Duration of diabetes-related complications and mortality in type 1 diabetes: a national cohort study

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review


  • Lasse Bjerg
  • Soffia Gudbjörnsdottir, Swedish National Diabetes Register, University of Gothenburg, Sweden
  • Stefan Franzén, Swedish National Diabetes Register, Sweden
  • Bendix Carstensen, Steno Diabetes Center, Copenhagen, Denmark
  • Daniel R Witte
  • Marit E Jørgensen, Steno Diabetes Center, Copenhagen, University of Southern Denmark
  • ,
  • Ann-Marie Svensson, Swedish National Diabetes Register, University of Gothenburg, Sweden

BACKGROUND: People with type 1 diabetes often live for many years with different combinations of diabetes-related complications. We aimed to quantify how complication duration and total complication burden affect mortality, using data from national registers.

METHODS: This study included 33 396 individuals with type 1 diabetes, registered in the Swedish National Diabetes Register at any time between 2001 and 2012. Each individual was followed and classified according to their time-updated diabetes-related complication status. The main outcomes were all-cause mortality, cardiovascular (CV) mortality and non-CV mortality. Poisson models were used to estimate the rate of these outcomes as a function of the time-updated complication duration.

RESULTS: Overall, 1748 of the 33 396 individuals died during 198 872 person-years of follow-up. Overall, the time-updated all-cause mortality rate ratio (MRR) was 2.25 [95% confidence interval (CI): 1.99-2.54] for patients with diabetic kidney disease, 0.98 (0.82-1.18) for patients with retinopathy and 4.00 (3.56-4.50) for patients with cardiovascular disease relative to individuals without complications. The excess rate was highest in the first period after a diagnosis of CVD, with an 8-fold higher mortality rate, and stabilized after some 5 years. After diagnosis of diabetic kidney disease, we observed an increase in all-cause mortality with an MRR of around 2 compared with individuals without diabetic kidney disease, which stabilized after few years.

CONCLUSIONS: In this cohort we show that duration of diabetes-related complications is an important determinant of mortality in type 1 diabetes, for example the MRR associated with CVD is highest in the first period after diagnosis of CVD. A stronger focus on time-updated information and thorough consideration of complication duration may improve risk stratification in routine clinical practice.

Original languageEnglish
JournalInternational Journal of Epidemiology
Pages (from-to) 1250–1259
Publication statusPublished - Aug 2021

    Research areas

  • type 1 diabetes, epidemiology, diabetes-related complications, mortality

See relations at Aarhus University Citationformats

ID: 208355716