Diabetologia

, Volume 60, Issue 7, pp 1234–1243

Data-driven metabolic subtypes predict future adverse events in individuals with type 1 diabetes

  • Raija Lithovius
  • Iiro Toppila
  • Valma Harjutsalo
  • Carol Forsblom
  • Per-Henrik Groop
  • Ville-Petteri Mäkinen
  • on behalf of the FinnDiane Study Group
Article

DOI: 10.1007/s00125-017-4273-8

Cite this article as:
Lithovius, R., Toppila, I., Harjutsalo, V. et al. Diabetologia (2017) 60: 1234. doi:10.1007/s00125-017-4273-8

Abstract

Aims/hypothesis

Previously, we proposed that data-driven metabolic subtypes predict mortality in type 1 diabetes. Here, we analysed new clinical endpoints and revisited the subtypes after 7 years of additional follow-up.

Methods

Finnish individuals with type 1 diabetes (2059 men and 1924 women, insulin treatment before 35 years of age) were recruited by the national multicentre FinnDiane Study Group. The participants were assigned one of six metabolic subtypes according to a previously published self-organising map from 2008. Subtype-specific all-cause and cardiovascular mortality rates in the FinnDiane cohort were compared with registry data from the entire Finnish population. The rates of incident diabetic kidney disease and cardiovascular endpoints were estimated based on hospital records.

Results

The advanced kidney disease subtype was associated with the highest incidence of kidney disease progression (67.5% per decade, p < 0.001), ischaemic heart disease (26.4% per decade, p < 0.001) and all-cause mortality (41.5% per decade, p < 0.001). Across all subtypes, mortality rates were lower in women compared with men, but standardised mortality ratios (SMRs) were higher in women. SMRs were indistinguishable between the original study period (1994–2007) and the new period (2008–2014). The metabolic syndrome subtype predicted cardiovascular deaths (SMR 11.0 for men, SMR 23.4 for women, p < 0.001), and women with the high HDL-cholesterol subtype were also at high cardiovascular risk (SMR 16.3, p < 0.001). Men with the low-cholesterol or good glycaemic control subtype showed no excess mortality.

Conclusions/interpretation

Data-driven multivariable metabolic subtypes predicted the divergence of complication burden across multiple clinical endpoints simultaneously. In particular, men with the metabolic syndrome and women with high HDL-cholesterol should be recognised as important subgroups in interventional studies and public health guidelines on type 1 diabetes.

Keywords

All-cause mortality Cardiovascular mortality Data-driven model Diabetic kidney disease Ischaemic heart disease Metabolic subtypes Self-organising map Sex difference 

Abbreviations

ESRD

End-stage renal disease

FinnDiane

Finnish Diabetic Nephropathy Study

IQR

Interquartile range

SMR

Standardised mortality ratio

SOM

Self-organising map

Supplementary material

125_2017_4273_MOESM1_ESM.pdf (742 kb)
ESM(PDF 742 kb)

Copyright information

© Springer-Verlag Berlin Heidelberg 2017

Authors and Affiliations

  • Raija Lithovius
    • 1
    • 2
    • 3
  • Iiro Toppila
    • 1
    • 2
    • 3
  • Valma Harjutsalo
    • 1
    • 2
    • 3
    • 4
  • Carol Forsblom
    • 1
    • 2
    • 3
  • Per-Henrik Groop
    • 1
    • 2
    • 3
    • 5
  • Ville-Petteri Mäkinen
    • 6
    • 7
    • 8
  • on behalf of the FinnDiane Study Group
  1. 1.Folkhälsan Institute of Genetics, Folkhälsan Research Center, Biomedicum HelsinkiUniversity of HelsinkiHelsinkiFinland
  2. 2.Abdominal Center NephrologyUniversity of Helsinki and Helsinki University HospitalHelsinkiFinland
  3. 3.Research Programs Unit, Diabetes and ObesityUniversity of HelsinkiHelsinkiFinland
  4. 4.National Institute for Health and Welfare, Chronic Disease Prevention UnitHelsinkiFinland
  5. 5.The Baker IDI Heart and Diabetes InstituteMelbourneAustralia
  6. 6.South Australian Health and Medical Research Institute, SAHMRI North TerraceAdelaideAustralia
  7. 7.School of Biological SciencesUniversity of AdelaideAdelaideAustralia
  8. 8.Computational Medicine, Faculty of MedicineUniversity of Oulu and BiocenterOuluFinland

Personalised recommendations