Investigating the Patterns of Abnormal Uterine Bleeding in Women Using Intramuscular Depot Medroxyprogesterone Acetate, a Copper Intrauterine Device or a Levonorgestrel Implant for Contraception

Abstract

Background: Access to contraception is a fundamental human right that significantly impacts women's health globally. In South Africa over 80% of women access contraception through the public sector and side effects, particularly abnormal uterine bleeding (AUB) remain a key reason for discontinuation. This subanalysis aimed to evaluate changes in self-reported menstrual bleeding patterns across three contraceptive methods, to inform tailored counselling and improve continuation rates. Methods: This subanalysis utilized data from the ECHO trial, which was a multicentre randomized open-label study comprising of HIV-negative women seeking contraception across nine South African sites. Participants in the ECHO trial were randomized to receive depot medroxyprogesterone acetate (DMPA-IM), copper intrauterine device (IUD), or levonorgestrel (LNG) implant. From the original dataset of 5743 participants, 3026 (53%) had usable data—defined as complete responses on bleeding-related questions at both first visit (1 month of method use) and later visit (six months of method use). Self-reported menstrual changes were collected from standardized questionnaires completed during the trial and assessed for four domains: bleeding frequency, duration, volume, and subjective acceptability. Demographic variables (age, BMI, smoking status) were included. Categorical analyses were conducted using Pearson’s chi-squared tests in R software; p-values <0.05 were considered significant. Results: Group distribution was as follows: copper IUD (n=1595, 53%), LNG implant (n=870, 28%), and DMPA-IM (n=561, 19%). Statistically significant changes in bleeding patterns were observed between first visit and later visit across all methods (p<0.001). Copper IUD users most frequently reported prolonged and heavy bleeding, while LNG implant and DMPA-IM users more often reported infrequent, short, and lighter bleeding. Obese participants had significantly higher probabilities of infrequent (up to 58%), prolonged, and light bleeding regardless of method. Smoking status was not significantly associated with any bleeding outcome. Among 323 participants who discontinued contraception, the most commonly cited reasons were heavy (29%) and prolonged bleeding (28%), followed by abnormal bleeding (26%) (χ2=80.17, df=4, p<0.001). Conclusion: This study confirms that contraceptive related menstrual changes vary by method and user characteristics. While these findings support existing knowledge, the large sample size and South African setting offer important contextual insight into method specific bleeding patterns and their role in discontinuation. Understanding these changes can help providers deliver more individualized counselling. However, improved counselling alone may not fully address discontinuation if broader sociocultural perceptions of menstrual changes are not also addressed.