Ethinyl Estradiol; Norgestrel

What it is

This is a combined oral contraceptive pill. Each active tablet of the common US product contains 0.3 mg norgestrel and 0.03 mg ethinyl estradiol; norgestrel is the progestin and ethinyl estradiol the synthetic oestrogen. It is taken as 21 active tablets followed by 7 inert tablets. Norgestrel is the racemic mixture whose active half is levonorgestrel, so the epidemiology on levonorgestrel-containing pills is the closest evidence to this product.

What the research says

It prevents pregnancy, mainly by stopping ovulation. In real use about 7% of pill users become pregnant within a year; with perfect use the estimate is about 0.3%. Beyond contraception, randomised trials show it reduces period pain, and large pooled epidemiology shows long-term users have substantially less ovarian and endometrial cancer for decades afterwards. Against that, cohort studies show roughly a doubling of the (low) absolute risk of venous thromboembolism, a smaller increase in stroke and heart attack, a small increase in breast cancer while using and shortly after, and an increase in cervical cancer in current users that fades after stopping. Smoking over 35 is the subject of the boxed warning.

Evidence grade: Well established.

How it works

Drug class: Combined oral contraceptive: a synthetic oestrogen (ethinyl estradiol) with a second-generation progestin (norgestrel, the racemic form of levonorgestrel)

Combined oral contraceptives work mainly by stopping the ovary from releasing an egg. The label states the mechanism in one sentence; the synthetic oestrogen and progestin together suppress the pituitary signals that drive ovulation. (Source 1)

Boxed warning

WARNING: CIGARETTE SMOKING AND SERIOUS CARDIOVASCULAR EVENTS Cigarette smoking increases the risk of serious cardiovascular events from combination oral contraceptive (COC) use. This risk increases with age, particularly in women over 35 years of age, and with the number of cigarettes smoked. For this reason, COCs are contraindicated in women who are over 35 years of age and smoke [see Contraindications].

(Source 2)

What it is used for

• In US national survey data the pill had a 12-month failure rate of 7%. A review of the literature on perfect use set the perfect-use estimate for the pill at 0.3%. The product's own trials, in 1,287 women over 11,085 cycles, reported about 1 pregnancy per 100 women-years. Evidence: established. (Source 3)
• A Cochrane review of 21 RCTs in 3723 women found combined pills reduce pain in women with dysmenorrhoea more effectively than placebo, a moderate reduction rated high-quality evidence (SMD -0.58, 95% CI -0.74 to -0.41). The same review found they probably increase the risk of any adverse events (RR 1.31, 95% CI 1.20 to 1.43) and increase irregular bleeding (RR 2.63, 95% CI 2.11 to 3.28). Evidence: established. (Source 4)
• A Cochrane review of 31 trials in 12,579 participants found that in all nine analysable placebo-controlled trials combined pills reduced acne lesion counts and severity. For a levonorgestrel-containing pill - the closest to norgestrel - total lesion counts were lower than placebo (MD -9.98; 95% CI -16.51 to -3.45). The review's own conclusion is that few important and consistent differences were found between pill types, and that how combined pills compare with other acne treatments is unknown. This specific product is not labelled for acne. Evidence: established. (Source 5)
• A collaborative reanalysis of 45 epidemiological studies, with individual data for 23,257 women with ovarian cancer and 87,303 without, found risk fell the longer women had used oral contraceptives, with protection persisting more than 30 years after stopping. In high-income countries 10 years of use was estimated to cut incidence before age 75 from 1.2 to 0.8 per 100 users. This is observational evidence, so it is an association, not proof of cause. Evidence: established. (Source 6)
• An individual-participant meta-analysis of 36 epidemiological studies with 27 276 women with endometrial cancer found a risk ratio of 0·76 per 5 years of use, persisting more than 30 years after use had ceased. Ten years of use was estimated to reduce the absolute risk before age 75 years from 2·3 to 1·3 per 100 women. Observational evidence only. Evidence: established. (Source 7)

Interactions

• St John's wort (Hypericum perforatum) (pharmacokinetic study): In a randomised crossover trial St John's wort did not change ethinyloestradiol levels but cut the progestin's AUC by about 43.9% and raised breakthrough bleeding from 35% to 77-88% of cycles. The authors concluded this could enhance the risk of unintended pregnancy. Note the trial used desogestrel, not norgestrel. (Source 8)
• St John's wort products, as the label records it (label): The label groups St John's wort with enzyme-inducing drugs that can lower pill hormone levels and reduce effectiveness, and advises back-up contraception continuing for 28 days after the inducer is stopped. (Source 9)
• Enzyme-inducing medicines (rifampicin, carbamazepine, phenytoin, topiramate, barbiturates, griseofulvin and others) (label): These induce CYP3A4 and other enzymes, lowering the pill's hormone levels, which can cause breakthrough bleeding and contraceptive failure. The label advises an alternative or back-up method, continued for 28 days after stopping the inducer. (Source 9)
• Vitamin C (ascorbic acid) and paracetamol/acetaminophen (label): Both may raise plasma ethinyl estradiol levels, the label says probably by blocking its conjugation. No clinical consequence is quantified. (Source 10)
• Grapefruit juice (label): The label lists grapefruit juice among CYP3A4 inhibitors that may increase plasma hormone concentrations. No outcome is attached to this; it is a label statement, not a trial result. (Source 10)
• Atorvastatin and rosuvastatin (label): Co-administration raises ethinyl estradiol exposure by roughly 20 to 25%. (Source 10)
• NSAIDs (ibuprofen, diclofenac and similar painkillers) (clinical trial): A national cohort found NSAID use was linked to extra venous thromboembolic events in the first week, and the excess was larger in women on higher-risk hormonal contraception than in women on none. (Source 11)
• Hepatitis C regimens containing ombitasvir/paritaprevir/ritonavir (clinical trial): Liver enzyme rises above five times the upper limit of normal were significantly more frequent in women on ethinyl estradiol-containing medicines, and the label says to stop the pill before starting that regimen. (Source 12)

Stopping it

• There is no withdrawal syndrome and no taper. The label records that the thromboembolic risk from combined pills gradually disappears after use is discontinued, and that risk is highest in the first year of use and when restarting after a break of 4 weeks or longer. (Source 13)
• Periods may be slow to return. In the product's trials 9% of participants reported amenorrhoea in one or more cycles, and the label notes some women experience amenorrhoea or oligomenorrhoea after stopping, especially if that was the case before starting. (Source 14)
• Fertility is not reported to be impaired by having used it, though the evidence is weaker than it sounds: a meta-analysis found 83.1% of women who stopped contraception conceived within the first 12 months, with no significant effect of how long they had used the pill. There was no comparison group of women who had never used contraception, and the sentence quoted here is the review authors' conclusion rather than a measured comparison. (Source 15)
• The cervical cancer excess fades after stopping, returning to never-user levels by 10 or more years. (Source 16)
• The breast cancer excess does not disappear immediately: among women who had used hormonal contraception for 5 years or more, risk remained higher after discontinuation than in women who had never used it. (Source 17)
• The ovarian cancer protection outlasts use by decades: the reduction persisted for more than 30 years after use had ceased, though it attenuated over time. (Source 6)

What goes wrong

Current use of a combined oral contraceptive roughly doubles the absolute risk of venous thromboembolism, from about 3 to about 6 events per 10 000 woman-years. (Source 18)

• Cohort study, Moderate certainty.
• Size: 10.4 million woman years; 4213 venous thrombotic events.
• Who: Danish women aged 15-49 with no history of cardiovascular or malignant disease, 1995-2005.
• How long: national cohort follow-up, Denmark.
• Result: Absolute risk per 10 000 woman years: 3.01 in non-users and 6.29 in current users. Rate ratio fell with duration of use (<1 year 4.17, 95% confidence interval 3.73 to 4.66, 1-4 years 2.98, 2.73 to 3.26, and >4 years 2.76, 2.53 to 3.02; P<0.001).
• Funding: not stated in this passage.

Limit of this finding: This is a national prescription-and-diagnosis registry cohort, not a trial: the figures are associations between what women were prescribed and what was later recorded, and prescribing is not random. The absolute numbers are the ones to hold on to - about 3 events per 10 000 woman-years without the pill and about 6 with it - because a rate ratio looks alarming next to a small absolute risk. Two defects come from the source feed and are reproduced faithfully in the passage: '75 mug desogestrel' is a mangling of 75 micrograms, and 'venous thrombembolism' is missing an o.

The overall absolute risk of venous thrombosis per 10 000 woman years in non-users of oral contraceptives was 3.01 and in current users was 6.29.

Current use of a 30-40 mcg ethinyl estradiol pill with levonorgestrel was associated with higher rates of thrombotic stroke and myocardial infarction, against low background absolute rates. (Source 19)

• Cohort study, Moderate certainty.
• Size: 1,626,158 women; 14,251,063 person-years; 3311 thrombotic strokes and 1725 myocardial infarctions.
• Who: Danish non-pregnant women aged 15-49 with no cardiovascular disease or cancer.
• How long: 15 years.
• Result: Background rates were 21.4 thrombotic strokes and 10.1 myocardial infarctions per 100,000 person-years. With 30 to 40 mcg ethinyl estradiol plus levonorgestrel the relative risks were 1.7 (1.4 to 2.0) for stroke and 2.0 (1.6 to 2.5) for myocardial infarction.
• Funding: Funded by the Danish Heart Association.

Limit of this finding: The progestin in these estimates is levonorgestrel, not the norgestrel in this product, though norgestrel is the racemate whose active half is levonorgestrel, so this is the nearest evidence available. Keep the absolute rates in view: thrombotic stroke ran at 21.4 and myocardial infarction at 10.1 per 100,000 person-years in this population, so a relative risk of 1.7 is a small addition to a small number. These are registry associations, not trial effects.

levonorgestrel, 1.7 (1.4 to 2.0) and 2.0 (1.6 to 2.5); norgestimate, 1.5 (1.2 to 1.9) and 1.3 (0.9 to 1.9)

Current or recent hormonal contraceptive use was associated with a 20% higher relative risk of breast cancer, equal to about 13 extra cancers per 100,000 person-years. (Source 17)

• Cohort study, Moderate certainty.
• Size: 1.8 million women, 19.6 million person-years, 11,517 breast cancers.
• Who: Danish women aged 15-49 with no prior cancer, venous thromboembolism or infertility treatment.
• How long: mean follow-up 10.9 years.
• Result: Relative risk 1.20 (95% CI 1.14 to 1.26), rising from 1.09 with under 1 year of use to 1.38 with more than 10 years. Absolute increase 13 (95% CI 10 to 16) per 100,000 person-years, about 1 extra breast cancer for every 7690 women using hormonal contraception for 1 year. Risk estimates for individual oral combination products varied between 1.0 and 1.6.
• Funding: Funded by the Novo Nordisk Foundation.

Limit of this finding: This counts all hormonal contraception together, coils and progestin-only methods included, and not norgestrel pills specifically; the figures for individual oral combination products are given only as a range of 1.0 to 1.6. It is a registry cohort, so the association is not proof of cause, and the absolute figure - about 13 extra breast cancers per 100,000 person-years, or one for every 7690 women using hormonal contraception for a year - is what tells you the size of the risk.

The overall absolute increase in breast cancers diagnosed among current and recent users of any hormonal contraceptive was 13 (95% CI, 10 to 16) per 100,000 person-years, or approximately 1 extra breast cancer for every 7690 women using hormonal contraception for 1 year.

Cervical cancer risk is higher in current users and rises with duration of use, but returns to that of never-users about 10 years after stopping. (Source 16)

• Case-control study, Moderate certainty.
• Size: Individual data for 16,573 women with cervical cancer and 35,509 without cervical cancer.
• Who: Women worldwide, including those positive for high-risk human papillomavirus.
• How long: pooled lifetime exposure histories.
• Result: Relative risk for 5 or more years' use versus never use, 1.90 (95% CI 1.69-2.13). Ten years' use from around age 20 to 30 was estimated to raise cumulative incidence by age 50 from 3.8 to 4.5 per 1000 in more developed countries and from 7.3 to 8.3 per 1000 in less developed countries.
• Funding: not stated in this passage.

relative risk for 5 or more years' use versus never use, 1.90 [95% CI 1.69-2.13]). The risk declined after use ceased, and by 10 or more years had returned to that of never users.

In the dysmenorrhoea trials combined pills increased the risk of irregular bleeding, of any adverse event, of headache and of nausea compared with placebo. (Source 20)

• Systematic review, High certainty.
• Size: 7 RCTs, 1025 women.
• Who: Women with primary dysmenorrhoea.
• How long: not stated in the abstract.
• Result: Any adverse events RR 1.31, 95% CI 1.20 to 1.43; I² = 79%; 7 RCTs, 1025 women; moderate-quality evidence. Serious adverse events RR 1.77, 95% CI 0.49 to 6.43; low-quality evidence. Irregular bleeding RR 2.63, 95% CI 2.11 to 3.28; I² = 29%; 7 RCTs, 1025 women; high-quality evidence. In women with a risk of irregular bleeding of 18% if using placebo or no treatment, the risk would be between 39% and 60% if using combined OCP.
• Funding: not stated in the abstract.

Women who received OCPs had an increased risk of irregular bleeding compared to women who received placebo or no treatment (RR 2.63, 95% CI 2.11 to 3.28; I² = 29%; 7 RCTs, 1025 women; high-quality evidence)

Combined oral contraceptive use was associated with a higher rate of first antidepressant prescription, with the largest relative risks in adolescents. (Source 21)

• Cohort study, Low certainty.
• Size: 1 061 997 women and adolescents.
• Who: Danish women and adolescents aged 15 to 34 with no prior depression diagnosis, antidepressant prescription, major psychiatric diagnosis, cancer, venous thrombosis or infertility treatment.
• How long: mean follow-up 6.4 years.
• Result: Combined oral contraceptive users RR of first antidepressant use 1.23 (95% CI, 1.22-1.25). Adolescents aged 15-19 using combined oral contraceptives RR 1.8 (95% CI, 1.75-1.84). The relative risk peaked at 1.4 six months after starting. Using never-users as the reference raised the combined pill estimate to 1.7.
• Funding: not stated in the abstract.

Limit of this finding: What was counted is a first antidepressant prescription or a first hospital diagnosis of depression in registry records, not an assessment of mood; women who start the pill also see a doctor, which raises the chance of either being recorded. The study cannot establish cause, and the authors' own suggestion that depression may be an adverse effect of hormonal contraception is their interpretation - it is recorded on this screen as one side of a disagreement, not as a result.

Compared with nonusers, users of combined oral contraceptives had an RR of first use of an antidepressant of 1.23 (95% CI, 1.22-1.25).

In the product's own trials weight increase was the most common adverse reaction and 8% of participants stopped early because of an adverse reaction. (Source 22)

• Official position, Low certainty.
• Size: 1,343 women across 9 clinical trials, 11,085 cycles.
• Who: Healthy women of child-bearing potential aged 15 to 40.
• How long: 429 women completed one year.
• Result: Weight increase 11%, cervical erosion 9%, weight decrease 6%, acne 4%, dysmenorrhea 4%, vaginal discharge 4%, abdominal pain/cramps/bloating 3%, appetite increase 3%, depression 3%, nervousness 3%, chloasma/melasma 2%, fatigue 2%, aggravation of varicose veins 2%. 8% discontinued prematurely due to an adverse reaction.
• Funding: manufacturer trials reported in the label.

A total of 8% of subjects discontinued the trials prematurely due to an adverse reaction, most commonly due to unscheduled bleeding, spotting, headache (including migraine), nausea, acne, changes in menstrual flow, weight increase, nervousness, high blood pressure, and depression.

Hepatic adenomas are associated with combined oral contraceptive use, with an estimated attributable risk of 3.3 cases per 100,000 users, and can rupture fatally. (Source 23)

• Official position, Certainty not rated.
• Size: not stated.
• Who: Women using combined oral contraceptives.
• How long: not stated.
• Result: Estimated attributable risk 3.3 cases/100,000 users. Rupture may cause death through intra-abdominal haemorrhage.
• Funding: not applicable.

Hepatic adenomas are associated with COC use. An estimate of the attributable risk is 3.3 cases/100,000 users.

Starting an NSAID was associated with a steep relative rise in venous thromboembolism in this cohort - incidence rate ratios of 7.2 without hormonal contraception rising to 11.0 with a high-risk hormonal contraceptive - while the absolute excess over the first week was 4 to 23 extra events per 100 000 women. (Source 24)

• Cohort study, Low certainty.
• Size: 2.0 million women followed for 21.0 million person years; 8710 venous thromboembolic events.
• Who: Danish women aged 15-49 with no prior thrombotic event, cancer, thrombophilia or infertility treatment, 1996-2017.
• How long: first week of NSAID treatment.
• Result: Adjusted incidence rate ratio of venous thromboembolism with NSAID use: 7.2 (95% confidence interval 6.0 to 8.5) in women not using hormonal contraception, 11.0 (9.6 to 12.6) with high risk hormonal contraception, 7.9 (5.9 to 10.6) with medium risk and 4.5 (2.6 to 8.1) with low/no risk. The absolute excesses over the first week of NSAID treatment, stated beside each of those by the paper itself, were 4 (3 to 5), 23 (19 to 27), 11 (7 to 15) and 3 (0 to 5) extra events per 100 000 women respectively.
• Funding: not stated in the abstract.

Limit of this finding: A rate ratio of 7 or 11 sounds enormous and the underlying risk is very small, so the two have to be read together: the measured excess was 4 extra clots per 100 000 women over a week of NSAID use without hormonal contraception and 23 per 100 000 with a high-risk one. This is a registry cohort, so these are associations rather than effects of the drugs, and NSAIDs are often prescribed for conditions that themselves raise clot risk.

Compared with non-use of NSAIDs, use of NSAIDs was associated with an adjusted incidence rate ratio of venous thromboembolism of 7.2 (95% confidence interval 6.0 to 8.5) in women not using hormonal contraception, 11.0 (9.6 to 12.6) in women using high risk hormonal contraception, 7.9 (5.9 to 10.6) in those using medium risk hormonal contraception, and 4.5 (2.6 to 8.1) in users of low/no risk hormonal contraception. The corresponding numbers of extra venous thromboembolic events per 100 000 women over the first week of NSAID treatment compared with non-use of NSAIDs were 4 (3 to 5) in women not using hormonal contraception, 23 (19 to 27) in women using high risk hormonal contraception, 11 (7 to 15) in those using medium risk hormonal contraception, and 3 (0 to 5) in users of low/no risk hormonal contraception.

St John's wort taken with a low-dose combined pill did not change ethinyl estradiol levels or stop ovulation suppression, but sharply increased breakthrough bleeding and cut progestin levels. (Source 8)

• Randomized trial, Low certainty.
• Size: 17 women analysed per cycle (13/17, 15/17 and 6/17 reported bleeding)
• Who: Healthy female volunteers on a low-dose combined pill with electronically monitored adherence.
• How long: three cycles (control, then St John's wort twice and three times daily)
• Result: Intracyclic bleeding in 13/17 (77%, P < 0.015) and 15/17 (88%, P < 0.001) versus 6/17 (35%) on the pill alone. Ethinyloestradiol AUC and Cmax unchanged. 3-ketodesogestrel AUC fell from 31.2 to 17.7 ng ml−1 h (43.9%; P = 0.001).
• Funding: not stated in the abstract.

Limit of this finding: The pill used in this trial was ethinyloestradiol 0.02 mg with desogestrel, not the norgestrel in this product, and the progestin level that fell was desogestrel's active metabolite; no norgestrel-specific interaction trial was found. The paper also prints two figures for the same result - cycle A bleeding is 77% (13 of 17 women) in the abstract and 76% in the discussion - and the figure quoted here is the abstract's 77%.

However, significantly more subjects reported intracyclic bleeding during cycles A (13/17 (77%), P < 0.015) and cycle B (15/17 (88%), P < 0.001) than with oral contraceptives alone (6/17 (35%)).

What the evidence supports

In US survey data the pill's typical-use failure rate was 7% in the first year, with most failures early in the year. (Source 3)

• Cohort study, Moderate certainty.
• Size: failure probabilities for the five most commonly used reversible methods.
• Who: US women aged 15-44 in the 2006-2010 National Survey of Family Growth.
• How long: first 12 months of use.
• Result: Pill failure rate 7%; injectable 4%; male condom 13%; withdrawal 20%; LARC 1%. 32% of annual failures occur in the first three months and 63% in the first six months.
• Funding: not stated in this passage.

The male condom had the next highest probability of failure (13%); the pill and the injectable had failure rates of 7% and 4%, respectively.

Longer oral contraceptive use was associated with progressively lower ovarian cancer risk, and the reduction persisted for more than 30 years after stopping. (Source 6)

• Case-control study, Moderate certainty.
• Size: 23,257 women with ovarian cancer and 87,303 controls from 45 studies in 21 countries.
• Who: Women in 21 countries; median year of diagnosis 1993.
• How long: average use 4.4 years in cases and 5.0 years in controls.
• Result: Proportional risk reductions per 5 years of use were 29% (95% CI 23-34%) if use ceased less than 10 years previously, 19% (14-24%) at 10-19 years and 15% (9-21%) at 20-29 years. In high-income countries 10 years of use was estimated to reduce incidence before age 75 from 1.2 to 0.8 per 100 users and mortality from 0.7 to 0.5 per 100.
• Funding: not stated in this passage.

In high-income countries, 10 years use of oral contraceptives was estimated to reduce ovarian cancer incidence before age 75 from 1.2 to 0.8 per 100 users and mortality from 0.7 to 0.5 per 100

Every 5 years of oral contraceptive use was associated with a lower endometrial cancer risk, a risk ratio of 0·76, persisting beyond 30 years after stopping. (Source 7)

• Meta-analysis, Moderate certainty.
• Size: 27 276 women with endometrial cancer and 115 743 controls from 36 studies.
• Who: Women in 36 epidemiological studies; median age of cases 63 years.
• How long: median durations of 3·0 years (IQR 1-7) in cases and 4·4 years (IQR 2-9) in controls.
• Result: Every 5 years of use was associated with a risk ratio of 0·76 (95% CI 0·73-0·78; p<0·0001). In high-income countries, 10 years use of oral contraceptives was estimated to reduce the absolute risk of endometrial cancer arising before age 75 years from 2·3 to 1·3 per 100 women. Stronger for carcinomas (RR 0·69, 95% CI 0·66-0·71) than sarcomas (0·83, 0·67-1·04).
• Funding: Medical Research Council, Cancer Research UK.

every 5 years of use was associated with a risk ratio of 0·76 (95% CI 0·73-0·78; p<0·0001)

Combined pills reduce period pain compared with placebo, but increase adverse events and irregular bleeding. (Source 4)

• Systematic review, High certainty.
• Size: 21 RCTs, 3723 women; pain outcome 6 RCTs, 588 women.
• Who: Women with primary dysmenorrhoea and no pelvic pathology.
• How long: not stated in the abstract.
• Result: Pain standardised mean difference -0.58, 95% CI -0.74 to -0.41; I² = 28%; 6 RCTs, 588 women; high-quality evidence. Pain improvement as a dichotomous outcome RR 1.65, 95% CI 1.29 to 2.10; low-quality evidence.
• Funding: not stated in the abstract.

the result can be interpreted as a moderate reduction in pain (standardised mean difference (SMD) -0.58, 95% confidence interval (CI) -0.74 to -0.41; I² = 28%; 6 RCTs, 588 women; high-quality evidence)

A levonorgestrel-containing combined pill reduced total acne lesion counts compared with placebo. (Source 5)

• Systematic review, Moderate certainty.
• Size: 31 trials, 12,579 participants; 9 analysable placebo-controlled trials.
• Who: Women with facial acne.
• How long: not stated in the abstract.
• Result: Levonorgestrel-COC total lesion counts MD -9.98; 95% CI -16.51 to -3.45. Norgestimate COC MD -9.32; 95% CI -14.19 to -4.45. Norethindrone acetate COC clinician global assessment OR 1.86; 95% CI 1.32 to 2.62.
• Funding: not stated in the abstract.

Limit of this finding: Two limits. The progestin tested here is levonorgestrel, not the norgestrel in this product - norgestrel is the racemic mixture of which levonorgestrel is the active half, so this is the nearest evidence rather than evidence about this pill. And the review's own conclusion is that few important and consistent differences were found between pill types for acne, so a good number for one progestin is not a reason to prefer it. The review's searches ended in January 2012.

A levonorgestrel-COC group had fewer total lesion counts (MD -9.98; 95% CI -16.51 to -3.45), inflammatory and non-inflammatory lesion counts

What the evidence does not support

Randomised trials do not support the widespread belief that combined contraceptives cause weight gain, though the evidence is too thin to settle it. (Source 25)

• Systematic review, Low certainty.
• Size: 49 trials with 85 weight change comparisons across 52 contraceptive pairs; only 4 trials had a placebo or no-intervention group.
• Who: Women using combination contraceptives in randomised trials of at least three cycles.
• How long: at least three treatment cycles.
• Result: The four placebo-controlled or no-intervention trials found no evidence supporting a causal association with weight change. Most comparisons between different combination contraceptives showed no substantial difference in weight.
• Funding: not stated in the abstract.

The four trials with a placebo or no intervention group did not find evidence supporting a causal association between combination oral contraceptives or a combination skin patch and weight change.

Stopping contraception, including the pill, is not followed by delayed fertility: most women conceive within a year. (Source 26)

• Meta-analysis, Low certainty.
• Size: Twenty two studies that enrolled a total of 14,884 women who discontinued contraception.
• Who: Women stopping hormonal and non-hormonal contraception, studies published 1985-2017.
• How long: first 12 months after discontinuation.
• Result: Pooled pregnancy rate 83.1% (95% CI = 78.2-88%) within the first 12 months. Not significantly different for hormonal methods and IUD users; duration of oral-contraceptive use did not significantly influence return of fertility.
• Funding: not stated in the abstract.

Limit of this finding: This is a pooled rate, not a comparison. The review reports how many women conceived within a year of stopping contraception and has no group of women who never used contraception to compare them with, so it can say that most women conceived within twelve months and it cannot by itself show that contraception has no effect on fertility. The review is thin for a claim of that weight: two authors, no registered protocol stated in the abstract, searches ending in 2017.

The pooled rate of pregnancy was 83.1% (95% CI = 78.2-88%) within the first 12 months of contraceptive discontinuation.

The same Cochrane review concluded that few important and consistent differences were found between combined pill types for acne, and that how combined pills compare with other acne treatments is unknown. (Source 27)

• Systematic review, Moderate certainty.
• Size: 31 trials with 12,579 participants; only one trial compared a combined pill with an alternative acne treatment.
• Who: Women with facial acne.
• How long: not stated in the abstract.
• Result: No between-type effect estimate is offered. Six combined pills were effective against placebo; comparisons between pill types produced no consistent winner.
• Funding: not stated in the abstract.

Limit of this finding: The version of this passage we first recorded stopped in the middle of the review's results, just after a run of favourable numbers for particular progestins, and left out both the rest of the results and this conclusion. Read whole, the review says combined pills beat placebo for acne and does not establish that any one type beats another.

Few important and consistent differences were found between COC types in their effectiveness for treating acne. How COCs compare to alternative acne treatments is unknown since only one trial addressed this issue.

Where the evidence is mixed

The 0% in this paper is the lowest rate reported by any single study of typical pill use, not an estimate of how often the pill fails in ordinary use; the author's own perfect-use estimate is 0.3%. (Source 28)

• Expert review, not systematic, Low certainty.
• Size: not applicable; a review of reported failure rates.
• Who: US women using the combined pill.
• How long: first year of use.
• Result: Perfect-use estimate set at 0.3%. The lowest pregnancy rate reported in any single study of typical use was 0%, and the author notes that pregnancies do occur, albeit rarely, during perfect use. The lowest reported rate for the progestin-only pill exceeds 1%.
• Funding: not stated.

Limit of this finding: Lifted out of its paragraph, 'the lowest reported pregnancy rate for the combined pill during typical use is 0%' reads as if the pill never fails in ordinary use. It does not mean that. It is the single lowest figure among the studies the author reviewed, quoted only to explain why he did not set the perfect-use estimate at zero. The typical-use figure to read beside it is 7% in the first 12 months, from the 2006-2010 National Survey of Family Growth; the same author's own Table 1, built from the 1995 and 2002 surveys, gives 9%. Both are far above 0%.

Hence, we set the perfect-use estimate for the pill at the very low level of 0.3%.

Pills containing levonorgestrel - the racemate of which is the norgestrel in this product - carried a lower venous thrombosis rate than pills with desogestrel, gestodene, drospirenone or cyproterone at the same oestrogen dose and length of use. (Source 18)

• Cohort study, Moderate certainty.
• Size: 10.4 million woman years; 4213 venous thrombotic events.
• Who: Danish women aged 15-49.
• How long: national cohort follow-up, Denmark.
• Result: Compared with levonorgestrel at the same oestrogen dose and length of use: norethisterone 0.98 (0.71 to 1.37), norgestimate 1.19 (0.96 to 1.47), desogestrel 1.82 (1.49 to 2.22), gestodene 1.86 (1.59 to 2.18), drospirenone 1.64 (1.27 to 2.10), cyproterone 1.88 (1.47 to 2.42).
• Funding: not stated in this passage.

Limit of this finding: Norgestrel itself was not one of the comparison groups. Norgestrel is the racemic mixture whose active half is levonorgestrel, so these levonorgestrel estimates are the nearest available evidence for this product rather than evidence about it. Note also what is being compared: every ratio here is one progestin against the levonorgestrel pills, not against taking no pill.

Compared with oral contraceptives containing levonorgestrel and with the same dose of oestrogen and length of use, the rate ratio for oral contraceptives with norethisterone was 0.98 (0.71 to 1.37), with norgestimate 1.19 (0.96 to 1.47), with desogestrel 1.82 (1.49 to 2.22), with gestodene 1.86 (1.59 to 2.18), with drospirenone 1.64 (1.27 to 2.10), and with cyproterone 1.88 (1.47 to 2.42).

A separate cohort in postpartum women found the direction of the depression association depends on the method, with progestin-only pills associated with lower risk. (Source 29)

• Cohort study, Low certainty.
• Size: 75,528 postpartum women enrolled in the US military medical system.
• Who: Postpartum women in the US military medical system who delivered between October 2012 and September 2014, excluding those with recent antidepressant use or depression.
• How long: the first 12months postpartum.
• Result: Antidepressants prescribed to 7.8% and depression diagnosed in 5.0%. Etonogestrel implant adjHR 1.22 (95%CI 1.06-1.41); ring 1.45 (1.16-1.80). Norethindrone-only pills were associated with lower antidepressant use 0.58 (0.52-0.64) and lower depression diagnosis 0.56 (0.49-0.64).
• Funding: not stated in the abstract.

Limit of this finding: This is an analysis of insurance claims, so what was measured is a recorded diagnosis or an antidepressant prescription, not a clinical assessment of major depressive disorder - the phrase 'major depression' appears only in the authors' conclusion, not in what they measured. Note also which methods the estimates belong to: the lower risks are for norethindrone-only pills and a levonorgestrel coil, the higher ones for etonogestrel implants and rings, none of them this product.

Use of norethindrone-only pills was associated with a lower risk of antidepressant use (0.58(0.52-0.64), p<0.001) and depression diagnosis (0.56(0.49-0.64), p<0.001).

Trussell's typical-use failure figures for the pill are a weighted average of two US national household surveys, corrected for under-reporting of abortion, not a trial result. (Source 30)

• Expert review, not systematic, Low certainty.
• Size: not applicable; the derivation note printed under the paper's Table 1.
• Who: US women in the 1995 and 2002 National Survey of Family Growth.
• How long: first year of use.
• Result: No new estimate. The note states which survey each method's typical-use figure is taken from and that each is corrected for under-reporting of abortion.
• Funding: not stated.

Limit of this finding: This matters because typical-use failure rates are survey estimates of what happens to people, not measurements of the drug. Different surveys give different numbers - 9% from the 1995 and 2002 surveys used here, 7% from the 2006-2010 survey quoted elsewhere on this screen - and both are corrected upwards for abortions that respondents did not report.

estimates for fertility awareness-based methods, withdrawal, the male condom, the pill, and Depo-Provera are taken from the 1995 and 2002 National Survey of Family Growth corrected for underreporting of abortion

The study authors concluded that venous thrombosis risk with combined pills falls with longer use and lower oestrogen dose, and that progestogen-only pills and hormone-releasing coils were not associated with any increased risk. (Source 31)

• Cohort study, Moderate certainty.
• Size: the whole Danish cohort; the event counts are in the Results passage.
• Who: Danish women aged 15-49 with no history of cardiovascular or malignant disease.
• How long: national cohort follow-up, Denmark.
• Result: No new estimate; this is the authors' summary of the rate ratios reported in the Results section of the same abstract.
• Funding: not stated in this passage.

Limit of this finding: This is the authors' interpretation, printed under a separate 'Conclusion' heading; in the version we first recorded it ran straight on from the results figures with that heading deleted. 'Not associated with any increased risk' means the study did not detect one, in a dataset where the progestogen-only estimates had wide confidence intervals. That is not the same as showing there is no risk.

For the same dose of oestrogen and the same length of use, oral contraceptives with desogestrel, gestodene, or drospirenone were associated with a significantly higher risk of venous thrombosis than oral contraceptives with levonorgestrel. Progestogen only pills and hormone releasing intrauterine devices were not associated with any increased risk of venous thrombosis.

Where the research disagrees

Whether hormonal contraception causes depression

• Skovlund and colleagues, Danish national cohort (2016), prospective national registry cohort of 1 061 997 women: Use of hormonal contraception, especially among adolescents, was associated with subsequent use of antidepressants and a first diagnosis of depression, suggesting depression as a potential adverse effect of hormonal contraceptive use. (Source 32)
• Horibe and colleagues, US military postpartum cohort (2017), retrospective insurance-records cohort of 75,528 postpartum women, in which norethindrone-only pills were associated with lower risk: The risk of major depression diagnosis and antidepressant use in the postpartum period varies with the type of hormonal contraception used. (Source 33)

Whether combined contraceptives cause weight gain

• Cochrane Fertility Regulation Group (2014), systematic review of 49 randomised trials, only 4 with a placebo or no-intervention arm: Available evidence was insufficient to determine the effect of combination contraceptives on weight, but no large effect was evident. (Source 34)
• The product label (effective 14 September 2026), reporting the manufacturer's own trials, pooled adverse-reaction reporting from 9 manufacturer trials in 1,343 women, with no placebo comparison: Common Adverse Reactions (≥ 2% of women): Weight increase (11%) Cervical erosion (9%) Weight decrease (6%) (Source 22)

How much

• Reference intake: Dosing is set by the prescriber, not by the reader. The label's position is one white active tablet daily for 21 consecutive days followed by one inert tablet daily for 7 consecutive days, taken at the same time each day and at intervals not exceeding 24 hours. (Source 35)
• Upper limit: There is no upper limit in the usual sense: the combined pill is a fixed-dose daily product, and the label states that the dosage is one tablet daily. Each active tablet contains 0.3 mg norgestrel and 0.03 mg ethinyl estradiol. (Source 36)
• Studied: The product's own efficacy studies ran 1,287 women through 11,085 cycles, equal to 852.7 women-years of usage. (Source 37)
• Studied: The St John's wort interaction trial used a lower-oestrogen pill: a low-dose oral contraceptive of 0.02 mg ethinyloestradiol with 0.150 mg desogestrel, against 300 mg St John's wort extract twice or three times daily. (Source 38)
• Studied: The arterial-risk cohort separated pills by oestrogen dose, comparing ethinyl estradiol at 30 to 40 mcg against 20 mcg; this product sits in the 30 to 40 mcg band. (Source 19)

A common belief, and what the research shows

The belief: The pill causes weight gain, and it makes it harder to get pregnant later.

What the research shows: Neither holds up well. On weight, the Cochrane review of 49 trials reports that "The four trials with a placebo or no intervention group did not find evidence supporting a causal association between combination oral contraceptives or a combination skin patch and weight change." - although it also says the evidence is insufficient to settle the question. On fertility, a meta-analysis found "The pooled rate of pregnancy was 83.1% (95% CI = 78.2-88%) within the first 12 months of contraceptive discontinuation." and that duration of pill use did not significantly influence the return of fertility. The real documented harms are elsewhere: venous thromboembolism, arterial events, and small increases in breast and cervical cancer while using.

Questions and answers

What is it?

It is a two-hormone contraceptive tablet. Each white active tablet of the common US product contains 0.3 mg norgestrel, a progestin, and 0.03 mg ethinyl estradiol, a synthetic oestrogen. The pack also contains inactive tablets. Norgestrel is a racemic mixture whose active half is levonorgestrel. (Source 36)

What does it do in the body?

The two hormones together suppress the hormonal signals that trigger ovulation, so no egg is released. The label states the mechanism as suppressing ovulation. Secondary effects on the cervical mucus and the womb lining are usually described as well, but the label itself puts the weight on ovulation. (Source 1)

Is it good or bad for you?

Both, in different measures. It is a highly effective contraceptive and, in observational data, long use is linked to substantially lower ovarian and endometrial cancer for decades. It also raises the absolute risk of venous thromboembolism from about 3 to about 6 events per 10 000 woman-years, raises stroke and heart attack rates modestly, and is linked to small increases in breast and cervical cancer while being used. The balance differs sharply by age, smoking status and clotting history. (Source 18)

How do you get more of it?

It is a prescription medicine, not something you accumulate. The label's position is one active tablet daily for 21 days then one inert tablet daily for 7 days, taken at the same time each day and never more than 24 hours apart. Dose and product choice are the prescriber's decision. (Source 35)

If it is harmful, what reduces it?

You reduce it by stopping, and the main harm reverses. The label records that the thromboembolic risk gradually disappears after use is discontinued. Cervical cancer risk returns to never-user levels about 10 years after stopping. Contraceptive protection, of course, stops immediately. (Source 13)

Why might someone be low in it or missing it?

This is a medicine, not a nutrient, so the question becomes why someone might not be getting the expected effect. The documented reason is drug interaction: enzyme-inducing medicines and St John's wort lower the hormone levels and can cause breakthrough bleeding or contraceptive failure. The label advises a back-up method while an inducer is used and for 28 days after stopping it. Missed or late tablets are the other common reason, which is why typical-use failure is far above perfect-use failure. (Source 9)

Which whole foods contain it or feed it?

No whole food contains it. Both components are synthetic steroids made for the purpose: norgestrel and ethinyl estradiol are defined in the label by their chemical names and are supplied only in manufactured tablets. One food does interact, though: grapefruit juice is listed among CYP3A4 inhibitors that may raise plasma hormone levels. (Source 36)

What happens if you do not have it?

Without it, or without another method, pregnancy becomes likely; in survey data the pill's own failure rate is 7% per year against much higher rates for condoms and withdrawal. Stopping also gives up the non-contraceptive effects. The ovarian and endometrial cancer reductions that come with long use persist for decades after stopping, so they are not lost the moment you stop, but they are only earned by years of use. (Source 3)

How can you test for it?

There is no routine blood test. Hormone levels can be measured in research - the St John's wort trial used liquid chromatography-mass spectrometry to quantify serum ethinyloestradiol and the progestin metabolite - but this is not done in clinical care. In practice effectiveness is judged by absence of pregnancy, and the research measure of ovulation suppression is follicle monitoring with serum oestradiol and progesterone, which that trial also used. (Source 39)

References

1. DailyMed / US FDA structured product label (Teva Pharmaceuticals). CRYSELLE (norgestrel and ethinyl estradiol) tablets - FDA prescribing information. 2026. Read the source
2. DailyMed / US FDA structured product label (Teva Pharmaceuticals). CRYSELLE (norgestrel and ethinyl estradiol) tablets - FDA prescribing information. 2026. Read the source
3. Perspectives on Sexual and Reproductive Health. Contraceptive Failure in the United States: Estimates from the 2006-2010 National Survey of Family Growth. 2017. PMID 28245088, DOI 10.1363/psrh.12017. Read the source
4. Cochrane Database of Systematic Reviews. Combined oral contraceptive pill for primary dysmenorrhoea. 2023. PMID 37523477, DOI 10.1002/14651858.CD002120.pub4. Read the source
5. Cochrane Database of Systematic Reviews. Combined oral contraceptive pills for treatment of acne — abstract, Main results. 2012. PMID 22786490, DOI 10.1002/14651858.CD004425.pub6. Read the source
6. The Lancet. Ovarian cancer and oral contraceptives: collaborative reanalysis of data from 45 epidemiological studies including 23,257 women with ovarian cancer and 87,303 controls — abstract, Findings. 2008. PMID 18294997, DOI 10.1016/S0140-6736(08)60167-1. Read the source
7. The Lancet Oncology. Endometrial cancer and oral contraceptives: an individual participant meta-analysis of 27 276 women with endometrial cancer from 36 epidemiological studies — abstract, Findings. 2015. PMID 26254030, DOI 10.1016/S1470-2045(15)00212-0. Read the source
8. British Journal of Clinical Pharmacology. Interaction of St John's wort with low-dose oral contraceptive therapy: a randomized controlled trial — abstract, Results. 2003. PMID 14616430, DOI 10.1046/j.1365-2125.2003.02005.x. Read the source
9. DailyMed / US FDA structured product label (Teva Pharmaceuticals). CRYSELLE (norgestrel and ethinyl estradiol) tablets - FDA prescribing information. 2026. Read the source
10. DailyMed / US FDA structured product label (Teva Pharmaceuticals). CRYSELLE (norgestrel and ethinyl estradiol) tablets - FDA prescribing information. 2026. Read the source
11. BMJ. Venous thromboembolism with use of hormonal contraception and non-steroidal anti-inflammatory drugs: nationwide cohort study — abstract, Conclusions. 2023. PMID 37673431, DOI 10.1136/bmj-2022-074450. Read the source
12. DailyMed / US FDA structured product label (Teva Pharmaceuticals). CRYSELLE (norgestrel and ethinyl estradiol) tablets - FDA prescribing information. 2026. Read the source
13. DailyMed / US FDA structured product label (Teva Pharmaceuticals). CRYSELLE (norgestrel and ethinyl estradiol) tablets - FDA prescribing information. 2026. Read the source
14. DailyMed / US FDA structured product label (Teva Pharmaceuticals). CRYSELLE (norgestrel and ethinyl estradiol) tablets - FDA prescribing information. 2026. Read the source
15. Contraception and Reproductive Medicine. Return of fertility after discontinuation of contraception: a systematic review and meta-analysis — abstract, Conclusion and recommendation. 2018. PMID 30062044, DOI 10.1186/s40834-018-0064-y. Read the source
16. The Lancet. Cervical cancer and hormonal contraceptives: collaborative reanalysis of individual data for 16,573 women with cervical cancer and 35,509 women without cervical cancer from 24 epidemiological studies. 2007. PMID 17993361, DOI 10.1016/S0140-6736(07)61684-5. Read the source
17. New England Journal of Medicine. Contemporary Hormonal Contraception and the Risk of Breast Cancer — abstract, Results. 2017. PMID 29211679, DOI 10.1056/NEJMoa1700732. Read the source
18. BMJ. Hormonal contraception and risk of venous thromboembolism: national follow-up study — abstract, Results. 2009. PMID 19679613, DOI 10.1136/bmj.b2890. Read the source
19. New England Journal of Medicine. Thrombotic stroke and myocardial infarction with hormonal contraception — abstract, Results. 2012. PMID 22693997, DOI 10.1056/NEJMoa1111840. Read the source
20. Cochrane Database of Systematic Reviews. Combined oral contraceptive pill for primary dysmenorrhoea. 2023. PMID 37523477, DOI 10.1002/14651858.CD002120.pub4. Read the source
21. JAMA Psychiatry. Association of Hormonal Contraception With Depression — abstract, Results. 2016. PMID 27680324, DOI 10.1001/jamapsychiatry.2016.2387. Read the source
22. DailyMed / US FDA structured product label (Teva Pharmaceuticals). CRYSELLE (norgestrel and ethinyl estradiol) tablets - FDA prescribing information. 2026. Read the source
23. DailyMed / US FDA structured product label (Teva Pharmaceuticals). CRYSELLE (norgestrel and ethinyl estradiol) tablets - FDA prescribing information. 2026. Read the source
24. BMJ. Venous thromboembolism with use of hormonal contraception and non-steroidal anti-inflammatory drugs: nationwide cohort study — abstract, Results. 2023. PMID 37673431, DOI 10.1136/bmj-2022-074450. Read the source
25. Cochrane Database of Systematic Reviews. Combination contraceptives: effects on weight — abstract, Main results. 2014. PMID 24477630, DOI 10.1002/14651858.CD003987.pub5. Read the source
26. Contraception and Reproductive Medicine. Return of fertility after discontinuation of contraception: a systematic review and meta-analysis — abstract, Results. 2018. PMID 30062044, DOI 10.1186/s40834-018-0064-y. Read the source
27. Cochrane Database of Systematic Reviews. Combined oral contraceptive pills for treatment of acne — abstract, Authors' conclusions. 2012. PMID 22786490, DOI 10.1002/14651858.CD004425.pub6. Read the source
28. Contraception. Contraceptive failure in the United States. 2011. PMID 21477680, DOI 10.1016/j.contraception.2011.01.021. Read the source
29. Contraception. Association of Hormonal Contraception with depression in the postpartum period — abstract, Results. 2017. PMID 28867443, DOI 10.1016/j.contraception.2017.08.010. Read the source
30. Contraception. Contraceptive failure in the United States — derivation note to Table 1. 2011. PMID 21477680, DOI 10.1016/j.contraception.2011.01.021. Read the source
31. BMJ. Hormonal contraception and risk of venous thromboembolism: national follow-up study — abstract, Conclusion. 2009. PMID 19679613, DOI 10.1136/bmj.b2890. Read the source
32. JAMA Psychiatry. Association of Hormonal Contraception With Depression — abstract, Conclusions and relevance. 2016. PMID 27680324, DOI 10.1001/jamapsychiatry.2016.2387. Read the source
33. Contraception. Association of Hormonal Contraception with depression in the postpartum period — abstract, Conclusion. 2017. PMID 28867443, DOI 10.1016/j.contraception.2017.08.010. Read the source
34. Cochrane Database of Systematic Reviews. Combination contraceptives: effects on weight — abstract, Authors' conclusions. 2014. PMID 24477630, DOI 10.1002/14651858.CD003987.pub5. Read the source
35. DailyMed / US FDA structured product label (Teva Pharmaceuticals). CRYSELLE (norgestrel and ethinyl estradiol) tablets - FDA prescribing information. 2026. Read the source
36. DailyMed / US FDA structured product label (Teva Pharmaceuticals). CRYSELLE (norgestrel and ethinyl estradiol) tablets - FDA prescribing information. 2026. Read the source
37. DailyMed / US FDA structured product label (Teva Pharmaceuticals). CRYSELLE (norgestrel and ethinyl estradiol) tablets - FDA prescribing information. 2026. Read the source
38. British Journal of Clinical Pharmacology. Interaction of St John's wort with low-dose oral contraceptive therapy: a randomized controlled trial — abstract, Methods. 2003. PMID 14616430, DOI 10.1046/j.1365-2125.2003.02005.x. Read the source
39. British Journal of Clinical Pharmacology. Interaction of St John's wort with low-dose oral contraceptive therapy: a randomized controlled trial — full text, Methods (assay paragraph). 2003. PMID 14616430, DOI 10.1046/j.1365-2125.2003.02005.x. Read the source