Study on relationship between menopausal hormone therapy and breast cancer
Objective To evaluate the relationship between the prescription of menopausal hormone therapy and the inci-dence of women's breast cancer(BC).Methods A total of 214 patients with BC who received medical treatment in Puyang Oilfield General Hospital from March 2020 to December 2022 and 214 patients who underwent health examina-tion in the outpatient department during the same period were selected.1∶1 case control study was used.The BC patients were selected as the case group and the health examination population as the health control group.Compare and analyze the general information and MHT of the two groups of patients.Logistic regression analysis was used to explore the corre-lation between MHT prescription and the onset of BC in patients,and to construct an early warning model for BC onset.Results Logistic regression:BMI,hypertension,diabetes,BC family history,pregnancy times,eating fried food≥3 times/week,eating pickled products ≥ 3 times/week,smoking,synthetic progesterone,etc.were all significant influ-encing factors for BC(P＜0.05).Pregnancy was a protective factor(OR＜1).Early warning model for BC(Log P model):Log P=Ln(P/1-P)=-0.173+0.516 × BMI(kg/m2)+0.363 × hypertension+0.442 × diabetes+0.604 × BC family history-0.297 × pregnancy times+0.227 × edible fried food ≥ 3 times/week+0.518 × edi-ble pickled products≥ 3 times/week+0.368 × smoking+0.691 × synthesis of progesterone.The ROC analysis showed that the ROC-AUC(95％CI)of the aforementioned early warning model for the onset of BC(including the use of synthetic progesterone factors)was 0.863(0.791 to 0.911),indicating that its significant early warning effectiveness is high.Conclusion Synthetic progesterone is the only MHT prescription drug associated with the onset of BC.When prescribing menopausal HT,micronized progesterone may be a safer progesterone.The predictive model constructed u-sing logistic regression has a good predictive effect on the incidence of BC in women.
NETL
NSTL
万方数据
