered in Finland throughout the pregnancy, based on the region of residency recorded in the Population Register Centre.; ; Exclusion criteria:; 1. Previously diagnosed T1DM recorded in the Medical Birth Register, or post-partum T1DM defined as at least one record of a diagnosis code for T1DM in HILMO or AvoHILMO registers after delivery.; 2. Dispensation of systemic glucocorticoids known to interfere with metformin or insulin recorded in the Prescription register during pregnancy, i.e. on the first day of the LMP or any time after it until the date of delivery.; 3. Dispensation of antidiabetic medications other than metformin or insulin (e.g. acarbose, thiazolidinediones, sulphonylureas, glinides, or glucagon-like peptide 1 (GLP-1) agonists, recorded in the Prescription register during pregnancy, i.e. on the first day of the LMP or any time after it until the date of delivery.
• Age groups:
  • Adolescents (12 to < 18 years)
  • Children (2 to < 12 years)
  • Infants and toddlers (28 days – 23 months)
  • Preterm newborn infants (0 – 27 days)
  • Term newborn infants (0 – 27 days)
  • Adults (18 to < 46 years)
  • Adults (46 to < 65 years)
• Special population of interest:
  Pregnant women
• Estimated number of subjects:
  64300

Study design details

• Main study objective:
  The main objective is to estimate longitudinally the prevalence, incidence and risk of diagnoses (obesity, hypoglycaemia, hyperglycaemia, hypertension, diabetes mellitus, PCOS (girls only), and diagnoses related to challenges in motor-social development) from the age of one week for as long as data are available in the four cohorts of interest.
• Outcomes:
  The primary outcomes are the long-term diagnoses collected from the age of one week for as long as data are available (maximum follow-up period 20 years). The secondary outcomes include the immediate effects at birth, up to one year of age and long-term growth-related effects (maximum follow-up period 16 years).
• Data analysis plan:
  For the primary objectives, the prevalence and incidence of children having a long-term diagnosis of interest will be estimated with 95% confidence interval separately within each study cohort and in the total population. The cumulative risk of permanent outcomes will also be characterised using the Kaplan-Meier estimator. In the analyses, three different Cox proportional hazards models will be used: unadjusted, adjusted, and adjusted using also propensity scores. The same three models will be used in logistic regression models to compare the yearly prevalence of the permanent long-term effects. For the secondary objectives, similar analyses will be conducted as for the primary objectives, considering the nature of the outcome and whether the outcomes are immediate at birth or long-term. For continuous outcomes, linear regression analyses will be conducted. In all regression analyses, the children with in utero exposure to insulin only will be considered as the reference group.