Interactions between race/ethnicity and anthropometry in risk of incident diabetes

Pamela L. Lutsey, Mark A. Pereira, Alain G. Bertoni, Namratha R. Kandula, David R. Jacobs

Research output: Contribution to journalArticlepeer-review

29 Scopus citations

Abstract

This study examined how adiposity influences racial/ethnic differences in diabetes incidence by exploring whether relations between anthropometric measures and incident diabetes vary by race/ethnicity. Data from the Multi-Ethnic Study of Atherosclerosis initiated in 2000 (n = 5,446 US men and women aged 45-84 years) were analyzed by using proportional hazards and Poisson regression. The diabetes incidence rate was 2/100 person-years (n = 479 cases). Interactions were present between race and anthropometry (P-interaction(race × body mass index) = 0.002). The slope of incident diabetes per anthropometric unit was greatest for Chinese, less for whites and Hispanics, and still less for blacks. For small waist, risk of incident diabetes was <1/100 person-years for all racial/ethnic groups. At intermediate waist levels, Chinese had the highest and whites the lowest rates of incident diabetes. At the respective 95th percentiles of waist circumference, risk of incident diabetes per 100 person-years was 3.9 for Chinese (104 cm), 3.5 for whites (121 cm), 5.0 for blacks (125 cm), and 5.3 for Hispanics (121 cm). Adiposity influenced relative diabetes occurrence across racial/ethnic groups, in that Chinese had a steeper diabetes risk per unit of adiposity. However, the generally low level of adiposity in Chinese led to a relatively low diabetes occurrence.

Original languageEnglish (US)
Pages (from-to)197-204
Number of pages8
JournalAmerican journal of epidemiology
Volume172
Issue number2
DOIs
StatePublished - Jul 2010

Keywords

  • Asian continental ancestry group
  • body mass index
  • diabetes mellitus
  • waist circumference

ASJC Scopus subject areas

  • Epidemiology

Fingerprint

Dive into the research topics of 'Interactions between race/ethnicity and anthropometry in risk of incident diabetes'. Together they form a unique fingerprint.

Cite this