Transcranial Alternating Current Stimulation for Major Depressive Disorder: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Transcranial Alternating Current Stimulation for Major Depressive Disorder (NCT02339285), a randomized na study. The primary outcome is Change in the Montgomery-Asberg Depression Rating Scale (MADRS) Score.

2. Clinical Question

This was a pilot clinical trial, so the posted results describe a relatively small set of participants and have substantial uncertainty around the group comparisons. The statistical question is about the pattern and precision of measured score changes, not a clinical recommendation. The record does not report a separate primary contrast as part of the stated endpoint: the endpoint is defined as MADRS change from baseline to F2, and the results include the three study arms. Further protocol details, including the planned statistical comparison and the rationale for the enrolled sample size, are not reported in the registry record.

3. Trial Design

The registry describes a randomized, quadruple-masked, parallel-group trial sponsored by the University of North Carolina, Chapel Hill. It lists an enrollment of 32 and a completed status. The recorded study start was 2015-05-07, and the primary completion date was 2017-06-19. The phase is listed as NA. The registry does not explain how the random sequence was generated, how allocation was concealed, or which groups of people were included in the quadruple-masking designation.

Three arms appear in the posted outcome results: tACS (Alpha), tACS (Gamma), and Sham Stimulation. The outcome group sizes are 10, 11, and 11, respectively. The active alpha intervention was described as 10 Hz stimulation at a peak-to-peak amplitude of 2 mA for 40 minutes. The active gamma intervention was 40 Hz at the same amplitude and duration. The sham description specifies an 80-second sequence of 10 Hz stimulation, including ramp-in and ramp-out periods, delivered with the alpha device.

The endpoint timeframe was baseline to F2, defined as four weeks after completion of the intervention. The outcome description also names intermediate assessment times, including Day 5 and F1, two weeks after completion, but the primary comparison is baseline to F2. The posted record lists seven outcome measures and seven analyses, with two secondary outcomes posted. The trial’s own sample-size justification, detailed visit schedule, adherence information, and analysis-population definitions are not reported in the registry record. As a result, the registry supports a description of the design and posted results, but not a full reconstruction of all operational or statistical decisions.

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: Participants were randomly assigned to 1 of 3 arms: tACS (alpha), tACS (gamma), or sham stimulation. All conditions were delivered during awake, resting state for 40 minutes on 5 consecutive days. 98 participants signed consent, 40 did not meet criteria, 6 declined participation, and 20 were excluded for other reasons. 32 were randomized.

Overall Study

Reasons for not completing this period:

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 Change in the Montgomery-Asberg Depression Rating Scale (MADRS) Score, measured over Baseline to F2 (4 weeks after completion of the intervention).

The primary endpoint was change in the Montgomery-Asberg Depression Rating Scale (MADRS) score from baseline, on Day 1 of stimulation, to F2, four weeks after intervention completion. MADRS is a 10-item measure of depressive episode severity. Each item is rated from 0 to 6, and the total score ranges from 0 to 60. Lower scores indicate fewer depressive symptoms. The registry defines a negative change as a decrease in symptoms.

A score-based endpoint provides a continuous measure, so the posted summaries describe the average change and its standard deviation in each group. The mean conveys the group’s average observed change, while the standard deviation describes how dispersed individual changes were around that average. An endpoint measured at a specified follow-up time makes the comparison interpretable in relation to a defined point after treatment completion. The registry also lists measurements at other time points, but the stated primary endpoint is the baseline-to-F2 change.

The outcome is a measured rating-scale change, not a direct measure of every aspect of a participant’s experience. The record does not report additional details about how missing endpoint assessments were handled, how the rating was administered, or whether the analysis adjusted for baseline score or other participant characteristics. Those details matter when interpreting group comparisons, but cannot be inferred from the posted description.

7. Results Posted in the Registry

• tACS (Alpha): mean -13.60 versus -8.27 on control; difference -5.33 (95% CI -13.36 to 2.70), standardized effect d = -0.56; z-test p = 0.193.
• tACS (Gamma): mean -8.82 versus -8.27 on control; difference -0.55 (95% CI -8.87 to 7.77), standardized effect d = -0.06; z-test p = 0.897.

8. Statistical Interpretation

9. Statistical Methods Explained

The outcome is continuous: each participant has a numerical change in MADRS score, and the posted group summaries use means and standard deviations. In general, a mean describes the average measured change in a group. A standard deviation describes the spread of individual observations around that mean; it is not the same as the uncertainty of the estimated group mean. The reported standard deviations are 7.75 for alpha, 8.94 for gamma, and 10.88 for sham stimulation.

What does the sponsor’s ANOVA test?

ANOVA is a general method for comparing average outcomes across multiple groups. An omnibus ANOVA addresses whether the group means, considered together, show evidence of variation beyond what would be expected under a model in which the means are equal. It does not, by itself, identify which particular pairs of groups differ. Pairwise follow-up tests, their adjustment for multiple comparisons, and the exact outcome associated with each sponsor-reported ANOVA result are not reported in the registry record.

The sponsor’s record contains three ANOVA entries, with p-values of >0.10, <0.001, and >0.10. The associated entries do not identify their particular outcomes or analysis questions, so those values cannot be assigned to the MADRS endpoint or to a specific pairwise comparison on the basis of the registry record alone. This matters because a p-value is meaningful only in the context of the test and question it addresses.

What are the pairwise calculations?

The pairwise results are simple unadjusted re-calculations from the posted arm summaries. A mean difference contrasts the average change in one active arm with the average change in sham. Negative contrasts indicate a more negative average change in the active arm under the stated ordering. For alpha versus sham, the estimated mean difference is -5.33; for gamma versus sham, it is -0.55. These are not adjusted estimates from a reported model.

A confidence interval gives a range of values compatible with the data and the method’s assumptions. The 95% intervals for the alpha and gamma comparisons are -13.36 to 2.696 and -8.872 to 7.772. Both include zero, but neither interval should be read as proving that the true difference is zero. They show that these small samples do not pin down the pairwise differences precisely. In general, the width of an interval reflects both the variability of observations and the amount of information available.

The unadjusted p-values for these re-calculations are 0.193 for alpha versus sham and 0.897 for gamma versus sham. In general, a p-value describes how unusual data at least as incompatible with a specified null hypothesis would be under that hypothesis and the test’s assumptions. It is not the probability that the null hypothesis is true, nor does it measure the size or importance of an effect. These re-calculated p-values are not substitutes for the sponsor’s reported analysis.

Why does the discrepancy matter?

The record marks a discrepancy between the sponsor analysis and the simple re-calculations. The sponsor’s own analysis takes precedence. The recalculations use posted group means and standard deviations to form unadjusted pairwise estimates, whereas the sponsor reports ANOVA entries whose outcome mapping and fuller analysis details are not given. Different analysis questions, models, data handling, or adjustment choices can produce different results; which, if any, explains this discrepancy is not reported in the registry record.

Random allocation can support comparisons between groups, but randomization does not eliminate uncertainty, especially when outcome groups are small. A parallel-group design compares separate sets of participants assigned to the different interventions. Quadruple masking indicates that the registry classifies the trial as masked at four levels, but the identities of those masked are not stated. The record also does not describe the allocation procedure or analysis population. These omissions limit how specifically the statistical results can be interpreted.

General statistical considerations apply to any analysis of multiple outcomes or comparisons. If many hypotheses are tested, the chance of at least one apparently unusual result can rise unless multiplicity is addressed. The registry does not report whether adjustments were made. Likewise, the standard errors, confidence intervals, and p-values depend on the chosen model and assumptions. The posted record does not provide sufficient detail to assess model diagnostics, missing-data procedures, or whether the sponsor’s ANOVA assumptions were checked. Thus, the numerical results can be described, but the full analytical pathway cannot be reconstructed from the registry entry.

10. Confidence Intervals and Precision

The pairwise confidence intervals are wide relative to the point estimates, so the estimated differences are not precise. For alpha stimulation versus sham, the unadjusted mean difference is -5.33, and its 95% confidence interval ranges from -13.36 to 2.696. For gamma stimulation versus sham, the estimate is -0.55, with a 95% confidence interval from -8.872 to 7.772. Each interval includes zero and also includes values on either side of zero.

In general, a confidence interval is a range generated by a statistical procedure that reflects the uncertainty in an estimate under its assumptions. A 95% confidence interval is not a guarantee that the specific interval contains the true value, nor is it a range containing 95% of individual participant results. Its width is informative: broader intervals indicate less precision. Here, the small outcome groups and the observed variability are relevant to the uncertainty, but the registry does not report a formal assessment of why the intervals have their particular widths.

These intervals accompany simple, unadjusted re-calculations from posted summary statistics. They are not identified as sponsor-reported intervals, and the registry does not provide confidence intervals for the sponsor’s ANOVA entries. Since the record flags a discrepancy, the sponsor’s analysis takes precedence. The intervals are useful for understanding the uncertainty in the descriptive pairwise recalculations, but should not be treated as a full replacement for the sponsor’s analysis or as evidence of equivalence.

11. 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 participants per arm for 80% power and participants per arm for 90% power; they are not total event counts. The planning illustration uses a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It sets the target effect to a stated fraction of the observed effect. These assumptions describe the illustration, not the trial’s original design. The trial’s own sample-size justification and original design assumptions are not reported in the registry record.

For alpha versus sham, planning for 100% of the observed effect gives 51 participants per arm for 80% power and 68 participants per arm for 90% power. For gamma versus sham, the corresponding figures are 5,146 and 6,888 participants per arm. The estimates are based on different observed effect sizes, so the figures should be read as contrast-specific illustrations rather than as a single enrollment recommendation.

The 80% column represents the estimated participants per arm needed, under the stated assumptions, for an 80% chance of detecting the target effect using the specified test. The 90% column applies a higher detection probability and therefore generally requires more participants. Planning for 75% or 50% of the observed effect also increases the required sample size: a smaller target effect is harder to distinguish from random variation. For example, the alpha comparison’s 50% illustration is 201 participants per arm for 80% power and 269 for 90% power; the gamma comparison’s 50% illustration is 20,581 and 27,552 per arm.

These calculations are planning illustrations based on observed estimates, not observed power. They do not describe the actual power of the completed trial, and they should not be interpreted as the study’s planned enrollment target.

12. 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 = -5.330; pooled standard deviation σ = 9.526.
2. n = 2 × (1.96 + 0.842)² × 9.526² / -5.330² = 50.1, rounded up to 51 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 <- -5.33                      # observed difference in means
sigma <- 9.52643551829385                     # 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(2339285)
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 = 51 (80% power), 68 (90% power)
 75% of observed effect: n per arm = 90 (80% power), 120 (90% power)
 50% of observed effect: n per arm = 201 (80% power), 269 (90% power)
power.t.test: 51.1 
Simulated power at n = 51 per arm: 0.799 (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.

13. Analyses Reported by the Sponsor

The sponsor reports an ANOVA across the three groups, with entries showing p-values of >0.10, <0.001, and >0.10. The registry does not identify the outcome or exact analysis question corresponding to each entry. It therefore is not possible to align those p-values unambiguously with the primary MADRS comparison or with either active-versus-sham contrast.

The recalculated alpha-versus-sham and gamma-versus-sham estimates are simple, unadjusted pairwise comparisons using posted means and standard deviations. Their p-values are 0.193 and 0.897, respectively. Because the record identifies a discrepancy, the sponsor’s own analysis takes precedence. The recalculations can still help describe the direction and uncertainty of the arm-level contrasts, but should not be presented as the sponsor’s inferential conclusion.

The registry does not include enough detail to establish why the reported results differ, including a complete mapping of each ANOVA result to its outcome and analysis specification. The sponsor’s analysis cannot be reconstructed from the posted entries alone.

14. Multiplicity, Missing Data and Analysis Populations

The record lists seven outcome measures and seven analyses, including the primary MADRS change and secondary outcomes concerning alpha oscillation power from resting-state EEG recordings. Two secondary outcomes are posted. The registry does not report whether the analyses were organized into a prespecified hierarchy, whether any correction for multiple comparisons was used, or which outcomes were included in each of the sponsor’s ANOVA entries.

In general, testing several outcomes or contrasts creates a multiplicity issue: the more hypotheses examined, the greater the chance of finding at least one result that appears unusual by chance. Methods such as a prespecified primary outcome or an adjustment to significance thresholds can address this concern, but the registry record does not state whether such methods were used. The unadjusted pairwise re-calculations should therefore be read as separate descriptive contrasts, not as multiplicity-controlled findings.

Handling of missing outcome data is also not reported in the registry record. The entry does not specify whether missing values were imputed, whether participants with unavailable F2 assessments were excluded, or what assumptions were made about missingness. The result-group sizes are listed, but they do not by themselves explain the number randomized, the number with observed endpoint data, or the analysis population used by the sponsor.

The record does not identify whether the sponsor used an intention-to-treat population, a per-protocol population, or another analysis set. It also does not report baseline adjustment, covariate adjustment, sensitivity analyses, or model diagnostics. These omissions do not invalidate the posted numbers, but they limit assessment of the analysis process and the robustness of the reported inference.

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.

The registry lists zero serious adverse events and zero other adverse events in each outcome group. The group counts at risk are 10 for alpha stimulation, 11 for gamma stimulation, and 11 for sham stimulation. The record also lists zero deaths in all three groups. No serious adverse event terms are listed.

These are the event counts reported for the groups in the registry record. They should be interpreted in the context of the number of participants listed and the trial’s pilot scale. A count of zero in the posted results means no events of that category were recorded in those entries; it does not establish that an event is impossible or that a treatment has no risk. The registry does not report additional details about adverse-event collection, observation duration, severity grading for other events, or whether all participants contributed the same duration of safety follow-up.

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

The posted arm means show average MADRS decreases from baseline in all three groups by the stated F2 endpoint. The mean changes were -13.6 for alpha stimulation, -8.82 for gamma stimulation, and -8.27 for sham. The registry defines negative values as decreases in depressive symptoms. These descriptive summaries do not, on their own, establish that the group differences are reliable or attributable to the intervention.

The simple unadjusted re-calculations estimate an alpha-versus-sham mean difference of -5.33 and a gamma-versus-sham difference of -0.55. Their 95% confidence intervals are broad and include zero. The sponsor’s analysis is reported as ANOVA with three p-value entries whose outcome mapping is not reported. Because the record flags a discrepancy, the sponsor’s analysis takes precedence over the re-calculations. The available record does not permit a definitive comparison between the sponsor’s analyses and the pairwise estimates.

These results do not establish that one stimulation frequency is superior to another, that either active intervention differs from sham, or that the groups are equivalent. Nor do they settle effects at time points other than the stated primary endpoint. Such conclusions would require clearly mapped analyses and adequate precision, neither of which can be inferred from the posted summaries alone.

The record describes a pilot trial with relatively small outcome groups. Its value is in documenting the design, endpoint, observed group summaries, and reported analyses, while also revealing uncertainty and reporting gaps that matter for interpretation. The posted findings are statistical observations from this trial, not medical advice or a basis for individual treatment decisions.

17. Why This Trial Matters Statistically

This trial illustrates why a reported average change and a between-group comparison answer different questions. Each arm had a negative average MADRS change, but the primary comparative question concerns whether the changes differ across randomized groups. The observed means are useful descriptions, while the uncertainty around pairwise contrasts is essential for judging how much the data distinguish among possible group differences.

The contrast between the alpha and gamma recalculations also demonstrates how effect size influences planning illustrations. The gamma-versus-sham estimate is close to zero, and the resulting planning figures based on that estimate are very large. Such figures are sensitive to the observed effect in a small pilot study; they are not reliable substitutes for a prospectively justified design target. The original sample-size rationale is not reported in the registry record.

The trial also highlights the importance of transparent statistical reporting. The record identifies an ANOVA and posts three p-values, but does not map each p-value to a particular outcome or analysis question. That ambiguity makes it difficult to reconcile the sponsor’s inference with the unadjusted pairwise calculations. Reporting the prespecified contrasts, analysis population, missing-data approach, multiplicity plan, and model details would make the statistical conclusions easier to evaluate.

Finally, a completed pilot trial can contribute information about study design and measured outcomes without resolving a clinical question conclusively. Small groups, broad intervals, and incomplete mapping of analyses call for careful interpretation rather than treating any single p-value or mean as a complete account of the evidence.

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.

The outcome groups are small: 10 participants in alpha, 11 in gamma, and 11 in sham. The pairwise intervals are broad, limiting precision. The registry does not provide participant-level data, baseline group summaries, or enough information to assess whether the posted outcome groups include every randomized participant.

The sponsor’s reported ANOVA entries are not mapped to specific outcomes or contrasts. The record flags a discrepancy between the sponsor analysis and simple re-calculations, and the sponsor’s analysis takes precedence. However, without complete analysis specifications, the reason for the difference cannot be determined. The recalculations are unadjusted and should not be confused with a reported sponsor model.

Other trial-specific details that could affect interpretation are not reported in the registry record. These include eligibility criteria, participant characteristics, the allocation procedure, the identities covered by quadruple masking, analysis-population definitions, missing-data methods, multiplicity handling, and model diagnostics. The registry does not report the study’s original sample-size justification or any interim-analysis plan. The absence of these details in the record does not show that such procedures were absent from the conduct of the trial; it means they cannot be evaluated from the registry entry.

The primary endpoint is limited to baseline-to-F2 MADRS change, four weeks after intervention completion. Other measurements are mentioned, but the posted primary contrast does not describe all time points or outcomes. Safety entries report zero events in the listed categories, but details of ascertainment and follow-up are not given. These limitations constrain interpretation of both efficacy-related measurements and safety counts.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Transcranial Alternating Current Stimulation for Major Depressive Disorder — NCT02339285.
• Registry record last updated 2018-08-20; status last verified 2018-01.
• Results first posted 2018-08-20.
• 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

This completed randomized, quadruple-masked, parallel-group pilot trial evaluated alpha- and gamma-frequency tACS against sham stimulation in people with major depressive disorder. Enrollment was 32. The primary outcome was change in MADRS score from baseline to four weeks after intervention completion. The posted mean changes were -13.6, -8.82, and -8.27 in the alpha, gamma, and sham groups.

Unadjusted pairwise re-calculations have broad confidence intervals that include zero. The sponsor reports ANOVA p-values, but their outcome mapping is not reported; the record flags a discrepancy, so the sponsor’s analysis takes precedence. The listed safety counts are zero serious adverse events, zero other adverse events, and zero deaths in each group. Important details about sample-size justification, missing data, multiplicity, and analysis populations are not reported in the registry record.