Why does ethnicity affect prevalence of gestational diabetes? The underwater volcano theory

Diabet Med. 1996 Aug;13(8):748-52. doi: 10.1002/(SICI)1096-9136(199608)13:8<748::AID-DIA164>3.0.CO;2-I.


To study why gestational diabetes (GDM) is more common in some ethnic groups than others, we tested the hypothesis that GDM is more common in people who are temporally closer to developing non-insulin-dependent (Type 2) diabetes mellitus (NIDDM). The prevalence of GDM and the mean age of affected women in each major ethnic group were determined. From our database of NIDDM 6052 patients, the mean age of onset in each ethnic gorup was calculated and the mean difference between age of developing GDM and age of developing NIDDM derived (NIDDM-GDM age gap). This age gap was used to adjust for the susceptibility to GDM of each group. The overall prevalence of GDM was 6.7% (CI 6.0%-7.4%). In Anglo-Celtic women it was 3.0% (CI 2.3%-3.7%). For the other ethnic groups the prevalence and odds ratio (OR) were: Chinese (15.0% CI 11.8%-18.2% OR 5.6), Vietnamese (9.6% CI 6.6%-12.5% OR 3.6), Indian (16.7% CI 9.8%-23.5% OR 6.4), Arabic (7.3% CI 4.6%-10.1% OR 2.5) and Aborigines (10.1% CI 3.8%-16.4% OR 3.7). The OR for susceptibility to GDM did not change after adjustment for BMI and maternal age and it correlated significantly with the NIDDM-GDM age gap (r = -0.85; p = 0.03). However, it fell substantially after adjustment for NIDDM-GDM age gap. For women of different ethnic origins there is a difference in the time gap between their pregnancies and the time at which they would on average be expected to develop diabetes. This difference may be an important factor underlying the higher prevalence of GDM in some ethnic populations.

Publication types

  • Comparative Study

MeSH terms

  • Age of Onset
  • Body Mass Index
  • Cohort Studies
  • Diabetes Mellitus, Type 2 / epidemiology*
  • Diabetes, Gestational / epidemiology*
  • Ethnic Groups / statistics & numerical data*
  • Female
  • Humans
  • Maternal Age
  • Odds Ratio
  • Pregnancy
  • Prevalence