Objective: To investigate an association of vaginal bleeding-affected deliveries with the long-term risk of cancer as compared with vaginal bleeding-unaffected deliveries and pregnancies ending in a termination or miscarriage.
Design: Registry-based cohort study in Denmark, 1995-2017.
Setting: Danish health and administrative registries.
Participants: Deliveries (N=37,085) affected by vaginal bleeding (VB) within 20 gestational weeks among 35,517 women, VB-unaffected deliveries (N=1,362,760) among 783,020 women, pregnancies ending in a termination (N=324,395) among 239,729 women or miscarriage (N=137,040) among 121,303 women.
Main outcome measures: Incidence rates (IR) per 10,000 person-years and cumulative incidence of cancer at the end of up to 24 years of follow-up, hazard ratios (HR) with 95% confidence intervals (CIs) adjusted for age, calendar year, reproductive history, history of chronic conditions, medication use, and socioeconomic factors using Cox proportional hazards regression.
Results: We observed 1,725 cancer events (IR=32.1, 95% CI: 30.6-33.6) following VB-affected deliveries, 52,620 events (IR=31.5, 95% CI: 31.2-31.7) following VB-unaffected deliveries, 12,925 events (IR=30.1, 95% CI: 29.6-30.6) following a termination and 6,080 events (IR=34.3, 95% CI: 33.4-35.1) following a miscarriage. We found no association between VB and any cancer in comparison with VB-unaffected deliveries (HR=0.98, 95% CI: 0.93-1.03), terminations (HR=1.00, 95% CI: 0.94-1.06) and miscarriages (HR=1.04, 95% CI: 0.94-1.14). Specifically, there was no increase in relative risk of breast (HR=0.94, 95% CI: 0.86-1.03), cervical (0.94, 0.77-1.14), ovarian and fallopian tube (1.16, 0.81-1.66), uterine cancer (0.78, 0.46-1.33) and other site-specific cancers across all comparisons and in sensitivity analyses.
Conclusions: Having a VB-affected pregnancy ending in a delivery was not associated with an increased risk of cancer in women in comparison with having a VB-unaffected pregnancy ending in a delivery, termination or miscarriage.