Clinical Question

Does use of hormonal contraception impact lactation and infant growth?

Evidence-Based Answer

Use of hormonal contraception does not appear to shorten breastfeeding duration or negatively impact infant growth, based on inconsistent evidence of moderate quality. It is unclear if hormonal contraception negatively impacts breast milk volume or composition. Overall, there was limited evidence regarding any particular hormonal contraceptive method. (Strength of Recommendation: B, based on inconsistent or limited-quality patient-oriented evidence.)

Practice Pointers

Contraception is an important health topic for breastfeeding women, because unintended pregnancy is associated with late entry into prenatal care, lower birth weight, and decreased breastfeeding rates. Similarly, a shortened interpregnancy interval is associated with low birth weight and preterm births.¹ This review sought to determine whether contraceptives adversely affect infant growth or breast milk supply and breastfeeding duration.

The authors identified 11 randomized trials with 1,482 women that examined the effects of combined oral contraceptives, the levonorgestrel-releasing intrauterine system (Mirena), progestin-only pills, or an etonogestrel-releasing implant vs. another method or placebo. Five studies were published before 1985, whereas six were published within the past 11 years. No studies examined the vaginal ring or transdermal patch. Inclusion and exclusion criteria varied across trials, although most included healthy women who delivered at 37 weeks' gestation or later and planned to breastfeed, had previous success with breastfeeding, or were actively breastfeeding. Follow-up ranged from 10 days to one year. The overall quality of the evidence was moderate, with several studies lacking information about randomization and allocation concealment; limited reporting of results; or many patients being lost to follow-up. Data could not be meta-analyzed because of differing protocols and outcomes.

Eight trials examined the effect of contraception on breastfeeding duration. One suggested a negative effect of combined oral contraceptives compared with placebo, but no adequate data were reported. Another trial showed fewer women were breastfeeding 75 days after insertion of the levonorgestrel-releasing intrauterine system compared with a nonhormonal intrauterine device (IUD; 56% vs. 79%; P < .05), but there was no significant difference between the groups at one year. Other head-to-head comparison trials showed no difference in breastfeeding duration between combined oral contraceptives and progestin-only pills, and between a levonorgestrel-releasing intrauterine system and nonhormonal IUD. In two trials that compared different insertion times for levonorgestrel-releasing intrauterine system and etonogestrel-releasing implant, there was no significant difference in breastfeeding duration between early postpartum insertion (within 48 hours in the levonorgestrel-releasing intrauterine system trial, and at one to three days postpartum in the etonogestrel-releasing implant trial) and standard insertion at four to eight weeks postpartum.

Milk volume or composition was assessed in six trials. One older trial reported decreased milk volume in participants using combined oral contraceptives compared with placebo, whereas another reported no difference. Neither trial provided data. Another trial (n = 171) conducted more than 30 years ago found diminished average milk volumes in patients using combined oral contraceptives vs. progestin-only pills after nine weeks (mean difference [MD] = −17.8 mL; 95% confidence interval [CI], −28.8 to −6.8), 16 weeks (MD = −24.0 mL; 95% CI, −34.5 to −13.5), and 24 weeks (MD = −24.9 mL; 95% CI, −36.0 to −13.8). Milk volumes decreased in both groups after study initiation, but average volumes decreased by a greater amount among combined oral contraceptive users compared with progestin-only pill users (42% vs. 12% reduction in volume, respectively) from weeks 6 to 24. Two studies showed no significant difference in milk volume or composition between progestin-only pills vs. placebo. Another trial showed no difference between etonogestrel-releasing implant insertion times.

Seven trials looked at infant growth. Six found no significant difference between progestin-only pills and placebo, combined oral contraceptives and progestin-only pills, or levonorgestrel-releasing intrauterine system and nonhormonal IUD. One study showed that use of an etonogestrel-releasing implant led to greater infant weight gain than no method at six weeks (MD = 426 g; 95% CI, 59 to 793), but less infant weight gain than depomedroxyprogesterone (Depo-Provera) from six to 12 weeks (MD = −271 g; 95% CI, −355 to −188).

The Centers for Disease Control and Prevention has medical eligibility criteria categories for contraceptives in breastfeeding women. Combined oral contraceptives, the combined hormonal patch, and ring are considered category 4 (i.e., unacceptable health risk) when the patient is less than 21 days postpartum because of the risk of venous thromboembolism. They are considered category 3 (i.e., theoretic or proven risks usually outweigh the advantages of the method) from 21 to 30 days, and category 2 (i.e., advantages generally outweigh the risks) after 30 days in women with no risk factors. Progestin-only pills, depomedroxyprogesterone, and the etonogestrel-releasing implant are category 2 during the first 30 days postpartum and category 1 (i.e., no restrictions for use) thereafter. The levonorgestrel-releasing intrauterine system is category 2 before four weeks postpartum, whereas the copper IUD is category 1 within 10 minutes of delivery of the placenta and category 2 up to four weeks postpartum. Both IUDs are considered category 1 after four weeks.¹