Bringing Relief to Adolescents Naturally Using Melatonin for Migraine: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Bringing Relief to Adolescents Naturally Using Melatonin for Migraine (NCT02344316), a randomized phase 2 study. The primary outcome is Number of Migraine/Migrainous Days Per 28 Day Period.

2. Clinical Question

The trial addressed whether nightly melatonin should be compared with placebo in adolescents with migraine, using migraine or migrainous days as the main outcome. It was a pilot study, so its posted results are most useful for understanding the observed estimates and their uncertainty, rather than as a definitive answer to the clinical question.

The main outcome was recorded using an online or mobile headache diary. The registry record does not report a more specific eligibility definition for the population represented in the primary outcome results. Nor does it state the trial’s original statistical hypothesis or the size of difference that investigators considered important. Those omissions matter when interpreting a pilot comparison: a group estimate can describe the observed data, but it cannot by itself establish what difference would be clinically meaningful or whether the study was designed to resolve that question.

3. Trial Design

This was a randomized, quadruple-masked, parallel-group Phase 2 pilot trial. The record lists a total enrollment of 31 and marks the study completed. Its listed start was 2015-05, and primary completion was 2016-04. The sponsor is Amy Gelfand. The registry record describes the study as remote and identifies melatonin and placebo as the two interventions.

Participants assigned to the melatonin group were to take 3 mg orally each night; the placebo group was to take placebo orally each night. “Quadruple” is the masking designation in the record. Which specific roles were masked, how masking was maintained, and whether masking was assessed are not reported in the registry record. The method used to generate the random sequence, allocation concealment procedures, and the planned allocation ratio are also not reported in the registry record.

The primary endpoint was measured during the final four weeks of treatment. The posted primary outcome groups contain 12 participants in the melatonin arm and 11 in the placebo arm. These endpoint counts differ from the overall enrollment figure; reasons for the difference are not reported in the registry record. The outcome entries do not establish whether these are all randomized participants, an analysis population defined by a particular rule, or participants with complete diary data.

The record lists four secondary outcomes: time to sleep onset, recruitment by strategy, medication adherence measured by openings, and headache diary compliance. Results are posted for time to sleep onset, adherence, and diary compliance. A numerical result for recruitment by strategy is not included in the posted outcome information. The trial’s protocol, detailed analysis plan, and sample-size justification are not reported in the registry record.

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: There was a 28 day baseline run-in prior to randomization.

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.

6. Primary Endpoint

The registry lists the primary outcome as Number of Migraine/Migrainous Days Per 28 Day Period, measured over final 4 weeks of treatment.

The primary endpoint counts migraine or migrainous days over a 28-day period, using entries from an online or mobile headache diary. It is reported as a mean number of days for each group, with standard errors, for the final four weeks of treatment. This makes the endpoint a measure of headache-day frequency over a defined observation window.

As a general statistical point, a frequency measure can be useful when the research question concerns how often an outcome occurs over time. Diary recording can capture repeated day-level information instead of relying only on a participant’s later recollection of an entire period. However, a diary-based measure also depends on recording behavior: missed entries and differences in diary completion can affect how much information is available. The registry record does not report the rule used to define a migraine or migrainous day, the minimum diary completion required, or how missing diary entries were handled.

The group means were 3.6 days for melatonin and 4.9 days for placebo. The posted mean difference was -1.3 days, defined by the comparison of melatonin with placebo. This estimate describes the observed difference in group means; it does not alone determine how compatible the data are with other possible differences. That uncertainty is conveyed by the confidence interval and should be considered alongside the small endpoint group counts.

7. Results Posted in the Registry

• Melatonin: mean 3.60 versus 4.90 on control; difference -1.30 (95% CI -5.07 to 2.47), standardized effect d = -0.29; z-test p = 0.499.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Number of Minutes to Sleep Onset

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Melatonin: mean 15.20 versus 17.20 on control; difference -2.00 (95% CI -10.74 to 6.74), standardized effect d = -0.18; z-test p = 0.654.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Melatonin and the control. The interval gives the range of effects that remain compatible with the data.

Medication Adherence as Measured by Number of Openings Per Participant

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Melatonin: mean 39.00 versus 60.00 on control; difference -21.00 (95% CI -43.01 to 1.01), standardized effect d = -0.76; z-test p = 0.061.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Melatonin and the control. The interval gives the range of effects that remain compatible with the data.

Headache Diary Compliance Rate

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Melatonin: 91.7% (11/12) versus 90.9% (10/11) on control; risk difference 0.8 percentage points (95% CI -22.3 to 23.8), risk ratio 1.01 (0.78–1.30); unadjusted two-proportion z-test p = 0.949.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Melatonin and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover distinct questions: sleep-onset time, medication-opening counts as an adherence measure, and diary compliance. Recruitment by strategy is also listed as a secondary outcome, but a numerical result for it is not included in the posted outcome information. These measures can add context about study conduct and participant reporting, but they do not measure the primary endpoint in the same way.

The secondary comparisons are uncertain and should be considered alongside, rather than as replacements for, the primary comparison. Their estimates do not provide a clear independent resolution of the uncertainty around migraine-day frequency. In particular, adherence and diary completion relate to how treatment and outcome information were recorded; they do not establish what the migraine-day result would have been under a different pattern of adherence or completion. The registry record does not report an analysis connecting these measures to the primary outcome.

As a general statistical point, multiple outcomes create multiple opportunities to observe apparently notable results by chance. If no adjustment or prespecified hierarchy is used, unadjusted secondary p-values can overstate how much evidence the collection of results provides. The record does not report a multiplicity plan, and no secondary analyses are listed. The secondary outcomes should therefore be regarded as exploratory descriptions, not as confirmatory tests or a substitute for a prespecified primary analysis.

The pattern across the secondary outcomes is also constrained by the small outcome groups and by the absence of reported methods for missing data and analysis populations. A measure based on recorded openings, for example, depends on the relationship between openings and actual medication-taking behavior; the registry describes the measurement but does not report validation details. Similarly, diary compliance describes completion, not the accuracy of each diary entry. These distinctions help explain why secondary measures can inform interpretation without settling the main clinical comparison.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Number of Minutes to Sleep Onset
• Number of Participants Recruited From Each Recruitment Strategy
• Medication Adherence as Measured by Number of Openings Per Participant
• Headache Diary Compliance Rate

9. Statistical Interpretation

10. Statistical Methods Explained

The registry identifies the primary outcome as a continuous measure reported as a mean, with standard error as the measure of dispersion. The treatment comparison is a mean difference, expressed as melatonin minus placebo. These summaries answer related but distinct questions: the group means describe the observed level in each arm, while the difference describes the observed contrast between arms.

What does the mean difference represent?

As a general statistical explanation, a mean difference is formed by subtracting one group mean from the other in the stated comparison order. A negative value in this trial’s comparison means the observed mean was lower in the melatonin arm. The posted primary estimate is -1.3 days. This is a summary of the observed groups; it is not a guarantee that the same difference would occur in another sample or setting.

The primary outcome’s standard errors are 0.9 for melatonin and 1.7 for placebo. A standard error describes the estimated uncertainty in a group mean, rather than the spread of individual participants’ values. The registry reports standard errors for this primary measure, whereas the sleep-onset and adherence outcomes are reported with standard deviations. These dispersion measures are not interchangeable: a standard deviation describes variability among observations, while a standard error describes uncertainty in an estimated mean.

What does a confidence interval add?

As a general explanation, a confidence interval gives a range of estimates compatible with the data under the method used to construct it. It is not a range containing a fixed proportion of individual participants, and it does not assign a probability to a particular true value after the data have been observed. The primary interval, -5.07 to 2.47 days, includes zero and spans differences in either direction. Its breadth is important: the point estimate is only one possible value within a substantially uncertain comparison.

How should the p-value be read?

The unadjusted p-value for the primary comparison is 0.499. In general, a p-value describes how unusual the observed data, or data more extreme under a specified test, would be if the test’s null hypothesis and assumptions held. It is not the probability that the null hypothesis is true, nor does it measure clinical importance. The registry record does not report the specific test used for the sponsor’s primary analysis. The posted comparison statistics are simple unadjusted re-calculations from the available summaries; the record does not list a separate reported analysis.

The discrepancy indicator is false. Where no discrepancy is identified, the posted recalculation can be described as consistent with the available summaries, but it does not provide missing details about the original analysis plan. The registry record does not report whether covariates were included, whether the analysis accounted for repeated measurements, or whether any other adjusted model was used.

Why does the pilot setting matter?

In general, smaller studies tend to yield less precise estimates, particularly when participant outcomes vary. Precision is reflected in the interval, not established merely by whether a p-value crosses a conventional threshold. A pilot study can contribute information about feasibility and the scale of observed outcomes, but its estimates may remain unstable. The record describes this study as a pilot and posts planning illustrations based on observed effects; it does not report the trial’s own original sample-size justification.

The primary endpoint is diary-based and measured over a specified treatment window. General considerations for diary outcomes include incomplete recording, differences in completion, and the possibility that the analyzed participants do not represent everyone randomized. The registry record does not report the missing-data procedure, the analysis population definition, or a sensitivity analysis for incomplete diaries. Consequently, the posted group summaries should be interpreted without assuming that missing observations were absent or handled in a particular way.

The registry marks the trial randomized, quadruple-masked, and parallel-group. These design labels describe important features, but they do not reveal the actual randomization sequence, concealment procedures, masking roles, or whether the planned procedures were followed. Those details are not reported in the registry record. The design labels therefore cannot resolve questions about implementation that are not documented in the posted record.

11. Confidence Intervals and Precision

The primary mean difference was -1.3 migraine or migrainous days per 28-day period, with a confidence interval from -5.07 to 2.47. That interval includes zero and extends in both directions. It therefore communicates uncertainty not captured by the point estimate: the posted data are compatible with a lower mean in the melatonin group, little or no difference, or a higher mean. The interval should be read as a range of estimates under the method used, not as a range of individual outcomes.

Group means were 3.6 days for melatonin and 4.9 days for placebo. Their standard errors were 0.9 and 1.7, respectively. The endpoint groups included 12 and 11 participants. These counts are relevant to precision, but the registry record does not report how the analysis population was selected or whether the displayed counts represent everyone with randomized assignments.

As a general statistical point, a confidence interval’s width reflects both the variability in the observations and the amount of information available, along with the method used to calculate the interval. A wide interval means the estimate is imprecise; it does not prove that there is no difference. Conversely, a point estimate in one direction does not establish that direction as the true effect. The uncertainty is especially important for this pilot comparison because the interval encompasses values on both sides of zero.

The posted unadjusted p-value is 0.499. It offers a different summary of the comparison, but it does not replace the interval or show the range of effect estimates compatible with the data. The record does not report the prespecified primary analysis method or a clinical threshold for interpreting the size of the difference. Accordingly, precision and possible importance cannot be judged by the p-value alone.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are illustrative sample sizes per arm for continuous or binary outcomes; they are not total event counts. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each illustration targets a fraction of the observed effect. The trial’s own original design assumptions and sample-size justification are not reported in the registry record.

For the primary outcome, planning for 100% of the observed effect gives 188 participants per arm for 80% power and 252 participants per arm for 90% power. Planning for 75% of the observed effect gives 335 participants per arm for 80% power and 448 participants per arm for 90% power. Planning for 50% of the observed effect gives 752 participants per arm for 80% power and 1,007 participants per arm for 90% power.

In plain language, the 80% power column represents the per-arm sample size that the stated calculation estimates would give an 80% chance of detecting the assumed effect under its assumptions. The 90% column uses a higher target chance and therefore indicates a larger required sample in these illustrations. Smaller assumed effects require more participants because they are harder to distinguish from random variation. The figures depend on the observed effect used as their starting point, so they should not be treated as a definitive design recommendation.

These calculations are planning illustrations, not observed power. They do not tell whether the completed trial had a particular probability of success after its results were known, and they do not document what the investigators originally intended. The calculation assumptions describe the illustrations, not necessarily the trial’s design or analysis.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -1.300; pooled standard deviation σ = 4.498.
2. n = 2 × (1.96 + 0.842)² × 4.498² / -1.300² = 187.9, rounded up to 188 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -1.3                      # observed difference in means
sigma <- 4.49772429229758                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(2344316)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 188 (80% power), 252 (90% power)
 75% of observed effect: n per arm = 335 (80% power), 448 (90% power)
 50% of observed effect: n per arm = 752 (80% power), 1,007 (90% power)
power.t.test: 188.9 
Simulated power at n = 188 per arm: 0.801 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome and four secondary outcomes. It does not report a strategy for controlling multiplicity, such as a prespecified sequence of tests, a correction for multiple comparisons, or a distinction between confirmatory and exploratory secondary analyses. The posted secondary comparisons are unadjusted. As a general statistical point, assessing several outcomes can increase the chance that at least one comparison appears unusual by chance, so isolated unadjusted results warrant caution.

The record lists no reported analyses. It nevertheless posts outcome summaries and comparison statistics. The registry does not identify the exact analysis population for the primary or secondary outcomes, whether analyses followed randomized assignment, or whether participants were excluded from any particular analysis. The endpoint group counts vary by outcome, but reasons for those counts are not reported in the registry record. They should not be interpreted as evidence of a specific withdrawal or exclusion process.

Missing-data handling is not reported in the registry record. For a diary-based endpoint, missing entries could affect the number of days available for measurement; for adherence and compliance outcomes, unavailable observations could affect the displayed counts or means. The record does not state whether incomplete records were omitted, imputed, or handled by another method. It also does not report sensitivity analyses that examine alternative assumptions about missing values.

The original statistical analysis plan, model specifications, covariate adjustments, and assumptions about outcome distributions are not reported in the registry record. Nor is a prespecified method for handling multiplicity reported. Without these details, the posted results can be described as group summaries and unadjusted comparisons, but they cannot establish how closely the reported calculations match a prespecified confirmatory analysis. These omissions do not change the numerical estimates; they limit the conclusions that can be drawn about their analysis context.

15. 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 0% reporting threshold.

Most frequent non-serious adverse event terms

The registry reports serious adverse events affecting 0 of 13 participants at risk in the melatonin group and 0 of 13 participants at risk in the placebo group. Deaths are reported as 0 in each group. These are the posted serious-event figures for the listed risk groups.

Other adverse events are reported as affecting 6 of 13 participants in the melatonin group and 4 of 13 in the placebo group. These counts are separate from the serious adverse event figures. The record does not provide serious-event terms, and it does not report further detail in the posted safety summary about the nature, timing, severity, or relatedness of the other events.

As a general statistical point, observing no serious events in a small group does not establish that the risk is absent. Safety counts describe what was recorded in the participants at risk; they do not by themselves compare risks precisely or resolve uncommon-event uncertainty. The registry record does not report additional safety analyses.

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

The posted primary result is a mean difference of -1.3 migraine or migrainous days per 28-day period for melatonin compared with placebo. The interval from -5.07 to 2.47 includes zero and values in both directions. The result therefore describes a lower observed group mean for melatonin, while also showing that the comparison is imprecise. The unadjusted p-value is 0.499; it does not determine whether the observed difference is important or establish that the groups are the same.

This result does not establish that melatonin does or does not prevent migraine in adolescents. The study was identified as a pilot, the endpoint groups were small, and the interval allows a range of possible comparisons. The registry record does not report a target difference for clinical importance, a detailed original analysis plan, or the handling of missing diary data. Those limitations make it inappropriate to treat a single estimate or p-value as a definitive answer.

Nor should the secondary outcomes be interpreted as proof that the primary result is explained by sleep onset, adherence, or diary completion. The posted information does not report analyses linking those measures to migraine-day outcomes. The secondary results are unadjusted, and the registry does not describe a multiplicity strategy. They are best understood as additional descriptive information about this pilot study.

The safety summary reports no serious adverse events or deaths in the listed risk groups, alongside other adverse events affecting participants in both arms. Those observations describe this study’s recorded participants and do not demonstrate that serious or other harms cannot occur. The record’s safety information is limited to the posted counts.

Overall, the data support a cautious reading: the registry presents a pilot comparison, an estimated group difference, and substantial uncertainty around that estimate. They do not provide a definitive treatment conclusion, a precise estimate of the effect, or a complete account of the analysis and missing-data procedures.

17. Why This Trial Matters Statistically

This trial is statistically informative as an example of the distinction between an observed estimate and the precision of that estimate. The primary group means differ, and the mean difference has a direction and magnitude. Yet the confidence interval is broad and includes values on either side of zero. A reader who focuses only on the point estimate or only on the p-value would miss that uncertainty.

The registry also makes clear why pilot studies and confirmatory studies serve different roles. A pilot comparison can document measured outcomes and provide estimates that may inform later planning. But when an estimate is imprecise, planning calculations that use it as a starting point can be sensitive to the assumed fraction of the observed effect. The illustrations are not evidence that the study had adequate power, nor do they establish the sample size that investigators originally intended.

Another statistical lesson is the importance of analysis transparency. Details such as analysis populations, handling of incomplete diaries, the original analysis plan, and multiplicity procedures affect how results should be interpreted. Those details are not reported in the registry record. Their absence does not erase the posted outcomes, but it limits assessment of how the outcome summaries relate to the study’s intended analysis.

Finally, secondary outcomes can add useful context while still carrying less evidential weight than a prespecified primary comparison. Measures of sleep onset, adherence, and diary completion address different questions, and unadjusted results across several outcomes should not be read as multiple independent confirmations. The record illustrates how to consider outcome estimates, uncertainty, design labels, and reporting gaps together rather than relying on any single statistic.

18. 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.

Several limitations specific to this record affect interpretation. The trial enrolled 31 participants, while the primary outcome groups list 12 participants for melatonin and 11 for placebo. The registry record does not report why those counts differ from overall enrollment or whether the endpoint groups represent a defined analysis population. The counts should therefore not be used to infer how many participants withdrew or why.

The primary outcome was obtained through an online or mobile headache diary, but the record does not report the definition of a migraine or migrainous day, completeness thresholds, or treatment of missing entries. Diary completion is also listed as a secondary outcome, making missingness and recording behavior relevant to interpretation. No analysis of the relationship between diary compliance and the primary endpoint is reported.

The study is described as a remote pilot, but operational details about remote participation, monitoring, and data verification are not reported in the registry record. Randomization and quadruple masking are recorded, but the sequence method, allocation concealment, masking roles, and masking assessment are not described. The original sample-size rationale and detailed statistical analysis plan are also not reported.

Only a limited set of safety counts is posted. The serious event count and deaths are zero in both listed groups, while the record does not provide terms for serious events or detailed descriptions of other events. The registry also lists recruitment strategy as an outcome without posting a numerical result for it. These reporting gaps restrict assessment of feasibility, data completeness, safety detail, and the extent to which the posted analyses were prespecified.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Bringing Relief to Adolescents Naturally Using Melatonin for Migraine — NCT02344316.
• Registry record last updated 2021-09-09; status last verified 2021-08.
• Results first posted 2021-09-09.
• 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.

22. Record Summary

The registry record describes a completed Phase 2 randomized, quadruple-masked, parallel pilot trial of nightly melatonin versus nightly placebo in adolescents aged 12–17 with migraine. Overall enrollment was 31. The primary outcome was migraine or migrainous days per 28-day period during the final four weeks of treatment, measured by an online or mobile headache diary.

The primary group means were 3.6 days for melatonin and 4.9 days for placebo, with a posted mean difference of -1.3 days and a confidence interval from -5.07 to 2.47. The unadjusted p-value was 0.499. The interval includes zero and values in both directions, indicating substantial uncertainty in this pilot comparison.

Secondary results include sleep onset, medication-opening counts, and diary compliance. The record does not report a multiplicity strategy, missing-data methods, analysis populations, or the original sample-size justification. Serious adverse events and deaths were reported as 0 in each listed group.