Association between vaginal bleeding in pregnancy that resulted in delivery and risk of cancer– a Danish registry-based cohort study

Highlighted results


Vaginal bleeding (VB) before 20 gestational weeks of a viable pregnancy is a manifestation of a threatened miscarriage. VB is associated with increased levels of pro-inflammatory cytokines such as interferon-gamma, tumour necrosis factor-alpha and interleukin-6. Increased levels of these cytokines and oxidative stress are risk factors for cancer. The risk of cancer following a VB-affected pregnancy that ended in childbirth is unknown. Objectives To investigate the associations between VB in pregnancy that resulted in delivery and risk of incident cancer. Methods - We conducted a cohort study (1995–2018) in Denmark using administrative and healthcare registries. We included 37,082 VB-affected deliveries, 1,363,614 VB-unaffected deliveries, 324,328 pregnancies ending in terminations, and 137,104 miscarriages. We computed the absolute risk of cancer and hazard ratios (HR) with 95% confidence intervals (CI) adjusted for age, calendar year, morbidities, and socio-economic factors using Cox proportional hazards regression. Multiple pregnancies to the same woman were accounted for in the analysis. Results The median (25th–75th percentile) follow-up in the study was 12.6 (6.9, 18.2) years. The prevalence of VB in the present study was 3%. At the end of the follow-up, there were 1320 cancer cases among the VB-affected delivery cohort, 40,420 among the VB-unaffected delivery cohort, 10,300 among the termination cohort and 4790 among the miscarriage cohort. HRs for any cancer among VB-affected deliveries were 1.03 (95% CI 0.97, 1.08) compared with VB-unaffected deliveries, 1.03 (95% CI 0.97, 1.09) compared with terminations and 0.90 (95% CI 0.84, 0.95) compared with miscarriages. There were no increased risks of premenopausal breast cancer, cervical cancer, ovary and fallopian tube cancer or uterine cancer following VB-affected deliveries vs. comparison cohorts. Conclusions We found no evidence of an association between vaginal bleeding in pregnancy and an increased risk of cancer.

In Paediatric and Perinatal Epidemiology
Elena Dudukina
Elena Dudukina

I am interested in women’s health, reproductive epidemiology, pharmacoepidemiology, causal inference, directed acyclic graphs, and R stats.