Repository logo
 

Native American ancestry and breast cancer risk in Colombian and Mexican women: ruling out potential confounding through ancestry-informative markers.

Published version
Peer-reviewed

Repository DOI


Change log

Authors

Zollner, Linda 
Torres, Diana 
Briceno, Ignacio 
Gilbert, Michael 
Torres-Mejía, Gabriela 

Abstract

BACKGROUND: Latin American and Hispanic women are less likely to develop breast cancer (BC) than women of European descent. Observational studies have found an inverse relationship between the individual proportion of Native American ancestry and BC risk. Here, we use ancestry-informative markers to rule out potential confounding of this relationship, estimating the confounder-free effect of Native American ancestry on BC risk. METHODS AND STUDY POPULATION: We used the informativeness for assignment measure to select robust instrumental variables for the individual proportion of Native American ancestry. We then conducted separate Mendelian randomization (MR) analyses based on 1401 Colombian women, most of them from the central Andean regions of Cundinamarca and Huila, and 1366 Mexican women from Mexico City, Monterrey and Veracruz, supplemented by sensitivity and stratified analyses. RESULTS: The proportion of Colombian Native American ancestry showed a putatively causal protective effect on BC risk (inverse variance-weighted odds ratio [OR] = 0.974 per 1% increase in ancestry proportion, 95% confidence interval [CI] 0.970-0.978, p = 3.1 × 10-40). The corresponding OR for Mexican Native American ancestry was 0.988 (95% CI 0.987-0.990, p = 1.4 × 10-44). Stratified analyses revealed a stronger association between Native American ancestry and familial BC (Colombian women: OR = 0.958, 95% CI 0.952-0.964; Mexican women: OR = 0.973, 95% CI 0.969-0.978), and stronger protective effects on oestrogen receptor (ER)-positive BC than on ER-negative and triple-negative BC. CONCLUSIONS: The present results point to an unconfounded protective effect of Native American ancestry on BC risk in both Colombian and Mexican women which appears to be stronger for familial and ER-positive BC. These findings provide a rationale for personalised prevention programmes that take genetic ancestry into account, as well as for future admixture mapping studies.

Description

Funder: Deutsches Krebsforschungszentrum (DKFZ) (1052)

Keywords

Ancestry-informative markers, Causal inference, Genetic admixture, Instrumental variables, Mendelian randomization, Female, Humans, American Indian or Alaska Native, Breast, Breast Neoplasms, Colombia, Mexico, Triple Negative Breast Neoplasms

Journal Title

Breast Cancer Res

Conference Name

Journal ISSN

1465-5411
1465-542X

Volume Title

25

Publisher

Springer Science and Business Media LLC
Sponsorship
National Cancer Institute (U19CA148065)
European Commission Horizon 2020 (H2020) Societal Challenges (634935)
European Commission Horizon 2020 (H2020) Societal Challenges (633784)
Cancer Research UK (16563)
Cancer Research UK (C1287/A8459)