Impact of phthalate exposure and blood lipids on breast cancer risk: machine learning prediction
Abstract Background Phthalates exposure and its potential link to cancer are increasingly drawing public attention, which are found in products frequently used by women, including plastic food packaging and cosmetics. Given the lack of consensus from existing studies on the association of phthalate...
Saved in:
| Main Authors: | , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2025-03-01
|
| Series: | Environmental Sciences Europe |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12302-025-01071-3 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract Background Phthalates exposure and its potential link to cancer are increasingly drawing public attention, which are found in products frequently used by women, including plastic food packaging and cosmetics. Given the lack of consensus from existing studies on the association of phthalate exposure with breast cancer, conducting large-scale, well-designed epidemiological studies is crucial for clarifying this potential risk. Methods Utilizing data from the National Health and Nutrition Examination Survey (NHANES), this study assessed the correlation between exposure to phthalates and the risk of breast cancer. The analysis included ten phthalate compounds selected based on their prevalence and potential health impact. Multiple logistic regression was used to examine the correlation between phthalate exposure or other risk factors and breast cancer. Furthermore, machine learning-based predictive models were constructed to evaluate the significance of different variables. Results In the multivariate logistic regression analysis, four types of phthalates including MEP, DEHP, MEHHP, and MEOHP were identified as risk factors of breast cancer. In addition, MIBP, MINP, MEHP were also recognized as risk factors after adjusting for age. Conversely, MNBP and MCPP exhibited protective effects against breast cancer. Notably, MIBP demonstrated the most significant predictive power in machine learning models. The predictive model’s accuracy, as indicated by the area under the ROC curve, was 87.1%. Furthermore, survival analysis indicated that breast cancer patients with higher levels of phthalate exposure experienced significantly poorer survival outcomes than those with lower exposure levels. Intriguingly, subgroup analysis revealed a significant inverse association between phthalate exposure and breast cancer risk, particularly among individuals with elevated blood lipid levels. Conclusions The study revealed that exposure to specific phthalates was significantly associated with an elevated risk of breast cancer. Conversely, a higher concentration of blood lipids appeared to be negatively correlated with this risk. |
|---|---|
| ISSN: | 2190-4715 |