If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Efficacy of combined oral contraceptives for depressive symptoms and overall symptomatology in premenstrual syndrome: pairwise and network meta-analysis of randomized trials
Combined oral contraceptives are often considered a treatment option for women with premenstrual syndrome or premenstrual dysphoric disorder also seeking contraception, but evidence for this treatment is scarce. We aimed to determine (1) the level of evidence for the efficacy of combined oral contraceptives in managing premenstrual depressive symptoms and overall premenstrual symptomatology and (2) the comparative efficacy of combined oral contraceptives (the International Prospective Register of Systematic Reviews registration number CRD42020205510).
Data Sources
We searched Cochrane Central Register of Controlled Trials, PubMed, Web of Science, PsycINFO, EMCare, and Embase from inception to June 3, 2021.
Study Eligibility Criteria
All randomized clinical trials that evaluated the efficacy of combined oral contraceptives in women with premenstrual syndrome or premenstrual dysphoric disorder were considered eligible for inclusion in this meta-analysis.
Study Appraisal and Synthesis Methods
A random effect Bayesian pairwise and network meta-analysis was conducted with change in premenstrual depressive symptoms and overall premenstrual symptomatology between baseline and 3 cycles as outcome. Certainty of the evidence was assessed using the Grading of Recommendations, Assessment, Development, and Evaluation approach.
Results
Of 3664 records, 9 eligible trials were included that studied 1205 women with premenstrual syndrome or premenstrual dysphoric disorder (mean age per study range, 24.6–36.5 years). The pairwise meta-analysis revealed that combined oral contraceptives were more efficacious than placebo in treating overall premenstrual symptomatology (standardized mean difference, 0.41; 95% credible interval, 0.17–0.67), but not premenstrual depressive symptoms specifically (standardized mean difference, 0.22; 95% credible interval, −0.06 to 0.47). However, none of the combined oral contraceptives were more effective than each other in reducing premenstrual depressive symptoms and overall premenstrual symptomatology.
Conclusion
Combined oral contraceptives may improve overall premenstrual symptomatology in women with premenstrual syndrome or premenstrual dysphoric disorder, but not premenstrual depressive symptoms. There is no evidence for one combined oral contraceptive being more efficacious than any other.
This study aimed to estimate the efficacy of combined oral contraceptives for premenstrual syndrome (PMS) and premenstrual dysphoric disorder (PMDD).
Key findings
Aggregated evidence from randomized, placebo-controlled clinical trials in a pairwise meta-analysis suggested that combined oral contraceptives may improve overall premenstrual symptomatology in women with PMS or PMDD, but not premenstrual depressive symptoms. However, the network meta-analysis showed that there is no evidence for a particular combined oral contraceptive being more effective than others.
What does this add to what is known?
This study suggests that if combined oral contraceptives are preferred for treating PMS or PMDD, there is no evidence for a particular formulation being more effective than any other.
Introduction
Symptoms that occur in the luteal phase of the menstrual cycle and resolve after the onset of menstruation characterize the premenstrual syndrome (PMS) and its more severe variant premenstrual dysphoric disorder (PMDD).
A broad range of physical and affective symptoms may be present, but in women with PMDD, severe affective symptoms such as depression are the primary complaint.
Of menstruating women, up to 12.2% were affected by PMS and up to 5.3% by PMDD. In both disorders, cyclic changes in the production of progesterone are thought to trigger symptoms.
Because combined oral contraceptives contain a synthetic estradiol and progestin that together suppress the hypothalamic-pituitary-ovarian axis, treatment with these drugs is often considered for women with PMS or PMDD who also seek contraception. However, the evidence for their efficacy remains scarce.
Currently, ethinylestradiol drospirenone (20 μg, 3 mg) in a 24-day regimen is the only combined oral contraceptive that has been approved by the US Food and Drug Administration for the treatment of PMDD.
This approval was based on the results of 2 randomized clinical trials in women with PMDD reporting that the use of this formulation reduced overall premenstrual symptomatology compared with placebo.
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
Notably, 2 trials investigating the continuous use of ethinylestradiol levonorgestrel (20 μg, 90 μg) in women with PMDD did not show a consistent beneficial effect of on premenstrual depressive symptoms and overall premenstrual symptomatology.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Altogether, these inconsistent findings question the efficacy of combined oral contraceptives for treating premenstrual symptomatology in women with PMS and PMDD.
Here, we sought to aggregate the results from clinical trials that randomized women with PMS or PMDD to combined oral contraceptives compared with any other hormonal or nonhormonal contraceptive treatment regimen or to placebo to examine the efficacy of combined oral contraceptives. We specifically aimed to determine (1) the efficacy of combined oral contraceptives for managing premenstrual depressive symptoms and overall premenstrual symptomatology in PMS and PMDD and (2) whether some combined oral contraceptives are more efficacious than others.
Methods
Per the recommended systematic review methodology, this review was developed and registered a priori with the International Prospective Register of Systematic Reviews (PROSPERO [registration number CRD42020205510]).
Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (editors). Cochrane handbook for systematic reviews of interventions version 6.2 (updated February 2021). Cochrane. 2021. Available from www.training.cochrane.org/handbook. Accessed May 1, 2021.
In the original PROSPERO protocol, the primary aim was to conduct a network meta-analysis, and only if this was not possible, to conduct a pairwise meta-analysis. Because all but 1 of the included randomized clinical trials had a placebo-controlled design, making it possible to conduct a (potentially better-powered) pairwise meta-analysis examining combined oral contraceptives as a group, the protocol was modified to include a pairwise meta-analysis as the primary analysis. We originally aimed to investigate whether studies with a greater proportion of women with PMDD reported larger effect sizes compared with placebo than those that included a smaller proportion of women with PMDD, using meta-regression analyses. However, because fewer than 10 studies could be included in this study, such analyses were not possible.
Eligibility criteria
The inclusion criteria for studies were as follows (within a Population, Intervention, Comparison, Outcomes, and Study framework):
1.
Population: Premenopausal women diagnosed as having PMS or PMDD. The diagnosis of PMDD was preferably made according to the Diagnostic and Statistical Manual of Mental Disorders, IV or 5 criteria.
Other assessments of PMDD through the use of validated questionnaires were accepted if the criteria used required a rise of at least 1 emotional symptom (including but not limited to depressed mood, anxiety or tension, affective liability, and anger or irritability) plus 4 additional symptoms during the luteal or premenstrual phase that subsided within the first 3 days of the menses. For PMS, the same criteria were used except that no additional symptoms were required for the diagnosis.
2.
Intervention: A specific combined oral contraceptive regimen
3.
Comparison: Another combined oral contraceptive regimen or nonhormonal or hormonal contraceptives (eg, hormonal or copper intrauterine device) or placebo
4.
Outcome: Depressive symptoms using a validated (self-report or observer-rated) scale
5.
Study design: Randomized controlled trials
Information sources, search strategy, and study selection
We conducted a search for studies published in English in the Cochrane Central Register of Controlled Trials, PubMed, Web of Science, PsycINFO, EMCare, and Embase from their inception to June 3, 2021. Reference lists of included trials and of related reviews (searched for by a separate search) were checked to identify other potentially eligible trials or ancillary publications. Supplement 1 lists the full searches in PubMed.
Two reviewers (C.S. and A.F.) independently scanned all retrieved citations by title, abstract, and full text according to the prespecified inclusion criteria. Any discrepancies were resolved through discussion or recourse to a third reviewer (A.E.d.W.). Two reviewers (C.S. and A.F.) extracted data on summary estimates independently for each eligible trial using a standardized, pilot tested data extraction form.
Data extraction
The reviewers (C.S. and A.F.) independently collected information on methodology (level of blinding, crossover or parallel group design), interventions (formulation, dose, frequency, regimen and route of administration), participants (number of subjects per group, number of dropouts, mean age, and proportion of participants with PMDD, with comorbidity, or who reported being sexually active), and outcomes (ie, tools or scales, reported time points, phase of cycle reported). Reviewers resolved discrepancies by discussion and, when necessary, through adjudication by a third party (A.E.d.W.). We contacted study authors and drug manufacturers to supplement incomplete data regarding outcomes. The relative effect per comparison was summarized using the standardized mean difference (SMD) adjusted for small sample sizes (Hedges g correction), with a 95% credible interval (CrI).
A CrI is an interval within which an unobserved parameter value falls with a particular probability given the evidence provided by the observed data. Using an SMD as an outcome, intervals containing 0 suggest that the data are compatible with no effect or no difference among groups, as with the analogous 95% confidence intervals in frequentist statistics.
The primary outcome was a change in premenstrual depressive symptoms or new-onset depression between baseline and 3 cycles. The secondary outcome was the change in overall premenstrual symptomatology measured over the same period. For the primary outcome, preference was given to validated depressive symptom severity questionnaires. However, when such questionnaires were not used in a trial, data from negative affect or depressive symptom subscales from scales measuring related concepts were accepted. Similarly, overall premenstrual symptomatology measures were accepted when they had been assessed with validated symptom questionnaires for premenstrual complaints. This questionnaire had to assess both the affective and physical domain of PMS and provide a score that reflects the severity of premenstrual complaints. When symptoms of the primary or secondary outcome had been measured with more than one standardized rating scale, the scale with the best psychometric properties was chosen. If results were not separately reported for the premenstrual phase, effect sizes across phases were taken. Similarly, when 3-cycle data were not available, other data were used (eligible range, 1–48 cycles) that were as close to this point as possible. Intention-to-treat data were used, whenever possible.
Assessment of risk of bias
We assessed the studies’ risk of bias using the modified Cochrane Risk of Bias tool 2.0.
In addition, we assessed the certainty of evidence contributing to network estimates of the primary outcome with the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) framework using the Confidence in Network Meta-Analysis tool.
Judgments on the certainty of evidence were made for each of the following domains: within-study bias, reporting bias, indirectness, imprecision, heterogeneity, and incoherence. Supplement 2 lists the full details on the GRADE assessment.
Statistical approach
First, we performed a random-effects Bayesian pairwise meta-analysis with uninformative priors and Markov chain Monte Carlo sampling, to determine the efficacy of combined oral contraceptives (as a group) compared with placebo, in managing premenstrual depressive symptoms and overall premenstrual symptomatology. The “brms” package in R (version 4.0.3, R Foundation for Statistical Computing, Vienna, Austria) was used to build the model. Simulations were run for 4 chains with an adaptive phase of 10,000 iterations and a sampling phase of 500,000 iterations, thinned such that every tenth iteration was retained. Model convergence was assured by checking whether the density of the replications was similar to the ones in the observed data and checking the potential scale reduction factor.
Second, a random-effects Bayesian network meta-analysis with uninformative priors was used to determine the comparative efficacy of combined oral contraceptives in managing premenstrual depressive symptoms and overall premenstrual symptomatology. The “gemtc” package in R (version 4.0.3) was used to build the model with the same method of sampling and running simulations as for the pairwise meta-analysis. Using multivariate distributions, the model was accounted for correlations induced by multi-arm studies. Model convergence was checked using trace plots, density plots, and the Brooks-Gelman-Rubin diagnostic. Each specific combined oral contraceptive (including regimen and dose) was treated as a separate node. A network plot was drawn, with thickness of the lines between nodes based on the number of direct comparisons investigated.
Results
Study selection
Our systematic search identified 3644 citations published between 1961 and June 3, 2021. One additional trial was identified through hand searching other reports. After removal of duplicates, 3581 articles were screened based on title and abstract, of which 98 were screened based on the full text. From these articles, 10 studies were considered eligible,
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
Does adding fluoxetine to combined oral contraceptives containing drospirenone improve the management of severe premenstrual syndrome? A 6-month randomized double-blind placebo-controlled three-arm trial.
Does adding fluoxetine to combined oral contraceptives containing drospirenone improve the management of severe premenstrual syndrome? A 6-month randomized double-blind placebo-controlled three-arm trial.
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
The 9 included studies examined 9 different interventions (8 different combined oral contraceptives and 1 placebo) among 1205 women (weighted mean age, 32.3 years; range, 24.6–36.5). Across comparisons, the sample size per arm ranged from 16 to 231 participants (Table 1). All studies evaluated the effect of the premenstrual intervention (in the final week before the onset of menses) after 3 months of treatment. During this follow-up period, dropout was generally high with a weighted mean dropout rate of 26.3% (range, 0.0%–40.2%). Most studies included women with PMDD,
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
This reference describes 4 studies including 2 randomized clinical trials (the multinational and the North American trial). Because the North American trial is also described in another reference included in this meta-analysis,14 we refer to the multinational trial when using this reference.13
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
Data are presented as median (SD), number, and percentage.
Ethinylestradiol norethindrone (35 μg, 0.5; 1; 0.5; 0 mg) is a multiphasic combined oral contraceptive that has 4 different dosages of norethindrone throughout a 4-week cycle.
BDI, Beck Depression Inventory; COPE, Calendar of Premenstrual Experiences; DRF, daily Rating Form; DRSP, Daily Record of Severity of Problems; PMDD, premenstrual dysphoric disorder; PMS, premenstrual syndrome; SD, standard deviation; WHAQ, Women’s Health Assessment Questionnaire.
de Wit. Efficacy of combined oral contraceptives in premenstrual syndrome. Am J Obstet Gynecol 2021.
a Number of women who were available for the analysis (so excluding the dropouts in case of complete cases analyses)
b This reference describes 4 studies including 2 randomized clinical trials (the multinational and the North American trial). Because the North American trial is also described in another reference included in this meta-analysis,
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
Notably, 7 trials provided data on the efficacy of interventions on overall premenstrual symptomatology; 5 trials assessed overall premenstrual symptomatology with the DRSP total score.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
The certainty of evidence (GRADE) for all comparisons examined in the network meta-analysis was judged to be “very low” (Supplement 2 Table 1). Within-study bias (Supplement 2.1) and imprecision of the effect sizes (Supplement 2.4) caused most of these judgments. The judgments made for each of the domains are described in Supplement 2.
Premenstrual depressive symptoms
Based on the results of 8 placebo-controlled trials,
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
the pairwise meta-analysis revealed no evidence to prove that combined oral contraceptives were effective for premenstrual depressive symptoms compared with placebo (SMD, 0.22; 95% CrI, −0.06 to 0.47) (shown in the upper panel of Figure 2).
Figure 2Forest plot of effects of combined oral contraceptives vs placebo
The effect sizes for each study are centered on the pooled effect size. Importantly, the effect sizes displayed are the estimates of the “true” effect size of study based on the Bayesian model (not the effect sizes reported in the original studies). The study by Freeman et al
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
describes 4 studies including 2 randomized clinical trials (the multinational and the North American trial). Because the North American trial is also described in the study by Halbreich et al,
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
This resulted in a network of 9 individual nodes where most interventions were linked only with placebo. Most comparisons were only studied once, but both ethinylestradiol levonorgestrel (20 μg, 90 μg) in a continuous regimen and ethinylestradiol drospirenone (20 μg, 3 mg) in a 24-day regimen were compared with placebo in 2 different studies (Figure 3). Therefore, comparisons between combined oral contraceptives and placebo were predominantly on the basis of direct evidence, whereas comparisons among combined oral contraceptives were predominantly based on indirect evidence. In line with the pairwise meta-analysis findings, the network meta-analysis showed that none of the combined oral contraceptives (including the specific regimens) were more effective than placebo or each other (comparative efficacy shown in lower panel of Table 2 and visualized in the upper panel of Figure 4).
Figure 3Network of treatment comparisons for premenstrual depressive symptoms
Each node represents a different combined oral contraceptive or placebo. The thickness of lines between the nodes is proportional to the number of randomly assigned patients contributing to direct comparisons. Ethinylestradiol norethindrone (35 μg, 0.5; 1; 0.5; 0 mg) is a multiphasic combined oral contraceptive that has 4 different dosages of norethindrone throughout a 4-week cycle. The dosages before the “slash” refer to the different dosages of the synthetic estrogen, and the ones after the “slash” to those of the synthetic progestin.
de Wit. Efficacy of combined oral contraceptives in premenstrual syndrome. Am J Obstet Gynecol 2021.
Table 2Comparisons for efficacy of combined oral contraceptives on premenstrual depressive symptoms (lower triangle) and overall premenstrual symptomatology (upper triangle)
None of the results were significant and all evidence was of low certainty.
(−1.26 to 0.53)
Placebo
Data are expressed as SMD with 95% credible intervals in the column-defining intervention compared with the row-defining intervention. Higher SMD values correspond with fewer symptoms in the column-defining hormonal contraceptive. Ethinylestradiol norethindrone (35 μg, 0.5; 1; 0.5; 0 mg) is a multiphasic combined oral contraceptive that has 4 different dosages of norethindrone throughout a 4-week cycle. The dosages before the “comma” refer to the different dosages of the synthetic estrogen, and the ones after the “comma” to those of the synthetic progestin.
SMD, standardized mean difference.
de Wit. Efficacy of combined oral contraceptives in premenstrual syndrome. Am J Obstet Gynecol 2021.
a None of the results were significant and all evidence was of low certainty.
The size of the SMD dots are proportional to the precision of the estimate (1/standard error). Ethinylestradiol norethindrone (35 μg, 0.5; 1; 0.5; 0 mg) is a multiphasic combined oral contraceptive that has 4 different dosages of norethindrone throughout a 4-week cycle. The dosages before the “comma” refer to the different dosages of the synthetic estrogen, and the ones after the “comma” to those of the synthetic progestin.
A total of 6 placebo-controlled trials were included in the pairwise meta-analysis on the efficacy of combined oral contraceptives in reducing overall premenstrual symptomatology.
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
In contrast to the findings on premenstrual depressive symptoms, combined oral contraceptives were found to be moderately effective compared with placebo (SMD, 0.41; 95% CrI, 0.17–0.67) (shown in the lower panel of Figure 2).
However, when estimating the comparative efficacy of combined oral contraceptives in 7 trials with the network meta-analysis,
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
none of the combined oral contraceptives were more effective in improving overall premenstrual symptomatology compared with placebo or compared with each other (comparative efficacy shown in upper panel of Table 2 and visualized in the lower panel of Figure 4).
Comment
Main findings
The results of this pairwise and network meta-analysis suggest that combined oral contraceptives, compared with placebo, may improve overall premenstrual symptomatology in both women with PMS and PMDD. However, we found no support for their efficacy in treating premenstrual depressive symptoms, and none of the combined oral contraceptives were more efficacious than any other.
Strengths and limitations
The main strength of this study is the use of a comprehensive literature search which enabled us to include 7 randomized clinical trials that were not examined in the previous meta-analysis.
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
In addition, the network meta-analysis enabled us to estimate the comparative efficacy of combined oral contraceptives using both direct and indirect evidence.
However, only 1 or 2 trials were available for each comparison and many trials had a relatively small sample size. The combination of these 2 factors limits the certainty of evidence for each comparison as was reflected by the GRADE assessment, which resulted in a “very low” certainty rating of the summary of evidence owing to relatively large imprecision of the estimates from each study. Moreover, there were only 2 trials that used combined oral contraceptives as a comparison group. Therefore, the estimates of efficacy of one combined oral contraceptive vs another are primarily based on indirect evidence (mainly through placebo). In addition, the limitations of individually included trials inevitably limited the quality of our analysis. Only 3 of the 9 trials included in the network meta-analysis, examined all participants randomized into a trial irrespective of what happened subsequently (also known as intention-to-treat analysis).
This may have increased the risk of overestimation of the efficacy of combined oral contraceptives, because the use of per-protocol or observed cases analysis violates the principles of randomization. Loss to follow-up was substantial in almost all trials, which further question the prognostic balance afforded by randomization. Finally, unregistered trials that were not published (as a full journal article) were not included in the search. In case of publication bias against null findings, this may have resulted in an overestimation of effects. However, because we mainly report null findings (except for the effect of oral contraceptives compared with placebo for overall premenstrual symptomatology), we do not expect this to have a large effect on our results.
Comparison with existing literature
We are aware of only one previous meta-analysis that attempted to estimate the efficacy of combined oral contraceptives in treating PMDD. This meta-analysis did not include measures on premenstrual depressive symptoms, but did report that, compared with placebo, ethinylestradiol drospirenone (20 μg, 3 mg) use in a 24-day regimen was effective in reducing overall premenstrual symptomatology in women with PMDD.
Therefore, this finding stands in direct opposition to the results reported here, showing that the same formulation was not superior to other types of combined oral contraceptives or placebo in treating PMS and PMDD. However, there are important differences in the way the data were evaluated that could explain the difference in findings. First, the previous meta-analysis used a fixed-effect rather than a random-effects model for analyzing the data. Fixed-effects models assume that there is a single underlying “true” effect size, which requires that all factors that could influence the effect size are the same in all trials. Because this assumption is rarely met, we argue that a random-effects model is more suitable. Second, the previous meta-analysis used frequentist inference testing rather than Bayesian inference testing. Although both approaches are reasonable, the contradictory results suggest that the efficacy of ethinylestradiol drospirenone (20 μg, 3 mg) in a 24-day regimen is not very robust to differences in analytical strategy.
Implications for current clinical practice and future research
This meta-analysis shows that combined oral contraceptives represent a reasonable therapeutic option to treat overall premenstrual symptomatology in women with PMS or PMDD also seeking contraception. However, combined oral contraceptives were not effective in treating premenstrual depressive symptoms. Physicians should take this into account when prescribing combined oral contraceptives, because depressive symptoms are often the main symptoms in women with PMDD. In women who predominantly complain about depressive symptoms, other commonly used treatment options for women with PMDD, such as selective serotonin reuptake inhibitors and cognitive-behavioral therapy, might be a more effective approach.
However, certainty of evidence for all available treatments is very low to moderate at best, which highlights that optimal treatment strategies are not yet well established.
In contrast to what is often thought, there was no evidence for a difference in efficacy among formulations containing drospirenone and other formulations or a 24-day regimen vs other regimens. This suggests that other arguments could justify the decision to choose 1 contraceptive over another in women with PMS or PMDD seeking treatment with a combined oral contraceptive. Because drospirenone is associated with an increased risk of blood clots compared with other hormonal contraceptives,
United States Food & Drug Administration FDA Drug Safety Communication: safety review update on the possible increased risk of blood clots with birth control pills containing drospirenone.
this may argue against preferentially prescribing a drospirenone-containing contraceptive. However, the insufficient evidence in favor of a specific combined oral contraceptive over another in treating PMS or PMDD may be caused by a lack of statistical power. This clearly highlights the need for additional large randomized clinical trials to answer the question of which treatment options are most effective in women with PMS and PMDD. Preferably, these trials should include both women with PMS and women with PMDD, to ensure that findings are generalizable to the full range of severity of premenstrual complaints which in addition would allow to examine whether combined oral contraceptives are equally effective in women with PMDD vs PMS.
Conclusions
The results of this pairwise and network meta-analysis show that combined oral contraceptives may improve overall premenstrual symptomatology in women with PMS or PMDD, but there is currently no evidence to favor a specific combined oral contraceptive over another. Moreover, there was no evidence to suggest that combined oral contraceptives are effective in treating premenstrual depressive symptoms, which are often the primary complaints of women with PMDD. Few trials were available for each formulation of combined oral contraceptives, and certainty of evidence for each comparison was rated as very low. Therefore, more randomized clinical trials with head-to-head comparisons of combined oral contraceptives are needed to determine which formulation is most effective for premenstrual complaints in women with PMS and PMDD.
Acknowledgments
We thank Jan W. Schoones, MA, information specialist from Leiden University Medical Center, for assistance in developing the search strategy. We are grateful for additional data that were provided to us by Cynthia A. Graham, professor at the University of Southampton; Cecilia Lundin, PhD student at Uppsala University; and Lynne Smith of Pfizer.
Treatment of premenstrual dysphoria with continuous versus intermittent dosing of oral contraceptives: results of a three-arm randomized controlled trial.
An overview of four studies of a continuous oral contraceptive (levonorgestrel 90 mcg/ethinyl estradiol 20 mcg) on premenstrual dysphoric disorder and premenstrual syndrome.
Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (editors). Cochrane handbook for systematic reviews of interventions version 6.2 (updated February 2021). Cochrane. 2021. Available from www.training.cochrane.org/handbook. Accessed May 1, 2021.
Combined oral contraceptive use is associated with both improvement and worsening of mood in the different phases of the treatment cycle—a double-blind, placebo-controlled randomized trial.
Does adding fluoxetine to combined oral contraceptives containing drospirenone improve the management of severe premenstrual syndrome? A 6-month randomized double-blind placebo-controlled three-arm trial.
Data sharing: On publication of this article, the full data set and the script for analyzing the data will be freely available at https://osf.io/8zbu3/.
00774-2&id=gr1.jpg){kind=link}
00774-2&id=gr2.jpg){kind=link}
00774-2&id=gr3.jpg){kind=link}
00774-2&id=gr4.jpg){kind=link}