Complicated contraception: Prescribing in women with obesity

In this new three-part mini-series, Dr Toni Hazell discusses some difficult scenarios when prescribing contraception, starting with the woman who has obesity

Which contraceptive methods should be avoided in women with obesity for reasons of safety?

In general, there are no safety concerns around intrauterine or progestogen-only contraception; the only exception is if obesity is one of multiple risk factors for cardiovascular disease (CVD), in which case the depot injection becomes a Faculty of Sexual and Reproductive Health (FSRH) UK Medical Eligibility Criteria (UKMEC) 3 (risks generally outweigh benefits). The progestogen only pill (POP) and sub-dermal implant (SDI) are a UKMEC 2 in this situation (benefits generally outweigh risks) and would be a better choice with multiple CVD risk factors.

For combined hormonal contraception (CHC), a BMI of 30-34 is UKMEC 2, and a BMI ≥35 is UKMEC 3. There is no BMI for which CHC is UKMEC category 4 (absolute contraindication). This restriction on CHC use (which includes the combined pill, patch and vaginal ring) is to do with the increased risk of venous thromboembolism; it is not thought that CHC increases the risk of arterial disease above background for those with obesity.

Which contraceptive methods might be less effective in a woman with obesity?

An intrauterine device (IUD) which has been successfully inserted will be just as effective regardless of weight. However, when the BMI is significantly raised, fitting can be technically challenging. A specialist gynaecology couch and specula in the appropriate sizes are helpful. If there is doubt as to whether a fitted IUD is correctly sited, an ultrasound should be arranged and the woman told not to rely on the IUD until after correct placement is confirmed.

The medroxyprogesterone depot injection is again effective at any weight if given correctly; the Depo-Provera preparation must be given intramuscularly, not subcutaneously. If there are concerns about the depth of subcutaneous fat causing inadvertent subcutaneous injection, a longer needle should be used, or the injection should be given in the deltoid instead of the usual gluteal region. An alternative would be to give Sayana-Press, formulated for subcutaneous use, which can be self-administered at home.

There is no reduced effectiveness of CHC when given by pill or vaginal ring at any weight, and no need to give two pills if weight is increased. There is limited evidence that women who weigh >90kg (regardless of height/BMI) might be more likely to have contraceptive failure with the combined patch. There is also some limited evidence of ovarian activity in the hormone-free interval (HFI) with obesity, leading to concerns about possible ovulation before the next pack is started; but no definite evidence of a higher failure rate. With increased use of tailored pill regimens, where the HFI is shortened to four days or eliminated, this will become less of a concern, and it is sensible to remind all women that tailored regimes are likely to give lower failure rates (see box 1).

The emergency contraception (EC) pill levonorgestrel may be less effective at a BMI ≥26 or a weight >70kg. A copper IUD should always be offered for EC; if refused, those with a weight or BMI above these parameters should be offered ulipristal, or a double dose of levonorgestrel.

Previous recommendations that those >70kg should have a double dose of the traditional POP (eg, norethisterone or levonorgestrel) have been dropped due to lack of evidence, and this has never been advised for the desogestrel or drospirenone POPs.

Which contraceptive methods cause weight gain?

Fear of weight gain is a common reason women refuse or discontinue hormonal contraception, but this is usually misguided – with the exception of the depot injection, there is no evidence for weight gain with any method. It is worth noting that most studies are done on a research cohort who are no more than 130% of ideal weight; our knowledge of further weight gain in those above this level before they use contraception is therefore more limited.

Those aged under 18 with a BMI ≥30 pre-depot are more likely to gain weight from using this form of contraception, with continued weight gain more likely in those who gain >5% of baseline body weight in the first six months of using the depot.

Are there any concerns about contraception after bariatric surgery?

Bariatric surgery is an umbrella term for a variety of operations of differing complexity. At its simplest, a band is placed around the stomach to reduce gastric size. More complex operations may involve removal of much of the stomach and re-routing of the digestive system to bypass the stomach and part of the small intestine. Women need effective contraception after bariatric surgery; pregnancy after such surgery has an increased risk of complications. A woman’s fertility may also improve with weight loss. For these reasons, the more effective long-acting reversible methods are recommended after bariatric surgery.

Concerns have been raised about poor absorption of oral contraceptives after bariatric surgery (including oral EC); this may be more likely in those with long-term diarrhoea after the surgery. Data is limited and of poor quality; to err on the side of caution, the FSRH suggests that women are advised of a potential loss of effectiveness and offered a non-oral method.

My practice is getting multiple letters daily about women who are taking GLP-1 agonists privately. Does this have any implications for their contraception? What about women taking orlistat?

Rapid weight loss for any reason can improve fertility, so conceptions could be expected to rise, however there are specific concerns related to tirzepatide, which reduced bioavailability of oral contraception in one study. This has not been shown to be clinically relevant with semaglutide or liraglutide. The FSRH advises that those using tirzepatide should switch to a non-oral method or add a barrier method for four weeks after initiation or dose increase.

It is suggested that women do not try to conceive until one month after stopping tirzepatide and two months after stopping semaglutide, due to a lack of safety data on use of GLP-1s in pregnancy.

Existing recommendations should be followed for those using oral contraception who have vomiting or severe diarrhoea as an adverse effect of a GLP-1 agonist, and all those taking these drugs should avoid conception whilst using them. The same applies for orlistat, which would only reduce the effectiveness of oral contraception if there is severe diarrhoea as an adverse effect.

Dr Toni Hazell is a GP in North London and a director at the Primary Care Women’s Health Forum