fMRI Study of Treatment Changes in Major Depression: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for fMRI Study of Treatment Changes in Major Depression (NCT01027559), a randomized na study. The primary outcome is Blood Oxygen-level Dependent Activations During an Emotional Distractor Task Between fMRI Scans of Depressed Participants in the CBT Group and Control Participants..

2. Clinical Question

3. Trial Design

This is a randomized, single-masked, parallel-group study registered as NCT01027559. Washington University School of Medicine is listed as the sponsor. The registry lists the study as completed, with a start date of 2009-02 and primary completion in 2014-01. The phase is recorded as NA.

The record describes depressed participants assigned to CBT or sertraline, alongside a healthy control group. The control group is not described as a treatment arm for depression. The CBT and sertraline arm descriptions specify an fMRI scan before treatment and a second scan after 12 weeks of therapy. Controls are described as having a second scan approximately 12 weeks after the first. The registry does not identify what was masked, or explain how masking was maintained.

The registry enrollment is 97. The primary outcome description says that MRI imaging was completed for 50 participants: 26 depressed participants randomized to CBT and 24 controls. This is the population described for the primary comparison; the registry does not provide a reconciliation of this analysis population with the enrollment figure. The listed adverse-event groups have separate at-risk counts, and those counts should not be treated as the primary fMRI analysis population.

The study’s broad stated aims include comparing CBT with antidepressant treatment and studying differences between depressed participants and healthy controls. The primary outcome title, however, specifies a comparison of depressed participants in the CBT group with controls. Sertraline is therefore part of the broader design but not the comparison identified in this primary outcome description. The record does not report the allocation procedure, the planned sample-size rationale, or detailed masking procedures.

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: Prior to being assigned to a group, 6 depressed participants dropped from the study before being assigned to treatment type.

Overall Study

Analysis population for the primary outcome, as defined in the registry: fMRI data from a total of 26 depressed participants randomized to receive CBT and 24 control participants was analyzed. The primary aim was to compare depressed and control participants to evaluate regional brain activation during an emotional task, both at baseline and 8-week follow-up. The primary regions analyzed were the amygdala and DLPFC.

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: A total of 97 participants enrolled in the study (63 depressed and 34 controls). Of the 63 depressed participants enrolled in the study, 57 (26 CBT and 31 SRT) were assigned to a treatment group and completed at least some study-related activities. Six depressed participants dropped from the study before being assigned to treatment type.

6. Primary Endpoint

The registry lists the primary outcome as Blood Oxygen-level Dependent Activations During an Emotional Distractor Task Between fMRI Scans of Depressed Participants in the CBT Group and Control Participants., measured over baseline visit and 8-week follow-up.

The primary endpoint concerns blood oxygen-level dependent (BOLD) activation during an emotional distractor task. BOLD is an fMRI signal used to characterize changes in regional brain activity while a participant performs a task. The registry identifies the amygdala and dorsolateral prefrontal cortex as a priori regions of interest and describes voxel-wise comparisons within masked regions. It states that positive values reflect BOLD activation and negative values reflect BOLD de-activation.

The measure is described as a mean of a continuous measure, with standard deviation as its dispersion measure and voxels as the unit. The endpoint is intended to characterize regional activation differences between the CBT depressed group and controls, rather than a clinical symptom score. Such an imaging measure can address a question about task-related brain activity, but it does not by itself establish what caused a difference or how it relates to clinical outcomes.

The outcome timeframe is stated as baseline visit and 8-week follow-up. The arm descriptions instead refer to scans after 12 weeks for CBT and approximately 12 weeks for controls. The record does not resolve this timing discrepancy. It also does not report group means, standard deviations, effect estimates, confidence intervals, or test results for the primary comparison.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes voxel-wise comparisons using ANOVAs within masked brain regions. This identifies a broad family of methods but leaves many details of the model and inference unreported. The explanation below distinguishes the general statistical meaning of the described approach from details that are not reported in the registry record.

What does a voxel-wise comparison do?

In general, an fMRI image contains measurements indexed by locations called voxels. A voxel-wise analysis compares the signal at corresponding locations across groups or conditions. The registry says that regions of interest were masked and that comparisons were conducted within those regions. This focuses the analysis on the amygdala and dorsolateral prefrontal cortex rather than describing an unrestricted comparison across the entire brain. The exact mask definitions, preprocessing steps, and voxel-level outcome construction are not reported.

In general, an analysis of variance (ANOVA) tests whether modeled group or condition means differ, given the structure of the model. The registry identifies ANOVAs but does not state the factors included, whether the model treated time as a repeated measure, or how participant-level dependence between scans was handled. It does not provide model coefficients, test statistics, degrees of freedom, or other numerical inferential results. Thus, the method description does not allow a reader to reproduce the analysis or determine how the reported comparison was estimated.

Why does testing many locations matter?

General statistical principle: conducting a separate test at many voxels creates a multiplicity problem. Even when each individual test uses a conventional threshold, a large set of tests can produce apparently notable findings by chance. Methods such as family-wise error control or false discovery rate control are used in some settings to address this issue. The registry record does not report whether any multiplicity correction was used, what threshold was applied, or whether testing across the two named regions was handled as a separate family of tests. No particular correction should be assumed.

The description mentions a priori regions, which indicates that the regions were specified in advance in the analysis description. It does not establish when or how that specification was made, nor does it replace the need to report the inferential procedure. The record also does not provide a full list of tested contrasts or identify which voxel-level results, if any, were selected for reporting.

How should the mean and standard deviation be understood?

The outcome is categorized as a continuous measure summarized by a mean, with standard deviation describing spread. In general, a mean gives a central summary of observed values, while a standard deviation describes how dispersed individual values are around that mean. These summaries can help readers understand the distribution of measurements, but they do not alone quantify uncertainty in a group difference. A confidence interval for a contrast, when reported, describes the range of effect values compatible with the data and model at the stated confidence level. No means, standard deviations, or confidence intervals for the primary endpoint are reported in the registry record.

The stated unit is voxels, while the description refers to BOLD activation values within masked regions. The record does not define the scale or units of the activation values at each voxel, or explain how values were aggregated to produce the named mean. It does specify the sign convention: positive values reflect activation and negative values reflect de-activation. Without numerical values and a more detailed measurement definition, the size of an activation difference cannot be interpreted.

What is known about timing and repeated scans?

The endpoint timeframe lists baseline and 8-week follow-up, while the arm descriptions state that the second scan followed 12 weeks of therapy for CBT participants and occurred approximately 12 weeks later for controls. Repeated measurements on the same participant are generally correlated. General statistical practice accounts for this dependence when estimating change over time or comparing trajectories; treating scans as independent when they are not can misstate uncertainty. The registry does not describe whether the primary analysis compared baseline values, follow-up values, within-person change, or a group-by-time pattern. It also does not report how this timing discrepancy was resolved.

The planning illustrations use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect described as a stated fraction of the observed effect. These are planning assumptions, not evidence of the trial’s original design rationale and not observed power. Planning sample sizes describe hypothetical future study requirements under assumptions; they do not measure the probability that this completed study detected an effect. The registry does not report its own sample-size justification.

Overall, the record identifies an imaging task, regions, and a broad analysis family. It does not provide enough detail to verify model specification, multiplicity handling, missing-data procedures, or numerical findings. Those gaps limit both reproducibility and assessment of statistical precision.

10. Multiplicity, Missing Data and Analysis Populations

The primary outcome description refers to voxel-wise comparisons within the amygdala and dorsolateral prefrontal cortex. Testing at many locations raises a general multiplicity concern: a reader needs to know whether and how the analysis controlled the chance of false-positive findings across the tested locations and contrasts. The registry record does not report a correction method, a significance threshold, or a prespecified hierarchy of tests. It also does not provide enough detail to determine how the outcome measures listed in the record relate to the primary analysis.

Missing-data methods and analysis populations are not described. The primary outcome description identifies MRI data from 50 participants, including 26 depressed participants randomized to CBT and 24 controls. The overall enrollment is 97, but the record does not explain the difference between enrollment and the MRI analysis population. Reasons for unavailable scans, exclusions, unusable images, or incomplete follow-up are not reported in the registry record. There is no description of imputation, sensitivity analyses, or a rule for including participants with only one scan.

The registry describes the primary comparison as CBT depressed participants versus controls, while the study also includes a sertraline group. It does not report whether the sertraline arm was excluded from this endpoint by design, analyzed separately, or included in another outcome analysis. The posted primary outcome description also gives a baseline and 8-week timeframe, whereas the arm descriptions refer to scans after 12 weeks or approximately 12 weeks. The analysis population and timing for each scan are therefore not fully clear.

The registry lists two outcome measures and zero analysis entries. It does not report a statistical analysis plan, an interim analysis plan, a missing-data plan, or the intended multiplicity strategy. Their absence from the record should not be interpreted as proof that none existed; the details are not reported in the registry record.

11. Safety Results

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

Other (non-serious) adverse events

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

Most frequent non-serious adverse event terms

The adverse-event information lists serious events separately from other adverse events for the three groups. For the CBT group, the registry reports 0 serious events among 26 at risk and 7 other adverse events among 26 at risk. For the healthy control group, it reports 0 serious events among 34 at risk and 2 other adverse events among 34 at risk. For the group labeled SSRI, it reports 0 serious events among 31 at risk and 21 other adverse events among 31 at risk. The intervention description identifies sertraline as the drug treatment.

These are registry counts, not estimates of comparative risk. The record does not provide event definitions, event types for the other adverse-event counts, ascertainment procedures, or exposure-adjusted summaries. It also does not report deaths: the death entries are null, which is not the same as a reported count of zero. The serious-event term list is empty. Zero serious events reported in these groups does not establish that serious events were impossible or that the groups had equivalent safety; the registry information alone is limited to the counts recorded.

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

The study addresses a question about task-related brain activation in depressed participants assigned to CBT compared with healthy controls. Its primary outcome is an fMRI measure, not a direct measure of depressive symptoms, daily functioning, or a clinical response. The record’s broader aims also mention comparing CBT with antidepressant treatment and examining differences related to depression, but the primary outcome title specifies the CBT-versus-control comparison.

The registry describes the measurement task, selected brain regions, and the MRI analysis population. It does not report numerical primary outcome findings. As a result, the record cannot establish whether BOLD activation differed between the groups, how large any difference was, or how precisely it was estimated. It also cannot establish that the groups were alike: absence of a reported result is not evidence of no difference.

Even if a difference in activation were reported, a statistical contrast would describe the measured groups under the analysis used; it would not by itself explain the biological mechanism, demonstrate that CBT caused a change, or show that an imaging difference corresponds to a clinical benefit. The primary outcome comparison is not with sertraline, so it cannot by itself answer whether CBT and sertraline have similar effects on the brain. That broader question would require results from a direct comparison and an appropriate analysis, neither of which is reported for this endpoint.

The discrepancy between the endpoint timeframe and the arm descriptions also matters to interpretation. The record does not clarify whether the stated follow-up was at 8 weeks or after 12 weeks. Until the numerical results and analysis details are available, the defensible conclusion is limited: the registry describes an fMRI comparison and its intended population, but does not report the result needed to evaluate the primary question.

13. Why This Trial Matters Statistically

This study illustrates why a clear distinction between a scientific aim and a reported statistical result is important. The study summary proposes questions about emotional circuitry, CBT, antidepressants, and healthy controls. The primary outcome narrows the focus to BOLD activation during an emotional distractor task in depressed CBT participants compared with controls. Those are related questions, but they are not interchangeable. A reader should identify the specific contrast before interpreting any claim about treatment or brain activity.

It also highlights the value of reporting the analysis population alongside the trial’s overall enrollment. The registry identifies 50 participants in the described MRI analysis, divided between depressed CBT participants and controls, while overall enrollment is 97. Making this distinction helps readers understand which participants contribute to a particular endpoint and prevents the mistaken assumption that every enrolled participant contributed usable data to every analysis.

Imaging outcomes create specific statistical demands. Voxel-wise analyses can involve many tests, repeated scans may be correlated, and interpretation depends on clear definitions of regions, contrasts, timing, and multiplicity control. The registry provides some of that context by naming the task, regions, and ANOVA approach, but does not report the inferential details or primary estimates needed to evaluate precision and robustness.

Finally, the timing mismatch shows why consistent endpoint descriptions matter. A stated 8-week follow-up and arm descriptions referring to 12 weeks can lead readers to different understandings of what was measured. The record is useful for identifying the broad design and intended imaging question, while also demonstrating that outcome results and analytic transparency are necessary for drawing statistical conclusions.

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 primary limitation for interpreting this registry record is the absence of numerical results for the fMRI endpoint. Group summaries, estimated differences, uncertainty intervals, and test results are not reported in the registry record. Without them, the direction, magnitude, and precision of the primary comparison cannot be evaluated.

The endpoint timing is inconsistent across descriptions. The primary outcome timeframe states baseline visit and 8-week follow-up, while the arm descriptions say the second scan occurred after 12 weeks of CBT or approximately 12 weeks for controls. The record does not clarify the timing used in the analysis. It also does not specify whether the central comparison concerns baseline measurements, follow-up measurements, change over time, or a group-by-time effect.

The analysis description is brief. It names voxel-wise ANOVAs and two regions, but does not report the model terms, covariates, dependence structure for repeated scans, voxel-level thresholds, or multiplicity correction. Details of image processing and the construction of the mean activation measure are also not reported. These omissions restrict reproducibility and make the statistical precision difficult to assess.

The MRI population is described as 26 depressed CBT participants and 24 controls, whereas the registry enrollment is 97. Reasons for the difference, missing scans, exclusions, and the analysis population rules are not reported. The primary endpoint does not describe the sertraline group, so this outcome cannot answer the broader CBT-versus-sertraline question on its own. Finally, the registry does not report the original sample-size rationale or a statistical analysis plan. These are limitations of the information in the record and do not establish what procedures may have been used outside it.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: fMRI Study of Treatment Changes in Major Depression — NCT01027559.
• Registry record last updated 2018-07-17; status last verified 2018-06.
• Results first posted 2018-06-13.
• 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

NCT01027559 is a completed, randomized, single-masked, parallel-group study sponsored by Washington University School of Medicine. It describes depressed participants assigned to CBT or sertraline and a healthy control group. The primary outcome concerns BOLD activation during an emotional distractor task, comparing depressed participants in the CBT group with controls in the amygdala and dorsolateral prefrontal cortex.

The MRI analysis is described as including 26 depressed CBT participants and 24 controls. The registry identifies voxel-wise ANOVAs, a mean continuous measure, and standard deviation as the dispersion measure. The outcome timeframe says baseline and 8-week follow-up, while arm descriptions refer to scans after 12 weeks or approximately 12 weeks.

Numerical primary outcome results and key analysis details, including multiplicity handling and missing-data procedures, are not reported in the registry record. The posted description therefore supports understanding the intended comparison, but not judging its result or precision.