Abstract
BACKGROUND: Anemia is a significant health concern for women in low- and middle-income countries. Hormonal contraceptives, which can reduce heavy menstrual bleeding and prevent pregnancy and its resulting blood loss, are an underexplored opportunity for anemia reduction. METHODS: Using Demographic and Health Survey data from 46 countries across four global regions, we examined country prevalence of anemia (i.e., hemoglobin below 12 g/dl via HemoCue testing) among women aged 15-49 years who were not pregnant. For each country, we used multivariable logistic regression to compare odds of anemia between women using hormonal contraceptive (users) to women using non-hormonal contraception or not using any contraception (nonusers). We merged data across countries to examine this effect stratified by contraceptive method and length of use and conducted a quantitative bias analysis of the anemia test. RESULTS: Anemia prevalence ranged from 12% in Rwanda to 63% in Gabon. In the merged sample, the crude odds ratio comparing anemia in users of hormonal contraception to nonusers was 0.57 (95% confidence interval: 0.56-0.58); the adjusted odds ratio was 0.55 (0.39-0.79) when controlling for age, education, wealth, and rurality. In the stratified analysis of merged data, we found that users of all hormonal contraception had lower adjusted odds of anemia compared to nonusers (AOR: 0.51, 95% CI: 0.46-0.57), and that the protective effect was strongest among users of contraceptive injections (AOR:0.37, 95% CI 0.30-0.45). In our bias analysis, the crude odds of anemia for users ranged from 0.29 to 0.57 when varying anemia test sensitivity and specificity between 75% and 100%. CONCLUSIONS: Future research should directly examine the causal link between hormonal contraceptive use and anemia as a high impact non-contraceptive benefit of hormonal contraception, especially with understudied methods such as injectable contraception and contraceptive implants.