Table 1.
Characteristics of the included randomised controlled trials
| Age, years | Intervention | Comparison | First-time users, | Number who dropped out | Duration | RCT type | Analysis | Depression scale | |||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean (s.d.) | Formulation | na | Formulation | na | %b | n (%) | Number of cycles | ||||
| Battaglia et al27 | 26.5 (2.4) | Ethinylestradiol/drospirenone (30 μg/3 mg) | 22 | Ethinylestradiol/etonogestrel (15 μg/120 μg) | 21 | ? | 3 (7.0) | 6 | Parallel, single blinded | Complete cases | BDI |
| Davis et al26 | 30.5 (7.4) | Oestradiol/dienogest (3;2;2;1 mg/0;2;3;0 mg) | 97 | Ethinylestradiol/levonorgestrel (30 μg/150 μg) | 103 | 0.0 | 13 (6.1) | 4 | Parallel, double blinded | Complete cases | PGWBI Depression |
| Elaut et al19 | 23.1 (4.3) | Ethinylestradiol/etonogestrel (15 μg/120 mg), desogestrel (75 μg) | 55 55 |
Ethinylestradiol/desogestrel (20 μg/150 μg) | 55 | ? | 40 (24.2) | 3 | Cross-over, unblinded | Intention to treat | SCL-90 Depression |
| Engman et al24 | 24.9 (4.2) | Ethinylestradiol/levonorgestrel (30 μg/150 μg) | 14 | Placebo | 15 | 0.0 | 4 (11.4) | 1 | Parallel, double blinded | Complete cases | MADRS |
| Graham et al, sample a20 | 32.4 (3.7) | Ethinylestradiol/levonorgestrel (30 μg/150 μg), levonorgestrel (30 μg) | 24 22 |
Placebo | 25 | 1.3 | 4 (5.3) | 3 | Parallel, double blinded | Complete cases | BDI |
| Graham et al, sample b20 | 32.2 (4.2) | Ethinylestradiol/levonorgestrel (30 μg/150 μg), levonorgestrel (30 μg) | 24 25 |
Placebo | 25 | 33.0 | 1 (1.3) | 3 | Parallel, double blinded | Complete cases | BDI |
| Greco et al7 | 19.8 (1.8) | Ethinylestradiol/norgestimate (35 μg/0.18;0.22;0.25 mg) | 31 | Ethinylestradiol/norgestimate (25 μg/0.18;0.22;0.25 mg) | 29 | ? | 12 (20.0) | 3 | Parallel, single blinded | Complete cases | BDI |
| Kelly et al8 | 26.5 (6.1) | Ethinylestradiol/drospirenone (30 μg/3 mg) | 193 | Ethinylestradiol/levonorgestrel (30 μg/150 μg) | 87 | ? | 144 (34.0) | 6 | Parallel, single blinded | Complete cases | MDQ Negative |
| Legro et al30 | 27.2 (5.1) | Ethinylestradiol/norethindrone acetate (20 μg/1 mg continuous) | 31 | Ethinylestradiol/norethindrone acetate (20 μg/1 mg) | 31 | ? | 11 (17.7) | 6 | Parallel, double blinded | Intention to treat | MDQ Negative |
| Lundin et al21 | 24.3 (4.2) | Oestradiol/nomegestrol acetate (1.5 mg/2.5 mg) | 80 | Placebo | 88 | 17.8 | 24 (11.9) | 3 | Parallel, double blinded | Complete cases | MADRS |
| O'Connell et al28 | 16.8 (2.1) | Ethinylestradiol/levonorgestrel (20 μg/100 μg) | 34 | Placebo | 34 | ? | 8 (10.5) | 3 | Parallel, double blinded | Complete cases | CES-D |
| Sangthawan and Taneepanichskul9 | 26.8 (4.5) | Ethinylestradiol/drospirenone (30 μg/3 mg) | 50 | Ethinylestradiol/levonorgestrel (30 μg/150 μg) | 49 | ? | 5 (4.8) | 6 | Parallel, unblinded | Complete cases | WHAQ Negative |
| Winkler et al23 | 28.3 (?) | Ethinylestradiol/desogestrel (20 μg/150 μg) | 404 | Ethinylestradiol/levonorgestrel (20 μg/100 μg) | 384 | 7.6 | 239 (23.3) | 6 | Parallel, unblinded | Complete cases | PGWBI Depression |
| Witjes et al22 | 27.6 (7.0) | Oestradiol/nomegestrol acetate (1.5 mg/2.5 mg) | 2631 | Ethinylestradiol/drospirenone (30 μg/3 mg) | 891 | ? | ? | 3 | Parallel, unblinded | Intention-to-treat | MDQ |
| Zethraeus et al25 | 23.7 (3.6) | Ethinylestradiol/levonorgestrel (30 μg/150 μg) | 162 | Placebo | 167 | ? | 11 (3.2) | 3 | Parallel, double blinded | Complete cases | BDI |
RCT, randomised clinical trial; BDI, Beck Depression Inventory; CES-D, Center for Epidemiologic Studies Depression Scale; MADRS, Montgomery-Äsberg Depression Rating Scale; MDQ neg, Menstrual Distress Questionnaire negative affect subscale; PGWBI dep, Psychological General Well-Being Index depression subscale; SCL-90 dep, Symptom Checklist 90 depression subscale; WHAQ neg, Women's Health Assessment Questionnaire negative affect subscale. Ethinylestradiol/dienogest (3;2;2;1 mg/0;2;3;0 mg) is a multiphasic combined oral contraceptive that has four different dosages of hormones throughout a 4-week cycle. The dosages before the ‘slash’ refer to the different dosages of ethinylestradiol, and the ones after the ‘slash’ refer to those of dienogest. The study by Graham et al included two different samples, of which sample a was from the Philippines and sample b was from Scotland.
Number of women who were available for the analysis (excluding those who dropped out) in case of complete-case analyses.
Percentage of the baseline sample that had no experience with hormonal contraceptive use before participating in the trial. As the majority of studies ran complete-case analyses, it is unknown what percentage of women included in the meta-analyses had previous hormonal contraceptive experience.