Background: Insulin has been suspected to influence breast cancer risk because of its mitogenic effects and impact on sex hormone levels. Epidemiological studies investigating breast cancer occurrence in insulin-treated patients with diabetes have produced conflicting results, but were often underpowered and lacked adequate control for important confounders.

Purpose: To investigate the impact of insulin treatment on mammographic density (MD) as intermediate phenotype for breast cancer risk, and to further explore causation by analyzing associations with a polygenic risk score (PRS) incorporating 18 single nucleotide polymorphisms (SNPs) associated with fasting insulin levels.

Design: We conducted a matched cohort study within a Swedish screening-based cohort including insulin-treated type 1 (T1D, N=122) and type 2 (T2D, N=237) diabetes patients and up to 5 non-diabetic women matched on birth year (N=1,780). Associations between diabetes status and duration of insulin use with volumetric mammographic density were analyzed using general linear models adjusting for a comprehensive set of potential confounders. PRS analyses were performed in an independent sample of non-diabetic women (N=9,437) from the same cohort.

Results: In multivariable analyses, T1D patients had higher percent (11.2 vs. 8.8%; P<0.001) and absolute dense volumes (66.6 vs. 60.8 cm3; P=0.09) and a lower absolute non-dense volume (513.2 vs. 617.7 cm3; P<0.001) compared to age-matched non-diabetics. Percent- and absolute dense volumes increased with increasing T1D duration, while no such trend was observed for the absolute non-dense volume. Similar associations but of smaller magnitude, were found in insulin-treated T2D patients. Genetically predicted higher fasting insulin levels among non-diabetes women were associated with increased percent and absolute dense volumes (% change per sd increase in PRS = 0.8 (0.0-1.6) and 0.9 (0.1-1.8) respectively), but no difference in absolute non-dense volume.

Conclusions: Our results demonstrate higher mammographic density among insulin-treated diabetes patients, and genetic analyses support an effect of insulin on volumetric mammographic density. Further studies into how the observed MD differences translate into breast cancer risk are warranted.

Citation Format: Borgquist S, Rosendahl AH, Czene K, Bhoo-Pathy N, Dorkhan M, Hall P, Brand JS. Insulin and breast cancer risk: Novel insights from mammographic density analyses [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P2-07-03.