Premenstrual Dysphoric Disorder and Antiepileptic Drugs: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Premenstrual Dysphoric Disorder and Antiepileptic Drugs (NCT00612235), a registered phase not stated study. The primary outcome is To Determine if the Frequency of Premenstrual Dysphoric Disorder Differs Among Various Antiepileptic Drug Monotherapies..

2. Clinical Question

The clinical question concerns whether PMDD designation is observed at different frequencies across groups of women with epilepsy taking different single antiepileptic drugs, compared with women without epilepsy. It is a group-comparison question, not a randomized treatment-effect question: the registry record does not state how participants were allocated, and it explicitly cautions that the arms are not a like-for-like control comparison.

The primary outcome description defines designation using the Endicott Daily Record of Severity of Problems (DRSP). Women had to meet the stated PMDD criteria in both consecutive menstrual cycles to count as meeting the primary designation. The record also describes a less stringent threshold, involving more severe premenstrual symptoms than midfollicular symptoms regardless of whether midfollicular symptom scores exceeded the Endicott cutoff, together with the other three Endicott criteria.

The record does not report the group-specific numbers meeting either threshold or an estimate of how the groups differed. Therefore, it defines the intended clinical comparison but does not give results with which to judge the size or precision of any group differences.

3. Trial Design

The registry record describes a completed study titled “Premenstrual Dysphoric Disorder and Antiepileptic Drugs.” Its official title specifies a comparison of different antiepileptic drug monotherapies for the occurrence of PMDD among women with epilepsy. The sponsor is Beth Israel Deaconess Medical Center. The record lists enrollment of 80, a start date of 2008-04, a primary completion date of 2010-06, and a completed status.

The listed groups are lamotrigine monotherapy, levetiracetam monotherapy, carbamazepine monotherapy, and a normal control group without epilepsy. The interventions list the three drugs and a no-intervention control. The safety record lists 20 participants at risk in each group, but those safety counts do not establish how participants entered or were assigned to the groups.

The registry caveat says this is not a randomized comparison, that allocation is not stated, and that the arms are not a like-for-like control comparison. This matters because group differences can reflect characteristics other than the drug or epilepsy status. The control group differs by the absence of epilepsy as well as by the absence of antiepileptic drug treatment, so it is not simply another drug-treatment group.

The primary outcome timeframe is assessment of PMDD designation after two consecutive menstrual cycles. The record lists two outcome measures and zero analyses. It does not describe a randomization procedure, masking, participant flow, treatment duration, or a statistical analysis plan. Those design details are not reported in the registry record, so the comparison should be interpreted from the stated group structure and outcome definition rather than from assumptions about how the study was conducted.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Since the original conceptualization of the study, phenytoin use declined among women with epilepsy in the general population to the point that we did not identify a sufficient base population in our initial screen to proceed with enrollment of this group and the decision was made at the start of the study to exclude this drug group.

Overall Study

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Women with epilepsy on antiepileptic drug monotherapies and normal controls. There was no difference in the number who started and completed.

6. Primary Endpoint

The registry lists the primary outcome as To Determine if the Frequency of Premenstrual Dysphoric Disorder Differs Among Various Antiepileptic Drug Monotherapies., measured over Assessment of PMDD Designation after two consecutive menstrual cycles.

The primary endpoint counts participants who meet the stated PMDD designation criteria. The registry describes the measure as the proportion of women with epilepsy and women in the control group who satisfy the Endicott Daily Record of Severity of Problems criteria for two consecutive menstrual cycles. A participant must meet the PMDD criteria in both cycles to qualify for the primary designation. Thus, the endpoint is a binary participant-level outcome: a participant either meets the stated designation or does not.

The description also mentions a less stringent designation threshold. Under that threshold, symptoms must be more severe during the premenstrual phase than during the midfollicular phase, regardless of whether midfollicular symptom scores exceed the Endicott cutoff, and the other three Endicott criteria must also be met. The record does not provide numerical results for this threshold, nor does it clarify how the two outcome measures relate to each other beyond listing them.

Using a defined symptom-recording criterion can make group comparisons more consistent than relying on an unspecified impression of mood. Requiring criteria in both consecutive cycles also makes the endpoint depend on repeated cycle-level observations rather than a single cycle. These are general measurement considerations, not evidence that the endpoint was measured completely or consistently in every participant.

The endpoint is a designation based on the stated criteria. It is not a measure of symptom severity across all participants, and a group-level frequency would not by itself describe the experiences of every individual. Because the registry record does not report the number of participants meeting the endpoint in each group, the observed frequencies and their uncertainty are unavailable.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record describes a binary primary outcome measured at the participant level. The outcome is whether a woman meets the stated PMDD designation criteria for two consecutive menstrual cycles. The record identifies the Endicott Daily Record of Severity of Problems (DRSP) and describes an additional, less stringent designation threshold. It does not report the statistical methods used to compare groups, so the discussion below is general statistical context rather than a description of the study’s analysis.

What does a binary endpoint represent?

A binary endpoint places each participant into one of two outcome categories, such as meeting or not meeting the stated designation. A basic summary would report the number meeting the criterion and the number assessed in each group. The proportion meeting the criterion then describes the frequency within that group. For a comparison, an analysis might report an absolute difference between proportions or a relative measure. These approaches answer different questions: an absolute difference describes how far apart the group frequencies are, while a relative measure compares their scale. The registry record does not report these summaries or state which comparative measure was planned.

Counts and denominators are essential to interpret a binary result. A proportion without its denominator can conceal how much information underlies it, and the same observed proportion can have different precision in groups of different sizes. In this study, the registry record lists an overall enrollment of 80 and safety groups with 20 participants at risk each, but it does not report the primary endpoint denominators or counts. Safety denominators cannot be assumed to be the endpoint analysis denominators.

Why does the cycle-based definition matter?

The primary designation requires meeting criteria across two consecutive menstrual cycles. In general, requiring repeated observations can make an endpoint depend on a pattern across time rather than on a single assessment. It also means that an outcome cannot be established in the same way when one of the required cycles is unavailable. The record does not explain how incomplete cycle records were handled, whether the assessment schedule was uniform, or whether every participant posted observations for both cycles.

The description of the less stringent threshold also matters. It changes the stated rule for designating PMDD by allowing the premenstrual-to-midfollicular symptom comparison regardless of whether midfollicular scores exceeded the Endicott cutoff, while retaining the other three Endicott criteria. Different operational definitions can classify different participants. The record lists two outcome measures, but it does not report their group results or explain whether one was designated as secondary, exploratory, or part of a hierarchy.

What can an interval or test tell a reader?

General statistical explanation: an estimate is a numerical summary of the observed data, while a confidence interval describes uncertainty around that estimate under a specified method and its assumptions. A narrow interval indicates more precision than a wide interval, but neither alone establishes that a difference is important. A p-value evaluates how unusual the data would be under a stated null hypothesis and model. It is not the probability that the null hypothesis is true, and it does not quantify the size of an effect.

For a binary outcome, the appropriate interval and test depend on such matters as group sizes, the measure of comparison, and the analysis assumptions. The registry record does not state a test, model, confidence level, adjustment strategy, or hypothesis-testing plan. No p-value or interval for the primary outcome is reported. It would therefore be inappropriate to attribute a particular inferential method to the study or to draw a significance conclusion.

How does the group structure affect interpretation?

The registry caveat says allocation is not stated and the arms are not a like-for-like control comparison. General statistical explanation: random allocation, when used and properly implemented, can help balance measured and unmeasured baseline characteristics across treatment groups on average. Without a stated allocation process, differences between groups may be entangled with differences in participant characteristics, clinical history, or other factors. The comparison with women without epilepsy has an additional structural distinction: that group differs in epilepsy status as well as in exposure to antiepileptic monotherapy. Statistical adjustment can sometimes address measured differences, but it cannot guarantee removal of unmeasured confounding, and no adjustment plan is reported.

Because the record gives no primary outcome results or analysis methods, it is not possible to assess how the investigators addressed these issues. The safety group counts do not resolve them. The design caveat should be treated as a constraint on the interpretation of any eventual group contrast, rather than as a statistical correction that can be applied from the registry record.

What is a planning illustration?

General statistical explanation: a sample-size planning illustration estimates how many participants might be needed under assumptions about the effect to detect, the significance level, allocation, and statistical method. It is not a measurement of what a study actually achieved. Any illustration based on an observed effect is conditional on that estimated effect and should not be described as observed power. The record’s planning assumptions specify a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the target effect expressed as a stated fraction of an observed effect. The registry record does not report the study’s own design justification, and no actual planning counts are included in the record details. Such assumptions should not be mistaken for the trial’s documented protocol.

Overall, the primary outcome definition can be explained, but the statistical analysis cannot be reconstructed from the record. A complete interpretation would require group-specific primary outcome data, denominators, handling of incomplete observations, and the analysis methods. Those details are not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists two outcome measures but does not report the full set of outcome definitions, their hierarchy, or which inferential comparisons were planned. It also lists zero analyses. The record does not describe whether there were multiple comparisons among the three drug groups and the control group, whether any outcome was designated secondary, or whether an adjustment for multiplicity was planned. These details are not reported in the registry record.

General statistical explanation: when many outcomes or group comparisons are tested, the chance of obtaining at least one apparently unusual result can increase. A prespecified testing hierarchy or multiplicity adjustment can help interpret a collection of tests. Without a documented analysis plan or posted analyses, the impact of multiple comparisons cannot be assessed. No claim about adjusted or unadjusted inference is supported by the record.

Missing data are also not described. The endpoint requires criteria to be met over two consecutive menstrual cycles, but the record does not say how participants with an incomplete cycle record, missing diary entries, or unavailable endpoint assessments were handled. It does not report the number excluded from the endpoint analysis or reasons for missingness. Thus, the primary endpoint denominators and the extent of missing outcome data are unknown.

The record does not define analysis populations, such as whether all enrolled participants or only participants with complete endpoint measurements were included. It also does not report participant flow between enrollment, group membership, and analysis. Enrollment of 80 and the safety counts do not establish the population used for the primary outcome. These omissions limit assessment of selection, missing-data bias, and the extent to which the posted outcome—if available—would represent all enrolled participants.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

The safety listing reports zero serious adverse events, zero other adverse events, and zero deaths in each named group. It lists 20 participants at risk in each group for serious and other adverse events. These are the posted safety counts; the registry record does not provide event terms because no serious event terms are listed.

Zero reported events should be read as a count in this record, not as proof that an event could not occur or that there was no safety risk. General statistical explanation: when no events are observed, the true event frequency remains uncertain, particularly when the amount of observation is limited. The record does not provide the duration or completeness of safety follow-up, details of event collection, or additional information needed to evaluate that uncertainty.

The safety figures do not answer the primary PMDD question. They also do not establish whether every enrolled participant was included in the safety assessment, or whether participants in the safety groups correspond exactly to the primary endpoint analysis population. The safety counts should therefore be interpreted separately from the unreported primary outcome results.

12. What the Result Does, and Does Not, Mean

The record establishes what the study set out to compare: PMDD designation among women with epilepsy taking lamotrigine, levetiracetam, or carbamazepine monotherapy, with a control group without epilepsy. It defines the primary designation using the Endicott DRSP criteria over two consecutive menstrual cycles and describes a less stringent threshold. It also records completed status, enrollment of 80, and safety counts for the named groups.

It does not establish whether PMDD designation was more or less frequent in any group. The registry record reports zero analyses and no group-specific primary outcome counts, proportions, effect estimates, confidence intervals, or p-values. A lack of posted analysis is not evidence that the groups had identical outcomes. It means the direction, size, and precision of any primary difference cannot be evaluated from the record.

The comparison also cannot by itself identify a causal drug effect. Allocation is not stated, and the registry caveat says the arms are not a like-for-like control comparison. Women without epilepsy differ from women with epilepsy in more than the named medication exposure. General statistical explanation: in nonrandomized comparisons, observed group differences may reflect factors that differ between the groups as well as the exposure of interest. The record does not report enough design or analysis detail to assess whether such factors were measured or addressed.

The safety listing reports no serious adverse events, other adverse events, or deaths in the listed groups. Those zero counts describe what the registry reports; they do not establish an absence of all risk, nor do they answer whether the primary endpoint differed.

In short, the registry record is informative about the question and its measurement, but not about the primary outcome findings. It can support a description of the study’s scope and limitations, not a conclusion that a particular antiepileptic drug did or did not change PMDD frequency.

13. Why This Trial Matters Statistically

This study raises a statistically important measurement question: how should a cyclical symptom pattern be converted into a participant-level outcome for comparison? The registry defines PMDD designation using criteria recorded across two consecutive menstrual cycles, rather than a single-cycle observation. That definition makes the endpoint rule explicit and highlights why the timing and completeness of repeated observations matter.

The study also illustrates the difference between a clinically relevant comparison and a causal treatment comparison. Comparing several monotherapy groups can describe how outcome frequencies vary across those groups, but an interpretation as a drug effect depends on design and analysis details. Here, the registry caveat says allocation is not stated and the arms are not a like-for-like control comparison. The control group without epilepsy adds a separate contrast involving epilepsy status, not just medication use.

From a statistical reporting perspective, the record shows why outcome definitions alone are not enough. To judge a binary outcome, readers need the group counts and denominators, an estimate of the comparison, and an account of uncertainty. They also need to know how incomplete cycle records were handled and which participants were analyzed. The record lists two outcome measures but zero analyses, leaving those elements unavailable.

The safety listing offers another useful distinction: event counts and primary efficacy outcomes address different questions. Zero recorded adverse events in the listed groups does not supply information about PMDD designation and cannot compensate for unreported primary outcome results.

The central statistical lesson is that a registered question and a well-described endpoint do not, on their own, establish a finding. Interpretation depends on posted data, transparent analysis, and a design appropriate to the comparison being made. In this record, the group structure and endpoint are described, while the results needed to quantify or test the primary comparison are not reported.

14. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The principal limitation for interpreting the primary question is that the registry record does not report primary outcome results. Although the endpoint is described, there are no group-specific PMDD counts, proportions, comparative estimates, confidence intervals, or p-values. The record lists two outcome measures and zero analyses, so neither the primary comparison nor the less stringent threshold can be evaluated quantitatively.

The registry itself cautions that this is not a randomized comparison, that allocation is not stated, and that the arms are not a like-for-like control comparison. The drug groups and the control group without epilepsy therefore should not be treated as though they differ only by assigned treatment. The record does not provide sufficient information to assess group comparability or whether measured differences were adjusted for.

The endpoint depends on meeting criteria during two consecutive menstrual cycles, but the record does not explain how missing diary entries, incomplete cycles, or unavailable assessments were handled. It does not report primary endpoint denominators, exclusions, participant flow, or analysis populations. Enrollment of 80 and the safety listing do not resolve these unknowns.

Other design details, including the recruitment approach, participant characteristics, masking, treatment duration, and statistical analysis plan, are not reported in the registry record. The record also does not state whether the two listed outcome measures were prioritized or how multiple comparisons were handled. These gaps prevent assessment of the risk of bias, the precision of the intended comparisons, and the robustness of any potential findings.

The safety figures are limited to counts of zero reported events and the listed numbers at risk. The registry record does not give follow-up duration or further details of event ascertainment. Those counts should not be interpreted as a comprehensive demonstration of safety.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Premenstrual Dysphoric Disorder and Antiepileptic Drugs — NCT00612235.
• Registry record last updated 2018-10-26; status last verified 2018-01.
• Results first posted 2018-10-26.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

The registry record describes a completed study of PMDD designation among women with epilepsy taking lamotrigine, levetiracetam, or carbamazepine monotherapy, compared with a control group without epilepsy. Beth Israel Deaconess Medical Center is listed as sponsor. Enrollment is 80; the listed start and primary completion dates are 2008-04 and 2010-06.

The primary outcome is participant designation under the Endicott DRSP criteria for two consecutive menstrual cycles. The record also describes a less stringent threshold. It lists two outcome measures and zero analyses, and primary outcome results are not reported in the registry record.

The registry caveat states that allocation is not stated and the arms are not a like-for-like control comparison. Safety listings report zero serious adverse events, zero other adverse events, and zero deaths in each group, with 20 participants at risk in each. These safety counts do not establish the primary outcome findings.