5.1.1 IUC effectiveness

The mechanisms of action of IUC are based on local effects and do not rely on systemic drug levels; therefore, a woman’s weight would not be expected to affect contraceptive effectiveness of the Cu-IUD or LNG-IUS.56–58

There is no evidence of impaired contraceptive effectiveness in IUC users with obesity, either with the Cu-IUD or the LNG-IUS. A prospective cohort study reported no statistically significant difference in contraceptive failure rate during the first 2 to 3 years of use among IUC users (Cu-IUD or 52 mg LNG-IUS) who were of normal BMI (n=1584), overweight BMI (n=1149) or obese BMI (n=1467).59 The overall IUC failure rate of less than one pregnancy per 100 woman-years did not vary by BMI; five women of normal weight, no women who were overweight, and seven women with obesity became pregnant. The initial efficacy trial of a lower dose 13.5 mg LNG-IUS included women weighing 38–155 kg (mean 68.7 kg) with BMIs from 16 to 55 kg/m² (mean 25.3 kg/m²) and did not report lower efficacy for obese study participants.60 The conclusions that can be drawn from these studies are limited by the low failure rate of IUC; there are too few pregnancies to determine statistically significant differences between weight groups.

5.1.2 IUC safety

Cu-IUD

No studies have specifically evaluated the safety of the Cu-IUD in women with raised BMI. There are no theoretical reasons why the Cu-IUD would pose health risks to women with raised BMI. According to the UKMEC, obesity does not restrict the use of the Cu-IUD (UKMEC 1).1

LNG-IUS

According to the UKMEC, obesity alone does not restrict the use of the LNG-IUS (UKMEC 1).1 Even when obesity is in the context of other risk factors for CVD (e.g. smoking, diabetes and hypertension), use of the LNG-IUS is UKMEC 2.

No studies have directly assessed whether women with raised BMI who use LNG-IUS are at increased risk of VTE and other adverse cardiovascular outcomes, as compared to their normal-weight counterparts. Data presented here are therefore not specific to women with raised BMI.

Those few studies that have looked at different progestogen-only contraceptives generally suggest that there is little or no increased risk of VTE,61–63 no statistically significant increased risk of MI64 and no statistically significant increased risk of stroke65 associated with use of POP, implants or the LNG-IUS. Data, however, are not available in relation to BMI or body weight. A Danish national registry-based cohort study61 sought to assess the incidence of VTE in current users of non-oral hormonal contraception; the study included all Danish non-pregnant women aged 15–49 years (n=1 626 158), with no history of thrombosis, and followed them from 2001 to 2010. The risk of confirmed VTE was not increased with use of hormone-releasing intrauterine devices (IUDs) (adjusted relative risk (aRR) 0.57; 95% CI 0.41–0.81). A subanalysis of a case-control study on risk factors for venous thrombosis among women aged 18–50 years, which included 29 LNG-IUS users (3 cases and 26 controls), found that use of an LNG-IUS was not associated with an increased risk of VTE compared to non-use of hormonal contraception (odds ratio (OR) 0.3; 95% CI 0.1–1.1).63 A meta-analysis of eight observational studies also showed no association between VTE risk and use of an LNG-IUS (aRR 0.61; 95% CI 0.24–1.53).62

Very little research has been conducted to assess the effect of progestogen-only contraception on cardiovascular risk factors (e.g. lipid and carbohydrate metabolism) in women with raised BMI or in relation to BMI/weight. Studies looking at this in healthy, non-obese progestogen-only users or progestogen-only users of normal weight have been reassuring with no clinically meaningful changes observed.53–61

5.1.3 Weight gain with IUC

There is no specific evidence relating to weight gain with IUC use by women with raised BMI. In the general population, there are no significant differences in weight gain when hormonal and non-hormonal intrauterine methods are compared and no evidence to support a causal association between IUC use and weight gain.

5.1.4 Health benefits of IUC

Obesity is associated with increased risk of endometrial hyperplasia and cancer.74 Although not directly studied in women with obesity, studies of women in the general population suggest that use of the LNG-IUS or Cu-IUD is associated with reduced risk of endometrial hyperplasia and cancer.75–79 The mechanism by which the Cu-IUD could reduce endometrial cancer risk has not been defined.

5.1.5 Practical considerations with IUC

IUC insertion for women with raised BMI is appropriate, safe and feasible. IUC is a highly effective and safe contraceptive option for women who are overweight and women with obesity. In practice, IUC insertion may be more challenging in women with obesity than in normal-weight women in terms of assessment of uterine position and gaining access to the uterus;80 however, raised BMI is not a significant factor in failed IUC insertions or expulsions,57 and insertion difficulties should not be presumed in women with raised BMI. Some practicalities may need to be considered in order to maximise the chances of insertion success. For example, having a supportive gynaecology couch and a range of speculum sizes is important. In addition, availability of a large blood pressure cuff for measuring blood pressure is necessary.

5.2.1 Implant effectiveness

The implant is a highly effective method of contraception, and true contraceptive failures are very rare (see Table 2). However, the manufacturer-sponsored clinical trials which determined this excluded women who were >130% of their ideal body weight, meaning that no or few obese women were included.82 Body composition could affect implant metabolism because of the increased volume of distribution, effects on plasma protein binding and altered plasma clearance in individuals with obesity.83

Though no studies have been specifically designed to answer the question of how obesity may impact implant effectiveness, currently available evidence suggests that both the ENG and levonorgestrel (LNG) contraceptive implants do remain highly effective in women with raised BMI. However, sufficient data on women with BMI ≥40 kg/m² are still lacking.

Taken together, recent pharmacokinetic and clinical outcomes data are reassuring that the implant is a highly effective contraceptive option for women of all BMIs. Available evidence is detailed below.

Pharmacokinetic studies

Some pharmacokinetic studies have shown an inverse relationship between body weight and ENG serum levels,84,85 raising concerns about the effectiveness of the method in women with raised BMI, and the duration for which the method is effective in women with high BMI.

In one small pharmacokinetic study85 of the ENG implant in the first 6 months of use, consistently lower serum concentrations of ENG were observed amongst those women with a BMI >30 kg/m² (n=13), compared with those with a BMI <25 kg/m² (n=4), although the difference was not statistically significant. In the group with a BMI >30 kg/m² (weight range 90–164 kg), projected serum levels at 1, 2 and 3 years of implant use were estimated to be 133,102 and 98 pg/mL, respectively.

In a pharmacokinetic study comparing 52 women (10 normal weight, 19 overweight, 23 with obesity) using the implant for at least 1 year,86 ENG levels were comparable across BMI categories and no participant fell below the suggested threshold for ovulation of 90 pg/mL.86

It is important to note that the absolute ENG serum level required for suppression of ovulation has not been defined and may be lower than the usually suggested 90 pg/mL.87

A recent study of 237 implant users (where participant BMIs ranged from 16.6 to 53.2 kg/m²; 29% (n=27) normal weight, 25% (n=23) overweight, 46% (n=42) obese or morbidly obese) showed no difference in median ENG levels between women of different BMI categories at 3 and 4 years of use.88 However, at those longer durations of use (3 and 4 years) some ENG levels among women who were overweight or who had obesity fell below 90 pg/mL (no data on how many women that included), although no pregnancies occurred. This cohort was also evaluated up to 5 years of implant use. A statistical difference in median ENG levels was noted at the end of 4 years of use, with women who were overweight having the highest serum ENG (195.9 (range 25.0–450.5) pg/mL) when compared to normal-weight women (178.9 (range 87.0–463.7) pg/mL) and women with obesity (137.9 (range 66.0–470.5) pg/mL) women (p=0.04). The clinical relevance of this is unclear given the nonlinear trend.89

While some pharmacokinetic data show that women with raised BMI using the implant may have lower serum ENG concentrations, clinical data do not suggest these women are experiencing more unintended pregnancies.

Studies considering contraceptive failure

A combined analysis90 of clinical trial and postmarketing surveillance data on 923 women and 1849 woman-years of implant use, stratified by body weight and duration of use, provides some data on contraceptive failure rates in women with raised BMI weighing up to 149 kg (134 women had a body weight of more than 70 kg; 79 of these women used the implant for 2 years or more and 11 for more than 3 years). The distribution of body weight among women with contraceptive method failure was similar to that of the general population of ENG implant users, though the numbers of users with obesity was small. Other studies82 that have included women weighing >70 kg (n=162 had BMI >26 kg/m²) have similarly not reported any pregnancies in this weight range, although none of these studies were designed specifically to compare women of differing weights. Findings of these studies are limited by the small number of women weighing >100 kg included, and the limited data in women weighing >70 kg for up to 3 years of implant use.

Two more recent analyses have included larger proportions of women with raised BMI and looked at extended implant use. A secondary analysis of 1168 implant users in the US Contraceptive CHOICE project showed that pregnancy rates in users of the ENG implant are extremely low and similar in women who were overweight and had obesity compared with normal-weight women over 4 years of implant use.59 In this cohort, 28% of women were overweight and 35% had obesity. Cumulative failure rates over 3 years were 0.00 per 100 woman-years for normal-weight and overweight women, and 0.23 per 100 woman-years for women with obesity. This study also assessed contraceptive failure by body weight; those women weighing >70 kg were not significantly more likely to experience implant failure compared to those weighing ≤70 kg (aRR 1.34; 95% CI 0.53–3.43), but the precision of this estimate was limited by the fact that pregnancy was a rare event and the small number of women in this weight group.

In a study of prolonged ENG implant use (up to 5 years), 237 women continued ENG implant use beyond 3 years, of whom 25% were overweight and 46% had obesity.88 No pregnancies occurred during the period of prolonged use, leading to an estimated failure rate of 0 (97.5% one-sided CI 0–1.61) over all BMIs. Serum ENG evaluation showed that median levels remain above the suggested ovulation threshold of 90 pg/mL for women in all BMI classes at 4 and 5 years. Some women with BMI ≥25 kg/m² experienced serum ENG levels below 90 pg/mL; however, the researchers did not specify how many women this included.

Overall these clinical outcomes data from recent prospective studies with appreciable numbers of women with raised BMI are reassuring that contraceptive failure is extremely rare with the ENG implant in women in all weight groups, even if use is extended beyond the licensed 3 years.

The Summary of Product Characteristics (SPC) for the ENG progestogen-only implant91 advises that “the clinical experience in heavier women in the third year of use is limited”. It therefore states that “it cannot be excluded that the contraceptive effect may be lower than for women of normal weight”. It advises that health professionals may therefore consider earlier replacement of the implant in ‘heavier’ women. The SPC91 does not specify a definition of heavier weight or after what duration of use replacement may need to be considered. The FSRH advises that there is no direct evidence to support a need for earlier implant replacement, and recent data assessing continued use in women with raised BMI beyond 3 years are very reassuring. Therefore, the GDG recommends that the ENG implant can be considered to provide very effective contraception for 3 years for women in all weight/BMI categories.

5.2.2 Implant safety

According to the UKMEC, obesity alone does not restrict the use of IMP (UKMEC 1).1 Even when obesity coexists with other risk factors for CVD (e.g. smoking, diabetes and hypertension), use of IMP is UKMEC 2.

No studies have directly assessed cardiovascular risk in women who are overweight or have obesity and use IMP. Data presented here are, thus, not specific to women with raised BMI. Studies looking at cardiovascular risk factors (e.g. lipids and carbohydrate metabolism) in healthy, non-obese implant users have been reassuring.56–58,60,61

Those studies that have looked at different progestogen-only contraceptives generally suggest that there is no increased risk of VTE,61,62 no statistically significant increased risk of MI,64 and no statistically significant increased risk of stroke associated with use.65 However, data are not available in relation to BMI or body weight, more research is required in high-risk women, and few studies have included the ENG implant.

A Danish national registry-based cohort study61 sought to assess the incidence of VTE in current users of non-oral hormonal contraception as well as different OC formulations. The study included all Danish non-pregnant women aged 15–49 years (n=1 626 158), with no history of thrombosis, and followed them from 2001 to 2010. Five confirmed venous thrombosis events were observed during the use of ENG implants, corresponding to an incidence rate of 1.7 per 10 000 exposure years and a non-significant aRR of 1.40 (95% CI 0.58–3.38) compared with non-users of hormonal contraception.

5.2.3 Weight gain with implants

In the general population of all implant users, there is no evidence of a causal association between implant use and weight gain. There is no specific evidence relating to weight gain with IMP use by women who are overweight or women with obesity.

5.2.4 Health benefits of implants

The main non-contraceptive benefit of IMP is that it may help alleviate dysmenorrhoea and ovulatory pain that are not associated with any identifiable pathological condition.92–94 While there is theoretically no reason why this would not be the case for women who are overweight or with obesity, this has not been specifically studied in women of different weight categories.

5.2.5 Practical considerations with implants

There are no data to suggest placement or removal of IMP is problematic in women who are overweight or women with obesity. Correct subdermal placement of the implant is important in women of all BMIs. Insertion or removal difficulties should not be presumed in women with raised BMI. Removal of appropriately placed implants (i.e. subdermal placement) should not be affected by BMI, including in the case of weight gain after insertion.

5.3.1 DMPA effectiveness

There is limited evidence regarding the effect of weight on the contraceptive effectiveness of progestogen-only injectable contraception, and no studies have specifically compared DMPA-IM or NET-EN failure rates in women with obesity versus women of normal weight.58 Overall, the available evidence (summarised below) indicates that the contraceptive effectiveness of DMPA is not affected by weight or BMI.

A study of the pharmacokinetics of DMPA-SC in women with BMI ≥30 kg/m² versus women with normal BMI observed that median medroxyprogesterone levels remained above the level needed to prevent ovulation. This study of 15 women (five per group) of normal weight, BMI ≥30 to <40 kg/m² and BMI ≥40 kg/m² found that medroxyprogesterone levels were lower in women with BMI ≥30 kg/m², and especially those with BMI ≥40 kg/m², compared to normal-weight women, although high enough to successfully suppress ovulation. In one woman with BMI ≥40 kg/m², levels dipped below the therapeutic level for the first 2 months of treatment and then rose.96 One multicentre study of DMPA-IM in 846 women (over 389.5 woman-years) reported a pregnancy rate of 0.7/100 woman-years across all women, and baseline body weight was not related to contraceptive failure. However, less than 5% of this population weighed >80 kg.97 An analysis of the contraceptive efficacy of DMPA-SC from two large phase III efficacy studies that included substantial numbers of women with obesity (n=601, BMIs ranging from 14.7 to 57.7 kg/m²; 18% with BMI >30 kg/m² in the trial sites in the Americas) reported no pregnancies in 16 023 woman-cycles of DMPA-SC use.98

5.3.2 DMPA safety

According to the UKMEC, obesity alone does not restrict the use of progestogen-only injectable contraception (UKMEC1).1 However, when obesity is one of multiple risk factors for CVD (e.g. smoking, diabetes and hypertension), use of progestogen-only injectable contraception becomes a UKMEC 3 (the theoretical or proven risks usually outweigh the advantages of using the method).1 The UKMEC notes that for women using NET-EN, the UKMEC categories are considered the same as for DMPA.

Studies relating effect of progestogen-only injectable use to risk of adverse cardiovascular events

No studies have specifically assessed whether women with raised BMI using progestogen-only injectable contraception are at increased risk of VTE and other cardiovascular outcomes, as compared to their normal-weight counterparts. Data presented here are therefore not specific to women with raised BMI.

One case-control study found a non-statistically significant increased risk of venous thrombosis in the small number of DMPA users compared to non-users of any type of hormonal contraceptive (11 cases and 34 controls); aOR for VTE 2.19 (95% CI 0.66–7.26).99 A larger case-control study,63 which included 446 cases and 1146 controls, found that DMPA use was associated with a 3.6-fold (95% CI 1.8–7.1) increased VTE risk compared with non-users of hormonal contraceptives. A meta-analysis of five case-control and three retrospective cohorts reported an increased risk of VTE associated with progestogen-only injectable use (VTE RR for users of an injectable progestogen versus non-users 2.67 (95% CI 1.29–5.53)).62 Only two studies, with a total of 31 VTE events, could be used to compute this value because no other studies reported the results separately for the subgroup of progestogen-only injectable users. More research is required before a causal relationship between progestogen-only injectable use and VTE can be confirmed or excluded. An international hospital-based case-control study99 comparing progestogen-only injectable users and non-users reported that current use did not affect risk of combined CVD, risk of stroke, or MI.

Studies relating to effect of progestogen-only injectable use on cardiometabolic markers

A prospective, non-randomised study100 of 14 women examined the effects of DMPA-IM or DMPA-SC on coagulation and inflammatory factors indicative of increased risk of thrombosis. Following four injections (i.e. 12 months of DMPA use), coagulation and inflammatory factors were not adversely affected by DMPA use, with concentration of D-dimer significantly reduced from baseline.100 The study was limited by very small numbers and no control group, making the findings difficult to interpret. However, the findings were not suggestive of DMPA having a negative effect on possible markers of thrombosis.

A prospective study101 examining the short-term effects of DMPA-SC on androgenic markers (testosterone, androstenedione, dehydroepiandrosterone sulphate, 3ɑ-androstanediol glucuronide and sex hormone-binding globulin) in women of differing BMIs (five normal-weight, five obese and five morbidly obese women) reported significant decreases in levels from baseline administration of DMPA-SC to week 26 among all three BMI groups and no differences between the groups at any time points. A Cochrane Review102 examined data from 31 randomised controlled trials (RCTs) to assess whether hormonal contraceptive use affected carbohydrate metabolism in healthy women and women who were at risk for diabetes due to being overweight. There were few data available for progestogen-only injectables, and one study showed a higher mean fasting glucose, glucose 2-hour response, and fasting insulin level amongst DMPA users compared to those using NET-EN. Overall the review suggested that there was little evidence on which to base conclusions about the impact of hormonal contraceptives in women with diabetes. Among women without diabetes, there did not appear to be any major differences in terms of carbohydrate metabolism. No studies were identified comparing DMPA-IM and DMPA-SC with regard to cardiovascular parameters.

5.3.3 Weight gain with DMPA

The data on progestogen-only injectable use and weight gain specifically in women who are overweight or women with obesity mainly derive from studies of adolescents. Overall, the available evidence (summarised below) indicates that use of DMPA is associated with some weight gain.

BMI prior to DMPA use in adolescents (aged <18 years) may predict weight gain with DMPA use, with higher initial BMI being predictive of increased weight gain, but evidence is insufficient to draw definitive conclusions.103–106 A systematic review of adverse events after contraceptive initiation among users of all ages and BMIs examined whether early weight gain among DMPA users was associated with later weight gain.107 Women who gained more than 5% of their baseline body weight in the first 6 months of DMPA use were more likely to experience continued weight gain.108,109

Significantly greater weight gain has been observed in the general population of new adolescent users of progestogen-only injectables compared to OC users, non-hormonal contraception users and discontinuers of any of the methods studied.103,110 A longitudinal study recruited 15–19-year-old initiators of DMPA (n=115), NET-EN (n=115), COCs (n=116) and non-users of contraception (n=144), and followed them for 4–5 years with 6-monthly assessments.103 Adolescents using DMPA or NET-EN throughout, or switching between the two, had gained an average of 6.2 kg compared to average increases of 2.3 kg in the COC group, 2.8 kg in non-users and 2.8 kg among discontinued users of any method (p=0.02). There was no difference in weight gain between adolescents classified as normal weight or overweight/obese at baseline in any of the four study groups.

Among young women with raised BMI, an observational study on weight gain with DMPA has reported that adolescents with raised BMI gain significantly more weight when using DMPA than when using oral contraception or no contraception.110 This prospective study of predominately black adolescents in the US of varying weight showed that those with a BMI >30 kg/m² had significantly increased weight gain over 18 months after initiating DMPA compared with obese adolescents initiating OC or using no hormonal contraception.110 At 18 months, mean weight gain was 9.4, 0.2 and 3.1 kg, respectively, for adolescents with obesity receiving DMPA, OC and no hormonal contraception (p<0.001).110 Weight gain over 18 months in adolescents with obesity using DMPA was also greater than weight gain among non-obese DMPA and OC users and non-obese users of no hormonal contraception.

5.3.4 Health benefits of DMPA

DMPA can be used as a treatment for the management of heavy menstrual bleeding (HMB), dysmenorrhea and for pain associated with endometriosis.111,112 DMPA may confer some protection against ovarian and endometrial cancers.95,113–116 While there is no reason to expect this would not be the case for women with raised BMI, the health benefits associated with DMPA have not been specifically studied in women of different weight categories.

5.3.5 Practical considerations with DMPA

The SPC for DMPA-IM117 states that it should be administered by deep intramuscular injection into muscle tissue, preferably the gluteus maximus but other muscle such as the deltoid (upper arm) may be used. Traditionally the dorsogluteal site (upper outer quadrant of the buttock) has been used. The ventrogluteal site (lateral thigh) has been investigated as an alternative site because the risk of sciatic nerve injury is reduced118,119 and the fat layer is thinner than in the dorsogluteal area. However, in women who are classified as overweight or obese it may be difficult to ensure intramuscular administration in either the dorsogluteal or ventrogluteal region. A retrospective study120 of 100 adult women reported that using standard-length ‘green’ 21-gauge needles (40 mm) into the ventrogluteal site resulted in 12% of injections being subcutaneous rather than intramuscular, and that in the dorsogluteal site 43% of injections failed to reach the muscle. For standard-length ‘blue’ 23-gauge needles (25 mm) the proportions were higher (26% and 72%, respectively). If there are concerns about the ability to administer an intramuscular injection due to body weight then the deltoid muscle in the upper arm may be considered as an alternative site117 or DMPA-SC could be a suitable alternative.

5.4.1 POP effectiveness

There is very limited evidence on POP effectiveness in women who are overweight or women with obesity. The available data (summarised below) have not shown reduced POP effectiveness in women with higher weight and/or BMI.

One UK observational cohort study found no association between body weight and contraceptive failure in traditional POP users;122 however, this study had significant limitations, including no report of the number of women with obesity or who were overweight, lack of statistical power, and measurement of weight only at recruitment (from 1968–1974). Data from an exploratory in vitro study of 16 women123 found that a single tablet of LNG 30 μg (n=8) or NET 350 μg (n=8) prevented sperm migration in cervical mucus 12 hours after ingestion, including in the three women who had a BMI >35 kg/m².123 As this study only examined the impact of a single POP dose, further evidence is required before conclusions can be drawn about the impact of weight or BMI on the cervical mucus effect of POPs.

No studies have examined the comparative effectiveness of the desogestrel (DSG) pill in women of differing weights or BMIs. One 12-month study124 of 71 women examining inhibition of ovulation with use of the DSG pill compared to an LNG pill reported that ovulation was suppressed in all but one of 59 DSG cycles. The weight range of the women in the DSG group was 46–78 kg, suggesting that the DSG pill is effective in women up to this weight. Some 16/57 LNG cycles were ovulatory, and women in this group weighed between 50 and 91 kg. However, no analysis was carried out by weight.

The SPCs for POPs125–127 do not advise dose adjustments based on weight or BMI.

5.4.2 POP safety

According to the UKMEC, obesity alone does not restrict the use of POP (UKMEC 1 indicating no restrictions on use).1 Even when obesity is in the context of other risk factors for CVD (e.g. smoking, diabetes and hypertension), use of POP is UKMEC 2.

No studies have directly assessed whether women with raised BMI who use POP are at increased risk of VTE and other cardiovascular outcomes, as compared to their normal-weight counterparts. Data presented here are therefore not specific to women with raised BMI but do not suggest a likely association between POP use and increased risk of CVD.

Those few studies that have looked at different progestogen-only contraceptives generally suggest that there is little or no increased risk of VTE,61,62 no statistically significant increased risk of MI,64 and no statistically significant increased risk of stroke65 with POP, implants or LNG-IUS. Data, however, are not available in relation to BMI or body weight. A Danish national registry-based cohort study61 sought to assess the incidence of VTE in current users of non-oral hormonal contraception as well as different OC formulations. The study included all Danish non-pregnant women aged 15–49 years (n=1 626 158), with no history of thrombosis, and followed them from 2001 to 2010. The risk of confirmed VTE was not increased with use of hormone-releasing IUDs (aRR 0.57; 95% CI 0.41–0.81) or progestogen-only subcutaneous implants (aRR 1.40; 95% CI 0.58–3.38) compared with non-users of hormonal contraception. This study did not include POPs.

Very little research has been conducted to assess the effect of progestogen-only contraception on cardiovascular risk factors (e.g. lipid and carbohydrate metabolism) in women with raised BMI or in relation to BMI/weight. Studies looking at this in healthy, non-obese POP users have been reassuring, with no clinically meaningful changes observed.66–73

5.4.3 Weight gain with POP

In the general population there is no evidence suggesting a causal association between POP use and weight gain. There is no specific evidence relating to weight gain with POP use by women with raised BMI.

5.4.4 Health benefits of POP

The main non-contraceptive benefits of POP are that they may help alleviate HMB and dysmenorrhoea with possible alleviation of premenstrual syndrome symptoms for the DSG-containing POP.128–130 While there is theoretically no reason this would not be the case for women with raised BMI, this has not been specifically studied in women of different weight categories.

5.5.1 CHC effectiveness

Combined oral contraceptive pill (COC)

The findings of studies of COC effectiveness in relation to increased body weight/BMI vary. In general, evidence relating to the effect of increased body weight/BMI on effectiveness of COC is limited to observational studies in which height, weight and pregnancy are often self-reported, and potential confounding factors such as contraceptive adherence and frequency of sexual intercourse are unknown. Most studies include relatively few women in the highest weight/BMI categories, and are unable to elucidate possible differences in weight-related effectiveness by type of progestogen.

A 2017 systematic review132 reported that 10 of the 14 studies of COC identified did not report a difference in COC effectiveness by body weight or BMI. In the remaining four studies, the magnitude of the reported increase in COC failure among overweight women and women with obesity compared to normal-weight women was very small.133–136 Potential differences between different COC formulations could not be distinguished. The studies included used a variety of measures and thresholds to examine associations between obesity and contraceptive effectiveness, making interpretation across studies difficult. All included studies were of fair to poor quality. Most of the studies reporting an association between higher BMI and COC failure are significantly limited by lack of data on pill adherence as well as self-reported weight recorded at a time distant from the time of contraceptive failure.

A total of four fair to poor quality studies identified an increased risk of contraceptive failure with COC in women with obesity compared to women without obesity. Two studies of low quality suggested that women with raised BMI using COC could experience up to a two-fold risk of contraceptive failure compared to their normal-weight counterparts. However, these studies relied on self-assessment of height and weight, recalled an average of 76.5 months after last COC use, and reported dichotomous outcomes using cut-off points that included a range of raised BMIs.133,134 Another study, a recent pooled analysis of the original data from seven phase III trials submitted to the US Food and Drug Administration (FDA), noted a small but statistically higher COC failure rate among women with obesity: women with a BMI ≥30 kg/m² had a slightly increased risk of contraceptive failure compared to those with a BMI <30 kg/m² (adjusted hazard ratio (aHR) 1.44; 95% CI 1.06–1.95). The magnitude of the difference reported (Pearl Index of 3.14 in users with obesity vs 2.53 in users with normal weight) is unlikely to be clinically important.135 This study may be biased due to limited control for confounding variables, no information about pill compliance, and pooling of data where women were exposed to COC containing different progestogens. In addition, one study found an association between women with more extreme obesity (BMI ≥35 kg/m²) and decreased COC effectiveness (aHR 1.5; 95% CI 1.3–1.8).136

By contrast, all 10 remaining studies in this systematic review did not show an association between increased BMI/weight and COC effectiveness. Four large, population-based studies did not demonstrate an association with more detailed analyses of weight and BMI classes.122,137–139 Further, four observational studies that incorporated more rigorous methods for capturing measures of height and weight, validated pregnancy status and included measures of contraceptive adherence also did not report any increased risk for COC failure.140–143 Two lower-quality case-control studies limited by small numbers and lack of information on contraceptive adherence also found no association between BMI and COC failure.144,145

In addition, a 2016 Cochrane Review58 concluded that in general, the evidence identified in that review did not indicate an association between increasing body weight or BMI and effectiveness of COC.

There is limited evidence that ovarian activity in the hormone-free interval (HFI) could be more pronounced in obese women.146,147 In one study, hormone profiles at the end of a 7-day HFI in 10 women with normal BMI and 10 women with obesity noted estradiol levels consistent with dominant follicles and progesterone levels consistent with ovulation in more women with obesity than women with normal BMI.146 A further study of women with obesity using 20 µg EE COC in a 21/7 regimen noted a reduction in ovarian activity when the woman changed to either a 30 µg EE COC as a 21/7 regimen or a 20 µg EE COC taken continuously.147

Overall, most high-quality studies suggest that COCs are as effective at preventing pregnancy in women with obesity as among non-obese women, but could potentially be less forgiving of imperfect use because the pharmacokinetics of steroid hormones appear to be altered in obese OC pill users compared with normal-weight users. However, ovulation is still suppressed in most OC pill users with obesity.148–150

Combined transdermal patch (patch)

The patch currently available in the UK, Evra^®, is an EE/norelgestromin (NGMN) patch.151 Data on the patch in the context of obesity are limited.

The limited evidence identified132,135,152 suggests that increasing body weight and BMI may be associated with increasing contraceptive failure rates of the EE/NGMN patch. Two studies have reported on contraceptive patch effectiveness and weight.135,152

A pooled analysis of phase III clinical trials between 2000 and 2012 submitted to the US FDA included one trial of an EE/NGMN patch (n=1523), with the same dosing as Evra.135 This study found that when adjusted for age and race, women with obesity (n=152) using the patch were at increased risk for contraceptive failure compared to women without obesity (aHR 8.8; 95% CI 2.54–30.5). Another pooled analysis of three multicentre, open-label studies of the EE/NGMN contraceptive patch (n=3319) reported that contraceptive failure was low and uniformly distributed across the range of body weights <90 kg, but suggested that the patch may be less effective in the subgroup of women weighing ≥90 kg. In total, 15 pregnancies were diagnosed during 6–13 cycles of follow-up. Five pregnancies occurred among women weighing ≥90 kg; women of this weight comprised less than 3% of the study population. Ten pregnancies were diagnosed among women weighing <90 kg.152 Thus, women weighing ≥90 kg reported proportionally more pregnancies compared to those of lesser weight. While results by BMI were not provided, the researchers noted a statistically significant association between baseline body weight and pregnancy. However, the pregnancy rates on the EE/NGMN patch were still low in both of these studies across all weight categories.

The SPC for the Evra patch151 states that contraceptive effectiveness could be decreased in women weighing ≥90 kg. As there are no new data to refute the manufacturer’s statement, the GDG recommends that additional precautions for pregnancy prevention or an alternative method of contraception should be advised for women weighing ≥90 kg.

Combined vaginal ring (ring)

The combined vaginal ring contains EE and ENG.153 Data on the ring in the context of obesity are limited. Most large trials assessing the effectiveness of the EE/ENG vaginal ring have been conducted in women without obesity.

One study directly assessed contraceptive failure by weight among women using the combined vaginal ring.154 This study of 128 pregnancies that resulted from a combined group of COC, patch or ring users concluded that 3-year failure rates among women using these methods were not different across BMI categories. However, the methodology used and the possible effect of non-adherence to the methods could have masked an actual effect of BMI on contraceptive failure in this study.155 The only study to specifically assess contraceptive failure of the EE/ENG vaginal ring in women with obesity versus women without obesity was a secondary analysis of phase III efficacy trials, but it included too few obese women to provide a precise answer. In this study, the pregnancy rate for women in the highest decile of weight (>75 kg/167 lbs) was 1.2%, with no pregnancies reported in the highest weight women (85–123 kg).156

Two pharmacokinetic analyses of the EE/ENG vaginal ring provide some indication that ring effectiveness could be similar in obese and non-obese women. In one analysis of 20 normal-weight women, while women with obesity had lower EE levels during ring use, ENG levels were similar, and follicular development was minimal in both groups. The researchers concluded that the findings predict that ring effectiveness will be similar in women with a BMI up to 39.9 kg/m² 157. In another pharmacokinetic analysis comparing 20 women with a normal BMI and 20 women with obesity, a single ring used for 6 weeks demonstrated therapeutic serum levels of EE and ENG among women with normal and obese BMI.158 Although reassuring, these studies were conducted over only one cycle in small groups of women.

This limited evidence suggests that the effectiveness of the ring is not affected by weight.

5.5.2 CHC safety

UKMEC recommendations relate to safety of use. CHC is UKMEC 3 for use by women with BMI ≥35 kg/m². Use of CHC is UKMEC 2 for use by women with BMI ≥30–34 kg/m².

These UKMEC 2 and 3 classifications, which indicate safety concerns for obese women using CHC, are related to cardiovascular risks from exogenous estrogen, including VTE, acute MI and stroke. They are based primarily on evidence that obesity and CHC use are both independent risk factors for thrombosis.

Venous thromboembolism (VTE)

Independent of CHC use, the risk of VTE rises as BMI increases over 30 kg/m² and rises further with BMI >35 kg/m².43,44 Baseline VTE risk in obese women is two-fold higher than VTE risk in normal-weight women.45,46 VTE risk also increases significantly with age, irrespective of BMI.159 CHC is associated with an increased risk of VTE, with use of CHC increasing VTE risk three-fold (in non-obese CHC users).160–163 Some research has suggested that there may be additional additive increased VTE risk in women who have raised BMI and use CHC.45,164–166

Among normal-weight women, evidence from observational studies suggests that current COC use is associated with a 3–3.5-fold increase in VTE risk compared with non-use.161–163 And, while data are conflicting, with some studies finding no increased risk of VTE in users of the contraceptive patch or ring when compared to COC users,167,168 other studies have reported a significant two-fold greater VTE risk among patch and ring users compared to COC users.61,169

A recent systematic review170 investigated whether CHC use modifies the risk of VTE in obese women, and also evaluated evidence for a dose–response relationship between BMI and VTE. No studies regarding the contraceptive patch or vaginal ring met the inclusion criteria in this review. Data from one pooled analysis,171 one cohort study172 and eight case-control studies45,164,166,173–179 showed that obese COC users consistently had VTE risk five to eight times that of obese non-users and approximately 10 times that of non-obese non-users. A single study180 of 129 cerebral venous thrombosis (CVT) cases suggested that COC users with BMI ≥25 kg/m² were at higher risk for CVT compared with normal-weight non-users (BMI <25 kg/m²), and that increasing BMI was associated with increasing CVT risk, but due to the small sample size, estimates of this association were imprecise.

Both COC use and higher BMI increase risk for VTE, and the greatest relative risks in this body of data were for those women with both risk factors. These studies are all limited by small numbers in BMI and COC-using subgroups, lack of information on the type/formulation of OC, and no adjustment for lifestyle factors or genetics. Data around BMI were frequently self-reported, and OC use was often based on prescriptions given, not actual use. Confidence intervals were often wide, particularly for the higher BMI groups (e.g. ≥35 kg/m²) because of the small numbers of OC users in those categories.

Based on the available data it is not possible to estimate absolute risk of VTE among women with both of these VTE risk factors: CHC and obesity. However, the absolute risk of VTE in healthy women of reproductive age is small (2 events per 10 000 women per year),181 as is the absolute risk of VTE associated with use of CHC (5–12 events per 10 000 women per year).181 The additional risk of VTE in the context of obesity in addition to CHC use is still likely to be less than the VTE risk that pregnancy/postpartum poses in a woman with obesity.181,182 However, women with obesity should be advised about effective methods of contraception that are not associated with increased risk of VTE.

There are no published studies focusing on VTE risk in contraceptive patch or ring users with obesity.

Other cardiovascular outcomes

The 2016 systematic review170 also investigated whether CHC use modifies the risk of other cardiovascular outcomes in obese women. No studies on the contraceptive patch or vaginal ring were identified that met the inclusion criteria in this review. This review included one pooled analysis183 and one case-control study184 of MI and one pooled analysis185 and one case-control study186 of stroke. The pooled analyses of case-control data found no increased risk for MI or stroke for COC users overall or stratified by BMI,183,185 whereas the other case-control studies that were included in the systematic review found significantly increased risk of MI184 and ischaemic stroke186 for obese COC users compared to normal-weight non-users, with the highest risk estimates for high-BMI COC users. These studies were of fair quality with diagnostic criteria for MI and stroke and adjustment for confounders, but with self-reported weight and height and inadequately systematic inclusion of MI and stroke fatalities, which could have led to bias. These studies all also considered a BMI of 27.3 kg/m² as the threshold for obesity, which is below the WHO definition of 30 kg/m², and could have led to an underestimation of the strength of the association between these cardiovascular outcomes and obesity.

5.5.3 Weight gain with CHC

In the general population there is no evidence that use of CHC causes weight gain. There is no specific evidence relating to weight gain with CHC use by women who are overweight or women with obesity.

5.5.4 Health benefits of CHC

Many non-contraceptive benefits are associated with CHC, including reduction of HMB and pain, alleviation of premenstrual symptoms, and management of symptoms associated with polycystic ovary syndrome. There is also a reduced risk of endometrial, ovarian and colorectal cancer.187–195

The only evidence relating to these benefits in women with raised BMI is from one large US prospective cohort study190 that found a significant 34% reduction in endometrial cancer risk associated with ≥10 years OC use compared to never-use or use for less than a year; the greatest risk reduction was observed amongst smokers, women who rarely exercised and women with obese BMIs.

5.6.1 Barrier method effectiveness

No studies have evaluated the effectiveness of barrier methods in women with obesity versus women without obesity or assessed weight-related effectiveness. Condoms (male and female), the diaphragm, cervical cap and contraceptive sponge are the most common forms of barrier contraception, and theoretically should be equally effective in women of all weights; however, there are no studies which make this direct comparison.

Following a large change in weight or childbirth, the GDG suggests good practice would be to check if a woman’s diaphragm still fits. A small magnetic resonance imaging (MRI) study has shown that following simulated intercourse, the single-size SILCS diaphragm remained in position covering the cervix in women with varying parity and BMI (n=2, BMI <25 kg/m²; n=2, BMI 25–30 kg/m²; n=2, BMI >30 kg/m²).196

Barrier methods are prone to user-related contraceptive failure due to non-adherence or incorrect use. This is reflected in their relatively higher contraceptive failure rates when compared to other contraceptive methods (see Table 2).

5.6.2 Barrier method safety

No studies have specifically evaluated the safety of barrier methods of contraception in women with raised BMI. There are no theoretical reasons for barrier methods of contraception to pose health risks to such women.

5.8.1 Cu-IUD

The Cu-IUD is the most effective method of EC and effectiveness of the Cu-IUD is not known to be affected by weight or BMI. Women overweight or with obesity desiring EC should be counselled on the effectiveness of all methods, including the Cu-IUD, which remains the most effective option for EC regardless of weight.202

5.8.2 Oral EC

Some studies have suggested that both levonorgestrel EC (LNG-EC) and ulipristal acetate EC (UPA-EC) could be less effective in women who are overweight or obese compared to women with normal or underweight BMI.203 The reported negative effect of obesity on effectiveness of LNG-EC is greater than that on effectiveness of UPA-EC.204 However, the European Medicines Agency concluded in 2014 that the available evidence was limited and not robust enough to support with certainty a conclusion that oral EC is less effective in women with higher body weight or BMI.205

UPA-EC

A 2012 meta-analysis197 suggests that UPA-EC is less effective for women with a BMI >30 kg/m² than for women with a BMI <30 kg/m² (pregnancy OR for women with obesity vs women without obesity 2.1 (95% CI 1.0–4.3; p=0.04); pregnancy OR for women weighing >85 kg vs women weighing <85 kg 2.2 (95% CI 1.1–4.6; p=0.03)). A 2011 meta-analysis198 reports a non-significantly greater pregnancy risk for women with obesity (BMI >30 kg/m²) using UPA-EC than for women with BMI <25 kg/m² (OR 2.62; 95% CI 0.89–7.00). The findings have limitations: none of the trials from which the data were taken206–208 were designed primarily to consider the effect of weight or BMI; weight and height were self-reported by women and may be inaccurate; numbers of pregnancies amongst women with obesity were small and confidence intervals are wide. Despite these limitations, the data suggest that UPA-EC could potentially be less effective for women weighing >85 kg or with a BMI >30 kg/m² than for women weighing <85 kg or with BMI <30 kg/m². A recently published pharmacokinetic study209 comparing serum UPA concentrations in 16 obese-BMI women and 16 normal-BMI women after taking UPA 30 mg found no significant difference between the two groups. This contrasts with the findings for LNG-EC (see below). There is no evidence that an increased dose of UPA-EC is more effective than the standard 30 mg dose in these women, as this has not yet been tested. Double-dosing of UPA-EC is not currently recommended.

LNG-EC

A pooled analysis199 of data from the LNG-EC comparator arms of two RCTs carried out comparing UPA-EC with LNG-EC206,208 demonstrates a sharp increase in pregnancy rates with LNG-EC for women weighing >70 kg or with BMI >26 kg/m². Considering the same data, women with raised BMI who took LNG-EC were at four times greater risk of pregnancy than women with BMI <25 kg/m² who took LNG-EC (OR 4.41; 95% CI 2.05–9.44; p=0.0002).198 These studies considered women from the UK and US; they were not designed primarily to assess the effect of weight or BMI on the effectiveness of oral EC, weights were sometimes self-reported, and the number of pregnancies amongst women with raised BMI were small.

One analysis of pooled data from three RCTs conducted by WHO concluded that there is no apparent effect of BMI or body weight on the effectiveness of LNG-EC.200 A second analysis of pooled data from these same three RCTs and a fourth WHO RCT suggested a greater risk of pregnancy after LNG-EC amongst women with a BMI >30 kg/m² than women with a BMI <25 kg/m². Again, the data were taken from studies that were not primarily designed to consider effect of weight or BMI on effectiveness of oral EC, weights were self-reported and the number of women included in the studies who had a BMI >30 kg/m² was small. A total of only six pregnancies occurred in women with obesity, all at the same Nigerian study site, and all of whom took oral EC after the expected date of ovulation.201

A study210 of the pharmacokinetics of LNG-EC in five women with obesity and five women without obesity demonstrates that obesity adversely impacts maximum serum concentrations of LNG. The authors postulate that this may explain a reduction in effectiveness of LNG-EC in women with obesity. In this study, doubling the dose of LNG-EC appears to correct the obesity-related pharmacokinetic changes without observed adverse effects. However, it is concluded that “additional research is needed to determine if this also improves EC effectiveness in obese women”. Another pharmacokinetic study comparing 16 women with obese-BMI and 16 with normal-BMI also demonstrated a significant difference in drug levels.209 The study concludes that after a single dose of LNG-EC, obese-BMI women are exposed to lower concentrations of LNG when compared to normal-BMI women. In a recent pharmacokinetic study, 1.5 mg LNG was given to 10 women with normal BMI and 16 women with obesity. All women studied were of reproductive age and ovulatory. Total serum LNG was measured over 0 to 96 hours; women with obesity were exposed to significantly lower total and bioavailable (free plus albumin-bound) LNG than women with normal BMI.211 The authors concluded that this lower LNG exposure could play a role in the purported reduced efficacy of LNG-EC in women with obesity. These pharmacokinetic findings contrast with those for UPA-EC (see above).

The evidence presented above suggests that LNG-EC could be less effective in women weighing >70 kg or with a BMI >26 kg/m². If a Cu-IUD is not indicated or not acceptable, such women can be offered UPA-EC. If UPA-EC is not suitable, a double-dose (3 mg) of LNG-EC can be used. Studies testing the effectiveness of double-dose LNG-EC are in progress, but its effectiveness is unknown. However, 3 mg LNG-EC is well tolerated and pharmacokinetic data210 suggest its potential ability to prevent unintended pregnancy more effectively than 1.5 mg LNG-EC in women weighing >70 kg or with a BMI >26 kg/m².

5.9.1 Effectiveness and safety

There is very limited evidence relating to sterilisation in women with obesity compared with women who are not obese. The available data suggest that women who are obese undergoing sterilisation may be at an increased risk of complications and/or technical failure of laparoscopic sterilisation,80,212 but that hysteroscopic sterilisation is unaffected by weight/BMI. Hysteroscopic tubal occlusion (Essure^®) is effective and may be a safer surgical sterilisation procedure than laparoscopic sterilisation for obese women because it can be performed under local anaesthesia and does not require entrance into the abdominal cavity, with its attendant morbidity. Essure had offered this alternative approach; however, it is no longer available in the UK, Europe or the US. Laparoscopic tubal sterilisation procedures are also effective, but are associated with longer operating times and more anaesthetic and surgical complications in women with obesity compared with women who are not obese.212–221

6.1.1 Orlistat

According to the National Institute for Health and Care Excellence (NICE),223 orlistat may be prescribed as part of an overall plan for managing obesity in adults who meet one of the following criteria: BMI ≥28 kg/m² with associated risk factors or BMI ≥30 kg/m², or to maintain or reduce weight before surgery for people who have been recommended bariatric surgery as a first-line option.

Orlistat 120 mg, an oral medication only available on prescription in the UK, is a gastrointestinal lipase inhibitor that acts by inhibiting the absorption of dietary fats, and is primarily metabolised within the gastrointestinal wall.224,225 A lower dose (60 mg) of orlistat is available over the counter. Orlistat is not recommended for use in pregnancy and breastfeeding224,226 so effective contraception with orlistat use is important.

There is almost no evidence on orlistat and OC. Based on its mechanism of action, orlistat could theoretically reduce the absorption of OC. One small pharmacokinetic study of poor quality, conducted by the makers of orlistat, concluded that orlistat has no effect on pharmacokinetics and/or pharmacodynamics of COC.227 In this study227 of normal-weight women, two groups of 10 women “on a stable regimen with OCs”, received either 120 mg orlistat three times daily or placebo three times daily on days 1–23 of the first cycle. Then, separated by a placebo washout period on days 24–28, women received the alternative treatment on days 1–23 of the second cycle. In both cycles, serum luteinising hormone was measured on days 12–16 and progesterone on days 12, 16 and 19–23. No changes in the ovulation-suppressing action of the COC were reported.

In seven double-blind, placebo-controlled clinical trials of orlistat,225 gastrointestinal symptoms, including diarrhoea and vomiting, were the most commonly observed (defined as >5% or an incidence in the orlistat group that is at least twice that of placebo) treatment-emergent adverse events associated with orlistat use. This could potentially have implications for the concomitant use of OCs and orlistat.

A secondary analysis228 of 319 women from two RCTs of women aged 18–40 years who were overweight or had obesity (BMI 27–42 kg/m²) and infertile with PCOS reported on serious adverse events and side effects with concomitant orlistat and COC use. In one arm of that study, 43 women received concomitant orlistat and COC (20 EE/1 NET) for 16 weeks prior to clomifene treatment (the other arms received either orlistat or COC alone). No serious adverse events were reported in the 43 women using orlistat and COC concomitantly, but women using both were more likely to have diarrhoea/steatorrhoea than women using one or the other alone.

6.1.2 Liraglutide

Liraglutide (Saxenda^®) is a long-acting glucagon-like peptide-1 receptor agonist. In a crossover RCT229 of liraglutide versus placebo in 21 postmenopausal women (who had had an oophorectomy or at least 1 year of amenorrhoea) with BMI 18–30 kg/m² and we re given a single dose of COC (30 EE/15 LNG) to assess COC pharmacokinetics. No clinically relevant reduction in bioavailability of EE/LNG occurred, though more women reported gastrointestinal side effects while on both treatments than when on COC alone.

6.1.3 Naltrexone/buproprion

There were no studies identified on the use of naltrexone/buproprion with contraception.

6.1.4 Laxatives

Some individuals use and/or abuse laxatives in order to lose or manage weight. For more information on contraception for women who self-induce diarrhoea and/or vomiting, refer to the FSRH CEU Statement Contraception for Women with Eating Disorders.230

6.2.1 Background

Bariatric surgery is provided to women with morbid obesity to obtain substantial weight loss and reduce obesity-related comorbidities.231 Bariatric procedures have their effect on weight loss and comorbidity improvement through either a restrictive or a malabsorption mechanism, or a combination of both.231 The NICE clinical guideline Obesity Prevention232 recommends that surgery should be an obesity management option in certain circumstances. The most common procedure performed in the UK is the gastric bypass, with gastric sleeve and gastric banding performed less frequently.233,234 Most procedures are performed laparoscopically.235

UKMEC 2016 identifies bariatric surgery in the past 2 years as a condition that exposes a woman to increased risk in the context of pregnancy.1 Effective contraception is important, as pregnancy should be avoided during the period of intensive weight loss, from 12 to 18 months after surgery.236,237 Pregnancy occurring after surgery may be associated with an increased risk of both maternal and child complications.237–239 The weight loss that accompanies bariatric surgery can enhance fecundity in women who were subfertile because of anovulation.237,240 Thus, contraception education, counselling and follow-up should be conducted prior to and at the time of bariatric surgery for all women with the potential to conceive.

The evidence regarding use of contraception after bariatric surgery, however, is very limited. Based on evidence from surveys conducted outside the UK, condoms and OC appear to be the most commonly utilised contraceptive methods after bariatric surgery in the US, whereas IUC is the predominant method used by women in Scandinavian countries.241–243 Gaps in contraception education and low levels of postoperative use were reported in all studies.241–243

A US survey244 of 574 bariatric surgeons reported that the majority provided a consistent message to patients to delay pregnancy after bariatric surgery, but far fewer provided appropriate referrals or were aware of whether patients initiated contraception. This supports the need for better communication between bariatric surgeons and women’s HCPs. In a survey amongst members of the British Obesity and Metabolic Surgery Society, bariatric surgeon respondents reported that contraceptive information was rarely provided in bariatric surgical clinics, practitioners had low levels of confidence in discussing contraception, and the majority of respondents requested further training, guidance and communication with contraceptive practitioners.245

6.2.2 UKMEC information on bariatric surgery

The UKMEC1 states that women with conditions that may pose a significant health risk during pregnancy (e.g. bariatric surgery within the past 2 years) should be advised to consider using the most effective long-acting reversible contraception (LARC) methods, which provide highly reliable and effective contraception (failure rate <1 pregnancy per 100 women in a year).1 The sole use of barrier methods and user-dependent methods of contraception may not be the most appropriate choice for women who have undergone bariatric surgery in the past 2 years given the relatively high typical-use failure rates of these methods (see Table 2).

Table 3 outlines UKMEC recommendations relating to contraceptive use by women in different BMI categories following bariatric surgery. The UKMEC makes no distinctions based on type of bariatric surgery. UKMEC categories relate to the safety, and not effectiveness, of use. With the exception of CHC methods, all methods of hormonal and intrauterine contraception (i.e. Cu-IUD, LNG-IUS, IMP, DMPA and POP) are UKMEC 1 for use in women with a history of bariatric surgery for all BMI categories (i.e. BMI <30, ≥30–34 and ≥35 kg/m²).

6.2.3 Effectiveness and safety of contraception after bariatric surgery

The evidence regarding effective and safe use of contraception after bariatric surgery is very limited. No RCTs or large prospective longitudinal studies have been conducted.

Individuals who undergo bariatric surgery require life-long calcium and vitamin D supplementation. Refer to Section 6.2.3.5 below and method-specific guidance when considering osteoporosis risk factors and contraception.

6.2.4 Contraception before bariatric surgery

Irrespective of bariatric surgery, CHC use is UKMEC 3 (a condition where the theoretical or proven risks usually outweigh the advantages of using the method) for women with BMI ≥35 kg/m² and UKMEC 2 for women with BMI ≥30–34 kg/m², due to concerns about cardiovascular risks from exogenous estrogen, including VTE. Most women undergoing bariatric surgery should not be on a CHC method prior to surgery. However, some may be.

Women who are immobile or have restricted mobility for an extended period of time (e.g. due to surgery) are at increased risk for VTE. This risk may be further increased with use of CHC. Women using CHC who are planning to undergo bariatric surgery should discontinue use at least 1 month before surgery to reduce the risk of postoperative thromboembolism. With regard to the use of CHC by women undergoing surgery, the BNF259 advises that “estrogen-containing contraceptives should preferably be discontinued (and adequate alternative contraceptive arrangements made) at least 4 weeks before major elective surgery. A non-oral progestogen-only contraceptive or the Cu-IUD should be offered as an alternative”. In general, progestogen-only contraceptives are considered more appropriate in women with obesity than combined hormonal methods, especially in the context of surgery.174,260