The morning-after pill works less well the later you take it — and at higher weight

Emergency contraceptive pills work, but not as reliably as their reputation as a foolproof backstop suggests, and two factors weaken them: how soon they are taken and the user's body weight. In randomised-trial data, the pregnancy rate after an emergency contraceptive pill taken within 72 hours was roughly 1-3% [s1]. In a pooled analysis, the risk of pregnancy despite treatment was more than three times higher in obese women than in those of normal weight — and higher still for the levonorgestrel pill specifically [s2].

That is the useful, uncomfortable detail behind the packaging: the pills reduce the chance of pregnancy but do not eliminate it, and the reduction is smaller for some women than others.

What the pills achieve

The clearest head-to-head evidence compared the two main oral options — ulipristal acetate and levonorgestrel — in a randomised trial with a meta-analysis [s1]. Among women treated within 72 hours of unprotected sex, there were 15 pregnancies in the ulipristal group (1.8%, 95% CI 1.0-3.0) and 22 in the levonorgestrel group (2.6%, 95% CI 1.7-3.9), a difference that favoured ulipristal but was not statistically decisive on its own (odds ratio 0.68, 95% CI 0.35-1.31) [s1].

Pooling the trials across the full 0-72 hour window, there were 22 pregnancies (1.4%) among 1,617 women given ulipristal and 35 (2.2%) among 1,625 given levonorgestrel [s1]. Both figures describe pills working — most treated women did not become pregnant — but neither is zero, and both are averages that hide meaningful variation.

Timing and weight change the odds

The follow-up analysis of the same trial data set out to find who remained at risk despite taking a pill, and it identified two clear factors [s2]. The first was body weight: "The risk of pregnancy was more than threefold greater for obese women compared with women with normal body mass index" (odds ratio 3.60) [s2]. The effect was not uniform across the two pills — for obese women, the risk was greater for those taking levonorgestrel (odds ratio 4.41) [s2].

The second factor was timing within the cycle: pregnancy risk was related to the cycle day of intercourse, with intercourse around the time of ovulation carrying substantially higher risk [s2]. The authors' practical conclusion pointed away from the pills for the highest-risk situations: women who have intercourse around ovulation "should ideally be offered a copper intrauterine device," and the analysis flagged concern for women with a body mass index above 25 kg/m² [s2]. The copper IUD, unlike the pills, is not known to lose effectiveness with weight.

How to read this without overreacting

The weight finding has limits and should not be overstated. It comes from a secondary analysis of trials not designed primarily to test the effect of weight, the number of pregnancies in any subgroup is small, and body weight is a crude proxy for the biology involved [s2]. What it establishes is a signal strong enough to have changed clinical advice — enough that heavier women are counselled that oral emergency contraception, and levonorgestrel in particular, may be less reliable for them, and that the copper IUD is the most effective emergency method regardless of weight [s2].

The timing point is more intuitive but just as important: emergency contraceptive pills mainly work by delaying or preventing ovulation, so their effectiveness falls when ovulation is already imminent or underway, which is precisely when the risk of conception is highest [s1] [s2]. That is why "sooner is better" is not merely marketing, and why a pill taken close to ovulation can fail even when used correctly.

The honest summary is that emergency contraceptive pills meaningfully lower the chance of pregnancy after unprotected sex but leave a real residual risk that is higher with delay, higher around ovulation, and higher at greater body weight — with the copper IUD the most effective option and the one unaffected by weight [s1] [s2].

This article is informational and is not medical advice. Anyone needing emergency contraception should seek advice promptly from a pharmacist or clinician, who can discuss which method is most appropriate.

Sources

• Ulipristal acetate versus levonorgestrel for emergency contraception: a randomised non-inferiority trial and meta-analysis — The Lancet, 2010-01-29
• Can we identify women at risk of pregnancy despite using emergency contraception? — Contraception, 2011-04-02