Inflammation-Induced CNS Glutamate Changes in Depression: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Inflammation-Induced CNS Glutamate Changes in Depression (NCT03004443), a randomized phase 4 study. The primary outcome is Central Nervous System (CNS) Glutamate.

2. Clinical Question

3. Trial Design

The registry describes a Phase 4, randomized, triple-masked, parallel-group study comparing infliximab with placebo in people with depression. The recorded enrollment is 22, and the trial status is terminated. The record gives a start date of 2017-05-15 and a primary completion date of 2019-11-27. It does not report the reason for termination.

The brief summary describes an intended study of people with high inflammation, defined there as CRP >3mg/L, and proposes one infusion of infliximab or placebo. It states a planned group size of n=30 per group. That stated plan differs in scale from the recorded enrollment, but the registry does not explain the difference or give a recruitment history. The enrollment figure should not be treated as the number who contributed usable data to every endpoint.

The primary outcome is left basal ganglia glutamate assessed with magnetic resonance spectroscopy. Its listed timeframe is baseline, Day 3, and Week 2. The summary also describes assessments of inflammatory biomarkers, motivation, and psychomotor activity at baseline and follow-up times, while the registered secondary outcomes include several symptom, task, fatigue, depression, and biomarker measures. The record lists 16 outcome measures but 0 analyses.

Some design details are not reported in the registry record, including the specific masking roles, allocation-concealment procedure, treatment dose, and the statistical analysis plan. The recorded details establish the basic comparison and design labels, but do not permit reconstruction of how the trial was conducted in every operational detail.

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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Eleven distinct participants in the placebo study arm were examined at least one time point but some had a poor signal scan resulting in non-usable data. One participant in the placebo study arm was removed from the study following the baseline assessment.

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 Central Nervous System (CNS) Glutamate, measured over Baseline, Day 3, Week 2.

The primary endpoint is central nervous system glutamate, expressed as a mean continuous measure in mmol/kg. The registry describes measurement of glutamate in the left basal ganglia using magnetic resonance spectroscopy. The planned primary-outcome time points are baseline, Day 3, and Week 2.

The rationale in the record is that left basal ganglia glutamate tends to be increased during inflammation and is associated with an increase in depressive symptoms. The research aim was to examine whether inflammation alters central nervous system glutamate and how glutamate relates to specific symptoms. This makes the endpoint a brain-based measure tied to the proposed biological pathway, rather than a direct measure of depression alone.

Because glutamate is listed as a mean with standard deviation, an analysis could describe the average measured level and the variability among observations. A treatment comparison would need to account for the repeated measurements over time and make clear which participants contributed usable scans at each time point. The registry record does not report group means, standard deviations, changes from baseline, treatment contrasts, confidence intervals, or p-values for the primary endpoint. Consequently, the endpoint’s rationale and measurement are described, but the posted record does not establish the observed pattern or the precision of a group comparison.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies glutamate as a continuous outcome summarized by a mean and standard deviation. It does not report the statistical procedure used to compare groups or time points. The descriptions below explain general statistical concepts that would be relevant to this type of outcome; they are not claims about the trial’s unreported analysis plan.

What does a mean and standard deviation describe?

For a continuous measure, a mean is the arithmetic average of the observed values included in a calculation. The standard deviation describes how dispersed those values are around their mean. A small standard deviation indicates values are more clustered around the average; a larger one indicates greater spread. These summaries describe the observed data, but do not by themselves quantify the uncertainty in a treatment comparison. The standard deviation is not the same as a confidence interval or standard error.

Why is a treatment contrast needed?

A clinical-trial question generally concerns a contrast between assigned groups, not whether one group’s mean changed in isolation. A contrast states how the outcome differs between groups under a defined analysis, such as comparing follow-up levels while accounting for baseline. An estimate of that contrast should be interpreted alongside its uncertainty. A confidence interval gives a range of values compatible with the data and the method’s assumptions; a wider interval indicates less precision than a narrower one. The registry record reports no contrast or confidence interval for glutamate.

How do repeated measurements affect analysis?

The primary endpoint has measurements planned at baseline, Day 3, and Week 2. Measurements from the same participant are generally related: a participant’s values over time are not independent observations. A repeated-measures method can account for that within-participant relationship and use the time structure explicitly. Alternative analyses may focus on a specific follow-up time or on change from baseline. The choice affects the estimand—the precise treatment effect being targeted—and should be described in the analysis plan. The registry record does not report the method, the target contrast, or whether baseline values were included as covariates.

When a trial measures outcomes at multiple time points, results can also be presented as separate contrasts at each time or as an overall pattern over time. Those approaches answer different questions. An overall treatment-by-time assessment asks whether the trajectories differ, while a single-time comparison focuses on one follow-up. The registry lists the primary time points but does not identify which contrast was designated for inference or how it would be interpreted.

Why do unusable scans and missing observations matter?

Magnetic resonance spectroscopy measurements can be unavailable for analysis when scan quality is inadequate; the registry explicitly notes poor signal scans among placebo participants. In general, excluding unusable measurements can reduce the information available and may affect the interpretation if the reasons measurements are missing are related to participant characteristics or outcomes. Statistical handling depends on how much data are missing, why they are missing, and the assumptions made about the unobserved values. Common approaches may use only observed data or model outcomes under stated missing-data assumptions. The registry does not report which approach was used, or the extent of missing primary-outcome data by group and time point.

A repeated-measures model can sometimes use participants with incomplete follow-up without requiring every participant to have every measurement, but this does not eliminate the need to examine missingness and justify assumptions. An analysis of only participants with usable scans at all specified times answers a different question from an analysis that includes all available measurements. The record does not say whether the primary analysis followed an intention-to-treat principle, used an evaluable-scan population, or followed another definition.

What do the planning assumptions mean?

The registry’s sample-size assumptions describe planning illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. Such illustrations concern the number of participants per arm for a selected power target and effect fraction; they are not observed power. No corresponding calculated sample-size values or observed primary-outcome effect are reported in the registry record. Therefore, those assumptions cannot establish how well the trial was powered for the primary outcome, and they should not be read as a description of the trial’s original design justification. The record does not report that justification.

The distinction between planning and inference is important. A planning calculation asks how many participants might be needed to detect a hypothesized effect under chosen assumptions. An analysis of trial data estimates an effect and its uncertainty from observed outcomes. A planning illustration based on an observed effect does not replace a treatment estimate, and it cannot resolve the absence of reported outcome analyses.

Why can over-dispersion matter?

In general, over-dispersion means that data vary more than a statistical model expects. If a model assumes less variability than is present, its uncertainty estimates can be too small and tests can appear more conclusive than warranted. For a continuous measure, the relevant issue may be whether the chosen model represents the observed distribution and variability adequately. The registry does not report a model or any assessment of its fit, so no claim about over-dispersion in this trial is possible. This general concept does not imply that over-dispersion occurred in the glutamate data.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 16 outcome measures, including the primary glutamate measure and secondary measures of pleasure, motivation, task performance, fatigue, depressive symptoms, and high-sensitivity C-reactive protein. Assessing multiple outcomes or multiple time points can raise multiplicity concerns: the more comparisons considered, the greater the chance that at least one appears unusual by chance if each is interpreted without adjustment. In general, investigators can specify a hierarchy of outcomes, limit confirmatory comparisons, or use an adjustment procedure. The registry record does not report a multiplicity strategy or identify which secondary comparisons, if any, were confirmatory.

The record lists 0 analyses. It therefore does not provide outcome-specific estimates, p-values, confidence intervals, or an analysis population for the primary measure. It also does not report whether analyses were planned separately for baseline, Day 3, and Week 2 or combined in a repeated-measures model. Without those details, it is not possible to assess the handling of multiple time points or distinguish prespecified from exploratory findings.

Missing-data information is limited. The primary-outcome description says eleven distinct placebo participants were examined at least one time point, some scans had poor signal and were unusable, and one placebo participant was removed after baseline. The record does not give counts of usable observations by group and time point, reasons for missingness for the other observations, or the approach used to account for missing scans. It does not report whether all randomized participants were included in the primary analysis, how withdrawal or removal was handled, or whether sensitivity analyses were conducted.

These omissions matter because a reported mean could represent all randomized participants, those with any usable scan, or a more restricted set. Those populations need not answer the same question. The statistical analysis plan, analysis population definitions, multiplicity handling, and missing-data methods are not reported in the registry record.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: serotonin syndrome after study completion and after receiving escitalopram (9.1%).

The registry reports adverse events by recorded group, with 11 participants at risk in each group for both serious and other adverse events. In the infliximab group, 1 participant was recorded as affected by a serious adverse event and 6 by other adverse events. In the placebo group, 0 participants were recorded as affected by a serious adverse event and 8 by other adverse events. Deaths were 0 in each group.

The record lists the serious-event term “Serotonin syndrome after study completion and after receiving escitalopram” with 9.1%. The wording places the event after study completion and after receiving escitalopram; it does not establish that infliximab or placebo caused it. The registry does not provide further details about timing, event adjudication, relatedness assessments, or the full clinical course.

These are descriptive counts, not a formal comparison of safety between groups. They do not account for differences in follow-up or exposure, and the record does not report a statistical analysis of adverse events. A count of affected participants also differs from a count of event episodes, since a participant may experience more than one event. The presented figures should therefore be read as the registry’s reported event totals and not as proof of a causal treatment effect.

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

The record establishes that investigators planned a randomized comparison of one infliximab infusion with placebo and selected left basal ganglia glutamate as the primary endpoint. It also describes a biological rationale linking inflammation, glutamate, and depressive symptoms. These details explain the question the trial was designed to address; they are not evidence that the proposed pathway was confirmed in the participants.

The primary outcome is not accompanied by group-level results or a treatment-effect estimate. Consequently, the posted information does not show whether glutamate changed over time, whether the groups differed at any follow-up, or how precisely any difference was estimated. A p-value, confidence interval, or observed effect size is not reported in the registry record. No conclusion about the direction or magnitude of a comparative effect follows from the endpoint description alone.

The record’s enrollment and placebo scan information also do not identify the number of participants who contributed usable primary-outcome data in each group at each time point. The stated enrollment of 22 and the summary’s intended n=30 per group describe different aspects of the record; neither substitutes for an outcome-specific analysis population. The reason for termination is not reported.

Safety counts offer limited descriptive information, but they are distinct from the efficacy question and do not establish causation. Similarly, the list of secondary outcomes indicates what was measured or planned, not what was found. The record does not report the analyses needed to assess those outcomes.

Overall, the registry record documents a mechanistic research question and basic trial design, while leaving the primary comparative result and its uncertainty unreported. It neither confirms nor rules out a treatment-related glutamate difference. Interpreting the study beyond that boundary would require outcome data and analysis details that are not reported in the registry record.

13. Why This Trial Matters Statistically

This trial is statistically notable because its primary outcome is a mechanistic measure rather than a direct clinical outcome. The investigators sought to examine whether inflammation-related changes in central nervous system glutamate might connect treatment assignment with symptoms. Such a question requires careful separation of biological measurements, symptom measures, and treatment comparisons: evidence about one does not automatically establish an effect on another.

The design combines randomized treatment assignment with repeated measurement of a continuous brain-imaging outcome. In general, randomization helps make assigned groups comparable on average, while repeated-measures analysis can account for the fact that measurements from the same participant are related. But these strengths are useful for interpretation only when the analysis population, time-point contrasts, missing observations, and uncertainty are reported. The registry record does not give those analytic results.

The discrepancy between the summary’s stated intended group size and recorded enrollment underscores the difference between design intentions and the data available for analysis. It is not possible to determine from the record why enrollment differed, why the trial was terminated, or how those circumstances affected precision. The placebo scan-quality note further illustrates that enrollment does not guarantee usable data for an imaging endpoint.

Finally, the record lists multiple outcome measures and follow-up times but does not report analyses or a multiplicity strategy. Statistical interpretation depends not only on what was measured, but on which comparisons were planned and how uncertainty and missingness were handled. This case therefore demonstrates why a trial’s design and rationale should be distinguished from its evidential result: without reported estimates and methods, readers can identify the question but cannot assess the size or precision of an answer.

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 trial is that the registry record does not report primary-outcome results. There are no arm-specific glutamate means or standard deviations, no comparative estimates, and no confidence intervals or p-values. The record lists 0 analyses. Therefore, the main question cannot be evaluated from the posted information.

The population description identifies eleven distinct placebo participants examined at least one time point, notes poor signal scans that were unusable, and reports that one placebo participant was removed after baseline. It does not give usable-data counts at each time point or corresponding details for the infliximab arm. The reason for removal and the handling of missing observations are not reported in the registry record. This prevents a clear assessment of the analysis population and the potential effect of incomplete imaging data.

The registry says the trial was terminated, but the reason is not reported. It records enrollment of 22 while the brief summary describes n=30 per group as the intended size; it does not explain the difference, recruitment history, or any consequences for statistical precision. The original sample-size justification is not reported.

Further statistical details are unavailable, including the planned primary contrast, treatment of repeated measurements, covariate adjustment, distributional assumptions, multiplicity handling, and any sensitivity analyses. The masking roles and the method of allocation concealment are also not reported. Although adverse-event counts are listed, the registry does not provide enough detail to infer causality or compare safety formally. These limitations concern what can be interpreted from the registry record and do not establish that any particular methodological problem occurred.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Inflammation-Induced CNS Glutamate Changes in Depression — NCT03004443.
• Registry record last updated 2023-10-27; status last verified 2023-10.
• Results first posted 2023-06-07.
• 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 describes a terminated, randomized, triple-masked, parallel-group Phase 4 study sponsored by Emory University. It compared one intravenous infusion of infliximab with one intravenous infusion of placebo in people with depression and high inflammation. Recorded enrollment was 22; the brief summary describes an intended n=30 per group. The primary outcome was left basal ganglia glutamate measured by magnetic resonance spectroscopy at baseline, Day 3, and Week 2.

The registry lists 16 outcome measures and 0 analyses. Some placebo scans were unusable because of poor signal, and one placebo participant was removed after baseline. Serious and other adverse-event counts are recorded by group, with 0 deaths in each group. The reason for termination, primary-outcome comparisons, statistical analysis plan, and missing-data methods are not reported in the registry record.