Answers are grounded in this paper and cite their sources.
To clarify the independent associations of abnormal vaginal discharge, family tumor history, menarche age, and first childbirth age with cervical cancer by comparing propensity score matching (PSM) and inverse probability treatment weighting (IPTW) for confounder control. In this retrospective case-control study, 125 cervical cancer patients and 375 healthy controls, all non-smoking women aged 35–64 years enrolled in the National Cervical and Breast Cancer Screening Program for Urban and Rural Women, were included. Propensity score matching (1:3 nearest-neighbor matching without replacement, caliper = 0.1) and inverse probability treatment weighting with stabilized weights were both employed for confounder control. Conditional logistic regression was used for the matched sample, and weighted logistic regression for the IPTW sample, with comparisons made to multivariate regression on the original unmatched data. After matching, 45 cases and 70 controls (115 participants in total) were retained. In the PSM analysis, abnormal vaginal discharge was excluded due to lack of variation in the matched sample, and none of the remaining variables showed significant associations. However, in the IPTW analysis which retained all 500 participants, abnormal vaginal discharge (aOR = 14.23, 95% CI 2.65–76.30), family history of tumors (aOR = 2.37, 95% CI 1.02–5.53), age at menarche (aOR = 0.83, 95% CI 0.71–0.96), and age at first childbirth (aOR = 0.90, 95% CI 0.82–0.98) remained significantly associated with cervical cancer after controlling for confounders. While PSM resulted in loss of statistical significance due to reduced sample size and elimination of exposure variation, IPTW analysis demonstrated that family history and reproductive factors maintained significant independent associations with cervical cancer after confounder control; the association for abnormal vaginal discharge should be interpreted cautiously, as it may partly reflect symptoms of existing disease. These findings require confirmation in prospective studies with detailed HPV data.
No takes yet. Share an insight, caveat, or question.
Bian et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: