PREDICTIVE- a global, prospective observational study to evaluate insulin detemir treatment in types 1 and 2 diabetes: baseline characteristics and predictors of hypoglycaemia from the European cohort

Diabetes Obes Metab. 2007 May;9(3):428-34. doi: 10.1111/j.1463-1326.2006.00677.x.


Aim: PREDICTIVEtrade mark (Predictable Results and Experience in Diabetes through Intensification and Control to Target: An International Variability Evaluation) is a large, multi-national, observational study assessing the safety and efficacy of insulin detemir. We report the study design, population characteristics and baseline observations, including cross-sectional analysis, from 19 911 patients with type 1 or 2 diabetes.

Methods: Patients with type 1 or 2 diabetes requiring basal insulin are prescribed insulin detemir and followed up for 12-52 weeks. Data on demographics, haemoglobin A(1c) (HbA(1c)), fasting glucose, within-subject fasting glucose variability and weight are collected from patient records (and/or recall for hypoglycaemia). A negative binomial distribution model is used to assess the influence of predictive/confounding variables on hypoglycaemic episodes in insulin-treated patients at baseline. Multi-factorial analysis of covariance is used to evaluate the association of the variables with current body weight and within-subject fasting glucose variability.

Results: Total hypoglycaemic episodes in the 4 weeks prior to study start were 47.5 per patient-year in patients with type 1 and 9.2 per patient-year in patients with type 2 diabetes. The frequency of hypoglycaemia in insulin-treated patients showed a significant, positive association with duration of diabetes, number of insulin injections and fasting glucose variability but was inversely related to HbA(1c), fasting glucose and body mass index. Weight showed a significant positive association with gender (male > female) and insulin dosage. Weight was also positively associated with fasting glucose variability in patients with type 1 diabetes, and age and number of injections in patients with type 2 diabetes.

Conclusions: These baseline data showed that, in addition to the established relationship with intensive treatment and HbA(1c), frequency of hypoglycaemia was positively associated with fasting glucose variability. Follow-up data from PREDICTIVE will provide insights on insulin detemir in diabetes management.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Blood Glucose / analysis
  • Body Mass Index
  • Body Weight / physiology
  • Cross-Sectional Studies
  • Diabetes Mellitus / drug therapy*
  • Diabetes Mellitus, Type 1 / complications
  • Diabetes Mellitus, Type 1 / drug therapy
  • Diabetes Mellitus, Type 2 / complications
  • Diabetes Mellitus, Type 2 / drug therapy
  • Female
  • Glycated Hemoglobin A / analysis
  • Humans
  • Hypoglycemia / chemically induced*
  • Hypoglycemia / complications
  • Hypoglycemic Agents / adverse effects
  • Hypoglycemic Agents / therapeutic use*
  • Insulin / adverse effects
  • Insulin / agonists
  • Insulin / analogs & derivatives*
  • Insulin / therapeutic use
  • Insulin Detemir
  • Insulin, Long-Acting
  • Male
  • Middle Aged
  • Prospective Studies
  • Sex Factors
  • Treatment Outcome


  • Blood Glucose
  • Glycated Hemoglobin A
  • Hypoglycemic Agents
  • Insulin
  • Insulin, Long-Acting
  • Insulin Detemir