Comorbidity patterns and risk of breast cancer: a case–control study exploring interaction with BMI using latent class analysis

Abstract

Background Most studies to date have examined individual comorbidities in relation to breast cancer risk but have largely ignored that these conditions often cluster in the same person. Looking at comorbidity patterns rather than isolated diseases may therefore offer a better understanding of breast cancer risk. Methods In this case–control study, latent class analysis (LCA) was performed in the pooled case–control sample based on 10 comorbidities/conditions, including diabetes mellitus, hypertension, coronary heart disease, benign breast disease, breast lump, nipple discharge, history of gynecological tumors, severe lobular hyperplasia of the breast, papillomatous lesions of the breast, and dysfunctional uterine bleeding. Multivariable unconditional logistic regression was then used to examine the association between the identified latent classes and breast cancer risk and to test for potential interaction with body mass index (BMI). Results The LCA identified three comorbidity patterns in the pooled study sample: a low-comorbidity pattern, a metabolic comorbidity pattern, and a gynecologic–breast comorbidity pattern, accounting for 53.90%, 19.90%, and 26.20% of the study participants, respectively. Compared with the low-comorbidity pattern, both the metabolic comorbidity pattern (aOR = 2.55, 95% CI: 1.69–3.84) and the gynecologic–breast comorbidity pattern (aOR = 3.32, 95% CI: 2.25–4.89) were associated with higher breast cancer risk. Furthermore, a significant interaction was observed between BMI and the metabolic comorbidity pattern, suggesting that the association between this pattern and breast cancer risk varied across BMI categories. Conclusion These findings suggest that a comorbidity-pattern-based approach can improve the understanding of breast cancer risk. Identifying specific comorbidity or clinical conditions, particularly the metabolic and gynecologic–breast comorbidity patterns, could inform personalized risk stratification and targeted prevention strategies for breast cancer. While BMI modified the association between the metabolic comorbidity pattern and breast cancer risk, no evidence of BMI-related effect modification was observed for the gynecologic–breast comorbidity pattern.