A Cochrane review [Abstract] 1 included 9 reports with data from 17 trials involving a total of 63 813 subjects. Most studies did not show a higher pregnancy risk among overweight or obese women. One of 5 studies using BMI found a higher pregnancy risk for overweight or obese women. In that trial, women assigned to an oral contraceptive (COC) containing norethindrone acetate 1.0 mg plus EE 20 µg and having a BMI at least 25 had greater pregnancy risk compared to those with BMI less than 25 (reported RR 2.49; 95% CI 1.01 to 6.13). The comparisons reported in the other 4 studies were not significantly different for pregnancy. A trial using a COC with levonorgestrel and EE reported a Pearl Index of 0 for obese women (BMI ≥ 30) versus 5.59 for nonobese women (BMI < 30). The same trial tested a transdermal patch containing levonorgestrel and EE. Within the patch group, obese women in the "treatment-compliant" subgroup had a higher reported Pearl Index than nonobese women (4.63 versus 2.15). Four studies from the 1990s used weight alone rather than BMI. Results were mixed. Among skin patch users, body weight was associated with pregnancy (reported P < 0.001) but BMI was not. Of 5 implant studies, 2 that examined the six-capsule levonorgestrel implant showed differences in pregnancy by weight. One study showed higher weight was associated with higher pregnancy rate in years 6 and 7 combined (reported P < 0.05). In the other, pregnancy rates differed in year 5 among the lower weight groups only (reported P < 0.01) and did not involve women weighing 70 kg or more. However, the groups of overweight women were smaller than groups of women with normal weight in all studies.
A systematic review 2 evaluated risk of acute myocardial infarction (AMI), stroke, cerebral venous thrombosis (CVT) and venous thromboembolism (VTE) in obese women using combined hormonal contraception including 3 pooled analyses, 11 case-control studies and 1 cohort study. One study showed no increased risk for AMI or stroke for COC users overall or stratified by BMI, another study found significantly increased risk of AMI and stroke, with the highest risk estimates for high-BMI COC users. A single study suggested that obese COC users may be at higher risk for CVT compared with normal-weight nonusers. For VTE, obese COC users consistently had a risk that was 5 to 8 times that of obese nonusers and approximately 10 times that of nonobese nonusers. 5 prospective cohort studies were identified as indirect evidence, and all found increased risk for VTE as BMI increased, suggesting a dose-response relationship between BMI and risk for VTE.
In another systematic review 3 3 fair quality and one poor quality study reported increased COC failure among a heterogeneous population of overweight and obese women compared with normal weight women, while 8 fair quality and 2 poor quality studies did not find an association. 2 fair quality studies reported on contraceptive transdermal patches. One pooled analysis described a higher proportion of pregnancies among women using the patch who weighed ≥90 kg; another secondary analysis suggested BMI>30 was associated with increased failure.
A meta-analysis 4 using individual participant data directly from 7 clinical trials with COCs included 2707 obese (BMI ≥30) and 11 317 non-obese women. The Pearl index (PI) for each trial varied: 2.05 to 5.08 for obese and 1.84 to 3.80 for non-obese women. The pooled PI estimated using direct weighted average method was 3.14 (95% CI 2.33 to 4.22) for obese and 2.53 (95% CI 1.88 to 3.41) for non-obese women. The overall HR of 1.44 (95% CI: 1.06-1.95; p=.018) in the meta-analysis suggested a 44% higher pregnancy rate during COC use for obese women after adjusting for age and race.
Comment: The quality of evidence is downgraded by study quality (post-hoc analysis and other shortcomings) and by inconsistency (variability in results across studies).
Primary/Secondary Keywords