Advanced

Long-term exposure to insulin and volumetric mammographic density : Observational and genetic associations in the Karma study

Borgquist, Signe LU ; Rosendahl, Ann H. LU ; Czene, Kamila; Bhoo-Pathy, Nirmala; Dorkhan, Mozhgan LU ; Hall, Per and Brand, Judith S. (2018) In Breast Cancer Research 20(1).
Abstract

Background: Long-term insulin exposure has been implicated in breast cancer etiology, but epidemiological evidence remains inconclusive. The aims of this study were to investigate the association of insulin therapy with mammographic density (MD) as an intermediate phenotype for breast cancer and to assess associations with long-term elevated circulating insulin levels using a genetic score comprising 18 insulin-associated variants. Methods: We used data from the KARolinska MAmmography (Karma) project, a Swedish mammography screening cohort. Insulin-treated patients with type 1 (T1D, n = 122) and type 2 (T2D, n = 237) diabetes were identified through linkage with the Prescribed Drug Register and age-matched to 1771 women without... (More)

Background: Long-term insulin exposure has been implicated in breast cancer etiology, but epidemiological evidence remains inconclusive. The aims of this study were to investigate the association of insulin therapy with mammographic density (MD) as an intermediate phenotype for breast cancer and to assess associations with long-term elevated circulating insulin levels using a genetic score comprising 18 insulin-associated variants. Methods: We used data from the KARolinska MAmmography (Karma) project, a Swedish mammography screening cohort. Insulin-treated patients with type 1 (T1D, n = 122) and type 2 (T2D, n = 237) diabetes were identified through linkage with the Prescribed Drug Register and age-matched to 1771 women without diabetes. We assessed associations with treatment duration and insulin glargine use, and we further examined MD differences using non-insulin-treated T2D patients as an active comparator. MD was measured using a fully automated volumetric method, and analyses were adjusted for multiple potential confounders. Associations with the insulin genetic score were assessed in 9437 study participants without diabetes. Results: Compared with age-matched women without diabetes, insulin-treated T1D patients had greater percent dense (8.7% vs. 11.4%) and absolute dense volumes (59.7 vs. 64.7 cm3), and a smaller absolute nondense volume (615 vs. 491 cm3). Similar associations were observed for insulin-treated T2D, and estimates were not materially different in analyses comparing insulin-treated T2D patients with T2D patients receiving noninsulin glucose-lowering medication. In both T1D and T2D, the magnitude of the association with the absolute dense volume was highest for long-term insulin therapy (≥ 5 years) and the long-acting insulin analog glargine. No consistent evidence of differential associations by insulin treatment duration or type was found for percent dense and absolute nondense volumes. Genetically predicted insulin levels were positively associated with percent dense and absolute dense volumes, but not with the absolute nondense volume (percentage difference [95% CI] per 1-SD increase in insulin genetic score = 0.8 [0.0; 1.6], 0.9 [0.1; 1.8], and 0.1 [- 0.8; 0.9], respectively). Conclusions: The consistency in direction of association for insulin treatment and the insulin genetic score with the absolute dense volume suggest a causal influence of long-term increased insulin exposure on mammographic dense breast tissue.

(Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Breast cancer, Diabetes, Insulin, Insulin genetic score, Mammographic density
in
Breast Cancer Research
volume
20
issue
1
publisher
BioMed Central
external identifiers
  • scopus:85054932395
ISSN
1465-5411
DOI
10.1186/s13058-018-1026-7
language
English
LU publication?
yes
id
b1d13494-9bc9-42de-ade2-f002031c136f
date added to LUP
2018-11-05 14:29:38
date last changed
2019-01-06 14:13:27
@article{b1d13494-9bc9-42de-ade2-f002031c136f,
  abstract     = {<p>Background: Long-term insulin exposure has been implicated in breast cancer etiology, but epidemiological evidence remains inconclusive. The aims of this study were to investigate the association of insulin therapy with mammographic density (MD) as an intermediate phenotype for breast cancer and to assess associations with long-term elevated circulating insulin levels using a genetic score comprising 18 insulin-associated variants. Methods: We used data from the KARolinska MAmmography (Karma) project, a Swedish mammography screening cohort. Insulin-treated patients with type 1 (T1D, n = 122) and type 2 (T2D, n = 237) diabetes were identified through linkage with the Prescribed Drug Register and age-matched to 1771 women without diabetes. We assessed associations with treatment duration and insulin glargine use, and we further examined MD differences using non-insulin-treated T2D patients as an active comparator. MD was measured using a fully automated volumetric method, and analyses were adjusted for multiple potential confounders. Associations with the insulin genetic score were assessed in 9437 study participants without diabetes. Results: Compared with age-matched women without diabetes, insulin-treated T1D patients had greater percent dense (8.7% vs. 11.4%) and absolute dense volumes (59.7 vs. 64.7 cm<sup>3</sup>), and a smaller absolute nondense volume (615 vs. 491 cm<sup>3</sup>). Similar associations were observed for insulin-treated T2D, and estimates were not materially different in analyses comparing insulin-treated T2D patients with T2D patients receiving noninsulin glucose-lowering medication. In both T1D and T2D, the magnitude of the association with the absolute dense volume was highest for long-term insulin therapy (≥ 5 years) and the long-acting insulin analog glargine. No consistent evidence of differential associations by insulin treatment duration or type was found for percent dense and absolute nondense volumes. Genetically predicted insulin levels were positively associated with percent dense and absolute dense volumes, but not with the absolute nondense volume (percentage difference [95% CI] per 1-SD increase in insulin genetic score = 0.8 [0.0; 1.6], 0.9 [0.1; 1.8], and 0.1 [- 0.8; 0.9], respectively). Conclusions: The consistency in direction of association for insulin treatment and the insulin genetic score with the absolute dense volume suggest a causal influence of long-term increased insulin exposure on mammographic dense breast tissue.</p>},
  articleno    = {93},
  author       = {Borgquist, Signe and Rosendahl, Ann H. and Czene, Kamila and Bhoo-Pathy, Nirmala and Dorkhan, Mozhgan and Hall, Per and Brand, Judith S.},
  issn         = {1465-5411},
  keyword      = {Breast cancer,Diabetes,Insulin,Insulin genetic score,Mammographic density},
  language     = {eng},
  month        = {08},
  number       = {1},
  publisher    = {BioMed Central},
  series       = {Breast Cancer Research},
  title        = {Long-term exposure to insulin and volumetric mammographic density : Observational and genetic associations in the Karma study},
  url          = {http://dx.doi.org/10.1186/s13058-018-1026-7},
  volume       = {20},
  year         = {2018},
}